People think that language is an entity inside someones head

>Language is much better explained as verbal behavior (anything that mediates the behavior of others, yes this includes things other speaking).

But that barely explains anything about the specific properties that human language has been found to have. The mind clearly has non trivial structure to many of it components that don't just sprout out of nowhere.

>But that barely explains anything about the specific properties that human language has been found to have
You mean the environment? Verbal Behavior/Language won't occur if no one is there to speak to the child at a young age (this is why heavily neglected children have trouble learning to speak). You don't need to rely on any assumptions regarding the mind to explain this.

i like readin', you got a starterpack for me?

nice bait. to anyone who knows the first thing about linguistics these kinds of beliefs are actually infuriating.

>I believe in innate ideas and that some races are innately more mentally developed than others
>i don't believe in innate ideas and language, arguably, the strongest evidence of innate ideas, falls short of the rationalist's standard

i'm conflicted /pol/ i din't know what to feel anymore.

i dont know much about linguistics but a quick read over the thread has left me cringing pretty hard. its like philosophy 101 faggots trying to be DEEEEP

also i respect chomsky as a linguist but he should stfu about politics cause his views are retarded

>respecting chomsky as a linguist
You should really read MacCorquodale's criticism on Chomsky's review of verbal behavior. Chomsky clearly knew nothing about Skinners position.

eh im not really that interested in linguistics but if you link something ill check it out

Read this.

ncbi.nlm.nih.gov/pmc/articles/PMC1333660/

tl;dr, thanks though

tl;dr: Chomsky has no idea what Skinners real positions are and attacks the position of a different group of behaviorists (ones that are heavily outdated even by the standards of the late 1950s).

so whats skinners position?

A good portion of Chomskys review attacked S-R Behaviorism, which is a group of behaviorists who believed that all behavior could be explained with classical conditioning. Skinner did not believe in this at all, and he is a major reason as to why behaviorists moved away from such ideas; he helped to discover the idea of operant conditioning (to describe this simply, this just means that behaviors can be controlled by their consequences) and verbal behavior is of course based heavily on operant conditioning. This should make the problem with Chomskys review clearer; if he doesn't understand the basics of Skinner's positions, then he is obviously going to make mistakes with the more complicated aspects (such as verbal behavior).

>Language is much better explained as verbal behavior (anything that mediates the behavior of others, yes this includes things other speaking).
lol

>should stfu about politics cause his views are retarded
Cringe

This response is really confused. The author is whining about hyperbolic arguments and extensions to ideas that Skinner did not explicitly profess, but he says very little regarding Chomsky's actual arguments, and when he does, he misses the point. For instance, he says that Skinner never said that "verbal behavior" required meticulous reinforcement. Chomsky was saying that even if verbal behavior required meticulous reinforcement, it wouldn't explain language development. The argument was later developed into what's been called "Poverty of the Stimulus," the observation that there are certain errors that are logically necessary (i.e. children must make these errors) in anything like a behaviorist system where learning takes place through generalization or association, latent or reinforced, but which children NEVER make. Auxiliary inversion is perhaps the most famous example, described here ling.upenn.edu/~ycharles/papers/tlr-final.pdf

>2017
>people STILL think we don't learn grammar, but are born with it

t.Sam Harris

All those 'evidence points' for a language prior can be disproved imo. Take for instance children learning to use a subject within a sentence quite late: that fits in perfectly with the idea that language acquisition corresponds with operant conditioning wherein children learn to map their internal representations of the world to words in order to communicate their thoughts and thereby receive reinforcement (attention, food, love, etc). As children don't have a strong theory of mind, the subject of a sentence is not very salient to them.

The use of infinitive verbs by children makes sense too. Since they are trying to receive reinforcement in the sense of approval by using correct language, they will tend to try to make the least amount of obvious mistakes. Using infinitive verbs is a 'safe choice' if you haven't fully learned how to conjugate.

I read your shitty paper (glorified blog post). And you clearly know very little about neuroscience, connectionist models of operant conditioning, or current progress in character-level machine learning using recurrent networks if you think there is some magic prior grammar in the brain or that language learning isn't completely driven by data and reinforcement.

i don't think that's what it means. it just means that there is innate structure and affinity to/for language, neurologically.

