Dogs, Squirrels, and attitudes

BDK: Can a creature have propositional attitudes without 'understood intentionality'? I can imagine it is possible: a creature engages in rational inference about the world, but never develops a theory of mind, doesn't have language. Indeed, I think dogs have propositional attitudes (they desire food, believe it is in the kitchen, so run to the kitchen). (Though to be fair to dogs, they do seem to learn word reference: don't say 'squirrel' around my dog).

VR: We might have to me a tad careful about how we are applying the term 'propositional attitude" or even 'rational inference." At least if we don't want to incur the wrath of a certain husband-wife philosopher team in Chargerland. First with respect to rational inference, I don't see how you can attribute rational inference without allowing for the possibility of fallacious rational inference, and how could you do that with a dog? Does a dog that follows a red herring commit a logical fallacy? Inference seems to me to logically require the understanding of the terms of the inference sufficiently well to be able to identify sameness of content and recognize difference of content as well. Some causal associations between, say the sound "squirrel" and squirrely behavior on the part of a dog seem to me insufficient to show that the dog understands the meaning of the term "squirrel."

One thing I am trying to do (since I am writing a piece on the AFR for a Blackwell companion to natural theology) is to try to explain eliminative materialism in a way that shows that they aren't quite as insane as they appear to be at first glance.