IntroductionInnate Knowledge of LanguageIntroductionIn Plato's dialogue "Meno", Socrates sets out to convince Meno that we have knowledge without experience. He uses a young slave from Meno's household, one who has not had any formal education, for this purpose. Socrates asks the boy questions about geometric shapes that he draws with a stick in the sand. Without teaching the boy anything, only by asking him questions, Socrates elicits the Pythagorean Theorem from the boy – that is, the square of the longest side of a right triangle (the hypotenuse) is equal to the addition of the squares of the two other sides. Socrates then concludes that knowledge is not taught but is recollection from a previous existence. In terms appropriate to today, we might say that knowledge is innate, that it is carried in our genes.Innate Knowledge of LanguageWe can approach the question of innate knowledge of language in the following way. Keep in mind that the knowledge we are referring to is not conscious knowledge anymore than the Pythagorean Theorem was conscious knowledge for Meno's slave boy.1. First, if we look at the acoustic record of a sentence it is best described as a continuum – what we call the stream of speech. In the physical record it is just uninterrupted noise. However, thespeaker/hearer "knows" that this noise is a linguistic event and different from other noises. The ability to distinguish linguistic noise from other noises is not taught. 2. Second, the stream of speech is heard mentally not as an undifferentiated stream of noise (which it is physically), but as made up of distinct phrases, distinct words, even distinct sounds. That is,unlikethesentencesonthepagewhichhavespacesbetweenthewordsandpunctuationbetweenphrasesandwhereeveryletterisdistinctfromeveryotherletterthestreamofsppechismorelikethisexample.Once again, we take it for granted that we hear distinct words and phrases when physically we hear a continuous stream of sound, and once again, when we take something like this for granted, it is likely innate. That is, no one has taught us to analyze the stream of speech into its linguistic components, but we do it all the time with sentences we hear for the first time of have heard before.3. Third, we can hear the same sequence of words in different contexts and interpret them with different meanings.1. Flying planes can be dangerous.4. Fourth, another example from Chomsky in Syntactic Structures points to another kind of innate knowledge.Chomsky's Syntactic Structures:2. Colorless green ideas sleep furiously.As Chomsky points out this sentence is meaningless, nevertheless it is quite different from the same words taken back to front.3. furiously sleep ideas green colorlessThat is, meaning does not determine grammaticality. Both (2) and (3) are meaningless, but (2) is grammatical. This knowledge is not taught, but we all know it.5. Fifth, consider the meaning of the pronoun he in the following examples.Chomsky's Syntactic Structures:4. John thinks he is smart.5. He thinks John is smart.6. Although he studied all night, John still failed the test.In (4) he can refer to John or to somebody else – say, Bill or Harry. In (5) he cannot refer to John at all. Notice that it is not simply that he followsJohn in (4) and precedes John in (5) that determines itsinterpretation. In (6), he precedes John as it does in (5) and still refers to John.As we will see in the discussion of syntax in the next module, the determination of pronoun reference in sentences like (4) through (6) depends on a structural property of sentences that is not taught, but must be innate.6. Sixth, we know that sentence like (7) and (8) mean the same thing, and those like (9) and (10) mean quite different things.Chomsky's Syntactic Structures:7. Mary opened the letter.8. The letter was opened by Mary.9. John is easy to please.10. John is eager to please.Though the words are in quite a different sequence in (7) and (8), the meaning of who did what is nevertheless the same. In (9) and (10), the order of words is the same, but the meaning of who will be pleased is completely different. This is knowledge that is not taught.As a final set of examples, consider the following:Chomsky's Syntactic Structures:11. Harry will study.12. Will Harry study? Sentences (11) and (12) are clearly related, the former is an answer to the question in the latter. Let us assume that the two sentences are related by a rule. A first approximation of the rule is: a question is formed by inverting the first two words of a statement. This rule will work for deriving (12) from (11), and it will work for deriving (14) from (13).Chomsky's Syntactic Structures:13. He will study.14. Will he study? Thus the rule – invert the first two words of a statement to form a question – is quite good. It will work for a very large number of sentences that start with a name like Harry (Mary, Sarah, Bill) or a pronoun like he (she, you, they) followed by a modal like will (can, must, might, would). However, it will fail for an even larger set of sentences like the following (an asterisk indicateds the sentence is ungrammatical):Chomsky's Syntactic Structures:15. The student will study.16. *student the will study. It seems obvious that a rule can't be based on counting the words in a sentence. No one tells children that this is the case. And it is not unreasonable to think that counting words might be the basis for a rule of grammar. Nevertheless, children don't make mistakes like (16) and counting wordsis irrelevant to grammatical knowledge.Thus, it appears that Plato's problem – how is it that we know so much with so little experience – has real bearing on knowledge of language, as it does on knowledge of geometry, on knowing what we see, or even more basically of how the organs of the body know how to carry out their functions. While this last case seems obvious - the digestive system is a physical system designed to do what it does - the knowledge of language or geometry is less obvious. For Plato, knowledge of geometry had to be recollection of memories from previous existence. For us, today, we see this recollection of memories as in our genetic structure. Similarly, for knowledge of language, we can conclude that untaught knowledge is in our genetic structure. As we proceed through the following modules on syntax, morphology, phonology, language change and acquisition, Plato's problem will continually come to the fore. It is central to the