Chomsky is considered one of the founders of cognitive science. He was the only person who was able to argue away Skinner's conceptualization of language. Were it not for him, behaviorism may still have been dominant.
Steve sounds like your typical Redditor posting about a famous person who has made monumental strides in academia. I'm guessing Steve also has ready-to-paste arguments for anyone who says that Stephen Hawking was intelligent.
Absolutely not. Your description of his theory is wrong. His argument is that humans innately possess universal grammar. That is not at all the same as “innately storing all grammar at birth.” It was obvious to him and anyone else at the time (and long before) that a kid born in America doesn’t spontaneously start speaking Swahili without first learning it. So what is universal grammar then? It’s the innate capacity for grammar itself.
Secondly, Chomsky did in fact present evidence for his theory. His most compelling evidence is summarized by the Poverty of the Stimulus argument. That is to say, human children begin to speak and understand their first language extremely rapidly given very few examples, mostly from listening to and interacting with their parents. More recently, given our experiences attempting to build AI in the form of Large Language Models, we’ve witnessed just how incredibly tiny the “training set” for human babies really is, compared to the sheer enormity of the text they feed into something like ChatGPT (a significant portion of all the text ever written in English).
This observation is what originally refuted Skinnerian behaviourist theories of language development. Skinner essentially tried to say that humans learn language by positive and negative reinforcement, the way we train a dog to do tricks. But that idea is obviously false to anyone who witnesses a child rapidly begin speaking. Language is compositional and infinitely productive, so to be able to acquire it so quickly implies that the brain is preconfigured or structured in a way that innately understands these features. That is Chomsky’s theory.
This is shocking to hear. I'm also extremely spooked that no experts have even attempted to dispute these claims. Was he really just some pseudo-academic celebrity? Is he merely the OG Jordan Peterson?
I think there is also something structural about academia, at least culturally, that supports the kind of monolithic views of how things work, even when they've never been shown to be demonstrated by the evidence. Linguistics is far from the only field where I've seen this scenario play out.
I mean, while I disagree with his thoughts on animal consciousness, which have been disproven in clinical settings (mirror-test), AI will obviously, at least not anytime soon, be conscious. We're not even sure how consciousness works, although we are sure that we are not sure consciousness works. Without that knowledge, I posit AI will never become conscious.
Chomsky is considered one of the founders of cognitive science. He was the only person who was able to argue away Skinner's conceptualization of language. Were it not for him, behaviorism may still have been dominant.
I urge anyone downvoting these comments to refute Steve. Downvoting without honest criticism is just petty
Steve sounds like your typical Redditor posting about a famous person who has made monumental strides in academia. I'm guessing Steve also has ready-to-paste arguments for anyone who says that Stephen Hawking was intelligent.
Absolutely not. Your description of his theory is wrong. His argument is that humans innately possess universal grammar. That is not at all the same as “innately storing all grammar at birth.” It was obvious to him and anyone else at the time (and long before) that a kid born in America doesn’t spontaneously start speaking Swahili without first learning it. So what is universal grammar then? It’s the innate capacity for grammar itself.
Secondly, Chomsky did in fact present evidence for his theory. His most compelling evidence is summarized by the Poverty of the Stimulus argument. That is to say, human children begin to speak and understand their first language extremely rapidly given very few examples, mostly from listening to and interacting with their parents. More recently, given our experiences attempting to build AI in the form of Large Language Models, we’ve witnessed just how incredibly tiny the “training set” for human babies really is, compared to the sheer enormity of the text they feed into something like ChatGPT (a significant portion of all the text ever written in English).
This observation is what originally refuted Skinnerian behaviourist theories of language development. Skinner essentially tried to say that humans learn language by positive and negative reinforcement, the way we train a dog to do tricks. But that idea is obviously false to anyone who witnesses a child rapidly begin speaking. Language is compositional and infinitely productive, so to be able to acquire it so quickly implies that the brain is preconfigured or structured in a way that innately understands these features. That is Chomsky’s theory.
This is shocking to hear. I'm also extremely spooked that no experts have even attempted to dispute these claims. Was he really just some pseudo-academic celebrity? Is he merely the OG Jordan Peterson?
I think there is also something structural about academia, at least culturally, that supports the kind of monolithic views of how things work, even when they've never been shown to be demonstrated by the evidence. Linguistics is far from the only field where I've seen this scenario play out.
I mean, while I disagree with his thoughts on animal consciousness, which have been disproven in clinical settings (mirror-test), AI will obviously, at least not anytime soon, be conscious. We're not even sure how consciousness works, although we are sure that we are not sure consciousness works. Without that knowledge, I posit AI will never become conscious.
Your position is:
We don't know what it is, but AI certainly won't have it?
Does not seem rigorous.
Lol. Lmao even.
Imagine thinking so rigidly. Plenty of things were known of before it was known how exactly they work. My position is:
We know something is causing this, but our understanding of it is currently infantile.
QED