So Close

I am honored to be in the RSS feed of the writer of Polyglot Conspiracy, a quite wonderful linguistics blog I enjoy reading. Here is the story of how I learned that, and one other little fun tidbit getting me just a hint closer to the big dog.

In the most recent Language Log post as I write this, Mark Liberman writes at the end that

[...] we can also learn on the web that PsyBlog is "the Language Log of psychology, maybe"

The thing I noticed here when you hover over "PsyBlog" is that it goes to Polyglot Conspiracy, but the search parameter is "catafora." I noted that this was the spelling I use for my website right here, and the normal English spelling is Cataphora, as it is in my protocol. I clicked the link, and noticed that two posts come up, one I mentioned before, and one I missed (!).

Two other points: First, it was lovely to see the connection again, and maybe some Language Log legionnaires will follow that link... and then the other link. Second, and most importantly, I need to wonder what was the basis for using that as the link? I would like to think that this is a subtle hat tip to my little ruminations here, and that ideally Mark Liberman takes a glance every so often in this direction. It's also possible that he just wanted a search parameter that was (more or less) unique to the page. There's a problem-set with that theory though, one, is that there is no problem just using a regular link to the page, and the second is that the mention of my site is not particularly close to the thing they were trying to reference on Language Log. On top of that, searching Psyblog will yield /only/ the desired result, and seems to be the natural choice. Either way, I should probably class it up here just to be safe.

Also! As of this post I am 2nd on google for a search of "catafora," and am mentioned in the 5th result if you search "paratactica." I'm the first result for "paratactic cataphora" sans quotes, and the first of just three (again, as a reference) for "catafora paratactica." I'm not really sure how proud of this I should be, but hey, it's something!

Phonetic Transcription For The Masses

The other night I was chatting about work and other business with this fellow I met, and linguistics came up of course, and he seemed to be familiar with several aspects, particularly Phonology. Apparently, he did some acting work, and he had to learn phonetic transcription, so he would be able to imitate dialects, and presumably to pronounce other things he might not be sure of. He gave a few casual names for characters, but sadly I don't remember them. Phonology might not be my favorite area of linguistics, but I guess he found it quite unpleasant.

Unpleasant as it might be, I'm always glad to see people learning about phonetic transcription, because I really do think that it's something useful to know. Especially because English does have such a poor relationship between spelling and pronunciation (well, if you learn a whole lot of rules about etymology and such, it's pretty predictable), I find it to be relatively practical. It's always the worst when I'm trying to figure out a pronunciation for a word and I need to read the dictionary format, or something like gahb-LEE-uh or something that I can't quite figure out the nuances of.

At the university I am at, they have elective seminars on IPA in case students want to brush up or get ahead of the curve, but maybe one day they'll work them into some kind of general classes. It's also a shame that most classes, for undergraduates at least, seem to spend a class or two early (or even more!) on going over IPA. It just seems wasteful; they should spread that joy around.

Spanish Syllable Structure

I heard that Kenneth Pike denies a possible syllabification of Spanish words like "transcribir" because they cannot subscribe to the necessary rules of syllabification, which are basically:

1. Onset First

2.Consistency with word boundary rules

3. Consistency with sonority scale

Specifically, the word boundary rule, in that worlds like transcribir could be split any of the following ways:

tra.nscribir (i)
tran.scribir (ii)
trans.cribir (iii)
transc.ribir (iv)
transcr.ibir (v)

Alright, (i) (iv) and (v) we can assume are ridiculous, because they have massive or strange clusters, very technical. The argument is that tran.scribir has a viable coda in the first, but not a viable onset in the second, and that trans.cribir has a viable onset in the second, but not a viable coda in the first.

My instinct, and the answer according to Jeff Stokes, is that you go with the latter (iii), because cr is a possible onset, and that scr is not. It's not perfect (if you assume the syllabification rules are truly universal, then this doesn't seem to quite work. I would argue personally that there is possibly some tiering involved, and that rule 2 is actually two rules, where the onset must be viable, and then a secondary rule would be that /if/ you can make the coda there work, great, if not, work with what you have.

I would predict that most Spanish speakers would syllabify it, without particular hesitation, as trans.cribir. It looks to me that the only potential problem words are words that would have -m in the coda, or -ns in the coda, both do not occur at word finals. So, perhaps there is a conventionalized mechanism for resolution, as there is not a widespread amount of this problem.

Catalan Rap

A Catalan speaker told me that Catalan rap is better than Spanish rap because Catalan tends to have shorter words. I might've missed something about some phonotactic aspects as well, but the gist was mainly for the former statement.

It's probably the case that languages with super long words don't lend themselves well to rap, but I don't know how many people would be rapping in Mohawk, or Chukchi anyway. It's probably different though, either way, because most poetry and such tends to be dependent on the language anyway, for metrical quality, versatility of rhyme, and such.

Old Norse poetry was always interesting because it really did seem reminiscent of modern rap and hip hop to me, weaving repeating sounds into sentences, rather than having ending rhyme. Japanese poetry often works differently as well, being mora-based, rather than having the same type of rhyming that one would expect from English poetry.

Perhaps Catalan rap can more easily resemble the American variety? I doubt that Spanish is so ill suited for the style that it just cannot produce anything worthwhile, they probably just need to change it up a little bit, and that's one of the things that makes vocal stylings so interesting in verse.

