NOTE-120: Better to be a pig dissatisfied than a plant satisfied

Contribution

The paper responds to the Animal Sentience commentary round on Segundo-Ortin and Calvo’s (2023) plant-sentience target article. It fills a gap the authors identify (pp. 7–8): critics addressed neither what would follow if plants were sentient, nor what kind of moral status would follow. The answer is a gradualist middle position. Plants may be quasi-sentient, and so may carry partial moral status, with a simple decision matrix (Table 1) and suggestions for how the weighing might go. The authors call the goal “modest” (p. 2).

Key insight

Moral status is not zero-sum (Douglas 2013; p. 12). If both sentience and status come in degrees, one can admit plants into the moral circle at a very low weight without lowering the standing of animals, and so avoid both “plants don’t matter” and “eating plants is as bad as eating pigs”. The title’s gloss: “it may be better to have a pig with low welfare that is still worth living than to have a plant with high welfare” (p. 13).

Assumptions

• Sentience concept. The authors adopt S_BROAD, “phenomenal consciousness or any type of valenced/subjective experience”, over S_NARROW, valenced positive and negative feelings (Browning & Birch 2022), because it matches Segundo-Ortin and Calvo’s usage (pp. 3–4). Footnote 3 concedes that these concepts have “fuzzy boundaries and come in degrees”.
• Sentience grounds moral status. “without sentience an entity would have no interests or welfare” (Timmer 2023; p. 9). Footnote 2 sets aside non-sentience grounds (agency, group membership, multi-criteria views), and footnote 13 notes that the premise is contested (Sachs 2011; Kammerer 2022).
• A1, gradualism about sentience (p. 8): sentience is “a spectrum that comes in degrees”, not on or off (citing Veit 2023a; Dennett’s “hemi-semi-demi-pseudo-proto-quasi-minds”).
• A2, plants may be quasi-sentient (pp. 8–9): plants sit at the far left of the continuum, “with such faint feelings that we could not even imagine them”. The support is “current inferential evidence” and the chemical similarity of auxin to serotonin, dopamine and adrenaline (Calvo & Lawrence 2023).
• A3, moral status is a matter of degree (pp. 9–10): the sliding-scale or scalar model (DeGrazia 2002, 2008; Timmer 2023). “the simpler the sentience, the lower the moral status” (McMahan 1996).
• Precautionary uptake (pp. 10–11): a non-negligible chance of plant quasi-sentience gives “a pro tanto moral obligation to consider that assumption” (Sebo & Schukraft 2021). A non-trivial credence is warranted under “some version of a confidence threshold, expected value principle, and/or the precautionary principle” (Ladak 2023).

Key results

These are what the paper argues.

• The conditional thesis (p. 3): if plants may have rudimentary subjective experiences, and sentience grounds status, then plants may have some moral status, “on a smaller magnitude relative to animals”.
• The evidence surveyed for taking plants seriously (pp. 5–7): plant awareness disparity; decentralised information processing with collective decisions but “no evidence for a central information integration system”; anaesthesia, anticipation, kin recognition, learning (Calvo & Lawrence 2023); Boquila leaf mimicry; Venus flytrap counting of action potentials; airborne stress sounds (Khait et al. 2023). Footnote 8: complex behaviour does not prove cognition or sentience, but “shift[s] the evidential burden”.
• The critics surveyed (pp. 7–8): parsimony (Mallatt et al. 2023), anthropomorphism, disanalogy between the nervous and vascular systems (Damasio & Damasio 2023; Dołęga et al. 2023), input-output sufficiency (Booth 2023), the demand for evidence specific to sentience (Dung 2023). Footnote 10: 7.23% of philosophers in the 2020 PhilPapers Survey “accept or lean toward” plant consciousness.
• Table 1 (p. 11): no sentience gives no moral status and no obligations. Quasi-sentience gives partial status and “Some, but outweighed by animal interests in most cases”. Sophisticated sentience gives full status and obligations.
• A toy case (p. 11): choosing between saving A (quasi-sentient) and B (non-sentient), save A, ceteris paribus.
• Policy (pp. 11–12): slightly strengthen protections for plants, bearing on land development, farming, flora conservation and food systems; ethical veganism remains best.
• Weighing (p. 13): “assigning a very low value to the partial moral status of plants that would make them ‘lose’ out against animal interests even when we compare a lot of plants against a few animals”, though “an entire forest” might outweigh one pig.
• Open fine-grained questions (pp. 12–13): variation between plant species and developmental stages; same-status entities of different kinds (annelids, insects); aggregation, given plant biomass (Bar-On et al. 2018).

