From gonzalor at jhu.edu  Mon Jan 12 02:14:06 1998
From: gonzalor at jhu.edu (Gonzalo Rubio)
Date: Sun, 11 Jan 1998 21:14:06 EST
Subject: Quechua en UPenn
Message-ID: <SUN.11.JAN.1998.211406.EST.GONZALOR@JHUNIX.HCF.JHU.EDU>

----------------------------Original message----------------------------
 
I got this from a friend. I think some of you may find it interesting,
____________________________
 
Gonzalo Rubio
Near Eastern Studies
The Johns Hopkins University
gonzalor at jhu.edu
____________________________
 
---------- Forwarded message ----------
Date: Wed, 7 Jan 1998 14:09:15 -0500
From: Ulises Zevallos <uzevallo at NIMBUS.OCIS.TEMPLE.EDU>
Subject: Quechua en UPenn
 
*Elementary Quechua I and II will be offered at The University  of
Pennsylvania in Summer  1998.*  These courses are open to graduate
students, advanced undergraduates, professionals and other interested
individuals.  They will be taught by Serafin Coronel-Molina, a native
speaker of Quechua, using a combination of traditional and multimedia
texts.  The course has been developed in close consultation with Dr.
Nancy H. Hornberger of the Graduate School of Education at the University of
Pennsylvania.
 
Elementary Quechua I (LING 140 for undergraduates; LING 508 for
graduates) will be offered in the first summer session (May 19 to June
26, 1998), and will meet five days a week from 10:00 am to 12:00 noon.
Elementary Quechua II (LING 141 for undergraduates; LING 509 for
graduates) will be offered in the second summer session (June 29 to
August 7, 1998), and will also meet five days a week from 10:00 am to
12:00 noon.
 
Elementary Quechua I is intended for students with no previous study
experience in Quechua.  It introduces students to the language and
culture of the Quechua people.  This is the language that was spoken by
the ancient Incas and is still spoken today by more than 10 million
speakers throughout the Andean countries of South America. The variety
taught will be from the Southern Quechua family spoken in Peru. The
course will promote the development of the four language skills:
speaking, listening, reading and writing, providing a good practical
command of oral and written skills appropriate for everyday situations.
In addition, classwork will include discussion of native Andean culture,
as well as the changing face of Quechua culture in light of recent
migration trends.
 
Elementary Quechua II is a continuation of Elementary Quechua I for
students who have taken the first session course or who have previously
studied Southern Peruvian Quechua at the beginning level.  The format
will be the same as for Elementary Quechua I, with continued building of
the four essential language skills:  speaking, listening, reading and
writing.  Grammatical structures will be continually reviewed throughout
this course, while a rich input of material in Quechua is provided with
the goal of increasing the range of vocabulary and linguistic structures
as well as knowledge of the culture.
 
For further information and registration, contact the Penn Language
Center, 401 Lauder-Fischer Hall, Philadelphia, PA  19104-6330.  Telephone
(215) 898-6039, fax (215) 573-2139.  E-mail <vassilie at sas.upenn.edu> (Ms.
Lada Vassilieva).


From larryt at cogs.susx.ac.uk  Mon Jan 12 02:16:29 1998
From: larryt at cogs.susx.ac.uk (Larry Trask)
Date: Sun, 11 Jan 1998 21:16:29 EST
Subject: Q: term
Message-ID: <SUN.11.JAN.1998.211629.EST.LARRYT@COGS.SUSX.AC.UK>

----------------------------Original message----------------------------
I'm looking for a term.  The phenomenon in question is extremely
familiar, but I don't know of an accepted name for it.
 
The phenomenon is this: a linguistic change which simplifies one
subsystem of a language may complicate another subsystem.
 
A typical example is the history of Spanish mid vowels.  Earlier
Spanish had two low-mid vowels and two high-mid vowels; the low-mid
vowels were *automatically* diphthongized under stress, while the
high-mid vowels were not.  But then the two low-mid vowels merged
with the two higher ones.  This change simplified the phonological
system by removing two phonemes, but it greatly complicated the
morphology: the formerly automatic and transparent diphthongizations
became totally unpredictable and opaque, since some instances of the
new /e/ and /o/ diphthongized while others did not.
 