Why would children have to make those errors?

So they can be corrected and be reinforced towards the right grammar.

Generalization is possible under a behavioral paradigm, its actually a pretty important component to treatment.

We see it in pigeons and other animals, in this picture, we see that a pigeon will peck more when a color similar to the original color used to establish pecking behavior (meaning that they would turn on a colored light (lets just say yellow) and whenever the pigeon pecked while the yellow light was on, it would receive food), then after pecking was established, researchers would test this with other colors (including yellow) without giving the organism food and we would see the pigeon would peck more often with colors similar to that of the original color than those that were less similar despite never being exposed to them before.

Unless children only use sentences, which their brain computes have a low chance of being wrong in terms of grammar. In that case the reinforced learning is going on in their brain, but they don't use a given type of sentence before they have seen sufficient evidence that its correct. In fact, I'm doing the same now while learning another language. Also one-trial/one-shot learning is a thing in conditioning literature, so you wouldn't necessarily see a graded transition of usage. Like people don't usually hold the middle-ground belief of 'touching hot stoves is kind of ok'. Its either 'oh I don't know what stoves are' (toddler) or 'its going to burn my finger'.

So you claim to be able to explain everything about one subject you don't understand with the tools of another subject you don't understand?

Because otherwise how are they supposed to know you can't do it that way? You can't say they just assume they can't do it that way because they never hear it; there are a ton of things that you don't hear just because they're rare, not because they're ungrammatical. You MUST assume innate knowledge, for instance in the auxiliary inversion case which this guy insists on remaining ignorant of, children never even try the simpler linear order-based rule for which they have direct evidence, and instead they use the structure dependence-based rule for which they have no direct evidence. When general learning techniques are applied to these problems they always predict that the learner entertains the incorrect but simpler model before moving on to the more complex but correct model. Children do not learn language that way.

See . There is a reason why children have problems with learning the exceptions to the rules of grammar (like saying runned away instead of ran away) because they over generalize. Assuming innate knowledge here is just an argument from ignorance (I don't know how it works, therefore it must be innate).

There is also the problem regarding children who are never exposed to verbal behavior in their life (usually neglected children) who are unable to learn to speak properly, this would also contradict the idea that language is innate.

I don't claim to be an expert in (I assume) your field, but I don't see why you couldn't answer my post with an argument instead of an insult, so I replied in a similarly non-constructive way.

There are multiple connectionist theories that can account for the type of structural learning you are talking about. For instance, multi-layer networks with higher layers representing some form of conjunction/disjunction of lower level features can produce the necessary abstraction. For instance, higher-abstraction language processing in the cortex is more proximate to the hippocampus/striatum/frontal cortex, i.e. its plausible that such higher-level language abstractions as grammar can be preferentially reinforced. Similarly there is a lot of work on how recurrent networks and similar models can learn to operate as modular components that process information using various learned gating structures incorporating hierarchies. Even normal RNNs (see Karpathy's github RNN that learns to spew English gibberish with some rudimentary grammar by reading .txt files) can produce some form of grammatical learning. There's a lot of work in the conjunction of category theory and neural networks showing how very abstract rules can be learned directly from data.

I have never heard of any hypothesis of how this universal grammar is actually encoded in the brain. On the other hand, people supporting the data-driven language learning hypothesis at least make falsifiable models with proposals of how to tie them to the brain's operation.

They make certain kinds of mistakes, like they overgeneralize regular inflections to irregular forms, but there are certain kinds of mistakes they DON'T make. It's not an argument from ignorance. If anything "generalization can do it all" is an argument from ignorance.

>Children who are never exposed to language don't speak it properly
That's called a critical period of development and is in fact evidence that language is an innate capacity. The same thing applies to vision.

>It's not an argument from ignorance. If anything "generalization can do it all" is an argument from ignorance.

Are you familiar with what an argument from ignorance is? Suggesting that something is innate because there is no other explanation is exactly an argument from ignorance. Suggesting that generalization occurs because the person learned other types of verbal behavior is not. You can disagree with the latter claim, but you cannot call it an argument from ignorance.

