Sunday, October 16, 2016

Baby talk

In his writing, Andrew Carnie presents the idea of language as a hard-wired element of human beings, or “universal grammar.”  He posits that as humans, we are equipped with an instinct for Language, and that with this toolkit, we should be able to almost effortlessly spot grammatical rules and differences between languages.

Carnie’s idea perhaps provides an answer to how humans communicate despite gaps in vocabulary, grammar, and fluency. One primary example is in the communication between young children and adults. A child might say “milk all gone” or “I goed,” both of which are grammatically incorrect. However, the meaning and intention behind the words are still very clear—somehow, it’s obvious to us that the child is trying to say “the milk is all gone” and “I went.”

So how exactly do we learn Language even without formal instruction? This past week in SYMSYS1, we read excerpts from Steven Pinker’s “The Language Instinct.” In a similar vein, Pinker talks about how the best way to represent sentences are not through word-chains, where we link together individual words based on how often we hear one word after another. Instead, sentences and language are better represented by trees and phrases. Parts of speech, he claims, make it easier for us to group words and fit them into structures.


This goes to show that a key element of our “language instinct” is in our ability to spot simple grammatical rules, such as adding “-ed” to make a word past tense. For example, though the phrase “milk all gone” doesn’t form a complete sentence, it does demonstrate knowledge of using an adjective to describe a noun. From this one rule, we can form many sentences that might not necessarily be correct, but will at least have a sound enough structure for people to recognize the meaning behind our words.

6 comments:

  1. Great post and connection! The one part of your argument I am hesitant to accept is that having a “language instinct” really means that language is an innate quality in humans. Though we might have some sort of instinct, is this instinct acquired through exposure to real-world events? Does it really count as a “language instinct” when babies can’t put together coherent sentences? These are just some lingering questions I have after reading your article.

    ReplyDelete
  2. I really like your point that often times, "good enough" works really well in being able to understand sentences that are grammatically incorrect. I'd really love to dig deeper to analyze where exactly the line "good enough" falls. For example, "milk gone" can be fairly easily interpreted, but if I said something like "I went milk" it's a bit more unclear -- do I mean "I went with the milk?" "I went to the milk?" Sometimes a preposition is absolutely needed for context and interpretation. So overall, at what point can we disregard grammar and at what point is grammar imperative to the understanding of a sentence?

    ReplyDelete
  3. This comment has been removed by the author.

    ReplyDelete
  4. I disagree with the idea that we learn Language "without formal instruction". Imagine someone who has not had a formal education but is able to speak English because he/she has been exposed to the language growing up. I am sure you could find instances of him/her making grammatical mistakes. Even as native speakers at school, we learn when to use "that" instead of "which". In the SATs for example, we are tested on our knowledge of this. So I don't think "the idea of language as a hard-wired element of human beings" is completely true. I more strongly agree with the alternate hypothesis hinted at in the problem sets, suggesting that humans pick up linguistic rules (components of the context-free grammar or FSA, perhaps) as they go along, and revise them based on the data. This is exactly how we see children pick up language and syntax - they make mistakes, are corrected, and then revise their mental model of the syntactical world.

    ReplyDelete
  5. I think you've made an excellent point about how babies use their innate language skills to form basic sentences, but similarly to the comments above, I find myself asking some questions. Although they can form sentences that can be understood by their parents, children still have to go through a more formal learning environment to be fully immersed in language and accurate grammar. I think the "sound enough structure' may only be applied to parents or close relatives that can understand "baby talk" because most of the time, when babies use incoherent sentences with strangers, people don't understand what the baby is asking for and usually end up asking the parents about what their baby needs.

    ReplyDelete
  6. Coincidentally, I have read Steven Pinker's "The Language Instinct," and you are right. He does provide examples where it seems as if children's amazing ability to learn to speak and comprehend a language is made possibly only by some innate quality inherent in our biology.

    He goes further to provide examples where the subject is unable to speak grammatically, and links the inability to the poor "wiring" of the brain and in some cases traces these links to genetically transferable conditions.

    I agree that is interesting to look at what provides a sentence really with semantic meaning - an interesting question would be to research how well these "implications" are understood by infants or children across different languages determined by an alternate set of parameters.

    ReplyDelete