Noncompositional

In the news today is China’s return to the discussion table with the US on human rights. Unless you read the Washington Post headline. Ah, what a difference a little word makes.

Read more »

This is not what I promised in an earlier post, but the topic is basically the same.

Now, I’m not widely read in (epic) fantasy novels, though it is often my preferred genre. Nonetheless, I think I can make a tentative generalization, which is that in any fantasy world where some either exotic or ancient race speaks a language unintelligible to contemporary folk, then there is somehow an inordinate amount of polysemy, connotation, or complexity to the words and sentences involved. Or, alternatively, it is approximately the same amount as in the modern language (usually “English”), but it is highlighted in such a way as to make it seem rather different from anything that might be familiar.

Just to pick a couple examples from Tad William’s book Shadowmarch (which I’m currently in the middle of reading; no spoilers):

The lady’s high house was called Shehen, which meant “Weeping.” Because it was a s’a-Qar word, it meant other things, too–it carried the intimation of an unexpected ending, and a suggestion of the scent of the plant that in the sunlight lands was called myrtle–but more than anything else, it meant “Weeping.”

…all the way down to the thrice-blessed fence that the mortals called Shadowline, and that the Qar themselves called A’shish-Yarrit Sa, which meant “Storm of Silence,” or, with a slightly different intonation of voice or gesture of the hand, “White Thoughts.”

I suppose that the second example is supposed to be significantly different from, say, tonal languages like those found in China and Africa. In this case, we are supposed to understand that the two meanings for A’shish-Yarrit Sa are somehow semantically related (in some deep way incomprehensible to mere humans). Either that, or somehow we’re dealing with a pun, or maybe just some philosophically interesting near-homophony which, perhaps, native speakers of Qar don’t even care about.

Now, this sort of thing is not in and of itself completely horrible. But for me, without an actual system apparent behind the words and their meanings, which could, with time, be discerned by the reader (and yes, this requires many more tokens in the books), it just seems…well, laughable.

Today’s Teen Jeopardy’s final question/answer was (paraphrasing)

This is the only [US] state that, when written correctly, has a diacritical mark [see below]

After going through my inventory of diacritics and possible parts of state names other than the proper name part (as in The State of California, or something like that), I came to the conclusion that it must be Hawai’i. And indeed this is the response Alex was waiting for.

It’s really too bad, because as far as I can tell, the ‘okina should be, and usually is, considered a separate character (a “letter”), expressing the glottal stop. It is not a diacritical mark, which intuitively is supposed to alter the pronunciation of a letter, not indicate a separate sound. Of course there are many cases where an a diacritic in fact does something rather more (e.g., the cedilla in several Turkic languages). And IIRC there are orthographies in which a true diacritic is used to mark glottal stops. But the ‘okina is not (in) one of them.

[edit: Some websites report the exact final Jeopardy answer as: "It's the only state name that when spelled officially contains a diacritical mark."]

It seems like one of those things that keeps getting “casually discovered”–that is, that I hear mention of at least once a year–is the result of applying the agentive -er suffix to particle verbs like pick up and clean out. What’s interesting is that the most common result (according to some study [studies? -I've only seen a 1978 manuscript by Moira Yip cited wrt this issue]) is picker upper and cleaner outer. There are some interesting observations made by David Mortensen in this post of his now archive-only blog.

But how about this: what do you do with take advantage of? Well, today, I somewhat consciously produced taker advantage ofer. A quick search on Google reveals:

Kira I Am: does it say that kkira is the numbe rone drunk girl taker advantage ofer? link

Yes! This seems to be a one-line extract from some sort of IRC or similar chat session. I especially like the semantic undergoer expressed as a pre-modifier.

In any case, this seems like an interesting test case for models of realizational morphology, as there is more than just a head verb and a particle. Exactly which bits of the word are eligible for the morphological process?

From time to time, I’ve observed that I mishear /p/ as /h/, in particular in initial position. It doesn’t happen all the time, but occasionally it does, and it hinders understanding (it might be that it happens more than I notice it, but in many cases it doesn’t really matter for parsing). Now, me and my office mate have a running joke that I do this because of my experience with Japanese, in which /h/ and /p/ alternate in certain morphophonological environments (in addition to h/b and b/p alternations). Historically this is due to intervocalic weakening of /p/ to /h/ via some intermediate steps.

One day, another colleague was walking by, and during a conversation with him, he evidently misheard one of our /p/s as as /h/ — and, crucially, he is very familiar with Japanese. So the theory lives on…

I wonder if others have experienced anything like this (assuming that there is a “this” to be experienced).