After having fielded quite a number of questions about this homework that I’ve labeled “midterm”, I have come to the conclusion that it really wasn’t very well constructed.
I did have some things in mind, but I also was not expecting in principle that everyone would have the same thoughts I did about it. However, let me just say a couple of quick words about the pieces of it.
Questions #1 and #4 were intended primarily to get at the fact that the proposed universals are statistical universals, they’re tendencies. No single example can ever by itself disprove them. Perhaps I kind of just handed you the answer I had in mind to #1, but so it goes.
Question #2 talks about syllables, but what I really had in mind was this: Parker talks about things in terms of “initial” and “non-initial” and I wanted to explore a slightly more abstract interpretation. On the handout about artificial languages, I provided the terminology for syllable components: onset, nucleus, coda. Though we know little else, we know that an initial consonant is going to be the onset of a syllable, so I was asking you to think about the propensity of /h/ to be initial in terms of what part of a syllable it would have to be, and then from there to think about whether /h/ and /?/ might each preferentially like to be a particular part of a syllable. Here too, I feel that I haven’t completely given away the answer that I had in mind, but pretty nearly. You can just kind of fill in the blanks, and you’ll have the idea that I had. Which is not to say that it is necessarily the right idea, there are further things one could check to see if it is.
Questions #3 and #4 ask you to suspend beliefs about the world and assume that the world is actually such that the universals that Parker reports are not in fact true if you took all possible languages into consideration. Then #4 asks: in that situation, what could have happened that would lead Parker to the findings he had? And #5 asks: in that same situation, if you had information about all the languages in the world, what would you find? It’s almost a silly question. The premise is that Parker’s generalizations weren’t actually true, and so you’d expect to find here that, well, Parker’s generalizations weren’t true—but more specifically, what patterns (or, I suppose lack of patterns) would you find in the whole data set, if they aren’t the ones that Parker reports?
Questions #5, #6, and #7 go together; question #5 is addressed in the paper itself, basically. Question #6 asks you to take a contrary view, and think of a possible alternative reason that you might find that words for ‘yes’ tend to have laryngeal consonants. What I had in mind here was that this correlation would exist, but it would not be because laryngeals are particularly appropriate for affirmative meanings, but that they correlate more by accident. But here too, I think there are probably a number of different things you could suggest. Whatever you suggest, #7 asks how you could tell if your alternative explanation is the right one. What would you have to look at, what would you expect to find?
Question #8 is just a lousy question, unclearly formulated. But all I really wanted you to think about is what bearing the universals Parker proposes have on languages that have only /h/ and not /?/ in their inventory. I don’t really have a single answer in mind here, I think that there are things you can say in favor of any of a number of views. I just wanted you to think about the applicability of these universals to a language that doesn’t have both mentioned phonemes in its inventory.
So: Don’t stress too much about these questions—more than anything else, I just wanted to lead you to think about a few of the aspects of Parker’s project about words for ‘yes’ in a little bit more detail. You don’t have to have read my mind to get the available credit for the homework, mainly I’m just looking to have you write things that show you’ve thought about the questions.