Most scholars beg to disagree, as context and conventions are crucial to understanding sentences such as "I like it" (what is it that I like?) or "We'll go tomorrow" (what day is tomorrow?) or "Do you know what time it is?" (whose answer is not a mere yes/no).

Next, Steven Pinker focuses on the different processes (presumably neural circuits) that implement words and rules: memorization/recognition of a word, application of rules. Brain activity has been shown to be different for the two tasks: it appears to involve different areas of the brain. Pinker basically claims that the language faculty is not one: it is two. There are two cognitive faculties that allow the human brain to understand language, one that understands rules and one that understands words, and they are separate. Words that need to be memorized include irregular verbs. Pinker sees the past tense of an irregular verb (e.g., "told") just like any other arbitrary word that must be learned in the language (e.g., "chair", "light",. "but"). On the other hand, the past tense of regular verbs does not need to be memorized: it can be derived from a rule. Pinker explains the very existence of irregular verbs as a "it doesn't cost anything" because the brain already has a mechanism to memorize/recognize(thousands of) individual words. Pinker mentions (and discredits) Chomsky's theory of irregulars (that would turn an irregular into a regular, based on rules for sound patterns), but does not mention Charles Yang's theory.

Charles Yang (at Yale University) has argued convincingly that irregular forms do not exist: there are simply different sets of rules. The irregular "got" is not irregular: it is the product of a rule, even though we may have forgotten that there used to be such a rule. Therefore Yang has no need for particular rules to memorize what are basically (in Pinker's theory) exceptions. Yang's brain has only one procedure to understand words. Yang needs, instead, a way to discriminate among all the available rules that could apply to the same word. Pinker assumes that only one rule applies, and, when it doesn't, the word is "irregular", and must be memorized in a different brain circuit.

Pinker's book is mostly about "irregularities" but he neglects to consider the simplest explanation for irregularities: that they used to be regularities. The invention of the printing press may have stopped a transition from one set of rules to another set of rules, and thus crystallized a language that was evolving. A lousy philosopher (and historian), Pinker also gets in trouble when he tries to put his theory in perspective (particularly in light of the rationalist-empiricist debate, which he totally misrepresents). In the end, his book does not provide any conclusive evidence one way or the other. The faculty of language may indeed involve more than one brain circuit. In fact, it may involve many. But this book does not solve the mystery.

The book is sprinkled with erudite detours about developmental psychology, western philosophy, neurobiology and genetics.

TM, ®, Copyright © 2005 Piero Scaruffi