Reliance, Trust, and Belief PETER RAILTON University of Michigan, USA (Received) AQ1 5 I. Creature Construction Paul Grice AQ2 introduced, and Michael Bratman brought to a high level of sophistication, a method for developing an understanding of a phenomenon of philosophical interest—such as being a person, acting intentionally or following a rule—by showing how one might build a creature step by step 10 from simpler components, which nonetheless could, at the end of the assem- bly, embody or manifest this phenomenon. Or near enough. 1 Successful creature construction is evidence that the phenomenon in ques- tion is possible—at least if the parts are possible. Skepticism about whether any ‘merely causal’ process could count as intentional action, for example, 15 could be met if it could be shown how ‘merely causal’ components could be fitted together into a working mechanism the behavior of which could be guided by a representation ‘in the right way’ to earn the name of intentional action. And even for those untroubled by skepticism, successful creature construction can help illuminate essential features of a phenomenon, provide 20 hints to conceptual analysis, or suggest hypotheses about its phylogenetic evolution or ontogenetic development. After all, we did, somehow, emerge from simpler organisms. And we are, somehow, built from simpler parts. II. Problem of Circularity or Regress The phenomena of interest for us here will be beliefs and believing. And the 25 components from which we will build our creature are states and processes that do not presuppose a creature already in possession of beliefs or belief- like attitudes. Correspondence Address: Peter Railton, Department of Philosophy, 2215 Angell Hall, 435 South State Street, Ann Arbor, MI 48109-1003, USA. Email: prailton@umich.edu 1 Grice, ‘Method in Philosophical Psychology’; Bratman, ‘Practical Reasoning’. Inquiry, 2013 Vol. 00, No. 00, 1–29, http://dx.doi.org/10.1080/0020174X.2014.858419 © 2013 Taylor & Francis from II. P The p uction s to conc olution or ontoge simpler orga into a w epresentation even for tho an help illum tual ana t if the par ocess could c d be shown h working mec ‘ i th on h non phenomeno on is ev ts are unt man broug understandi a person, ac ht build a ld ht to of Q Q1 Q Q2