Interview with Sylvia Sotomayor
Sylvia Sotomayor has been conlanging since she was 14. She has a B.A. in Linguistics from UC Berkeley. A Kēlen grammar, dictionary, and some texts can be found at http://www.terjemar.net/kelen.php. Notes on the planet Terjemar can be found there, too. She currently lives in Claremont, California, with too many cats and not enough computers.
Sylvia is one of my favorite conlanging people, if that makes sense (for those familiar with the linguistic concept of iconicity as espoused by John Haiman, I’m not using the term “conlanger” here for a specific reason: to emphasize Sylvia’s person-hood). I was glad to have met her at LCC1, ecstatic to see her again at LCC2, and am looking forward to seeing her once again at LCC3.
The nice thing about talking with Sylvia—or listening to this interview, for example—is that there’s so much that’s immediately recognizable. To an extent, this is true about all conlangers, but I in particular find that in many ways, Sylvia and I are on the same wavelength. For example, about John Quijada’s fantastic talk at LCC1, she says (starting at around 25:04):
“John Quijada’s talk was…was like, ‘Oh, that’s right, I know all this stuff; I’m supposed to take all this into consideration’, but do I? No. Of course not… (Laughter.)
“I try to, and I will probably consciously try to a little more now, for…for a while at least, until I forget again.”
Ain’t that the truth! In conlanging—especially in the naturalist school—there are so many variables to keep track of, and, if you’re like me, which of those variables you devote the most attention to seem to be directly related to what you’re most interested in at the time. Then you’re reminded of the fact that, for example, metaphor is everywhere, and if you’re not careful, you’ll unconsciously copy the structural metaphors of your own language, or those you know well. It’s definitely something to keep in mind, but there’s just so much going on all the time…
Another fascinating correspondence is something that I think is generally true of a number of conlangers. In response to Sai’s question about whether or not she’s fluent in Kēlen, Sylvia says (at around 10:03):
“I have basically the structure memorized. … But the vocabulary not so much. So, you know, I can like figure out, okay, this noun would go here, and that noun would go there, and I may not know exactly what they are, but I know all the little things in between.”
Now, granted, there are a few exceptional conlangers who are fairly fluent in one or more of their languages (Sally Caves?), but most of us—myself included—experience exactly what Sylvia describes. For each of my languages, for example, I have a few nouns and verbs I can reliably reproduce, but for the most part, it’s all structure. I know the structures of my languages even better than I know the orthographies—and I love my orthographies!
What makes this most interesting is its relationship to language-learning, in general. In one of my previous lives as a graduate student of linguistics, I served for three years as a teaching assistant for an undergraduate class on first and second language acquisition. Granted, there are a number of competing theories about everything, undergraduates don’t get the full story on anything, and it seems like there’s an asterisk at the end of every fact and/or universal, but when it comes to acquisition, there is a generally agreed-upon tendency amongst ordinary language learners:
- Children are great at learning rules/structures; they’re not as great at learning vocabulary items.
- Adults are great at learning vocabulary items; they’re not as great at learning rules/structures.
Perhaps you’ve seen something like this if you’ve attempted to learn a new language late in life. I, for example, can probably rattle off a hundred words and more in Arabic, Russian and even Middle Egyptian, but ask me to produce a grammatically correct conjunctive in German, and I’m sunk.
If this is true (let’s say it is, for the sake of argument), what can we say about language creators and their languages? We should be able to rattle off hundreds of words in our languages—that’s supposed to be the easy part! And yet, in Kamakawi, for example, I can sooner produce a relative clause that requires the embedded verb to be both passive and applicative than fill out what the nouns and verbs are in that same sentence (I usually end up with the same words in every sentence: “woman”, “fish” and “hug”).
There are a couple of things that can be said about this. First, is there some sort of inherent difference between learning natural language structures and learning invented language structures? If there’s something we do that better allows us to learn the structures of our invented languages, can we isolate it, and apply it to natural language learning? And honestly, what’s the deal with conlang vocabulary? Adults are supposed to be able to learn about twenty new words a day! Perhaps we should start working with flashcards… Couldn’t hurt, I suppose.