Chapter 3 of Language Instinct

Language instinct has always been heard. I browsed a few pages in the library before, because the translation of the last edition had to be given up. Some time ago, I heard that there was a new translator, so I quickly started. I have to say that the translation seems young, but the translation pen is still very accurate.

I think that when I started this blog, the book had already read the fourth chapter, and I was too lazy to summarize all the previous chapters, so I started from the third chapter and officially entered the "dry goods" of linguistics (the first two chapters were to prove the instinct of language, and the second chapter involved general linguistics such as language characteristics, language ability and language expression).

After introducing how human beings use the mechanism of "mind language" (similar to Chomsky's "universal grammar" theory) to learn and use new languages, Pinker conveniently entered the field of syntax and further analyzed how the mind constitutes a language.

At the beginning of the chapter, two concepts are mentioned, which explain why human beings can assemble thousands of sentences with language. One is Saussure's "arbitrariness of symbols", that is, the sound and meaning in the language are completely arbitrary; The second is wilhelm humboldt's "Infinite Application of Finite Fields", which means that we have a compiling system in our mind, which can transform words and ideas into each other. This is called "Generative Grammar". The relationship between generative grammar here and the previous concepts: psychological grammar and universal grammar is that generative grammar is about equal to psychological grammar or a psychological grammar with specific functions, which is aimed at individuals. And universal grammar is aimed at groups.

As mentioned above, the concept of "finite field" is a "discrete combination system" for the "finite field" of grammar: the grammar system contains limited "components", and a larger system can be formed by random arrangement and combination, which is called "infinite application", and this process depends on the "arbitrariness of symbols" of language.

Unfortunately, however, our minds belong to another system-* * * mixed system. In the * * * mixed system, various elements are mixed, dispersed and mixed well, losing their individual characteristics (such as a bowl of broken egg paste). The reason why the mind is such a system involves the concept of "categorization" in cognitive linguistics. I think it will be mentioned later in this book, so I won't say it here.

So after a long speech, what Pink wants to talk about is what kind of mechanism exists in our language to connect these two systems.

Of course, Pinker still has a lot of nonsense to say before formally entering the discussion of syntax. He mentioned the most primitive and basic discrete combination system-"string machine", which simulated the process of combining words into sentences by string machine. The lyrics creation process of Weibo's joke "Wang Feng" is to randomly select, arrange and combine words such as "hesitation", "life", "ideal" and "anger". It is a working mode of string machine. So to put it simply, the string machine relies on its own logic to extract words from different vocabularies and form sentences.

Pinker went to great lengths to tell such a bunch of seemingly identical things, just to prove that human language mechanism is much more powerful than artificial language mechanism. For example, according to the logic of string theory, you can never make such a completely ungrammatical sentence as "colorless and green sleeping anger" exemplified in the book. Moreover, in the face of complex sentence patterns, the string machine is at a loss.

This difference is due to the characteristics of language, but also the embodiment of language characteristics.

In this chapter, Pinker explains this theory step by step. First, start with traditional grammar.

When humans combine sentences, they use phrases as units. In traditional syntax, a sentence consists of a noun phrase and a verb phrase. Sounds subversive. "I read less. Don't lie to me. The sentence is so complicated, how can it be composed of only two things? " But if you think about it, you will find that it is true. No matter how complicated a sentence is, it can be simplified into this form, which is why a sentence can have no object, but it must have a subject predicate, such as "I overslept", "I ate during the day" and "Fuck me".

So what this sentence expresses is:

S→NP VP(→ means "composed of")

A noun phrase consists of a dispensable determiner, an infinite number of adjectives and a noun, namely:

NP→ (det) A* N

Verb phrases are similar and consist of a verb and a noun phrase:

VP→V NP

These principles are enough to form a simple sentence weighing several tons. But the problem is that language can't be just a simple sentence, so this phrase system stipulates where to insert something so that the sentence can continue to extend. In the process of this extension, there will be ambiguity because a word can be attributed to multiple phrases, or a phrase can be attributed to multiple sentences.

To discuss NP and VP, we must first understand what is a noun and what is a verb. The noun verbs mentioned here do not refer to the category of meaning, nor to the part of speech. The category of word meaning corresponds to the concept represented by words. Just like a fish cooked, stewed and fried by a squirrel, the same concept can be expressed in different parts of speech, but its semantic category is still the same. So this noun can actually mean an action (his death made her very sad. ); Verbs are not always an action (I'm busy writing this damn blog. Engage is not i) the actual action.

After this point is made clear, we can begin to discuss the composition of phrases.

The meaning of a phrase depends on the meaning of the head language. A phrase has its center of gravity, which is the head language. When the head word changes, the meaning of the phrase changes. "This is a GV without code" and "This is a GV with code" are all about GV, while "You are shooting a GV with code" is about the action of "shooting".

With the central phrase, you can point to a single concept, but this is not enough. Therefore, "argument" gives phrases the ability to describe the relationship between a series of objects and the head language. For example, the argument in "A GV You Shoot" is "you", and the relationship between you and the head language is the relationship between the lens and the lens.

When the center word and argument are crowded together, a secondary phrase under a phrase is formed, which is called "X bar", and X is the name of the phrase.

Parameter has something that looks like a modifier. For example, in the sentence "You shoot GV straight", "You are straight" is a modifier, and the core of this sentence still lies in you and your GV. Whether you are straight or not is just a dispensable message. In addition to guessing by feeling, you can also guess scientifically. As I said before, the head language and arguments will warm up and are not allowed to intervene. Therefore, if inserting elements into two components of a phrase will change the meaning of the phrase, it means that they are secondary phrases, and the secondary phrase is either the central component or the argument. The reason why this judgment method is scientific lies in the relationship between the head language and the argument that one is an actor and the other is a role. They constitute "what" and modifiers describe "how". An actor plays a role, and this role can be like. Therefore, Pinker also called the argument "role-playing".

The last component in the phrase is a demonstrative word (short for SPEC, Specifier). It corresponds to the subject in traditional grammar, and the subject is also regarded as a special role in syntax, because in many cases the subject is the sender of the behavior in the phrase. But noun phrases can also have subjects.

The above four phrase components, 1) head, 2) argument, 3) modifier and 4) subject, are all noun phrases, but they are also applicable to verb phrases. It is precisely because of this similarity that sentence jurists further simplify the expression of phrases and merge verbs and noun phrases into "X phrases":

XP→ (specification) X-bar YP*

A phrase consists of a dispensable subject, an X- bar and any number of modifiers YP.

X-bar→X ZP*

An X- bar phrase consists of a head word X plus any number of teleological elements.

The research object of linguistics is all languages after all. The arrangement of words in different languages may be completely opposite. For example, in English, verbs come before objects, while in Japanese, verbs come after objects. Therefore, linguists have modified these two rules and added a pair of braces {} to the right of the expression, indicating that the elements in brackets can be transposed at will. In specific applications, only specific applicable conditions need to be added, which are called parameters. This theory is Chomsky's latest research achievement and is called "principle-parameter theory".

I'm dying at this point. Let's talk about the rest next time.

Just sauce.