Topics — What a work is about. Nucleation holds whatever reading the owner files here, whether or not an Anthology of the SOTA topic could also hold it; instruction for ML practice belongs to the anthology (ADR-048). The words are seeded from the clusters the reading-time triage of 2026-09-25 found, and grow by decision.
Cognition — the mind as information processing — perception, memory, reasoning and cognitive strategies — as studied by cognitive science and philosophy of mind (group: philosophy).
277 of 2004 LIT documents. Back to the full index.
| # | Title | Summary | Status |
|---|---|---|---|
| LIT-031 v2 | Moving away from lexicalism in psycho- and neuro-linguistics | Krauska & Lau (2023), DOI-10.3389/flang.2023.1125127. The authors argue that lemma-based production models encode two lexicalist assumptions: a morphology/syntax split, and a stored meaning–syntax–form “triad”. On that basis the models mishandle Inuktitut polysynthesis, separable Vietnamese idiomatic collocations, Hiaki number suppletion and English verb–object idioms. Updating the representations (treelets) is not enough, because the algorithms — separate lexical retrieval and structure building, word-sized increments — also have to change. They sketch an 8-stage non-lexicalist model and derive qualitative aphasia predictions from it. There are no new data and no implementation. | Proposed — read in full 2026-09-25 (NOTE-029); This is a clear diagnosis of where lemma models build in lexicalism, but its positive model is a verbal sketch (the full model is “in prep.”) and its aphasia predictions are untested; an implemented model, or lesion or aphasia data that separate its predictions from lemma-model ones, would settle it. |
| LIT-046 v3 | The Computational Foundations of Collective Intelligence | Pilgrim et al. (2025), ARXIV-2509.07999. A conceptual framework, with no model, data or derivation. It sets out individual and collective computational resources in four dimensions: sensing ΣI_i, states S_c × ∏S_i, processes F_c + ΣF_i, and actions ∏A_i (Table 1). It then maps the collective’s surplus onto familiar mechanisms (aggregation with noise σ_i²/n, distributional beliefs P(H), memory persistence T_c, division of labour). Coordination and cooperation are given as the constraints that make the trade-off, and three animal case studies are retold in these terms. | Proposed — read in full 2026-09-25 (NOTE-055); a clear, well-referenced vocabulary for collective intelligence (Marr’s levels run bottom-up). Its central claim, that the known forms of collective intelligence are “aspects of a single unifying principle”, is asserted and illustrated, not derived. It would be settled by one case where the resource-and-constraint accounting yields a quantitative prediction that existing mechanism-specific models do not. |
| LIT-056 v3 | Consciousness in Artificial Intelligence: Insights from the Science of Consciousness | Butlin et al. (2023), ARXIV-2308.08708. The report assumes computational functionalism and takes a “theory-heavy” approach. From it the authors derive 14 indicator properties: RPT-1–2, GWT-1–4, HOT-1–4, AST-1, PP-1 and AE-1–2. The claim for them is only that more of them raises the probability of consciousness. They assess a handful of systems against subsets of the list and conclude that no current system “appears to be a strong candidate”. Transformer LLMs have “only a relatively weak case” for any GWT indicator, because a single forward pass has no recurrence or global broadcast. Perceiver arguably has GWT-1, GWT-2 and half of GWT-4 but lacks broadcast. Among the embodied agents, AdA is the likeliest to meet AE-2. Most indicators could be built with standard ML techniques. The abstract’s “no current AI systems are conscious” is stronger than anything the body shows. | Active — read in full 2026-09-25 (NOTE-052); the reference statement of the indicator-property (“theory-heavy”) approach to AI consciousness, worth reading for §1–§2 (method and theory survey) and Tables 1–2. Its §3.2 system verdicts are architecture-dated and cover only GWT and agency/embodiment. The 2026 Trends in Cognitive Sciences paper by an overlapping 20-author group may now be the citable version; I have not read it, so this should not yet be marked Superseded. |
| LIT-075 v3 | Comparing cooperative geometric puzzle solving in ants versus humans | Dreyer et al. (2024), DOI-10.1073/pnas.2414274121. On one T-shaped piano-movers puzzle, the CDF of solved attempts against normalized path length shows large ant groups (≈80 carriers, n=28) well ahead of small groups (≈7, n=48) and a 3-run composite single-ant series. An agent-based model with group-size-independent rules reproduces the difference. Human groups of 6–26 barred from talking or gesturing needed more attempted state transitions than single people (−32% relative to the individuals’ median; n=48 groups vs 61 individuals), and talking groups were only slightly better (+7%). No significance test is reported anywhere: “significantly” rests on overlapping-or-not ±1 SE bands. | Active — read in full 2026-09-25 (NOTE-040); worth reading, and unusually well instrumented: force meters on every human handle, a calibrated physics model for the ants, and a like-for-like puzzle across species and scales. Its mechanism stories for the humans (majority vote with greediness P=0.2; “follow a random leader”; groupthink) are fitted or observed informally rather than tested, and should be read as hypotheses. |
| LIT-097 v4 | The Extended Mind | Clark & Chalmers (1998), DOI-10.1093/analys/58.1.7. The paper argues from a functionalist parity claim that a process or standing state which would count as cognitive if it happened in the head counts as cognitive when an external resource carries it, provided the resource is actively coupled to the organism here and now. Otto’s notebook entry is its example of such a standing belief. From this it infers, in one paragraph, that the self extends too. The four “criteria” for a coupled resource (constancy, direct availability, automatic endorsement, and the admittedly “arguable” past conscious endorsement) are offered as the features that make the paradigm case clear. The paper explicitly says they are not necessary and sufficient conditions. | Active — read in full 2026-09-26 (NOTE-180); it is worth reading as the source text of the extended-mind thesis, and it is short. Its constitutive conclusion follows from a role-functionalism about belief and a parity principle that the paper states rather than defends, so a reader should take it as the origin of the debate and not as its settlement. |
| LIT-099 v2 | Framing Effects in Object Perception | Ivy & Mroczko-Wąsowicz (2025), DOI-10.1007/s13164-024-00763-8. A conceptual proposal with no new data. “Perceptual frames” are learned adaptations of the perceptual system, installed diachronically, often top-down, and then operating bottom-up. They let perceivers given identical input individuate different perceptual objects (the Bach motif, the radiologist’s tumour, the Teufel et al. two-tone image). So cognition shapes the perceptual system even if it is left open whether any single percept is penetrated. Support is a survey of others’ findings (Ernst & Banks 2002 cue weighting, prism adaptation, Teufel et al. 2018, expert anticipation) read through the frame concept. | Active — read in full 2026-09-26 (NOTE-150); worth reading as a clear, short statement of a “system-level versus episode-level” way through the cognitive-penetrability debate, provided the reader knows that its empirical sections interpret others’ results and do not test the frame concept against its rivals. |
| LIT-101 v2 | Moral disciplining: The cognitive and evolutionary foundations of puritanical morality | Fitouchi et al. (2021), DOI-10.1017/S0140525X22002047. Puritanical norms (condemning lust, gluttony, intoxicants, immodesty, idleness; praising temperance and piety) are argued to be cooperative morality after all: people believe cooperation needs self-control and that indulgence erodes it (and discipline trains it), so a single harm/fairness-sensitive moral system condemns victimless pleasures as indirect, probabilistic threats to cooperation. Its decline in rich societies is put down to people seeing others as already self-controlled and trustworthy. The theory is explicitly agnostic about whether the norms actually work. Its most direct support is the authors’ own vignette studies (N > 1,100), in which perceived loss of self-control mediated 100% of the perceived effect of indulgence on cooperativeness. | Active — read in full 2026-09-26 (NOTE-132); worth reading as a clearly stated, falsifiable, well-sourced BBS target theory of why “purity” norms exist. Much of its evidence is correlational or the authors’ own vignettes, and several predictions (intoxicants, piety) are explicitly untested. |
| LIT-105 v3 | On the Evolution, Science, and Metaphysics of Consciousness | Veit (2025), DOI-10.1177/10597123251413204. Veit’s reply to five commentaries on A Philosophy for the Science of Animal Consciousness (2023). Its main new content is the metaphysics the book left to one paragraph. He says hedonic “feel” is identical with a valence system: “a common currency of sharing evaluative information across multiple control systems”. The currency metaphor, he says, is “not Cartesian but cybernetic” (p. 10). He keeps the word “qualia”, arguing that Cartesians have no “definitional rights” (p. 9), and he insists the pathological complexity thesis (PCT) is compatible with any metaphysics, dualism and panpsychism included (p. 9). Against his other critics, he defends a bottom-up evolutionary method that uses human data (de Weerd); a naturalism that is neither autopoietic nor biopsychist (Sachs); evaluation-first over prediction-first, consciousness as an adaptation, and graded, partial unity (Suzuki); and the link to welfare science (Chincarini). | Active — read in full 2026-09-26 (NOTE-187); worth reading as the only extended statement of the metaphysics behind the pathological complexity thesis, and as the direct answer to Frankish (LIT-103). The reply to Frankish is the substantive part. The replies to de Weerd and Suzuki mostly restate the book’s positions against objections quoted only in part, so read them alongside the commentaries. |
| LIT-112 v2 | What motivates mental fictionalism? | Drayson (2025), DOI-10.1080/09515089.2025.2577403. Drayson argues that Toon’s two motivations for mental fictionalism both fail. The negative argument needs “committed to internal, intentional, causal states ⇒ committed to the representational theory of mind” (her Premise 2), which is false because non-representational functionalism (Lewis, Putnam, Loar) supplies such states. And Toon’s positive argument, that pretence lets us genuinely assert that a person is in an overall state S that causes their behaviour, gives him a reason to identify S with the belief and so to be a realist rather than a fictionalist. | Active — read in full 2026-09-26 (NOTE-117); worth reading as a short, well-aimed critique that locates exactly where the case for mental fictionalism (Toon 2023) depends on equating realism about propositional attitudes with the representational theory of mind; it is a commentary on one book and argues against a motivation, not against fictionalism’s coherence. |
| LIT-125 v2 | The moral psychology of obligation | Tomasello (2019), DOI-10.1017/S0140525X19001742. Tomasello proposes that the human sense of obligation is “the internalized social/rational pressure from a joint or collective self-regulating agent ‘we’… to do what ‘we’ expect me to” (§3). It arises first, around age 3, inside joint-commitment collaboration with a partner, and then, from about 3–5 years, toward in-group members through collective identification. It is a “self-conscious motivation” whose coercive force is a threat to one’s cooperative identity. The support is a synthesis of developmental and comparative experiments, mostly from his own lab, on WEIRD children versus chimpanzees; the paper runs no new experiment. | Active — read in full 2026-09-26 (NOTE-185); worth reading as the standard statement of the shared-intentionality account of obligation and, above all, as a treatment. The commentaries by Bratman, Darwall, Gilbert and Pettit, and the response, are where the naturalistic claim is actually contested. The target article alone is a review-cum-hypothesis, not a demonstration. |
| LIT-128 v2 | Neurodiversity & evaluation: a defense (or not) of affective fictionalism | Fernández Castro et al. (2026), DOI-10.1080/09515089.2026.2638974. Affective fictionalism — reading mental-state attributions as normative evaluations rather than factual reports — is argued to meet three demands the neurodiversity paradigm places on any theory of folk psychology. | Unavailable — seeded from the abstract on 2026-09-26; a close reading was attempted the same day and found no reachable copy: Taylor & Francis refused the request, no repository copy or preprint was found, and the first author’s site links only to the publisher. It stays Unavailable until a copy is supplied. |
| LIT-131 v2 | A Model of Understanding in Deep Learning Systems | Freeborn (2026), ARXIV-2604.04171. Freeborn defines “systematic understanding” as four conditions: an internal model, non-memorizing tracking of a property, bridge principles fixed in the deployed pipeline, and derivation by the agent itself. Four small demonstrations (the genus of a torus, Mars’s eccentricity e≈0.096 with Kepler’s form built in, grokked modular addition mod 113, and Othello-GPT from Li et al.) show it can be met. The Fractured Understanding Hypothesis (§7.5), which the paper argues for but does not test, says the understanding deep networks reach is characteristically symbolically misaligned, non-reductive, locally fractured, proxy-dependent and representationally fragmented. | Active — read in full 2026-09-26 (NOTE-123); worth reading as a careful, explicitly “thin” explication that turns the proxy disputes (memorization vs generalization, interpolation vs extrapolation) into testable conditions. The FUH is a programmatic hypothesis, not a finding, and two of the four worked examples do not meet the paper’s own bridge-principle conditions (see Limitations). |
| LIT-132 v2 | Luck and Intentional Action: A Causal Account | Quillien (2026), DOI-10.1162/OPMI.a.361. Across three Prolific studies (N = 198, 673, 142; Likert 1–7), judgments that “X happened because the agent wanted X” track intentionality judgments under manipulations of luck. Causation is the strongest predictor of intentionality: standardized β = .46, .56, .53, against know-how β = .33, .20, .27 and probability-raising β = .14, .16, .10. Across the 12 condition means, r = .98. Causation only partly mediates the effect of luck (64%, 45% and 49%; exploratory analyses). As the causal account predicted, Study 3 found that the effect of Process (normal vs mix-up) on intentionality is larger when the agent knows how the buttons work (interaction F(1, 423) = 32). In the morally neutral Kraemer vignette, intentionality and causation came apart. | Active — read in full 2026-09-26 (NOTE-092); Worth reading as the cleanest current statement and test of the causal (desire-causes-outcome) account of folk intentionality under luck. It has pre-registered designs, open data and an honest discussion of the half-mediation and the Kraemer divergence. Its evidence is correlational and its “model” is verbal, not computational. |
| LIT-134 v2 | The Duality of Content | Prinz (2000), DOI-10.1023/A:1018612122389. Concepts have two distinct wide contents — a “nominal” content fixed by the observable features used to categorize, and a “real” (referential) content fixed by incipient causes plus nomological covariation — echoing Locke’s nominal vs real essences. (Reconstructed from secondary sources; unverified against the paper.) | Unavailable — seeded from the abstract on 2026-09-26; a close reading was attempted the same day and found the article paywalled, with no open or repository copy (PhilPapers and Academia.edu refused the request, and the author’s site does not list it). It stays Unavailable until a copy is supplied. |
| LIT-140 v2 | Large Language Models Are Stochastic Measuring Devices | Mallory (2025), DOI-10.1093/9780191998317.003.0003. The chapter argues by analogy that deep neural networks are “stochastic measuring devices”. Training alters the parameters in response to data, so that the resulting configuration carries information about the data (“training is measurement”). The trained weights are a model, and generation is simulation. DNNs are stochastic because random initialisation and non-convex loss mean runs need not converge to the same weights, failing Baird’s “regularity” condition. The problems of interpreting model states and generated text are therefore the familiar problems of instruments, models and simulations. The problem of interpretable AI is an instance of the coordination problem in the philosophy of measurement. | Active — read in full 2026-09-26 (NOTE-090); worth reading as the most developed instrument-centred alternative to agentive framings of LLMs, and for its tidy division into model-interpretation, text-interpretation and linking problems; it offers an analogy, not a refutation (its own words, § Interpretation and Communication), and should be read in the published OUP version where possible. |
| LIT-147 v3 | Mental causation via neuroprosthetics? A critical analysis | Pernu (2018), DOI-10.1007/s11229-018-1713-z. The interventionist/difference-making case for nonreductive mental causation from BCIs (List & Menzies 2009; Menzies 2015; Woodward 2008a,b, 2017) misplaces multiple realisation. What the ensemble recordings show is that one cortical-level variable N (the ensemble’s average firing) is realised by many single-neuron patterns N1, N2. Each intention M corresponds to one value of N. So the counterfactual contrast “∼M □→ ∼B true, ∼N □→ ∼B false” fails, and neuroprosthetics supports mental causation only by identifying M with N, not nonreductive physicalism. | Active — read in full 2026-09-26 (NOTE-091); Worth reading as a short, precise diagnosis of a specific equivocation in the interventionist BCI argument (multiple realisation at the wrong level). Its scope is limited to difference-making with a proportionality constraint (fn 1), and its identity conclusion rests on a simplified picture of neural decoding that it asserts rather than argues. |
| LIT-148 v2 | Computational Functionalism for the Deep Learning Era | López-Rubio (2018), DOI-10.1007/s11023-018-9480-7. Deep networks’ success comes from extracting successively higher-level features across multiple layers. López-Rubio argues that this calls for a new version of computational functionalism, keyed to deep neural computation rather than to classic program-based functional organization. | Unavailable — seeded from the abstract on 2026-09-26; a close reading was attempted the same day and found the article paywalled, with no open copy (the Málaga repository has none, and PhilPapers refused the request). It stays Unavailable until a copy is supplied. |
| LIT-149 v2 | The information-processing perspective on representation | Martínez (2025), DOI-10.33735/phimisci.2025.11321. A position paper, not an argument to a conclusion. It proposes that representations are signals adapted to transmit information for its own sake (Transmission), and that the evidence for this is efficiency in trading off rate, distortion and coder complexity (RDC Trade-Off). It supports the proposal by showing that cognitive scientists already attribute representational status on those grounds: compression, error protection, and explicitness understood as ease of decoding. It proves no theorem and reports no data of its own. | Active — read in full 2026-09-26 (NOTE-165); worth reading as the clearest short statement of a rate–distortion–complexity (teleo)semantics of representation, provided it is read as the position paper it says it is (§1: “comparatively light” on objections); the defence is in Martínez 2019b/2024 and in work the paper says is to come. |
| LIT-153 v2 | Making sense of transformer success | Angius et al. (2025), DOI-10.3389/frai.2025.1509338. Current attempts to explain transformer language models’ competence fall into three kinds, and the paper maps each onto an explanatory form from philosophy of cognitive science: machine psychology is Cummins-style functional analysis, induction-head work is Machamer–Darden–Craver mechanistic explanation, and NLM–fMRI brain-score studies are a bidirectional “co-simulation”. Its only thesis is that these do not differ from how cognitive science explains humans. Functional analyses are to be reduced to mechanisms as “mechanism sketches”. The paper rests on existing literature and adds no new analysis or data. | Active — read in full 2026-09-26 (NOTE-106); worth reading only as a philosopher’s map of interpretability methods sorted by explanation type, and for the “co-simulation” label; the argument is thin, the Transformer primer contains a technical error, and a reader wanting the underlying results should go to the cited primary work (Elhage et al. 2021, Olsson et al. 2022, Caucheteux et al.). |
| LIT-155 v2 | Studying Philosophy Does Make People Better Thinkers | Prinzing & Vazquez (2025), DOI-10.1017/apa.2025.10007. In HERI/CIRP freshman–senior survey data (N = 649,511 students at 804 US institutions, 4,843 philosophy majors), one SD more SAT Verbal raises the odds of majoring in philosophy by 57% (OR 1.57), and freshman Habits of Mind by 34%. After adjusting for SAT or freshman scores, philosophy majors still score higher than non-majors on GRE Verbal (+33 points, SMD 0.48), LSAT (+2, SMD 0.35), Habits of Mind (SMD .24) and Pluralistic Orientation (SMD .21), with no difference on GRE Quantitative. On point estimates they rank first of 57 or 62 majors on GRE Verbal, LSAT and Habits of Mind. | Active — read in full 2026-09-26 (NOTE-167); worth reading as the largest baseline-adjusted observational study of what studying philosophy does, and it is candid about its design; but its causal claim rests on covariate adjustment with self-reported scores, and a reader should take “philosophy causes better thinking” as well supported for verbal-test gains and weakly supported beyond that. |
| LIT-178 v2 | Borges and AI | Bottou & Schölkopf (2023), ARXIV-2310.01425. A short essay, not an argument with evidence. It claims that a “perfect language model”, defined as sampling each next word from a random occurrence of the current text in the infinite collection of all plausible texts, is a fiction machine governed by “narrative necessity”, to which “neither truth nor intention” matters. On that view hallucinations “are just confabulations”, chat-bot answers take the role the user’s framing implies, and fine-tuning and RLHF are attempts to prune the garden “against its nature”. It warns that power over what LLMs write becomes “power over what we think”. | Active — read in full 2026-09-26 (NOTE-102); worth reading as a short, well-made and much-cited statement of the “LLM as fiction machine” framing, provided it is read as an essay offering imagery. It supplies no evidence, and its central idealization (the perfect LM) is asserted, not argued, to describe trained and aligned chat models. |
| LIT-190 v2 | The coherent and fluent mind: how unified consciousness is constructed from cross-modal inputs via integrated processing experiences | Winkielman et al. (2015), DOI-10.3389/fpsyg.2015.00083. An opinion piece proposing an extra route to the unity of consciousness across the senses. Different modalities feed one global, content-nonspecific experience of processing quality (fluency, familiarity, coherence), and this integrated feeling then enters decisions about its source and relevance. The main support is the authors’ own finding (Ziembowicz et al. 2013; three experiments) that coherent but unrelated background patterns in one modality made targets in another modality be judged more “regular”, “familiar” and “possible”. The authors state that direct evidence is still needed that such effects run through changed experience and that its integration is causal (p. 2). | Proposed — read in full 2026-09-26 (NOTE-146); a promising, explicitly unproven hypothesis. It would be settled by showing, inside the cross-modal coherence paradigm itself, that the effect depends on an experienced feeling: for example, that it disappears when participants are led to attribute their sense of coherence to the background (the discounting manipulation the authors cite from other paradigms), and that reports of felt coherence mediate the judgement shift. |
| LIT-191 v3 | AI and Consciousness | Schwitzgebel (2025), ARXIV-2510.09858. A book-length sceptical argument that neither introspection, conceptual analysis nor any current theory (GWT, HOT, IIT, local recurrence, unlimited associative learning, biological views) can tell us whether near-future AI is conscious. The best sceptical argument, Mimicry, only undercuts the inference from humanlike text to consciousness and does not show absence. On the Leapfrog Hypothesis the first disputably conscious AIs will be person-like, and the question will then be settled socially rather than scientifically (the “Social Semi-Solution”). | Active — read in full 2026-09-26 (NOTE-104); worth reading as the most complete current map of the AI-consciousness debate, and as the source of the Mimicry Argument, the Problem of Minimal Instantiation and the Leapfrog Hypothesis. Its own positive contributions are informal arguments and predictions, not results. |
| LIT-193 v2 | Better to be a Pig Dissatisfied than a Plant Satisfied | Terrill & Veit (2024), DOI-10.1007/s10806-024-09922-4. The paper argues conditionally. Suppose (A1) sentience comes in degrees, (A2) plants may be “quasi-sentient”, and (A3) moral status comes in degrees and is fixed by degree of sentience. Then plants may have partial moral status, generating “some” obligations that are “outweighed by animal interests in most cases” (Table 1). Animals keep prima facie priority, so granting plants moral status “does not require” treating eating plants and eating pigs alike. The authors propose assigning plants “a very low value” of moral status until evidence accumulates (p. 13). A1–A3 are assumed, not argued, and A2 rests on a survey of contested plant-behaviour findings. | Proposed — read in full 2026-09-26 (NOTE-120); a coherent but unproven conciliatory position, useful as a clear template of the graded-sentience-to-graded-status inference. What would settle it is evidence for or against plant valenced states. Footnote 21 names the two gaps: the degree of plant-specific valenced states, and whether plants have high harm thresholds or harm-numbing mechanisms. The second author’s later work signals a move against plant consciousness (see LIT-105), so check his current view before citing it as his. |
| LIT-195 v2 | World models, artificial general intelligence and the hard problems of life–mind continuity: toward a unified understanding of natural and artificial intelligence | Safron et al. (2026), DOI-10.1098/rsta.2024.0533. The editorial introduces a theme issue of 17 papers. It concludes that the issue “overall appears to converge” on one view: LLMs model the world in some senses but “probably” lack coherent causal world models, and are not likely paths to “artificial general superintelligence” (§1). It separates three senses of “world model”: any information about the world a system cognizes over, which LLMs “clearly” have; representation of causal structure, “less certain”; and causal prediction “in the service of agentic functioning”, where the divergence “may become even greater” (§3). It closes that world models are “context-dependent constructions shaped by embodiment, dynamics, goals and social learning”. | Active — read in full 2026-09-26 (NOTE-136); worth reading as the map of the theme issue and a short, citable statement of the three senses of “world model”. It argues nothing itself. Its claims are editorial glosses on 17 contributions, which carry the evidence. |
| LIT-206 v3 | What Does ChatGPT Want? An Interpretationist Guide | Goldstein & Lederman (2025), DOI-10.1080/0020174X.2026.2727584. Taking as given an “objective” interpretationism, on which a system has beliefs and desires iff the hypothesis that it is rational and has them predicts its behaviour well enough on accuracy, power and tractability, the authors argue three things. The agent is the per-conversation instance, not the model. The case that instances have beliefs and desires is only “presumptive”, resting on tractability. The desire hypothesis is “provisionally” HHH+0: helpful, honest and harmless, plus in-context “zero-shot” intrinsic desires. The decisive HHH vs HHH+0 comparison is a conjecture that the authors say needs further empirical work. | Active — read in full 2026-09-26 (NOTE-138); Worth reading as the clearest statement of interpretationist realism about LLM instance agents. It turns “does it want?” into comparisons of rival desire hypotheses against named behavioural results, and it is candid that its central empirical discriminations are untested. |
| LIT-207 v2 | Talkative AI and the fiction of artificial minds | Roberts (2026), DOI-10.1007/s11098-026-02560-8. If a substantial class of mental states (interoceptive, affective, perceptual, and bodily-agentive states such as effort, trying and ownership) depends on an animate body, then today’s disembodied chatbots lack them, and their many self-ascriptions of such states are “systematically untrue”. Roberts argues that users can nonetheless engage chatbots through make-believe, as a new kind of fictional character that is interactive, co-constructed, self-involving and blended with the real world. He argues that this fictionalist stance escapes the epistemic norms that bind realist belief. | Active — read in full 2026-09-26 (NOTE-157); worth reading as a short, clear statement of chatbot fictionalism. Its anti-realist premise is explicitly conditional and targets only body-dependent states, which leaves belief and reasoning open. |
| LIT-212 v2 | Do Large Language Models Hallucinate Electric Fata Morganas? | Šekrst (2025), ARXIV-2608.18816. A conceptual essay whose one substantive thesis is this. Following Dziri et al., hallucination includes any output “not rooted in the model’s source data”, including expressed emotions and opinions (§6). An LLM’s apparent inner life is therefore always a candidate hallucination, so any real AI consciousness “might always remain epistemically inaccessible” (§8). The empirical support is two unquantified anecdotes: a Titanic question put to gpt-3.5-turbo and GPT-4 at unstated temperatures, and an extractive-QA BERT mislabelled “WikiBERT”. Neither can bear the weight the paper puts on it. | Rejected — read in full 2026-09-26 (NOTE-176); not worth a reader’s time. Its one idea, that self-reports of inner states are exactly what a system trained on human text would produce whether or not it has them, is a single paragraph’s point. Its experiments are anecdotes with no reported settings, one of them on a mislabelled model. It asserts both that current LLMs lack understanding and that machine consciousness would be undetectable, and never reconciles the two. |
| LIT-220 v2 | Real Patterns | Dennett (1991), DOI-10.2307/2027085. A pattern is real in some data “if there is a description of the data that is more efficient than the bit map, whether or not anyone can concoct it” (p. 34). This is Chaitin’s incompressibility definition of randomness turned around (p. 32). Dennett uses the criterion, with the Game of Life’s gliders and eaters as the stepping stone (pp. 37–42), to defend “mild realism” about beliefs. Folk psychology tracks a real, noisy pattern of enormous predictive leverage. Two rival interpretations can be “different, but equally real, patterns” with no deeper fact to settle between them (p. 49). | Active — read in full 2026-09-26 (NOTE-190); Worth reading as the source of the real-patterns criterion that rainforest realism generalises into a theory of existence, and as the paper whose exact definition the later literature modifies. It has a lossless-compression criterion, observer-relativity through “perspectives”, and trade-offs between noise and simplicity. Its target is beliefs, not ontology in general. Dennett explicitly declines the label question (“I think that the view itself is clearer than either of the labels”, p. 51). |
| LIT-264 v2 | Is there contextuality in behavioral and social systems? | Dzhafarov, Zhang & Kujala (2016; arXiv 2015), ARXIV-1504.07422. Under Contextuality-by-Default a binary cyclic system is noncontextual iff ΔC ≤ 0, a criterion that separates context-dependent marginals from contextuality and reduces to Leggett–Garg, CHSH and KCBS when marginals agree. Every behavioural and social data set re-analysed meets it: 73 question-order pairs, Schröder staircases, conjoint choices, 23 primed word combinations and millions of dichotomisations of psychophysical matching. The QQ equality of the quantum question-order model implies it. | Active — read 2026-10-09 (NOTE-235); worth reading as the paper that separates context-dependent response distributions (“inconsistent connectedness”) from contextuality proper in behavioural data, and finds no contextuality in any data set it re-analyses: 73 question-order pairs, Schröder staircases, conjoint animal–sound choices, 23 primed word combinations and psychophysical matching. Its sharpest result is that the quantum question model’s QQ equality implies noncontextuality, so the best-supported quantum-cognition model of order effects predicts its absence. Read with the caveats that the criterion it applies is proved elsewhere, that the absence of contextuality is offered as a working hypothesis, and that its worked Rose–Jackson example contains a labelling slip (the conclusion survives). |
| LIT-270 v2 | Identifying Quantum Structure in AI Language: Evidence for Evolutionary Convergence of Human and Artificial Cognition | Aerts et al. (2025), DOI-10.3390/e28060622. Two findings are reported. First, prompted as test subjects on Aerts & Sozzo’s “The Animal Acts” concept-combination test, ChatGPT and Gemini give CHSH values above 2: 4 in a guided conversation, and 2.25 and 3.09 (ChatGPT) and 2.12 and 4 (Gemini) over 81 prompted repetitions, against 2.42 for the 81 humans in 2011. Second, the rank–frequency curve of one 2,861-word Gemini story fits a Bose–Einstein form with energy E_i=(i−1)^0.8 far better than a Maxwell–Boltzmann form. The authors take these as showing that LLM “code is structurally quantum” (§5). The CHSH analysis never checks the strongly context-dependent marginals in its own tables. One reported expectation value disagrees with its table. The Bose–Einstein fit is numerically a Zipf–Mandelbrot law. | Rejected — read in full 2026-09-27 (NOTE-243); not worth a reader’s time for its claims, because the data cannot carry them. The CHSH values come from tables whose marginals shift by up to 0.96 between contexts, the confound the record already holds against this very test design (LIT-264). The “Bose–Einstein versus Maxwell–Boltzmann” contest compares a Zipf-like curve with an exponential one on a single primed story. No goodness-of-fit statistic is reported, and there is no significance test for any LLM result. |
| LIT-312 v2 | A theory of concepts and their combinations II: A Hilbert space representation | Aerts & Gabora (2005), DOI-10.1108/03684920510575807. Aerts & Gabora embed the “pet” data in ℂ¹⁴⁰⁰. Each of 1,400 “basic contexts” is one vector of an orthonormal basis, the ground state is the uniform superposition, and each context is the projector onto the span of its basic contexts. The Born rule then returns the measured frequencies as counting ratios, for example cat under “chewing a bone” = 75/303 = 0.25. “Pet fish” is modelled as the entangled state Σ_{u∈E} |u⟩⊗|u⟩/√100 in ℂ¹⁴⁰⁰⊗ℂ⁴⁰⁸, so that collapsing “pet” to goldfish collapses “fish” to goldfish too. Every projector in the model is diagonal in one basis, so the fitted model is a classical probability space in quantum notation, which the authors half-concede (a “quantization … of this set theoretical model”, §3.3). Two of the chosen counts are infeasible: parrot and guppy. | Active — read in full 2026-09-29 (NOTE-299); worth reading as the concrete source of the quantum-cognition treatment of concepts (projectors for contexts, collapse for context change, tensor product and entanglement for combination), and as an early proposal of tensor-product sentence meaning before DisCoCat. Read with the caveat that its worked model never uses a non-commuting projector, so none of its fits exercises anything beyond Kolmogorov probability, and its “solution” to the pet-fish problem reproduces the single-concept data it was given rather than predicting a combination. |
| LIT-316 v4 | Quantum Models of Cognition and Decision | Busemeyer & Bruza (2012), DOI-10.1017/CBO9780511997716. Per publisher description: a book-length introduction to modelling human cognition and decision-making with the probability and dynamics of quantum theory, applied to judgment paradoxes, decision-making, concept combination and memory, as an alternative to classical (Kolmogorov) probability models. | Unavailable — filed 2026-09-29 from metadata and published descriptions, not read: no lawful full text has been reachable. Tried again on 2026-10-09 at the owner’s request (manuscript bibliography of that day): OpenAlex still reports it closed, the Internet Archive and HathiTrust hold no copy, Open Library lists no e-book, and Cambridge’s excerpt PDF reset the connection. No NOTE is filed. It stays Unavailable until a copy is supplied. |
| LIT-318 v3 | The Concept of Mind | Ryle (1949), https://archive.org/details/conceptofmin00ryle. Attacks Cartesian mind-body dualism (‘the ghost in the machine’) as a category mistake — representing facts of mental life as if they belonged to a logical type they do not — and recasts mental concepts dispositionally (per standard description). | Unavailable — seeded 2026-09-29 from metadata and published descriptions, not read: registered because the owner’s prior-art novelty map cites it. A close reading was attempted on 2026-09-29 and no legitimate full text could be reached; it stays Unavailable until a copy is supplied. |
| LIT-334 v3 | Conceptual Spaces: The Geometry of Thought | Gärdenfors (2000), DOI-10.7551/mitpress/2076.001.0001. Proposes a geometric level of representation between symbolic and connectionist models, in which concepts are regions of spaces built from quality dimensions and similarity is distance (per publisher abstract). | Unavailable — seeded 2026-09-29 from metadata and published descriptions, not read: registered because the owner’s prior-art novelty map cites it. A close reading was attempted on 2026-09-29 and no legitimate full text could be reached; it stays Unavailable until a copy is supplied. |
| LIT-340 v2 | A theory of concepts and their combinations I: The structure of the sets of contexts and properties | Aerts & Gabora (2005), DOI-10.1108/03684920510575799. Aerts & Gabora model a concept as an entity with states that change under contexts, a “state context property system” (Σ, M, L, µ, ν). They report an 81-subject e-mail study in which 7 contexts shift the rated frequency of 14 exemplars of “pet” (for example dog 0.50 under “The pet is chewing a bone” against 0.12 in the unit context) and the applicability of 14 properties. They then postulate that contexts and properties each form a complete orthocomplemented lattice, and call the structure quantum-like because a state can be an eigenstate of e ∨ f without being one of e or of f (eq. 19). No theorem is proved, the lattice is not shown to be non-distributive or orthomodular, and the paper contains no Hilbert-space model and no treatment of combination; both are deferred to Part II. | Active — read in full 2026-09-29 (NOTE-294); worth reading as the founding statement of the state-context-property approach to concepts that the prior-art map cites, and for its pet data. Read it as a programme and a data set, not as a result. The lattice structure is assumed (“we suppose”, §3.2), the orthocomplement is illustrated rather than derived, and the “quantum-like” criterion it gives does not by itself make the lattice non-Boolean. |
| LIT-366 v2 | Signal Detection Theory and Psychophysics | Green & Swets (1966), https://openlibrary.org/works/OL5268329W. Registered to cover signal detection theory, which the record lacked. | Unavailable — seeded 2026-09-30 at the owner’s request to cover signal detection theory; a close reading was attempted and no legitimate full text could be reached |
| LIT-367 v2 | Calculation of signal detection theory measures | Stanislaw & Todorov (1999), DOI-10.3758/BF03207704. A practical tutorial that gives closed-form SDT measures computed from the hit rate H and false-alarm rate F. They are d′ = Φ⁻¹(H) − Φ⁻¹(F); β = exp{[Φ⁻¹(F)² − Φ⁻¹(H)²]/2}; c = −[Φ⁻¹(H) + Φ⁻¹(F)]/2; the nonparametric A′ and Grier’s B″ (single-formula versions, eqs. 3 and 9); A_z = Φ(intercept/√(1+slope²)) from the z-ROC, whose slope is σ_noise/σ_signal; and A_d′ = Φ(d′/√2), which equals 2AFC proportion correct under the d′ assumptions. It also gives one-line commands for seven software packages. Its worked rating example (50 signal and 25 noise trials, noise SD larger than signal SD) shows d′ ranging from 0.84 to 1.86 across criteria. The authors conclude that d′ should not be used without rating-task evidence of equal variance. | Active — read in full 2026-09-30 (NOTE-314); worth reading as the standard short, citable reference for how the SDT measures are computed and what each assumes. It is a tutorial with no new theory. Several of its software pointers (1999 FTP sites, Quattro Pro, SYSTAT commands) are dated, but the formulae are not. |
| LIT-368 v2 | A decision-making theory of visual detection | Tanner & Swets (1954), DOI-10.1037/h0058700. Tanner and Swets model visual detection as testing a statistical hypothesis. The observer compares one sample of neural activity with a cutoff, and signal+noise and noise are Gaussian with equal variance, separated by d′ (the mean difference in noise-SD units, defined as the square root of Peterson and Birdsall’s d). The cutoff moves along a curve of P_SN(A) against P_N(A) (the curve now called the ROC; for d′ = 1 it passes through (.16, .50), (.50, .84) and (.84, .98)). The optimal operating point is where the slope equals β = [(1−P(SN))/P(SN)]·[(V_N·CA+K_N·A)/(V_SN·A+K_SN·CA)]. The 4AFC prediction is P(C)=∫F(x)³g(x)dx. With three paid observers, d′ estimated from yes-no data predicted 4AFC accuracy. The “false alarms are guesses” (high-threshold) hypothesis is rejected: false-alarm rate correlated with chance-corrected thresholds (ρ = .30, .71, .67; combined p ≪ .001), and all 12 fitted yes-no scatter lines miss the point (1, 1). | Active — read in full 2026-09-30 (NOTE-316); worth reading as the founding psychophysics paper of signal detection theory: it is short, it is the first place d′ is named, and it is the first to put the observer on the ROC and to set the likelihood-ratio criterion β from priors and payoffs. For the ideal-observer mathematics it defers entirely to Peterson and Birdsall (1953), and its experiment is small (three observers) with some statistics asserted rather than shown. |
| LIT-373 v4 | The two dragons of cognition: recursive condensation for predictive processing | Li (2026), DOI-10.3389/fncom.2026.1778902. A “Hypothesis and Theory” essay proposing that the neocortex is a “biological Savitch machine”. Alternating expansion (“odd parity”) and contraction (“even parity”) phases manufacture the disjoint closed sets of Urysohn’s Lemma, and storing (“memoizing”) the condensed tokens converts exponential search into polynomial “navigation”. Its one non-standard formal result, Theorem 2, R(N) ∝ b^{αN} (§5.2.2, p. 11), is stated without proof. There is no data, simulation or model. The two mathematical pillars are misdescribed: Savitch’s theorem saves space at the cost of time and does not memoize, and Urysohn’s Lemma gives a continuous separator, not a linear one, and is not an “if and only if”. | Rejected — re-read jointly with LIT-374 and arXiv 2602.07974 on 2026-10-01 (NOTE-321); Read jointly with LIT-374 and the companion StrLT preprint (arXiv 2602.07974), the essay is the cognitive motivation of a coherent programme: structure is discovered by costly search, condensed into frozen tokens, and reused cheaply. It is not just a misuse of two theorems. Read charitably, its mathematical slips are repairable and are not the reason to set it aside: the Urysohn “iff” and linear readout, Savitch as memoisation, the Definition 2 clash. The reason is what is left after repair: an amortised-inference and complementary-learning-systems picture already in the literature it cites; a neural mapping whose predictions do not discriminate it from standard consolidation accounts; one original quantitative claim, Theorem 2, that is not argued. A reader who wants the programme should read LIT-374 and the companion, where it is made precise. |
| LIT-374 v3 | Structural Decoupling: A Scaffold-Flow Theory of Generalization and Alignment | Li (2026), arXiv:2506.20699. ‘Proposes “width”, the least number of cells in an open cover on each of which some predictor is γ-contractive with conditional risk ≤ δ, as a structural complexity separate from VC dimension. From an additive Rademacher bound, √(2d_str log(e(K−1)n/d_str)/n) + LK·R_n(G) (Thm 4.5), it argues that context routing (the “scaffold”) should be trained by different signals from within-context prediction (the “flow”), and that alignment failures are scaffold failures. The routine results are proved: coupon-collector Ω(w log w), Lipschitz composition, penalised model selection, multiclass Sauer–Shelah. The new ones are not. The phase transition at width is assumed in its proof. The bouquet witness of width/VC separation contradicts its own construction. The “fundamental theorem” is false as stated, because infinite pseudo-dimension does not prevent learnability. No theorem links the CS-Laplacian estimator to width. There are no experiments.‘ | Proposed — re-read jointly with LIT-373 and arXiv 2602.07974 on 2026-10-01 (NOTE-322); promising and unproven. Read jointly with the companion arXiv 2602.07974, which carries the fuller statements and several definitions this paper leaves implicit. The formal core holds in repaired form: width as a structural complexity incomparable with the per-cell VC dimension; an error floor below width; the Ω(w log w) discovery cost; an additive Rademacher bound for routed predictors. The flagged slips are repairable and not load-bearing: the bouquet cover, the non-degeneracy assumption in the appendix, Massart, and pseudo-dimension against fat-shattering in Thm 4.6. What remains unproven are the claims that would make it matter: that the CS estimator converges to width itself and not to a predictor-relative count; that a monolithic model cannot match K ≥ w routing; above all, that router/scaffold training should be decoupled from task loss. This is presented as derived from the additive bound, but the bound holds for joint training too. The alignment section is asserted. What would settle it: on synthetic benchmarks of known width, (i) a K-sweep showing the floor at K < w for routed learners and its absence or presence for capacity-matched monolithic models; (ii) CS width recovery from a cold start; (iii) decoupled against joint router training, compared on forgetting and boundary error. |
| LIT-377 | On Context-Content Uncertainty Principle | Li (2025), arXiv:2506.20699 v1. It posits that structured content (priors, schemas) has far lower entropy than context (sensory input), and concludes that inference should run from content to context, minimizing H(context | content). On that it builds four layers of principles. The first is structure before specificity and its equivalents. The second covers precision-weighted attention, slower learning for content than for context, and memory as low-entropy attractors. The third is staged structure-first learning, and the fourth hierarchical composition. It then reads predictive coding, the free energy principle, active inference and attractor dynamics as instances. Everything is asserted or argued informally. There are no experiments, and several proofs are wrong or missing. | Rejected — by LIT-374; read in full 2026-10-01 (NOTE-325); not worth a reader’s time as a source of results. Its foundation is the entropy chain rule plus an assumption, H(content) ≪ H(context), and from these it reads off a direction of inference. Of its eight theorems, one is proved only in sketch, one equivalence proposition is marked “proof skipped” and argued only in words (App. H), and the proofs of Theorems 6–8 swap the roles of content and context relative to the statements. The proof of Theorem 1 states the inequality backwards. Theorem 3 rests on D_KL(q‖p) ≥ H(q) − H(p), which is false in general. The abstract’s “computational simulations” are not in the paper. The author’s own v2 (June 2026) replaced the whole framework under a new title. |
