Our next challenge is to determine whether we need to appeal to mental representations of some sort when we speak, philosophically, about language, communication, knowledge, and truth. The unquestionable phenomenon is that we use words to name things and that we syntactically articulate the way things appear. When we try to explain how this happens, and how we can be right or wrong in what we say, must we postulate in our minds some sort of mental entity, some sort of meaning, concept, or proposition, that corresponds to the things we speak about? What do our words express: do they express the things we speak about or these mental entities? Are we forced to posit these mental intermediaries, or can we somehow speak philosophically about cognition without invoking them?

Mental Representations are Problematic

Such mental representations or mental entities are troublesome things, for two basic reasons. First, it is hard to say what sort of things they are. They never present themselves to us directly. We don't experience them, even when we reflect on our experiencing. We do know that there is something like internal imaging, which we obviously experience in dreaming and daydreaming, but we never directly experience our “thoughts” or mental representations or meanings. We have to posit them philosophically because we don't know how else we can account for knowing and speaking about things; they seem needed as a kind of bridge that links our minds to the objects that we know, and they seem to be what is behind our words; they seem to be the “meaning” that makes the sounds we utter to be words.