logy, directly analogous to that which we employ in the psychological explanation
of languageusing creatures. But this assumption might be rejected. The theorist may seek to find differences between the type of thoughts involved in psychological explanations of the behavior of nonlinguistic creatures and the type of thoughts that we have been discussing up to now. If these differences are significant enough, then it may well be that the problems we have been considering cease to loom so large. In chapter 31 examine a minimalist or deflationaryyway of understanding the thoughtss that might be attributed to nonlinguistic creatures. On the minimalist view, it is per fectly correct to describe nonlinguistic creatures as thinkers, but we need to be care ful about how we describe the type of thinking going on. We should not assimilate it to the thinking engaged in by languageusing creatures, or think of it in terms of beliefdesire psychology. When we appreciate the differences between these types of thinking we see that the apparent incompatibility with the Fregean approach does not really arise. Nonlinguistic thinking does not involve propositional attitudes— and, a fortiori, psychological explanation at the nonlinguistic level is not a variant of beliefdesire psychology. So there is no need to give an account of how a nonlin guistic creature might believe something—or indeed of how it might grasp a thought at all. It is not, according to the minimalist view, appropriate to describe nonlinguis tic thoughts as having determinate contents. Nor should the types of reasoning en gaged in at the nonlinguistic level be modeled on those at the linguistic level. For all these reasons, then, the failure of the Fregean approach to extend to nonlinguistic creatures should not be a cause for concern. As I show in chapter 3, however, there are powerful arguments suggesting that the minimalist conception cannot be the whole story about nonlinguistic thought. For the moment, the point to extract is that we have at least established a condi tional. If it is indeed the case that psychological explanations applied to nonlin guistic creatures are broadly similar in structure and intent to psychological expla nations of languageusing creatures, then we will need to find answers to our four key questions. Here they are again. 1. What can we take as our paradigm for grasping a thought (if we cannot model it as understanding a sentence) and taking a propositional attitude to it? 2. How should we understand the structure of nonlinguistic thought? 3. How can we go about identifying and attributing propositional attitudes without the resources provided by linguistic interaction? 4. How can we understand the inferential transitions by which beliefs and other propositional desires can lead to action (without taking those beliefs to have linguistic form)? Of course, it may turn out (as the minimalist view suggests) that nonlinguistic thoughts are so different in type from linguistic thoughts that these questions do not need to be answered. I turn to this suggestion in the next chapter. In the next two sections of this chapter, however, I consider the extent to which the language of thought hypothesis can deal with these four questions.