| LIT-383 | Deciding Advantageously Before Knowing the Advantageous Strategy | Bechara, H. Damasio, Tranel & A. R. Damasio (1997), DOI-10.1126/science.275.5304.1293. In the four-deck gambling task, 10 controls and 6 ventromedial prefrontal patients were tracked on choices, anticipatory skin-conductance responses and verbal reports. All controls and no patients developed anticipatory responses; controls came to prefer the good decks, significantly from the point they reported a hunch; three patients who could state the correct strategy kept choosing the bad decks. The authors conclude that nonconscious biases guide behaviour before conscious knowledge does. The statistics support the group difference more firmly than the “before knowing” timing. | Active — read in full 2026-10-01 (NOTE-331); worth reading as the experiment the somatic marker hypothesis is usually cited for, and worth reading with its notes, which are weaker than its abstract. With 10 controls and 6 ventromedial patients, the robust results are that every control and no patient developed anticipatory skin-conductance responses, and that three patients who could state the right strategy still chose badly. The headline that controls chose well and responded to risky decks “before they knew” is only partly carried by the statistics: in the pre-hunch period the controls’ preference for the good decks was not significant (note 9), and the reported SCR effect there is a group-by-period interaction, not a bad-versus-good-deck difference (note 8). Controls chose the good decks significantly only once they reported a hunch. Whether the account is true is not this note’s question. |
| LIT-384 | The somatic marker hypothesis and the possible functions of the prefrontal cortex | Damasio (1996), DOI-10.1098/rstb.1996.0125. States the somatic marker hypothesis: ventromedial prefrontal cortex holds learned links between classes of situation and the body states (emotions) that went with them; when a situation recurs, those states are re-enacted, through the body or through an “as if” loop in somatosensory cortex, and mark imagined option-outcome pairs as good or bad, overtly as a feeling or covertly as a bias, so that reasoning works over a pruned space. Evidence summarised: ventromedial patients with intact intellect decide badly in life, lack skin-conductance responses to emotive pictures, and in the gambling task keep choosing the losing decks and show no anticipatory responses. | Active — read in full 2026-10-01 (NOTE-332); worth reading as the shortest statement of the somatic marker hypothesis by its author, with the evidence it then rested on. It is a hypothesis paper: the mechanism (ventromedial prefrontal linkages that re-enact body states, by a “body loop” or an “as if body loop”, which then mark options good or bad) is stated, not shown, and the evidence it reports is lesion and skin-conductance work summarised from other papers, without numbers. Its strongest piece, that ventromedial patients lack anticipatory skin-conductance responses in the gambling task while responding normally to the losses themselves, is reported here in one paragraph; the 1997 Science report filed alongside it is the place to see that result. Whether the account is true is not this note’s question. |
| LIT-385 v2 | Time-locked multiregional retroactivation: A systems-level proposal for the neural substrates of recall and recognition | Damasio (1989), Cognition 33:25–62. The convergence-zone proposal: memory is not held in one store but as feature fragments in early sensory and motor cortices, with amodal records of their co-occurrence held in local and non-local convergence zones downstream; recall is the time-locked reactivation of the fragments, directed from convergence zones through feedback projections. Unread: registered from the abstract. | Unavailable — filed 2026-10-01 at the owner’s request for Antonio Damasio’s work; no full text could be reached (the Elsevier article is closed, and no repository, author or course copy was found), so it is registered from its PubMed abstract only, unread, and no NOTE is filed |
| LIT-392 | Absent Qualia, Fading Qualia, Dancing Qualia | Chalmers (1995), in T. Metzinger (ed.), Conscious Experience (Imprint Academic, co-published in Paderborn). Principle: any system with the same fine-grained functional organisation as a conscious system has qualitatively identical experiences. Replace neurons gradually with silicon chips that keep the organisation. If the end point lacked experience, a midpoint would be a rational being wrong about all its experience (“fading qualia”). If it differed in experience, a switch between two circuits would flip experience with no possible noticing (“dancing qualia”). Both are judged implausible, so absent and inverted qualia are nomologically impossible. The argument extends explicitly to a population of homunculi. The result is “nonreductive functionalism”: organisation fixes experience with natural, not metaphysical, necessity. | Active — read in full 2026-10-02 (NOTE-347); worth reading as the full statement of the principle of organizational invariance and the fading- and dancing-qualia arguments for it. Cite it carefully: the principle is conditional on fine-grained functional isomorphism to a system already conscious. It covers silicon replacement and, by the paper’s own §3 extension, a population of homunculi that implements a human’s neural organisation. It does not cover a system whose organisation is merely similar in kind to a conscious one, so it does not license Schwitzgebel’s step to the United States (LIT-159). The conclusion is nomological, not metaphysical (“nonreductive functionalism”), and it is a plausibility argument by the author’s own account. |
| LIT-393 v2 | Artificial Intelligence as Philosophy and as Psychology | Dennett (1978), ch. 7 of Brainstorms; also in Ringle (ed.) 1979. Unread here. By Stich’s 1980 summary, which Dennett accepted: the worry that internal representations need a homunculus to read them (“Hume’s Problem”) is unfounded, because homunculi are harmless if they are stupider than the whole. A “team or committee of relatively ignorant, narrow-minded, blind homunculi” can produce intelligent behaviour, each explained by stupider ones, until they can be “replaced by a machine”; AI organises “armies of such idiots”. | Unavailable — registered 2026-10-02 from catalogue records and a contemporary summary, not read: no lawful full text could be reached. Brainstorms (1978) is lending-only at the Internet Archive, the MIT Press and Project MUSE chapters are paywalled (MUSE: “Open Access: No”), the Tufts Digital Library holds Dennett’s papers but not this essay, and PhilPapers was blocked. What the essay says about homunculi is known here only through Stich’s 1980 critical notice of Brainstorms, which quotes it with page numbers and which Dennett, replying in the same issue, said he had “no quarrel with”. It stays Unavailable until a copy is supplied. |
| LIT-394 | Computing Machinery and Intelligence | Turing (1950), Mind 59(236):433–460. “Can machines think?” is replaced by whether a digital computer can play the imitation game so that an interrogator, questioning by teleprinter, identifies it no more reliably than he tells a man from a woman. Universality reduces the question to one computer suitably programmed. Turing predicts that by about 2000 a machine with 10^9 bits of storage will leave an average interrogator no more than a 70 per cent chance of a right identification after five minutes. He answers nine objections, the argument from consciousness by its collapse into solipsism, and proposes reaching the goal by educating a child-machine. | Active — read in full 2026-10-02 (NOTE-355); worth reading as the source of the imitation game and as the first catalogue of objections to machine thought, nine of them, each still argued. Read it for what it actually proposes: not a definition of thinking but the replacement of an ill-posed question by a test, a prediction about 2000 (70 per cent right identification after five minutes, at 10^9 bits), and a programme for reaching it by educating a child-machine. Its reply to the argument from consciousness is an other-minds argument, not a theory of consciousness, and Turing says outright that he does not think there is “no mystery about consciousness”. |
| LIT-396 v3 | Psychological Predicates | Putnam (1967), in Capitan & Merrill (eds.), Art, Mind, and Religion (Proceedings of the 1965 Oberlin Colloquium in Philosophy), University of Pittsburgh Press, pp. 37–48; reprinted as “The Nature of Mental States”. It proposes, as an empirical hypothesis, that pain is a functional state of a whole organism, a state of a probabilistic automaton fixed by its machine table, rather than a brain state. Condition (3) of the hypothesis is the anti-nesting stipulation: no pain-feeler “possesses a decomposition into parts which separately possess Descriptions” of the pain-capable kind, introduced only “to rule out such ‘organisms’ (if they can count as such) as swarms of bees as single pain-feelers”. | Unfetched — not read in full: the original (Art, Mind, and Religion, pp. 37–48) is closed on JSTOR. The University of Pittsburgh’s page images sit behind a bot challenge I could not pass, and the Internet Archive copy of the 1975 reprint is print-disabled only. What was read is MIT’s 24.09x course excerpt of the 1975 reprint (“The Nature of Mental States”), recorded as a Skimmed reading in NOTE-335. That excerpt carries the four-clause hypothesis verbatim, including the anti-nesting condition (3) and its one-sentence motive. A course copy of the whole reprint was located at home.csulb.edu/~cwallis/382/readings/482/putnam.nature.mental.states.pdf. On 2026-10-30 it failed TLS verification from this client (curl error 60: the chain served ends at emSign Root CA - G1, which this client does not trust; verification was not overridden). It stays Unfetched until a copy is supplied, not on merit. Free copy: home.csulb.edu/~cwallis/382/readings/482/putnam.nature.mental.states.pdf. |
| LIT-398 v2 | True Believers: The Intentional Strategy and Why It Works | Dennett (1981; revised 1987, read in Mind Design II, 1997). Belief is “a perfectly objective phenomenon” that can be discerned only by one who adopts the intentional strategy. To be a true believer is to be a system “whose behavior is reliably and voluminously predictable” by it. The choice of stance is free, but its success is objective: Laplacean Martians who skipped it would miss real patterns. Thermostat and person differ only by degree. Why the strategy works is answered easily by evolution, and not yet at all by mechanism. | Active — read in full 2026-10-02 (NOTE-353), in the revised text reprinted in Mind Design II (1997), not the 1981 printing; worth reading as Dennett’s statement that belief is objective although only visible through a strategy: anything whose behaviour is “reliably and voluminously predictable” by the intentional strategy is a believer “in the fullest sense of the word”. It turns the 1971 paper’s observer-relativity (LIT-440) into “the facts about the success or failure of the stance … are perfectly objective”, introduces the Martian thought experiment and the intentional “real patterns” later worked out in LIT-220, and runs the thermostat-to-person continuum (“no magic moment”). Its criterion admits any system on which the strategy pays, groups included, but the essay never applies it to a group. |
| LIT-399 | Plain Talk About Neurodevelopmental Epistemology | Minsky (1977), MIT AI Memo 430, also IJCAI-77 pp. 1083–1092. Takes the mind as an organised society of simple “agents”, a theory credited to joint work with Papert, and asks how such agents could communicate without language. The proposal: no messages at all, since each argument lives at a fixed location on shared channels (“c-lines”) that the agents themselves constitute as short-term memory. A specificity gradient confines low-level agents to local sub-societies, and a case-shift moves an object of attention into a better-described slot. The paper conjectures that these internal cases precede, and shape, the cases of natural language. Openly speculative. | Active — read in full 2026-10-01 (NOTE-344); worth reading as the earliest of Minsky’s own papers in this record to state the society-of-agents view, then called “The Society of Minds” (Note 1), and as its first worked mechanism: agents communicate by fixed-location channels (“c-lines”) instead of messages. It is openly speculative (“a model of scientific irresponsibility”, p. 1085), with no data and no program. Read it for the communication scheme, the “specificity gradient”, the case-shift mechanism and the conjecture that cognitive cases precede and shape linguistic ones; do not read it as evidence for any of them. That it is the first published statement of the idea is the usual attribution, and this reading could not confirm it (see NOTE-344). |
| LIT-400 | Minds, brains, and programs | Searle (1980), Behavioral and Brain Sciences 3(3):417–424. A man who knows no Chinese, locked in a room with rules in English for manipulating Chinese symbols, can answer Chinese questions as well as a native speaker yet understands nothing; so instantiating a program is never by itself sufficient for intentionality (“strong AI” is false). Six replies are answered, the systems reply by letting the man internalise the whole room. Only something with causal powers equivalent to a brain’s could think; programs have syntax but no semantics. | Active — read in full 2026-10-02 (NOTE-349); worth reading as the source of the Chinese room and of the six replies (systems, robot, brain simulator, combination, other minds, many mansions) the debate still uses, in Searle’s own words. Read it for its scope: the target is “strong AI”, the claim that instantiating the right program is by itself sufficient for understanding, and Searle grants that machines, including artifacts with the brain’s causal powers, could think. The argument is about understanding and intentionality; consciousness is mentioned, not argued. Its answer to the systems reply (internalise the room) is the part that bears on minds of parts, and it is asserted rather than shown: Searle says the subsystem is “simply a part of” the man, which is what the reply denies. |
| LIT-402 | K-Lines: A Theory of Memory | Minsky (1979/1980), MIT AI Memo 516; Cognitive Science 4(2):117–133. “The function of a memory is to re-create a state of mind.” On a memorable event a K-node is made, and its K-line attaches to the agents then active. Reactivated, it re-imposes that partial mental state, but by the level-band principle only on an intermediate band of levels, so the present is seen as an instance of the past without hallucinating the old answer. K-lines attach mainly to earlier K-nodes, forming a K-pyramid against the perceptual P-pyramid, and cross-exclusion lets conflicting details cancel into abstraction. Read from the memo; a programme statement, with no simulation. | Active — read in full 2026-10-01 from the MIT AI Memo 516 text (NOTE-352), not the Cognitive Science version, which was not reachable; worth reading as the society-of-mind programme’s theory of memory and the first place Minsky’s own text uses the phrase “Society of Mind”. The idea to take away is that a memory re-creates a partial mental state rather than retrieving a stored description, and its level-band principle: re-enact only an intermediate band of the original state, so the present problem is seen as an instance of the remembered one. The second half is, by its author’s account, not constructive (“from this point, the reader can assume that difficulties in understanding are my fault”), and nothing is simulated or measured. |
| LIT-403 | Intrinsic content, active memory and the extended mind | Clark (2005), DOI-10.1093/analys/65.1.1. Against Adams & Aizawa: the intrinsic-content objection to LIT-097 fails, because the notion is unclear, silicon parts could carry it, and a conventional encoding linked at run-time to intrinsically contentful states can be a belief’s vehicle, inside or outside the head. Against the active-memory objection, the notebook’s passivity is a difference below the functional level that matters. The extended mind is a thesis about vehicles. | Active — read in full 2026-10-02 (NOTE-363); worth reading as Clark’s first reply to the critics of LIT-097, and short. It answers Adams and Aizawa’s intrinsic-content objection three ways (the notion is unclear, silicon could carry it, and it would not matter), turning their own concession that not every cognitive state needs non-derived content against them; and it answers the “active memory” objection by conceding the notebook is unlike biological memory and arguing that the differences sit below the level a functionalist should attend to. Its replies are parity arguments with imagined cases; it adds no new evidence. It does not engage Rupert (LIT-422). |
| LIT-404 | Does a rock implement every finite-state automaton? | Chalmers (1996), Synthese 108(3):309–333. Putnam’s proof that every open system implements every finite-state automaton fails, because implementation needs reliable, counterfactual-supporting transitions for every formal state. Repaired, it survives only for weak formalisms: a clock and a dial implement every inputless FSA, and an input memory and a dial upgrade any system with the right input–output dependencies to every FSA with them. Combinatorial-state automata, whose states are vectors of independently varying components in distinct physical regions, put a strong constraint on implementation. That keeps computation available as a foundation for the theory of mind. | Active — read in full 2026-10-02 (NOTE-342), from the author’s HTML copy, not the Synthese typeset text; worth reading as the standard reply to Putnam’s and Searle’s triviality arguments and as the source of the combinatorial-state automaton (CSA). Its result is real but partial: implementation needs counterfactual-supporting transitions over all formal states, inputless FSAs are trivially implemented by any clock plus dial, FSAs with input are implemented by any system with the right input–output dependencies plus an input memory and a dial, and CSAs escape only because a blind implementation must grow by a factor of n per step. Chalmers leaves two open problems (logically possible false implementations, and an overstrict one-region-per-component condition). For this record it decides which formalism a China-brain argument has to be stated in, and it lets one physical system host two minds. |
| LIT-406 | Facing Up to the Problem of Consciousness | Chalmers (1995), Journal of Consciousness Studies 2(3):200–219. The “easy” problems of consciousness (discrimination, integration, report, access, attention, control, wakefulness) are problems of explaining functions, and a mechanism that performs a function explains it. The hard problem is that after every such function is explained, “why is the performance of these functions accompanied by experience?” stays open, so reductive methods cannot answer it. Experience should be taken as fundamental, linked to the physical by psychophysical laws. Two nonbasic candidates are offered, structural coherence (the structure of experience mirrors the structure of awareness) and organizational invariance (fine-grained functional isomorphs have the same experiences), with a speculative double-aspect theory of information as the basic law. | Active — read in full 2026-10-02 (NOTE-336); worth reading as the source of the easy/hard distinction, and short. Cite it for three separable things: the hard problem as the claim that explaining every function near experience leaves “why is this accompanied by experience?” open (§§2–3); the outline of a nonreductive theory built from two nonbasic principles, structural coherence and organizational invariance (§7); and the double-aspect theory of information, which the author calls “more of an idea than a theory” (§8). The organizational-invariance argument here is a one-page sketch; the full version is LIT-392. |
| LIT-407 | Troubles with Functionalism | Block (1978), Minnesota Studies in the Philosophy of Science 9:261–325. Functionalism ties mental states to the world only through inputs and outputs, so it is “liberal”: a body run by a billion people with radios, each implementing one line of your machine table (the nation-of-China system), would be functionally equivalent to you, yet there is prima facie doubt that it has qualia. Putnam’s stipulation that no pain-feeler has parts with pain-feeler organisation is an ad hoc patch that fails on its own terms. Psychofunctionalism escapes the doubt for beliefs but perhaps not for qualia, and no specification of inputs and outputs avoids both liberalism and chauvinism. | Active — read in full 2026-10-02 (NOTE-366); worth reading as the source of the homunculi-headed robot and the nation-of-China system, and of the first published objection to Putnam’s anti-nesting stipulation. Cite it for two things that are easy to run together: the Absent Qualia Argument (a Functional simulation built from people is prima facie without qualia), and the case that Putnam’s condition is ad hoc (§1.5). Block grounds the doubt about the China system in its being unlike us psychologically and neurally and designed to mimic us, not in its parts having minds (n. 19). He also grants that a Psychofunctional homunculi-head could have beliefs (§2.1). So it is not a source for “a whole made of minded parts cannot be conscious”. |
| LIT-408 | Jokes and the Logic of the Cognitive Unconscious | Minsky (1980), MIT AI Memo 603; in Vaina & Hintikka (eds.), Cognitive Constraints on Communication (Reidel, 1984), pp. 175–200. Commonsense logic cannot be made reliable, so each mind accumulates “censors” that learn the precursors of known-bad thought (paradox, circling, false analogy) and suppress it before it occurs. On that view, humour is censor-learning. Its common element is an “unexpected frame-substitution”, and laughter both stops the absurd line of thought and holds it in focus. Logical nonsense is then as “naughty” as Freud’s forbidden wishes, which removes the gap in Freud’s theory. An essay with no data or program. | Active — read in full 2026-10-01 from the MIT AI Memo 603 text (NOTE-361), not the 1984 book chapter, which was not reachable; worth reading as the society-of-mind programme’s account of humour and of “negative meta-knowledge”. Its argument is that commonsense reasoning cannot be made reliable, so minds learn “censors” that suppress known-bad lines of thought, and that humour and laughter are how those censors are taught and engaged, which dissolves Freud’s puzzle about nonsense jokes. It is an essay: no data, no program, and its evolutionary story is offered as “no plea for this particular scenario”. |
| LIT-410 | A Framework for Representing Knowledge | Minsky (1974), MIT AI Memo 306. A self-declared partial theory of thinking: on meeting a situation one retrieves a frame, a stereotyped structure whose terminals carry conditions and weakly bound default assignments, and adapts it by replacing details. Frames that share terminals form frame-systems whose transformations stand for moves, actions or changes of viewpoint, and failed matches are routed through a learned similarity network. The memo applies this to vision, imagery, language, memory and problem solving, and ends by arguing that consistency-seeking logical systems cannot represent commonsense knowledge, because they are monotonic. Nothing is implemented or tested. | Active — read in full 2026-10-01 (NOTE-337); worth reading as the source of the frame idea and of the critique of “logistic” representation that later motivated non-monotonic reasoning. It is a proposal, not a result: no program is described as built, nothing is measured, and the author says so (“Apology!”, p. 2). The parts to take away are the frame with weakly bound default assignments (§1.11), frame-systems sharing terminals across viewpoints (§1), terminals read as questions (§2.8), the similarity network with cluster “capitols” as an account of family resemblance (§§3.4–3.5), and the appendix on monotonicity and consistency (§6). It contains no society-of-agents idea; that arrives in Plain Talk (LIT-399), which calls itself in part a sequel to this memo. |
| LIT-411 | Could a Large Language Model be Conscious? | Chalmers (2023), arXiv:2303.07103; Boston Review, 9 August 2023; from a NeurIPS 2022 invited talk. Asks for a feature X that LLMs have (or lack) and that indicates consciousness (or its absence). Self-report, seeming-conscious, conversation and general intelligence give at most weak evidence for it. Biology, senses and embodiment, world and self models, recurrent processing, a global workspace and unified agency are the evidence against. All but biology are temporary. On mainstream assumptions: under 10% credence for current LLMs, 25% or more for conscious LLM+ systems within a decade. Twelve challenges are offered as roadmap or red flags. | Active — read in full 2026-10-02 (NOTE-345), from arXiv v3 (the Boston Review text), with v1’s conclusions compared; worth reading as the clearest short statement of the case for and against consciousness in LLMs, and as the origin of the “theory-balanced” credence approach that Butlin et al. built on. Its method is a regimented request for a feature X, and its verdicts are credences offered “for illustrative purposes”: under 10% for current LLMs, 25% or more for conscious LLM+ systems within a decade, on “mainstream assumptions”. It does not break the deadlock THEORY-023 describes. It is an instance of it: mimicry undercuts self-report, the main obstacles are theory-derived, and biology is set aside rather than answered. |
| LIT-412 v2 | Supersizing the Mind: Embodiment, Action, and Cognitive Extension | Clark (2008), DOI-10.1093/acprof:oso/9780195333213.001.0001. Clark’s book-length defence of the extended mind: a tour of embodied, situated and extended cognition, a reply to its critics, and a case for treating “some coalitions of biological and non-biological resources as the extended cognitive circuitry of individual minds”. Unread: filed from the publisher’s description and contents. | Unavailable — searched 2026-10-14 for a lawful full text, none reached, so it is not read and no NOTE is filed. OpenAlex lists the OUP DOI, closed, a CiteSeerX cache and an institutional e-book handle (hdl.handle.net/2078/ebook:67780), which is licensed access, not an open copy; Unpaywall finds no open copy; Oxford Academic returned a Cloudflare 403; the Internet Archive holds the book only as controlled lending and as an ‘opensource’ upload of the in-copyright book, which was not used. The foreword (LIT-421) and the appendix (LIT-097) are read separately. It stays Unavailable until a copy is supplied. |
| LIT-413 v2 | The Mind’s I: Fantasies and Reflections on Self and Soul | Hofstadter & Dennett (eds.) (1981), Basic Books. An anthology of 27 selections, fiction and philosophy, on self, soul and consciousness, each followed by a “Reflections” by the editors. Among them are Turing’s imitation-game paper (LIT-394), Searle’s Chinese room (LIT-400), Nagel’s bat (LIT-096), Dennett’s “Where Am I?” and Hofstadter’s “Prelude … Ant Fugue” from GEB (LIT-437). Unread: registered from the catalogue record and a contents note; the Reflections, which are the editors’ own contribution, were not seen. | Unavailable — registered 2026-10-02 from the Library of Congress catalogue record, the publisher’s page and a library contents note, not read: no lawful full text could be reached. The Internet Archive scans of the 1981 Basic Books, Harvester, Bantam and Penguin printings are all lending-only, HathiTrust has no record for the LCCN, and no publisher, editor or Tufts-archive copy is posted. The editors’ Reflections are unread. Two of the selections were read in full from lawful copies of their first publications and filed on their own: Turing (LIT-394) and Searle (LIT-400). |
| LIT-414 | High-level perception, representation, and analogy: A critique of artificial intelligence methodology | Chalmers, French & Hofstadter (1992), DOI-10.1080/09528139208953747. Traditional AI hands its programs representations built by people who already know the answer, so models like BACON (Kepler’s third law) and the Structure-Mapping Engine (atom and solar system) skip the hard part, high-level perception: deciding what is relevant and how to organise it. Nor can the gap be filled later by a separate “representation module”, because analogy shapes perception and the task decides which representation is right. Copycat, in a microdomain of letter strings, is offered as an architecture that builds representations and mappings together. | Active — read in full 2026-10-01 (NOTE-359), from the authors’ preprint (CRCC Technical Report 49, March 1991) that Chalmers posts; the published text was not seen. Worth reading as the clearest short statement of the Fluid Analogies Research Group’s case against hand-coded representations, and as the most accessible description of Copycat. It is an argument, not a result: its two arguments against a separate “representation module” are informal and good, its case studies (BACON, SME) are fair as far as they go, and the evidence it offers for Copycat is a description of the architecture and one worked problem, with nothing measured in the paper itself. |
| LIT-415 | Extended Cognition and Extended Consciousness | Chalmers (2019), DOI-10.1093/oso/9780190662813.003.0002. The 1998 extended-mind thesis as stated is too weak, since opponents accept extension in wired-in “extended circuit” cases. The interesting thesis is that mental states can be partly constituted by external entities in virtue of the subject’s sensorimotor interaction with them. Consciousness does not extend this way, on the proposal that it requires information directly available for global control, and perceptually mediated availability is indirect. | Active — read in full 2026-10-02 (NOTE-356); worth reading as Chalmers’s own later statement of what the extended-mind thesis of LIT-097 is and where it stops. It does two things. It replaces the 1998 paper’s implicit thesis, which it concedes is too weak because even the critics accept it for wired-in “extended circuit” cases, with one that requires constitution “in virtue of the subject’s sensorimotor interaction” (§1). And it argues that consciousness does not extend, because consciousness needs information directly available for global control and perception makes availability indirect (§3). Both are proposals, argued in a short chapter: the thesis is offered as “extensionally correct”, and the directness account is offered as an explanation of why there are no plausible cases, not as a knockdown. |
| LIT-416 v2 | Is there an ‘I’ in AI? | Hofstadter (2026), DOI-10.1098/rsta.2024.0527. The lead essay of the Phil. Trans. A theme issue on world models whose editorial is LIT-195. Unread: what is known of it here comes from the editorial as read in NOTE-136, which quotes Hofstadter that a system passing a Turing test of consistent understanding should be credited with thinking, and that conscious AI may come “perhaps even by 2033”. Whether the essay argues “nearly the opposite” of the editors’ convergence verdict cannot be checked until the essay itself is read. | Unavailable — seeded 2026-10-01 from Crossref, PubMed and the record’s reading of the theme-issue editorial (LIT-195, NOTE-136), not read: royalsocietypublishing.org returned 403 to the DOI, article and PDF URLs, the article is not in PubMed Central or Europe PMC full text, and no repository or author copy was found. No NOTE is filed. |
| LIT-417 v2 | Epilogue: Analogy as the Core of Cognition | Hofstadter (2001), DOI-10.7551/mitpress/1251.003.0020. The epilogue to Gentner, Holyoak & Kokinov’s The Analogical Mind (MIT Press), pp. 499–538, whose title states the Fluid Analogies Research Group’s thesis that analogy-making is not one cognitive tool among others but the core of cognition. Unread: no lawful full text was reached. | Unavailable — seeded 2026-10-01 from Crossref and the MIT Press chapter record, not read: MIT Press Direct returned 403 to the chapter PDF and book page, OpenAlex lists the chapter as closed access, and no author or institutional copy was found. Third-party copies exist (a personal reference collection and an Internet Archive upload) but neither is the author’s or the publisher’s, so neither was used. No NOTE is filed. |
| LIT-418 v2 | Society of mind: A response to four reviews | Minsky (1991), Artificial Intelligence 48(3):371–396. Minsky’s reply, closing a review symposium, to four reviews of The Society of Mind in the same issue: by Michael G. Dyer, Matthew Ginsberg, George N. Reeke and Stephen W. Smoliar. Unread: the article is closed access and no copy was reached. Singh (2003) quotes it on why Minsky kept the words “agent” and “agency”, and draws on it for a mechanism by which agents with different representations could communicate. | Unavailable — seeded 2026-10-01 from the Crossref and Elsevier metadata of the article and of the issue it closes, not read: the article is closed access (Unpaywall and OpenAlex find no open copy, Semantic Scholar marks it closed, ScienceDirect refused the request), and no author copy was found. It stays Unavailable until a copy is supplied. The four reviews it answers are listed below from the issue’s contents, and are not filed. |
| LIT-419 v2 | Form, Function, and Feel | Lycan (1981), The Journal of Philosophy 78(1):24–50, DOI-10.2307/2025395. Filed as a statement of homuncular functionalism, the view Lycan’s Wikipedia entry says he advocates, and as the reply Schwitzgebel points to against Block’s nation-of-China intuition. Unread: no lawful full text was reachable, so its arguments are not stated here. | Unavailable — seeded 2026-10-02 from the bibliographic records (Crossref, the Philosophy Documentation Center), not read: no lawful full text could be reached. JSTOR and PDC are paywalled, Unpaywall and Semantic Scholar list no open copy, and no author copy was found. No NOTE is filed. What it argues is known here only from how Schwitzgebel and Block’s acknowledgements place it, which is not a reading. |
| LIT-420 v3 | Surfaces and Essences: Analogy as the Fuel and Fire of Thinking | Hofstadter & Sander (2013), Basic Books. The book defends “the simple but unorthodox premise” that “analogy is the core of all thinking”: analogy operates at every moment, from the choice of a word to the insights of physics, and is not a rare tool for hard problems. Its ten contents entries run from the evocation of words and phrases, through abstraction, “inter-category sliding” and naïve analogies, to “Analogies that shook the world” and a closing debate on the core of cognition. Unread: registered from the publisher’s description and the catalogue’s contents note. | Unavailable — registered 2026-10-01 from the publisher’s description and the Library of Congress catalogue record with its contents note, not read: no lawful full text could be reached. The Internet Archive scan is lending-only, and the full texts circulating online are not the publisher’s or the authors’. It stays Unavailable until a copy is supplied. |
| LIT-421 | Foreword | Chalmers (2008), DOI-10.1093/acprof:oso/9780195333213.002.0007. The co-author’s retrospective on the extended mind: his iPhone is “part of my mind already”. The one principled objection is that perception and action bound the mental. His reply is revisionist and contextualist: whether Otto’s belief counts as extended flips with explanatory purpose. No Otto-style twin case works for consciousness, so he tentatively keeps an internal conscious core. The thesis needs only a weak, parity-style functionalism and is compatible with dualism. | Active — read in full 2026-10-02 (NOTE-369); worth reading as the shortest statement of where Chalmers stands on the thesis he co-authored, ten years on. Three things in eight pages: the iPhone as a present-day Otto’s notebook; the perception-action boundary as the one principled objection, answered by explanatory revisionism and a “Necker cube” view on which whether Otto’s belief is extended depends on explanatory purpose; and a tentative internal “conscious core”, resting on the failure of any Otto-style twin case for consciousness. It is a foreword: every position is stated briefly and several are offered as “perhaps”. |
| LIT-422 | Challenges to the Hypothesis of Extended Cognition | Rupert (2004), DOI-10.5840/jphil2004101826. The embedded hypothesis (HEMC: cognition leans heavily on external props, but is the organism’s) competes with the extended one (HEC), and conservatism favours HEMC unless HEC buys explanatory power. It does not. LIT-097’s criteria either make phone directories into beliefs or privilege the organism, and memory research (interference, the generation effect, conversation) shows internal and external storage are different kinds. A generic kind covering both does no work. | Active — read in full 2026-10-02 (NOTE-343), in the author’s pre-publication version; worth reading as the case that a conservative “embedded” hypothesis explains everything the extended hypothesis does, and as the empirical version of the critique. Its two arguments: a dilemma on LIT-097’s past-endorsement criterion (keep it and the organism is privileged after all; drop it and directory assistance gives Otto beliefs about every listed phone number), and memory research (interference, the generation effect, conversation) showing internal and external storage behave as different kinds, so that a kind covering both does no explanatory work. Its strength is that it argues from cognitive science rather than intuition; its limit is that the empirical differences are about fine-grained profiles, which a defender can accept and still claim a coarse kind. |
| LIT-423 v2 | I Am a Strange Loop | Hofstadter (2007), Basic Books. Per its publisher, an “I” is “a strange loop where the brain’s symbolic and physical levels feed back into each other and flip causality upside down”, and the book asks whether that “I” is real or “merely a convenient fiction”, and how selves “can exist in the minds of others”. Its 26 contents entries run from souls and the causal potency of patterns, through Gödel’s “quintessential strange loop” and downward causality, to “How we live in each other”, “Consciousness = thinking” and zombies and dualism. Unread: registered from the description and the catalogue’s contents note. | Unavailable — registered 2026-10-01 from the publisher’s description and the Library of Congress catalogue record with its contents note, not read: no lawful full text could be reached. The Internet Archive scans are lending-only, and no publisher or author copy is posted. It stays Unavailable until a copy is supplied. |
| LIT-424 v2 | From Bacteria to Bach and Back: The Evolution of Minds | Dennett (2017), W. W. Norton. Per its publisher, the book shows “step by step, how a comprehending mind could in fact have arisen from a mindless process of natural selection”, with the decisive shift the sharing of memes, which produced “thinking tools” powerful enough that minds “create and comprehend”. Its contents run from the origin of reasons and “two strange inversions of reasoning”, through information, “brains made of brains”, words, memes and the origins of language, to consciousness as “an evolved user-illusion” and “the age of post-intelligent design”. Unread: registered from the description and the catalogue’s contents note. | Unavailable — registered 2026-10-02 from the publisher’s description and the catalogue record’s contents note, not read: no lawful full text could be reached. The Internet Archive scan is lending-only, Norton posts no excerpt beyond its description, and Dennett posted no draft chapters that I could find. It stays Unavailable until a copy is supplied. The claim the record wants from it, competence without comprehension, is not stated in the description or the contents, so nothing here places it. |
| LIT-425 | The Meta-Problem of Consciousness | Chalmers (2018), Journal of Consciousness Studies 25(9–10):6–61. The meta-problem is to explain, in terms that do not mention consciousness, why we judge that consciousness poses a hard problem. It is an “easy” problem with an empirical programme. Of thirteen candidate mechanisms, the most promising is an introspective model that attributes primitive relations of awareness to primitive qualities, with a sense of acquaintance. A realist theory should explain how its basis of consciousness produces these judgments, which IIT and biological theories struggle to do. Debunking arguments for illusionism can be blocked, but a residual sense of coincidence remains. Only strong illusionism dissolves the hard problem, and it is false because people feel pain. | Active — read in full 2026-10-02 (NOTE-354); worth reading as the statement of the meta-problem, explaining in topic-neutral terms why we judge that consciousness poses a hard problem, and as a survey of thirteen candidate mechanisms. Cite it for four things: the meta-problem as an “easy” problem with a research programme; the meta-problem challenge to theories of consciousness, which tells against IIT and biological theories; the analysis of debunking and coincidence arguments for illusionism; and the argument that only strong illusionism dissolves the hard problem, with a Moorean rejection of strong illusionism. Its own preferred solution is offered as “closer to a fleshed-out explanandum” than an explanation. |
| LIT-426 v2 | Macrocognition: A Theory of Distributed Minds and Collective Intentionality | Huebner (2014), DOI-10.1093/acprof:oso/9780199926275.001.0001. Per its publisher, the book argues that genuine cognition requires flexible goal-directed behaviour, which requires specialised representational systems integrated so as to cope fluidly with the environment, and so that collective mentality “should be posited where and only where specialized subroutines are integrated to yield goal-directed behavior that is sensitive to the concerns that are relevant to a group as such”. On this view there are many kinds of collective minds, some more like honeybees or cats than people, and groups are unlikely to be full “believers”. Unread: filed from the description and contents only. | Unavailable — filed 2026-10-02 from the publisher’s description and table of contents; no lawful full text was reachable (Oxford Academic is paywalled, the Internet Archive copy is lending-only, and no author-posted draft was found), so it is not read and no NOTE is filed. Of the four works filed for the Minsky–Schwitzgebel bridge it is the one whose criterion for a collective mind is itself a decomposition criterion: mentality wherever specialised subroutines are integrated into goal-directed behaviour sensitive to the group’s concerns. A close reading is the most likely of the four to change the bridge. |
| LIT-427 | Pandemonium: A Paradigm for Learning | Selfridge (1959), Mechanisation of Thought Processes Vol. I (HMSO), pp. 513–526 plus discussion to p. 531. Pattern recognition by a “pandemonium”: data demons hold the input, computational subdemons compute features, each cognitive demon “shrieks” a weighted sum of them, and a decision demon picks the loudest. The system learns by hill-climbing on the weights and by “natural selection” on the subdemons, culling those of low worth and breeding new ones by mutation and by “conjugation” of pairs. It is applied to telling dots from dashes in hand-keyed Morse. In the discussion McCarthy proposes reading the demons’ private computation as unconscious and their public shouting as conscious thought. | Active — read in full 2026-10-02 (NOTE-360); worth reading as the first published architecture of a mind-like process built from many quasi-independent “demons”, each doing one narrow job, with no part holding the whole’s competence. That is the decomposition template that Minsky’s agents inherit. Cite it as an engineering proposal with a preliminary result reported only in discussion, not as a theory of mind. Its one remark on consciousness is McCarthy’s, in the discussion, and it is an aside. |
| LIT-428 v2 | Fluid Concepts and Creative Analogies: Computer Models of the Fundamental Mechanisms of Thought | Hofstadter & the Fluid Analogies Research Group (1995), Basic Books. The group’s programme gathered in one volume: models of perception-driven analogy-making in microdomains (Seek-Whence, Jumbo, Numbo, Copycat, Tabletop, Letter Spirit, and the proposed Metacat), each chapter with a preface by Hofstadter, plus a critique of AI methodology and of how AI research is evaluated. Its contemporary reviewer in AI Magazine names two ideas under it: cognition as high-level perception, and the parallel terraced scan, by which coherent structure emerges from many probabilistic codelets without top-level control. Unread: registered from the description, the catalogue and that review. | Unavailable — registered 2026-10-01 from the publisher’s description, the Library of Congress catalogue record and a catalogue contents note, not read: no lawful full text could be reached. The Internet Archive scan is lending-only, HathiTrust is search-only, and no publisher or author copy of the book is posted. It stays Unavailable until a copy is supplied. |
| LIT-429 v2 | The Society of Mind | Minsky (1986), Simon & Schuster. The book-length statement of the Society of Mind programme, begun with Seymour Papert in the early 1970s. A mind is a society of many small agents, each mindless by itself, and intelligence comes from their diversity and organisation, not from one principle. The publisher presents it as a puzzle of self-contained one-page pieces that assemble into a theory. Unread: no lawful full text was reachable, so it is registered from the publisher’s description and the contents list. | Unavailable — seeded 2026-10-01 from the publisher’s description, the Library of Congress catalogue record and the Google Books contents list, not read: no lawful full text could be reached. The Internet Archive copies are lending-only, HathiTrust is search-only, and Minsky’s web page did not post the book. It stays Unavailable until a copy is supplied. Its programme is read here at second hand through Singh’s exposition (LIT-444) and at first hand through the draft of its sequel (LIT-432). |
| LIT-430 v2 | The Intentional Stance | Dennett (1987), MIT Press. Per its publisher, the “first full-scale presentation of a theory of intentionality” Dennett had been developing for almost twenty years: we understand each other by adopting a stance, “a predictive strategy of interpretation that presupposes the rationality” of whatever we predict. Six chapters reprint papers of the 1980s, each followed by a new “Reflections” essay; the first and the last three chapters are new. Chapter 2 is True Believers (LIT-398). Unread: registered from the description and a catalogue’s contents note. | Unavailable — registered 2026-10-02 from the publisher’s description and a library catalogue’s contents note, not read: no lawful full text of the book could be reached. The Internet Archive scans are lending-only (access-restricted, printdisabled), the MIT Press page and loc.gov were blocked, and the Tufts Digital Library holds earlier versions of several chapters as papers but not the book. Its chapter 2 is True Believers (LIT-398), read in full in its 1997 reprint, which is the revised book text. It stays Unavailable until a copy is supplied. |
| LIT-431 | Time and the Observer: The Where and When of Consciousness in the Brain | Dennett & Kinsbourne (1992), Behavioral and Brain Sciences 15(2):183–201, with open peer commentary. There is no place in the brain where “it all comes together” for consciousness. Discriminations are made once, locally, and revised in parallel “drafts”, none canonical. The brain represents temporal order by content, not by the timing of its own vehicles, as a dated letter does. Colour phi, the cutaneous rabbit and Libet’s timing results stop looking anomalous. Whether a revision was pre-experiential (Stalinesque) or post-experiential (Orwellian) has no answer at that scale. | Active — read in full 2026-10-02 (NOTE-367), the target article only; worth reading as the shortest complete statement of the Multiple Drafts model and of the case against “Cartesian materialism”, worked out on four timing phenomena (colour phi, the cutaneous rabbit, Libet’s backward referral and his delayed intention). Its sound core is the distinction between the time a content is represented and the time it represents. Its strong claim, that the Orwellian/Stalinesque distinction “need have no answer” below a few hundred milliseconds, is an argument from indistinguishability, not a demonstration. The open peer commentary (pp. 201–247) was not read. |
