![Sample our Humanities journals, sign in here to start your FREE access for 14 days]({"type":"image" , "src" : "/sda/5939/TOC-Subject-Sample-2021-Humanities.jpg"})
ABSTRACT
This paper provides a naturalistic account of inference. We posit that the core of inference is constituted by bare inferential transitions (BITs), transitions between discursive mental representations guided by rules built into the architecture of cognitive systems. In further developing the concept of BITs, we provide an account of what Boghossian [2014] calls ‘taking’—that is, the appreciation of the rule that guides an inferential transition. We argue that BITs are sufficient for implicit taking, and then, to analyse explicit taking, we posit rich inferential transitions (RITs), which are transitions that the subject is disposed to endorse.
Notes
1 Siegel [Citation2017] is an exception, in that she offers an epistemological account that allows for unconscious inference.
2 We follow the convention of using small caps to denote structural descriptions of concepts.
3 And in fact, when subjects think that they're stupid (or have lowered self-esteem), effort justification ceases to change attitudes [Glass Citation1964], just as one would predict would occur if a premise in a deductive chain of reasoning was deleted.
4 A further difference is that associative links are ideally symmetric: ceteris paribus, activating one associate will activate the other and vice versa. We lack space to develop this point here (although see Mandelbaum [Citation2017]).
5 Whether the logic is good old-fashioned classical logic, some non-classical logic, or rather a proprietary mental logic [Braine and O'Brien Citation1998] is immaterial—although we suspect that mental logic is indeed proprietary and inconsistent with wide swathes of classical logic (e.g. people don't seem to reason by the principle of explosion).
6 ‘Association’ is a technical term whose meaning is understood against the background of associationist psychology going back to Pavlov [Mandelbaum Citation2017]. It also has a very loose meaning in ordinary language, which is best avoided in a careful discussion of the difference between inference, association, and other sorts of transitions. For example, we might wonder whether there's an innate ‘association’ between spiders and fear. But if this innate link between spider and the activation of fear cannot be modulated by counterconditioning or extinction, it simply isn't an association in the sense that figures in psychology. In fact, these innate associations (such as taste aversions) were the first empirical counterexamples to associationism (e.g. Garcia and Koelling [Citation1966]). If it turned out that no structures were modulable by counterconditioning and extinction, then we'd have to conclude that there were no associative structures.
7 ‘Explicit’, as we use it, doesn't mean ‘conscious’. An explicit representation is just a concrete mental token, which may be conscious or may be unconscious.
8 We add ‘modulo logical constants’ because BITs will be sensitive to elements of thoughts that aren't purely syntactic, such as if and then in a conditional, negation, etc. Since logical constants can be given narrow identity conditions, and since the only semantics involved is that of logical constants, this condition doesn't undermine the formal computational character of BITs.
9 Note that there being some indeterminacy is very different from the more radical Kripkensteinian claim that there's no naturalistic way of delineating performance errors as opposed to instances of following some bizarre rule.
10 There's a very loose reading of ‘disposition’, on which our account might be dispositional simply in virtue of our appeal to counterfactual support. But, on that loose reading, even an arch-representationalist and anti-dispositionalist like Fodor [Citation1975] provides a dispositional account. We have in mind the more robust notion of disposition employed in, for example, Schwitzgebel's [Citation2002] account of belief (cf. Quilty-Dunn and Mandelbaum [Citationms.]).
11 One option between the propositional and the associative poles holds that the concept apple functions as a pointer that enables access to various predicates, such that activating the propositional structure apples are red facilitates access to the predicate coloured via the pointer apple. We lack space to develop this intriguing possibility here (but see Green and Quilty-Dunn [Citationforthcoming]).
12 In addition to the question of whether the rules involved in inferential transitions are followed, there's the question of whether the rules are followed by the thinker, or are followed instead by the cognitive system. We intend talk of rules being built into the architecture to be neutral on whether the rule is thereby followed by the thinker or just a subsystem of the thinker (or even, as noted, whether it's ‘followed’ in some more robust sense at all). Perhaps, since thinking is something that the agent does even when it's involuntary (e.g. you are involuntarily thinking about polar bears now that we've mentioned them), BITs are things that agents do despite the BITs being architectural. We lack the space to pursue this question here.
13 Helpful comments and criticisms were offered by Tim Bayne, Joseph Bendaña, Jake Berger, Paul Boghossian, Ryan DeChant, Zoe Jenkin, David Papineau, Jesse Rappaport, David Ripley, Susanna Siegel, and the NYU Consciousness discussion group. Special thanks to the editor and referees at AJP for their useful and open-minded comments.
