#### Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!

Options

# Introduction: Anna Mai

Hi All, I'm a graduate student in linguistics studying representation in phonology. A fellow graduate student introduced me to category theory through the DisCoCat model of semantics a couple of months ago, and I've been excited to learn more ever since. My hope is to be able to concretise the mathematical relationship between rule-based phonology and constraint-based phonology, and I think category theory might be useful tool to that end. I'm looking forward to learning with everyone here!

• Options
1.

Hi Anna, your work sounds interesting! I'm dealing with math and music (including some references to singing voice). Categories are a really helpful tool. I hope we can share some research ideas!

Comment Source:Hi Anna, your work sounds interesting! I'm dealing with math and music (including some references to singing voice). Categories are a really helpful tool. I hope we can share some research ideas!
• Options
2.

Welcome to the course, Anna! Where are you a grad student, if I may ask? And what's "representation" in phonology - the problem of representing phonemes using symbols, or something like that?

I didn't know phonology had rule-based and constraint-based versions. It sounds interesting. Right now I'm just dipping my toe in the sea of linguistics, since I've been trying to understand Lambek's work and the DisCoCat model. (I've included some easy links for other people who might read this and wonder what I'm talking about.)

Comment Source:Welcome to the course, Anna! Where are you a grad student, if I may ask? And what's "representation" in phonology - the problem of representing phonemes using symbols, or something like that? I didn't know phonology had rule-based and constraint-based versions. It sounds interesting. Right now I'm just dipping my toe in the sea of linguistics, since I've been trying to understand [Lambek's work](https://www.cs.cmu.edu/~fp/courses/15816-f16/misc/Lambek58.pdf) and the [DisCoCat model](https://golem.ph.utexas.edu/category/2018/02/linguistics_using_category_the.html). (I've included some easy links for other people who might read this and wonder what I'm talking about.)
• Options
3.

Maria, your work looks super cool! I like the concept of gestural similarity you talk about between flute & vocal vibrato-- It made me think about similarity between speech sounds, especially in the context of dialect / accent variation. For example, I've heard the word 'brother' pronounced with a v, z, or th sound by different speakers, but I think typically a single speaker is consistent in their use of a particular variant unless they are code-switching. A Londoner who uses the v-variant has z in their phonological inventory, and a French speaker who uses the z-variant has a v in their phonological inventory, but each speaker makes a different choice when modelling the th sound in the word 'brother.' I wonder what aspects of articulatory and acoustic similarity influence this kind of choice, and I wonder if thinking of two speakers with different phonological backgrounds (i.e. native Londoner, native French speaker) as different instruments performing the same gesture (vibrato / the English word 'brother') in your framework of gestural similarity might lead to interesting insights.

Comment Source:Maria, your work looks super cool! I like the concept of gestural similarity you talk about between flute & vocal vibrato-- It made me think about similarity between speech sounds, especially in the context of dialect / accent variation. For example, I've heard the word 'brother' pronounced with a v, z, or th sound by different speakers, but I think typically a single speaker is consistent in their use of a particular variant unless they are code-switching. A Londoner who uses the v-variant has z in their phonological inventory, and a French speaker who uses the z-variant has a v in their phonological inventory, but each speaker makes a different choice when modelling the th sound in the word 'brother.' I wonder what aspects of articulatory and acoustic similarity influence this kind of choice, and I wonder if thinking of two speakers with different phonological backgrounds (i.e. native Londoner, native French speaker) as different instruments performing the same gesture (vibrato / the English word 'brother') in your framework of gestural similarity might lead to interesting insights.
• Options
4.

Hi John, I'm at UC San Diego, just down the road from you :) I would say questions of representation in phonology concern what the basic unit of phonological analysis is. The typical shape of a representational question in phonology might be: Which set of articulatory/acoustic/abstract features are necessary to uniquely define all the phonemes in the phonemic inventory of Language X? or Should contour tones in Language Y be represented as a concatenation of level tones or as a single unit in themselves? For me, questions of representation in phonology also concern the level of abstractness of phonological units of analysis (i.e., features, phonemes). For example, most/generative phonologists work from the assumption that spoken words are derived from abstract, underlying forms of those words, either by rules or by optimally satisfying a set of constraints. In this way, if you come across the novel object 'wug' and someone asks you what the plural is, you know how to derive (via rules or constraint satisfaction) that the correct plural ends with a voiced alveolar fricative (z). However, in lexicalist views of phonology, knowledge of a network of related spoken forms (i.e. dog, dogz, pen, penz, law, lawz, cat, cats) allows the speaker to arrive at the correct form of novel items by analogy, without making reference to abstract forms. Right now, I'm doing work in generative phonology to connect rule-based and constraint-based formalisms, but I'm most interested in what the brain does and being able to relate what the brain does to what phonologists have observed in the languages of the world.

Comment Source:Hi John, I'm at UC San Diego, just down the road from you :) I would say questions of representation in phonology concern what the basic unit of phonological analysis is. The typical shape of a representational question in phonology might be: Which set of articulatory/acoustic/abstract features are necessary to uniquely define all the phonemes in the phonemic inventory of Language X? or Should contour tones in Language Y be represented as a concatenation of level tones or as a single unit in themselves? For me, questions of representation in phonology also concern the level of abstractness of phonological units of analysis (i.e., features, phonemes). For example, most/generative phonologists work from the assumption that spoken words are derived from abstract, underlying forms of those words, either by rules or by optimally satisfying a set of constraints. In this way, if you come across the novel object 'wug' and someone asks you what the plural is, you know how to derive (via rules or constraint satisfaction) that the correct plural ends with a voiced alveolar fricative (z). However, in lexicalist views of phonology, knowledge of a network of related spoken forms (i.e. dog, dogz, pen, penz, law, lawz, cat, cats) allows the speaker to arrive at the correct form of novel items by analogy, without making reference to abstract forms. Right now, I'm doing work in generative phonology to connect rule-based and constraint-based formalisms, but I'm most interested in what the brain does and being able to relate what the brain does to what phonologists have observed in the languages of the world.