Claims

id	claim	strength	support
C1	If sentience and moral status are both graded, plant quasi-sentience implies plant partial moral status.	moderate	valid conditional from A1–A3 (pp. 8–10); the conclusion depends entirely on the premises
C2	Plants may be quasi-sentient.	weak	assertion that “current inferential evidence offers support” (p. 9); evidence surveyed at second hand, mainly Calvo & Lawrence 2023; the auxin–neurotransmitter analogy is chemical similarity only
C3	Moral status comes in degrees determined by degree of sentience.	weak	assumption (A3) backed by citations; contested grounds acknowledged in fn. 13
C4	Granting plants partial moral status does not trivialise animals’ moral status.	moderate	non-zero-sum argument (Douglas 2013), p. 12
C5	Plants should receive “a very low value” that loses to animals even in many-to-few comparisons.	weak	a recommendation under uncertainty (p. 13), with no aggregation rule given; it sits uneasily with the “forest vs one pig” concession
C6	Gradualism is “likely the only game in town” against line-drawing debates.	weak	assertion (p. 13)

Concepts

• S_BROAD / S_NARROW: sentience as any phenomenal or valenced experience, versus valenced positive and negative feeling specifically.
• Quasi-sentience: a rudimentary form of sentience with only some of its properties, low in intensity and relevance.
• Partial moral status: “minimal morally relevant interests that provide some reason for moral agents to have minimal moral obligations” (Terrill 2021; p. 2). It sets plants apart from stones.
• Sapience versus sentience: sapience is rationality without affect, as in DeGrazia’s angels or the Vulcans. It is treated as separable from sentience.
• Plant awareness disparity: the tendency to overlook plants, due to timescale, visual processing and historical bias.

Connections

The account of consciousness is gradualist and evolutionary about sentience, with sentience understood as phenomenal or valenced experience (S_BROAD). Attribution works by credences and markers, as for animals. It stays agnostic about the grounding account of plant sentience (p. 9).

• LIT-146 (Gruen, SEP The Moral Status of Animals) surveys the grounds and weighing of moral status that A3 relies on. Its reading note flags that the entry is not neutral on weighing, since it argues for a multi-factor utilitarian view, so cite it for the map of views rather than as support for any one sliding-scale model.
• LIT-120 (Bruckner, welfare subjects and autopoiesis) grounds welfare subjecthood in adaptive self-maintenance rather than sentience. That is the kind of non-sentience ground this paper sets aside in footnote 2, and it would reach plants by a different route.
• LIT-105 (Veit 2025): the second author later declines to count Paramecium or Chlamydomonas as conscious. He objects to extending cognition arguments to bacterial sentience, calls plant ‘experiences’ predictive control that would make prediction-based criteria “implausibly liberal”, and cites his own forthcoming “a case against plant consciousness”. That is a clear tension with this paper’s openness to plant quasi-sentience.
• LIT-159 (Schwitzgebel, over-generation from liberal materialist criteria) bears on A2. Any criterion liberal enough to include plants on behavioural and information-processing grounds invites the same over-generation worry. The paper does not address it.
• LIT-166 (Schwitzgebel’s review of Birch, Sebo and Keane; seeded only when this was written, since read in NOTE-116) concerns the precautionary moral-circle arguments that this paper borrows in its “non-negligible chance ⇒ pro tanto obligation” step (Sebo & Schukraft 2021). I have not read that review, so I note the topical link only.

