Please post your comments on the readings for Monday Nov. 18th and Wednesday Nov. 20th and the class discussion here.
8 thoughts on “Age-related changes in speech perception”
Monica Bennett
The Wang & Kuhl (2003) proceedings paper used learning to identify tones as its measure. However, couldn’t one argue that this is just like learning to identify any kind of perceptual feature (like different colors) , and is not comparable to the massive undertaking that is acquiring an entire language? And, it also seems like tone in particular could easily be reduced to acoustic learning and not processing specific to language.
Anecdotally, I have been exposed to occasional Mandarin for much of my life, and I can perceive the differences in tones in isolation, but I can’t label them accurately, nor am I confident in my ability to perceive their differences in continuous speech. If I have some cortical connections that have been maintained to perceive differences, but did not fully learn it, how much should that facilitate my potential learning of Mandarin as an adult?
Building off of Monica’s comment, how can we think of phonological learning experiments, including but not limited to Wang and Kuhl, as an analog for language learning generally? What exactly are the steps between perceiving differences and learning the language, even putting aside production for now?
I had a question relating to a thought I had during yesterdays (Mondays) lecture. It was mentioned that the evidence from a study with chinchillas being able to learn the same thing that infants could showed that the ability wasn’t innate. I’m not sure how finding the same ability in an animal would show that this process is not innate, in fact I would think that it would strengthen the argument. With that being said, have there been studies done with closer descendants of humans such as apes looking at speech perception? I know that in a lot of developmental psychology work, evidence that certain processes exists in apes is taken as support for innate processes.
(i) The version of the critical period hypothesis that Wang & Kuhl are arguing against here seems very strong. I think there are less extreme versions (that I seem to recall from my introductory linguistics classes not too long ago..) that would *not* say that learning of any novel second language contrast is impossible. Are[n’t] W&K to some extent arguing against a straw man?
(ii) I found it unsurprising that adults should perform a task that they were trained to do better than children perform the same task after training; there are many [other] strategy games/memory-based tasks for which we would expect better adult than child performance, too. In line with Monica and Tina’s above comments, it seems worth asking: Wouldn’t a fairer assessment of adult vs child language learning abilities would be a long-time-course experiment employing an implicit learning paradigm, with lots of fillers/distractors in training?
In response to Ashley’s question: I agree that the finding that chinchillas have categorical perception is not an argument against the innateness of the ability. I think that linguists originally found those studies very informative because they showed that categorical perception was not unique to humans, and thus we could *not* argue that it is part of that innate and human-language-specific faculty that is supposed to aid us in learning and using language like no other species.
Tina: There’s a pretty cool paper, Pallier et al. (2003) which looks at something kind of related. Participants were people born in Korea but who were adopted at an early age and grew up in France. They were all fluent French speakers, and had no conscious language ability in Korean. The control group was monolingual French native speakers. They performed a couple of behavioral tasks (recognizing which language speech was, and Korean word recognition), plus fMRI scanning.
And… on the behavioral tasks, the two groups performed about the same. The fMRI data also didn’t really differentiate the groups that much, though there was more widespread activation to French in native French speakers (compared to other languages) than for the Korean-born participants.
Here’s the article: Pallier, C., Dehaene, S., Poline, J. B., LeBihan, D., Argenti, A. M., Dupoux, E., & Mehler, J. (2003). Brain imaging of language plasticity in adopted adults: Can a second language replace the first?. Cerebral Cortex, 13(2), 155-161. http://dx.doi.org/10.1093/cercor/13.2.155
So, this doesn’t really address your question about acquisition, since this is more about testing the “hidden knowledge” that the adopted Koreans had rather than how easy it would be for them to learn Korean. But just going on intuition, I think it might be easier for you to learn tone contrasts because you’re already more able to perceptually distinguish them better than naive learners are… probably. Wanna do some lab tests?
Does anyone know of a paper more directly related to acquisition?