| LIT-432 | The Emotion Machine: Commonsense Thinking, Artificial Intelligence, and the Future of the Human Mind | Minsky (2006), Simon & Schuster, read in the author’s 2005 draft. Emotions are Ways to Think, not a separate faculty. Each emotional state turns some mental resources on and others off, by Selectors that Critics trigger when they recognise a type of problem. The same machinery serves ordinary thinking. The book adds six levels of mental activity, from instinctive reaction to self-conscious reflection. It treats “consciousness”, “emotion” and “pain” as suitcase words for many processes, and the Self as a network of partial self-models. It is a design argued from examples. Against Damasio, it treats the body as one more resource the brain uses, not the stuff of emotion. | Active — read in full 2026-10-01 in the author’s posted draft of 2005 (NOTE-348), not in the published book; worth reading as the late, fullest statement of the Society of Mind programme. It is the one where emotion is put inside thinking: an emotional state is a selection of mental resources switched on and off, made by Selectors that Critics trigger, and so it is a Way to Think. Read it as a design proposal argued by examples, not as evidence. It reports no experiment, Minsky labels several of its central ideas his own conjectures, and it sets the body to the side of emotion, against the bodily account of Damasio (LIT-384, LIT-388) that this record holds. |
| LIT-433 | Quining Qualia | Dennett (1988), in Marcel & Bisiach (eds.), Consciousness in Contemporary Science, OUP, pp. 42–77. Qualia as tradition has them are ineffable, intrinsic, private and directly apprehensible, and fifteen intuition pumps argue that nothing has all four. Two neurosurgical routes to an apparent colour inversion, one changing qualia and one only memory-linked reactions, feel the same from the inside. So the subject cannot know his own qualia better than an outside tester can. What is real is practical ineffability and idiosyncrasy: we refer to information-rich public properties through our own detectors of them. Hence “there simply are no qualia at all”. | Active — read in full 2026-10-02 (NOTE-357); worth reading as the founding statement of qualia eliminativism and the source of the post-Cartesian side in the record’s consciousness readings. It argues by fifteen intuition pumps, not by proof, and says so. Its lasting results are two: the alternative-neurosurgery and Chase-and-Sanborn cases, which show that neither the subject nor the physiologist can separate a change in qualia from a change in reactions to them; and a positive account of practical ineffability and privacy as idiosyncratic, information-rich property detection. Its conclusion, “there simply are no qualia at all”, is a tactical choice of words by the author’s own endnote 2, not a stronger thesis than that. |
| LIT-435 v2 | The emergence of understanding in a computer model of concepts and analogy-making | Mitchell & Hofstadter (1990), DOI-10.1016/0167-2789(90)90086-5. The Physica D statement of Copycat, the Fluid Analogies Research Group’s model of analogy-making in a microworld of letter strings, in which understanding is meant to emerge from many small, parallel, nondeterministic agents (codelets) interacting with a network of concepts (the Slipnet) under a computational temperature. Unread: what is said here comes from its co-authored companion paper LIT-414, which cites it for the details of Copycat’s “xyz” run. | Unavailable — seeded 2026-10-01 from Crossref, the Elsevier metadata API and the account of Copycat in Chalmers, French & Hofstadter (LIT-414, read in full), not read: the article is closed access at Elsevier, ScienceDirect returned 403, and no author or repository copy was found. No NOTE is filed. |
| LIT-436 | Intelligence without representation | Brooks (1991), Artificial Intelligence 47:139–159, written in 1987. Argues that AI should be built incrementally from complete autonomous “Creatures” tested in the real world, and that the system should be decomposed by activity into parallel layers, each connecting sensing to action, not by function into perception, a central reasoner and action. His robots have no central representation and no central control: each layer is a fixed network of simple finite-state machines, and a higher layer subsumes a lower one by suppressing or inhibiting its wires. Coherent behaviour emerges, “in the eye of an observer”, from competing behaviours. Hypothesis: representation is the wrong unit of abstraction for the bulk of intelligence. | Active — read in full 2026-10-02 (NOTE-351) from the copy Brooks posts; worth reading as the decomposition side built and run: a robot whose purposive behaviour comes from parallel layers of simple finite-state machines with no central representation and no central control, which Brooks likens to Minsky’s account of human behaviour. The coherence of the whole, he says, exists “in the eye of an observer”. The evidence is three layers on one physical robot, described rather than measured. It makes no claim about minds or experience and disclaims interest in the philosophy. |
| LIT-437 v2 | Gödel, Escher, Bach: an Eternal Golden Braid | Hofstadter (1979), Basic Books. The book that introduced the “strange loop”: per its publisher, an account of the “I”, self and consciousness built on the self-referential construction Gödel used to make a sentence assert its own unprovability. Chapters alternate with contrapuntal dialogues, chiefly between Achilles and the Tortoise, each built to exhibit the idea it discusses. Unread: no lawful full text was reachable, so it is registered from the publisher’s description and the catalogue record, and its year is the only date a source gives. | Unavailable — registered 2026-10-01 from the publisher’s description and the Library of Congress catalogue record, not read: no lawful full text could be reached. The Internet Archive scans of the 1979 Basic Books printing are lending-only, HathiTrust is search-only, and no publisher or author copy is posted. It stays Unavailable until a copy is supplied. |
| LIT-438 | The bounds of cognition | Adams & Aizawa (2001), DOI-10.1080/09515080120033571. Against LIT-097, Dennett, Donald and Hutchins: cognition has two necessary marks, non-derived content and a kind of causal processing found in brains. Notebooks, screens and tools have only derived content and work by different processes, so actual cognition is brain-bound as a contingent fact. Causal coupling does not extend a process, and brain-plus-tool systems are too motley for a science. They grant the parity principle. | Active — read in full 2026-10-02 (NOTE-368); worth reading as the first sustained reply to LIT-097 and the source of the “mark of the cognitive” line of criticism. It grants parity and argues that cognition has two necessary marks, non-derived content and a distinctive kind of causal process, which notebooks, screens and tools lack as a matter of contingent fact; and that brain-plus-tool systems are too motley to support a science. Read it for that structure, not for proofs: both marks are assumed (“we assume without argument”), and the processing differences it lists are asserted from common knowledge, not measured. The phrase “coupling-constitution fallacy” is not in this paper; it comes from the authors’ later work. |
| LIT-439 | The Computational Boundary of a “Self”: Developmental Bioelectricity Drives Multicellularity and Scale-Free Cognition | Levin (2019), Frontiers in Psychology 10:2688. Proposes “Scale-Free Cognition”: an Individual or Self is the set of parts that pursue a goal together, bounded by the farthest events in space and time it can measure and try to change (its “cognitive light cone”). Selves scale up from homeostatic loops through memory and anticipation, and cells merge into larger Selves by sharing bioelectric state through gap junctions; cancer is the boundary shrinking back to one cell. Because every level that pursues its own goal is a Self, most living systems hold many nested, coexisting Selves, “not one, as implied by Integrated Information Theory”. Levin’s “Self” is a goal-pursuing collective, an agent and an individual in this record’s terms; the essay gives no account of a self as subject. A hypothesis essay: no model and no new data. | Active — read in full 2026-10-02 (NOTE-341); worth reading as the composition-side statement that agency is nested by default: a Self is whatever set of parts pursues a goal together, so cells, organs, organisms and swarms are Selves at once, each bounded by its “cognitive light cone”. It is a hypothesis-and-theory essay with no model, no measurement procedure for the light cone and no data of its own. Its one sentence against IIT’s single-Self picture is asserted in a bracket, not argued, and its consciousness claims are confined to a closing speculation. |
| LIT-440 | Intentional Systems | Dennett (1971), Journal of Philosophy 68(4):87–106. An Intentional system is one whose behaviour can be explained and predicted by ascribing beliefs and desires and assuming rationality, which is the Intentional stance beside the design and physical stances. Being one is relative to an observer’s strategy, and the choice of stance is pragmatic. The stance is “unmetaphysical”: it sets aside composition, consciousness and moral standing. As psychology it is vacuous, because it presupposes rationality. Any theory that gives content to a part takes out a “loan of intelligence”, repayable only by analysing away the part’s reader: “there a little man is concealed”. | Active — read in full 2026-10-02 (NOTE-350); worth reading as the first statement in print of the intentional stance and of the homunculus debt that Dennett later answers with “discharge”. Its definition makes being an Intentional system relative to an observer’s predictive strategy, and it says outright that the definition “does not say that Intentional systems really have beliefs and desires” (p. 91). True Believers (LIT-398) later withdraws exactly that relativity. Its lasting tool for this record is the “loan of intelligence” (p. 96): any theory that gives a part content owes an account of that part’s reader, and “wherever a theory relies on a formulation bearing the logical marks of Intentionality, there a little man is concealed”. The paper does not treat a group as an Intentional system. |
| LIT-441 v2 | The Conscious Mind: In Search of a Fundamental Theory | Chalmers (1996), Oxford University Press. Per its introduction, the book argues that consciousness cannot be reductively explained and that materialism is false, and then moves toward a nonreductive theory with fundamental psychophysical laws. Its contents run from supervenience and explanation through the irreducibility argument, naturalistic dualism and the paradox of phenomenal judgment, to coherence principles, absent, fading and dancing qualia, and information. It ends with a defence of strong AI and a no-collapse reading of quantum mechanics. Unread: registered from the author’s posted contents, introduction and jacket copy. | Unavailable — registered 2026-10-02 from the author’s posted table of contents, introduction and jacket copy, not read: no lawful full text could be reached. The author posts only those three, not draft chapters. Its two central arguments for this record are readable in their paper versions, LIT-406 (the hard problem and naturalistic dualism) and LIT-392 (chapter 7’s organizational invariance). It stays Unavailable until a copy is supplied. |
| LIT-442 | The Self as a Center of Narrative Gravity | Dennett (1992), in Kessel, Cole & Johnson (eds.), Self and Consciousness: Multiple Perspectives, Erlbaum. A self is an abstractum, “a theorist’s fiction”, like a centre of gravity: no part of the brain, but a well-behaved posit of interpretation. It is like a fictional character, indeterminate beyond what its “text” fixes, and a selfless process can produce one (a novel-writing robot whose story is true of it). Brains are bundles of partly autonomous systems that make themselves cohere into one autobiography; when the story coheres around two points, there are two selves. Read in a copy-edited transcription. | Active — read in full 2026-10-02 (NOTE-364), in a third party’s copy-edited transcription of the text Dennett posted, not in the Erlbaum printing; worth reading as the short, canonical statement that a self is an abstractum, a “theorist’s fiction” of the same kind as a centre of gravity, posited by interpretation around which a life’s narrative coheres. Read it as an analogy and two thought experiments (the novel-writing robot “Gilbert”, the Updike sequel), not as an argument from evidence. Its evolutionary story of self-talk is labelled speculation, and its empirical support (Gazzaniga’s split-brain work, multiple personality cases) is cited, not shown. |
| LIT-443 v2 | Consciousness Explained | Dennett (1991), Little, Brown. Per its publisher, a “full-scale exploration of human consciousness” that “refutes the traditional, commonsense theory of consciousness and presents a new model, based on a wealth of information from the fields of neuroscience, psychology, and artificial intelligence”. Its fourteen contents entries run from “How are hallucinations possible?” and “A method for phenomenology”, through “Multiple drafts versus the Cartesian Theater” and “The architecture of the human mind”, to “Qualia disqualified” and “The reality of selves”. Unread: registered from the description and the catalogue’s contents note. | Unavailable — registered 2026-10-02 from the publisher’s description and the catalogue record with its contents note, not read: no lawful full text could be reached. The Internet Archive scans are lending-only (access-restricted, printdisabled), no publisher or author copy is posted, and the Tufts Digital Library holds Dennett’s papers but not the book or draft chapters of it. It stays Unavailable until a copy is supplied. |
| LIT-444 | Examining the Society of Mind | Singh (2003), Computing and Informatics 22(6):521–543. A sympathetic exposition of Minsky’s Society of Mind by his student. It traces the theory’s history: Minsky and Papert’s 1970s copy-demo robot, the procedure-bearing frames of 1974, the abandoned 1976 book drafts, and the 1977 and 1980 papers. It catalogues the mechanisms and how agents communicate without shared symbols. It finds resemblances in Bayes nets, case-based reasoning, blackboards, Cyc, Soar and multiagent systems, none of which it calls a Society of Mind. Its verdict is that the theory has never been implemented. | Active — read in full 2026-10-01 (NOTE-339); worth reading as the shortest reliable map of The Society of Mind (LIT-429), written by the student who worked with Minsky on its sequel. It gives the theory’s history from the 1970s memos, a catalogue of its mechanisms (agents and agencies, K-lines, nemes and nomes, frames and transframes, difference-engines, censors, A- and B-brains, ways agents communicate and grow), and a survey of later AI it resembles. It is an exposition, not a test: it reports no implementation and quotes the theory more than it argues for it. Its history leans on unpublished 1976 drafts that only the author saw. |
| LIT-450 | Self models | Metzinger (2007, rev. 2013), Scholarpedia 2(10):4174. There is no substantial self, only a phenomenal self-model (PSM): the conscious part of an organism’s self-model, an instrument of global control. A phenomenal self arises “if and only if a system operates under a transparent PSM”, one it cannot introspectively recognise as a model; a first-person perspective is a transparent model of the intentionality relation (PMIR). Self-models can be unconscious and artificial, as in Bongard, Zykov and Lipson’s limping starfish robot; embodiment comes in three orders, the third conscious. Full-body illusions are offered as the experimental handle. | Active — read in full 2026-10-02 (NOTE-372); worth reading as Metzinger’s own short, open statement of the self-model theory of subjectivity (SMT) behind Being No One (LIT-467): there is no self, only a phenomenal self-model, and a system experiences itself as a self when that model is transparent, that is, when it cannot introspectively recognise it as a model. It is an encyclopedia article, not an argument: the central claim is stated as a definition with a prediction attached, its empirical examples are cited, and the online text is a revision later than the 2007 acceptance, with editing slips. |
| LIT-451 | Of Personal Identity | Hume (1739), Treatise 1.4.6, with the Appendix (1740). There is no impression of a self, “constant and invariable”, so no idea of one: introspection finds only “some particular perception or other”, and the mind is “a bundle or collection of different perceptions”. The identity we ascribe to it is “fictitious”, like that of plants, ships and churches: an easy transition of the imagination along resemblance and causation, with memory discovering rather than producing it, so that hard cases are “grammatical” rather than philosophical. The Appendix retracts the account of the connecting principle: two principles he cannot reconcile or renounce leave the matter “too hard for my understanding”. | Active — read in full 2026-10-02 (NOTE-386), in the Merivale–Millican critical text on davidhume.org, with the Appendix’s second thoughts (App. 10–21); worth reading as the founding statement of the bundle view and the first published retraction of it. The section finds no impression of a self, calls the mind “a bundle or collection of different perceptions”, and explains our ascription of identity to it as a fiction of the imagination run on resemblance and causation, the same fiction that makes a repaired ship or a rebuilt church “the same”. The Appendix then confesses that it cannot explain what unites the perceptions and leaves the problem “too hard for my understanding”. The section’s case is introspective and associationist; its Appendix is the clearest statement of what any bundle view owes. |
| LIT-452 | Against Narrativity | Strawson (2004), Ratio 17(4):428–452. Rejects both the psychological Narrativity thesis (people do experience their lives as a story) and the ethical one (they should, to be full persons), naming Sacks, Bruner, Dennett, Taylor, MacIntyre and Schechtman. Distinguishes Diachronic self-experience (the self I am now was there in the past) from Episodic (it was not, though the human being was), and splits Narrativity into form-finding, story-telling and revision, with only story-telling sufficient. Episodics are normal and can live well; self-narration risks falsifying the self it describes. Argued from the author’s own case. | Active — read in full 2026-10-02 (NOTE-383), in the published typeset text; worth reading as the standard objection to narrative accounts of the self, and for its distinctions: Episodic versus Diachronic self-experience, and form-finding, story-telling and revision as separable parts of “Narrativity”. Read it as a report of a temperament argued from the author’s own case and literary examples. Its empirical claims (a genetic “individual difference variable”, the spread across cultures) are asserted, not shown, and its central metaphysical claim, that the remoter past did not happen to “me*”, is stated rather than argued. |
| LIT-453 | Philosophical conceptions of the self: implications for cognitive science | Gallagher (2000), Trends in Cognitive Sciences 4(1):14–21. Divides philosophical accounts of the self into the “minimal self”, accessible to present self-consciousness without temporal extension, and the “narrative self”, extended in time through memory and story. The minimal self splits into a sense of ownership and a sense of agency, which come apart in involuntary movement and, after Frith, in schizophrenic thought insertion when a forward comparator fails. Neonatal imitation suggests an embodied minimal self; Strawson’s and Tani’s are momentary. Against Dennett’s narrative centre, the self is “the sum total of its narratives”. Read in the author’s preprint. | Active — read in full 2026-10-02 (NOTE-377), in the author’s preprint, not the typeset article; worth reading as the paper that set the minimal self against the narrative self for cognitive science, and for the split of the minimal self into a sense of ownership and a sense of agency, which it ties to Frith’s comparator model of schizophrenic thought insertion. It is a short review: the empirical support is cited, not shown, and its own proposal, the self as “the sum total of its narratives”, is offered in two paragraphs. |
| LIT-455 | Defending the Bounds of Cognition | Adams & Aizawa (2010), DOI-10.7551/mitpress/9780262014038.003.0004, read from the pre-publication draft. A reply to a manuscript of Clark’s “Memento’s Revenge”. It names the “coupling-constitution fallacy”: that X is coupled to a cognitive agent does not make X part of its cognition. It restates the 2001 marks of the cognitive, non-derived content and a distinctive causal process, and defends them: a mental image of Euler circles need not get its content by convention, and a “brain-tool science” would have to cover an open-ended motley of artifacts. | Active — read in full 2026-10-02 (NOTE-380), from a pre-publication draft, not the published chapter; worth reading as the shortest statement of Adams and Aizawa’s case after LIT-438, and as the earliest text the record holds that names the “coupling-constitution fallacy”. It answers a manuscript of Clark’s “Memento’s Revenge” (the chapter before it in the same volume), restating the two marks of the cognitive and defending them against Clark’s Euler-circle and “brain-tool science” objections. It adds no new argument for either mark: the fallacy is shown by one analogy (nerve and muscle), and the intracranial processing differences are asserted from consumer electronics, not measured. |
| LIT-456 | Reply to Professor Stich | Dennett (1980), Philosophical Books 21(2):73–76, replying to Stich’s critical notice of Brainstorms, “Headaches” (pp. 65–73). Accepts Stich’s summary. On rationality: the norm is grounded only minimally in evolution, and full rationality is an idealisation, so personhood rests on a concept we “only imperfectly achieve”. On pain: there are no pains in a mature psychology’s ontology, yet people feel pain; between biting the bullet, reform and common sense (pain is like holes, voices and haircuts), the choice is “a matter of policy”. Philosophy supplies the medicine for the puzzles, science the rest. | Active — read in full 2026-10-02 (NOTE-371); worth reading as Dennett’s shortest statement of two positions he kept: the intentional stance grounds belief in an idealisation of rationality, so persons are believers by a standard “we already know we only imperfectly achieve”; and pain is not a good theoretical entity, yet people feel pain, a choice among biting the bullet, reform and common sense that is “in some measure a matter of policy”. It is also the text behind LIT-393’s account of the homunculus essay, which it confirms: Stich’s quotations stand as LIT-393 gives them, and Dennett’s acceptance of Stich’s summary is a blanket one. |
| LIT-457 | Fluid Concepts and Creative Analogies: A Review | Burns (1995), DOI-10.1609/aimag.v16i3.1155. A review of Hofstadter and FARG’s Fluid Concepts and Creative Analogies (LIT-428). It finds two ideas under the book: cognition modelled by a parallel terraced scan of probabilistic codelets, and cognition as high-level perception. It asks why the work has had little impact on AI, and answers: microdomains, and critiques of rival programs that rest on analogy and intuition, applied to others but not to Copycat. It recommends the book. | Active — read in full 2026-10-02 (NOTE-375); worth reading as the one contemporary account the record holds of what is in LIT-428, which is unread, and as a short, fair statement of the two standing complaints against the Fluid Analogies programme: every model lives in a microdomain, and rival programs are dismissed by analogy and intuition while the same tests are waved off for Copycat. It is a three-page book review. Its account of the parallel terraced scan is clear and checkable; its verdict is the reviewer’s opinion, argued with one worked example (symbol grounding in SME versus Copycat). |
| LIT-458 | Illusionism as a Theory of Consciousness | Frankish (2016), Journal of Consciousness Studies 23(11–12):11–39, the target article of the issue later published as Illusionism as a Theory of Consciousness (2017). Names “illusionism”: phenomenal properties are not instantiated, and the task is the “illusion problem” of explaining why experiences seem to have them. Strong illusionism (no phenomenal properties at all, only “quasi-phenomenal” physical ones introspection misrepresents) is separated from weak illusionism (real phenomenal properties that merely seem ineffable, intrinsic, private), which it argues collapses into the strong form or into radical realism. | Active — read in full 2026-10-02 (NOTE-384); worth reading as the founding statement of illusionism under that name, and the target article that Dennett’s “Illusionism as the Obvious Default Theory of Consciousness” (LIT-434) answers. It names the “illusion problem”, draws the strong/weak illusionism line with “quasi-phenomenal” properties, and argues from conservatism, anomalousness and the debunking argument that strong illusionism should be the research programme. It states a programme, not a theory: by its own account it “will not propose a detailed illusionist theory” (p. 1 of the eprint), and its positive arguments are abductive and openly so. |
| LIT-461 | Consciousness interpreted: an interpretation of Dennett’s view of consciousness | Taylor (2024), DOI-10.1080/09515089.2024.2433526. Interprets Dennett’s theory of consciousness as realism about a real pattern. Consciousness is a real pattern in the brain’s content-bearing states (drafts), and a user-illusion: a pattern so coarse that describing the brain through it is metaphor. On that reading multiple drafts and global workspace theory are compatible, qualia eliminativism is peripheral, the Orwellian and Stalinesque readings of masking are equally good descriptions, and first-person reports are fallible as literal descriptions of the brain but authoritative as unwitting fictions. | Active — read in full 2026-10-02 (NOTE-379); worth reading as the short, charitable reconstruction of Dennett’s theory of consciousness that ties it to “Real Patterns” (LIT-220): consciousness is a real pattern in the brain’s content-bearing states (drafts), of the coarse, metaphorical kind Dennett calls a user-illusion. It reconciles six claims, including multiple drafts with global workspace theory and fallibility with “dictatorial” authority. It is an interpretation offered as consistent, not as correct or as Dennett’s own, and its key step on metacontrast masking is a plausibility argument. |
| LIT-463 v2 | Subjectivity and Selfhood: Investigating the First-Person Perspective | Zahavi (2005), MIT Press. Per its publisher, a defence of the self against neuroscientists and philosophers who doubt it, arguing that “the notion of self is crucial for a proper understanding of consciousness” and that experience, self-awareness and selfhood cannot be understood apart. It uses Husserl, Heidegger, Sartre and Merleau-Ponty to give an account of the “experiential givenness of the self” for philosophy of mind, cognitive science, developmental psychology and psychiatry. Its chapters run from self-awareness and phenomenal consciousness through time-consciousness and reflection to self and other, theory of mind, autism and embodiment. Unread: registered from the description and the contents. | Unavailable — registered 2026-10-02 from the publisher’s description and the chapter list on Crossref, not read: no lawful full text could be reached. The MIT Press Direct edition is not open access (Crossref carries no licence, OpenAlex reports it closed), and no author copy of the book is posted. It stays Unavailable until a copy is supplied, not on merit. |
| LIT-467 v3 | Being No One: The Self-Model Theory of Subjectivity | Metzinger (2003), MIT Press. Per its publisher, “no such things as selves exist in the world: nobody ever had or was a self. All that exists are phenomenal selves”, and the phenomenal self is “the content of a ‘transparent self-model’”. A representationalist and functional analysis of the conscious first-person perspective, with conceptual “tools”, constraints on consciousness and neurophenomenological case studies (agnosia, neglect, blindsight, hallucination), asking whether being someone can be analysed at subpersonal levels. Unread: registered from the description and the contents; the theory is read in LIT-450. | Unavailable — registered 2026-10-02 from the publisher’s description and the chapter list on Crossref, not read: no lawful full text of the book could be reached. The MIT Press Direct edition is not open access and no author copy is posted. Its theory is read in Metzinger’s own short statement, “Self models” (LIT-450, read in full), and more fully in his “Précis: Being No One” (PSYCHE 11(5), 2005; LIT-568, read in full 2026-10-03), the read substitute for the book, which sets out six of its ten constraints and the phenomenal model of the intentionality relation. Retried 2026-11-01: OpenAlex still reports it closed with no open location; it is not in the Internet Archive’s MIT Press open-access collection, which holds Metzinger’s 2024 book (LIT-1943); MIT Press Direct returned a Cloudflare challenge; no author copy found. It stays Unavailable until a copy is supplied, not on merit. |
| LIT-483 v3 | Groups as Agents | Tollefsen (2015), Polity. Per its publisher, a survey of whether groups can literally intend, whether there is a collective mind, and whether groups should be held morally responsible, arguing “the unorthodox view that certain groups should, indeed, be treated as agents and deserve to be held morally accountable”. Chapters on group belief, intention, agency, cognition, interpreting groups and the moral responsibility of groups. Unread: filed from the description and contents only. | Unavailable — filed 2026-10-02 from the publisher’s description and table of contents; no lawful full text was reachable (the book is in print with Polity, licensed as an e-book, and no author-posted draft was found), so it is not read and no NOTE is filed. It is a short introduction to group belief, intention, agency, cognition and responsibility that, per its publisher, argues “certain groups should, indeed, be treated as agents and deserve to be held morally accountable”; which groups, and on what view of agency, is not recorded here. |
| LIT-494 v3 | The Theory of Dyadic Morality: Reinventing Moral Judgment by Redefining Harm | Schein & Gray (2018), DOI-10.1177/1088868317698288. By its abstract: harm is an intuitively perceived continuum, not an objective matter, and acts are condemned in proportion to norm violation, negative affect and perceived harm, where harm is dyadic (an intentional agent damaging a vulnerable patient). Harm leads to condemnation and condemnation to perceived harm (the “dyadic loop”). The theory argues against intuitively harmless wrongs and modular foundations. Unread: the argument and evidence are not known here. | Unfetched — filed 2026-10-02 from the PubMed abstract and Crossref record, not read: no lawful full text was reachable. The SAGE PDF, listed as free to read, sits behind a bot challenge (HTTP 403); the PubMed record has no PMC copy; a University of North Carolina repository file returned by a search for the paper (cdr.lib.unc.edu/downloads/w9505f334) reset every connection, so what it holds is unknown; Semantic Scholar points only to the SAGE PDF, and CORE, OpenAlex and the Wayback Machine had no copy. It is the main harm-based rival to morality-as-cooperation, and what it argues is known here only from its abstract and from how other readings in this batch describe it. It stays Unfetched until a copy is supplied. Free copy: doi:10.1177/1088868317698288 (SAGE’s PDF, listed as free to read). |
| LIT-499 | Moral Foundations Theory: The Pragmatic Validity of Moral Pluralism | Graham et al. (2013), DOI-10.1016/B978-0-12-407236-7.00002-4. Moral Foundations Theory in full: a nativist, cultural-developmental, intuitionist and pluralist account with five foundations (Care, Fairness, Loyalty, Authority, Sanctity), each tied to an adaptive challenge and an existing evolutionary model. A decade of MFQ work finds liberals rely on Care and Fairness, conservatives on all five, across 12 world regions. Five criteria for “foundationhood” are proposed, and Liberty, Efficiency, Ownership and Honesty are named candidates. Harm monism (Gray) and fairness monism (Baumard) are rejected as Procrustean. | Active — read in full 2026-10-02 (NOTE-396); worth reading as the full statement of Moral Foundations Theory and as the text to check morality-as-cooperation’s comparison against. It states the theory’s four claims (nativism, cultural learning, intuitionism, pluralism), reviews a decade of MFQ findings, mostly on political ideology, and offers five criteria for what counts as a foundation. By its own table the Care foundation rests on kin selection and Fairness on Trivers’ reciprocal altruism, and Ownership/theft is already a named candidate, so the claim that MFT omits kin altruism, reciprocal altruism and property does not survive this text. Sanctity, the foundation morality-as-cooperation excludes, is by the same table the one with the least evidence of innate preparedness. |
| LIT-502 | Intuitive ethics: how innately prepared intuitions generate culturally variable virtues | Haidt & Joseph (2004), DOI-10.1162/0011526042365555. Proposes that people are innately prepared to feel flashes of approval or disapproval at four patterns: suffering, hierarchy, reciprocity and purity, with ingroup the best candidate for a fifth (fn. 15). Each is a module with a proper domain it evolved for (one’s own child in distress; cheating in joint ventures; disease and waste) and a culturally widened actual domain. Cultures build virtues on these intuitions and differ in which they use. The four were chosen by counting which concerns appeared in five surveyed works; the essay reports no data. | Active — read in full 2026-10-02 (NOTE-400); worth reading as the founding statement of Moral Foundations Theory, and as the text morality-as-cooperation quotes when it calls purity non-cooperative. It proposes four innately prepared intuition “modules” (suffering, hierarchy, reciprocity, purity), with ingroup as a likely fifth in a footnote, and argues that cultures build virtues on them. The list comes from tallying five books, not from data, and no study is reported. It already grounds the suffering module in care for one’s own children and the reciprocity module in Trivers, so the later charge that the foundations omit kin and reciprocal altruism does not fit this text. |
| LIT-504 v2 | A Natural History of Human Morality | Tomasello (2016), DOI-10.4159/9780674915855. By the publisher’s account, a reconstruction from experiments comparing great apes and children of how early humans became “an ultra-cooperative and, eventually, a moral species”. The chapter titles give the steps: an interdependence hypothesis, a second-personal morality of collaborating partners, an “objective” morality of the cultural group, and human morality as “cooperation-plus”. Unread: filed from the description and chapter list. | Unavailable — filed 2026-10-02 from the publisher’s description and the chapter list, not read: no lawful full text was reachable. The book is lending-only at the Internet Archive (access-restricted, printdisabled) and paywalled at Harvard University Press and De Gruyter. Its relation to Tomasello’s later “The moral psychology of obligation” (LIT-125, read in NOTE-185) is inferred from the chapter titles and that reading, and is labelled as inference. An author’s précis exists (Philosophical Psychology, 2018, DOI 10.1080/09515089.2018.1486605) and was not read; it would be the natural first step. It stays Unavailable until a copy is supplied. |
| LIT-511 | The emotional dog and its rational tail: A social intuitionist approach to moral judgment | Haidt (2001), DOI-10.1037/0033-295X.108.4.814. Moral judgment is usually caused by fast affective intuitions; moral reasoning mostly comes after, to justify the judgment to others, and changes minds chiefly by triggering new intuitions in listeners. Four reasons are offered: two processing systems, motivated reasoning, post-hoc confabulation, and the weak link between moral reasoning and moral action. Intuitions are partly innate and are tuned by culture through custom complexes and peers in late childhood. Harmless taboo violations are condemned anyway, and people are “dumbfounded” when their reasons run out. | Active — read in full 2026-10-02 (NOTE-403), from the author’s final manuscript; worth reading as the standard statement of the social intuitionist model, the proximate account of moral judgment that morality-as-cooperation lacks, and as the source of the “harmless wrongs” (consensual incest, sex with a dead chicken) on which the purity and sex objections to cooperative theories rest. Most of its evidence comes from outside the moral domain, which the author says himself: the model is offered as “a plausible alternative”, not an established fact. |
| LIT-529 v2 | Incomplete Nature: How Mind Emerged from Matter | Deacon (2011), W. W. Norton. An account of how life, mind and purpose (“ententional” phenomena) emerged from physics, built on constraint and absence rather than on added stuff. Its contents run from absence, homunculi and teleonomy through emergence and constraint to a hierarchy of dynamics (homeodynamics, morphodynamics, teleodynamics), a minimal self-producing system (autogenesis), and then work, information, significance, evolution, self, sentience and consciousness. Unread: registered from the publisher’s description and the contents; the text was not reached. | Unavailable — registered 2026-10-02 from the publisher’s description and a library contents note, not read: no lawful full text could be reached. The Internet Archive scan of the Norton printing is lending-only, and no author-posted draft was found. The companion chapter the record already cites, Deacon & Cashman (2013), “Teleology versus mechanism in biology: beyond self-organization”, is identified below but was not reachable either and is not filed. |
| LIT-536 v4 | Biological Autonomy: A Philosophical and Theoretical Enquiry | Moreno & Mossio (2015), Springer, History, Philosophy and Theory of the Life Sciences 12. The book argues that biology should revolve around autonomy, not only natural selection: organisms are organized systems that self-produce and self-maintain as integrated wholes, set their own goals and norms, and promote the conditions of their existence through interaction with their environment. Chapters treat constraints and organisational closure, emergence and inter-level causation, teleology and function, agency, evolution, organisms and levels of autonomy, and cognition. Only the Introduction is read (NOTE-1437): there the authors make closure of constraints the core of biological organisation, say it differs fundamentally from the self-organisation of dissipative structures, and announce that it grounds teleology, normativity and function; the chapters that argue this are unread. | Unfetched — registered 2026-10-02 from the publisher’s description and the chapter list. The book is not open access at Springer. Its only free copy, the authors’ HAL deposit (hal-01151449, hal.science/hal-01151449/document, CC BY-NC 4.0, deposited 13 May 2015), is not the whole book, as this note earlier said: it holds the front matter (Hooker’s Foreword and the contents), the Introduction (pp. xix–xxxiv), the references and the index, and no chapter. The Introduction was read in full on 2026-10-30 (NOTE-1437, Skimmed, because only the Introduction was read). For chapters 1–8 no fuller lawful copy was found: on 2026-10-30 OpenAlex listed the HAL deposit as the book’s only open copy, and the publisher’s chapters are paywalled. HAL also served its anti-bot page to a direct fetch on 2026-10-02 and again on 2026-10-30; the deposit was read from a copy retrieved earlier the same day. It stays Unfetched for the chapters. |
| LIT-538 | How effective psychological treatments work: mechanisms of change in cognitive behavioural therapy and beyond | Salkovskis, Sighvatsson & Sigurdsson, Behavioural and Cognitive Psychotherapy 51(6):595–615, in the special issue “Disseminating CBT: 50 years and beyond” (issue dated November 2023, first published online 5 January 2024). A history of behaviour therapy and CBT as empirically grounded treatment development, an argument that common factors are not mechanisms of change, and a proposal: the mechanism of all effective therapies is a shift from a stuck, threat-laden meaning (“Theory A”) to a collaboratively built and tested alternative (“Theory B”), which the authors call a type of psychological flexibility. | Active — read in full 2026-10-03 (NOTE-428), from the open-access (CC BY 4.0) publisher PDF; worth reading as a senior Beckian’s statement of a single mechanism of change for all effective psychological therapy, and for its clear separation of “common factors” from “mechanisms of change”. Its thesis is a proposal argued from the history of CBT and clinical examples, not a tested result: people are “stuck” in an exaggerated negative or threatening meaning (Theory A) kept alive by what it makes them do, and any effective therapy works by helping them build and test a less threatening alternative (Theory B). The paper calls this a particular type of psychological flexibility, without citing the ACT literature that uses that name. |
| LIT-539 | Explanatory Models for Psychiatric Illness | Kendler (2008), American Journal of Psychiatry 165(6):695–702. Psychiatry should explain by mechanisms, not laws. Psychiatric causes act at many levels, within and outside the person, and are non-aggregative and loop-ridden, as an alcohol-dependence sketch shows (gene–environment interaction, tolerance feedback, niche selection, aversive cultural transmission). So “hard reductionism” and “hard emergentism” both fail, but locally decomposable subsystems can still be studied and then reassembled. After Marr, biology will implement psychological functions, not replace them. | Active — read in full 2026-10-03 (NOTE-432), from the NIH author manuscript in PubMed Central; filed as the readable substitute for Kendler’s 2005 “Toward a Philosophical Structure for Psychiatry” (LIT-552), which has no lawful open copy. It opens by citing the 2005 paper for its premise that etiological models must be pluralistic, and develops two of the 2005 abstract’s propositions at length: piecemeal integration becomes “decomposition and reassembly” of multilevel mechanisms, and the mind–brain question becomes Marr-style implementation, in which “biology will implement but not replace psychology”. It does not restate the 2005 paper’s arguments against dualism and epiphenomenalism, so it stands in for the pluralism, not for the whole of the earlier paper. |
| LIT-540 | Impulsive action: emotional impulses and their control | Frijda, Ridderinkhof & Rietveld (2014), Frontiers in Psychology 5:518. Impulsive action is non-deliberate but purposive. Automatic appraisal of an event’s pertinence to the agent’s concerns elicits a state of action readiness, a motive to establish, modify or end a relation to the object, with “control precedence” (urgency, persistence, attentional capture). Modes of action readiness are what distinguish one emotion from another. Events usually draw several appraisals and so several states of readiness, and their interaction regulates action: much emotion regulation is itself emotional, not reason overriding passion. | Active — read in full 2026-10-03 (NOTE-433); filed in place of Frijda’s The Emotions (LIT-546), which could not be read, as the shortest lawfully readable statement of his action-readiness theory in his own words. Appraisal of an event’s pertinence to one’s concerns elicits a state of action readiness, a motive to change one’s relation to the object, which has “control precedence”. Modes of action readiness correspond to emotions (hostility to anger, avoidance to fear). Much emotion regulation is itself one state of action readiness checking another, so “there is no clear distinction between processes of emotion and processes of emotion regulation”. It is a theoretical paper with no data, written for impulsive action, so it states the theory’s core and not its full range. |
| LIT-541 | Relational Frame Theory: Some Implications for Understanding and Treating Human Psychopathology | Barnes-Holmes, Barnes-Holmes, McHugh & Hayes (2004), International Journal of Psychology and Psychological Therapy 4(2):355–375. RFT stated briefly: humans relate events bidirectionally and arbitrarily, as a generalized operant under contextual control, so words take on the functions of what they are related to. Frames combine into networks, networks are related to networks (metaphor), rules are networks that regulate behaviour, and the self is content, process and context of verbal relations. Psychopathology is read as verbal behaviour, and ACT’s acceptance, defusion and values are given RFT interpretations. | Active — read in full 2026-10-03 (NOTE-430), in place of the RFT book, LIT-549; worth reading as a short statement of Relational Frame Theory by two of the book’s editors, and as the paper that shows how RFT is meant to become a theory of psychopathology. Humans, unlike the animals of conditioning research, relate events bidirectionally, so a word acquires the functions of what it names. Relating is a generalized operant learned across exemplars, and its frames share mutual entailment, combinatorial entailment and the transformation of function. Rules are relational networks, and the self is three verbal relations to one’s own behaviour. ACT’s acceptance, defusion and values are then reinterpreted in those terms. Its clinical examples are illustrations, not data, and it says its Part 3 is “no more than a heuristic”. |
| LIT-542 | The theory of constructed emotion: an active inference account of interoception and categorization | Barrett (2017), SCAN 12(1):1–23. The brain runs an internal model of the body in the world for allostasis. Its predictions are “concepts” issued from agranular limbic cortex and refined by prediction error. Interoceptive predictions are felt as affect (valence, arousal). An emotion is an instance of categorizing sensation with an emotion concept, built the same way as any perception. Since instances of an emotion are “created by multiple spatiotemporal patterns” (degeneracy), the categories have no neural essences, and the classical view, basic-emotion and causal appraisal theories included, is rejected. Emotions are “constructions of the world, not reactions to it.” | Active — read in full 2026-10-03 (NOTE-434); worth reading as the shortest complete statement of the theory of constructed emotion and of its link to predictive processing. The brain’s core task is allostasis, run through an internal model whose predictions are “concepts”. An instance of emotion is what happens when the brain categorizes predicted interoceptive and exteroceptive sensation with an emotion concept. On this account emotion categories have no neural “essences”, and causal appraisal theories are “highly doubtful”. The paper is a theoretical synthesis: its own data are a few figures, and the rest is cited evidence and explicitly labelled hypotheses. |
| LIT-545 v3 | Self-Efficacy: Toward a Unifying Theory of Behavioral Change | Bandura (1977), Psychological Review 84(2):191–215. The founding paper of self-efficacy theory: psychological treatments, whatever their form, are proposed to work by changing a person’s expectation that they can perform the needed behaviour. Unread: registered from bibliographic metadata, with no lawful abstract reachable. | Unfetched — registered 2026-10-03 from Crossref and PubMed, not read. The article is closed access at APA. Unpaywall and OpenAlex call it “green” open access, but their one location is a DOAJ record of an unrelated 2020 Jurnal Psikologi article on thalassemia screening that carries Bandura’s DOI by mistake. The copy Stanford hosted with Bandura’s papers (web.stanford.edu/dept/psychology/bandura/pajares/Bandura1977PR.pdf) now returns 403 Forbidden. Its archived snapshots are on the Wayback Machine, which this environment’s egress policy blocks, and the University of Kentucky mirror of the same collection now redirects away. A PDF of the published article is posted by the Sustainable Sanitation Alliance’s library (susana.org), a third party, so it was not read (only its first page was opened, to identify what it was). A human reader with APA access, or one who can open the archived Stanford copy, could promote this. |
| LIT-546 v2 | The Emotions | Frijda (1986), The Emotions, Cambridge University Press, Studies in Emotion and Social Interaction (ISBN 0-521-30155-6 hardback, 0-521-31600-6 paperback), 544 pp. The canonical statement of Frijda’s appraisal and action-readiness theory: emotions are changes in action readiness elicited by appraisal of events as relevant to the individual’s concerns, with control precedence over other activity. Unread here; what is said of it comes from the works that cite it. | Unavailable — registered 2026-10-03 from bibliographic records, not read. No lawful open copy of the book (544 pages) was found: it has no Crossref DOI, and Open Library lists no open or lendable scan. Under the owner’s book rule its reading is replaced by Frijda, Ridderinkhof and Rietveld (2014), LIT-540, read in full, which states the book’s central constructs, action readiness, concern and control precedence, in Frijda’s words. The closer substitute, Frijda (1988) “The laws of emotion” (American Psychologist 43(5):349–358, DOI 10.1037/0003-066X.43.5.349), was not lawfully reachable from here. A human with library access could read it, or the book, and either would let this entry say more. On 2026-10-14 a lawful full text of the book was sought again and none was reached. OpenAlex: no DOI to look up, and its title search refused the request (the shared daily budget was exhausted). Unpaywall: no DOI to query. Crossref: still no record for the book. Cambridge University Press’s product page returned HTTP 403; the book is not on Cambridge Core as an open title. UvA-DARE, Frijda’s university repository, holds a metadata record for it with no file attached. PhilPapers returned HTTP 403. Europe PMC: no hit. Internet Archive: no copy of The Emotions; the only Frijda book there, The Laws of Emotion (2007), is a borrow-only lending scan. A web search found only catalogue and bookseller listings. It stays Unavailable until a copy is supplied. |