Alternatives To IPA Characters

I have heard complaints about problems with IPA, mainly for on-the-fly transcription, or how to actually produce the characters. In handwriting, several characters can be ambiguous {s} and {ʃ} can look very similar if not made obvious, {m/ʍ} and {r/ɾ} are other examples. There are solutions if you are careful, or make habit of writing characters a certain way. Generating {ɾ} with a serif along the bottom (like in r, but without the upper left hook) is quite common, to the point where a room of about 12-15 linguists looked at me with sideways tilted heads when I insisted that the character actually does not have a line at the bottom. Alternative symbols are quite common also, APA {š} is used for handwriting IPA {ʃ} to prevent confusion.

Sometimes I'll use a shorthand or nonstandard diacritics to redundantly decide characters, but not too often, since I'm usually not in too much of a hurry or likely to confuse what I'm writing. For quick work, I could see why it might be a problem. I just discovered this page on Wikipedia about non-standard characters, and many characters include so-called "secondary" features of articulation. Sinologists, Japanologists, and Koreanologists use a lot of these characters, and the common feature among them is that the symbols they use tend to be much simpler, and probably easier for notating a symbol that has phonemic quality. {ʮ} is supposed to be represented with {z̩ʷ} in IPA, which seems a bit tedious. If secondary articulation is a primary feature, why shouldn't it be represented in a single character? Already, most symbols already represent several qualities abstractly. {p} for example represents three features, a voiceless, bilabial, plosive, sound. If a language has a very productive phonemic distinction of aspiration, it seems useful to perhaps distinct symbols for it, especially if there are many other qualities that are commonly associated with phonemes.

I do feel that for reading purposes, and clarity of transcription, especially for a wider audience, say linguistics classes, or pronunciation guides in books/online, standard IPA is ideal, everybody can figure it out, and probably quicker than referencing all of the (non-)standard characters used by a specific group.

There aren't any languages I work with (that I can think of) where this would particularly help out, but some languages that have high amounts of aspirated, or ejective, or labialized consonants that are distinct phonemes, or they even don't have the "basic" form. If a language has the phonemes {kʲʰ/kʷʰ/kʼʲ/kʼʷ} and not {k}, what sense does it make to keep writing {k} over and over? If the crucial distinctions are aspirated versus ejective and labialized versus palatalized. Perhaps there could be symbols for aspirated and ejective, such as ʞ and ᴋ respectively, and then use the diacritics on them, or some other variation, to keep it visually simple and more representative of the phonemic aspects of the language.

I would like to think that anybody who had this problem has probably worked out there own solutions, but it might be nice to see endorsed versions of such consonants as "notational alternatives," even exclusively. There already plenty of (arguably) redundant diacritics, so I see no reason why a new paradigm could be set up for people to standardize their notes for passing around.

Keyboards, Left-handedness,

I mentioned recently that I would comment on the frequency of keys typed on my laptop (that was pretty much the whole post, so why bother linking?) and there is a little bit of a story too it, but be forewarned, it's not a very good one.

Regularly I use a whiteboard to do my linguistic analysis, or for whatever things I need to be accomplishing that day, I find it to be a great way to help my thinking and work with data. So much so that I carry a smaller board and markers around with me (not everywhere, or course) but for when I might wind up sitting someplace for a time or know I will be working. Now, the finer nuances of how white boards work escapes me, but basically, when you wipe away the marker, it has to go somewhere, right?

Well, apparently I use my board enough that little bits get on my fingers and usually it is of no consequence to anyone or anything. After a while though, I noticed that my lovely white keyboard and touch pad are turning greenish, and not just the yellowish that is to be expected from MacBooks that were bought during a certain period of time.

The keys that are turning green are {a, s, e, r, t, h, n, u, i, o} and to a lesser extent {y, l} and where my thumb would sit on the space bar and right shift. Those probably seem to be pretty regularly used letters, and I think they coincide roughly where my hand sit on the keyboard. Something like:

LP - on {a}
LR - on {s}
LM - on {e}
LI - on {r, t}
LT - on space bar, near c/v break

RT - on {n}
RI - on u/i break
RM - on i/o break
RR - on o/p break
RP - on right shift

It's not too surprising that it worked out that way, that's right about where your fingers are supposed to line up. My keyboard looks something like:

QWERTYUIOP
 ASDFGHJKL
  ZXCVBNM

The ergonomic keyboards with the break in the middle always threw me off because I am right handed, but left hand dominant, so (well, I'm not sure if it's causative or not, but) I type much further over with my left hand that is intended, specifically characters {y, h}.

Incidentally, the department I'm in has an unusual amount of left-handed linguists. Less than 10% of the adult population is expected to be left-handed, but it's easily over 20% of faculty, and I don't actually know the orientation of a good number. Weird stuff, eh?

Why I Am Bad At Phonological Analysis

Just a quick note about phonological analysis, I almost never think to use a sound that doesn't appear in a language as the template for a distribution of different phonemes. I've seen the contexts where it really seems to work well, where there is an unusual gap, but in a lot of cases it feels like drawing a line around the exterior of the maze, and claiming that not only did you figure out the maze, but you have the most efficient method.