Throughout the semester we have had discussions on whether components of phonological learning are domain specific or domain general. I am wondering this in response to Rivera-Gaxiola and colleagues’ (2005) term neural commitment as a possible explanation for their ERP findings. Would a robust effect such as perceptual narrowing within face differentiation/processing be explained by neural commitment as Rivera-Gaxiola and colleagues explain it? Are there other motion-(action) related experiences of learning that may also reflect neural commitment?
Additionally, I would have liked further discussion on the process of neural commitment (to auditory patterns of an infant’s native language). Is the commitment taken in the form of shifts from fast occurring perceptual responses to later attentional or memory processing of auditory patterns? Lastly, are there multiple periods either as a function of biological maturation or experience-based maturation that would lend to multiple occurrences of neural commitment? Because there is still some flexibility post the onset of neural commitment, when acquiring a second language later in life an instance for a reemergence or re-organization of neural processing?
One thing I’ve never been entirely clear on is if there’s a consistent position on whether this sort of perceptual narrowing is actually helpful, or just a byproduct of maturation.
I mean, is there any reason not to attempt to keep the ability to distinguish these tones? I’ve heard contradictory things about the effect of learning multiple languages. Does narrowing help you learn a single language more effectively as a child? Does lack of narrowing impede language learning?
Monica and Tina’s comments sound very much like the reaction I had to the Wang and Kuhl paper — simply being able to learn the tones is very different from the undertaking of learning an entire language. This task was, to my mind, much more akin to an auditory learning task. I would be interested, for example, to see how children and adults performed on a similar but more “linguistic” task — perhaps one incorporating syntactic structure, such as learning tone sandhi from training items and applying it to novel forms.
In the article, Wang and Kuhl say their study shows evidence that “language learning is not a strictly timed developmental process with rigid cut-off periods.” However, as we know, child language learning has very little to do with explicit training — indeed, children in the critical period often ignore explicit linguistic training. Researchers have to be very careful when asserting that experiments with explicit training give insight into critical period language learning.
The Wang & Kuhl (2003) proceedings paper used learning to identify tones as its measure. However, couldn’t one argue that this is just like learning to identify any kind of perceptual feature (like different colors) , and is not comparable to the massive undertaking that is acquiring an entire language? And, it also seems like tone in particular could easily be reduced to acoustic learning and not processing specific to language.
Anecdotally, I have been exposed to occasional Mandarin for much of my life, and I can perceive the differences in tones in isolation, but I can’t label them accurately, nor am I confident in my ability to perceive their differences in continuous speech. If I have some cortical connections that have been maintained to perceive differences, but did not fully learn it, how much should that facilitate my potential learning of Mandarin as an adult?
Building off of Monica’s comment, how can we think of phonological learning experiments, including but not limited to Wang and Kuhl, as an analog for language learning generally? What exactly are the steps between perceiving differences and learning the language, even putting aside production for now?
I had a question relating to a thought I had during yesterdays (Mondays) lecture. It was mentioned that the evidence from a study with chinchillas being able to learn the same thing that infants could showed that the ability wasn’t innate. I’m not sure how finding the same ability in an animal would show that this process is not innate, in fact I would think that it would strengthen the argument. With that being said, have there been studies done with closer descendants of humans such as apes looking at speech perception? I know that in a lot of developmental psychology work, evidence that certain processes exists in apes is taken as support for innate processes.
In reaction to the Wang & Kuhl article:
(i) The version of the critical period hypothesis that Wang & Kuhl are arguing against here seems very strong. I think there are less extreme versions (that I seem to recall from my introductory linguistics classes not too long ago..) that would *not* say that learning of any novel second language contrast is impossible. Are[n’t] W&K to some extent arguing against a straw man?