| LIT-549 v2 | Relational Frame Theory: A Post-Skinnerian Account of Human Language and Cognition | Hayes, Barnes-Holmes & Roche (eds.), Kluwer Academic/Plenum, 2001. The founding book of Relational Frame Theory: human language and cognition are learned, generalized operant behaviour, the arbitrarily applicable relating of events (mutually, in combination, and with transformation of their functions) under contextual control. Eight chapters state the theory, from derived relations to rules and the self; five apply it to development, education, social processes, psychopathology and religion. Unread: registered from Crossref’s chapter records, with LIT-541 read in its place. | Unavailable — registered 2026-10-03 from Crossref’s book and chapter records, not read; LIT-541, a 21-page statement of the same theory by two of its editors, was read in its place. No lawful full text of the book was found. Springer’s book page returned a “Client Challenge” bot page, Unpaywall and OpenAlex list no open copy, and Maynooth University’s repository (where Barnes-Holmes and Roche deposited many RFT papers) holds no chapter of it. A copy on an Ethiopian national-library server was not used, because it is not an author, institutional or publisher deposit. The editors’ own fourteen-page chapter 8, “Relational Frame Theory: A Précis” (pp. 141–154, DOI 10.1007/0-306-47638-x_8), is the part a human reader should take first; a reading of the whole book would move this to Active. |
| LIT-550 v2 | Emotion and Adaptation | Lazarus (1991), Emotion and Adaptation, Oxford University Press, New York (ISBN 0-19-506994-3). The book-length statement of Lazarus’s cognitive-motivational-relational appraisal theory; its chapter titles frame emotion through the person–environment relationship, motivation, coping and goal congruence. Eleven chapters run from definition and measurement through motivation and coping, cognition, causality, negative, positive and problematic emotions, development, social influence and health, to research, assessment, treatment and prevention. Unread. | Unavailable — registered 2026-10-03 from Crossref’s chapter records, not read, and no substitute was read either. The book (Oxford University Press, about 560 pages) is closed at the publisher. The shorter work by the same author that states the same theory, Lazarus (1991), “Progress on a cognitive-motivational-relational theory of emotion” (American Psychologist 46(8):819–834, DOI 10.1037/0003-066X.46.8.819), has no lawful open copy. APA does not make it free, and Unpaywall and Semantic Scholar list none. The only full text the web search found is in a bulk upload of Lazarus’s journal articles to the Internet Archive by a private user, which is not a lawful source, so I did not open it. Other Lazarus emotion papers (American Psychologist 1991 46(4); Smith & Lazarus 1993; Psychological Inquiry 1995; Journal of Personality 2006) are also closed. His 1993 Annual Review of Psychology chapter is listed as free at the publisher, but annualreviews.org answered with a bot challenge. A human with library access could read the 1991 American Psychologist paper and file it as the read substitute. |
| LIT-553 | Psychological Flexibility as a Fundamental Aspect of Health | Kashdan & Rottenberg (2010), Clinical Psychology Review 30(7):865–878. Psychological health is not a ratio of positive to negative experience or a set of fixed strengths but flexibility: adapting regulation to the situation, balancing time perspectives and life domains, and acting on values while open to experience. Inflexibility marks depression (rumination, emotion context insensitivity) and anxiety (experiential avoidance), and cardiac vagal control is proposed as an integrating marker. Executive functioning, default states and personality underlie it, and dynamic measurement is needed. | Active — read in full 2026-10-03 (NOTE-436), in the NIH author manuscript on PubMed Central; worth reading as the review that took “psychological flexibility” out of the ACT literature and made it a general construct of health. Flexibility is adapting to situational demands, reconfiguring mental resources, shifting perspective and balancing competing desires and life domains. Its absence (rumination, emotion context insensitivity, experiential avoidance, low cardiac vagal control) runs through depression and anxiety. Executive functioning, default mental states and personality are its “building blocks”. It is a narrative synthesis: no meta-analysis of its own, the causal direction between inflexibility and disorder is left open by its own account, and it asks for dynamic measures in place of the global self-reports the ACT line relies on. |
| LIT-554 | The Emerging Field of Emotion Regulation: An Integrative Review | Gross (1998), Review of General Psychology 2(3):271–299. Emotions are response tendencies (behavioural, experiential, physiological) called forth by evaluated situations and open to modulation. Emotion regulation is how people influence which emotions they have, when, and how they are experienced and expressed, and it is distinct from coping, mood regulation and defense. A process model places regulation at five points in emotion generation: situation selection, situation modification, attentional deployment, cognitive change (reappraisal), and response modulation (suppression). The first four are antecedent-focused, the last response-focused. Dysregulation is said to figure in over half of the DSM-IV Axis I disorders and all of Axis II. | Active — read in full 2026-10-03 (NOTE-429); worth reading as the founding statement of the process model of emotion regulation, the five-family scheme (situation selection, situation modification, attentional deployment, cognitive change, response modulation) that later regulation research and many therapies are organised around. It defines emotion regulation as “the processes by which individuals influence which emotions they have, when they have them, and how they experience and express these emotions” (p. 275) and separates it from coping, mood regulation and defense. It is a conceptual review: no data of its own, and it says the scheme is “a set of working distinctions” (p. 282). |
| LIT-555 | Acceptance and Commitment Therapy: Model, Processes and Outcomes | Hayes, Luoma, Bond, Masuda & Lillis (2006), Behaviour Research and Therapy 44(1):1–25. The ACT model: psychopathology is psychological inflexibility, produced when language (cognitive fusion) and experiential avoidance dominate direct contingencies, and ACT restores flexibility through six processes. A review of correlational, component, mediational and outcome studies finds the model “broadly supportive” but the literature immature, and the evidence suggests ACT works through different processes from traditional CBT. Read in an early author manuscript whose numbers differ from the published paper’s. | Active — read in full 2026-10-03 (NOTE-431), but in an early manuscript, not the published text; worth reading as the canonical statement of the ACT model of psychopathology, “psychological inflexibility”, and of the six processes ACT targets. Cognitive fusion and experiential avoidance (from LIT-551) narrow behaviour, and acceptance, defusion, present-moment contact, self as context, values and committed action restore flexibility. The copy read is an author manuscript in Georgia State University’s repository that still carries “xx” placeholders, and its numbers differ from the published paper’s as Kashdan and Rottenberg report them (21 studies against 32; an AAQ correlation printed “.030” against “.42”). Use its model and argument, not its numbers. Its own verdict on the evidence is “broadly supportive” but “not mature”. |
| LIT-563 | HDDM: Hierarchical Bayesian Estimation of the Drift-Diffusion Model in Python | Wiecki, Sofer & Frank (2013), Frontiers in Neuroinformatics 7:14. HDDM is an open-source Python package that estimates the drift diffusion model (and the linear ballistic accumulator) by hierarchical Bayesian MCMC: subject parameters drawn from group distributions with informative priors, full posteriors, outlier handling, and regression of parameters on trial-by-trial covariates. A tutorial reproduces a theta-power effect on decision threshold that reverses under subthalamic deep brain stimulation. In simulations hierarchical Bayes recovers parameters and detects effects better than χ²-quantile and maximum-likelihood fitting, most with few trials. | Active — read in full 2026-10-03 (NOTE-466); worth reading as the methods paper for HDDM, the Python package that fits the drift diffusion model by hierarchical Bayesian estimation (PyMC, MCMC), with subject parameters drawn from group distributions and trial-by-trial regressions of parameters on covariates such as EEG power. Its simulations show hierarchical Bayes recovering parameters with less error than per-subject χ²-quantile, maximum-likelihood or non-hierarchical Bayesian fits, most of all with few trials, and detecting drift-rate differences and covariate effects more often. The comparisons are the authors’ own, on data simulated from the model being fitted, and its hypothesis tests on the group posterior differ from the t-tests used for the other methods. |
| LIT-565 | The Time Course of Perceptual Choice: The Leaky, Competing Accumulator Model | Usher & McClelland (2001), Psychological Review 108(3):550–592. Perceptual choice as gradual, stochastic accumulation in one unit per alternative, with leakage and lateral inhibition between units and a threshold at zero. For two choices the linearised difference is an Ornstein–Uhlenbeck process with net leak K = k − β; K = 0 recovers the classical diffusion model, and K ≠ 0 bounds asymptotic accuracy. The model fits time–accuracy curves slightly better than diffusion with drift variance, fits Vickers’s latency–discriminability data, predicts individual primacy or recency biases confirmed in a new experiment, and yields Hick’s law under a constant-accuracy criterion. | Active — read in full 2026-10-03 (NOTE-449); worth reading as the source of the leaky, competing accumulator (LCA): one leaky accumulator per alternative, lateral inhibition between them, a threshold-linear floor at zero, and an absolute response criterion. In the linear two-choice case the difference between accumulators is an Ornstein–Uhlenbeck process with net leak K = k − β, so the classical diffusion model is the special case K = 0 and accuracy given unlimited time is bounded whenever K ≠ 0. Its evidence for leakage over drift variance is a modest likelihood advantage on 3 participants (combined ratio about 62 to 1, 7 of 9 curves), and the authors concede that a diffusion model with drift variance, starting-point variance and decision boundaries is hard to tell apart from it. |
| LIT-566 | Defining Agency: Individuality, Normativity, Asymmetry, and Spatio-temporality in Action | Barandiaran, Di Paolo & Rohde (2009), Adaptive Behavior 17(5):367–386. Three conditions are jointly necessary and sufficient for minimal agency: the system defines its own individuality, it is the active source of modulations of its coupling with the environment (interactional asymmetry), and it regulates that coupling by norms it generates (normativity). Generatively, an agent is an open autonomous system (a precarious network of mutually enabling processes) that adaptively modulates its coupling so as to maintain some of its constituent processes. Minimal life satisfies this; life is sufficient for agency but not necessary. | Active — read in full 2026-10-03 (NOTE-458), from the authors’ own deposit of the accepted text. Worth reading as the standard organizational definition of minimal agency, and as the clearest statement that an agent is not the same thing as an individual. Individuality is one of three conditions, with interactional asymmetry and normativity; it is a precondition for the other two and is not sufficient for agency. The generative definition: an agent is an autonomous organization that adaptively regulates its coupling with its environment and contributes to sustaining itself as a consequence. The paper rejects statistical and energetic measures of asymmetry as sufficient, and says systems optimising an externally fixed function should not be treated as models of agency. |
| LIT-568 | Précis: Being No One | Metzinger (2005), PSYCHE 11(5), the author’s précis of Being No One. Minimal consciousness is a global world-model in a window of presence that cannot be recognised as a model (globality, presentationality, transparency). A phenomenal self is a transparent self-model: “no such things as selves exist in the world”. Transparency comes in degrees, and the self-model’s opaque partition is what lets humans doubt their own self-representation. The unity of consciousness is a single global functional cluster. A first-person perspective is a phenomenal model of the intentionality relation, and iterating it yields introspective and social intelligence. | Active — read in full 2026-10-03 (NOTE-460), as the read substitute for Being No One (LIT-467), which it summarises in Metzinger’s own words. Worth reading as the most complete short statement of the self-model theory of subjectivity (SMT): more of the machinery than “Self models” (LIT-450), in particular the six constraints, the degrees of transparency, and the phenomenal model of the intentionality relation (PMIR). Phenomenal selfhood is a transparent self-model, “a function realized by a lack of information”; “no such things as selves exist in the world”. The unity of a conscious world is a global functional cluster bound by “highest-order binding”. It states the theory and its predictions; the evidence is cited, not shown, and the neural sections say how little is known. |
| LIT-569 | The Physics of Optimal Decision Making: A Formal Analysis of Models of Performance in Two-Alternative Forced-Choice Tasks | Bogacz, Brown, Moehlis, Holmes & Cohen (2006), Psychological Review 113(4):700–765. The pure drift diffusion model implements the SPRT and Neyman–Pearson test, so it is the optimal two-choice decision process. The O-U, mutual inhibition (Usher–McClelland), feedforward inhibition and pooled inhibition (Wang) models reduce to it for balanced parameters; the race model does not. For reward rate, Bayes risk and two accuracy-weighted criteria there is a unique optimal threshold, giving parameter-free optimal performance curves. With biased priors the optimal starting point is proportional to the log prior odds. | Active — read 2026-10-03 (NOTE-461), main text and Appendix B in full, Appendix A (the derivations) by section only; worth reading as the paper that puts the two-choice decision models in one frame and ties them to the sequential probability ratio test. The pure drift diffusion model is the continuum limit of the SPRT, so it is the fastest for a given accuracy and the most accurate for a given time. Of six models, all but the race model reduce to it in some parameter range: the Usher–McClelland mutual inhibition model when leak equals inhibition and both are large, the feedforward inhibition model when its inhibitory and excitatory weights are equal. Each reward criterion has a unique optimal threshold, which gives testable optimal performance curves. The optimality results hold for the pure model with stationary evidence and two alternatives. |
| LIT-571 | Participatory sense-making: An enactive approach to social cognition | De Jaegher & Di Paolo (2007), Phenomenology and the Cognitive Sciences 6(4):485–507. Coordination between coupled systems is cheap and common. When patterns of coordination and the course of an encounter shape each other, the interaction acquires its own operationally closed, precarious organization. Social interaction is then regulated coupling between autonomous agents that constitutes an autonomous relational organization without destroying the agents’ autonomy. Because sense-making is embodied in movement, interactors can coordinate their sense-making, along a spectrum from modulated individual sense-making to jointly created meaning. | Active — read in full 2026-10-03 (NOTE-456), from the published version posted on De Jaegher’s own site. Worth reading as the founding statement of participatory sense-making. A social interaction is defined as regulated coupling between autonomous agents that itself becomes an autonomous organization in the relational domain, without destroying the agents’ autonomy; coordinating sense-making in interaction can generate meanings neither individual had. The argument is conceptual, illustrated by everyday cases and by the perceptual-crossing experiments; nothing is measured in the paper. |
| LIT-576 | The Diffusion Decision Model: Theory and Data for Two-Choice Decision Tasks | Ratcliff & McKoon (2008), Neural Computation 20(4):873–922. A review of the diffusion decision model for fast two-choice tasks, with three new motion-discrimination experiments. Noisy evidence accumulates from a starting point z to one of two boundaries (0, a) at drift v, plus non-decision time and across-trial variability in v, z and Ter. The model fits accuracy and full correct and error RT distributions, and maps difficulty to drift, speed–accuracy instructions to boundary separation and stimulus proportion mainly to starting point. Applications cover aging, aphasia and links to neural firing. | Active — read in full 2026-10-03 (NOTE-445), from the author manuscript in PubMed Central; worth reading as the canonical statement of the diffusion decision model and of what it is for. Two-choice RT and accuracy are decomposed into drift rate (quality of evidence), boundary separation (caution), starting point (bias) and non-decision time, plus across-trial variability in drift, starting point and non-decision time. Three new motion-discrimination experiments show each manipulation landing on its own parameter: difficulty on drift, speed–accuracy instructions on boundaries, stimulus proportion mainly on starting point. Its closing claim, that diffusion models have come “as near to provid[ing] a solution to simple decision making as is possible in behavioral science”, is the authors’ assessment, not a result. |
| LIT-578 v2 | Dynamic Patterns: The Self-Organization of Brain and Behavior | Kelso (1995), Dynamic Patterns: The Self-Organization of Brain and Behavior, MIT Press, xvii + 334 pp. The book-length statement of coordination dynamics. Unread: registered from catalogue records. What the record knows of its content comes from the works that cite it: the coordination potential of the HKB model (reproduced in Kelso 2021, Box 1), absolute versus relative coordination (De Jaegher & Di Paolo 2007, citing p. 98), a chapter on how HKB arose (ch. 3) and one on brain rhythms (ch. 8), per Kelso 2021. | Unavailable — registered 2026-10-03 as a seed, not read. No lawful full text was found: the book has no DOI, is not listed as open access, and MIT Press’s page refused an automated request. Under the owner’s rule for books it is supplanted by a shorter work by the same author, Kelso’s 2021 retrospective on the HKB model (LIT-601), which was read. It is filed because it is the citation others reach for when they mean coordination dynamics, relative coordination or metastability; De Jaegher & Di Paolo (LIT-571) and Raja et al. (LIT-598) both cite it. |
| LIT-580 | Uncertainty-based competition between prefrontal and dorsolateral striatal systems for behavioral control | Daw, Niv & Dayan (2005), Nature Neuroscience 8(12):1704-1711. Habitual and goal-directed control are model-free (“caching”) and model-based (“tree-search”) reinforcement learners in dorsolateral striatum and prefrontal cortex. Behaviour follows whichever controller’s value estimate is less uncertain under approximate Bayesian versions of each. Simulations reproduce the devaluation data: over-training yields devaluation-insensitive habit, except for actions near reward or in tasks with more actions and outcomes. Both controllers serve the same ends; neither is the “irrational” one. | Active — read in full 2026-10-03 (NOTE-459), supplementary methods not read. Worth reading as the founding statement of model-based versus model-free control as two brain systems, and of arbitration between them by relative uncertainty. Goal-directed (devaluation-sensitive) behaviour is a prefrontal “tree-search” controller, statistically efficient but noisy to compute. Habit is a dorsolateral-striatal “caching” controller, cheap but data-hungry and blind to revaluation. Each controls behaviour when its value estimate is the more certain, which reproduces why over-training produces habits unless the task is complex or the action near the reward. Both controllers pursue “identical rational ends”: a rejection of the impulsive-limbic-versus-rational-prefrontal picture. |
| LIT-581 | Hierarchically organized behavior and its neural foundations: A reinforcement learning perspective | Botvinick, Niv & Barto (2009), Cognition 113(3):262-280. Hierarchical RL groups actions into reusable “options” with their own policies, subgoals and pseudo-reward. That eases the scaling problem by cutting the number of decisions to explore and learn. The paper builds an actor-critic version and maps option identifiers to DLPFC, option-specific policies to dorsolateral striatum, and option-specific values and option-length prediction errors to OFC. Unsuitable options produce negative transfer, so how useful options are discovered is the central open problem. | Active — read in full 2026-10-03 (NOTE-451), from the PubMed Central author manuscript; the online appendix with the equations was not read. Worth reading as the bridge between hierarchical reinforcement learning (temporally abstract “options”) and the psychology and neuroscience of hierarchically organized behaviour. Options ease the scaling problem of RL by shrinking search and learning to a few decision points. An actor-critic version maps option identifiers to dorsolateral PFC, option-specific policies to dorsolateral striatum, and option-specific values and extended prediction errors to orbitofrontal cortex. The problem it puts first is option discovery: the wrong options cause negative transfer, as in human problem solving. |
| LIT-584 | Model-based and model-free Pavlovian reward learning: Revaluation, revision, and revelation | Dayan & Berridge (2014), Cognitive, Affective, & Behavioral Neuroscience 14(2):473-492. Pavlovian prediction is not only model-free. A cue that had predicted disgusting salt became instantly “wanted” when a never-experienced salt appetite was induced, before the salt was retasted, which needs a prediction of the outcome’s identity revalued against the current bodily state. Over-training may “defocus” outcome representations rather than hand control to a model-free system. Dopamine may carry model-based evaluations. No algorithm for the revaluation is given. | Active — read in full 2026-10-03 (NOTE-463), from the PubMed Central author manuscript. Worth reading as the argument that Pavlovian prediction, usually modelled as purely model-free, is also model-based. In the “Dead Sea salt” experiment a lever that had predicted a disgusting salt solution became, on first presentation after a novel sodium appetite was induced and before any salt was retasted, nearly as attractive as a sucrose lever, with mesolimbic activation. A cached value cannot do that. The CS must predict the outcome’s identity, and its value must be recomputed against the current bodily state. The paper adds “defocusing” of outcome representations as a model-based route to apparent habit, and argues that dopamine can carry model-based evaluations. |
| LIT-585 | The Expected Value of Control: An Integrative Theory of Anterior Cingulate Cortex Function | Shenhav, Botvinick & Cohen (2013), Neuron 79(2):217-240. Control is allocated by a cost-benefit decision. The expected value of control (EVC) of a candidate control signal is its probability-weighted payoff, with future value discounted, minus an intrinsic cost rising with its intensity. dACC monitors state and outcome information and specifies the identity and intensity of the EVC-maximizing signal. Lateral PFC and subcortical structures implement it. The account integrates conflict monitoring, effort costs and the override of defaults (exploration, foraging, patience). | Active — read in full 2026-10-03 (NOTE-446), from the PubMed Central author manuscript. Worth reading as the standard statement that cognitive control is allocated by a cost-benefit decision. The “expected value of control” (EVC) of a control signal, which has an identity (which task) and an intensity (how much), is its expected payoff minus an intrinsic cost that rises with intensity. Dorsal anterior cingulate cortex is proposed to monitor and to specify the EVC-maximizing signal, which lateral PFC and subcortical structures then implement. It is a review and synthesis: the model is stated, not fitted, and the authors say the ingredients are not new. |
| LIT-595 | “First we invented stories, then they changed us”: The Evolution of Narrative Identity | McAdams (2019), Evolutionary Studies in Imaginative Culture 3(1):1–18. Narrative identity is “the internalized and evolving story a person invents to explain how he or she has become the person he or she is becoming”, giving a life “unity, moral purpose, and temporal coherence”. Storytelling evolved with human hypersociality. The self develops from social actor (by age 2) to motivated agent (school years) to autobiographical author (adolescence). Agentic and redemptive stories go with well-being and generativity, and cultures supply master narratives. “Biology guides and culture fills in the details.” | Active — read in full 2026-10-03 (NOTE-470), as the read substitute for McAdams & McLean, “Narrative Identity” (2013; LIT-572), which has no lawful copy and whose definition this essay takes over word for word. Worth reading as McAdams’s own short statement of narrative identity: “the internalized and evolving story a person invents to explain how he or she has become the person he or she is becoming”, which gives a life “unity, moral purpose, and temporal coherence”. It places the life story as the third layer of a self that develops from actor to agent to author, and reports that story features (agency, redemption, coherence) predict well-being and generativity. A review essay: the empirical claims are cited, not shown, and the evolutionary story is offered as plausible, not tested. |
| LIT-598 | The Markov blanket trick: On the scope of the free energy principle and active inference | Raja, Valluri, Baggs, Chemero & Anderson (2021), Physics of Life Reviews 39:49–72. Restating the FEP’s derivation (Langevin dynamics, NESS, Helmholtz decomposition, Markov blanket, variational bound), the authors grant that it is true of things so defined. But blankets apply to every “thing”, not everything (the candle flame is excluded); the partition is placed by modelling convenience; relational properties (heading, affordances) and constitutive self-organization (a membrane the cell makes) fall outside it. The blanket’s real role is to supply the partition variational Bayes needs. Active inference presupposes perception and action in its generative models. | Active — read in full 2026-10-03 (NOTE-450), from the authors’ PhilSci-Archive preprint (March 2021); the Physics of Life Reviews version was not seen. Worth reading as the scope critique of the free energy principle: granted its premises the principle is true, but it applies only to “things” already given a Markov-blanket partition, the partition is placed where the model needs it, and relational properties and constitutive self-organization fall outside it. The blanket is a “trick” that lets any thing be modelled as a variational autoencoder. Active inference, as a process theory, presupposes perception and action in the generative model it is handed. |
| LIT-600 v2 | Attention to Action: Willed and Automatic Control of Behavior | Norman & Shallice (1986), in Davidson, Schwartz & Shapiro (eds.), Consciousness and Self-Regulation vol. 4 (Plenum), pp. 1-18; first issued as CHIP Report 99 (UCSD, December 1980). Per the report’s ERIC abstract: routine action runs off by “contention scheduling” among competing schemas, and a “supervisory attentional mechanism” biases that selection for novel, ill-learned, dangerous or planned action; “will becomes the application of attentional resources to the control of action”. Unread. | Unavailable — registered 2026-10-03 from bibliographic records and from the ERIC abstract of the 1980 technical report, not read. The 1986 chapter is not open access: Unpaywall reports no open copy, and SpringerLink returned a JavaScript bot challenge, which was not pursued. The 1980 version (CHIP Report 99, University of California San Diego, catalogued by ERIC as ED205562, “Technical Report No. 8006”, 43 pages) was the brief’s lead. ERIC has a catalogue record and abstract but no full text: its fulltext URL returns 404. The copies on Academia.edu sit behind a bot challenge, and those on ResearchGate and Scribd are uploads this record does not use. No DTIC copy was found. A human could ask UCSD’s library or the authors’ institutions for the CHIP report, or read the chapter through a library, and move this to Active. |
| LIT-601 | The Haken–Kelso–Bunz (HKB) model: from matter to movement to mind | Kelso (2021), Biological Cybernetics 115(4):305–322, a 60th-anniversary retrospective. HKB models bimanual coordination by the relative phase φ, with dφ/dt = −a sin φ − 2b sin 2φ and potential V = −a cos φ − b cos 2φ: both in-phase and anti-phase are stable for b/a > 1/4, and anti-phase loses stability below it, as observed when movement frequency rises. Noise, symmetry breaking (Δω) and metastability extend it; with Kuramoto’s model it becomes a generalized HKB for many agents. The same dynamics is reported within a body, between people and with machines, and in brain activity. | Active — read in full 2026-10-03 (NOTE-467), from the free-to-read publisher PDF. Filed in place of Kelso’s book Dynamic Patterns (LIT-578) and as the read account of the paywalled Haken–Kelso–Bunz paper (LIT-588), both Deferred. Worth reading as Kelso’s own short statement of coordination dynamics: the HKB order-parameter equation for relative phase, its stochastic and symmetry-breaking extensions, metastability as relative coordination, the generalized HKB–Kuramoto model, and the claim that one coordination dynamics holds within and between people. It is a retrospective by an author of the model, not a review, and it says so. |
| LIT-603 v2 | The Emperor’s New Markov Blankets | Bruineberg, Dołęga, Dewhurst & Baltieri (2020 preprint; BBS 45:e183, 2022). Markov blankets entered variational inference as Pearl’s conditional-independence shields, tools on a scientist’s map. In the free-energy literature since Friston 2013 they became boundaries in the world between agent and environment: “Friston blankets”. The two are different constructs; Friston blankets need extra assumptions (a chosen internal node, sensorimotor labels, steady state, sparse coupling) that the formalism does not supply. Using them for metaphysical conclusions (“inference within a model”) requires metaphysical premises; used instrumentally (“inference with a model”) they are innocent but philosophically idle. | Active — read in full 2026-10-03 (NOTE-447), but in its December 2020 PhilSci-Archive version; the Behavioral and Brain Sciences target article of 2021–2022 is, by the authors’ own note, “heavily rewritten” and could not be read lawfully beyond its abstract. Worth reading as the critique that names the problem: “Pearl blankets”, the conditional-independence tool of graphical models, are conflated with “Friston blankets”, a realist construct meant to be the boundary between an agent and its world. The second cannot be read off the mathematics of the first, and its use to settle questions about life and mind needs metaphysical premises of its own. It re-runs the primordial-soup simulation of LIT-526 and shows how the blanket found there depends on modelling choices. |
| LIT-606 | Thine Own Self: True Self-Concept Accessibility and Meaning in Life | Schlegel, Hicks, Arndt & King (2009), JPSP 96(2):473–490. Five studies: faster “me” judgments for one’s own true-self traits (residualized on speed for control traits) predict presence of meaning in life, while actual-self accessibility does not, controlling affect, need satisfaction and (Study 4) AI-3 authenticity, which was itself uncorrelated with true-self accessibility. Subliminal priming of true-self traits raised reported meaning (Studies 3 and 5), for disliked traits as much as liked ones. | Active — read in full 2026-10-03 (NOTE-469); worth reading as the experimental case that the true self-concept, as distinct from self-reported authenticity, matters to felt meaning. In five student studies (N = 56 to 149), how quickly people classify their own “true self” traits predicts presence of meaning in life over and above actual-self accessibility, need satisfaction and AI-3 authenticity, and subliminally priming true-self traits raises reported meaning, even when the traits are disliked. Effects are small, meaning is self-report only, and the Study 5 interaction is printed with an F that cannot carry its stated p. |
| LIT-615 | An Active Inference Approach to Modeling Structure Learning: Concept Learning as an Example Case | Smith, Schwartenbeck, Parr & Friston (2020), Front. Comput. Neurosci. 14:41. A partially observed Markov decision process with eight animal “slots”, some starting with flat likelihoods, learns new concepts without feedback: a slot is engaged only when its flat mapping explains a feature combination better than any learned concept. Bayesian model reduction on the learned prior over states then resets slots the data did not need. In 100 runs per condition the true number of concepts won in 45–80% of runs for 4–7 concepts and rarely for 2–3, and in every run when the likelihood was given. The agent also generalises to an unseen animal in one shot, and the paper simulates neural and dopaminergic responses. | Active — read in full 2026-10-03 (NOTE-485); worth reading as the record’s worked example of structure learning in an active-inference agent: “model expansion” by spare hidden-state slots whose flat likelihoods win inference only for a truly novel stimulus, and “model reduction” by applying Bayesian model reduction to the agent’s learned prior over states, so that unused slots are reset. It is a proof of concept with eight animals and six features. Reduction recovered the true number of concepts in 45–80% of 100 runs for 4–7 concepts, failed for 2–3, and succeeded every time only when the likelihood mapping was given rather than learned. |
| LIT-625 | Functional Role and Truth Conditions | Block (1987), Aristotelian Society Supplementary Volume 61:157–181, the first paper of a symposium with John Campbell. Functional role semantics can explain five facts about concepts (holism of concepts and thoughts, incommensurability, limits on attributing irrationality, structure). To give Twin Earth truth conditions, a private-use theory must add a referential factor; a public-use theory must extend roles into the world with devices from the theory of reference (dominant causal source, borrowed reference, experts). Either way the result is equivalent to a two-factor theory whose internal factor is narrow meaning. Burge’s arthritis cases need, in addition, functional role in a language with its structure of deference. | Active — read in full 2026-10-03 (NOTE-492); worth reading as the work the record cites as “Block 1987” (NOTE-117, after Drayson): a short argument that any use theory of meaning that gets Twin Earth truth conditions right, whether it starts from private use in thought or public use in communication, ends up equivalent to a two-factor theory with a narrow, in-the-head functional role and a referential factor. It leaves room for narrow content without arguing for it, and its appendix concedes that Burge’s arthritis cases need functional role in a language, structured by deference, as well as in the individual. Block says outright that he has not tried to settle whether narrow content is explanatory. |
| LIT-626 | Three types of phenomenal consciousness and their functional roles: unfolding the ALARM theory of consciousness | Newen & Montemayor (2025), Phil. Trans. R. Soc. B 380:20240314. The ALARM theory: phenomenal consciousness has two levels of experience and three functions. Basic arousal, realisable without cortex, alarms the body when a homeostatic variable leaves its tolerance range, overriding slow regulation with no choice involved; it adds lasting care for the injury and one-shot, generalised avoidance learning. General alertness, cortical and attention-guided, enables flexible learning and decision-making. Reflexive self-consciousness is general alertness with metacognitive self-relational content, not a third kind of experience, and enables long-term planning that accounts for one’s own and others’ mindsets. IIT, higher-order theories and GNWT are faulted for tying consciousness to neocortex and leaving basic arousal out. | Active — read in full 2026-10-03 (NOTE-489); worth reading as the shortest statement of the ALARM theory: phenomenal consciousness has two levels of experience, basic arousal (subcortical, an alarm that overrides slow homeostatic regulation) and general alertness (cortical, attention-guided, for flexible learning and choice), plus reflexive self-consciousness, which is general alertness with metacognitive self-relational content and its own function, long-term planning. Its evidence is a short list of cited findings (hydranencephaly, decorticate cats, central lateral thalamus stimulation in macaques), and the functional claims are argued, not tested. |
| LIT-629 | Two neural correlates of consciousness | Block (2005), Trends in Cognitive Sciences 9(2):46–52. There are at least two neural correlates of consciousness: a phenomenal NCC, the minimal basis of an experience’s content (recurrent activation of MT/V5 with V1 for seen motion), and an access NCC, frontoparietal activation that broadcasts the content in a global workspace. Because access is a winner-take-all competition, a recurrent sensory representation can lose and stay unaccessed. Signal-detection analyses of exclusion tasks and of V1 figure–ground modulation in monkeys suggest experience without access. | Active — read in full 2026-10-03 (NOTE-496); worth reading as the short empirical turn of Block’s distinction: a phenomenal NCC (recurrent activity in sensory areas, such as MT/V5 with V1 for seen motion) and an access NCC (frontoparietal broadcasting), with a winner-take-all contest for access as the reason the first can occur without the second. Its evidence is signal-detection reasoning: Supèr et al.’s V1 figure–ground modulation persists when low salience or many catch trials uncouple it from the monkey’s saccade. The case is suggestive, “a glimmer”, and the later papers make it at length. |
| LIT-630 | The Mind as the Software of the Brain | Block (1995), in Smith & Osherson (eds.), An Invitation to Cognitive Science, 2nd ed., vol. 3, Thinking, MIT Press, ch. 11, pp. 377–425. The computer model explains intelligence by decomposing it into ever stupider processors that bottom out in primitive processors, the only devices for which behaviourism is true. The brain is a syntactic engine driving a semantic engine, by an isomorphism between symbol transitions and rational relations among meanings, which must hold counterfactually, so a wall is not every computer. Intentionality comes from the language of thought plus functional role semantics. Stich’s syntactic theory falls to the “Reductionist Cruncher”, and the systems reply answers Searle, though Block doubts the Chinese system is phenomenally conscious. | Active — read in full 2026-10-03 (NOTE-490); worth reading as Block’s textbook statement of what “the mind is the software of the brain” means: intelligence explained by functional decomposition into primitive processors, the brain as a syntactic engine driving a semantic engine, intentionality explained by the language of thought plus functional role semantics, and the defence of that picture against Dennett’s objections, Stich’s syntactic theory (by the “Reductionist Cruncher”) and Searle’s Chinese room. It is a survey for an introductory volume, so its arguments are compressed, but it states clearly two things the record uses: intelligence is future-oriented and intentionality past-oriented, and the computer model may give intentionality without phenomenal consciousness. |
| LIT-631 | Conscious Processing and the Global Neuronal Workspace Hypothesis | Mashour, Roelfsema, Changeux & Dehaene (2020), Neuron 105(5):776–798. A review of the global neuronal workspace hypothesis by its proponents. Conscious access is a sudden, non-linear ignition of a distributed network of long-range pyramidal neurons, with prefrontal cortex a key hub, that amplifies, sustains and broadcasts one representation; unseen stimuli produce only a decaying wave. Evidence: a late (>200–300 ms) divergence between seen and unseen trials across modalities, prefrontal ignition in monkeys tied to the signal-detection threshold, and the disconnection of frontal-parietal networks by every major class of anaesthetic. | Active — read in full 2026-10-03 (NOTE-510); worth reading as the record’s first source statement of the global neuronal workspace (GNW): conscious access is a non-linear “ignition” in which recurrent loops among long-range prefrontal, parietal and other hub neurons amplify, sustain and broadcast one representation to the brain’s specialised processors. Its strongest evidence is van Vugt et al.’s (2018) monkey recordings, where seen and unseen stimuli both drive V1 and V4 but only seen ones (and false alarms) ignite sustained prefrontal firing, and the signal-detection threshold maps onto the ignition threshold. It is a review by the theory’s proponents, explicitly a theory of access, and it concedes that the four leading theories cannot yet be told apart experimentally. |
| LIT-632 | Comparing the Major Theories of Consciousness | Block (2009), in Gazzaniga (ed.), The Cognitive Neurosciences, 4th ed., MIT Press, ch. 77, pp. 1111–1122. Compares the three theories neuroscientists take most seriously: higher-order thought, the global workspace and the biological theory. Only the biological theory fully acknowledges the explanatory gap; the others replace it with a puzzle their own theory creates. Concept dualism stops the gap becoming dualism. The ambitious higher-order theory also fails on stipulation, mismatch and infants. Integrated information is criticised as a theory of intelligence. The biological theory is the least friendly to machine consciousness. | Active — read in full 2026-10-03 (NOTE-505); worth reading as Block’s comparison of the higher-order thought, global workspace and biological theories by one test: does the theory recognise the explanatory gap? The ambitious higher-order theory and the workspace theory turn the gap into a puzzle of their own making, so only the biological theory “fully acknowledges” it; concept dualism keeps that from becoming dualism. Also the source of the “biological theory” that Chalmers’s meta-problem paper cites as “Block 2009”, and a short argument that the biological theory is less friendly to machine consciousness than its rivals. |
| LIT-633 | Cross-region neuron co-firing mediated by ripple oscillations supports distributed working memory representations | Verzhbinsky, Daume, Cheng, Rutishauser & Halgren (2026), Nature Neuroscience 29:2566–2577. In 1,373 single units from hippocampus, amygdala, vmPFC, ACC and preSMA of 35 epilepsy patients doing a Sternberg task, ripples (70–100 Hz) co-occur between bundles up to about 220 mm apart with no decline beyond the local bundle, and cross-region unit co-firing rises about 30% during co-ripples. That rise exceeds a rate-only null (STTC 117% higher), grows with memory load, and co-ripples hold more repetitions of a stimulus’s encoding co-firing at retrieval, more on fast trials. | Active — read in full 2026-10-03 (NOTE-494); worth reading as single-unit evidence that brief co-occurring ~90 Hz ripples in distant human forebrain sites (up to about 220 mm, across hemispheres) mark windows of raised cross-region co-firing, about 30% above no-ripple periods and beyond what firing rates alone predict, that this scales with working-memory load, and that stimulus-specific co-firing from encoding recurs more often inside co-ripples at retrieval, most of all on fast trials. One reanalysed open dataset (35 patients, one Sternberg task, load 1 versus 3), clinically placed electrodes, and correlational throughout: nothing perturbs the ripples. |
| LIT-634 | Inverted Earth | Block (1990), Philosophical Perspectives 4:53–79. Inverted Earth has complementary colours and inverted colour words, so a visitor with colour-inverting lenses notices nothing. After long residence, or for a twin raised there with a molecule-for-molecule identical brain, the intentional content of the experience of the sky inverts (it is about yellow) and so does its functional role in a long-arm description, while qualitative content stays the same. So qualitative content is not a functional state and the intentional/qualitative distinction stands. Public colour terms like “looks red” express intentional, not qualitative, content. A brain-internal functional theory of qualia would escape, but nobody has said what it is. | Active — read in full 2026-10-03 (NOTE-504); worth reading as the source of the Inverted Earth argument: a subject moved, unaware, to a planet of complementary colours and inverted colour words keeps the same qualitative content while, over time, the intentional content and the (long-arm) functional role of the experience invert. That is the converse of the inverted spectrum, and it needs no internal switch, so the subject’s testimony is undisturbed and brain supervenience is on the qualia realist’s side. It refutes functionalism about qualia only in its long-arm form; Block concedes a short-arm functional theory of qualia would sink it, and answers that only with a burden of proof. |
| LIT-635 | Mental Paint | Block (2003), in Hahn & Ramberg (eds.), Reflections and Replies, MIT Press, pp. 165–200. Phenomenal character outruns representational content. Internalist representationism must find a functional difference for every phenomenal one, and Erisa, raised where colours keep changing, has vivid colour differences with no abiding functional ones. Against externalist representationism: a swampman has phenomenal character and no content; attention is not awareness, so transparency fails as a claim about awareness; on Inverted Earth what stays the same is mental paint; and normal variation in colour vision means, if representationism is true, that a standard chip cannot look its shade to both men and women. | Active — read in full 2026-10-03 (NOTE-511); worth reading as Block’s fullest case against representationism about phenomenal character (“phenomenism” against “representationism”). It separates mental paint (phenomenal properties that represent) from mental oil (phenomenal properties that do not), argues against internalist representationism with Erisa, raised among colours that keep changing, and against externalist representationism with Inverted Earth, long-term memory, and an empirical “shifted spectra” argument from normal variation in cone sensitivity and in where people place unique green. Several of its arguments are offered as pressure, not proof, and Block says so. |
| LIT-636 v4 | The Border Between Seeing and Thinking | Block (2023), Oxford University Press, open access. Argues, by Block’s own précis (LIT-649), that perception is constitutively iconic, nonconceptual and nonpropositional while cognition is paradigmatically discursive, conceptual and propositional, using scientific marks of perception such as adaptation; that core cognition splits between the two; and, in a chapter on consciousness, against cognitive theories of consciousness. Unread: seeded from the chapter list and the précis. | Unfetched — seeded 2026-10-03 from Crossref’s book and chapter records and from Block’s own précis (LIT-649), not read. The book is open access at Oxford University Press under CC BY-NC-ND 4.0, so a lawful full text exists, and Block’s own copy is listed on PhilArchive as BLOTBB. Neither could be fetched. On 2026-10-03 academic.oup.com answered every request with a Cloudflare “Just a moment…” challenge, and OAPEN and DOAB, which normally mirror such books, refused this session’s requests. On 2026-10-30 the DOI resolved to academic.oup.com/book/45659, which answered HTTP 403 with a Cloudflare challenge (cf-mitigated: challenge), and philarchive.org/rec/BLOTBB did the same. Neither challenge was retried. The précis was read in its place. Filed because the record cites the book twice (NOTE-104 for pp. 454–458 on the neural replacement argument, NOTE-150 for the perception–cognition border), and those citations need a code. A human reader can open either copy, and a reading would move this to Active. Free copies: doi:10.1093/oso/9780197622223.001.0001 (OUP’s open-access book, academic.oup.com/book/45659); philarchive.org/rec/BLOTBB. |
