I defend an account of thought on which non-linguistic beings can be thinkers. This result is significant in that many philosophers have claimed that the ability to think depends on the ability to use language. These opponents of my view note that our everyday understanding of our own cognitive activities qua thought bestows upon those activities the propositional structure of sentences and the inferential norms of public linguistic practice. They hold that our attributions of thought to non-linguistic beings project non-existent structure onto the cognitive activities of those beings, and assess the beings’ activities according to standards to which the beings bear no responsibility. So, despite the complex neural and behavioral activities of many non-linguistic beings, my opponents hold that those beings are not properly described as thinkers.
To respond to my opponents successfully, one must not merely cite scientific and folk practices of thought attribution that permit thought to be attributed to some non-linguistic beings. My opponents’ insights might be taken to demonstrate a need to revise those practices, or to treat the attributions of thought to non-linguistic beings made within those practices as instrumentally valuable but technically false. Instead, my strategy is to acknowledge the language-like structure and norms of thought, and show that a non-linguistic being’s cognitive activities might nonetheless have that structure and be subject to those norms. I identify seven features pertaining to the structure and normativity of thought – intensionality, extensionality, control, reflection, objectivity, conceptual composition, and the institution of standards – that those who share my opponents’ intuitions might deem necessary for any cognitive activity to count as thought. I argue that the motivations for the reflection requirement rest on confusions about the nature of entertaining a proposition, and I reject some interpretations of the institution requirement. But I accept the other five requirements, and a properly-interpreted institution requirement, and show how a non-linguistic being can meet them. On the account I defend, a non-linguistic being counts as a thinker given that it obtains information from at least two sensory modalities, behaves in ways that are not immediate, invariant responses to stimuli, has been through a learning process in which its cognitive system underwent modification to better conform to standards of truth and rationality and track flourishing-relevant categories, and displays a pattern of cognitive and behavioral activity in which a rational pattern can be found.
I am guided in this process by reflection on the practical purposes that are served by our folk practice of thought attribution, chief among them the coordination of rational social living. I argue that these purposes place strictures upon the nature of thought that could potentially conflict with the structure cognitive science might find in cognition, but that the folk notion requires no legitimization from science. Furthermore, which cluster of attributions correctly describes a subject is always partially determined by attributers’ varying conceptions of ideal rationality and interpretations of how to render odd behavior rationally explicable. These commitments yield my approach to the theory of thought, “Non-Scientific Interpretationism”.