Giacomo Melis
Affiliation: University of Stirling
Category: Philosophy
Keywords: Inference, Comparative psychology, Disjunctive syllogism, Logic vs Reasoning, Non-human animals
Date: Friday 5th of September
Time: 16:00
Location: GSSR Plenary Hall (268)
View the full session: Causal & Theoretical Reasoning
I begin by comparing how the notion of inference is characterized in philosophy and comparative psychology. Simplifying, in comparative psychology, inference is contrasted with associative learning and it is often described along two distinct dimensions. On the one hand, a representational characterization has it that whereas associative learning involves associating two perceivable events, inferring involves conjoining a perceivable and an imagined event. On the other hand, a characterization in terms of explanations has it that while reasoning involves the search for causes of unexplained events, learning consists in the mere identification of regularities and the association of external events (Premack 1995; Call 2004). Simplifying again, in philosophy inference is commonly taken to be a transition between doxastic attitudes in accordance with a rule or in response to relevant reasons (Boghossian 2014, Quilty-Dunn & Mandelbaum 2018).
A clear disciplinary difference is that, as doxastic attitudes are commonly taken to involve concepts, the philosophical notion poses higher cognitive demands on inferring subjects than the comparative characterizations—which are compatible with non-conceptual (i.e. imagistic or iconic) thought. Moreover, the philosophical notion of inference is sometimes taken to involve propositional self-consciousness (Marcus 2021) or appreciation that the premises support the conclusion (Boghossian 2014). Yet, philosophical and comparative notions have a substantial common core. As per the representational characterization in comparative psychology, transitions between doxastic attitudes are typically transitions to a new (not-previously perceived or represented) content. Similarly, to the extent that transitions between doxastic attitudes are reasons-based or inherently aim to conform to a rule, they involve sensitivity to relevant evidence, in accord with the explanatory characterization in comparative psychology. Some philosophers (e.g. Quilty-Dunn and Mandelbaum 2018; Siegel 2017, Sreckovic 2024) recognize the importance of the common core just outlined, and have defended theories of inference that acknowledge some minimal features of rational or intelligent update of one’s set of representations in response to changing information, that may well be common to human and non-human animals.
The foregoing provides one example in which the philosophical study of inference has to gain by engaging with research in comparative psychology: it enables to investigate the relationship between human and animal rationality and cognition. In the second part of the presentation, I highlight a way in which comparative psychology may gain from engaging with philosophy.
It’s fair to say that current evidence in comparative psychology suggests that at least some non-human animals are capable of inferential reasoning, at least in the minimal sense illustrated by the representational and explanatory characterizations (e.g. Call 2022, Engelmann et al 2023). It is less clear whether non-human animals can perform more cognitively demanding inferences, such as those discussed by philosophers. Accordingly, empirical research in the last 20 years or so has focused primarily on whether apes and other non-human animals perform deductive inferences, such as disjunctive syllogism.
In the comparative literature, types of inferences are commonly distinguished in terms of the degree of confidence with which the conclusion is reached. For example, Call (2022), following Mody and Carey (2016), characterizes abductive inferences as those involving a weak degree of confidence, while deductive inferences are described as those in which subjects are certain of the conclusion. Operationally, ascriptions of deductive or abductive reasoning are made on the basis of indicators of degrees of confidence, such as the percentage of success in a given task, or the hesitation in making the relevant choice. Distinguishing forms of inferences in this way is surely helpful in finding measurable indicators of types of inference, but it comes at the price of obscuring the distinction between logic and reasoning (Harman 1986).
In short, logic pertains to probabilistic and entailment relations between propositions, which obtain regardless of whether anyone considers them. By contrast, reasoning concerns how actual subjects go about forming and revising beliefs. The two are related. Careful reasoners ought to be sensitive to the constraint on reasoning posed by logic—in essence: avoid incoherence (Christensen 2004). Yet, while the necessity that binds relations of entailment is inescapable, a subject who performs a deductive inference may well fall short of being certain of the conclusion—say because one is less than certain that the premises is true, or is unsure whether the premises really entail the conclusion. The lesson for comparative psychologists is this: given that uncertainty is consistent with deductive reasoning, measures of uncertainty in performing relevant tasks needn’t be evidence of failure to reason deductively or, even worse, failure to possess deductive skills at all.
In the third and final part, I illustrate the point just made with a discussion of how the 2- and 3-cup tasks have been used to study disjunctive syllogism in primates. After outlining the task—something I cannot do in the present abstract—I focus on Enegelmann et al (2023) as a sort of case-study. In short, Engelmann et al (2023) take subjects’ inconsistent rate of success in the relatively complex three-cup task to suggest that subjects’ excellent performance in the simpler two-cup task is unlikely to have been due to disjunctive syllogism. In the light of the distinction between logic and reasoning, I argue that since uncertainty is compatible with deductive reasoning, we should be wary of taking failure to consistently succeed in a complicated task that could be solved with disjunctive syllogism as a sign that: (i) subjects are unlikely to be reasoning deductively simpliciter; and (ii) that subjects are unlikely to have reasoned deductively even in a simpler but roughly similar task where their performance was close to ceiling. Given the widespread fallibility of all reasoners, caution is needed in using degrees of confidence as a measure for deductive reasoning. In particular, it's important to propose tasks that are as simple and ecologically suitable as possible, in order to enable subjects to easily individuate relevant premises and to attach high degrees of confidence to them.
My overarching conclusion is that philosophical and empirical studies of reasoning stand to benefit from each other.
Boghossian, P. 2014. What is Inference? Philosophical Studies, 169: 1–18.
Call, J. 2004. Inferences about the location of food in the great apes (Pan paniscus, Pan troglodytes, Gorilla, and Pongo pygmaeus). Journal of Comparative Psychology, 118(2), 232–241.
Call, J. 2022. The “Avoid the Empty Cup” Hypothesis Does Not Explain Great Apes’ (Gorilla gorilla, Pan paniscus, Pan troglodytes, Pongo abelii) Responses in Two Three-Cup One-Item Inference by Exclusion Tasks. Journal of Comparative Psychology, 136, No. 3, 172–188.
Christensen, D. 2004. Putting Logic in its Place. Oxford: Oxford University Press.
Engelmann, J. M. et al. 2023. Do chimpanzees reason logically? Child Development. 94: 1102–1116.
Harman, G. 1986. Change in View: Principles of Reasoning. Cambridge, MA: MIT Press.
Marcus, E. 2021. Belief, Inference and the Self-Conscious Mind. Oxford: Oxford University Press.
Mody, S., & Carey, S. 2016. The emergence of reasoning by the disjunctive syllogism in early childhood. Cognition, 154: 40-48.
Premack, D. 1995. Cause/induced motion: Intention/spontaneous motion. In J. P. Changeux & J. Chavaillon (Eds.), Origins of the human brain. Oxford: Oxford University Press, 286–308.
Quilty-Dunn, J. & Mandelbaum, E. 2018. Inferential Transitions. Australasian Journal of Philosophy. 96 (3): 532-547.
Siegel. S. 2017. The Rationality of Perception. Oxford: Oxford University Press.
Srećković, S. 2024. Transitive Inference over Affective Representations in Non-Human Animals. Review of Philosophy and Psychology, 15: 1247-1279.