>That's called a critical period of development and is in fact evidence that language is an innate capacity
That suggests the exact opposite, if language is innate, why do certain environmental factors have to be in place for it to occur. That suggests that something in the environment is responsible as well, meaning that language is not innate.

I assumed you were OP/a troll.

I'm not saying, nor is anyone, that general learning mechanisms are incapable of honing in on things like general structure. Nor am I saying, nor is anyone, that general learning mechanisms don't have a role to play. Given enough evidence, such procedures can be expected to come to learn all kinds of very complex and abstract patterns. Nobody objects to that fact. The objection is that there is simply a paucity of evidence in the case of language. It would be theoretically really nice and parsimonious if we could just say "general learning" for language, but it just doesn't work. The fact is that the people who are making these general learning models are not addressing Poverty of Stimulus problems. The models which do address, say, auxiliary inversion, in fact predict that learners entertain incorrect hypotheses before moving onto the correct ones.

Everything requires input of some kind. Vision requires visual input, language requires linguistic input. There is a critical period for vision too. Does that mean visual perception is not innate?

>Everything requires input of some kind. Vision requires visual input, language requires linguistic input. There is a critical period for vision too. Does that mean visual perception is not innate?
That would imply that it is not. Its just that the situations required for vision (for people) are so ubiquitous that almost no one has a problem developing it. Suggesting that something requires factors from the environment to occur contradicts the idea that it is innate.

Cant wait for biosemiotics to come up with theoretical models and can be data-fitted

Why do you think that in order for something to be innate it can't require any extra input? Human infants have the innate capacity to develop a human language, but they need input from a language to do so. Similarly, human infants have the innate capacity to develop a human visual system, but they require visual input to do so. Kittens have the innate capacity to acquire a cat visual system, but they also require visual input to do so. Notice that the same visual input in both organisms leads to the development of different visual systems. The same applies to language; kittens exposed to the same linguistic input as human infants do not develop a language.

So then your theory is correct by default: we know that humans are more capable of learning language due to our brain being different. I think most people associate Chomsky's claim with a stronger proposition: human language capabilities don't arise from simply scaling up processes extant in other mammalian brains (e.g. larger cortex), but rather through some unique architectural element which encodes some important structural information about grammar, which sensory experience then somehow interacts with.

To some extent that is certainly true: we know that different genes are expressed differently throughout the brain. we also know that neuronal circuitry is slightly different across many areas (e.g. prevalence of inhibitory connections, microglia, pyramidal neurons, etc). However, there are multiple studies which show that in the case of injury/degeneration the cortex is very flexible in compensating for losses in functionality in one area through activity in other areas. So to me it seems that there is something akin to 'general learning' in the cortex, but that different areas specialize in different sorts of processing through the influence of both experience-driven neuroplasticity and evolved differences in genetic expression producing slightly different architectures/neuronal behaviours.

That definition of innate is completely meaningless then, as it means that literally anything an organism does is innate as the organism has the capacity to do it. You could argue that typing is innate because a human can do it with the correct environmental factors in play (someone showing you how to type). If we extend this argument even further, it would suggest shaped behavior in animals (like teaching a dog how to ride a unicycle) is also innate because the organism has the capacity to do so with someone shaping the dogs behavior.

A much better explanation is to discard the concept of "innate" and to focus on which environmental factors are responsible for producing behaviors of interest. Like focusing on verbal behavior (understanding how speaking can be reinforced as an example) as opposed to relying on the assumption that language is innate.

Yes. The fact that language is an innate capacity in humans is little more than a tautology. The arguments that language is a specialized capacity is another matter and is subject to debate, largely over how little we can get away with assuming is innate. The most parsimonious hypothesis is that there is no language-specific capacity, and that language can be explained through general intelligence/general learning mechanisms. Unfortunately there seems to be good reason to believe that's implausible. I see the view you're espousing, where different cortical regions specialize in different kinds of processing, as totally compatible with the idea that humans have innate knowledge relating to language (Universal Grammar). UG would be equivalent to the specializations of the language region(s). A language-specialized region might for instance favor structural generalizations and disfavor generalizations of linear order. That's basically the Poverty of Stimulus argument in the auxiliary inversion data; the data become straightforward if you just endow the learner with a preference for structural generalization.