Does anybody know of an accepted label for this phenomenon, which I
suppose we might elevate to the status of a "principle"?  If not,
wuld anybody like to propose one?
 
Larry Trask
COGS
University of Sussex
Brighton BN1 9QH
England
 
larryt at cogs.susx.ac.uk


From jharvey at ucla.edu  Mon Jan 12 22:04:50 1998
From: jharvey at ucla.edu (Jasmin Harvey)
Date: Mon, 12 Jan 1998 17:04:50 EST
Subject: Quechua en UCLA
Message-ID: <MON.12.JAN.1998.170450.EST.JHARVEY@UCLA.EDU>

----------------------------Original message----------------------------
For those on the west coast who are interested in studying Quechua, it is
also offered fairly often at UCLA as part of the Indigenous Languages of the
Americas section of the Linguistics department.  I don't know whether it
will be offered this following quarter, but the online schedule of classes,
accessible from their web page at http://www.registrar.ucla.edu/schedule/
will allow you to search, and the course description may be found in the
catalogue at http://www.registrar.ucla.edu/catalog/catalog-Indigeno.html
 
Jasmin Harvey
Germanic Linguistics, UCLA
jharvey at ucla.edu


From vbubenik at morgan.ucs.mun.ca  Mon Jan 12 20:18:37 1998
From: vbubenik at morgan.ucs.mun.ca (Vit Bubenik)
Date: Mon, 12 Jan 1998 15:18:37 EST
Subject: Q: term
In-Reply-To: <m0xr1yG-0006MXC@rsunx.crn.cogs.susx.ac.uk>
Message-ID: <MON.12.JAN.1998.151837.EST.VBUBENIK@MORGAN.UCS.MUN.CA>

----------------------------Original message----------------------------
I am using the term "trade-off" between phonology and morphology (or
between morphology and syntax) for this phenomenon.
Vit Bubenik, Linguistics, Memorial University of Newfoundland, St.John's,
Canada.
 
On Sun, 11 Jan 1998, Larry Trask wrote:
 
> ----------------------------Original message----------------------------
> I'm looking for a term.  The phenomenon in question is extremely
> familiar, but I don't know of an accepted name for it.
>
> The phenomenon is this: a linguistic change which simplifies one
> subsystem of a language may complicate another subsystem.
>
> A typical example is the history of Spanish mid vowels.  Earlier
> Spanish had two low-mid vowels and two high-mid vowels; the low-mid
> vowels were *automatically* diphthongized under stress, while the
> high-mid vowels were not.  But then the two low-mid vowels merged
> with the two higher ones.  This change simplified the phonological
> system by removing two phonemes, but it greatly complicated the
> morphology: the formerly automatic and transparent diphthongizations
> became totally unpredictable and opaque, since some instances of the
> new /e/ and /o/ diphthongized while others did not.
>
> Does anybody know of an accepted label for this phenomenon, which I
> suppose we might elevate to the status of a "principle"?  If not,
> wuld anybody like to propose one?
>
> Larry Trask
> COGS
> University of Sussex
> Brighton BN1 9QH
> England
>
> larryt at cogs.susx.ac.uk
>


From tvn at cis.uni-muenchen.de  Mon Jan 12 20:17:09 1998
From: tvn at cis.uni-muenchen.de (Theo Vennemann)
Date: Mon, 12 Jan 1998 15:17:09 EST
Subject: TUNNEL VISION
Message-ID: <MON.12.JAN.1998.151709.EST.TVN@CIS.UNIMUENCHEN.DE>

----------------------------Original message----------------------------
Today I opened the following message signed by Larry Trask.
I would like to respond.
 
>----------------------------Original message----------------------------
>I'm looking for a term.  The phenomenon in question is extremely
>familiar, but I don't know of an accepted name for it.
>
>The phenomenon is this: a linguistic change which simplifies one
>subsystem of a language may complicate another subsystem.
>
>A typical example is the history of Spanish mid vowels.  Earlier
>Spanish had two low-mid vowels and two high-mid vowels; the low-mid
>vowels were *automatically* diphthongized under stress, while the
>high-mid vowels were not.  But then the two low-mid vowels merged
>with the two higher ones.  This change simplified the phonological
>system by removing two phonemes, but it greatly complicated the
>morphology: the formerly automatic and transparent diphthongizations
>became totally unpredictable and opaque, since some instances of the
>new /e/ and /o/ diphthongized while others did not.
>
>Does anybody know of an accepted label for this phenomenon, which I
>suppose we might elevate to the status of a "principle"?  If not,
>wuld anybody like to propose one?
 
