Throughout most of my career as a linguist, I have enjoyed doing descriptive work. Practically, that means I describe what is found in languages, rather than prescribing what should be in a given language (as a teacher might). But descriptive work can also find itself set against more theoretical work. That is, figuring out what makes up language more generally, as opposed to how things work out in a particular language. So far, I have been content to figure out what’s going on in a particular language, since that’s my concern: I want to see particular people groups able to read and write. While studying for my M.A., a teacher once described theories as tools in a toolbox. One may be useful for solving one problem, and another may be useful for solving another, so we use the tool that is useful, then set it down when we finish that problem.
Needless to say that for some linguists, this is a horrifying idea. For some, the whole purpose of doing linguistics is to understand the principles of language generally, not the details of a given language. The desire is to understand the language functions of the human mind, not how a particular subset of humans uses it. But there is another reason the concept of theory I picked up at Oregon horrifies some: for many, a linguistic theory is not just helpful (or not) in a particular instance, it is supposed to be True (note the capital ‘T’; this is not the ‘my truth’ v ‘your truth’ kind of truth). So if a theory is supposed to reflect what’s actually going on in the mind, then it shouldn’t be dropped in favor of another theory to deal with the next problem.
During my last several years at UT Arlington, I’ve had plenty of chances to interact across this divide; few people here are interested in descriptive linguistics. And I often get the question, after presenting a good (IMHO) description, “so what does this mean (i.e., about language in general)?” And “it means that this is how these people use language” is not a sufficient answer.
I’ve grown more competent in developing more theoretically interesting conclusions (i.e., interesting to those interested in theory development, not only theoretically interesting…:-)), but I have remained mostly uninterested in developing theory, nonetheless. So I continue to do my descriptive work, showing what’s in a particular (set of) language(s), and I throw in more theoretically interesting conclusions for those more interested in theory.
Until recently, when I got hit by something I hadn’t expected. What would happen if you were working on a complex job, and you realized you needed two different tools to make it work? What if those tools were not designed to work together? What if you could only figure out a couple ways of making them work together at all, and any one of those ways limited the operation of each of the two tools? This is something like what I’ve run up against.
For over a decade, I’ve been using a theory of what makes up tone that has been described in The Geometry and Features of Tone, by Keith Snider. It builds on earlier work in Feature Geometry (how the bits that change one meaningful sound into another interrelate) and Autosegmental theory (how things like consonants and vowels interact with things like tone and stress). Perhaps the most important contribution of GFT is that tone is made up of two features that distinguish sound, and each has its own contribution to how a word with tone is pronounced.
More recently, as I’ve interacted more with languages with Consonant-Tone Interaction (see this blog entry for more on these consonants), I’ve been using a theory ([L/voice], from Bradshaw 2000) that says that low tone is intrinsically bound to voicing (when your voice box vibrates, more or less). This theory has proved helpful as well, though perhaps in a smaller selection of languages. But in any case, those are the languages I’m working on now, so this tool is in my hand, as it were.
But as I was writing up some of my work based on these two theories, I realize that GTF says that low tone is really two features (low tone and lower register), while the [L/voice] theory says that low tone is intrinsically linked to your voice box. But neither theory explicitly addresses the other, despite the fact that I need to use them both at the same time. I originally thought that I would sit down one afternoon and sketch out a number of different ways to possibly make them work together, but I could only come up with two. The longer I beat my head against this problem, the more I found arguments against any other way of making the two theories fit together. That is, voicing must be bound to either low tone, or lower register. Period.
I had been assuming that I would be able to make both tone and register features bound to voicing, but that simply can’t work, at least not without making radical adjustments to either or both theory.
Anyway, I’ve been presenting this information to my committee, and for the first time in my career as a linguist, I’m seeing not only very practical implications of our theoretical assumptions, but I’m seeing things that need to be figured out on a theoretical level, and I might be the person best placed to do it. Which is to say that I entered this Ph.D. program to help develop a number of languages in eastern DRC in a more strategic way, but I may well end up with something of broader implications (with whatever theoretical claims I make informed by, and informing, the description). But maybe that’s why they call this degree Doctor of Philosophy, and not Doctor of a couple things I wanted to write about.
Anyway, as I make this turn, I’m looking forward to see how these musings can benefit a larger set of people groups, which remains the point of what I’m doing at all. I remain committed to linguistic description, and trust that the languages I describe in my dissertation will have good tone work represented there, at least, and that the people who speak those languages will be better able to read and write as a result of it. And as a result of that, that they would have better access to information about this life, and the next.
This work by Kent is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International
0 Comments