right back at you buddeh

he acted like a stubborn child during their discussion and just used strawman arguments. but even if that never happened, his politics are still retarded

explain why his politics "are retarded'
im fine with you disagreeing with chomsky but dont do it like a /b/rainlet with brain damage. say something subsantive and fact based or gtfo

nah

he's pretentious and all his arguments boil down to "america is evil"

try again

explain why i should "try again'
im fine with you disagreeing with me but dont do it like a /b/rainlet with brain damage. say something subsantive and fact based or gtfo

all you did was replace retard with pretentious, with no justification as to why, he seems to be a character much like spock from star trek with a few giggles personality wise. In a very fact oriented way, he says alot more than america is evil, he says the state and their friends are bad for our well being.
personally i agree with him.
the economic and military action that is instigated by and for the united states and co. are pretty evil

It's hard to not call someone retarded when they lost a debate to some loser from /a/.

all you did was confirm what i said, that his points basically all come down to "america is evil". its a juvenile position to make him feel morally superior

>people think psychology isn't a real science and its because of people who refuse to discard the notions of mentalism
because ITS NOT you fucking moron!

>UG would be equivalent to the specializations of the language region(s). A language-specialized region might for instance favor structural generalizations and disfavor generalizations of linear order. That's basically the Poverty of Stimulus argument in the auxiliary inversion data; the data become straightforward if you just endow the learner with a preference for structural generalization.

I feel that this is the best way to describe it. The human tendency to learn language is biased towards certain kinds of structures and away from others, favoring those that recursively 'wrap around' prticular linguistic constructs, assumably due to some structural component that arises that subconsciously keeps track of the recursion.

>your mind is infinite
t-thank you, user-kun~
en.wikipedia.org/wiki/Epsilon_numbers_(mathematics)

you havent watched or read much of chomskys political work

I hate to be an internet psychologist, but I can't help but think that a large chunk of Chomsky's views on foreign policy come from being bullied. That he can't seem to let that go is where people get the best of him, like Foucault.

its mostly facts though

It's a laughably naive conception of the world. There's a great deal of room for debate between realists and idealists. Chomsky is far outside that debate. I don't think he could discuss a security dilemna without calling one side a bully without any legitimate reason, sans malice for their actions.

It is, you just have to pick the right paradigm.

ive read enough to know he's not worth investing any more time into

What fuck do all these stupid quotes mean?
1.Are you saying languages don't exist? And if they exist where the fuck do you think they exist? In the soul or something?
Languages exist in the head in the same way your retarded comment exists in my computer.

2. More valuable for what?

3. Mind is a useful word... Are you saying you don't have a mind? You don't have any consciousness?

You're a fucking faggot OP. Chomsky could be wrong about a lot of things but his work provides actual decent fucking ideas that you can use to think about language and make predictions.

Language is just another way to say words people say. They don't have to exist anywhere for people to say them.
>valuable than what
Valuable for studying.
>Mind is a useful word
We don't have a mind, everything we do is just a result of our environment and genetics.

For the longest time I just assumed Noam Chomsky was dead because the first time I heard his name I was like "yeah, that guy sounds dead".

I'll assume that you agree with OP btw so I'm arguing against that.

Are constructs more valuable than behavior? Well... Let's just say that they're pretty fucking important. What are you going to do with just behavior? You're trying to get a scientific model of Language, you're going to end up constructs.
If you're trying to understand gravity "actual behavior" is all well and good but you need to start abstracting and simplifying things into vector fields, force vector, acceleration vectors etc.

Yes, and rigid bodies don't exist. All bodies are a result of the accumulated behavior of individual atoms moving. Wait, the behavior of the atoms is actually the results of individual subparticles moving. Wait, it's inaccurate to call them subparticles, they're quantum mechanical objects.

Mentalism is experiential; you can't prevent the acausal trades of shared experience from happening and being accepted by those that are convergent on the affair, so you need to resort to better language to understand these things. See for a real-time example of what I mean.

Behaviorism died hundreds of years ago pal

>what meme psychology believes

>dead psychology calling real psychology meme psychology