There are 7 messages in this issue.
Topics in this digest:
1.1. Re: Field vs armchair linguistics (was:OT YAEPT -omp, -onk_
From: Dirk Elzinga
2a. Re: Ancient languages reconstructed by computer program
From: Carsten Becker
2b. Re: Ancient languages reconstructed by computer program
From: Alex Fink
2c. Re: Ancient languages reconstructed by computer program
From: Aidan Grey
2d. Re: Ancient languages reconstructed by computer program
From: Matthew Turnbull
3.1. Re: OT YAEPT -omp, -onk
From: David McCann
3.2. Re: OT YAEPT -omp, -onk
From: Allison Swenson
Messages
________________________________________________________________________
1.1. Re: Field vs armchair linguistics (was:OT YAEPT -omp, -onk_
Posted by: "Dirk Elzinga" [email protected]
Date: Wed Feb 13, 2013 8:59 am ((PST))
I agree with what Roger said about field work; there's nothing like it. Of
course, I don't travel to primitive and exotic locales to do field work
(unless you consider small towns in the Great Basin and Southeastern Utah
to be either, or both). But the feeling of cultural dislocation is very
strong on Indian reservations. That coupled with a traditional mistrust
(and even active dislike) of white people makes for some tense moments when
being introduced into a community. [1]
It is true that, of the conlangers that I've met, most have been interested
in field work. I wonder, though, if it isn't really Field Work they're
interested in so much as finding out new and unusual things about language.
Since the bulk of the theoretical work in linguistics that I've seen is
confined to well known and well documented languages, it may be that
conlangers, in their search for the weird and exotic, are naturally drawn
to reference grammars and other products of field work.
Dirk
[1] I attended a meeting of the cultural preservation committee for a tribe
in which one of the senior members took my presence as an opportunity to
berate me, as a representative of White People Everywhere, for the losses
and depredations suffered by his people. Most of the other committee
members were embarrassed--most, but not all. I ended up not getting a whole
lot done there, but I did score some pity points and made a couple of good
friends in the community.
On Wed, Feb 13, 2013 at 8:59 AM, Roger Mills <[email protected]> wrote:
> --- On Tue, 2/12/13, And Rosta <[email protected]> wrote:
> Sai, On 12/02/2013 02:27:
> > I'm not a field linguist so I have no idea what the perceptual
> > distribution is,
>
> I'm not a field linguist either. I'm a sworn adherent of armchair
> linguistics. I do do a bit of field linguistics, but only what can be done
> from the comfort of my own armchair (or classroom). [Discussion topic for a
> new thread: Among professional linguisticians, armchair linguists are many
> and field linguists are few; among conlangers interested in careers or
> advanced academic studyin linguistics, would-be field linguists are many
> and would-be armchair linguists are few. How come?]
>
> Sai, On 12/02/2013 02:27:
> > I'm not a field linguist so I have no idea what the perceptual
> > distribution is,
>
> I'm
> not a field linguist either. I'm a sworn adherent of armchair
> linguistics. I do do a bit of field linguistics, but only what can be
> done from the comfort of my own armchair (or classroom).
> RM-- HaHa. I
> began as a field linguist but in my ultimate lack of an academic
> appointment have perforce become an armchair linguist......
>
>
> [Discussion
> topic for a new thread: Among professional linguisticians, armchair
> linguists are many and field linguists are few; among conlangers
> interested in careers or advanced academic study in linguistics,
> would-be field linguists are many and would-be armchair linguists are
> few. How come?]
> ============================================================
>
> OK--I'll start it here: Not sure that's
> entirely accurate. Most academic training in linguistics (as well as
> Anthropology) will include field work (if only in a "Field Methods"
> course, where some us catch the bug-- that course shifted my whole
> academic/intellectual orientation from Romance to Malay-Polynesian!).
> OTOH most conlangers AFAICT are armchair linguists-- although one could
> say that inventing a language has certain elements of field linguistics
>
> Field
> work is truly fascinating-- you're not only encountering a whole new
> language but often a new culture as well (same in a lot of conlanging).
> It can be rather uncomfortable-- primitive surroundings, weird foods,
> difficulties relating to the people, or the converse, going a bit
> "native" ;-) -- and sometimes frustrating when you find variation even
> with a single small group. Or working with a moribund language, whose
> speakers are few and elderly, maybe lacking some teeth and having memory
> problems....
>
> Even the SILers, with their religious motivation, would probably confess
> to some occasional discomfort in their field work.
>
> You
> need a firm grounding in phonetics (even for English! as we discover in
> these YAEPT threads), and some kind of
> theoretical underpinning to make sense of the phonology and grammar you're
> discovering. Such field work has formed the basis of many a doctoral
> dissertation!! (Unless you're Chomsky & Co., who apparently saw no
> reason to work on anything other than English (so I'm told...))