(ii) I found it unsurprising that adults should perform a task that they were trained to do better than children perform the same task after training; there are many [other] strategy games/memory-based tasks for which we would expect better adult than child performance, too. In line with Monica and Tina’s above comments, it seems worth asking: Wouldn’t a fairer assessment of adult vs child language learning abilities would be a long-time-course experiment employing an implicit learning paradigm, with lots of fillers/distractors in training?
In response to Ashley’s question: I agree that the finding that chinchillas have categorical perception is not an argument against the innateness of the ability. I think that linguists originally found those studies very informative because they showed that categorical perception was not unique to humans, and thus we could *not* argue that it is part of that innate and human-language-specific faculty that is supposed to aid us in learning and using language like no other species.
Tina: There’s a pretty cool paper, Pallier et al. (2003) which looks at something kind of related. Participants were people born in Korea but who were adopted at an early age and grew up in France. They were all fluent French speakers, and had no conscious language ability in Korean. The control group was monolingual French native speakers. They performed a couple of behavioral tasks (recognizing which language speech was, and Korean word recognition), plus fMRI scanning.
And… on the behavioral tasks, the two groups performed about the same. The fMRI data also didn’t really differentiate the groups that much, though there was more widespread activation to French in native French speakers (compared to other languages) than for the Korean-born participants.
Here’s the article: Pallier, C., Dehaene, S., Poline, J. B., LeBihan, D., Argenti, A. M., Dupoux, E., & Mehler, J. (2003). Brain imaging of language plasticity in adopted adults: Can a second language replace the first?. Cerebral Cortex, 13(2), 155-161. http://dx.doi.org/10.1093/cercor/13.2.155
So, this doesn’t really address your question about acquisition, since this is more about testing the “hidden knowledge” that the adopted Koreans had rather than how easy it would be for them to learn Korean. But just going on intuition, I think it might be easier for you to learn tone contrasts because you’re already more able to perceptually distinguish them better than naive learners are… probably. Wanna do some lab tests?
Does anyone know of a paper more directly related to acquisition?
Throughout the semester we have had discussions on whether components of phonological learning are domain specific or domain general. I am wondering this in response to Rivera-Gaxiola and colleagues’ (2005) term neural commitment as a possible explanation for their ERP findings. Would a robust effect such as perceptual narrowing within face differentiation/processing be explained by neural commitment as Rivera-Gaxiola and colleagues explain it? Are there other motion-(action) related experiences of learning that may also reflect neural commitment?
Additionally, I would have liked further discussion on the process of neural commitment (to auditory patterns of an infant’s native language). Is the commitment taken in the form of shifts from fast occurring perceptual responses to later attentional or memory processing of auditory patterns? Lastly, are there multiple periods either as a function of biological maturation or experience-based maturation that would lend to multiple occurrences of neural commitment? Because there is still some flexibility post the onset of neural commitment, when acquiring a second language later in life an instance for a reemergence or re-organization of neural processing?
One thing I’ve never been entirely clear on is if there’s a consistent position on whether this sort of perceptual narrowing is actually helpful, or just a byproduct of maturation.
I mean, is there any reason not to attempt to keep the ability to distinguish these tones? I’ve heard contradictory things about the effect of learning multiple languages. Does narrowing help you learn a single language more effectively as a child? Does lack of narrowing impede language learning?
Monica and Tina’s comments sound very much like the reaction I had to the Wang and Kuhl paper — simply being able to learn the tones is very different from the undertaking of learning an entire language. This task was, to my mind, much more akin to an auditory learning task. I would be interested, for example, to see how children and adults performed on a similar but more “linguistic” task — perhaps one incorporating syntactic structure, such as learning tone sandhi from training items and applying it to novel forms.
In the article, Wang and Kuhl say their study shows evidence that “language learning is not a strictly timed developmental process with rigid cut-off periods.” However, as we know, child language learning has very little to do with explicit training — indeed, children in the critical period often ignore explicit linguistic training. Researchers have to be very careful when asserting that experiments with explicit training give insight into critical period language learning.