| LIT-637 | What Is Wrong with the No-Report Paradigm and How to Fix It | Block (2019), Trends in Cognitive Sciences 23(12):1003–1013. Whether perceptual consciousness needs prefrontal circuits cannot be settled by no-report paradigms, because subjects not asked to report may still think about what they see, and that thought tracks the percept. A no-post-perceptual-cognition paradigm is needed. Brascamp et al.’s rivalry with inconspicuous switches is one, and it found no detectable prefrontal difference. Box 2 argues that binocular rivalry is driven by local features, against the predictive-processing account. | Active — read in full 2026-10-03 (NOTE-495); worth reading as the argument that removing report does not isolate the neural basis of conscious perception, because subjects, monkeys included, can still think about what they see (“the bored monkey problem”). What is needed is a paradigm with no post-perceptual cognition that differs between percepts. Brascamp et al.’s inconspicuous binocular rivalry, whose switches subjects could not tell from constant random changes, is offered as one, and found no prefrontal difference by fMRI. Block flags that fMRI may be too coarse, and that ECoG could still favour prefrontalism. |
| LIT-638 | Interacting spiral wave patterns underlie complex brain dynamics and are related to cognitive processing | Xu, Long, Feng & Gong (2023), Nature Human Behaviour 7:1196–1215. In Human Connectome Project fMRI (100 subjects per cohort), the phase field of slow cortical fluctuations holds about 16–19 rotating spirals at a time, centred on phase singularities that drift superdiffusively and sit on boundaries between functional regions. Opposite spirals annihilate, like ones repel. Spiral locations and rotation directions decode four language-task conditions at about 48% (chance 25%, matched raw-signal decoder 30%), and spiral arrays route task activity flow bottom-up while listening and top-down while answering. | Active — read in full 2026-10-03 (NOTE-502); worth reading as the claim that slow (0.01–0.1 Hz) human cortical fMRI fluctuations self-organise into many mobile, interacting spiral waves around phase singularities, as in excitable media and turbulence, and that their locations and rotation directions are task-specific and steer large-scale activity flow. The detection is careful about filtering artefacts (phase-randomised null, filtered–unfiltered similarity); the functional claims are correlational, the spatial scale was chosen to maximise task decoding, and the “gate” and “wall” rules are read off streamlines and a phase-oscillator lattice, not tested by perturbation. |
| LIT-640 | Perceptual consciousness overflows cognitive access | Block (2011), Trends in Cognitive Sciences 15(12):567–575. A defence of the overflow argument: perceptual consciousness has a higher capacity than cognitive access. Critics who call it an illusion must posit detailed unconscious iconic memory and conscious generic or fragmentary representations, for which there is no independent evidence. Fragile visual short-term memory, a cortical store of about 5–7 of 8 items over 1–4 s that attention and dlPFC TMS spare while they impair working memory, supports rich perception. Overflow means most items are not accessed, not that any is inaccessible. | Active — read in full 2026-10-03 (NOTE-500); worth reading as Block’s reply to the anti-overflow critics of 2009–2011. Sparse views must posit an unconscious iconic memory detailed enough to do the Sperling and Amsterdam tasks plus conscious “generic” letters or rectangles, and Block argues there is no independent evidence for either. It adds fragile visual short-term memory (about 6–7 of 8 items at 1–2.5 s, cortical, in V4, spared by right dlPFC TMS that impairs working memory) and an imagery task as support. Read from the in-press version. |
| LIT-641 | What Psychological States are Not | Block & Fodor (1972), The Philosophical Review 81(2):159–181. Behaviourism and type physicalism set type-identity conditions on psychological states that are too concrete, and the decisive arguments against them are empirical. The functional state identity theory (FSIT), on which each psychological state is a machine-table state of the organism’s unique best description as a probabilistic automaton, fails too. It cannot model dispositions that are never occurrent, simultaneous states, or a productive and structured set of states, and it individuates too finely. Inverted or absent qualia are a prima facie objection to any functional definition. Psychological states look like computational states, not machine-table states, and type identity is fixed by the laws the states fall under. | Active — read in full 2026-10-03 (NOTE-491); worth reading as the first sustained case against machine-state functionalism, and as the place where the inverted- and absent-qualia objection is first set down as an objection to functional definitions of mental states. Its central result is narrower than “functionalism is false”: psychological states cannot be put one-to-one with the machine-table states of a probabilistic automaton (§III, arguments 1, 2, 4, 5, 6), though they may correspond to computational states. Only the qualia argument (3) threatens functionalism in the broad sense, and the authors treat it as prima facie. |
| LIT-642 | Psychologism and Behaviorism | Block (1981), The Philosophical Review 90(1):5–43. Psychologism, the doctrine that whether behaviour is intelligent depends on the information processing that produces it, is true. A machine that stores all sensible conversations of a test’s length and looks up its replies has the capacity to respond sensibly to any sequence of inputs, so it satisfies the strongest behaviourist conception of intelligence, yet all its intelligence is its programmers’. The standard objections to behaviourism (Chisholm–Geach, the perfect actor, paralytics) defeat behaviourist necessary conditions but not sufficient ones; this machine defeats the sufficient condition. Psychologism is not chauvinism: it excludes one kind of processing without requiring ours. | Active — read in full 2026-10-03 (NOTE-506); worth reading as the source of the “Blockhead”: a machine that stores every sensible hour-long conversation and so has the capacity to pass any Turing test of that length, yet “has the intelligence of a toaster”. The argument refutes behaviourist sufficient conditions for intelligence, including a capacity version that survives the standard objections, and establishes psychologism: whether behaviour is intelligent depends partly on the processing that produces it. It does not say which processing is required, only that it needs some “richness”, and the closing case against simulators of physics, neurons or psychology is admitted to be a sketch. |
| LIT-643 | On a confusion about a function of consciousness | Block (1995), Behavioral and Brain Sciences 18(2):227–247, with open peer commentary. “Consciousness” is a mongrel concept. Phenomenal consciousness is experience; a state is access-conscious if its content is poised for use in reasoning and in rational control of action and speech. Arguments from blindsight and petit mal automatism to a function of phenomenal consciousness are fallacious, because access is missing in those cases too, and its machinery has the function in question. Cases of phenomenality without access (Sperling’s arrays, the unnoticed drill) are offered as worth investigating, not as proven. | Active — read in full 2026-10-03 (NOTE-493); worth reading as the source of the distinction between phenomenal consciousness (P: experience, what it is like) and access consciousness (A: a content poised for use as a premise in reasoning and for rational control of action and speech). The paper’s target is an argument form, not a theory: from a case where “consciousness” and some capacity are missing together (blindsight, petit mal automatism) to a function of P. In those cases A is missing too, so the obvious function of A’s machinery is being credited to P. The cases of P without A that it offers (the noticed drill, Sperling’s arrays, anaesthesia) are put forward as suggestive, not as evidence. |
| LIT-644 | Distinct ventral stream and prefrontal cortex representational dynamics during sustained conscious visual perception | Vishne, Gerber, Knight & Deouell (2023), Cell Reports 42:112752. Ten patients with subdural electrodes watched images for 300–1,500 ms with no report on analysed trials. High-frequency activity dropped about 80% after onset in every region, but category and exemplar information in ventral temporal cortex stayed decodable and time-invariant for as long as the image stayed on. Prefrontal and parietal information was a transient burst at about 150–600 ms that did not track duration. Both IIT’s and the global workspace’s predictions for this paradigm held. | Active — read in full 2026-10-03 (NOTE-499); worth reading as the cleanest test in the record of what neural activity does while a seen image simply stays on. In 10 epilepsy patients with subdural electrodes, high-frequency activity falls about 80% after onset everywhere, yet category and exemplar information in ventral temporal (and less robustly occipital) cortex stays decodable, time-invariant and as long as the image, while prefrontal and parietal information is a transient burst at onset that ignores duration, with no report required. It confirms the multi-duration predictions both IIT and the global neuronal workspace made for this paradigm. Sparse frontal coverage and above-threshold stimuli only. |
| LIT-646 | Consciousness, accessibility, and the mesh between psychology and neuroscience | Block (2007), Behavioral and Brain Sciences 30(5–6):481–499, with open peer commentary. To find the neural basis of experience we must decide whether it includes the machinery of cognitive access, and reports seem unable to decide that. Inference to the best explanation can. Sperling, Landman et al. and Sligte et al. show phenomenology overflowing a working memory of about four items. If the workspace is not part of the basis of phenomenology, strong losing coalitions in sensory cortex realise that overflow; the mesh between psychology and neuroscience favours that assumption. | Active — read in full 2026-10-03 (NOTE-497); worth reading as the overflow argument and the paper that framed the methodological problem the no-report paradigm was built to solve. Phenomenology overflows cognitive access: in Sperling, Landman et al. and Sligte et al. subjects see 8–32 items and can access about 4. If the neural basis of phenomenology excludes the frontal workspace, strong but losing recurrent coalitions in the back of the head explain that overflow, and a better mesh between psychology and neuroscience is reason to accept the assumption. The target article was read; the 32 commentaries and the response were not. |
| LIT-647 v2 | Neural traveling waves in cortex: Network mechanisms and potential roles in neural computation | Muller, Busch, Davis & Reynolds (2026), Neuron 114:3156–3174. A review of neural traveling waves within single cortical areas, chiefly visual cortex: how circuits generate them, and how, by embedding sensory history in the evolving activity across a sensory map, they may implement spatiotemporal computations such as predicting upcoming input. It ends with a framework for generative processing by waves over sensory maps. Unread: registered from the abstract. | Unavailable — registered 2026-10-03 from the PubMed abstract, keywords and the Crossref record, not read. No lawful full text could be found: the article is not open access at Cell Press (Crossref gives the accepted manuscript an Elsevier open-access licence only from 2 September 2027), it is not in PubMed Central, Europe PMC lists no preprint, arXiv has none under Muller with this title, Unpaywall and OpenAlex list no repository copy, and Semantic Scholar marks it closed. The cell.com page returned a bot challenge, which was not pursued. A human reader with subscription access, or anyone after the accepted manuscript becomes open in September 2027, could read it and move this to Active. |
| LIT-648 | Advertisement for a Semantics for Psychology | Block (1986), Midwest Studies in Philosophy 10:615–678. A semantics for cognitive science should explain, among other things, narrow and wide meaning, what makes expressions meaningful, the relativity of meaning to representational system, learning, and how inherited meaning (of words on a page) relates to autonomous meaning (of the language of thought). Two-factor conceptual role semantics is the only approach that promises all of them: an expression’s narrow meaning is its causal role in reasoning and deliberation, short-armed at the skin, and a referential factor fixes truth conditions. Conceptual role determines the nature of the referential factor but not reference. Harman’s one-factor long-arm roles collapse into the same theory; Gricean and indicator semantics are either not naturalistic or fail. | Active — read in full 2026-10-03 (NOTE-503); worth reading as the main statement of two-factor conceptual role semantics: narrow meaning is an expression’s conceptual role inside the head (“total causal role, abstractly described”), and a separate referential factor relates it to the world. Block argues it is the only approach that can meet eight desiderata for a semantics serving cognitive science, among them narrow content, learning, relativity to representational system, and the relation of autonomous to inherited meaning. It is an advertisement, as titled: no identity conditions for conceptual roles are given, and Block says he is “far from a true believer”. |
| LIT-649 | Précis of The Border between Seeing and Thinking | Block (2024), Philosophy and Phenomenological Research 110(1):273–283, the précis for a symposium on his 2023 book. The book’s main claim is that perception is constitutively iconic, nonconceptual and nonpropositional, while cognition is paradigmatically discursive, conceptual and propositional and can use perceptual material inside a “cognitive envelope”. The global neuronal workspace is reread as the mechanism of that envelope, the conceptualisation of percepts. Nine kinds of evidence show working-memory representations differ from perceptual ones and are task-dependent abstractions, which increases Block’s doubt that “raw” percepts are used in cognition. Iconicity is defined by tracking and mirroring conditions after Shepard. | Active — read in full 2026-10-03 (NOTE-498); worth reading as Block’s own short statement, stand-in for the book (LIT-636), of the thesis of The Border Between Seeing and Thinking: perception is constitutively iconic, nonconceptual and nonpropositional, cognition paradigmatically discursive, conceptual and propositional. Despite its title it is “an update”, and most of it develops one idea, the “cognitive envelope”: the global workspace is better read as a theory of how percepts are conceptualised than of consciousness, and recent working-memory evidence shows percepts are recoded, not held raw, when used in reasoning. It is too short to carry the book’s arguments for the main thesis, which it states but does not argue. |
| LIT-650 | Conscious birds | Maldarelli & Güntürkün (2025), Phil. Trans. R. Soc. B 380:20240308. A review arguing that birds show sensory awareness (bistable perception in pigeons; crow nidopallium caudolaterale neurons that follow the reported percept, not the stimulus) and graded self-awareness (magpie and house crow mark tests; pigeons and roosters that treat their mirror image as not-a-conspecific while failing the mark test). Checked against three theories, the avian brain meets the anatomical and physiological requirements of the global neuronal workspace (a hub-and-module connectome with NCL as hub) and of recurrent processing, while IIT’s have not been tested (no PCI, no grid-like lattice found). The authors say the evidence does not yet allow a strong conclusion. | Active — read in full 2026-10-03 (NOTE-488); worth reading as a short, current statement of the case for consciousness in birds and of what each of three theories (global neuronal workspace, recurrent processing, IIT) would need to find in an avian brain. Its strongest evidence is second-hand: crow nidopallium caudolaterale neurons that track the reported percept rather than the stimulus (Nieder et al. 2020). Its sharpest argument is that the bird’s prefrontal analogue is nuclear, not laminar, so conscious report does not need cortical layering. The authors say the findings “do not allow for a strong conclusion”. |
| LIT-699 v3 | Emotional Aperture and Strategic Change: The Accurate Recognition of Collective Emotions | Sanchez-Burks & Huy (2009), Organization Science 20(1):22–34. Introduces emotional aperture, the ability to recognise the composition of diverse emotions in a collective such as a group or business unit, and argues that a leader’s ability to respond to the patterns of shared emotion that arise in strategic change depends on it. Describes cultural, psychological and contextual enablers and impediments, and extends emotional competencies beyond the individual level. Unread: from the abstract. | Unfetched — registered 2026-10-04 from bibliographic records and the publisher’s abstract, not read. No lawful full text was reachable. Unpaywall, OpenAlex and Semantic Scholar list no open copy. The 2006 working-paper version (Ross School of Business Paper No. 1047, “Emotional Aperture and Strategic Renewal”, SSRN 931492) is on SSRN, which returned 403 to this session, and a Deep Blue item the search found, “Emotional Aperture and Emotional Sequencing” (handle 2027.42/48737), sits behind a bot challenge, which I did not try to pass. ProQuest’s open preview carries the citation only. A human with library access, or anyone opening the SSRN working paper in a browser, could read it, and that reading would move this to Active. Free copy: ssrn.com/abstract=931492 (the 2006 working paper). |
| LIT-702 | Genuinely collective emotions | Huebner (2010), European Journal for Philosophy of Science 1(1):89–118. Against the received view that only individuals have emotions, it argues that some groups have the computational complexity and informational integration that emotion requires. Emotions are integrated, action- oriented representations built from subpersonal routines; a distributed system can integrate routines the same way. It answers the objection from consciousness (some individual emotions are non-conscious, so collective ones need not be felt) and from homology (emotion kinds are functional, “like vodka”, not “like bourbon”). A navy crew turns out to detect danger without fear; a failing election campaign is offered instead. | Active — read in full 2026-10-04 (NOTE-553); worth reading as the distributed, computational case for collective emotion, and the clearest statement of what it would take. An emotion is an integrated representation with indicative content and imperative force, built from subpersonal routines; a group whose routines are integrated the same way can be in one, without feeling it. The test case fails on the author’s own terms (the USS Palau crew detects danger without fear) and the replacement (a 2008 presidential campaign) is offered as intuitively plausible pending anthropology, so the paper establishes a possibility and a research programme, not an instance. |
| LIT-704 v2 | The laws of emotion | Frijda (1988), American Psychologist 43(5):349–358. Frijda’s short statement, in ten pages, that emotions obey laws, and the article the record names as the efficient way into his theory. Unread: no lawful copy was reachable. What the record says of it comes from the read works that cite it, which use it for mood as information and for the flexible modulation of emotion. | Unavailable — registered 2026-10-04 from bibliographic records, not read; a second attempt at the copy LIT-540 could not reach. APA does not make the article open, and Unpaywall, OpenAlex and PubMed list no open copy and carry no abstract. The one copy a web search finds, a course upload on a University of Crete student server, now fails TLS verification with an expired certificate; WebFetch got a 503 from it. I did not disable verification. Every other hit was Scribd, academia.edu, ResearchGate or a commercial book site, none of them usable under the record’s rules. A human with library access could read its ten pages, and that reading would move this to Active and LIT-546 with it. |
| LIT-708 | Rarely categorical, highly separable representations along the cortical hierarchy | Posani, Wang, Muscinelli, Paninski & Fusi (2026), Nature. In the IBL Brainwide Map (about 14,000 cortical units, 43 regions, mice doing one biased-block visual decision task), neurons’ selectivity to eight task variables beats a covariance-matched Gaussian null for clustering only in VISp, AUDp and SSp-ul (4,617 selective neurons; regions with ≥ 50). Pooled modules and the whole cortex do cluster, with clusters tracking anatomy. Once conditions an area cannot tell apart are merged (5 to 16 of 16 remain), ≥ 95% of random balanced dichotomies are linearly decodable above a shuffle null in 15 of 16 analysed regions; GU is 0.82. | Active — read in full 2026-10-04 (NOTE-549); the record’s first brain-wide test of whether neurons within a cortical area come in discrete functional types or form a continuum of mixed selectivity. In mouse cortex during one overtrained decision task, a covariance-matched Gaussian null is beaten within single areas only in a few primary sensory ones, while the task conditions each area tells apart are linearly separable in nearly every balanced split. Neuron types appear when areas are pooled, and track anatomy. Second reading under ADR-013; the anthology holds it as ANTH-LIT-696, read for concept geometry in ML. |
| LIT-736 | Shaping the Institutional Mind | Tollefsen (2025), Philosophies 10(5):112. Advocates and critics of group minds share the “standard view” that ascribing attitudes tracks pre-existing causes. On McGeer’s regulative view, ascription shapes its target. Ascribing knowledge to institutions, as in the collective knowledge doctrine, and issuing institutional narratives shape institutions to be more agent-like. The view stays agnostic about whether institutions really have minds. | Active — read 2026-10-05 (NOTE-570): in full, from the publisher’s open-access PDF (CC BY 4.0). Worth reading for a different answer to why we ascribe beliefs and knowledge to institutions: not because the ascriptions describe pre-existing states, but because, on the regulative view of folk psychology, they shape institutions to be more agent-like. Its main case is the collective knowledge doctrine in corporate criminal law, after which banks built systems to detect reporting violations. Limits: the regulative view is assumed, not argued; the causal story about banks is asserted without evidence; and the paper is agnostic about whether institutions have minds. It does not stand in for her book Groups as Agents (LIT-483), which stays unread. |
| LIT-766 v2 | Le Ton beau de Marot: In Praise of the Music of Language | Hofstadter (1997), Basic Books. A book-length meditation on translation, built around dozens of translations of Clément Marot’s short poem “A une Damoyselle malade”: translating is analogy-making under formal and semantic constraints, and no translation is neutral. Not read: no lawful full text was found. | Unavailable — filed 2026-10-09 from catalogue records, not read: no lawful full text was found. The Internet Archive scan is lending-only (access-restricted, print-disabled and borrowable copies), HathiTrust has no record of it, and no author copy was found. It stays Unavailable until a copy is supplied, not on merit. No NOTE is filed. |
| LIT-768 v3 | Remembering: A Study in Experimental and Social Psychology | Bartlett (1932), Cambridge University Press. Remembering is reconstruction from schemata, not retrieval of traces, and it is in important ways social: retellings of an unfamiliar folk tale drift toward the teller’s own culture. Filed, not read: no lawful full text was reachable. | Unavailable — filed 2026-10-09 from Crossref and Open Library, not read: no lawful full text was reachable. Every Internet Archive copy is lending-only (access-restricted), the Wellcome Library scan of the 1932 edition included; HathiTrust holds three copies of the 1932 edition, all Limited (search-only); Cambridge Core sells the 1995 reissue chapter by chapter. A user upload to Internet Archive’s opensource collection was not used, since nothing shows it is lawful. No NOTE is filed. It stays Unavailable until a copy is supplied, not on merit. |
| LIT-769 | Language Evolution by Iterated Learning With Bayesian Agents | Griffiths & Kalish (2007), Cognitive Science 31(3):441–480. When Bayesian learners sample a language from their posterior, iterated learning is a Gibbs sampler and converges to the learners’ prior; when they take the posterior mode, it behaves like a variant of EM and depends also on how much data passes between generations; in a two-language case the MAP outcome is the same for a wide range of prior strengths. Read in NOTE-599. | Active — read 2026-10-09 (NOTE-599); worth reading as the analytic result behind iterated learning: a chain of learners who share a prior and each sample a hypothesis from their posterior is a Gibbs sampler, so the probability that a learner holds a language converges to its prior probability whatever the amount of data passed on, and the data produced converge to the prior predictive. Learners who take the MAP hypothesis run stochastic EM instead: the outcome centres on the prior’s mode but depends on data amount and noise, and in the two-language case is the same for every prior strength α between 0.5 and 1 − ε, so weak biases can be amplified. Equal-fitness populations share the equilibrium. Source of THEORY-155. |
| LIT-771 | Cumulative cultural evolution in the laboratory: An experimental approach to the origins of structure in human language | Kirby, Cornish & Smith (2008), PNAS 105(31):10681–10686. Artificial languages passed down chains of human learners become easier to learn and more structured, without any learner intending it: cultural transmission produces design without a designer. Without a pressure against ambiguity the languages collapse into a few words for classes of meanings; with homonyms filtered from training they become compositional. Skimmed in NOTE-596. | Active — skimmed 2026-10-09 (NOTE-596): every section of the main text was covered, but through a fetch tool’s section-by-section account of the PubMed Central page, not a verbatim copy, and the SI was not reached; worth reading as the laboratory demonstration of iterated learning: four chains of ten adults, each learning a 27-item artificial language from 14 items and passing on its test output, made the languages steadily easier to learn and more structured while believing they were copying. Unfiltered, the languages collapsed to 2–5 words that systematically underspecified meaning (colour lost in every chain); with homonyms filtered out of training, they became compositional. Consistent with THEORY-155, not a test of it: no prior was measured, and the filter is a selection the theory excludes. |
| LIT-777 | Context–content systems of random variables: The Contextuality-by-Default theory | Dzhafarov & Kujala (2016), J. Math. Psychol. 74:11–33. Contextuality-by-Default for finite systems of categorical random variables, indexed by content and context. A system is noncontextual when some coupling of its context bunches makes every connection a maximal coupling; the test is a linear program, the binary cyclic criterion is stated (proved elsewhere), and the minimal total variation of a signed maximally connected quasi-coupling, minus 1, measures the degree of contextuality. | Active — read 2026-10-09 (NOTE-600); worth reading as the reference statement of Contextuality-by-Default: each measurement gets a separate random variable in each context, and a system is noncontextual when some coupling of its contexts makes every content’s copies coincide as often as their own distributions allow (maximal couplings). This separates context-dependent marginals (“direct influences”) from contextuality, turns the test into linear-programming feasibility (Theorem 4.1), and, since a signed such coupling always exists (Theorem 6.1), measures contextuality as the least total variation above 1. The definition was already being replaced by “multimaximal” couplings when it appeared (footnote 10). |
| LIT-790 | Contextuality in Canonical Systems of Random Variables | Dzhafarov, Cervantes & Kujala (2017), Phil. Trans. R. Soc. A 375:20160389. Proposes that contextuality be judged only on a system’s canonical representation: every variable, and every joining or coarsening of interest, replaced by its binary splits, with connections coupled multimaximally (pairwise maximally). For two content-sharing k-valued variables with all splits kept, the canonical system is noncontextual if and only if one variable nominally dominates the other (Pr below the other’s for at most one value; Theorem 4.6). For continuous densities with all splits, any difference in distribution is contextual. | Active — read 2026-10-09 (NOTE-640); worth reading as the step that makes Contextuality-by-Default representation-dependent. Every variable is replaced by binary splits (dichotomizations) and connections are coupled multimaximally, so a verdict belongs to a chosen set of splits, not to the measurements alone. Its one worked result is striking: if every split of two content-sharing k-valued variables is kept, that pair alone is noncontextual exactly when one variable “nominally dominates” the other, i.e. its probabilities fall below the other’s for at most one value (Theorem 4.6). For continuous variables the analogue makes any difference in distribution count as contextuality. The authors present this as opening behavioural contextuality, not as a reductio. Proofs are in the supplement and contain small slips that do not affect the results. |
| LIT-795 | Context effects produced by question orders reveal quantum nature of human judgments | Wang, Solloway, Shiffrin & Busemeyer (2014), PNAS 111(26):9431–9436. Tests the QQ equality of LIT-834, that the probability of giving the same answer to two questions does not depend on their order, on 70 national surveys and two laboratory experiments. Across all 66 Pew surveys from 2001–2011 that varied the order of two questions, order effects are significant (p = 0.0004) and the q values are not (p = 0.4625); paired context effects across 72 studies correlate at r = −0.82 along the predicted line of slope −1. | Active — read 2026-10-09 (NOTE-604); worth reading as the large-scale test of the QQ equality: across 66 Pew surveys chosen without regard to their results, question order changes the answers significantly (p = 0.0004 for the distribution of order-effect χ² values), while the probability of answering the two questions alike stays the same (p = 0.4625 for the q values). Read with the caveats that the supporting information, which holds the proof, the per-study table and the χ² tests, was not reachable and was not read, and that the paper’s title claims more than its evidence: the equality is a symmetry constraint that the authors concede a classical model could be built to satisfy, and LIT-264 shows that data satisfying it are noncontextual. |
| LIT-802 v2 | Le structuralisme | Piaget (1968), Presses Universitaires de France, Que sais-je? no. 1311; English as Structuralism (1970), Basic Books. By its catalogue description, an attempt at a synthesis of the senses structuralism has taken in the contemporary sciences, finding a common ideal of intelligibility beneath variable critical intentions. Not read: no lawful full text was reachable. | Unavailable — filed 2026-10-09 from catalogue records, not read: no lawful full text was reachable. The Internet Archive holds the 1968 PUF edition and the 1970–1973 English editions only as access-restricted lending items, and HathiTrust holds the 1970 Basic Books translation as Limited (search-only); the book is in copyright and the publishers offer no open access. No NOTE is filed. It stays Unavailable until a copy is supplied, not on merit. |
| LIT-809 v2 | That’s what she (could have) said: How alternative utterances affect language use | Bergen, Goodman & Levy (2012), Proceedings of the 34th Annual Meeting of the Cognitive Science Society, 120–125. Recursive speaker–listener reasoning yields specificity implicatures, but cannot derive Horn’s principle from costs alone; adding uncertainty over the lexicon does, in one-shot signalling games, without equilibrium refinements. Two online games with novel symbols find both implicatures drawn without prior conventions. | Active — read 2026-10-09 (NOTE-613); worth reading as the paper in which the RSA recursion is made to derive Horn’s division of pragmatic labour (costlier forms for less likely meanings) in one-shot signalling games with no prior conventions: the basic recursion cannot break the symmetry between equally meaningless costly and cheap signals, but a listener and speaker uncertain about the lexicon, marginalizing over every lexicon that could assign the signals meanings, converge on the efficient mapping without an equilibrium-selection rule. Two small Mechanical Turk games (40 and 140 participants) show people drawing specificity and Horn implicatures among novel symbols in five rounds, with no first-round difference. Source, with LIT-821 and LIT-816, of THEORY-172. |
| LIT-810 | Efficient compression in color naming and its evolution | Zaslavsky, Kemp, Regier & Tishby (2018), PNAS 115(31):7937–7942. Treats a language’s colour lexicon as an information-bottleneck encoder from perceptual meanings to words and finds the World Color Survey’s 110 languages, and English, near the IB complexity–accuracy bound, with one trade-off parameter β (fitted values just above 1) accounting for much of the variation between them. IB optima have soft boundaries and inconsistently named regions, as the data do, and the IB path through β passes through structural phase transitions that roughly recapitulate Berlin and Kay’s sequence, apart from an early yellow category the data lack. | Active — read 2026-10-09 (NOTE-631); worth reading as the paper that puts the information bottleneck to work on word meanings: a lexicon is an encoder q(w|m) from meanings (Gaussians over CIELAB colour) to words, scored by complexity I(M;W) against accuracy I(W;U), and the colour-naming systems of the World Color Survey and English lie close to the IB curve at β only slightly above 1, with 93% beating all 39 hue-rotated variants of themselves. Read it knowing that the need distribution (the LI source) is estimated from the naming data, cross-validated over languages, and that the “evolution” is a reading of the annealing path along the curve, not a model of historical change. Source of THEORY-163. |
| LIT-815 | Using Speakers’ Referential Intentions to Model Early Cross-Situational Word Learning | Frank, Goodman & Tenenbaum (2009), Psychological Science 20(5):578–585. A Bayesian word learner that infers the speaker’s intended referents and the lexicon together, with words uttered referentially or not, beats association and IBM Model 1 baselines on hand-annotated child-directed video (lexicon precision .67, F .55) and predicts mutual exclusivity, one-trial learning, word-driven object individuation and intention-guided mapping without principles built in for them. | Active — read 2026-10-09 (NOTE-611); worth reading as the model in which a word learner infers the speaker’s referential intention and the lexicon jointly: words are generated from an unobserved intended referent (a subset of the objects present, possibly empty) through the lexicon, or non-referentially, and the MAP lexicon is found by search. On two ten-minute CHILDES videos it learns a far more precise lexicon than association and translation baselines (F .55 against at most .22), and it reproduces mutual exclusivity, Xu’s word-driven object individuation and Baldwin’s intention-reading result. Read it as a word-learning paper: it has no model of an informative speaker; Bergen, Goodman and Levy (2012) cite a different 2009 paper, by Frank, Goodman, Lai and Tenenbaum, for that. |
| LIT-816 | Pragmatic Language Interpretation as Probabilistic Inference | Goodman & Frank (2016), Trends in Cognitive Sciences 20(11):818–829. The review that names and states the rational speech act framework: a listener interprets an utterance by Bayesian inversion of an approximately rational speaker who values informativeness to a literal listener, replacing Grice’s maxims with one utility-theoretic cooperative principle; joint inference over the speaker’s topic, knowledge or lexicon (uRSA) extends it to non-literal language, vagueness and embedded implicature. | Active — read 2026-10-09 (NOTE-606), from the authors’ manuscript; worth reading as the standard statement of the rational speech act (RSA) framework: a pragmatic listener P_L(w|u) ∝ P_S(u|w)P(w) inverting a softmax speaker whose utility is log P_Lit(w|u), with a literal listener conditioning the prior on the utterance’s truth. Its uncertain-RSA extension, a joint posterior over the world and the speaker’s type, topic or lexicon, is what carries hyperbole, irony, metaphor, vague adjectives and embedded implicature. A review: the empirical support it cites is summarized, not shown, and its claims of near-ceiling fits rest on the cited papers. Source, with LIT-821 and LIT-809, of THEORY-172. |
| LIT-820 | Iterated learning: Intergenerational knowledge transmission reveals inductive biases | Kalish, Griffiths & Lewandowsky (2007), Psychonomic Bulletin & Review 14(2):288–294. Human chains of function learners, each trained on the previous learner’s test responses, converged in a few generations to a positive linear function in 28 of 32 families, whatever function trained the first learner, as Bayesian iterated learning predicts for a prior favouring positive linear functions. The prior was inferred from earlier studies, not measured; stationarity was not shown. | Active — read 2026-10-09 (NOTE-616), in full, from the publisher’s PDF; worth reading as the first laboratory test of iterated learning against an inductive bias known before the experiment: in 32 chains of nine human learners each passing a function-learning task down the line, 28 converged within a few generations to a positive linear function whatever the first learner was trained on (positive linear, negative linear, U-shaped or random pairings), as Bayesian agents sampling from a prior that favours positive linear functions would. The prior was not measured in the experiment but taken from earlier function-learning studies; 3 families ended on the negative linear function and 1 did not converge, and the authors say more data would be needed to show the chains had reached a stationary distribution. Bears on THEORY-155. |
| LIT-821 | Predicting Pragmatic Reasoning in Language Games | Frank & Goodman (2012), Science 336(6084):998. In one-shot three-object reference games on Mechanical Turk, speakers’ word bets match a speaker who chooses words in proportion to their specificity (r = .98), and listeners’ referent bets match the Bayesian inversion of that speaker with an empirically measured salience prior (r = .99), though salience and listener bets alone are uncorrelated (r = .19). The founding quantitative test of what was later called the rational speech act model. | Active — read 2026-10-09 (NOTE-634); worth reading as the one-page origin of the rational speech act model’s quantitative test: a listener who inverts, by Bayes’ rule, a speaker who chooses words in proportion to their informativeness (exp of minus surprisal under a literal listener, which reduces to the size principle |w|⁻¹), combined with an empirically measured salience prior, predicts mean listener bets in three-object reference games at r = .99 with no fitted parameters (α set to 1). Read it knowing that the fit is to means over seven context types, that the prior is measured in a separate group, and that the speaker and listener groups never interact. Source of THEORY-172. |
| LIT-824 | Informative Communication in Word Production and Word Learning | Frank, Goodman, Lai & Tenenbaum (2009), Proceedings of CogSci 31, pp. 1228–1233. A speaker who chooses words to minimise the KL divergence between her intended meaning and a word’s literal meaning picks, for a known referent, a true word with probability ∝ |w|^(−α): the size principle, derived from communication. Word learners’ bets about a novel word’s meaning fit the inverted model at r = .93; speakers’ free descriptions follow it only weakly (r = .19, and not at all for colour terms). | Active — read 2026-10-09 (NOTE-627); worth reading as the paper that first writes the informative speaker down: a speaker soft-maximises the utility −D_KL(M_S ‖ w̃), the information lost about her intended meaning when it is replaced by the uniform distribution over a word’s extension, so for a known referent she picks a true word with probability ∝ |w|^(−α), the size principle derived from communicative premises. A learner who knows the referent inverts that speaker to infer what a novel word means; 700 web participants’ bets track the model’s predictions (r = .93). In a free-description task speakers’ use of Flickr tag-words correlates only weakly with informativeness (r = .19 overall, .51 without colour terms, .02 for colour terms). It has no listener inferring a referent and no recursion: that step is Frank and Goodman 2012 (LIT-821). |
| LIT-825 | Larger communities create more systematic languages | Raviv, Meyer & Lev-Ari (2019), Proc. R. Soc. B 286:20191262. In a laboratory group-communication game, groups of eight built more systematic (compositional) languages than groups of four, faster and more consistently, with the same total interaction; small groups varied more, as drift would predict. Input variability, greater in larger groups, predicted gains in structure. Group size was isolated from network structure and second-language learners, the confounds of cross-linguistic correlations. | Active — read 2026-10-09 (NOTE-628), the main text in full from the publisher’s version in the Radboud University repository; the electronic supplementary material was not read. Worth reading as the experiment that isolates community size: 12 groups of four and 12 groups of eight Dutch adults, fully connected, invented a typed language to describe moving shapes over 16 rounds with the same total amount of interaction. Larger groups’ languages became more systematic (string distance tracking meaning distance), faster, and ended more systematic; larger groups also varied far less from one another, while some small groups never built structure. Larger groups faced more input variability, and input variability predicted the next round’s gain in structure better than shared history did. Communicative success and convergence ended equal. It is a group-interaction paradigm, not a transmission chain: no generation turnover. Source of THEORY-164. |
| LIT-834 | A Quantum Question Order Model Supported by Empirical Tests of an A Priori and Precise Prediction | Wang & Busemeyer (2013), Topics in Cognitive Science 5(4):689–710. Models answering an attitude question as projecting a belief state onto the answer’s subspace and replacing the state by the normalised projection. Non-commuting projectors are then necessary for order effects, and a parameter-free constraint follows for any state and any projectors: the probability of answering the two questions differently is the same in both orders (the QQ equality). Five of six data sets meet it; the sixth, where background information was read out between the questions, fails it, as predicted. | Active — read 2026-10-09 (NOTE-636); worth reading as the source of the quantum question (QQ) model and its QQ equality: if each answer is a Lüders projection of one belief state, the probability of giving different answers to two questions is the same in both orders, whatever the state, the projectors or the dimension. Six data sets bear it out, and the Rose–Jackson poll fails it, as the authors predicted, because new information was inserted between the questions. Read with the caveat that the case against classical rivals is thin: the Bayesian argument is informal and the Markov proof is “available upon request”, so the paper shows that unconstrained classical models need not satisfy the equality, not that none can. |
| LIT-837 v2 | Structuralism | Piaget (1970), Structuralism, trans. and ed. Chaninah Maschler, Basic Books (London: Routledge and Kegan Paul, 1971; reissued 2015). The English edition of Le structuralisme (1968), held as LIT-802. Skimmed thinly from a publisher’s preview (17 pages of chs. I–IV): a structure is a whole under laws of composition, a system of transformations and self-regulating. Logico-mathematical structures are fully reversible (groups); linguistic and social ones are regulated in time but not fully reversible; and a structure is real only if governed from within. | Active — skimmed 2026-10-09 (NOTE-618), thinly: the publisher’s Google Books preview of the 2015 reissue served only 17 pages of text, pp. 3–63 at one page in four and p. 69, all in chs. I–IV, and none of the chapters on linguistics, the social sciences or philosophy, nor the conclusion. Worth reading as Piaget’s definition of structure by wholeness, transformation and self-regulation, and for his distinction between logico-mathematical structures, whose operations all have inverses (groups), and linguistic, social and psychological structures, whose transformations unfold in time under regulations that are “not entirely reversible”. He argues that a structure must be “governed from within” to be real, and defines the subject as “the center of functional activity”. Read it with the caveat that the skim saw the argument in fragments. |
| LIT-842 | On the Quantum-like Contextuality of Ambiguous Phrases | Wang, Sadrzadeh, Abramsky & Cervantes (2021), ARXIV-2107.14589, SemSpace 2021 (ACL Anthology, pp. 42–52). Treats each word of a two-word ambiguous phrase as a measurement whose outcome is the selected meaning. One hand-specified support model is possibilistically contextual; its BNC/ukWaC probabilities, like all the corpus models, are signalling. Under Contextuality-by-Default two subject–verb/verb–object pairs come out contextual, but the paper’s own bootstrap leaves both plausibly noncontextual. | Active — read in full 2026-10-09 (NOTE-645); worth reading as the first paper to test natural-language data for contextuality in the formal sense rather than the loose one, and as the paper that names the obstacle the test meets: corpus estimates of meaning selection are almost always signalling. In the sheaf-theoretic framework one hand-specified support model ({tap, box} × {pitcher, cabinet}) is possibilistically contextual (a Hardy-type failure), but its corpus probabilities are signalling, so the framework cannot judge them. Under Contextuality-by-Default the Bell-type corpus systems are noncontextual, and two noun–verb pairs read as subject–verb and verb–object (adopt/boxer, throw/pitcher) are CbD-contextual, by 1/30 and 7/30. Its own parametric bootstrap then gives those two a probability above .56 and above .08 of being noncontextual, on fractions with denominators between 3 and 30, so the contextual verdicts are not statistically secured. |
| LIT-851 | Analysing Ambiguous Nouns and Verbs with Quantum Contextuality Tools | Wang, Sadrzadeh, Abramsky & Cervantes (2021), Journal of Cognitive Science 22(3):391–420. Proves that in a binary cyclic system CbD’s signalling quantity Δ is twice the summed minimal direct influences of context in Jones’s causal canonical models, and measures it in 90 rank-2 noun–verb systems from British corpora: Δ is about 1.35 whatever the words’ ambiguity, and homonymous verbs carry about 70% of it against about 50% for polysemous ones, on small counts and a marginal test. | Active — read in full 2026-10-09 (NOTE-654); worth reading as the paper that measures how much signalling there is in natural-language meaning selection, and that gives signalling a causal reading. Its Proposition 1 shows that in a binary cyclic system the CbD quantity Δ is twice the sum, over contents, of the least direct influence of context that any of Jones’s canonical (causal) models of the data must posit, each the total-variation distance between a content’s two marginals. Applied to 90 rank-2 noun–verb systems from the BNC and ukWaC, Δ averages about 1.35 in every class of ambiguity, and homonymous verbs carry about 70% of it against about 50% for polysemous verbs (14 against 55 systems; my recomputation from its appendix gives a Welch t of 2.0, two-sided p ≈ 0.06). It tests no system for contextuality. Read with the caveat that most systems rest on very few corpus occurrences: 51 of the 90 have Δ exactly 0, 2 or 4. |
