Last week I participated in a workshop trying to answer the above question. The initial proposal was that thoughts are not abstract but rather mental entities for which there is currently no definition available. However, most of my readers will agree that something is going on in our heads that we usually refer to as “thinking”, a process involving something called “thoughts”.
Well, what are they, thoughts? Usually people describe having emotions or imaginations, intentions and perceptions as something they “think about”. The same holds for beliefs, propositions and, maybe, mental representations. Thus, the entire collection can be seen as a set of instances described by the more general term “thought”—and still, we are far form a proper definition.
Can the notion of consciousness help us define a thought? Are thoughts and thinking—the manipulation of thoughts—necessary to make a being conscious? Or, on the reverse, it is necessary for having and manipulating thoughts to be conscious? What can phenomenology teach us about thought?
One approach to the problem might be to appeal to “reasoning” when talking about thinking. Reasoning seems to be closely connected to rationality, i.e. in order to reason a being must have a concept of what it is to be rational. But what is rationality? And would an explanation of thought that involves rationality presuppose thinkers to be rational?
I like the way Max Coltheart puts it with reference to delusions: Humans are not rational, they think and behave not rational, so why should their thoughts be? Where is a role to play for rationality?
Still, there might be one if we assume that thinking is the systematic manipulation of thoughts according to rules of logic. But in order to do so, we would have to take for granted that thoughts can be manipulated in this (the formal, logical)—and perhaps no other—way. Consequently, a thought is to be seen as some kind of coded proposition. The question to ask is then how to code a thought most appropriately. Is it language that we need for this purpose, as José Bermúdez suggests? By the way, I think it is not.
We can try to circumvent these matters if we find a way to define thinking without reasons and rationality. According to a proposal by John Campbell, thinking can be a motor-process:
Unconscious pre-conceptual motor-instructions are brought about by beliefs, memories, interests, and perhaps, a given task. They are monitored in a feed-forward control mechanism and finally cause a possibly personal-level, i.e. conscious, thought accessible to introspection. Introspection of a thought influences the monitoring mechanism and thereby has an impact on further thoughts—progressing this way some thinker can solve a task.
There are two things I would like to draw the reader’s attention to. First, in order to adopt the above account of thinking, we need a proper theory of causation. Since this is not the matter of debate here, let us take this as unproblematic. Second, we must clarify why a thought is caused by unconscious pre-conceptual motor intentions. This claim roots in the quasi-motor-thesis of thought according to which thoughts are not cognitive representations but motor representations. Thoughts thus carry features analogous to those of motor processes in the brain.
Does this mean that all our thought works the same way—or at least analogous to the way—our motor system does? How far does the analogy between motor performance and thinking take us? And how fruitful is it?