Please post your comments and questions on the class Monday October 21st here.
3 thoughts on “Phonological learning and concept learning – wrap-up and modeling”
Josh
I’ve been thinking about how various types of conditioned phonological variation relate to one another, what counts as conditioned variation, and how conditioned variation relates to free variation. My concise questions are at the bottom, but here is my train of thought:
Just as allophonic alternations are conditioned by a phonetic/phonological context and sociolinguistic variation is conditioned by social context (e.g. register, class, gender), could phoneme learning be thought of as learning of phonemic alternations conditioned by meaning? You could then add free variation (noise) on top of each of these alternations, with the most applied to sociolinguistic variation, less to allophonic alternations, and a very small amount to phonemic alternations. I imagine the amount of noise could be coded in the constraint weights of the MaxEnt-OT grammar.
Is it wrong to (1) characterize phoneme learning, allophonic alternations, and sociolinguistic variation as lying on the same continuum (i.e. phonological learning), (2) lump the three phenomena mentioned above into ‘conditioned variation’, and (3) characterize conditioned variation as signal and free variation as noise?
I don’t quite get the details of what you’re laying out in the second paragraph, but it doesn’t strike me as right to characterize sociolinguistic pressures as on a continuum with other types of linguistic pressure.
Take, for example, production difficulties, which could lead to allophonic variation, and compare this to some type of sociolinguistic pressure. An example of this might be early adopters’ decision that it’s funny to pronounce vowels a certain way. It seems far-fetched to suppose that these could be on the same scale.
I agree with Megan, Josh, that it really makes no sense to me to place sociolinguistic variation on the same continuum as the others.
In my opinion, there are so many more factors influencing sociolinguistic variation than the things you mentioned (which I also agree with), and, unlike phonetic/phonological context, people can have many multiple sociolinguistic contexts. For example, I imagine multi-racial or otherwise multi-cultural people frequently code-switch. While the factors you mentioned (class, gender) are influences that are probably more immutable to social context, I see most social factors more as a matter of choice (e.g., you find yourself intentionally using non-rhotic speech when home in working class Boston) compared to something like phonemic alternation (which people do without realizing). I suppose it comes down to me believing that sociolinguistic pressures upon language can be explicit choices, whereas phonological context and pressures are implicit.
I’ve been thinking about how various types of conditioned phonological variation relate to one another, what counts as conditioned variation, and how conditioned variation relates to free variation. My concise questions are at the bottom, but here is my train of thought:
Just as allophonic alternations are conditioned by a phonetic/phonological context and sociolinguistic variation is conditioned by social context (e.g. register, class, gender), could phoneme learning be thought of as learning of phonemic alternations conditioned by meaning? You could then add free variation (noise) on top of each of these alternations, with the most applied to sociolinguistic variation, less to allophonic alternations, and a very small amount to phonemic alternations. I imagine the amount of noise could be coded in the constraint weights of the MaxEnt-OT grammar.
Is it wrong to (1) characterize phoneme learning, allophonic alternations, and sociolinguistic variation as lying on the same continuum (i.e. phonological learning), (2) lump the three phenomena mentioned above into ‘conditioned variation’, and (3) characterize conditioned variation as signal and free variation as noise?
I don’t quite get the details of what you’re laying out in the second paragraph, but it doesn’t strike me as right to characterize sociolinguistic pressures as on a continuum with other types of linguistic pressure.
Take, for example, production difficulties, which could lead to allophonic variation, and compare this to some type of sociolinguistic pressure. An example of this might be early adopters’ decision that it’s funny to pronounce vowels a certain way. It seems far-fetched to suppose that these could be on the same scale.
I agree with Megan, Josh, that it really makes no sense to me to place sociolinguistic variation on the same continuum as the others.
In my opinion, there are so many more factors influencing sociolinguistic variation than the things you mentioned (which I also agree with), and, unlike phonetic/phonological context, people can have many multiple sociolinguistic contexts. For example, I imagine multi-racial or otherwise multi-cultural people frequently code-switch. While the factors you mentioned (class, gender) are influences that are probably more immutable to social context, I see most social factors more as a matter of choice (e.g., you find yourself intentionally using non-rhotic speech when home in working class Boston) compared to something like phonemic alternation (which people do without realizing). I suppose it comes down to me believing that sociolinguistic pressures upon language can be explicit choices, whereas phonological context and pressures are implicit.