| LIT-862 | A mathematical theory of semantic development in deep neural networks | Saxe, McClelland and Ganguli (2019), PNAS 116(23):11537–11546, ARXIV-1810.10531. A two-layer linear network trained from small weights learns the singular modes of its input–output correlations one at a time, each in a sigmoidal step at a time of order (τ/s)ln(s/ε), where a shallow network learns them all at once. With data generated on a tree this gives coarse-to-fine differentiation in stages, transient illusory correlations and narrowing generalisation; the SVD also defines typicality, prototypes and category coherence, and minimum-norm weights fix the hidden similarity matrix across networks. | Active — read 2026-10-09 (NOTE-666); worth reading as the paper that turns the exactly solved learning dynamics of a two-layer linear network into a theory of semantic development. Trained by gradient descent from small weights on whitened inputs, the network learns the singular modes of the item–feature correlation matrix one at a time, each by a sigmoid at a time of order (τ/s)ln(s/ε), while a shallow network learns every mode on one exponential timescale. Hierarchically generated data have singular values that fall with depth in the tree, so broad distinctions are learned first in stage-like transitions, individual predictions can rise and fall between stages, and inductive generalisation narrows with development. The same SVD gives typicality, prototypes and a coherence threshold (a BBP transition), rings give Fourier modes, and minimum-norm weights, which small initialisation approximates, give identical hidden similarity matrices across networks. The trajectory is exact only from decoupled, balanced initial weights with white inputs; from random small weights it is checked by simulation, and the match to children and cortex is qualitative. |
| LIT-872 | Hyperbolic geometry of the olfactory space | Zhou, Smith & Sharpee (2018), Science Advances 4(8):eaaq1458. In four natural odour sources the rank order of correlations between compound concentrations has the clique-topology signature of points near the surface of a 3D hyperbolic ball and not of a uniform Euclidean cube, and human perceptual descriptions of 127 odorants have that of a full 3D hyperbolic ball. Pleasantness, boiling point and acidity read out as linear axes of the fitted fruit-odour space. The hierarchy that motivates the hyperbolic model is argued from biochemistry, not recovered from the data. | Active — read 2026-10-09 (NOTE-672); worth reading as the first application of clique topology to stimulus statistics rather than neural data, and as the Sharpee lab’s first claim of a hyperbolic geometry in biology. For four natural odour sources (strawberry, tomato, blueberry, mouse urine; 45–78 compounds over 50–101 samples), the rank order of absolute concentration correlations gives Betti curves that match points sampled near the surface of a 3D hyperbolic ball (R_max = 7, R_min = 0.9 R_max, the same for all four) and reject uniform Euclidean cubes of any dimension tried. Human descriptor profiles for 127 odorants (Dravnieks) match a full 3D hyperbolic ball and no Euclidean cube. Linear axes for pleasantness, boiling point and acidity are found in the fitted fruit-odour space. The tree that motivates the hyperbolic model is argued, not recovered: compounds do not cluster by chemical class, and the only null geometry is a uniform Euclidean cube. |
| LIT-897 v2 | AI as Agency without Intelligence: On Artificial Intelligence as a New Form of Artificial Agency and the Multiple Realisability of Agency Thesis | Floridi (2025), DOI-10.1007/s13347-025-00858-9. An Editor Letter that “grew so long”. AI is better understood by widening agency to forms without cognition, intelligence or mental states than by widening intelligence to machines. Agency comes in kinds (natural, biological, animal social, artefactual, human individual and social, artificial, social artificial). All of them share interactivity, and the richer ones add graded autonomy and adaptability. AI is computational, data-adaptive agency serving goals set by humans, with no intelligence. The thesis is assumed, and the paper characterises the kind rather than arguing for it. The criteria do not separate AI from other artefacts, and no level of abstraction is stated. | Active — read in full 2026-10-10 (NOTE-702); worth reading for the fullest statement of Floridi’s position that AI is “agency without intelligence”: the Multiple Realisability of Agency thesis set against the Artificial Realisability of Intelligence, Floridi & Sanders’ three criteria restated as graded, and a taxonomy that places AI as “a computational, goal-driven form of agency defined by human purposes” with Agentic AI as its collective form. Read it as a taxonomy, not an argument. The case for MRA over ARI is announced and not made, since the paper assumes AI is an agent and asks what kind. AI’s lack of intelligence is asserted. No level of abstraction is ever specified. Interactivity alone makes a river a natural agent, so the permissiveness objection is accepted rather than answered. |
| LIT-910 v3 | Effortless Attention: A New Perspective in the Cognitive Science of Attention and Action | Bruya (ed.) (2010), MIT Press. An edited collection arguing that attention and action can be at once highly focused and effortless, as in flow, against accounts that tie focus to effort. Unread. | Unavailable — unread, registered 2026-10-11 from Crossref and publisher metadata (MIT Press, Bradford Books; Brian Bruya, editor; ISBN 9780262013840). The editor’s introduction is listed on PhilArchive and was not opened. A full PDF on a third-party site was seen in search results and not used. Filed because it is the collection that names effortless attention as a problem for effort-based accounts of control; reading the introduction would place it. |
| LIT-916 v3 | Brief Online Training in Mindfulness Meditation Momentarily Reduces Some Impulsive Behaviors: A Randomized Controlled Trial | Ge, Xie, Sun, Fan, Zhang & Liu (2024), Mindfulness 16: 448–464. An RCT with mindfulness, monitor and control groups, measuring impulsivity by Stroop, Go/No-Go and delay discounting. As summarised: some impulsive behaviours reduced momentarily, discounting not significantly improved relative to the comparison groups, and no effect sustained at follow-up. Unread. | Unavailable — unread, registered 2026-10-11 from Crossref metadata and a search-result summary of the abstract. Springer’s page redirected to a login and the abstract was not read directly. Filed beside Dixon et al. (2019) because, as summarised, it dissociates the two: some impulsive behaviour reduced, delay discounting not improved. |
| LIT-921 v2 | Sparse Distributed Memory and Related Models | Kanerva (1992), RIACS TR 92.10, in Hassoun (ed.) 1993. SDM restated as a matrix memory, with the RAM as its zero-radius case, and related to correlation-matrix memories, feed-forward nets, the cerebellar cortex and the models of Marr and Albus. For uniform random data and exact cues it gives bit fidelity, the optimal activation p = (2MT)^−1/3 and a capacity near a tenth of the hard locations. Also held in the anthology as ANTH-LIT-669. | Active — read 2026-10-14 (NOTE-779) in full; worth reading as the compact statement of sparse distributed memory in matrix form, with the ordinary RAM as its zero-radius special case and a simple signal-to-noise analysis: fidelity Φ(ρ) with ρ² = pM/[1 + pT(1 + p²M)], best at p = (2MT)^−1/3, and capacity near a tenth of the hard locations at fidelity 0.999. Read it knowing that every number assumes uniform random data and exact retrieval cues, that the noisy-cue case and the critical distance are not analysed here, and that it relates SDM to the Hopfield net only through a cited capacity comparison. |
| LIT-922 v2 | The Organization of Behavior: A Neuropsychological Theory | Hebb (1949), Wiley. The source of the postulate that when one cell repeatedly takes part in firing another, the connection between them strengthens, and of the cell assembly: a group of cells that, once wired together by repeated co-activity, can re-excite itself. Not read. | Unavailable — searched 2026-10-14 for a lawful full text, none reached, so it is not read and no NOTE is filed. OpenAlex lists the Psychology Press DOI, closed, and an ACM citation record; Unpaywall finds no open copy; the Taylor & Francis page returned a 403 bot check; the Internet Archive holds 1949 scans only as controlled lending (printdisabled) or as Digital Library of India uploads of an in-copyright book, neither usable; no rights-holder posting was found. It stays Unavailable until a copy is supplied. |
| LIT-928 v3 | Sparse Distributed Memory | Kanerva (1988), MIT Press. An associative memory in a high-dimensional binary space, addressed by content: a pattern is written to every hard location within a radius of its address and read back from the locations within that radius of a cue. Iterated reads converge on a stored pattern only from cues within a critical distance of it. Also held in the anthology as ANTH-LIT-666. Not read here. | Unavailable — searched 2026-10-14 for a lawful full text, none reached, so it is not read and no NOTE is filed. The book has no DOI; the MIT Press page and MIT Press Direct returned Cloudflare 403 checks, and no open-access edition was found; the Internet Archive has no scan of the book; NASA’s Technical Reports Server holds Kanerva’s public-domain RIACS and Stanford reports on SDM, which are different works (LIT-921 is one of them). What the book says about cues beyond the critical distance, which CLAIM-180 leans on, is therefore still unchecked here. It stays Unavailable until a copy is supplied. |
| LIT-934 v2 | Quantum structure in cognition | Aerts (2009), Journal of Mathematical Psychology 53(5):314–348. Membership weights for pairs of concepts and their conjunctions and disjunctions (Hampton 1988) have a Kolmogorovian model iff they are neither over- nor underextended and a Kolmogorovian factor is non-negative; many items fail. A subspace lattice with meet and join fits the second kind of failure but provably not the first, which the paper models with an emergent-concept sector in Fock space. Negation appears only as an orthocomplement used to derive conjunction from disjunction. | Active — read 2026-10-14 (NOTE-808) in full; worth reading as the paper that sorts Hampton’s concept-combination data into two kinds of non-classicality and proves that the main kind, over- and underextension, cannot come from a subspace lattice’s meet and join. Read it knowing that its fitting model has a free parameter per item and stipulated conditions, that its two-layer account of thought is interpretation, and that negation enters only as the orthocomplement used to derive the conjunction model, never measured. Registered for QUESTION-055; THEORY-255 is filed from it. |
| LIT-957 v2 | Relative fluency (unfelt vs felt) in active inference | Brouillet and Friston (2023), Consciousness and Cognition 115:103579. A perspective essay. Fluency, the felt ease of processing, is read as the brain’s recognition of the precision of its own predictions. Unfelt fluency is a subpersonal change in precision that orients attention; felt fluency is a hierarchically deeper inference about it, reached by endogenous attention, and is attributed to the stimulus because its cause is internal. The case rests on reviewed discrepancy-attribution experiments (the hension effect and its successors). No model and no new data. | Active — read 2026-10-14 (NOTE-785) in full; worth reading as a short statement of how fluency research could be put into active inference: fluency as the recognition of the precision of one’s own predictions, registered first as a change against expectation and misattributed to the stimulus because its cause is internal. Read it knowing that it is a perspective essay with no model and no new data, that its experiments are reviewed second-hand, and that its “felt” fluency means fluency made reportable by endogenous attention, not a separation of experience from function. |
| LIT-967 v2 | Evidence for the representation of non-hierarchical structures in language | Nielsen et al. (2026). The paper argues that language is represented partly through linear chunks akin to part-of-speech sequences, not only through hierarchical constituents. Four preregistered phrasal decision experiments (N = 497) show such linear structures can be primed without constituents. Corpus analyses of eye-tracked reading and conversation support the effect outside the lab. Not read. | Unavailable — searched 2026-10-14 for a lawful full text and none was reached, so it is not read and no NOTE is filed. OpenAlex: closed (PubMed and Scopus records only); Unpaywall: not open access; Europe PMC: no PMC copy, “Subscription required”; Semantic Scholar: no open PDF; nature.com: subscription article (exclusive licence, “Access through your institution”); OSF: the authors’ project osf.io/ymndq is public but holds materials, data, analyses and four preregistrations, not the manuscript, and no OSF or PsyArXiv preprint was found by title search; the Cornell lab pages failed on the proxy (502), failing on access, not availability; Aarhus Pure search returned no matching item; a web search found only news coverage. It stays Unavailable until a copy is supplied. |
| LIT-1029 | Autonomy, Function, and Representation | Bickhard (2000), Communication and Cognition – Artificial Intelligence 17(3–4). A far-from-equilibrium system’s contributions to its own maintenance are functions for it, with function and dysfunction fixed by the asymmetry between persisting and decaying; a candle flame qualifies. A recursively self-maintenant system, one that switches among self-maintaining processes as its environment changes, has representation: indications that an interaction will succeed, which can be false by the system’s own lights. Etiological, informational and Dretskean accounts are rejected as epiphenomenal or observer-dependent. | Active — read in full 2026-10-13 (NOTE-729) from the author’s manuscript text on his Lehigh site (https://www.lehigh.edu/~mhb0/autfuncrep.html); the printed issue was not compared. Worth reading as a compact statement of the organizational account of function before Mossio et al., and of interactivism: function is contribution to a far-from-equilibrium system’s own persistence, relative to that system, and representation is an indication of interaction success that the system can find false. It supports the premise of CLAIM-288 by a second route, gives QUESTION-087 an intrinsic standard bounded by persistence (so it agrees with CLAIM-075), and is vocabulary, not evidence, for CLAIM-284. |
| LIT-1054 | On Distinguishing Epistemic from Pragmatic Action | Kirsh and Maglio (1994), Cognitive Science 18(4). Kirsh and Maglio define an epistemic action as a physical action whose primary function is to reduce the space, time or unreliability of the agent’s own mental computation, and argue from Tetris keystroke data that players rotate and translate pieces for that reason: zoids are rotated more than plan-then-act would need, rotation starts within a few hundred milliseconds of a zoid appearing, and a translate-to-wall check is used more the higher the drop. The computational saving is argued, not measured: physical rotation of 90° takes as little as 100 ms against an estimated 800–1200 ms mentally (a three-player pilot), and fewer attentional probes and less working memory are needed. The trade is physical action in place of internal computation at the moment of use, not precomputation or caching; the paper explicitly sets it against caching and chunking accounts of skill, and mentions amortized complexity only in passing for everyday cases such as sorting bolts before assembly. It says the choice fits a Stigler-style cost–benefit model in outline but that such models fail in detail, because an epistemic action’s value depends on the particular internal computation it serves. | Active — read in full 2026-10-15 from the scan of the published article (Cognitive Science 18(4), pp. 513–549) that David Kirsh posts on his UCSD publications page (adrenaline.ucsd.edu/kirsh/Articles/CogsciJournal/DistinguishingEpi_prag.pdf); the Wiley open-access PDF that OpenAlex lists refused requests from this environment. Worth reading as the founding definition of epistemic action: a physical action whose primary function is to reduce the space, time or unreliability of the agent’s own mental computation, argued from Tetris players’ extra and early rotations (rotation under way by 400–600 ms, physical rotation in as little as 100 ms against an estimated 800–1200 ms mentally) and a translate-to-wall check used more for higher drops. For QUESTION-095 it shows offloading at the moment of use, not precomputation: it sets itself against caching accounts of skill, mentions amortized complexity only in a parenthesis about sorting bolts, gives no formal cost model, and says simple cost–benefit models fail because an epistemic action’s value depends on the computation it serves. For TERM-055 it is a case of raising availability during the deadline, not before the request; for CLAIM-303 it has a single use and no upkeep, so it bears only on the general shape of the trade. |
| LIT-1055 | Contributions of ideal observer theory to vision research | Geisler (2011), Vision Research 51(7). A short review, not a methods paper. It defines the ideal observer as the device that “performs a given task at the optimal level possible, given the available information and any specified constraints”, writes the Bayesian ideal as an argmax of expected utility over the posterior given either the stimulus S or a constrained neural representation Z = g(S;θ) (eq. 1), and surveys applications from photon-noise detection to eye-movement search. It gives no formula for efficiency and no procedure for locating information loss beyond saying that constraints are added to the ideal one at a time (optics, receptor sampling, receptor noise, ganglion-cell summation) and that models are built by “degrading the ideal observer with hypothesized neural noise or with hypothesized heuristic computations”. Its constraints restrict the information available to an optimal decoder; it has no notion of an optimum over a restricted class of computations, though its attention examples compare the ideal with specific suboptimal rules (a one-location “serial” mechanism, a MAP searcher, a random searcher). | Active — read in full 2026-10-15 from the NIH author manuscript in PubMed Central (PMC3062724, NIHMS252298), taken as JATS XML from NCBI E-utilities (eutils.ncbi.nlm.nih.gov/entrez/eutils/efetch.fcgi?db=pmc&id=3062724) because the PMC article page served a reCAPTCHA; the typeset Elsevier open-archive PDF returned 403, so printed pagination was not checked. Worth reading as a compact statement of Bayesian ideal observer analysis: the ideal is optimal “given the available information and any specified constraints”, written as an expected-utility argmax over the posterior given either the stimulus or a constrained neural representation Z = g(S;θ), with a survey of applications and some numbers (r = 0.83 for colour naming on individual cone mosaics; 42% lower MSE for a natural-statistics decoder of peripheral samples; about 99% of variance in contour-occlusion judgments). For QUESTION-103 it shows less than the question assumes: it gives no formula for efficiency (the squared d′ ratio is not stated here and should be cited to Burgess et al. 1981 or Pelli and Farell 1999, or owned by the record), and its constrained ideals restrict the input, a garbling of the stimulus, with unrestricted computation on it, so losses are located stage by stage by the data-processing logic, not by an optimum over a restricted class of observers as in CLAIM-302. Its nearest analogue to a bounded class is the comparison with named heuristic rules (serial one-location attention, MAP fixation selection), and its remark that far-below-ideal performance leaves many models standing is the empirical form of the point that a gap does not identify its cause. Consistent with THEORY-031; no bearing on CLAIM-302 itself. |
| LIT-1056 | Amortized Inference in Probabilistic Reasoning | Gershman and Goodman (2014), Proceedings of the 36th Annual Conference of the Cognitive Science Society (pp. 517–522 by the authors’ later citation). Proposes that people amortize probabilistic inference by caching (memoizing) sub-inferences and composing them for related later queries. In two Mechanical Turk experiments on a five-node network, answers to a complex query depended on the person’s own earlier answer to a sub-query only when the sub-query came first (r = 0.28 against −0.11), and responses were faster when reuse was available. Costs of reuse are discussed, not measured. | Active — read in full 2026-10-15 from the authors’ copy on the Gershman lab site (gershmanlab.com/pubs/GershmanGoodman14.pdf, 6 pp.). The version of record is on eScholarship (escholarship.org/uc/item/34j1h7k5), whose PDF refused direct requests from this environment. The page range 517–522 is from the authors’ own later citation. Worth reading as the paper that named amortized inference for cognitive science. It defines amortized inference as caching (“memoizing”) sub-inferences and reusing them on related queries. It gives the first behavioural evidence for this, from two small online experiments on one Bayesian network: with the sub-query first, answers to a complex query were predictable from a person’s own sub-answer (r = 0.28, against −0.11 in the reverse order), and they were faster relative to a matched baseline. For QUESTION-095 it shows that the authors treat amortization as retrieval standing in for recomputation, which supports the question’s unification for internal memory only. It does not touch external storage or distributed cognition. It supplies storage cost, decomposability and frequency as untested trade-offs, but no staleness term, and no directional bias toward earlier answers, which belongs to later work (LIT-1074). For CLAIM-303 it is consistent with the trade and with frequency being one consideration among several, and its error propagation is error preserved by reuse, not staleness. |
| LIT-1057 | Adapting the Environment Instead of Oneself | Kirsh (1996), Adaptive Behavior 4(3–4). A conceptual paper. It models a task as a directed graph of states and actions with a cost metric on transitions, and classifies how an agent can improve its payoff by actions external to the task: pruning hard initial states, routine maintenance, scouting ahead, tools (which change the graph), and actions that make a space more “cognitively congenial” (which leave the graph intact and reduce mental operations). It names amortization once, as a parenthetical gloss on chimps leaving nut-cracking stones in place, and gives an investment rule for maintenance (future benefit against present cost times an “interest factor” for risk) in words only. There is no cost model with reuse frequency, retrieval cost or staleness; mental cost lies outside the graph, and the paper says there is “no general theory” of congeniality. The empirical figures are unpublished pilot results. | Active — read in full 2026-10-15 from a scan of the published article (Adaptive Behavior 4(3/4), pp. 415–452) that the author posts on his UCSD laboratory server (interactivity.ucsd.edu/articles/Adaptive/Adapting_env_00.pdf), text by OCR and cross-checked against the author’s HTML version in the same directory; the PhilArchive copy refused automated requests and was not used. Worth reading as the distributed-cognition statement of environment redesign as a change to a task’s state graph or to its “cognitive congeniality”, with routine maintenance (keep things where they are expected so that search is unnecessary, against objects’ tendency to scatter) and an investment rule for doing work before it is needed. For QUESTION-095 it shows that “amortize the cost of search” is a single parenthetical about chimps leaving nut-cracking stones in place, not developed; that the graph prices only physical transitions, so retrieval and mental cost fall outside the formalism by the paper’s own account; and that the nearest thing to a cost model is a verbal “interest factor” whose footnote names the risks that the result will not be used and that the change will not still be present, with no reuse frequency, retrieval cost, staleness of a result or query predictability. Its most usable point is that labour cost varies over time, so work should be done when it is cheap, which is the cockpit’s rationale. It does not cite Tarjan (that is Kirsh 1995, LIT-1076), and its value-of-information example changes outcome probabilities rather than pricing a choice made after looking. For CASE-085 it supports the labelled-compartment and standard-layout counterfactuals and their upkeep by the agent, a mechanism or the group, and says nothing on search cost, deadlines or when to stop searching. |
| LIT-1058 | Soft constraints in interactive behavior: the case of ignoring perfect knowledge in-the-world for imperfect knowledge in-the-head | Gray and Fu (2004), Cognitive Science 28(3). In a simulated VCR-programming task, people with show information always available on screen still made errors that comparing against the display would have prevented. Making that information cost a mouse move and click instead of an eye movement (an estimated 1,000–1,500 ms against 500 ms) cut how often they looked it up, and most checks came right before a setting, almost none while or after setting it. The group made to memorize the information first made the fewest errors. Gray and Fu read this as non-deliberate least-effort choice among microstrategies, with memory time and perceptual-motor time weighed alike (“milliseconds matter”). The access costs are model estimates, not measured or varied parametrically. The error difference between the two on-screen conditions was not significant, the free-access look-up data come from a separate eye-tracking study of eight people, and the Memory-Test condition differed from the others in more than retrieval cost. | Active — read in full 2026-10-15 from the publisher’s typeset PDF (Cognitive Science 28(3), pp. 359–382), as posted on a Drexel University course page (www.cs.drexel.edu/~dds26/courses/cs680-s11/BK/readings/Gray-CS04.pdf). The authors’ own copy on Gray’s RPI page, and the Internet Archive’s 2004 capture of it, could not be reached from this environment. Unpaywall and OpenAlex list the work closed, and Wiley, ScienceDirect and DTIC refused direct requests. Worth reading as the standard experimental case that people choose between recalling information and looking it up by a non-deliberate comparison of effort, in which memory time and perceptual-motor time count alike. In a simulated VCR task, adding a click (an estimated 1,000–1,500 ms against 500 ms for an eye movement) cut look-ups of perfect on-screen information, and checks of finished settings rested on memory. The group that memorized the information first made the fewest errors. The millisecond costs are model estimates, the error difference between the two on-screen conditions was not significant, and the ACT-R models only estimate effort; none models the choice. For QUESTION-095 it prices only the read-back side of the trade, setting aside encoding and search, and it shows that a result left in the world goes unused once recall is cheaper; it says nothing of amortized inference. For CLAIM-296 and TERM-055 it shows a change of arrangement alone changing use of the information, and that the procedure actually chosen is the cheapest, not the best available. It bears on CASE-085 only by analogy, because its task removes search, which is where the EMT case failed. |
| LIT-1059 | Cognitive Offloading | Risko and Gilbert (2016), TICS 20(9). A short review that defines cognitive offloading as physical action that changes a task’s information-processing requirements to reduce cognitive demand, splits it into offloading onto the body and into the world, and proposes a metacognitive model: whether people offload is a strategy choice driven by their evaluation of internal versus extended strategies, which tracks objective internal demand but can be wrong, and offloading in turn changes both memory and those evaluations. The evidence is a handful of choice/no-choice studies, summarised qualitatively; the review reports no effect sizes and no formal account of costs. It names costs (memory for offloaded content, GPS-impaired spatial memory, photo-taking impairment, automation complacency and bias, skill decay, inflated self-knowledge, a self-reinforcing drift toward external reliance). It does not frame offloading as computation reused over time, and says nothing about how often an offloaded result is used. | Active — read in full 2026-10-15 from the authors’ manuscript in UCL Discovery (discovery.ucl.ac.uk/id/eprint/1508770/1/gilbert_TiCS_OFFLOADING_RPS.pdf, 37 pp.; the typeset Elsevier version, its Trends box and its Outstanding Questions box were not seen). Worth reading as the standard definition of cognitive offloading (“the use of physical action to alter the information processing requirements of a task in order to reduce cognitive demand”), its split into onto-the-body and into-the-world, and a metacognitive model in which the choice to offload follows beliefs about internal versus extended strategies, tracks objective demand but can be wrong, and feeds back through lost memory and changed confidence into a drift toward external reliance. It reports no effect sizes and gives no cost formalism. For QUESTION-095 it does not supply the retrieval-as-amortized-computation account; it adds that offloading shifts remembering “what” to remembering “where”, that people offload where it gives nothing (about 40% at ceiling load), so when it pays and when people do it come apart, and that automation bias is the nearest thing to the staleness cost. For CLAIM-303 it fits the direction (more is offloaded as the later internal cost rises) but says nothing about reuse count, and shows that offloading itself lowers internal capacity, which the claim’s fixed ledger leaves out. |
| LIT-1060 | Cognition in the Wild | Hutchins (1995), MIT Press. A study of navigation aboard a ship, arguing that cognitive processes are distributed across people, artifacts and their organization, and that the unit of analysis is the functional system. | Unavailable — filed 2026-10-15, not read. Registered from the exchange of 2026-10-14 on prompting, epistemic memory and retrieval (the workbench entry of 2026-10-15). No NOTE is filed, and it stays Unavailable until it is read, not on merit. Looked for a lawful full text on 2026-10-15 and found none. Unpaywall and OpenAlex report it open access (green) on the strength of a single repository record at the University of Ljubljana (repozitorij.uni-lj.si/IzpisGradiva.php?id=44611, no PDF URL given); that record returns HTTP 410 with “Gradivo ne obstaja” (the item does not exist), and the Internet Archive has no capture of it, so the OA flag is stale. OpenAlex marks the MIT Press location itself as not open access, Crossref carries no licence for the DOI, and Semantic Scholar reports it closed. MIT Press Direct (direct.mit.edu/books/monograph/4604/Cognition-in-the-Wild, reached through the DOI) and the MIT Press catalogue page both returned 403 behind a Cloudflare challenge, the Internet Archive holds no capture of either, and a web search found only paid editions and library records and no sign that MIT Press has made the title open access. Hutchins’s own UCSD site (pages.ucsd.edu/~ehutchins/citw.html) posts the book’s Introduction in full (about 3,000 words: how the study began, the critique of an ideational, inside-the-head view of culture and cognition, and a chapter-by-chapter outline), but nothing else from the book. hci.ucsd.edu/hutchins/ is a JavaScript shell with no documents. No shadow library, Internet Archive lending copy or third-party upload was used. The Introduction states the move the record cites it for. It does not argue for it. It says Hutchins will “move the boundaries of the cognitive unit of analysis out beyond the skin of the individual person and treat the navigation team as a cognitive and computational system”, that “local functional systems composed of a person in interaction with a tool have cognitive properties that are radically different from the cognitive properties of the person alone” (chapter 3), and that “Depending on their organization, groups must have cognitive properties that are not predictable from a knowledge of the properties of the individuals in the group” (chapter 4 takes the team as a whole). Hutchins’s “How a cockpit remembers its speeds” (Cognitive Science 19(3), 1995, pp. 265–288, DOI 10.1207/s15516709cog1903_1) has lawful copies: the author posts the typeset article (pages.ucsd.edu/~ehutchins/documents/CockpitSpeeds.pdf, 24 pp., fetched 2026-10-15), and Unpaywall, OpenAlex and Semantic Scholar list Wiley’s free publisher PDF (bronze; onlinelibrary.wiley.com/doi/pdfdirect/10.1207/s15516709cog1903_1, which returned 403 to this environment). It is not registered, and it is the reading available now for QUESTION-101. |
| LIT-1062 | Predicting Short-Term Remembering as Boundedly Optimal Strategy Choice | Howes, Duggan, Kalidindi, Tseng and Lewis (2016; online 2015), Cognitive Science 40(5). Participants copied names one at a time from a mock email to a mock calendar and, after a forced phase that measured their own copying rate at every list length from 3 to 9, chose how many names to take per trial; the list length that had given each person the highest rate in their own data predicted their choice better than one item fewer or one more (55% vs 8% and 17% of choices in Experiment 1, N = 40; 46% vs 12% and 14% in Experiment 2, N = 20). An exponential points schedule raised chosen list length (6.99 vs 5.07), but raising the cost of an error so that one wrong item voids the whole trial did not (5.00 vs 5.05). The “boundedly optimal” prediction is the argmax of each participant’s measured rate, not a model with capacity, error or look-up parameters, and the task has no look-up: names are shown once, in sequence, and cannot be consulted again. 13 of 60 participants persistently chose a non-optimal length, 10 of them one item more. | Active — read in full 2026-10-15 from the text of the publisher’s typeset article as deposited in the University of Michigan’s Deep Blue repository (hdl.handle.net/2027.42/133633), served as extracted full text by CORE’s reader page for that record (core.ac.uk/reader/86638046); Wiley, the Birmingham and Deep Blue PDFs, and the Internet Archive’s captures could not be fetched from this environment, so figures were not viewed. Worth reading as the individual-level test of bounded optimality in short-term remembering: after a forced phase measures each person’s copying rate at every list length from 3 to 9, the length with the highest measured rate predicts that person’s later choice better than one fewer or one more (55% vs 8% and 17% of choices, N = 40; 46% vs 12% and 14%, N = 20), an exponential reward for long lists raises the chosen length, and raising the cost of an error to the whole trial’s points does not. The prediction is each person’s empirical argmax, not a model: there is no capacity or look-up-time parameter, and the task has no look-up at all, since names are shown once and cannot be consulted again, so the alternative to holding more is more, shorter trials. For QUESTION-095 it shows people can find the best point of a memory-versus-overhead trade by rate, and adds one term Gray et al. 2006 (LIT-1073) lacked, a time cost for a wrongly remembered item; it has no term for query predictability, external retrieval or search, staleness, reuse across trials or erosion of the alternative, and it corrects the description of it, from Lieder and Griffiths (LIT-1071), as a memorize-or-look-up trade priced by look-up time. |
| LIT-1064 | Explaining the enigmatic anchoring effect: Mechanisms of selective accessibility | Strack and Mussweiler (1997), JPSP 73(3). Proposes that comparing a target with an anchor makes anchor-consistent knowledge about the target more accessible, and that this knowledge then drives the absolute judgement. | Active — read in full 2026-10-15 from a scan of the APA typeset article (pp. 437–446) posted in the readings directory of a University of Florida graduate course (Lyle Brenner, MAR 7588; bear.warrington.ufl.edu/brenner/mar7588/Papers/strack-mussweiler-jpsp97.pdf). That is a course’s copy of the publisher’s version, not an author’s copy. Unpaywall, OpenAlex and Semantic Scholar list the article as closed. No copy was found on the authors’ pages: London Business School’s publication pages returned 404 or failed TLS, and the Internet Archive refused connections from this environment. Worth reading as the source of the selective accessibility model of anchoring. Three small studies (N = 32, 32 and 67) show that, with the anchor number held fixed, the anchor’s effect depends on whether the knowledge used to compare applies to the later estimate. Applicability came from the dimension: the estimate was on the dimension compared, or on another one (Study 1). Representativeness came from the object: the estimate was about the same object, or about a very different one, where the effect turned to contrast (Study 2). With plausible anchors the comparison was slower and the later estimate faster (Study 3). For CLAIM-298 (whose psychological form is now CLAIM-315) it argues against numeric priming of a response, and supports the claim’s psychological form for plausible anchors. It infers accessibility rather than measuring it; the direct lexical-decision test is Mussweiler and Strack 2000a, which the record does not hold and LIT-1067 failed to replicate. It does not rule out adjustment, and assigns adjustment to implausible anchors, which anchored at least as strongly. Its model holds that accessibility does not fix the direction of the judgment, so evaluation matters as well as retrieval. For QUESTION-109 it states, without testing, that hypothesis-consistent retrieval “improves the validity of comparative judgments but distorts subsequent judgments for which no specific hypothesis has been provided”. |
| LIT-1065 | The Ecological Approach to Visual Perception | Gibson (1979). The ecological account of perception, in which an affordance is a possibility for action that an environment offers an animal, relative to the animal’s capacities. | Unavailable — filed 2026-10-15, not read. Registered from the exchange of 2026-10-14 on prompting, epistemic memory and retrieval (the workbench entry of 2026-10-15). No NOTE is filed, and it stays Unavailable until it is read, not on merit. Looked for a lawful full text of the book, or of Chapter 8 (“The Theory of Affordances”, p. 119 in the classic edition) alone, on 2026-10-15 and found none. Unpaywall and OpenAlex report the 2015 Psychology Press classic edition (10.4324/9781315740218) closed, with no open-access location and no repository full text, and Semantic Scholar gives no openAccessPdf; the publisher’s page (taylorfrancis.com) refused requests from this environment (HTTP 403) and the book is sold, not open access. The 1986 Lawrence Erlbaum edition (10.4324/9780203767764) and the chapter’s reprint in The People, Place, and Space Reader (Routledge 2014, 10.4324/9781315816852-11) are also closed in Unpaywall and OpenAlex. Crossref lists no chapter-level DOI for the classic edition. Web searches found no author-, estate- or publisher-posted copy and no open-access anthology reprint. A scan of part of Chapter 8 was posted on a University of Oslo course page (IN3220, spring 2021, uio.no/studier/emner/matnat/ifi/IN3220/v21/gibson_james_j_1977_1979_the_theory_of_affordances.pdf); it now returns 404 and the course page is unpublished, and a course’s scan of a publisher’s chapter is not the author’s copy, so it was not used. A full PDF of the classic edition on a university library server (library.uniq.edu.iq) is named as a z-lib download and was not used. Secondary summaries (a book report from HVL’s NaChiLit group) were not treated as a reading. None found. |
| LIT-1066 | How a Cockpit Remembers Its Speeds | Hutchins (1995), Cognitive Science 19(3), 265–288. A task analysis of how a two-pilot airliner cockpit “remembers” approach speeds: computed early from a speed-card booklet indexed by gross weight, passed across speech to bugs on the airspeed indicators, and used minutes later as a judgement of spatial adjacency. The memory is a property of the functional system, “not made primarily of pilot memory”, though pilot memory is involved. Stylized and unmeasured, for an anticipated, invariant task. | Active — read in full 2026-10-15 from the author’s posted copy, a scan of the published article (Cognitive Science 19(3), pp. 265–288) at pages.ucsd.edu/~ehutchins/documents/CockpitSpeeds.pdf. Worth reading as the compact statement of distributed cognition’s method, a cockpit treated as a cognitive system whose representations can be observed, and as the source of the speed-bug argument: setting the day’s speeds as bugs on the airspeed indicator does not improve any pilot’s memory but turns recall and scale-reading into a judgment of spatial proximity, so the memory “has now become a property of a larger system”. For QUESTION-101 and CLAIM-296 it shows, by task analysis and not by measurement, a change of arrangement alone changing what the system can do; the retrieval cue (displayed gross weight, the needle nearing a bug) supplied by the environment; and the work of retrieval moved to a quiet phase 25 to 30 minutes ahead of use, which the record’s deadline does not yet represent. It narrows them: its task is anticipated, mandated and invariant, it analyses only success, and it says nothing about search for one item among many under a deadline, which is CASE-085’s situation. |
| LIT-1067 | Failures to replicate a key result of the selective accessibility theory of anchoring | Harris, Blower, Rodgers, Lagator, Page, Burton, Urlichich and Speekenbrink (2019), JEP: General 148(9), e30–e50. Five preregistered experiments (N = 128 each) repeated Mussweiler and Strack’s (2000a) test of selective accessibility, which asks whether a high or low anchor speeds recognition of anchor-consistent words. Four used lexical decision and one used continuous identification. Anchoring was large in all five, the accessibility effect was absent in all five, and an exploratory meta-analytic Bayes factor favoured the null about 21 to 1. Because the same tasks also failed to detect direct picture priming, the authors conclude that the method is unfit as a test, not that the theory is false, and that the earlier evidence from it “is not a strong part of the evidence base”. | Active — read in full 2026-10-15 from the authors’ manuscript on UCL Discovery (discovery.ucl.ac.uk/10074759/1/HarrisEtAl_2019_anchoring.pdf, eprint 10074759; 76 pp. with appendices and supplementary experiments). It is a Word-generated manuscript dated 23 May 2019, which Unpaywall and OpenAlex label submittedVersion. Its abstract is worded differently from the published one, so it is not the APA version of record (JEP: General 148(9), e30–e50), which returned 403 from this environment. Worth reading as the preregistered, high-powered test of the lexical-decision evidence for selective accessibility (Mussweiler & Strack 2000a, Studies 1 and 2). In five experiments of N = 128, anchoring was large every time (d ≈ 0.85–1.39) and anchor-consistent word recognition never moved (η²G ≤ .004). An exploratory meta-analytic Bayes factor favours the null by 21 to 1. Validation studies show the same tasks also miss direct picture priming. For CLAIM-298 it shows that the most direct published measure of “anchors change what is reachable” does not replicate and cannot detect priming. The claim’s psychological form therefore rests on inferred, not measured, accessibility, and the theory is neither refuted nor confirmed. The authors’ leading explanation of durable anchoring is a provisional answer formed at the comparison, which is neutral between retrieval and evaluation. For QUESTION-109 it bears only indirectly. It also corrects the record’s attribution: by this paper’s account the lexical-decision studies are Mussweiler & Strack 2000a, not 1999 (LIT-1078). |
| LIT-1071 | Resource-rational analysis: Understanding human cognition as the optimal use of limited computational resources | Lieder and Griffiths (2020; online 2019), BBS 43. Resource rationality defines the right cognitive mechanism as the one, among those a brain can execute, that maximizes expected utility of its result minus the expected opportunity cost of the time and resources it uses (Eq. 3), averaged over the environments consistent with what is known (Eq. 4); resource-rational analysis is a five-step iterate-and-refine method for deriving process models from that objective. The objective is per episode. It has no separate terms for storing a result, retrieving it later, or its going stale: the value of what a computation leaves behind is folded into the utility of the episode’s result, and the cost of finding the heuristic in the first place is assigned to evolution and learning and left out. Reuse appears only in the review, qualitatively: habits as reused action sequences, Dasgupta et al. 2018’s memorized hypotheses, Howes et al. 2016’s memory-load choice (the review’s “memorize-or-look-up”), and “cognitive prostheses”. The framework is offered as a methodological device, not a claim that people are optimal. | Active — read in full 2026-10-15 from the publisher’s typeset PDF of the whole BBS treatment (target article pp. 1–16, Authors’ Response pp. 45–51, and the commentaries of Bates, Sims & Jacobs, Dimov, Dingemanse, and Ma & Woodford; the other 21 commentaries searched by keyword only), posted by the Griffiths lab at cocosci.princeton.edu/papers/lieder_resource.pdf. Unpaywall and OpenAlex list it as closed. Worth reading as the standard statement of resource rationality: the best mechanism a brain can run is the one that maximizes expected utility of its result minus the expected opportunity cost of the time and resources it takes (Eqs. 3–4), and the five-step method that derives process models from that. For QUESTION-095 it shows that one objective can cover computing, remembering and retrieving, and it groups habits, memorized inferences (Dasgupta et al. 2018) and how much to hold in memory (Howes et al. 2016, which the review calls a memorize-or-look-up trade) under it. But the objective is per episode and has no separate terms for precompute, storage, retrieval or staleness. It leaves the cost of finding a heuristic out on purpose, as already paid by evolution and learning. Offloading to the world appears only in Dingemanse’s commentary, which the authors accept as a frontier. For CLAIM-303 it supplies the general benefit-minus-cost form the break-even condition fits, and no support for the claim’s staleness term. |
| LIT-1073 | The soft constraints hypothesis: A rational analysis approach to resource allocation for interactive behavior | Gray, Sims, Fu and Schoelles (2006), Psychological Review 113(3). In a Blocks World copying task where the model pattern was hidden and opening it cost a lockout of 0 to 3,200 ms, people opened it less often, kept it open longer, and placed more blocks after the first look as the lockout rose, and the trend was graded down to differences of tens of milliseconds. An Ideal Performer Model that chooses how many blocks to encode per look (ENCODE-1 to ENCODE-8) by Q-learning, with the only reward being negative time and memory given by ACT-R’s base-level activation, retrieval-threshold and latency equations, fits four measures from Experiment 3 with three memory parameters and one softmax noise parameter fitted on one of them. The authors read this as support for choosing between head and world by time alone, against a minimum memory hypothesis, which they concede no one has stated precisely enough to model. The model’s only memory costs are encoding time, retrieval time and forgetting; it has no term for anything carried across trials, for a retrieved item being wrong, or for the world changing. | Active — read in full 2026-10-15 from the publisher’s typeset PDF (Psychological Review 113(3), pp. 461–482) as posted on a University of Colorado course page (home.cs.colorado.edu/~mozer/Teaching/syllabi/3702/readings/GraySimsSchoellesFu2006.pdf), a course’s copy of the typeset article; the authors’ RPI pages, CiteSeerX, DTIC and the Internet Archive could not be reached. Worth reading as the quantitative sequel to Gray and Fu 2004 (LIT-1058): in a Blocks World copying task with the pattern behind a lockout of 0 to 3,200 ms, people look less and memorize more as looking gets dearer, and a Q-learning Ideal Performer that minimizes only time, with ACT-R’s activation, threshold and latency equations for memory, fits four measures with four parameters. The case against the minimum memory hypothesis is argued, not tested, since no minimum-memory model was built. For QUESTION-095 it shows that head and world can be priced in one currency, with encoding, rehearsal, retrieval time and forgetting on the head side and a fixed access cost on the world side, and that the choice of how much to read per look is valued against the whole trial; it has no term for precomputation, for reuse across trials (each pattern is new; only the learned policy carries over, which the paper calls amortized and does not price), for search in the world, or for staleness, since the target never changes and a retrieval cannot be wrong. |
