Here and Now

None of this makes any sense so long as we persist in thinking of the mind as ideally rational, and perfectly self-transparent or unified. [emphasis added] What good could talking to yourself do, if you already know what you intended to say? But once we see the possibility of partial understanding, imperfect rationality, problematic intercommunication of parts, we can see how the powerful forces that a language unleashes in a brain can be exploited in various forms of bootstrapping, some of them beneficial, and some of them malignant.

Here is an example.

You are magnificent!

Here is another:

You are pathetic!

You know what these sentences mean. You also know that I have just introduced them out of the blue, as an aid to making a philosophical point, and that they are not the intended speech acts of anyone. Certainly I am neither flattering you nor insulting you, and there is no one else around. But could you flatter yourself, or insult yourself, by helping yourself to one or the other of my sentences, and saying it to yourself, over and over, “with emphasis”? Try it, if you dare. Something happens. You don’t believe yourself for one minute (you say to yourself), but you find that saying the words to yourself does kindle reactions, maybe even a little reddening of the ears, along with responses, retorts, disclaimers, images, recollections, projects. These reactions may go either way, of course. Dale Carnegie was right about the power of positive thinking, but like most technologies, thinking is easier to create than to control. When you talk to yourself, you don’t have to believe yourself in order for reactions to set in. There are bound to be some reactions, and they are bound to be relevant one way or the other to the meaning of the words with which you are stimulating yourself. Once the reactions start happening, they may lead your mind to places where you find yourself believing yourself after all—so be careful what you say to yourself.

The philosopher Justin Leiber sums up the role of language in shaping our mental lives:

“Looking at ourselves from the computer viewpoint, we cannot avoid seeing that natural language is our most important ‘programming language.’ This means that a vast portion of our knowledge and activity is, for us, best communicated and understood in our natural language…. One could say that natural language was our first great original artifact and, since, as we increasingly realize, languages are machines, so natural language, with our brains to run it, was our primal invention of the universal computer. One could say this except for the sneaking suspicion that language isn’t something we invented but something we became, not something we constructed but something in which we created, and recreated, ourselves.”

The hypothesis that language plays this all-important role in thinking might seem at first glance to be a version of the much-discussed hypothesis that there is a language of thought, a single medium in which all cognition proceeds. There is an important difference, however. Leiber aptly calls natural language a programming language for the brain, but we may distinguish high-level programming languages (such as Lisp and Prolog and Pascal) from the basic “machine language” or slightly less basic “assembly language” out of which these high-level languages are composed. High-level languages are virtual machines, and they create (temporary) structures in a computer that endow it with a particular pattern of strengths and weaknesses. The price one pays for making certain things “easy to say” is making other things “hard to say” or even impossible. Such a virtual machine may structure only part of the computer’s competence, leaving other parts of the underlying machinery untouched. Bearing this distinction in mind, it is plausible to maintain that the details of a natural language—the vocabulary and grammar of English or Chinese or Spanish—constrain a brain in the manner of a high-level programming language.

But this is a far cry from asserting the dubious hypothesis that such a natural language provides the structure all the way down. Indeed, Fodor and others who defend the idea of the language of thought typically insist that they are not talking about the level at which human languages do their constraining work. They are talking about a deeper, less-accessible level of representation. Fodor once made the point with the aid of an amusing confession: he acknowledged that when he was thinking his hardest, the only sort of linguistic items he was conscious of were snatches along the lines of “C’mon, Jerry, you can do it!” Those may have been his “thoughts,” and we have just seen how they may in fact play an important role in helping him solve the problems that confronted him, but they are hardly the stuff out of which to fashion perceptual inferences, hypotheses to be tested, and the other postulated transactions of the ground-level language of thought. Keynes was right to resist the words-versus-pictures choice; the media used by the brain are only weakly analogous to the representational media of public life.” ”