Growing a tree of words

In the reading, Andrew Carnie begins to explain the multitude of concepts and nuances of morphology, or as I like to think about it, the philosophy of language structure. Striving to answer the questions about the universality of grammar across languages, Carnie explains a concept that lives up to its name: Universal Grammar. Universal Grammar (UG) says that the human facility for language is innate. From a young age, we are able to recognize grammatically incorrect sentences that we have never heard and have never learned rules about. Personally, this concept fascinates me and begs interrogating the concept that humans are wired for language. When measuring intelligence, many standards relate to language abilities, and that may not be coincidental. Right now I’m thinking about Artificial Intelligence. Many targets in the field of AI revolve around the ability to produce intuitive sentences, to speak a language without the programming of responses to particular prompts.

Explaining our intuition for language lends itself to theories of the rules that govern structure. When creating and evaluates theories of syntax, Chomsky explains the importance of looking at what he calls “levels of adequacy.” The first, “observationally adequate grammar” explains what exists in the corpus today, but that is not enough to describe the sentence possibilities. So we then have “descriptively adequate grammar,” a standard than can generalize rules we see in language for the future use. Finally, “explanatorily adequate grammar” accounts for how children acquire language.

It becomes apparent that language is structured in a hierarchical manner. The formation of sentences is not linear. When broken down, sentences appear like a tree; they arise from overarching structures that bifurcate into specificities that end up with words. For example, take the sentence “The boy thinks that Sally thinks that the dog is cute.” In this sentence, we have the noun phrase (NP) “The boy” and the verb phrase (VP) “thinks that Sally thinks that the dog is cute.” This VP can then be broken down into another NP and VP, where the NP is just the noun “Sally” and the VP is “thinks that the dog is cute.” Then this VP can broken further and further until we get the NP “the dog” and the VP “is cute,” and then finally the verb (V) “is” and the adjective (Adj) “cute.” Yayyy were done! Language exists in a tree-like structure, and some of its nuances lead to certain complex qualities. As we can see from the previous example, a VP can be broken down into another VP and NP. So a category of classification can exists within that same category of classification, a quality of logic that is referred to as recursion.

Due to the potential synonymy between language and intelligence, I’m curious how understanding the hierarchical structure of language can inform us about the wiring in our brains. When forming sentences, what parts of the sentence arise first in our brains? Do we subconsciously create the tree-structure that is used to map our words?