Reflections on Language (Chomsky, 1975) With two updates

     Noam Chomsky Reflections on Language (1975) Chomsky’s famous book defending his

view that there is some innate language-learning capability, and that details of its nature are at least in part accessible to empirical research. A dense text, made more so by NC’s irritating habit of using letters where fictitious names would do just as well, or words would do better. Also, several of his examples purportedly showing some universal grammatical rule don’t in fact do so, but merely demonstrate some of the quirks of English grammar.
     His general thesis is IMO valid enough. He points out that it’s a specific example of the general rule that theories are under-determined by evidence, in that language as experienced by the child does not offer any transparent clues to its nature, content, and form. Clearly, children must have some sort of decoding capacity built in, else they could not arrive at language competence (which they do).
     The maxim of indeterminacy of theories is one that critics and supporters of science would do well to remember. Many people believe, falsely, that science deals in certainties, that if something is scientifically proven, it’s certain. It’s not. It’s just proven. OTOH, non-scientific beliefs aren’t proven. At best, there are grounds for belief, a phrase that means that supporting evidence in the speaker’s opinion outweighs refuting evidence.
     This is not the best book to give a person who wants to find out something about why the preponderant opinion is that children have an instinct to learn language and will do so with a minimal amount of environmental input (and often in despite of it!) ** (2000)

     Update 2012: On reflection, I think that Chomsky has made a number of errors because he focuses on written rather than on spoken language. His famous distinction between surface and deep structure IMO demonstrates that he has a tin ear for speech. Intonation differentiates what he calls surface and deep structure very nicely. Intonation is in fact essential. Chomsky should have asked himself, Why do English speakers agree pretty well 100% on which bits of sound form a word?

     Update 20250502: I discovered a few weeks ago that NC was unilingual. I think that's the reason for the erroneous claim of the surface/deep structure of  language. Any multilingual knows that intonation is an essential aspect of every language spoken by humans. Intonation defines both the structure (syntax) of an utterance, and defines the semantics of what would otherwise be identical utterances.  For example, in English a rising inflection signals a question, or incompleteness, or an invitation to agree, etc, depending on context and dialect.

     For an example of intonational semantics, say the following sentence 5 times, stressing a different word in it each time: "He arrived here this morning." Then do the same with a  rising inflection.

     One of the problems in every stage production is the right or best intonation for every speech. Consider also the semantic role of intonation in Mandarin and related languages.