Consciousness tag: justified. The paper’s central variable, sentience in the S_BROAD sense, is phenomenal consciousness. Primary topic ethics is correct.

Bearing on the record

• It could serve a THEORY or position document on graded moral status as a clean example of the graded-sentience ⇒ graded-status template, and of that template’s dependence on an assumed scalar link (A3). The same inference is used in AI moral-status arguments (footnote 11 raises the AI parallel but leaves it open).
• It should not be cited for evidence of plant sentience. Its evidence is secondary and it concedes that “the burden of proof shifts to the plant neurobiologists” (p. 13).
• ML practice: none. There is nothing for the Anthology of the SOTA.

Limitations

• All three premises are assumed. The paper argues the conditional and says so (“our conditional argument”, p. 3; “nothing … settles the final question”, p. 13).
• The account is admittedly “coarse-grained” (p. 10): there is no scale, no aggregation rule and no mapping from degree of sentience to weight. That leaves the forest-versus-pig and plants-versus-insects questions (p. 13) open.
• The evidence for A2 is drawn largely from a popular book (Calvo & Lawrence 2023) and from commentaries. The chemical-analogy argument for “diminished intensity” does not bear on intensity.
• Footnote 21 names the missing evidence: the precise degree of plant-specific valenced states, and whether plants have high harm thresholds, harm-numbing mechanisms or a different kind of harm experience.

Open questions

• Is there any plant valenced state at all? Evidence of motivational trade-offs in plants, the marker Veit uses for animals in LIT-105, would bear directly on A2.
• How should partial status aggregate across the very large numbers of plants (Bar-On et al. 2018)?
• Do different plant taxa or developmental stages differ in quasi-sentience (for instance, fast-moving carnivorous plants, p. 13)?

Corrections to the seeded skim

• Three assumptions, not two. The dossier lists A1 (gradualism) and A2 (plants may be quasi-sentient) under a section it calls “Between Sentience and Moral Status” (p. 8). The section is titled “Assumptions: Gradualism, Degrees, and the Conceptual Link Between Sentience and Moral Status” (pp. 8–10), and it has a third premise, A3: moral status is a matter of degree, with degree of moral status “determined by the strength and complexity of the degree of sentience” (p. 9). A3 answers the dossier’s question: the inference from graded sentience to graded status is assumed (A3), supported by citations to DeGrazia, Timmer and McMahan, not argued.
• Section names and pages. The dossier’s “Moral Status (pp. 10–11)” is “Moral and Policy Implications of Plant Quasi-Sentience and Partial Moral Status” (pp. 10–13). Table 1 is on p. 11. The Conclusion is on p. 13. The dossier also omits the “Why Take Plants Seriously?” and “Criticisms” sections (pp. 5–8), which carry the paper’s only discussion of evidence.
• Which “Veit 2023a” supports the evolutionary premise. The dossier gives “complex traits come in degrees (Veit 2023a)”. In this paper’s reference list, Veit 2023a is “Confidence levels or degrees of sentience?” (Asian Bioethics Review 15:93–97), and the book A Philosophy for the Science of Animal Consciousness is Veit 2023b. The dossier’s wording implies the book.
• The dossier’s summary overstates the conclusion. The paper claims only that plants may have partial moral status “if the proponents of plant sentience are possibly right” (p. 3). It says “Nothing we … have said settles the final question” (p. 13), and it endorses ethical veganism as “still the best moral option” (p. 11).
• The consciousness tag and the primary topic are both right. ethics leads correctly, since the paper’s thesis is about moral status. The consciousness tag is justified because the sentience concept it adopts, S_BROAD, is explicitly “phenomenal consciousness or any type of valenced/subjective experience” (p. 3).