>
> My
> own field work was done in comfortable surroundings, but if I'd had
> time, could well have involved some serious trekking into the wilds....
> At one point early on I thought (with my background in Spanish) that it
> might be interesting to work on S.American native languages, but an
> audited Anthro. course under N.Y.Chagnon (of Yanomamo fame) disabused me
> of that idea.
>
> Since becoming an armchair/non-affiliated
> linguist, I've discovered another problem: something I'd spent several
> years writing,
> and considered worth publishing, was rejected on the basis that it was
> based only on book research, not on actual field work. Bah humbug. I
> could have pointed out that one of the early stars in the MP field,
> Renward Brandstetter, never did field work AFAIK, yet managed to do
> important work from published sources....(Like Einstein, he clerked in
> some Swiss govt. office.)
>
> So let's see what sort of discussion this starts. I may have unfairly
> categorized some conlangers.....
>
Messages in this topic (39)
________________________________________________________________________
________________________________________________________________________
2a. Re: Ancient languages reconstructed by computer program
Posted by: "Carsten Becker" [email protected]
Date: Wed Feb 13, 2013 9:39 am ((PST))
If anyone cares, I have access to it through the university network and
downloaded it. But I have zero time at the moment because I'm preoccupied with
my BA thesis, so I won't be likely to read it very soon.
I really wonder if this could be made useful for conlanging.
Carsten
Messages in this topic (6)
________________________________________________________________________
2b. Re: Ancient languages reconstructed by computer program
Posted by: "Alex Fink" [email protected]
Date: Wed Feb 13, 2013 9:46 am ((PST))
On Wed, 13 Feb 2013 08:53:57 -0800, Roger Mills <[email protected]> wrote:
>However, two people on my AN-List commented, just sort-of semi-favorably.
>One said-- QUOTE an interesting and technically sophisticated piece of work.
>However,
note that their algorithm requires a tree as input and that their
inferences are still a long way from those of professionally trained
linguists. The
algorithm does better than just picking a reflex in an extant language,
but not by as much as one might hope. END QUOTE
The algorithms of the paper are inspired by those biologists use in doing
phylogenetic reconstruction, and certainly in biological applications one
strives to algorithmically reconstruct the phylogenic tree of a bunch of
organisms given only the DNA sequences for cognate sets of genes in each of
them. And there are in fact good algorithms for this out there. Accordingly I
don't see any intrinsic reason that one shouldn't be able to write a similar
algorithm that reconstructs the tree. I'd guess that the main reason this
paper didn't go for it in one swoop is just that we'd probably need a lot more
data to get a comparably good (passable?) result.
Actually, I know a number of mathematicians, including academic siblings, who
work on this kind of phylogenetic tree reconstruction. My reaction on learning
of this paper included a little bit of "man, why didn't _I_ do that?"
> I'm a little curious if they did any sort of checking for accuracy; for
>instance, inputing modern Romance langauges and seeing if the program spat out
>Latin.
>============================
>Don't know; interesting idea.
They don't mention having done so. My guess would be that this is because of
availability of amenable datasets. The Austronesian database
<http://language.psy.auckland.ac.nz/austronesian/> is a beautifully curated
thing (y'all should have a wander around in it, if you haven't yet), and I
don't know where you'd find a comparably curated, semantically-aligned vocab
list for Romance languages. You could probably assemble it from various
general-purpose multilingual dictionary sources for at least the five national
Romance languages, with a tradeoff between accuracy of semantic alignment and
time spent in arduous hand-correction, but if their Figure 1C is anything to go
by, five languages would get you a fairly poor reconstruction.
Alex
Messages in this topic (6)
________________________________________________________________________
2c. Re: Ancient languages reconstructed by computer program
Posted by: "Aidan Grey" [email protected]
Date: Wed Feb 13, 2013 4:05 pm ((PST))
You downloaded the study, or the program? Because honestly, I would love
both... But I'd be happy with just the one.
>________________________________
> From: Carsten Becker <[email protected]>
>To: [email protected]
>Sent: Wednesday, February 13, 2013 10:39 AM
>Subject: Re: Ancient languages reconstructed by computer program
>
>If anyone cares, I have access to it through the university network and
>downloaded it. But I have zero time at the moment because I'm preoccupied with
>my BA thesis, so I won't be likely to read it very soon.
>
>I really wonder if this could be made useful for conlanging.
>
>Carsten
>
>
>
Messages in this topic (6)
________________________________________________________________________
2d. Re: Ancient languages reconstructed by computer program
Posted by: "Matthew Turnbull" [email protected]
Date: Wed Feb 13, 2013 5:25 pm ((PST))
Thanks for pointing out this study, I read an article about it in Nature,
but the link helped me actually find it. I'm not much for math, but it
looks really cool. There is a part in the paper where they describe how to
build a tree no? page 2 paragraph 7, or did I misunderstand that? Also they
do a test where they do a random tree and the reconstruction fails
considerably more, with only about 68% of words agreeing instead of 85%. It
appears to be more affected by the size of the dataset than the tree, as
shown in figure 1 A, no? (I understand the concepts here but not the actual
implementation, so any help is appreciated if I've got this wrong). I
couldn't find a link to the program, my guess is they will publish it with
the paper if they do so, this is published ahead of print.
