I think the inaccessibility evidence is definitely the hardest for poverty of stimulus arguments to address. Even Chomsky himself suggests that taking on the task of recording all the possible environmental input a child hears would be a ridiculous undertaking. However, I agree with the authors of the article that perhaps it doesn’t need to be to this extent in order to prove validity to the poverty of stimulus argument empirically although it is certainly incredibly difficult. But what would be considered a sufficient criterion to show that a certain acquirendem is a rare enough input? Although there may be some constancy across texts, how generalizable are these when we account for individual variability and how much children must passively hear? Attention to stimuli also plays a role in deciding what counts as input as well, how would this factor feasibly be controlled?
The authors state that they are not implying that innately primed theories are incorrect; they are saying that the proponents of APS have yet to provide crucial empirical evidence that would disprove a data-driven hypothesis of language acquisition.
“The lack of a complete corpus of the input for any given learner will not be an insuperable obstacle provided only that corpora of English text do not vary dramatically with respect to the constructions that they contain. Our default assumption will be this: a construction inaccessible to infants during the language acquisition process must be rare enough that it will be almost entirely absent from corpora of text quite generally.”(13)
I thought it was interesting that they proposed that more work be done in corpus linguistics, considering the difficulties they mentioned with collecting these corpora that will reflect the types of linguistic information children will be exposed to, especially when many high frequency linguistic structures in spoken language would never be included in written works. It’s possible that items that are considered low frequency and thus inaccessible according to corpus work may not actually be so in the spoken-language environment of the child. Using corpora that reflect that child’s language exposure is crucial in order to support or refute the inaccessibility and/or the indispensability arguments
The poverty of stimulus argument is about the absence of negative evidence in the child’s input in support of the innate mechanism of UG. Pullum & Scholz poke holes in four specific arguments for linguistic nativism, but in doing so, they seem to miss the essence of what Chomsky has to say about negative evidence.
I know that the APS is mostly about child L1 language acquisition, but second language learning and L2 end-state grammar may also be an interesting area of investigation regarding the role and nature of UG in language acquisition when certain linguistic properties of the target L2 violate the groundwork laid by the L1.
Yes, absolutely, I also agree that the question of contradictions between LV1 and Lv2 are really interesting. Actually, I would even say that it is when 2 languages with contradictory rules are learned, for example french and english (des timbres-poste but the mouse-eaters). Because when comparing Lv1 to Lv2, one could claim that the LV1 learnt in childhood will be the one who dominates, given that the ability to use grammar rules properly without negative evidence has been documented only in children.
I am quite happy this article was in the readings, as I have always been curious about how the poverty of stimulus argument could have ever been substantiated with empirical evidence. Despite having no knowledge of linguistics anyone would be surprised by the fact that a claim that is so ambitious (since linguists imply that they have checked the absence of counter examples) seems rarely contradicted (it is ignored sometimes, but rarely attacked frontally). It made especially little sense to me, since English is not my first language and I was not aware of the existence of rules about plural compound words. I think Pullum is right in that more empirical/systematic evidence is needed, if only to strengthen the linguistic nativist claim.
My biggest problem with the evidence from empirical testing of inaccessibility claims, is the simple fact that the researchers are not showing if any children even use the linguistic phenomena they are explaining. For example, there is no evidence of significant amounts of children spontaneously using auxiliary sequences like “It may have been raining,” complex uses of the anaphoric one or specific idioms. Children do not talk like that. As a child grows up they start to use these types of phrases, but that’s after exposure to books, media, conversation..etc and the poverty of the stimulus argument no longer applies.
Sure, children may be able to recognize whether the uses of these linguistic phenomenon are correct or incorrect, but that’s not really impressive or special. A mass amount of direct exposure to a type of utterance is not necessarily required to notice if an utterance is right or wrong. The large amounts of linguistic data that a child indirectly experiences, such as from TV/ media/ passerby/ in their daily life, like at the school yard or waiting in a store with their parents, is constantly being compiled by the brain to unconsciously form a schemata which guides determinations of what sounds ‘good’ and ‘bad.’ The same thing happens with pitch and key in music — even if a child has never been musically trained they know when a cadence finally sounds ‘right’ or ‘wrong.’