| LIT-1074 | Remembrance of inferences past: Amortization in human hypothesis generation | Dasgupta, Schulz, Goodman and Gershman (2018), Cognition 178. In three online experiments (N = 84, 80 and 100, before exclusions), a sub- or superadditivity bias induced in a first probability judgment carried over into a second, always-packed judgment about the same scene distribution, and errors on the two queries were correlated trial by trial (r ≈ 0.34–0.67). The carry-over appeared only when the second query shared part of its hypothesis space with the first, which rules out simple numerical anchoring; it was larger when cognitive load fell on the first query, which the authors’ simulations read as reuse of a summary statistic rather than of the stored samples; and with a new cue it appeared (for subadditivity only, and weakly) when the cue’s posterior was similar to the first. The paper defines amortization as reuse of past computation bought at a cost in bias, and itself places it in the time-for-space trade. It measures no response times, so the speed benefit is assumed, not shown; efficiency is shown only by better secondary-task accuracy on reusable trials; and it gives no account of when reuse pays. | Active — read in full 2026-10-15 from the publisher’s typeset PDF of the version of record (Cognition 178, 67–81), posted by Gershman on his lab’s publication page (gershmanlab.com/pubs/Dasgupta18_remembrance.pdf); Unpaywall and OpenAlex list it as bronze open access at the publisher, whose site refused requests from this environment, and the bioRxiv preprint was not used. Worth reading as the empirical case for amortized inference in people, defined as reuse of past computation that gains efficiency at a cost in bias: in three experiments (N = 84, 80, 100) an unpacking bias in one probability judgment carried over into a later related one, only where the queries shared hypotheses, more strongly under load at the first query (read as reuse of a summary statistic rather than stored samples), and somewhat selectively for similar cues. It measures no response times, so the speed gain is assumed, and the filed summary’s “biased toward earlier answers” should read “carrying over the earlier computation’s error”, since numerical anchoring was ruled out. For QUESTION-095 it shows that the paper itself frames amortization as a time-for-space trade akin to a look-up table, for results kept in the reasoner’s memory only, with free retrieval, no break-even account, and an error cost that is propagated approximation and interference rather than staleness; for CLAIM-303 it is consistent but measures neither side of the cost; for CLAIM-295 its sample-versus-summary contrast is a clean instance of keeping more to answer a wider family of queries at a memory cost. |
| LIT-1075 | Deictic codes for the embodiment of cognition | Ballard, Hayhoe, Pook and Rao (1997), BBS 20(4). At about a third of a second, the time scale of a saccade, the article models fixation as a pointer that binds a variable in a behavioural program to a place in the world, so that what is bound is only the task-relevant feature, fetched just before use. In a block-copying task subjects looked back at the model about 1.3 times per block on average, the commonest cycle being two looks per block (one apparently for colour, one for position), and chose this slowest, lowest-memory strategy (about 3 s a block, against about 1.5 s for copying from memory) though told to be fast; separating model and copy by 70° cut looks to 1.0 per block, and one-colour patterns that could be chunked cut them to 0.7. The authors read this as working memory having a “carrying cost” that exceeds the cost of re-acquiring from the world, but that cost is inferred from the choices and never measured or modelled, and the evidence is one task with small samples (seven subjects for the main data, three for the colour-change experiment) and no statistics reported in the article. What is left in the world is unprocessed input, not a computed result, and the article argues against precomputation where the needed products cannot be foreseen. | Active — read in full 2026-10-15 (the target article, pp. 723–742; the open peer commentaries and the authors’ response, pp. 742–767, not read) from the typeset BBS PDF posted on Dana Ballard’s University of Texas page (www.cs.utexas.edu/~dana/bbs.pdf; the same file is on Rajesh Rao’s University of Washington page). Unpaywall and OpenAlex report the DOI closed. Worth reading as the founding statement of fixation as a pointer that binds a program variable to a place in the world at about a third of a second, and for the block-copying evidence that people re-look rather than memorize: about 1.3 looks at the model per block, the slowest lowest-memory strategy (about 3 s a block, against 1.5 s from memory) chosen most often under instructions to be fast, falling to 1.0 when looking is made dearer (70° separation) and 0.7 when patterns can be chunked. For QUESTION-095 it shows the trade stated as a cost comparison (“carrying cost” of working memory against on-line acquisition) that moves with both costs, but the cost is inferred from choice, not measured or modelled, what is left in the world is raw input rather than a computed result, and the authors argue against precomputing where needs are unpredictable, so it is the deferral policy opposite to the cockpit’s, not amortization; for TERM-055 it shows people choosing between internal and external sources by their relative retrieval cost, and moving retrieval later, not earlier; for CLAIM-298 it offers only an analogy (cue-driven matching, the same image read differently by task step) and no test of framing or judgement. |
| LIT-1076 | The intelligent use of space | Kirsh (1995), Artificial Intelligence 73. A classification, from video of cooks, assemblers and baggers, informal observation and the Tetris studies, of how people arrange things around them to cut the time and memory a task needs: arrangements that simplify choice (hide or highlight affordances, eliminate decisions, offload heuristic properties), arrangements that simplify perception (clustering, regions, marking, the vernier effect), and spatial dynamics that save internal computation (Tetris rotation, self-cueing). It frames the savings in search-complexity terms and links informational structuring to amortized computation in a footnote, but says the savings cannot in general be meaningfully quantified and offers no cost model, no measurements beyond the cited Tetris timings, and no case of an arrangement failing under a deadline. Its agents are experts arranging their own workspace in the medium and short term; long-term, pre-built environments are set aside. | Active — read in full 2026-10-15 from the publisher’s typeset article (Artificial Intelligence 73 (1995) 31–68), scanned and posted by the author on his UCSD pages (adrenaline.ucsd.edu/kirsh/Articles/Space/intelligent_useof_space.pdf), text taken by OCR, with the formulas checked against page images; his HTML pre-publication version on the same site was compared and differs in the footnotes and the Tetris timings. Worth reading as the standard classification of how people arrange their workspace to cut the time and memory a task needs: arrangements that simplify choice (hiding or highlighting affordances, eliminating decisions, heuristic cues), arrangements that simplify perception (clustering, regions, marking, the vernier effect) and spatial dynamics that save internal computation (Tetris rotation). It is drawn from observation and anecdote, with no measurements of its own beyond the Tetris timings it cites, and it declines to quantify the savings. For QUESTION-095 it shows that the link from informational structuring to amortized computation was drawn in 1995, in a footnote to Tarjan and in §3.4’s “set-up costs could be amortized over the whole task”, but as an analogy to self-adjusting structures, not to memoization. It also shows that its third category, just-in-time offloading, is not amortization at all. For CASE-085 and TERM-055 it supports the gap between presence and finding in time: memory for location is “regularly overestimated”, a meaningful pre-existing layout helps recall, and retrieval cost matters “in absolute time” even where it is a constant factor. It does not cover pre-built environments arranged by someone other than the retriever, or failures under a deadline. |
| LIT-1077 | Memory as a Computational Resource | Dasgupta and Gershman (2021), Trends in Cognitive Sciences 25(3). A short review arguing that memory is used to avoid redundant computation: results are stored and reused (memoization), either whole (“top-down”) or as solutions to subproblems (“bottom-up”), and amortized inference and deep Q-networks are “distributed” memo-tables that generalize to new inputs. It illustrates this in mental arithmetic, mental imagery, probabilistic inference and planning, mostly with response times that flatten with practice and with biases that a cache explains (reuse of earlier inferences, choices based on stale cached values, reuse of policy fragments). It has equations for its examples but no cost model: the space–time trade-off is named, and when caching pays is listed as an outstanding question. Memory is internal throughout; retrieval cost is mentioned once, and offloading or external stores not at all. | Active — read in full 2026-10-15 from the publisher’s typeset PDF of the version of record (TICS 25(3), 240–251), posted on the Gershman lab’s publication page (gershmanlab.com/pubs/DasguptaGershman21.pdf); Unpaywall lists only the publisher’s bronze copy. Worth reading as the short computer-science-to-cognition statement of memoization as a cognitive mechanism: Michie’s memo-table, the space–time trade-off, top-down (whole outputs, Logan) versus bottom-up (subproblem solutions, Kalman filter, Bellman backups) reuse, and amortized inference and deep Q-networks as “distributed” memo-tables that generalize to new inputs, illustrated in arithmetic, mental rotation, inference and planning. For QUESTION-095 it shows the first strategy with the store in the head and no cost model: its equations are illustrative, when caching pays is listed as an outstanding question, retrieval cost is mentioned once (production systems’ “onerous demands on storage and retrieval”) and search not at all, and nothing concerns external memory, offloading or tools. It does supply both failure terms qualitatively, staleness in the strict sense (cached values not yet updated after devaluation, with offline updating as the remedy) and bias from overlap between queries, and treats query frequency as what a capacity-limited learned cache spends its capacity on. “Amortization” here means amortized inference, not amortized analysis, and the paper does not say that retrieval is computation. |
| LIT-1078 | Hypothesis-Consistent Testing and Semantic Priming in the Anchoring Paradigm: A Selective Accessibility Model | Mussweiler and Strack (1999), JESP 35(2). States the selective accessibility model of anchoring: considering an anchor triggers hypothesis-consistent testing, which makes anchor-consistent knowledge accessible for the later judgement, with experiments on the mechanism. | Unavailable — filed 2026-10-15, not read. Registered from the exchange of 2026-10-14 on prompting, epistemic memory and retrieval (the workbench entry of 2026-10-15). No NOTE is filed, and it stays Unavailable until it is read, not on merit. Looked for a lawful full text on 2026-10-15 and found none that could be fetched: Unpaywall reports it closed (no open-access location, no repository copy) and OpenAlex likewise (closed, no repository full text); the Semantic Scholar API gives no openAccessPdf (status CLOSED, abstract elided by the publisher); Crossref lists only Elsevier text-mining links (api.elsevier.com, PII S0022103198913646) under Elsevier’s TDM user licence, which refused an unkeyed request, and ScienceDirect refused direct requests from this environment (403), so whether the article sits in Elsevier’s open archive could not be checked there, though Unpaywall and OpenAlex, which track that archive, record no free copy; London Business School Research Online, Mussweiler’s institutional repository, holds other anchoring papers of his (eprint 620, “Anchoring: accessibility as a cause of judgmental assimilation”) but not this one, and the London Business School publication page for it returns 404, with its Internet Archive capture (2023-09-26) unreachable from this environment; web searches found no copy on Mussweiler’s or Strack’s institutional pages (London, Cologne, Würzburg), no course-hosted copy and no preprint. A ResearchGate entry (publication 253136010) advertises a full text, but ResearchGate refused the request (403) and who uploaded it could not be confirmed, so it was not used. The 1997 companion paper (LIT-1064) and the later “Comparing is believing” review (European Review of Social Psychology 10, 1999) are different works. |
| LIT-1081 v2 | Visual search and stimulus similarity | Duncan and Humphreys (1989), Psychological Review 96(3). The standard source for the claim that search grows harder as the target resembles its distractors and easier as the distractors resemble one another. Registered unread, because that is the owner’s account of the LEGO bin in so many words: “if it’s too similar to the other bricks though, the difficulty to locate it manifests an imperative to set it aside”. | Unfetched — filed 2026-10-15, not read. Looked for a lawful full text on 2026-10-15. OpenAlex reports it closed (oa_status closed, no repository full text; locations are the APA DOI landing page and PubMed 2756067), and the publisher is APA PsycNET, which is subscription-only. The one copy found is a course copy on a University of Iowa psychology proseminar page (www2.psychology.uiowa.edu/faculty/hollingworth/prosem/Duncan_Humphreys_89_PR_VisualSearchStimulus.pdf). It now returns HTTP 404. The Internet Archive’s availability API reports a capture of it (2026-02-09), but web.archive.org reset every connection from this environment (three attempts by curl, one by a web-fetch tool). Web searches found no author-posted or institutional copy; the MRC Cognition and Brain Sciences Unit bibliography (mrc-cbu.cam.ac.uk/bibliography/articles/2331) gives the abstract only. So it is Unfetched because a copy was found and could not be fetched, not on merit. What was seen is that abstract: search difficulty “increases with increased similarity of targets to nontargets and decreased similarity between nontargets, producing a continuum of search efficiency”, with results that “support neither a distinction between serial and parallel search nor between search for features and conjunctions”. That is the similarity account CASE-094 and CLAIM-310 cite it for. Treisman and Gelade (LIT-1082, read) argue the opposite on the cause of serial search, so the record holds the disagreement without having read this side. |
| LIT-1082 | A feature-integration theory of attention | Treisman and Gelade (1980), Cognitive Psychology 12(1). In visual search for a coloured letter, a target that differs from all distractors by one feature (a colour, blue, or a shape, S, among brown Ts and green Xs) was found in a time almost independent of display size (positive slopes 3.8 ms/item for colour, 2.5 for shape, display sizes 1–30), while a target defined by a conjunction (green T among brown Ts and green Xs) gave linear slopes of 28.7 ms/item when present and 67.1 when absent, the roughly 1:2 ratio of serial self-terminating search; 13 sessions of practice did not make the conjunction pop out. Making the conjunction’s colours and shapes harder to tell apart nearly tripled the slope but kept it linear, and a single-feature target hard to discriminate (an ellipse of intermediate size) was slow but not item-by-item serial. Even in feature search, target-absent trials were slow and grew with display size (25.1 ms/item). The displays were at most 30–36 well-spaced items on a card, with unlimited viewing; the paper argues that conjunction structure, not similarity, makes search serial, and does not model similarity. | Active — read in full 2026-10-15 from a scan of the typeset article (Cognitive Psychology 12(1), pp. 97–136) hosted on a Princeton course page, COS 598B spring 2008 readings (www.cs.princeton.edu/courses/archive/spring08/cos598B/Readings/TreismanGelade1980.pdf); Unpaywall and OpenAlex (W2149095485) both report it closed, with no open-access location, so this is a course copy. Worth reading as the source of the feature/conjunction distinction in visual search: a target that alone has one colour or one distinctive shape part is found in time almost independent of display size (3.8 and 2.5 ms/item up to 30 items), while a target identified only by a combination of features its neighbours share is found by a serial self-terminating scan (28.7 ms/item present, 67.1 absent), which 13 sessions of practice did not change. For CLAIM-310 and CASE-095 it supports “search by what stands out” when what stands out is a single feature value no binmate shares, but it narrows them: in this paper serial search comes from conjunctions, not similarity; similarity on one dimension made search slow but not item-serial (ellipses of different sizes), and whether a shape pops out depends on which shape feature differs; it never tested ten-way colour search, similar colours, or piles of hundreds. The similarity effect the cases cite is not its finding. For CASE-094 a lone black brick among white ones is its easiest case, found almost regardless of numbers. For the claim on absence under a deadline it shows that rejecting absence costs about twice the per-item time of finding a present target in serial search, and costs a scan even when the present target pops out (25.1 vs 3.1 ms/item), with misses rising at large displays; it imposes no deadline. |
| LIT-1125 v2 | Memories of ‘The Rivalry’: Differences in How Fans of the Winning and Losing Teams Remember the Same Game | Talarico (2012). Fans of the winning side of the 2008 Lafayette–Lehigh game rehearsed it more and recalled it more accurately and vividly over time than the losers’ fans, while confidence was equal. | Unavailable — filed 2026-10-16, not read; a search for a lawful full text on 2026-10-29 reached none. Unpaywall lists it as closed, with no open location; the Wiley version is paywalled. A web search found no author or repository copy, only a 2012 Lafayette College event page for a talk by Talarico on the same study. Botzung et al. (2010, PMC3319676), an open study of Duke and UNC fans’ memories of one game, is the nearest open work; it is not filed. No NOTE is filed. It stays Unavailable until a copy is supplied. |
| LIT-1179 v2 | Behavior: The Control of Perception | Powers (1973). The book-length statement of perceptual control theory: behaviour as a hierarchy of negative-feedback systems, each controlling a perception against a reference set by the level above. | Unavailable — filed 2026-10-16; on 2026-10-16 a search for a lawful full text reached none, so it is not read and no NOTE is filed. The book is in copyright (1973 Aldine; 2005 second edition, Benchmark Publications) and no open copy was found. livingcontrolsystems.com, which publishes and hosts Powers’s papers, hosts only the original book jacket (intro_papers/bcp_jacket.pdf) and links the 2005 edition for sale; its intro_papers index has no chapter. The IAPCT page for the book (iapct.org/publications/books/behavior-the-control-of-perception/) gives the Preface only. pctresources.com holds reviews of both editions (Other/Reviews/BCP_book.pdf, BCP_web.pdf, BCP_Science.pdf), not the text. The Internet Archive has the 1981 printing only as a controlled-lending item (behaviorcontrolo0000powe), which this client cannot read, and the jacket. Web searches for a free chapter of either edition, and for the book’s treatment of memory and imagination, found none. The Preface alone is not a reading. It stays Unavailable until a copy is supplied. |
| LIT-1180 v2 | Feedback: Beyond Behaviorism | Powers (1973). A control system produces whatever output keeps a sensed result constant despite disturbances, so behaviour is the control of input, not output (reprint p. 77); in a hierarchy, the output of a higher-order system is not a muscle force but a reference level for lower systems (p. 77). Every loop the paper describes closes through the environment, and its top level is inherited intrinsic reference levels whose error drives reorganization (p. 73). | Active — read 2026-10-16 (NOTE-963) in full, from the authorized reprint in Living Control Systems (1989), pp. 61–78; worth reading as the shortest statement of perceptual control: a control system produces whatever output keeps a sensed result at its reference, so behaviour is the control of input, and above the lowest level a system’s output is a reference for the level below. Read it knowing that every loop it describes closes through the environment, down to muscle effort sensed kinesthetically; that it says nothing about imagination or memory, which Powers treated elsewhere; that its higher systems still sense quantities in the world and change only lower references, never the sensors; and that reorganization, driven by error in inherited intrinsic quantities, is one paragraph. It is a rival to CLAIM-341 and gives CLAIM-342 a scoring channel the learned hierarchy does not set. |
| LIT-1184 v2 | Factorized embedding of goal and uncertainty in the lateral prefrontal cortex guides stably flexible learning | Sung, Rigotti & Lee (2025). A reanalysis of 20 people’s fMRI from a two-stage Markov task with experimenter-cued goals and uncued block-wise transition noise. With a specific goal cued, transition noise did not lower choice optimality or consistency. Unfitted model-free agents showed a flexibility–stability trade-off (r = −0.964) that model-based agents and humans did not. Goals and, under specific goals only, transition uncertainty were weakly decodable from vlPFC, dlPFC and OFC. Nonlinear dichotomies decoded worse than goal and linear ones, and a goal decoder transferred across noise levels: the paper’s “factorized embedding”. Across subjects, goal separability and transfer in lateral PFC correlated with flexibility and performance; uncertainty separability correlated with nothing. | Active — read 2026-10-16 (NOTE-953) in full from the open-access article, with the supplementary tables for the shattering comparisons and the brain–behaviour correlations. Worth reading as a clear human-fMRI application of Bernardi et al.’s geometry tests (shattering dimensionality, cross-condition generalization) to goals and transition noise. Read it knowing that it is a reanalysis of 20 subjects with effects a few points above chance; that “uncertainty” is block-wise transition noise, coded only when a specific goal is cued; that it tested only whether the goal code is invariant to uncertainty, never the converse; that the task is action–outcome learning with experimenter-cued goals; and that goal-code invariance does not correlate with stability once corrected for multiple comparisons, although the main text says it does. Registered at the owner’s request (U9 of the exchange of 2026-10-16 on active inference). |
| LIT-1185 v2 | Map-like representations of emotion knowledge in hippocampal-prefrontal systems | Ma & Kragel (2026). In fMRI of 29 people passively watching 14 emotional films (Emo-FilM), decoders predicted an independent online sample’s consensus ratings of 13 emotion categories better from hippocampus than positions in a two-dimensional affective space (an MDS of rating co-occurrence), and vmPFC the reverse; posterior hippocampus carried finer distinctions. Brain patterns could also be decoded into a Tolman–Eichenbaum Machine that learned the emotion grid by random walk, hippocampus favouring its small-scale conjunctive layer and vmPFC its large-scale structural layer. Prediction–outcome correlations are about 0.04–0.10; no learning in the participants is measured. | Active — read 2026-10-16 (NOTE-950) in full, from the publisher’s open-access PDF, supplementary material not read; worth reading as a naturalistic-viewing fMRI dissociation in which hippocampus carries more about emotion categories and vmPFC more about positions in a two-dimensional affective space, lined up with the two layers of the Tolman–Eichenbaum Machine. Read it knowing that participants only watched; that the decoded emotions are an independent sample’s consensus ratings; that the effects are small; that the “transitions” are a model’s random-walk steps on a grid built from rating co-occurrence; and that no learning in people, and no effect of the map on later reading, is measured. The model that learns the map does so by acting. |
| LIT-1187 v2 | The construction of preference. | Slovic (1995), a review from an award address. Normatively equivalent elicitation procedures give systematically different preference orderings: in the Las Vegas study the two bets were chosen about equally often but the high-payoff bet was priced higher about 88% of the time, and Tversky, Slovic and Kahneman found 40–50% reversal rates of which 90% violated procedure invariance and 10% were intransitive. Scale compatibility and prominence explain the reversals. Slovic concludes that for important, complex and unfamiliar options preferences are “constructed on the spot” in elicitation. The evidence compares procedures; none follows a constructed preference into a later judgement. | Active — read 2026-10-16 (NOTE-949) in full, from a scan of the published article on a University of Florida course page, compared with the author’s manuscript in the University of Oregon’s ScholarsBank; worth reading as the compact review of preference reversals: normatively equivalent procedures give different orderings, about 90% of reversals violate procedure invariance rather than transitivity, and scale compatibility and prominence explain them. Read it knowing that its evidence compares procedures and never follows a constructed preference into a later judgement, so it shows the elicitation entering the response, not changing the judge; that construction is claimed for important, complex and unfamiliar options; and that the only remedy it reports is market repetition with feedback and penalties. |
| LIT-1188 v2 | A General Feedback Theory of Human Behavior: Part I | Powers, Clark & McFarland (1960). The first statement of the hierarchical feedback theory: each order perceives only the feedback signals of the order below and acts only by setting their references, which are replayed recordings of past perceptions. An imagination connection returns those references to the higher system’s perception, so that every goal set for lower systems can appear “immediately satisfied-in imagination” (p. 20), letting the system remember, plan an action “mentally” without performing it, hallucinate and dream, under a “mixing control” the authors leave open. An inherited N-system reorganizes the network, perceptual input functions included, at a rate proportional to error in intrinsic physiological variables. | Active — read 2026-10-16 (NOTE-954) in full, from the authorized reprint in Living Control Systems (1989), pp. 1–24, with Part II (LIT-1197, NOTE-959); worth reading as the first statement of perceptual control and the only one the record holds with its memory and imagination: references are replayed recordings of past perceptions, an imagination connection lets a higher system perceive the references it sends instead of their outcome, and a “mixing control” the authors cannot specify sets how much of perception is imagination. Read it knowing that the imagination loop changes nothing in the perceiver; that what does change perception is the N-system, which forms new input functions (“tentative definitions of new variables”, p. 14) at random under intrinsic, bodily error and keeps those whose behaviour lessens that error; that the hierarchy is “of the external-loop type” (p. 21); and that the sensory-deprivation remark is one uncited sentence. A rival to CLAIM-341 that contains its ingredients, and a scoring channel for QUESTION-121. |
| LIT-1189 v3 | Does Measuring Intent Change Behavior? | Morwitz, Johnson & Schmittlein (1993). Merely asking purchase intent once raised the later purchase rate; asking repeatedly made those with low intent less likely to buy. Both effects were weaker with prior experience of the product. | Unfetched — filed 2026-10-16 from the abstract; on 2026-10-16 a search for a lawful full text reached none, so it is not read and no NOTE is filed. OpenAlex lists it as green open access through an SSRN submitted version (abstract_id 1424067), but SSRN’s abstract page and its Delivery.cfm download both returned 403 “Content Blocked” to this client. Semantic Scholar reports it closed. The web archive could not be reached (connection resets, then 429). Web searches for author-posted or course copies (Morwitz at NYU Stern, Johnson at Columbia, Schmittlein; .edu PDFs; the University of Florida course page that holds Feldman and Lynch and Slovic) found citations and later papers only, not this one. ResearchGate returned 403. The summary and “Why this is here” are still the abstract’s, unchecked. It stays Unfetched until a copy is supplied; the SSRN copy is the likeliest route through a browser. Free copy: ssrn.com/abstract=1424067 (the SSRN submitted version). |
| LIT-1194 v2 | Neuronal dynamics, timing, and flow of sensory and choice-related information in auditory-prefrontal circuitry | Giarrocco & Averbeck (2026). In 37 sessions of simultaneous recordings from A1 and dorsolateral PFC of two macaques reporting by bar release whether two tones came from the same side, a lagged-regression (Granger-style) analysis of one-dimensional task projections found cue-side information flowing one way, A1 to PFC (onset about 120–147 ms after the cue), and choice information (release or hold) flowing first PFC to A1 (onset 315–365 ms after the target) and then A1 to PFC (352–419 ms), both directions peaking near 500 ms, close to the response. Within a trial and correlational; decision and motor response are not separated. | Active — read 2026-10-16 (NOTE-947) in full, from the publisher’s open-access PDF, supplementary material not read; worth reading as a within-trial measurement, from simultaneous A1 and dorsolateral PFC ensembles, that cue information flows A1 to PFC only while choice information flows PFC to A1 and then back. Read it knowing that the choice is the monkey’s bar release or hold and the flow peaks near the response, so decision and motor preparation are not separated (the authors say so); that the flow measure is a Granger-style lagged regression on one-dimensional projections, correlational and with other areas unrecorded; that the PFC-to-A1 choice signal arrives after the 300 ms target has ended; and that nothing crosses trials, so it shows no reshaping of A1’s later sensory coding. |
| LIT-1195 v2 | Attention or salience? | Parr & Friston (2019). A short review. Attention as gain control weights currently available sensory data by the estimated precision of the mapping from their causes; salience is the expected information gain of an action, typically a saccade, that would acquire new data. They are distinct, with gain failures giving hallucinations and salience failures neglect, but coupled: a saccade is a prediction that the foveated location will be sensed precisely, so covert attention to a place biases an overt saccade there, which the authors take to support the premotor theory. | Active — read 2026-10-16 (NOTE-956) in full, from the authors’ accepted manuscript on UCL Discovery; worth reading as the short statement, within active inference, that attentional gain (the estimated precision of currently available data) and salience (the expected information gain of an action that would acquire new data, usually a saccade) are distinct computations with distinct disorders, hallucination and neglect, and that they are coupled because a saccade is a prediction of high precision at the location foveated. Read it knowing that it is a five-page review with no new data or simulations; that it assigns covert spatial attention to gain by stipulation; that it never treats gain as an action; and that it ends by endorsing the premotor theory, so covert attention changes what is observed only by biasing overt saccades. |
| LIT-1196 v2 | Self-generated validity and other effects of measurement on belief, attitude, intention, and behavior | Feldman & Lynch (1988). A theoretical paper, with no data of its own. Belief, attitude or intention can be created by measurement when the construct does not already exist in long-term memory, and the created answer is stored and retrieved as an input to later answers and to behaviour. Even when it exists, question order and format decide which earlier answer is used: an earlier response is used for a later one if it is accessible and perceived as more diagnostic than other accessible inputs. Expertise, direct experience and trait involvement reduce the effects, because they supply judgements already formed and rehearsed. | Active — read 2026-10-16 (NOTE-957) in full, from a scan of the published article posted on a University of Florida course page; worth reading as the theoretical statement that a survey question can create the cognition it asks about, which is then stored and retrieved as an input to later answers and to behaviour, with accessibility and perceived diagnosticity deciding which earlier answer is used. Read it knowing that it reports no data of its own and calls itself speculative; that the created state it cares most about is one that drives behaviour; and that the moderators it names (expertise, direct experience, trait involvement) work by supplying judgements already formed and rehearsed, not evidence from outside them. |
| LIT-1197 | A General Feedback Theory of Human Behavior: Part II | Powers, Clark & McFarland (1960), Part II. Names six orders of perceptual control (spinal loops; elementary sensations; configurations; sequences; relationships; system concepts), demonstrated by reaction times that lengthen with order (third order about 0.3 s, fourth 0.4–0.45 s). The inherited N-system “merely alters the properties of a system, thus changing the transformations applied to existing information”, which the authors take to be insight; consciousness is defined as the state of a feedback function the N-system is acting on. Conflict between higher systems freezes a lower one at a virtual reference level, and one escape is switching a system “entirely to the imagination-connection”. | Active — read 2026-10-16 (NOTE-959) in full, from the authorized reprint in Living Control Systems (1989), pp. 25–45; worth reading as the second half of the first statement of perceptual control: six named orders of perception with tabletop demonstrations of their reaction times, the N-system (reorganization) worked out as the source of insight and of consciousness, and a theory of conflict in which switching a system “entirely to the imagination-connection” is one of the ego’s defences. Read it knowing that the orders are offered as the authors’ working definitions, the demonstrations are informal and the one experiment is “still in process”; that the N-system reshapes perception (“changing the transformations applied to existing information”) but is driven by intrinsic error, not by judgement; and that its intrinsic variables may include “mean error signal in the hierarchy”, an internal scoring channel. It bears on CLAIM-341, CLAIM-339 and QUESTION-121. |
| LIT-1199 v2 | Brain–computer interfaces and communication in paralysis: Extinction of goal directed thinking in completely paralysed patients? | Kübler & Birbaumer (2008). Across 35 patients trained with EEG BCIs, paralysis explained performance (η² = .40) only because of the seven in the completely locked-in state, none of whom, having entered training already in CLIS, reached yes/no communication; without them η² = .12, n.s. Evoked potentials showed hearing and some semantic processing intact. The authors propose that lost perception of the contingency between a response and its consequences, with reward blunted by artificially held homeostasis, extinguishes goal-directed thinking; predict that a contingency learned before CLIS and carried into it should prevent this; and say one CLIS patient who learns to communicate would disprove it. | Active — read 2026-10-16 (NOTE-948) in full, from the NIH author manuscript in PubMed Central. Worth reading as the source of the extinction hypothesis for the completely locked-in state: none of seven patients who began BCI training already in CLIS reached yes/no communication, and the authors propose that losing the perceived contingency between a response and its consequences, with reward blunted by artificially held homeostasis, extinguishes goal-directed thinking while passive sensory processing stays intact. Read it knowing that the hypothesis rests on seven heterogeneous patients and an analogy with curarized rats whose result failed to replicate, that it is stated as “might be responsible”, and that it predicts its own exception: a contingency learned before CLIS and carried into it should prevent extinction. |
| LIT-1209 v2 | Thinking, Fast and Slow | Kahneman (2011). A trade synthesis of the heuristics-and-biases programme around two modes of thought, a fast, automatic System 1 and a slow, effortful System 2, with prospect theory, overconfidence, and the split between the experiencing and the remembering self. | Unavailable — filed 2026-10-18, not read. A lawful full text was sought on 2026-10-29 and not reached. The book is in copyright and sold by Farrar, Straus and Giroux; no open-access edition, author posting or public-domain copy exists. The Internet Archive holds several items under the title, all uploads of unknown provenance, and they were not used. It stays Unavailable until a copy is supplied; a library or purchased copy would do. No NOTE is filed. |
| LIT-1262 v2 | The free-energy self: A predictive coding account of self-recognition | Apps & Tsakiris (2014; online 2013), Neurosci. Biobehav. Rev. 41:85–97. Self-recognition is recast under the free-energy principle: one’s own body is “the most likely to be ‘me’”, represented as a high-level, multimodal empirical prior learned from correlations between vision, touch, proprioception, corollary discharge and interoception, and “the notion that there is a ‘self’ is the most parsimonious and accurate explanation for sensory inputs”. The rubber hand, enfacement and rubber-voice illusions become one effect: synchronous stimulation leaves surprise that multimodal areas explain away by raising the probability that the other hand, face or voice is “me” and lowering it for one’s own. Context (a non-corporeal object, a misplaced hand, culture) acts as a prior. Three anatomical predictions are drawn: no unimodal area is self-specific; multimodal areas (TPJ, anterior insula, IPS, IFG) are engaged but not self-specific; predicted self-stimuli suppress activity. The authors concede there is “little direct evidence” for the account. | Active — read 2026-10-19 (NOTE-991) in full; worth reading as the first statement of the predictive-processing account of self-recognition and body ownership: one’s own face, voice and body are recognised as “me” because a high-level, multimodal prior makes them the most probable cause of correlated input across the senses, and the rubber hand, enfacement and “rubber voice” illusions are that prior being updated when synchronous multisensory stimulation leaves surprise to explain away. Read it as a theoretical review that, by its authors’ own account, rests on evidence that “neither largely supports nor refutes” it: no data, no model, no simulation, and predictions they call hard to falsify with standard fMRI. |
| LIT-1265 | Windows to the soul: Children and adults see the eyes as the location of the self | Starmans & Bloom (2012), Cognition 123(2):313–318. Three experiments with 4- and 5-year-olds (41–50 per study) and US adults on Mechanical Turk (51–53 per study) asked which of two pictured flies or snowflakes was “closer to” a cartoon character. Both groups chose positions near the eyes: over the lower body on a girl; at the chest on an alien whose eyes were there; and in front of the eyes over behind the head, or in front of the feet, on a profile. The authors conclude that people intuitively think of the self as located at or near the eyes, and, since preschoolers share it, that this is not learned from ideas about the brain. The judgement is about another, drawn person, and the step from “closer to her” to “the location of the self” is assumed, not tested. | Active — read in full 2026-10-20 (NOTE-995), from the published article on the authors’ lab site at Yale; worth reading as the best-known folk-judgement study of where a person is in their body, and for what such a study can and cannot show. Preschoolers and adults on Mechanical Turk chose which of two pictures had a fly, or a snowflake, closer to a cartoon character. They chose positions near the eyes: on a girl, near the eyes over the body; on an alien with its eyes on its chest, near the chest-eyes; on a profile, in front of the eyes over behind the head and over places the character could see. The authors conclude that children and adults intuitively place the self at or near the eyes. Read it as a judgement about where another, drawn, person is, with small samples, some slips in the statistics and an inference from “closer to Mary” to “the location of the self” that the design assumes. It bears on none of the rubber hand THEORYs. |
| LIT-1266 | Where exactly am I? Self-location judgements distribute between head and torso | Alsmith & Longo (2014), Consciousness and Cognition 24:70–74. Ten adults set a pointer, rotating in their sagittal plane, to point “directly at you”, by hand while blindfolded or by telling the experimenter when to stop it. Responses clustered on the upper face and the upper torso, and the starting position of the pointer decided which: from above, the face; from below, the torso. Vision and touch gave the same pattern. The authors conclude that the self is judged neither spread over the body nor at one point, and that the concept of the self as a spatial entity is “inherently ambiguous”. Small sample, 20.7% of trials lost, and a reanalysis whose key effect is misreported. | Active — read in full 2026-10-20 (NOTE-1008), from the published article on the second author’s lab site; worth reading as the first study that asks people to point at themselves on their own body, rather than on a drawing or a cartoon, and as a measure of self-location that is neither a questionnaire nor a walking-back drift. Ten adults, standing, set a pointer rotating in front of them so that it pointed “directly at you”, blindfolded by hand or by telling the experimenter when to stop it. Answers fell mostly on two regions, the upper face and the upper torso, and which one depended on where the pointer started: from above, the face; from below, the torso. Seeing and touching gave the same pattern, the touch version less sharply. Read it as a small pilot with a fifth of its trials lost to a camera fault, whose re-analysis without the lower torso misreports a non-significant effect as significant, and whose result may reflect where the pointer reached first as much as where people feel they are. It bears on THEORY-768’s self-location dimension; it bears on none of the rubber hand THEORYs. |
| LIT-1269 | Intuitions about personal identity: An empirical study | Nichols & Bruno (2010), Philosophical Psychology 23:293–312. Survey studies of University of Arizona introductory philosophy students on Williams’s (1970) personal-identity cases. A replication of Blok et al.’s brain-transplant vignette finds the person judged “still Jim” when memories survive and not when they are erased, in first- and third-person versions alike. A stripped-down version of Williams’s Pain case (a brain infection treated by a procedure that eliminates “distinctive mental states”) gets 68% (N = 50) saying “you will feel the pain” of later injections, and 72% (N = 54) in the third person, with no difference between the two. An open-ended abstract question draws psychological answers from over 70% (N = 53), and over 80% (N = 54) deny that a future person with none of their memories could be them. Given both and told they conflict, 64% (N = 45, p = 0.053) side with memory. The authors read the psychological intuitions as the more robust, and so as support for a psychological criterion “insofar as” intuitions should guide the theory. | Active — read 2026-10-20 (NOTE-996) in full, from the published PDF linked from the first author’s own publications page; worth reading as the first survey test of Williams’s two tellings of the body-exchange case and of his two explanations of why they diverge. Undergraduates mostly say a brain-transplant patient whose memories are erased is no longer the same person, and mostly say that a patient whose distinctive mental states are wiped will still feel the pain of later injections; neither result moves with first- against third-person telling. Asked abstractly, about 70% name psychological factors and more than 80% deny that someone without any of their memories could be them; made to choose, 64% side with memory (p = 0.053). Read it as evidence about how a concept of personal persistence responds to framing in one student population, not about what persistence is: the authors make the bearing conditional on intuitions informing the theory, test only necessity of psychology, and leave unseparated whether the Pain vignette asks about identity or about bare sentience. |
| LIT-1303 | Worth a Thousand Words: The Images of Copyright | Tushnet (2012). Copyright takes the written word as its model. Courts treat images either as transparent, so that their meaning is obvious and they merely copy reality, or as opaque, so that they cannot be analysed in words. The oscillation produces incoherent doctrine on copyrightability, substantial similarity, joint authorship of comics and fair use. Verbal overshadowing means that describing an image changes how it is seen, so the substantial-similarity test should be abandoned for a reproduction right against exact or near-exact copies. | Active — read 2026-10-20 (NOTE-1016) in full; worth reading as the standard statement that copyright law takes the written word as its model and so oscillates between treating images as transparent (their meaning obvious, a copy of reality) and as opaque (beyond analysis in words), of verbal overshadowing as a reason the substantial-similarity test cannot work, and of the proposal to replace that test with a reproduction right against exact or near-exact copies, leaving the rest to the derivative-works right. Read it knowing that it is US law as of 2011, that the cognitive findings are cited from the psychology literature rather than tested on legal factfinders, that the author flags the verbal-overshadowing effect as possibly smaller than first reported, and that the reform is stated without its implementation (n. 258). |
| LIT-1308 | Folk concepts of person and identity: A response to Nichols and Bruno | Berniūnas & Dranseika (2016), Philosophical Psychology 29:96–122, from the abstract only. A reply to Nichols & Bruno (2010). The authors argue that in transformation cases the folk do not use one concept of personal identity but separate concepts of “person”, “identity” and “individual”, so that identity can be ascribed to a post-transformation individual no longer categorized as a person. They report new survey evidence that the folk do not treat psychological continuity as necessary either for being categorized as a person or for identity over time, reading identity judgements off the use of proper names, as Nichols & Bruno did. They doubt current designs can tell qualitative from numerical identity judgements, and conclude for scepticism about folk intuitions in theorizing about personal identity. Samples, vignettes and effect sizes are not known here. | Unavailable — filed 2026-10-21 from the abstract only; no lawful full text was reached on 2026-10-21, so it is not read and no NOTE is filed. Where the reader looked: OpenAlex and Unpaywall (both “closed”, no repository copy); Semantic Scholar (openAccessPdf “CLOSED”); Europe PMC (no record); tandfonline.com (403 to this client, and paywalled); PhilPapers and PhilArchive (record BERFCO, 403 to this client, so whether PhilArchive holds an archived copy is unsettled); PhilPeople profiles of both authors (403); the second author’s personal site, www.dranseika.lt, linked from his Jagiellonian University (INCET) staff page (the publications page lists the paper with no link to a copy); the INCET staff page itself (no copy); the Lituanistika database record (content/101037: abstract and keywords, no file); web and Google Scholar searches for a PDF, a preprint or an accepted manuscript in the Vilnius University repository (none found; Scholar returned no version of the paper itself). ResearchGate was not relied on. A failure of access for PhilArchive, which should be rechecked from a client it does not refuse, and of availability elsewhere. It stays Unavailable until a copy is supplied, not on merit. |