-Matt
On Wed, Feb 13, 2013 at 8:36 AM, Петр Кларк <[email protected]> wrote:
> The BBC has an article (http://www.bbc.co.uk/news/science-
> environment-21427896) about computer reconstructions of proto-languages
> (specifically Austronesian), with an 85% match with what linguists had
> reconstructed "by hand". The full report is published in the Proceedings of
> the National Academy of Science
> (http://www.pnas.org/content/early/2013/02/05/1204678110) -- anyone have
> access to it?
> I'm a little curious if they did any sort of checking for
> accuracy; for
> instance, inputing modern Romance langauges and seeing if the program spat
> out
> Latin.
> :Peter
>
Messages in this topic (6)
________________________________________________________________________
________________________________________________________________________
3.1. Re: OT YAEPT -omp, -onk
Posted by: "David McCann" [email protected]
Date: Wed Feb 13, 2013 9:50 am ((PST))
On Tue, 12 Feb 2013 18:00:24 -0500
Tim Smith <[email protected]> wrote:
> You haven't been mispronouncing it! I've just looked in two print
> dictionaries (as opposed to online dictionaries), and they both show
> both pronunciations (although they both list the /Nk/ pronunciation
> first, indicating that it's "preferred" (but preferred by whom?)).
I've always said /konk/, but then I know the Latin concha and the
mathematicians' conchoid. My ancient copy of Jones (specifically RP)
lists /konč/ as rare and my (almost as ancient) Kenyon & Knott only
lists /konk ~ kank/. I suspect /conč/ is a modern spelling
pronunciation. The sense "person of low social status in Florida" was
(naturally) new to me: yet another word I'll never get the chance to
use!
Messages in this topic (39)
________________________________________________________________________
3.2. Re: OT YAEPT -omp, -onk
Posted by: "Allison Swenson" [email protected]
Date: Wed Feb 13, 2013 10:21 am ((PST))
Is there a difference between CLOTH and CAUGHT for others, then? I believe
I say them the same, or nearly so.
On Tuesday, February 12, 2013, Herman Miller wrote:
> On 2/12/2013 5:17 PM, And Rosta wrote:
>
> The likeliest form of (4b) is (4c):
>> (4c) FATHER in all -omp words and some -onk words and CAUGHT in other
>> -onk words.
>>
>> I presume -ong words pattern with -onk words.
>>
>
> All -ong words have the CAUGHT vowel for me.
>
> I had been trying to work out how many phonologically short vowels North
>> American accents have. There are grounds for counting only 5 (KIT,
>> DRESS, TRAP, STRUT, FOOT), or for counting 6, or (especially for (4b/c)
>> dialects) for counting 7. (Not counting extras due to e.g. BAD/LAD split.)
>>
>
> I'd count only the 5 you've mentioned, but then CAUGHT would be the only
> "long" vowel that can appear before /ŋ/ (not counting foreign borrowings).
> (Historically it would have been the CLOTH vowel.)
>
> My tally of responses (based on info given in responses):
>> Zach (1)
>> Tony (2) [perplexing! -- Tony: what words have the COT vowel? Sob? Bomb?
>> Blond? Sconce? Mop? Or does BOTHER not have the COT vowel?]
>> Tim (4a)
>> Stevo (4a)
>> Roger (4a) ((4c) counting _conch_)
>> Allison (4a) ((4c) counting _conch_)
>> Gary (4c)
>> Herman (4c)
>>
>> For the 4c-ers, it seems as though the incidence of FATHER in -onk --
>> and -ong?? -- is sufficiently rare that they might be considered special
>> exceptions (like e.g. _boing_ and _oink_ are).
>>
>
> There are few enough -onk words as it is, but I think probably (for me)
> the majority are pronounced with CAUGHT exclusively, and the others can be
> pronounced either way. So it seems fair to treat them as exceptions.
>
Messages in this topic (39)
------------------------------------------------------------------------
Yahoo! Groups Links
<*> To visit your group on the web, go to:
http://groups.yahoo.com/group/conlang/
<*> Your email settings:
Digest Email | Traditional
<*> To change settings online go to:
http://groups.yahoo.com/group/conlang/join
(Yahoo! ID required)
<*> To change settings via email:
[email protected]
[email protected]
<*> To unsubscribe from this group, send an email to:
[email protected]
<*> Your use of Yahoo! Groups is subject to:
http://docs.yahoo.com/info/terms/
------------------------------------------------------------------------