I think the inaccessibility evidence is definitely the hardest for poverty of stimulus arguments to address. Even Chomsky himself suggests that taking on the task of recording all the possible environmental input a child hears would be a ridiculous undertaking. However, I agree with the authors of the article that perhaps it doesn’t need to be to this extent in order to prove validity to the poverty of stimulus argument empirically although it is certainly incredibly difficult. But what would be considered a sufficient criterion to show that a certain acquirendem is a rare enough input? Although there may be some constancy across texts, how generalizable are these when we account for individual variability and how much children must passively hear? Attention to stimuli also plays a role in deciding what counts as input as well, how would this factor feasibly be controlled?
ReplyDeleteThe authors state that they are not implying that innately primed theories are incorrect; they are saying that the proponents of APS have yet to provide crucial empirical evidence that would disprove a data-driven hypothesis of language acquisition.
ReplyDelete“The lack of a complete corpus of the input for any given learner will not be an insuperable obstacle provided only that corpora of English text do not vary dramatically with respect to the constructions that they contain. Our default assumption will be this: a construction inaccessible to infants during the language acquisition process must be rare enough that it will be almost entirely absent from corpora of text quite generally.”(13)
I thought it was interesting that they proposed that more work be done in corpus linguistics, considering the difficulties they mentioned with collecting these corpora that will reflect the types of linguistic information children will be exposed to, especially when many high frequency linguistic structures in spoken language would never be included in written works. It’s possible that items that are considered low frequency and thus inaccessible according to corpus work may not actually be so in the spoken-language environment of the child. Using corpora that reflect that child’s language exposure is crucial in order to support or refute the inaccessibility and/or the indispensability arguments
The poverty of stimulus argument is about the absence of negative evidence in the child’s input in support of the innate mechanism of UG. Pullum & Scholz poke holes in four specific arguments for linguistic nativism, but in doing so, they seem to miss the essence of what Chomsky has to say about negative evidence.
ReplyDeleteI know that the APS is mostly about child L1 language acquisition, but second language learning and L2 end-state grammar may also be an interesting area of investigation regarding the role and nature of UG in language acquisition when certain linguistic properties of the target L2 violate the groundwork laid by the L1.
Yes, absolutely, I also agree that the question of contradictions between LV1 and Lv2 are really interesting. Actually, I would even say that it is when 2 languages with contradictory rules are learned, for example french and english (des timbres-poste but the mouse-eaters). Because when comparing Lv1 to Lv2, one could claim that the LV1 learnt in childhood will be the one who dominates, given that the ability to use grammar rules properly without negative evidence has been documented only in children.
DeleteI am quite happy this article was in the readings, as I have always been curious about how the poverty of stimulus argument could have ever been substantiated with empirical evidence. Despite having no knowledge of linguistics anyone would be surprised by the fact that a claim that is so ambitious (since linguists imply that they have checked the absence of counter examples) seems rarely contradicted (it is ignored sometimes, but rarely attacked frontally). It made especially little sense to me, since English is not my first language and I was not aware of the existence of rules about plural compound words. I think Pullum is right in that more empirical/systematic evidence is needed, if only to strengthen the linguistic nativist claim.
ReplyDeleteMy biggest problem with the evidence from empirical testing of inaccessibility claims, is the simple fact that the researchers are not showing if any children even use the linguistic phenomena they are explaining. For example, there is no evidence of significant amounts of children spontaneously using auxiliary sequences like “It may have been raining,” complex uses of the anaphoric one or specific idioms. Children do not talk like that. As a child grows up they start to use these types of phrases, but that’s after exposure to books, media, conversation..etc and the poverty of the stimulus argument no longer applies.
ReplyDeleteSure, children may be able to recognize whether the uses of these linguistic phenomenon are correct or incorrect, but that’s not really impressive or special. A mass amount of direct exposure to a type of utterance is not necessarily required to notice if an utterance is right or wrong. The large amounts of linguistic data that a child indirectly experiences, such as from TV/ media/ passerby/ in their daily life, like at the school yard or waiting in a store with their parents, is constantly being compiled by the brain to unconsciously form a schemata which guides determinations of what sounds ‘good’ and ‘bad.’ The same thing happens with pitch and key in music — even if a child has never been musically trained they know when a cadence finally sounds ‘right’ or ‘wrong.’