29 April 2008

WTF, Cupertino???

i have just stumbled upon a most peculiar Cupertino effect trap.  i am writing a paper for my phonology ii class on raddoppiamento in Italian.  it's a continuation of some previous work that i did last semester.  when writing my previous paper, i was running Mac OS X 10.4.  the system-wide spellchecker had no clue what this big, long, Italian word was, and offered up no possible corrections.  to avoid lots of little red underlines, i added the word to the dictionary and went merrily on my way.  since then i suffered a hard drive failure, which had two consequences: 1) the dictionary "forgot" that raddoppiamento is a word, as far as i'm concerned and 2) i've upgraded to Mac OS X 10.5.  no problem, i just have to re-teach the dictionary the word.


and that's when this happened:


whaaaaaaaaat?  where did that "d" come from??  i fired up dictionary.app (which presumably is the same dictionary resource used by the spellchecker) and, no surprise, raddoppiamdento is decidedly not in the dictionary either.  nor, as far as i can tell, is any substring that would fool the spelling suggestions algorithm (which i know for a fact does sometimes produce novel suggestions, especially when two words are accidentally conjoined) into thinking that raddoppiamdento was not only a possibility, but in fact preferable to raddoppiamento.

just for confirmation that i wasn't going crazy, i turned to Google, which seems to know which is the real word and which is the imposter (what the hell is going on, imposter just got flagged, despite the fact that dictionary.app gives both imposter and impostor):

um, yes?

if anyone has any hypothesis whatsoever as to what the great "improvement" Apple made to its spellchecker algorithm such that it produces these shenanigans, please leave a comment.

26 April 2008

i can haz fotos?

seriously, flickr? seriously?



a quick handful of refreshes gave me greetings in languages that i expected: Portuguese, Tagalog, Arabic, French, and English.  although i guess flickr isn't totally hanging on formality—both "hello" and "yo" are given as English greetings.

just as long as they don't publish an entire site localization in lolcat.

24 April 2008

on the origin of sign languages?

with sign languages, the sign is not necessarily arbitrary.  which can lead to hilarity.


i stumbled on this from a "don't click here!" link on mgoblog today.  i'm a little embarrassed to say that i enjoyed it quite a bit, and there are several other related videos on youtube.


the performer is David Armand, in character as Johann Lippowitz.  this was recorded in 2005, so don't get upset with me for being behind on my memery.

this also reminds me of the "fake language" videos going around youtube last year (which seem to have all been taken down since).  in those, the best you can do to fake a language is try to make up nonsense words that accurately reflect the phonology and prosody of the language, but with fake sign language you can actually put some semantic content into the mix.  just think about it: it's possible to "translate" an English song into fake sign language, but it's impossible to translate ASL speech into fake English!

23 April 2008

everything's...segmented differently in texas?

i've been watching a lot of hockey on the versus recently (probably to the detriment of my academics, but it's the playoffs!)  as a result, i've seen the same eight commercials dozens and dozens of times over the past couple weeks.  mostly they're unremarkable, but one stuck out.  it's an ad by the Texas tourism board or something, and what caught my attention the first time i saw it is their tagline, which is displayed proudly at the end of the spot:

i meant to write on "split another" last fall, but never got around to it, after i realized that there's already been quite a bit done on it (which i somehow can't find now via quick google search; odd).  the fact of the matter is that i didn't have anything too exciting to say.  my ideolect has split another—i say "a whole 'nother" without exception, so when i saw the Texas ad, it seemed wrong to me.  i wonder how many Texans actually say "a whole other."  my bet is that it's very few, and the only reason we got "whole other" in the tagline is because an editor at some point in the creative process nixed a spelling with nother.

so why prefer nother at all?  this may be just due to the influence of orthography getting the best of me, but it seems that if you take it upon yourself to reconstruct the unmodified phrase by removing whole, the results are better with nother.
a whole other --> *a other --> ?an other
a whole 'nother --> a 'nother --> another
when you get to the intermediate stage a other, allomorphy kicks in and you get an other, but does it get to go on to fully become another afterwards?  it seems to me that a 'nother is more likely to reform into a single word.  like i said, i realize this is relying on the spelling somewhat, but i have a feeling that unless i'm totally artificially imposing it there is some difference in the prosody of an other vs. a 'nother.

regardless of whether this little reconstruction exercise was sound or not, i still think that Texas should have stuck with 'nother.

21 April 2008

LL made me do it

look at this here, a new post.  back from the dead.  Eric Bakovic posted a call for comment on a couple sentences over on Language Log, and I couldn't resist putting in a quick two cents.  i'm sure his comments will be more in-depth, but here's a quick shot before i head to bed.


the sentences in question are the following:
I’ll never forget how he must have felt. (overheard)
Aren’t you glad you archived instead of deleted? (over-read)

1. I'll never forget how he must have felt.
yeah, this is definitely semantically odd.  the intent and necessary pragmatic context seems pretty obvious to me, and i doubt that i would do a serious double-take at hearing it.  the oddity lies in the complement of forget, "how he must have felt."  there's no syntactic trouble, since similar constructions like "i'll never forget how that sunset looked" are unproblematic.  there is a semantic disconnect, though.  the predicate forget includes a presupposition, namely that whatever is to be forgotten was at some point known or experienced by the subject.  this is not realized in the complement "how he must have felt," because the experiencer in the complement is not the same as the experiencer of forget (the forgetter, if you will).

nevertheless, i can come up with two scenarios where this sentence could be used felicitously.  the first, and i think more plausible, interpretation is that the sentence means "i will never forget witnessing/experiencing his outward reaction of how he felt."  for example, someone just dropped a full cup of coffee all over himself; he makes an unforgettable yelp and a look of shock comes over his face.  it's an event not to be forgotten, and furthermore the onlooker has a pretty good idea of how a hot cup of coffee in the lap feels.  the second, and a little more strained, interpretation is that the forgetter has had a strikingly parallel experience, and therefore "how he must have felt" actually refers to the forgetter's experience, which is in keeping with the requirements of the predicate.  an example scenario: "his brother died of cancer last month.  i'll never forget how he must have felt (because my brother died of cancer too)."

2. Aren't you glad you archived instead of deleted?
maybe i've just been organizing my email too much recently, but this seems perfectly normal to me.  of course, if i look at it, i can see the object drop that's going on, which is uncommon for english.  ew, i thought simple reconstruction would do the trick, but that just creates more syntactic problems.  taking "your email" as the understood object of each predicate gives
*aren't you glad you archived your email instead of deleted your email/it?
bad, bad, bad.  this is definitely WTF coordination.  the closest grammatical paraphrase with reconstruction would be something like:
aren't you glad you archived your mail instead of deleting it?
the question then is whether this is still perceived as an oddity with truly unaccusative predicates.  are the following weird?
aren't you glad you walked instead of ran?
aren't you glad you whispered instead of shouted?
the walked/ran sentence seems more natural to me; it might have something to do with the fact that the forms are not morphologically parallel.  anyway, there is something about the "instead of..." clause that wants a tenseless verb form.  i'll leave it to Eric to flesh that out further.


more posts to come soon in the future?  maybe.  because i quit blogging when it starts to feel like work.  but when i have papers to write, blogging definitely becomes a preferable form of writing.  (huzzah, procrasti-tasks!)