Subscribe to:
Post Comments (Atom)
-
Harnad, S. (2003b) Categorical Perception . Encyclopedia of Cognitive Science . Nature Publishing Group. Macmillan. http://www.youtube.com...
-
(10a. Comment Overflow) (50+)
-
Psychology PSYC 538, Winter 2018: Categorization, Communication and Consciousness Time : Tuesdays 2:35-5:25 Place : STBIO S3/3 ...
This comment has been removed by the author.
ReplyDeleteIn this text, Pullum writes that the conclusion that is supposed to follow from the argument of the poverty of the stimulus is that;
ReplyDelete“infants are equipped with innate mental mechanisms with specific linguistic content that assist the language acquisition process(…)”
However, these conclusions are false since the only thing that follows from this argument is that since the language learning child has a lack of negative evidence(non-UG compliant phrases), then the rules of universal grammar must be innate. There is no connection to these rules of universal grammar aiding the language acquisition process, although it may be a possibility. I think that this false conclusion is a result of the ambiguity surrounding the rules of universal grammar, which is also apparent in Pullum’s claim concerning the various authors who cite “the argument of the poverty of the stimulus.” Pullum states that these authors cite it and fail to state what it is, however this is as just wrong as his supposed conclusion. All of these errors seem to arise from the fact that Chomsky himself could not give a full explanation of the adaptive value of universal grammar, making Pullum’s attempt superficial at best.
Additionally, I find that the real error in the arguments presented by Pullum (which he alludes to in an attempt to counter linguistic nativism), is their confusion of universal grammar with ordinary grammar. This is obvious by the fact that they say there are many examples of ungrammatical sentences(which Pullum calls degeneracy), before concluding that grammar is innate.
ReplyDeleteThe evolutionary explanation of language was very compelling to me. In addition to all the research cited in the paper, there are theories that since language specialized areas are closer to the outer portions of the brain, language must have developed later as the cortex grew. The encephalization quotient of human brains are much larger than primate brains, which may also help to explain the gap between our language capabilities. Additionally, I think its plausible that through evolution, humans developed a tendency to learn language, rather than a "language mental organ". However, this brings up the same issue of Fodor's paper, that correlation does not equal causation. Can evolutionary findings be interpreted causally? Or are they too correlational like neuroanatomy. Additionally, if neuroanatomy were to give a causal explanation, how would it excuse all the variance in the brain? Specifically, language lateralization is typically in the left hemisphere, but can sometimes be found in the right hemisphere or even represented bilaterally in left handed people. Additionally, as stated in the paper, with SLI's, language can be the only aspect of intelligence affected, while all other capabilities survive. Further, fossilization studies, especially those demonstrating an inability to attain native like mastery over phonology, seem to suggest a critical period for language acquisition. This data implies that there is something different and unique to human language and how it is developed in the brain. From all the variance between people, neuroanatomy cannot explain why this came to be. What causal mechanism could explain these discrepancies, if not evolution or neuroanatomy?
ReplyDeleteI also really enjoyed the "problems" section at the end of the article. I only wonder if researchers are overestimating the language acquisition process. While obviously the learning of language is effortful and inevitable, some theories seem to overcomplicate things. For example, I don’t know if I would agree that the child learning (whether aware of this or not) is attending to whether or not their father wrinkles his nose to indicate proper grammar. Because of this, I think the biological, innate theories of learning acquisition are very tempting, though obviously flawed.
Whorf proposed a hypothesis that the categories we use to understand the world are originated through our language, so speakers of different languages conceptualize the world differently. In this way, language acquisition is learning to think, not only learning to talk. The language-thought interface was also mentioned in Harnad's paper: "symbols (words) are inseparable from their meanings. They express thoughts…. The structure of UG is bound by the structure of thought." Pinker argues against the idea that language acquisition is equivalent to thought acquisition by using evidence that thinking predates language acquisition, and that before language we thought with images. I agree that everything in our environment can be thought of without the presence of language. But maybe Whorf's proposition isn't so far-fetched - since the ability to utilize propositions to describe and understand our world is a unique human ability, and only humans have language, maybe it is possible that language can transform the way we perceive the world and changes the quality of our thought. The fact that only thinkable thoughts can be expressed UG-compliantly supports this idea, and that the two are perhaps interdependent.