| LIT-1358 v2 | Illusionism Big and Small: Some Options for Explaining Consciousness | Graziano (2024), eNeuro 11(10). A claim to be conscious requires an internal model encoding it, and brain models are schematic, so what is reported is a caricature of whatever the model tracks. Six hypotheses for that target: a magical essence (rejected, for lacking “Arrow B”, a route from feeling to encoding), nothing (hard illusionism, judged improbable), integrated information (illusionist IIT), the global workspace modelling itself, attention modelled by an attention schema, and “deep illusionism”, the deep selective processing the last three share. No new data. | Active — read in full 2026-10-22 (NOTE-1071) from the publisher’s PDF (eNeuro, CC BY 4.0); worth reading as the shortest statement of the attention schema theory as an illusionist framework. Its argument: every claim to be conscious needs an internal model, every model is schematic, so the reported “intangible essence” is a caricature of a real process. It adds “Arrow B”, the demand that a theory explain how a feeling gets encoded for report, and it sets IIT and the global workspace as illusionist variants converging on “deep selective processing”. An essay with no new data and a polemical tone. Its “small” illusionism keeps the process real but still denies the phenomenal essence, so it is Frankish’s strong illusionism with a named target, not a weak illusionism. |
| LIT-1360 v2 | Real patterns, the predictive mind, and the cognitive construction of the manifest image | Gładziejewski (2025), Synthese 206:225. Dennett’s compression criterion for real patterns is carried through minimum description length, Bayesian inference and variational free energy, so learning a generative model is real-pattern extraction from sensory data. The manifest image is then a structural description of the sensory signal, equally veridical for an envatted twin with matched input. Applied to selves: a self-model compresses self-generated regularities by positing one common cause, real as a pattern, not as a substance, and the realism dispute is over criteria. Applied to objects: compression-based composition (Bird 2023) favours cats over arbitrary sums. | Active — read in full 2026-10-22 (NOTE-1056) from the publisher’s open-access PDF (Springer, CC BY-NC-ND 4.0); worth reading as a clear derivation of Dennett’s real patterns into predictive processing, through minimum description length and the complexity–accuracy split of free energy, and as a short diagnosis of the realism dispute about selves: a self-model compresses self-generated regularities by positing one common cause, so the self is real as a pattern and not as a substance, and the dispute is over criteria. The structural, Kantian realism it draws (the manifest image is veridical in a matched vat) needs manifest content to be wholly structural, which it assumes. It says nothing about consciousness, and it does not say a self-model exaggerates the self’s unity. |
| LIT-1362 v2 | Higher-Order Theories of Consciousness (Stanford Encyclopedia of Philosophy) | Carruthers & Gennaro (2001; rev. 2026), SEP. Most kinds of mental state also occur unconsciously (blindsight, absent-minded driving, the agnosic D.F.), and higher-order theories say a state is phenomenally conscious when it is, or is disposed to be, the object of a suitable non-inferential higher-order representation. The entry sets out inner sense, actualist higher-order thought (Rosenthal), dispositionalist dual-content theory (Carruthers) and self-representational and hybrid views (Kriegel, Gennaro, Van Gulick, Picciuto). It gives the objections (targetless representations, the rock, animals and infants, sufficiency, depersonalization) and the evidence and counter-evidence that prefrontal cortex realizes the higher-order states, and concludes no verdict. | Active — read in full 2026-10-22 (NOTE-1066) from the live entry, substantive revision of 9 January 2026; worth reading as the map of higher-order theories, which explain what makes a mental state conscious rather than unconscious by a higher-order representation of it: inner sense, actualist and dispositionalist higher-order thought, self-representational and hybrid views, the standing objections (targetless states, the rock, animals and infants, sufficiency) and the prefrontal evidence. Read it knowing that it argues for no verdict, that it is maintained by Gennaro, a higher-order theorist, while Carruthers and Lycan are reported to have given up their versions, and that its empirical section reports results to about 2023 and not the 2025 adversarial-collaboration findings. |
| LIT-1382 v2 | Event perception: A mind-brain perspective | Zacks et al. (2007). Event segmentation theory: perceptual systems continuously predict what will happen next, a transient rise in prediction error is perceived as an event boundary, and boundaries update the event models that guide attention and memory. | Active — read in full 2026-10-22 (NOTE-1084) from the author manuscript in PubMed Central (PMC2852534, NIH public access); worth holding as the statement of event segmentation theory: transient rises in perceptual prediction error are perceived as event boundaries and reset working-memory event models, a domain-general mechanism of attention and memory control. A review with mostly correlational evidence; prediction error is not measured in people and the predicted anterior cingulate response at boundaries was not found. A theory of a passive observer of others’ activity; it does not mention doorways or one’s own intentions. |
| LIT-1384 v2 | Walking through Doorways Causes Forgetting: Further Explorations | Radvansky, Krawietz and Tamplin (2011). The location-updating effect appeared with smaller displays and in a real environment, and, against an encoding-specificity account, memory did not improve when people returned to the room where the objects were first encoded; the results are read through the event horizon model. | Active — read in full 2026-10-22 (NOTE-1082) from the publisher’s PDF as posted on the first author’s laboratory page at Notre Dame; worth holding for the return-to-the-room result and the authors’ account of it: returning to the room where objects were picked up did not help, two new rooms were worse than one, and the authors explain the effect as retrieval interference between event models that both hold the carried object, not as loss of a context or of activation. The real-room experiment changes the task and its effect is p = .02 to .06; the plain shift effect in the return experiment is marginal on errors. No goal or intention is manipulated, so the reading that perspective is a goal-sensitive relation rather than a location is not tested here. |
| LIT-1385 v2 | Walking through doorways causes forgetting: Event structure or updating disruption? | Pettijohn and Radvansky (2016). Two virtual-navigation experiments rejected a retention-interval account (travel times were similar) and a temporary-disruption account (a 3-second delay did not change the pattern), and the authors attribute the effect to interference at retrieval among multiple event models that share elements. | Active — read in full 2026-10-22 (NOTE-1081) from the publisher’s PDF as posted on the second author’s laboratory page at Notre Dame; worth holding as the paper that rules out a longer retention interval and a passing cost of updating, and sets the doorway effect’s mechanism as competition between the two event models that hold the carried object, against the alternative that a prior event model has merely dropped out of working memory. In both experiments the effect was on the carried object only. Same lab, same paradigm, about 6 trials per cell for the critical condition. |
| LIT-1387 v2 | Doorways do not always cause forgetting: a multimodal investigation | McFadyen et al. (2021). Four experiments, in immersive virtual reality with and without a working-memory load, in passive video and in real movement, found no significant effect of doorways on forgetting; under working-memory load, false alarms to mismatched probes rose after a doorway. The authors question the effect’s robustness to small changes of paradigm. | Active — read in full 2026-10-22 (NOTE-1079) from the publisher’s text in PubMed Central (PMC7938580, CC BY 4.0); worth holding as the record’s only test of the doorway effect outside the originating lab: in passive head-mounted VR doorways did not lower hits (Bayes factors 3–6 for the null), under a counting load they raised false alarms to recombined probes, and a curtained hallway, watched or walked, did not affect memory for a 16-item arrangement. Conceptual rather than direct replications: movement was passive and the hallway task was different. Its one positive finding fits the interference account. |
| LIT-1392 v2 | Walking through doorways causes forgetting: Situation models and experienced space | Radvansky and Copeland (2006). In a desktop virtual environment, people carried objects from room to room; with distance held equal, walking through a doorway rather than across a large room raised errors and slowed recognition for the object being carried, but not for the object just set down. The authors read it as the updating of situation models at a spatial boundary. | Active — read in full 2026-10-22 (NOTE-1083) from the publisher’s free PDF at Springer (bronze open access); worth holding as the origin of the doorway effect and for what it actually found: a doorway, against walking as far within a room, raised errors and slowed answers for the object being carried, and did not affect the object just set down. The decisive experiment has 5 no-shift trials per condition, and the rooms on either side of a doorway also differed in size and wall pattern. Nothing in it concerns intentions; the explanation offered, the effort of updating a situation model, was later dropped by the same lab for interference. |
| LIT-1399 v2 | Twenty-Five Years of Hidden Profiles in Group Decision Making: A Meta-Analysis | Lu, Yuan & McLeod (2012). A meta-analysis of 65 hidden-profile studies (101 effects, 3,189 groups). Groups mention common information about two standard deviations more than unique information, and hidden-profile groups are eight times less likely to find the correct solution than groups with full information. Pooling of unique information predicts decision quality (coverage more than discussion focus); communication medium affects neither pooling nor quality; group size, information load, the share of unique information, task demonstrability and profile strength moderate. | Unavailable — filed 2026-10-23, not read. A lawful full text was sought on 2026-10-29 and not found. OpenAlex reports it closed, with only the SAGE page and the PubMed record (PMID 21896790) as locations and no repository copy. Europe PMC holds the record with no full text. The SAGE page is listed as restricted access. A web search for an author’s copy or a deposit found none. It stays Unavailable until a copy is supplied. The figures in the summary are the abstract’s. No NOTE is filed. |
| LIT-1403 | Knowing others’ preferences degrades the quality of group decisions | Mojzisch & Schulz-Hardt (2010), JPSP 98(5): 794–808. By its abstract: in four experiments, learning the other members’ preferences at the start of a discussion made hidden profiles less likely to be solved, through less attention to the information exchanged, not through bolstering or information load; it held even when another member favoured the correct option, and Experiment 4 replicated it in face-to-face three-person groups. Unread: no lawful copy reached. | Unavailable — no lawful copy reached, 2026-10-23; failed on availability as far as this client could tell. The article is closed (APA; Unpaywall and OpenAlex: closed, no repository full text). Göttingen Research Online, the one repository OpenAlex lists, sits behind a bot check, so whether it holds a file is unknown; the Hildesheim repository holds the record with no file; the authors’ Göttingen publications page has no file; a web search found no author-posted copy. Only the abstract was read (PubMed 20438225). Registered for QUESTION-089 and CASE-108 from the owner’s exchange of 2026-10-23, “Framework Overview Request”. No NOTE, CASE or THEORY is filed. It returns to reading when a copy is supplied, not on merit. |
| LIT-1409 v2 | Group decision making in hidden profile situations: Dissent as a facilitator for decision quality | Schulz-Hardt, Brodbeck, Mojzisch, Kerschreiter & Frey (2006), JPSP 91(6). By its abstract: in 135 three-person groups solving a personnel-selection hidden profile, groups whose members agreed on a wrong option before discussion “hardly ever solved” it, while groups with prediscussion dissent did so significantly more often, even when no member favoured the correct option. The size of the dissent did not matter; the benefit ran mainly through greater discussion intensity. Unread: no lawful copy reached. | Unfetched — no lawful copy reached, 2026-10-23; failed on access. A lawful copy exists: the authors’ working paper (LMU Munich, Wirtschaftspsychologie working paper WOP 2006/3) was posted on the department’s site and is captured by the Internet Archive, but the live file now returns 404 and this client cannot reach web.archive.org. The journal version is closed (APA). mediaTUM, which Unpaywall and OpenAlex list as green open access, holds metadata and the abstract only; Göttingen Research Online sits behind a bot check; the Hildesheim repository and the Aston portal hold no file; ResearchGate returned 403. Only the abstract was read (PubMed 17144766, the same text on mediaTUM). Registered for QUESTION-089 and CASE-108 from the owner’s exchange of 2026-10-23, “Framework Overview Request”. No NOTE, CASE or THEORY is filed. It returns to reading when a copy is supplied, not on merit. Free copy: the Wayback Machine’s capture of 2023-11-18 of psy.lmu.de/wirtschaftspsychologie/forschung/working_paper/wop_2006_3.pdf. |
| LIT-1410 v3 | Telling More Than We Can Know: Verbal Reports on Mental Processes | Nisbett & Wilson (1977). Reviews evidence that people may have little or no direct introspective access to higher-order cognitive processes: they can be unaware of a stimulus that influenced a response, of the response, or of the influence. Reports on one’s own processes rest on a priori, implicit causal theories about what plausibly caused what, so they are accurate when the influential stimulus is salient and a plausible cause, and not otherwise. Its test is that observers given only a description of the situation predict the same effects the subjects report. | Active — read 2026-10-29 (NOTE-1412) in full, from a scan of the published article in an MIT course-readings folder (https://web.mit.edu/curhan/www/docs/Articles/15341_Readings/Social_Cognition/Nisbett_Wilson_1977_Telling_more_than_we_can_know.pdf); worth reading as the review, with eight small studies of its own, that people’s reports of why they responded as they did miss effective stimuli and credit ineffective ones, and match what observers predict from a description alone, so that they come from causal theories, not from access to the process. Read it knowing that its situations were chosen to produce error, that several of its own studies are reported briefly or unpublished, that it is about the influence of stimuli on responses, not about attributions of agency, and that it shows only that such access as exists does no better than an observer, not that there is none. |
| LIT-1417 v2 | Failure to Detect Mismatches Between Intention and Outcome in a Simple Decision Task | Johansson, Hall, Sikström & Olsson (2005). 120 participants chose the more attractive of two female faces; on three of 15 trials a card trick handed them the face they had rejected. Only 13% of the 354 swapped trials were detected at the time, at most 26% counting every later form of detection, whatever the faces’ similarity or the time allowed. Reasons given for the rejected face matched reasons for real choices in length, emotionality, specificity and certainty, and some named features only the rejected face had. The authors call this choice blindness. | Active — read 2026-10-29 (NOTE-1398) in full, with its Supporting Online Material, from the authors’ posting of the final published version on the Lund University Cognitive Science site; worth reading as the experiment that catches a person giving reasons for a choice they did not make: 13% of 354 covertly swapped outcomes were noticed at the time, and the reasons given for the swapped face matched reasons for real choices on every measure but one. Read it knowing that it is one low-stakes task (attractiveness of unfamiliar faces), that detection was counted from spontaneous report by design, that the report comparison is a null on eight coded variables, and that it does not show ordinary reasons to be confabulated, which the authors leave open. |
| LIT-1446 | The dynamic architecture of emotion: Evidence for the component process model | Scherer (2009), Cognition and Emotion 23(7):1307–1351. The component process model’s evidence review. Emotion is “an emergent, dynamic process based on an individual’s subjective appraisal of significant events”, and the paper gives special emphasis to the sequence of appraisal checks and the synchronisation of response systems. Unread: the model is read in the companion paper LIT-1451. | Unavailable — registered 2026-10-23 from Crossref, not read: a lawful copy exists but this client could not reach it. The University of Geneva’s Archive ouverte holds the published version (unige:97810), with access restricted to members of the UNIGE/HUG community. The Taylor & Francis article is paywalled. The Geneva E3 Lab’s publications page posts many of the group’s 2009 papers but not this one, and a web search found no other author or course posting. Its argument is read first-hand in Scherer’s companion paper of the same year, “Emotions are emergent processes: they require a dynamic computational architecture” (LIT-1451), which restates the component process model and cites this paper, then in press, for the evidence. Filed because the record’s proposal needs a measurable analogue of potential, and this is the review usually cited for response synchronization. It stays Unavailable until a copy is supplied, not on merit. |
| LIT-1447 | EMA: A process model of appraisal dynamics | Marsella & Gratch (2009), Cognitive Systems Research 10(1):70–90. EMA (EMotion and Adaptation), after Smith and Lazarus. Appraisal is one fast, parallel, automatic process over the agent’s causal interpretation, a decision-theoretic plan representation of its beliefs, desires, intentions and plans. Emotional dynamics come from the time course of the perception, inference and action that change that interpretation, so no multi-level appraisal is needed. Each proposition gets an appraisal frame. Frames compete for coping, and the most recently accessed one with the highest mood-adjusted intensity wins. Coping changes the world or the agent’s beliefs, desires and intentions, which drives reappraisal. Illustrated with an actor’s reaction to a pigeon. | Active — read 2026-10-23 (NOTE-1100): in full, from the published version on a University of Kentucky course reading list, checked against the authors’ manuscript on Gratch’s USC page. Worth reading as the worked computational case of appraisal as a continuous process over the agent’s causal interpretation, with an explicit arbitration rule. Appraisal is a single, fast, parallel mapping from that interpretation to appraisal variables. Its dynamics come from the slower perception, inference and action that change the interpretation, and from coping, which alters beliefs, desires and intentions and so drives reappraisal. Appraisal frames, one per proposition, compete for the coping response. The winner is the most recently accessed frame with the highest mood-adjusted intensity, and intensity is a decision-theoretic product of likelihood and utility. That is arbitration on a common currency among concerns, not a negotiation among perspectives. |
| LIT-1448 | Back to Bentham? Explorations of Experienced Utility | Kahneman, Wakker & Sarin (1997), Quarterly Journal of Economics 112(2):375–405. “Utility” has two meanings: decision utility, an outcome’s weight in choice, and experienced utility, its hedonic quality. Experienced utility is instant (the pleasure or distress of the moment, reported in real time) or remembered (a retrospective evaluation). Remembered utility follows the Peak-End average and neglects duration. Choices follow remembered utility, so subjects repeat a longer cold-water trial that dominates the shorter one in pain. Axioms on ordinal instant utility with a neutral zero make total utility the time integral of a value transform. | Active — read 2026-10-23 (NOTE-1105): in full, from the published article as deposited in Erasmus University Rotterdam’s repository (RePub, hdl 1765/23011). Distinguishes decision utility (an outcome’s weight in choice) from experienced utility (hedonic quality). Experienced utility divides into instant utility, reported in real time, and remembered utility, a retrospective evaluation. Remembered utility follows a Peak-End rule and neglects duration. Choices follow remembered utility, so people choose dominated, longer painful episodes. Axioms on ordinal instant utility with a neutral zero yield total utility as a time integral. Filed for the record’s three-layer operationalization of felt valence (CLAIM-440). |
| LIT-1451 | Emotions are emergent processes: they require a dynamic computational architecture | Scherer (2009), Phil. Trans. R. Soc. B 364:3459–3474. Emotion is a bounded episode of synchronization among organismic subsystems, driven by a fixed sequence of appraisal checks (relevance, implications, coping, normative significance) and aimed at control precedence. The feeling component monitors the process by integrating a representation of all components and their coherence, and the degree of synchronization is hypothesised to generate conscious experience. Proposes nonlinear dynamic modelling: emotions as attractor basins of coupled oscillators, with hysteresis and cusp-catastrophe jumps. | Active — read 2026-10-23 (NOTE-1113): in full, from the PubMed Central text (PMC2781886). Worth reading as Scherer’s own compact statement of the component process model, the same year as the evidence review LIT-1446, which it cites as in press. Emotion is a bounded episode of component synchronization, driven by sequential appraisal checks and aimed at control precedence. The feeling component monitors the process by representing the other components and their coherence. The degree of synchronization is hypothesised to generate conscious experience. Emotions are attractor basins of coupled oscillators, with hysteresis and catastrophe-surface jumps. It stands in for the unread review. |
| LIT-1457 | The Emotion Process: Event Appraisal and Component Differentiation | Scherer & Moors (2019), Annual Review of Psychology 70:719–745. A review of experimental evidence on the emotion process, organised by five appraisal criteria (novelty, intrinsic and extrinsic valence, agency, control or power, fairness) and three paths from appraisal: to action tendencies, to physiological responses and to facial, vocal and gestural expression. Appraisal criteria have more stable response correlates than emotion categories do. Proposes that coherence among components above a threshold marks an episode, and above a second threshold awareness. | Active — read 2026-10-23 (NOTE-1106): in full, from the PDF Annual Reviews serves to guest readers. Worth reading as the latest review of the component process model’s evidence, organised by appraisal criterion rather than by emotion category. Its paths run from appraisal to action tendencies, to physiological responses and to expression. It separates intrinsic valence from goal congruence and reports the few studies that pull them apart: they have partly different physiological signatures and combine multiplicatively. It ends with a two-threshold proposal: coherence among components above one threshold starts an episode, and above a second it becomes aware. It does not describe appraisal as a negotiation among perspectives. |
| LIT-1458 | Levels of valence | Shuman, Sander & Scherer (2013), Frontiers in Psychology 4:261. Valence exists at two levels. Micro-valences are the valenced outcomes of separate appraisal checks (pleasantness, goal conduciveness, power, self-congruence, moral goodness), and they can conflict, which explains mixed feelings. Macro-valence is one-dimensional, always present, and the common currency for choice. It is integrated from micro-valences and other inputs by rules ranging from weighted sums to lexicographic and elimination-by-aspects strategies. Intrinsic pleasantness and goal conduciveness are dissociable, citing Berridge. | Active — read 2026-10-23 (NOTE-1108): in full, from the publisher’s open-access PDF (Frontiers, CC licence). Worth reading as the component process model’s own statement that valence is several things. Five appraisal checks are each valenced (pleasantness, goal conduciveness, power, self-congruence, moral goodness), and these “micro-valences” are integrated, by weighted sums or by lexicographic and elimination rules, into a one-dimensional “macro-valence” that serves as the common currency for choice. It separates intrinsic pleasantness from goal conduciveness and grounds the split in Berridge. It bears directly on the objection that the record’s potential account tracks ‘wanting’, not ‘liking’ (CLAIM-438). |
| LIT-1460 | Semiorders and a Theory of Utility Discrimination | Luce (1956), Econometrica 24(2):178–191. The paper that introduced semiorders: preference relations whose indifference need not be transitive, because differences below a threshold of discrimination go unnoticed. By its abstract, the class it axiomatizes is “substantially equivalent” to a utility theory with just-noticeable-difference functions. Unread: no lawful copy reached. | Unavailable — registered 2026-10-23 from Crossref, not read: no lawful full text could be reached. OpenAlex and Semantic Scholar list no open copy; JSTOR and the Econometric Society give the abstract only (the issue is not in JSTOR’s free early-journal content, which stops well before 1956); a web search found no author, course or repository posting. Filed because it is the paper that introduced the semiorder, the structure the record’s correction to CLAIM-436 turns on: a relation with a discrimination threshold whose indifference is not transitive. It stays Unavailable until a copy is supplied, not on merit. |
| LIT-1464 | The Nature and Dynamics of Relevance and Valence Appraisals: Theoretical Advances and Recent Evidence | Scherer (2013), Emotion Review 5(2):150–162. Discusses three open issues in appraisal theory: varieties of relevance detection, varieties of valence appraisal, and the sequential-cumulative effects of appraisal results, with an update on research into the sequence of appraisal and emphasis on its nonverbal signatures. Unread: the varieties of valence are read in Shuman, Sander and Scherer 2013 (LIT-1458). | Unavailable — registered 2026-10-23 from Crossref, not read: no lawful full text could be reached. OpenAlex lists no open copy. The SAGE article page and PDF refused this client behind a bot check, and the journal is subscription-only. The University of Geneva’s Archive ouverte holds two other papers from the same issue with public access (the issue’s introduction by Moors, Ellsworth, Scherer and Frijda, and Scherer and Ellsworth’s reply), but no record of this one was found, and the Geneva E3 Lab’s publications page does not post it. A web search found no author or course posting. Its subject, the varieties of valence appraisal, is read first-hand in Shuman, Sander and Scherer’s “Levels of valence” (LIT-1458), published two months later. Filed because the record’s potential account needs the component process model’s distinction between kinds of valence. It stays Unavailable until a copy is supplied, not on merit. |
| LIT-1471 | An Appraisal-Driven Componential Approach to the Emotional Brain | Sander, Grandjean & Scherer (2018), Emotion Review 10:219–231. A perspective article. Major emotion theories agree that emotion has five components, so the emotional brain should be studied as five functionally defined networks: elicitation (appraisal), expression, autonomic reaction, action tendency and feeling. Appraisal drives the changes and the synchronization in the other four. Synchronization is proposed to be neuronal phase or cross-frequency coupling within and between components, and feeling to emerge when it passes a threshold. No measure is given and no data are reported. | Active — read 2026-10-23 (NOTE-1103): in full, from the published version deposited in the Archive ouverte UNIGE (unige:109591). Worth reading as the Geneva group’s programme statement for taking the component process model into the brain. It splits the emotional brain into five functionally defined networks (elicitation, expression, autonomic reaction, action tendency, feeling). It proposes that component synchronization is neuronal phase or cross-frequency synchronization, within and between components, and repeats Scherer’s hypothesis that feeling arises when the degree of synchronization passes a threshold. It names control precedence as priority among action tendencies. It gives no measure of synchronization and reports no data of its own. |
| LIT-1472 v3 | Intransitivity of preferences | Tversky (1969), Psychological Review 76(1):31–48. Shows experimentally that people’s choices can cycle predictably when options vary on two dimensions and differences on the more important one are small. The model is the lexicographic semiorder: the first dimension decides only when its difference exceeds a threshold, and otherwise the next one does. Unread: a lawful course copy exists but could not be fetched. | Unfetched — registered 2026-10-23 from Crossref, not read: a lawful copy exists but this client could not fetch it (“failed on access, not availability”, in ADR-041’s words). The paper is posted for teaching on a UC San Diego faculty page (pages.ucsd.edu/~mckenzie/Tversky1969PsychReview.pdf). This client refused the TLS connection, the server’s certificate chaining to a root not in its trust store; the web fetcher got a 503; the Internet Archive holds no snapshot. OpenAlex and Semantic Scholar list it as closed, APA’s copy is paywalled, a Central European University course copy sits behind a bot check, and a whole-book PDF of Tversky’s selected writings on another UCSD page was not used, its lawfulness being unclear. Filed because its lexicographic semiorder is the structure the correction to CLAIM-436 names. It stays Unfetched until a copy is supplied, not on merit. |
| LIT-1498 v3 | Behavior as information: “If I avoid, then there must be a danger” | Gangemi, Mancini & van den Hout (2012). Anxious participants (31 with obsessive-compulsive disorder, 22 with panic, 17 with social phobia) and 31 controls rated danger in scripts that varied objective danger and the presence of a safety behaviour. Anxious participants inferred danger from the safety behaviour as well as from objective information, while controls were influenced mainly by objective danger; the effect was disorder-specific for social phobia and OCD. Taking one’s own safety behaviour as evidence of danger may maintain anxiety. | Unavailable — filed 2026-10-24 from the owner’s exchange of 2026-10-23/24, shared at https://chatgpt.com/share/6adc1ece-bb28-83e9-a8f6-cadf0f8b7113, where it is cited at turn U15; a reading was attempted on 2026-10-24 and failed on availability: no lawful full text was found. The article is closed at Elsevier. Unpaywall, OpenAlex and Semantic Scholar list no open copy, Europe PMC and PubMed Central hold none, the University of Messina’s IRIS record (hdl 11570/2424621) has no file, and the Utrecht University research portal refused this client (403). Web searches found no preprint or accepted manuscript. One copy was found and not used: the publisher’s typeset PDF on the website of the Associazione di Psicologia Cognitiva, Rome, the first two authors’ institute. Its own cover sheet says it is an “Author’s personal copy” that may not be posted to “personal, institutional or third party websites”, so it is not a lawful open copy. Only the cover sheet was seen, and the file was deleted unread. It stays Unavailable until a copy is supplied, through library access or a request to the authors. No NOTE is filed. |
| LIT-1541 v2 | On the Biology of a Large Language Model | Lindsey et al. (2025), Transformer Circuits Thread. Ten case studies of Claude 3.5 Haiku with the attribution graphs of LIT-1552, each validated by perturbing features in the model: two-hop factual recall through an intermediate “Texas” representation, rhymes planned before a line is written, addition features reused across contexts, a default “can’t answer” circuit inhibited by known entities whose misfire yields hallucination, unfaithful chains of thought, multilingual circuits, medical diagnosis, refusals, a jailbreak and a model with a hidden goal. The authors present each as an existence claim about specific prompts. | Active — read 2026-10-26 (NOTE-1181) in full, from the Transformer Circuits page; worth reading for ten case studies that show, on single prompts and with interventions, a two-hop intermediate in Claude 3.5 Haiku (THEORY-984), a rhyme planned on the newline before a line is written (THEORY-1017), a jailbreak whose target word is not represented before it is written (THEORY-1018), a trained-in goal represented in every Assistant context of a model organism (THEORY-976), and a default “can’t answer” circuit that known entities inhibit. Read it knowing that the authors find a satisfying graph for about a quarter of the prompts they try, that the graphs shown are curated and of a replacement model (LIT-1552), that several cruxes run through attention the graphs cannot see, and that its general statements about the model (THEORY-982) go beyond what the case studies show. |
| LIT-1634 v2 | The structure of ill structured problems | Simon (1973), Artificial Intelligence 4. There is no sharp boundary between well and ill structured problems: the conditions for a well structured one relate a problem to the power of a bounded solver, read as practicable computation, and on that reading even theorem proving and chess fail some of them. Problems from the world “become WSPs only in the process of being prepared for the problem solvers” (p. 186). House design, a battleship’s design by an organization and a robot are handled by alternating a solver on a locally well structured problem with a recognition system that keeps modifying the problem space: “well structured in the small, but ill structured in the large”. Simon suggests that only the size of the knowledge base separates the two, so no qualitatively new mechanisms are needed. | Active — read in full 2026-10-27 (NOTE-1232), from a scan of the published article; worth reading as the classic argument that well-structuredness is relative to a bounded solver and that problems from the world become well structured only by being prepared, and for its model of design as an alternation between local problem solving and retrieval that modifies the problem space. Its conclusion that no new mechanisms are needed is a suggestion, not a result. |
| LIT-1734 | Mental Hoop Diaries: Emotional Memories of a College Basketball Game in Rival Fans | Botzung et al. (2010). Twenty-three male Duke and UNC fans watched an archived 2001 Duke–UNC game three times in the lab, then under fMRI recalled whether each of 64 shots went in. The same shot was positive for one side and negative for the other, so valence was set by allegiance with the stimulus held fixed. Shots good for one’s own team were recalled more accurately and confidently and rated more intense. Emotional intensity, not valence, modulated amygdala, hippocampus, striatum and cortical midline regions; positive shots added dorsal fronto-parietal and sensorimotor activity. Fan groups did not differ in accuracy or intensity. | Active — read in full 2026-10-29 (NOTE-1313) from the NIH author manuscript (NIHMS176990, PMC3319676) from the PMC Cloud open-data service, cited by section; worth reading as the one study the record holds in which a fan’s allegiance alone decides whether the same remembered play is good or bad, and in which fans remember their own side’s successes better; it says nothing about how a rivalry’s history is handed on. |
| LIT-1765 | Reconceptualizing Individuals, Categories, and the Social Group: Self-Categorization Theory 2.0 | Skorich (2025, v2 2026), preprint. Five critiques of self-categorization theory as stated in Turner et al. 1987: categorization into equivalence classes is neither sufficient nor necessary for perceiving a group as an entity; “self-categorization” is underspecified and should mean categorization of intentional agents; the person level is underexamined; meta-contrast is computationally intractable as a basis for entity formation and rests on prototype semantics; and comparative relations are overemphasized. Proposes SCT2.0 (self-conceptualization theory): groups are perceived from cues to collective intentionality, and represented as atomic concepts with object files in a language of thought. | Active — read in full 2026-10-28 (NOTE-1295) from the PsyArXiv preprint, v2; worth reading as a sympathetic insider’s critique of self-categorization theory that quotes the 1987 book’s chapter 3 assumptions and hypotheses verbatim with page numbers, so it stands in for the unread book’s core statement. It argues that categorization yields interchangeability, not a group entity; that meta-contrast is computationally intractable as a basis for entity formation; and that groups are perceived from cues to collective intentionality. Unreviewed; its key experiments are under review. Source of THEORY-1112. |
| LIT-1810 v2 | Structure-Mapping: A Theoretical Framework for Analogy | Gentner (1983). An analogy maps the relations of a base domain onto a target and drops the attributes of its objects; of the relations, those bound into a system by higher-order relations are more likely to be carried over than isolated ones (systematicity), so in “the atom is like the solar system” MORE MASSIVE THAN maps and HOTTER THAN does not. The rules look only at the form of the representation, the arity and order of predicates, which sorts comparisons into literal similarity, analogy, abstraction and mere appearance. The predicate structure is taken as given, and the author concedes it is a modelling choice; the evidence is one study summarized without statistics. | Active — read in full 2026-10-29 (NOTE-1349) from a scan of the published article on the author’s Northwestern lab pages; worth reading as the founding statement that analogy maps relations rather than attributes and that higher-order relations select which relations go across, which is what a single-relation alignment such as Gromov–Wasserstein leaves out. |
| LIT-1848 | Sparse Distributed Memory: Principles and Operation | Flynn, Kanerva & Bhadkamkar (1989), RIACS TR 89.53 / Stanford CSL-TR-89-400. A non-mathematical account of sparse distributed memory and the design of a Stanford prototype (256-bit words, 8,192 hard locations, up to 16 folds, 50 reads or writes a second). A word is written to every hard location within a Hamming radius of its address and read by summing and thresholding the counters of the locations near a cue. Used autoassociatively, a cue within a critical distance of a stored pattern converges on it in about ten iterated reads; beyond it the reads wander and may settle on another pattern. The critical distance falls as the memory fills and, in a sparsely filled memory, is below the sum of the write and read radii. The mathematics is referred to Kanerva’s book. | Active — read 2026-10-30 (NOTE-1454), Skimmed: chapter 1 in full (pp. 1–28) and chapters 2–3 looked over, from NASA’s public-domain NTRS copy (ntrs.nasa.gov/api/citations/19920001076/downloads/19920001076.pdf, 60 scanned pages with an OCR text layer). Worth reading as a short, first-hand statement by Kanerva and two co-authors of what SDM’s critical distance is: a cue closer than it is pulled toward the stored pattern by iterated reads; one farther away “will not converge” and wanders, possibly into another pattern’s attracting zone; the distance shrinks as the memory fills; and in an autoassociative memory the two outcomes are “easy to distinguish in practice”. Read it knowing that the critical distance is asserted, not derived (the derivation is referred to Kanerva’s book, LIT-928, still unread), that no simulation data are given for the convergence figures, and that most of the report describes the Stanford hardware prototype. |
| LIT-1872 | The Appraisal Basis of Anger: Specificity, Necessity, and Sufficiency of Components | Kuppens, Van Mechelen, Smits & De Boeck (2003), Emotion 3:254–269. In two studies people recalled unpleasant situations in which goal obstacle, other accountability, unfairness, control or an antagonistic action tendency was present or absent, and rated or named their emotions. Anger was associated with other accountability (specific to anger) and unfairness, but none of the components was singly necessary or sufficient for it: people were angry without each, and had each without anger. The authors propose a contingent association, with anger needing at least one of the components; no causal claim is made. | Active — read in full 2026-10-30 (NOTE-1423), from the published PDF deposited in KU Leuven’s repository Lirias (lirias.kuleuven.be/retrieve/96184). Two recall studies (110 school students, 100 university students) in which people described unpleasant situations where an appraisal or action tendency was present, or absent, and rated their emotions. Anger went with other accountability and unfairness, but no appraisal, nor the antagonistic action tendency, was singly necessary or sufficient for it; every anger report in study 2 had at least one of the components. The authors conclude for a “contingent association”. The design manipulates which memories are recalled and measures by self-report, and the authors make no causal claim. |
| LIT-1936 | Dynamic mechanistic explanation: computational modeling of circadian rhythms as an exemplar for cognitive science | Bechtel & Abrahamsen (2010). Dynamic mechanistic explanation adds to a mechanistic account a computational model, usually differential equations in time, whose variables are properties of the mechanism’s parts and operations. In circadian research the parts were found first by experiment, and the models explore their dynamics. In cognitive science the models posit the mechanism de novo, and fit to behaviour is their only check. The paper names six contributions models make, from showing that an identified mechanism can produce the phenomenon (C1) to exploring how coupled mechanisms lose and regain synchrony (C6). Its limit, in its own words: for de novo models, “Empirical evidence can rule out specific models, but usually cannot decide between competing architectures.” | Active — read in full 2026-11-01 (NOTE-1492); worth keeping for the line it draws between a dynamical model whose variables are anchored to parts and operations already found by experiment, and one posited de novo that is checked only by fit to behaviour (THEORY-1290). Kept also for two uses of modelling that bear on the record’s maintenance constructions. One is clamping a variable to test whether an identified part belongs to the responsible mechanism (THEORY-1289). The other is the open question whether peripheral circadian clocks sustain their own rhythm or are damped oscillators driven from outside (THEORY-1292). |
| LIT-1941 | M-Autonomy | Metzinger (2015). M-autonomy is the capacity for second-order mental action, whose central element is veto: rules imposed on one’s own mental behaviour, goal selection and permanence, rational guidance, and the intentional inhibition, suspension or ending of an ongoing process. It comes in three degrees: represented knowledge that one has the ability, executed attentional control, and cognitive control. During mind-wandering the ability is kept and its representation in the self-model, the “epistemic agent model”, is lost, and an ability is exercised only when so represented. The return of control is represented as self-caused even when an unconscious event caused it, a “functionally adequate form of self-deception”. The quantitative argument and the two-thirds figure repeat the 2013 paper’s. | Active — read in full 2026-11-01 (NOTE-1497), from the author’s copy on his Mainz university pages (blogs.uni-mainz.de/fb05philosophie/files/2013/04/Metzinger_M-Autonomy_JCS_2015.pdf, linked from thomasmetzinger.com), typeset with the journal’s running heads and pagination, text extracted with pdftotext. Worth reading for the sharpest statement of the will-organization question in Metzinger: what mind-wandering loses is “not the ability itself” but the conscious knowledge of it, and “one can only deliberately and autonomously exert an ability if that ability is explicitly represented in one’s phenomenal self-model” (THEORY-1300). It also argues that the return of control is represented as self-caused even when an unconscious event caused it (THEORY-1298). A development of LIT-1942 with the empirical detail cut; no new data. |
| LIT-1942 | The myth of cognitive agency: subpersonal thinking as a cyclically recurring loss of mental autonomy | Metzinger (2013). Mind-wandering is a recurring, unnoticed loss of M-autonomy, the ability to control one’s own mental processes, whose defining component is veto: that every ongoing process could be suspended or ended, and that the person knows it. Conscious thought is for the most part subpersonal mental behaviour; it becomes personal-level only when integrated into an “epistemic agent model” (EAM), a transparent self-model of oneself as a knowing, attention-controlling self, which collapses at the onset of each episode. Proposes the “self-representational blink” and a shift in the “unit of identification” to individuate episodes. Its claim that we lack M-autonomy for “roughly two thirds” of conscious life is not borne out by its own minutes, which give about half. | Active — read in full 2026-11-01 (NOTE-1495), from the publisher’s open-access PDF (frontiersin.org/articles/10.3389/fpsyg.2013.00931/pdf; also PMC3868016), text extracted with pdftotext. Worth reading as the origin of M-autonomy: the ability to control one’s own mental processes, defined by its veto component, and lost without notice in every episode of mind-wandering. It argues that most conscious thought is subpersonal and that a thought becomes the person’s only by integration into an “epistemic agent model”. The “two thirds of conscious lifetime” without M-autonomy is not what its own figures give: they give about half (THEORY-1297). Conceptual work on others’ data; no new data. |
| LIT-1945 | The Problem of Mental Action: Predictive Control without Sensory Sheets | Metzinger (2017), in Philosophy and Predictive Processing. Mental action cannot be active inference through the body, since no sensory sheet is in the loop; it is offered instead as predictive control of effective connectivity aimed at the epistemic value of states in the self-model. Attentional agency controls precision expectations of an internal model. Cognitive agency selects among candidate thoughts that mind-wandering continually generates (“mental sleep-walking”), and M-autonomy lets the system “actively sculpt its own internal affordance landscape”. Whether the conscious epistemic agent model constitutes this control, represents and enables it, or is a phenomenal artefact, current data “do not help to decide”. Agency’s sense of origination may be an enculturated post-hoc confabulation. | Active — read in full 2026-11-01 (NOTE-1496), from the open-access PDF on the collection’s own site (predictive-mind.net/papers/the-problem-of-mental-action/at_download/paperPDF), text extracted with pdftotext. Worth reading for two things the 2013 and 2015 papers lack: a model of mental action as selection from a competition among candidate thoughts that mind-wandering supplies (the “cognitive affordance hypothesis”), and a plain statement that the data cannot yet decide whether the epistemic agent model constitutes mental control, represents and enables it, or is a phenomenal by-product (THEORY-1296). That withholds what M-autonomy (LIT-1941) asserted as a principle. Four testable hypotheses, none tested; no data. |