Does UG exist?

Andrew Carnie introduces how generative linguists deal with syntax in this introductory textbook. He provides a well-balanced presentation of most of the issues in a student-friendly way.

In Chapter 1, he points out that the study of language could be done from a scientific perspective. Linguists could apply the scientific method to study sentence structure: observe some data, make some generalizations, develop hypotheses, and then test the hypotheses (revise the hypotheses if necessary).  Regarding the sources of data, corpora are not enough because they never contain all the grammatical forms of a language. Instead, linguists need to rely on their knowledge of their native language, which is subconscious knowledge. Generative linguists argue that much of Language is an ability hard-wired into our brains by genes. In other words, there is an innate human facility for Language, which is called Universal Grammar (or UG). The author also provides some arguments to account for the existence of UG. In Chapter 2, Carnie tells us how generative linguists determine part of speech and introduces the major parts of speech, as well as their subcategories and features. Chapter 3 introduces constituents (groups of words that functions together as a unit), such as Noun phrase, Adjective phrase, Adverb phrase, Preposition phrase, Verb phrase and so on, and the methods to test these constituents. This chapter also shows us that how to draw syntax trees.

In my opinion, generative linguists are very ambitious. They are not satisfied with the work of describing languages as structural linguists did in the past. Instead, they attempt to study the world languages from a deeper level and hypothesize that all the languages in the world share certain properties, which are called universals of Language. However, the arguments of generative linguists to account for the existence of UG do not seem to be very convincing.

First, Linguists Daniel Everett discovers that Pirahã, the indigenous language of the isolated Pirahã people of Amazonas, Brazil, is a finite language without embedding and without recursion. For example, Pirahã cannot say, "John's brother's house" but must say, "John has a brother. This brother has a house." in two separate sentences. Everett’s statement presents a challenge for the proposal by Noam Chomsky that all languages should show evidence of recursive (and similar) grammatical structures.

Second, the example that parents fail to correct their children’s syntax errors but children still acquire language, in my opinion, does not necessarily mean that children acquire language because of the existence of UG. Children might not pay attention to parents’ “instruction” when they are eager to get the spoon they want badly. They might learn the correct rule after they hear others say the correct form for many times.

Additionally, generative linguists argue that our study of Language should rely on the subconscious knowledge of native speakers. However, native speakers often have different intuitions for a certain sentence. If the UG does exist and Language is hard-wired into people’s brains by genes, why do people speaking the same language have different intuitions for the same sentence.