ReplyDeleteRe: 6.3 Motherese
ReplyDelete"Nonetheless, those children, like all children, grow up to be fluent language speakers. It surely must help children when their parents speak slowly, clearly, and succinctly to them, but their success at learning can't be explained by any special grammar-unveiling properties of parental babytalk."
The use of Motherese is not universal between cultures, as there have been studies done which look at the way that mothers in Africa interact with their children, compared to the ways that mothers in the USA interact with theirs. Interestingly, mothers in Africa hold a similar view to the one mentioned in the article -- that children aren't worth speaking to because they don’t understand, and therefore do not speak to their children nearly as often. That being said, there are no significant developmental milestone differences, the children who are not spoken to -- especially not in motherese -- still develop language much like the ones who are spoken to. This is interesting because it illustrates again how there must be something within babies' minds that allow them to learn the properties of language without them being broken down like they are in motherese.
I find that the argument of positive evidence for the input is the most appealing with regards to language learning. By being surrounded by a social environment, the need of language emerges and thus, children develop it by a sort of instinct. On the other hand, children who are not in social contexts and do not communicate until a certain age will never develop the ability to talk. Going along with evolutionary theory, humans may have developed the capacity for UG because of this social need. From my understanding, we could also integrate prosody in the positive evidence because without the exposure of language, the prosody will not develop either, even if some words or very simple phrases could be produced. On the same line of thoughts, context is equally important and could also be considered as an element of positive evidence from what I understand of it. Indeed, I strongly doubt that if one is stuck in a black room for years with a child and tries to teach him language, that he will succeed. The child will have no reference to the world and won’t be able to make any guesses about the what the sounds would represent. There would be no temporal contiguity, not any sensorial input to help the learning of the meaning of the words. Thus, I think that as human, we evolved in a way to be ready to learn a language (innate UG), when the conditions are normal to our development (not in a black room without anything). This is the reason why positive evidence would best describe this fact, because it maintains that if the environment is favorable, a language will develop no matter what, even if it is not a recognized one.
ReplyDeleteIn section 6, Pinker argues that children need linguistic input to acquire language, and children growing up in an isolated environment are most likely to be mute. In addition, empirical evidence suggests that children need enough sources of positive evidence in order to produce grammatical sentences within the community, but negative evidence is not required for them to fully grasp the language. In short, young children learn language through social interactions with family and other children in the community. I believe this is a strong evidence supporting UG -- despite recognizing that input from the environment and unsupervised learning do play a fairly important role in language development, it is not the whole story how children acquire languages. There must be some innate and inborn mechanisms for children to pick up the rules of grammar so quickly (before the age of three).
ReplyDeleteFor this reading I particularly enjoyed when Pinker was talking about the critical period for language acquisition in children. This made me think of Genie the feral child who was found at the age of 13 chained and was never socialized. Genie was never able to acquire a first language and was tested by researchers on this topic. Her case really embodies things that Pinker talks about the critical time I child needs to learn language because of synaptic plasticity. This ties in socialization, brain maturation, and the hypothesis on negative feedback. It gives rise to more questions about how adults can learn a second language, even if it’s harder for them than a child to learn a second language, but how can we learn a second language but Genie couldn’t learn one language. This theory proves Pinker’s hypothesis of critical period but also proves how socialization at a young age is important and having stimulus.
