nucleation

THEORY-538: A system believes and desires when its behaviour is reliably and voluminously predictable by treating it as a rational agent with the beliefs and desires it ought to have, and the success of that strategy is an objective fact though adopting it is free; the more richly a system is connected to its world, the fewer interpretations fit, so thermostat and person differ in degree

4 min read

Properties4
StatusProposed
Tagscognition, agency, metaphysics
SourceLIT-398 — True Believers: The Intentional Strategy and Why It Works
RivalsTHEORY-092 — Behavioural capacity does not fix intelligence: a system with the capacity to respond sensibly to every possible input can lack it, so whether behaviour is intelligent, or the mark of any mental property, depends partly on how it is produced, THEORY-638 — Instantiating a formal program is never by itself sufficient for understanding or intentionality, since a person can carry out a program for answering questions in Chinese by hand without understanding Chinese; whatever thinks does so in virtue of causal powers at least equivalent to the brain's, and the reply that the whole system understands is answered only by an asserted internalisation

Graph View

Source

Dennett, True Believers: The Intentional Strategy and Why It Works, LIT-398, read in full in NOTE-353. The text read is the revised 1987 text as reprinted in Mind Design II (1997), not the 1981 printing, so claims about what the 1981 essay contained are not supported here.

What was actually shown

The strategy (pp. 61–64). Treat the object as a rational agent; ascribe the beliefs and desires it ought to have given “its place in the world and its purpose”; predict what practical reasoning recommends. A true believer is “a system whose behavior is reliably and voluminously predictable via the intentional strategy” (p. 59), a believer “in the fullest sense”.

Objectivity (pp. 66–67). The lectern is excluded because the strategy gives no predictive gain over what “we antecedently” knew. Against the worry that cleverer observers would not need it: “The decision to adopt the intentional stance is free, but the facts about the success or failure of the stance, were one to adopt it, are perfectly objective.”

Real patterns (pp. 68–71). Martian super-physicists who predict the ticker tape but not the intentional patterns “would be missing something perfectly objective”: the interchangeable ways of ordering 500 shares, and the “societal pivot points” where an outcome turns on what people believe.

Degree (pp. 72–75). A thermostat has about half a dozen belief-like states that can be reinterpreted. Enriching its sensors and behaviour shrinks “the class of indistinguishably satisfactory models” until a unique interpretation is “practically (but never in principle)” fixed. “There is no magic moment.” In degenerate cases there is no fact of the matter which beliefs a person has, and that this is so “is itself a matter of objective fact” (p. 71).

Mechanism (pp. 75–77). Why the strategy works is answered by design for a thermostat and by evolution for people. Whether a language of thought produces the pattern is a separate empirical claim, for which the reasons are “not overwhelming”.

What could have come out otherwise. Little, since the essay argues and reports nothing. What it predicts is that the strategy pays on people, animals and chess programs and not on lecterns, which is common ground.

What this does not say

  • Anything about experience. A well-predicted machine is a believer; the essay says nothing about whether it feels (NOTE-353, on THEORY-023).
  • That groups are believers. The criterion would admit a firm or a state whose group-level stance pays, and Dennett never applies it to one. Its institutional examples are reached through individuals.
  • How much prediction is enough. “Reliably and voluminously” is not quantified, and the baseline the lectern fails against is informal.
  • That intentional patterns are invisible to physics. The Martian case assumes it.
  • That most beliefs are true, as a count. The “less than ten percent” false is a guess that depends on a way of counting beliefs Dennett doubts can be made (p. 63).
  • That the inner mechanism is a language of thought. Dennett expects “some version” to prove correct and keeps that claim separate from the reality of the pattern.

Connections

  • Rival: psychologism (THEORY-092). Block holds that whether behaviour is intelligent, “or the mark of any mental property”, depends partly on how it is produced, and his lookup machine is the case. This account fixes belief by predictive success, with no condition on production. The only escape the essay offers is its baseline: whether a lookup table’s behaviour is predicted more cheaply some other way. It does not discuss the case. THEORY-092’s promote_when asks for exactly this reading.
  • Rival: causal powers (THEORY-638). Searle holds that a system running a formal program has no intentionality whatever its behaviour, and that ascription to a humanlike robot is withdrawn once the program is known. On this account it would not be withdrawn if the strategy still paid.
  • Turing (THEORY-433) gives a behavioural criterion of the same family for thinking, without a theory of belief behind it.
  • Real patterns. THEORY-036 examines the later “Real Patterns” criterion, compressibility against the bit map, which this essay’s Martians anticipate.
  • Selves. Dennett’s self as a centre of narrative gravity (THEORY-691) applies the same interpretive method to the self.
  • Agency under composition. THEORY-043 cites this criterion for why every reading it surveys grants agency in both directions.
  • Language models. Goldstein & Lederman (LIT-206) restate the criterion without a stance, measuring the gain against a mechanical explanation rather than against what was known before; they exclude thermostats, which Dennett includes as degenerate cases.

Backlinks