ReplyDelete1. “One suggestion is that parameter settings are ordered, with children assuming a particular setting as the default case, moving to other settings as the input evidence forces them to (Chomsky, 1981). But how would the parameter settings be ordered? One very general rationale comes from the fact that children have no systematic access to negative evidence. Thus for every case in which parameter setting A generates a subset of the sentences generated by setting B (as in diagrams (c) and (d) of Figure 1), the child must first hypothesize A, then abandon it for B only if a sentence generated by B but not by A was encountered in the input (Pinker, 1984; Berwick, 1985; Osherson, et al, 1985). The child would then have no need for negative evidence; he or she would never guess too large a language.” (page 28)
ReplyDeleteThis example makes sense logically, however it fails to address the mechanism by which these children would be able to know all of these possible parameters. It is very easy to say that these are all innate and to suggest the use of Universal Grammar (UG). However, we must keep in mind that there are many of these parameters, and for nearly every aspect of language. For example, Werker and Hensch (2015) assert that the first sensitive period in language acquisition is one of language discrimination (i.e. being able to identify which language the child is hearing in the input). However, as was noted earlier in this reading, this occurs before the child is even able to make sense of the meaning of the input (within the first year of life). At this stage in the child’s life their brain has not fully finished developing and they are likely not actively aware of the parameters that the model suggested above proposes that they use in order to determine which language in the input is the language that they have been previously exposed to. Furthermore, I don’t see how this would account for the case of a bilingual or multilingual child, where different parameters are activated depending on the language. Overall, I feel that Pinker’s argument has a lot of merit but that it avoids its main flaw by placing a lot of the weight of the work on to UG.
2. The ‘pilking’ task by Gropen et al (1991b) described on page 27 reminds me of the ‘wug’ test (Gleason, 1958) where children were taught the nonsense word ‘wug’ was a noun referring to a type of bird. They found that the children used the term productively, showing that they have some implicit knowledge of morphology. The study by Gropen et al seems to be an extension of this, using the nonsense word ‘pilking’ as a verb, and resulting in the same response by the children (i.e. acceptance of the new term as part of the lexicon, and productive use).
ReplyDelete“language acquisition depends on an innate, species-specific module that is distinct from general intelligence.”
ReplyDeleteI am in accordance with Chomsky’s conclusion here, that it is a neural difference which underlies our ability to acquire language skill as contrasted by the rest of the animal kingdom. This idea contrasts one which was presented early in this paper with regards to the right-angle between the throat and the respiratory tract and the relative nuisance of increased choking hazard attributed to this anatomical change. Monkeys also display this anatomical change and I doubt it has anything to do with language acquisition, and more to do with the fact we are bipeds with forward facing eyes. Furthermore, recent research by William Tecumseh Fitch has concluded that macaques have the anatomical capacity to produce complex speech sounds with the same amount of variety as we do. This indicates that the evolutionary change is not at the anatomical level, but at the neural level, which is concentrated around the sylvian fissure as discussed as well as the connection between the motor and the auditory cortices. These changes can no doubt be equated to what Chomsky refers to as our species-specific module. The fact that these neural changes are appreciable in humans and that neurogenesis is most active during the first seven years of life – coupled with the finding of a “window of opportunity” for language acquisition during this time, reinforces that notion that we learn language. This has very important implications for the theory of the poverty of the stimulus and Universal Grammar – ideas that Chomsky also advanced.
If you make an OG error you know why it’s wrong because you’ve learned the OG rules either through supervised or unsupervised learning. On the other hand, if you make a UG error you know it’s wrong but you can’t say why other than it sounds wrong, because there are no rules you’ve learned for it. This is a conundrum. How is it they we know UG without learning it?
ReplyDeleteI haven’t heard many explanations, but the one that makes the most sense to me calls on the intimate link between thoughts and language. It says that sentences that are non-UG-compliant are also unthinkable thoughts: sentences that when thought are just a collection of words that don’t mean anything. For example, the sentence: "colorless green ideas sleep furiously.” may be nonsensical, but it still can be thought and mean something. On the other hand, the sentence: “furiously sleep ideas green colorless” is both nonsensical and can not be thought and mean anything. There is no way to concieve what that sentence means, thus it’s non-UG-compliant.
I don't know if this helps explain how language as a whole developed, but I think it could offer insight on how UG developed…as a link between thinkable, meaningful thoughts and language.