Doesn't the example given above fall under the principle that all lan-
guage change is LOCAL improvement, i.e. change on the parameter
it is working on (and may thus cause complications on other parame-
ters)?
 
See "Language change as language improvement", in Vincenzo Orioles,
ed., Modelli esplicativi della diacronia linguistica: Atti del Convegno
della Società Italiana di Glottologia, Pavia, 15-17 settembre 1988. Pisa
(Giardini Editori e Stampatori), 1989, 11-35. Reprinted in: Language
change as language improvement", in Charles Jones, ed., Historical
linguistics: Problems and perspectives. London (Longman), 1993,
319-344.
 
The consequence is so self-evident that I do not really think a name is
needed. If things were different, languages would be optimal on all
parameters, which is impossible. Thus, the principle also follows
from the observation that languages keep an overall identical level
of complexity, at least as long as we do not talk about language de-
velopment in terms of the evolution of the species.
 
One could say that language change is, in general, blind to its own side
effects.Thus, if a name were needed, one could call it the blindness
principle. Or if it is to be emphasised that the principle only looks in
its own direction, but neither left nor right, one could call it the
principle of tunnel vision.
 
Theo Vennemann,
12 January 1998.


From bhk at hd1.vsnl.net.in  Tue Jan 13 14:33:32 1998
From: bhk at hd1.vsnl.net.in (Bh. Krishnamurti)
Date: Tue, 13 Jan 1998 09:33:32 EST
Subject: Q: term
Message-ID: <TUE.13.JAN.1998.093332.EST.BHK@HD1.VSNL.NET.IN>

----------------------------Original message----------------------------
>Date: Mon, 12 Jan 1998 22:10:13
>To: Larry Trask <larryt at cogs.susx.ac.uk>
>From: "Bh. Krishnamurti" <bhk at hd1.vsnl.net.in>
>Subject: Re: Q: term
>
>I noticed a similar phenomenon in the history of the Telugu language. I
called the process involved 'morphologization of sound change'. In short,
the process was as follows: Verb roots of the type (C)Vn- (some 7 in number)
when followed by a dental stop in modern Telugu change it to a retroflex,
e.g. an- 'to say' +tu: --> aN-Tu:'saying' (caps for retroflexes). There is
no phonetic motivativation for a dental stop to become a retroflex after a
non-retroflex nasal (dental-alveolar). Historically, the final n represents
a merger of an alveolar n and a retroflex N. The intermediate rules were (1)
n+tt-->n+t't' (a dental t following an an alveolar n became an alveolar t';
(2) some roots ended in N, so N+tt--> N+TT ( a dental t became aa retroflex
T  after a retroflex N): (3) Alveolar t't' merged with retroflex TT
unconditionally; (4)N-->n in all environments except before retrolexes
[phonetically]. So we have roots ending in n (< n and N); t't'-->TT; these
two sound changes are followed by another regular sound change,(5)Geminate
reduction: CC-->C/C+__. Rules (1) to (5)are all automatic. As a consequence
of these we have a complicated morphophonemics in modern Telugu, e.g. an+tu:
--> aN-Tu: 'saying' (older *aN-TTu: < *an-t't'u:); kon-Tu: ('buying' from
*koN+TTu:; older root was *koN-). The problem here is that we have to shift
the cause (conditioning factor) of retroflexion of t to T to the canonical
form of the root (i.e. (C)Vn-) which is not either phonetically or
phonologically a natural cause for a dental to become a retroflex.
>
>In the Spanish case, one solution is to treat diphthongized e and o as a
separate class of stems from those with e and o  which do not diphthongize.
This classification reflects their history. This is also a case of a regular
sound change creating morphological conditioning; simplification achieved in
phonology has a cost in morphology. This phenomenon is common in many languages.
>
>With regards, Bh.K.
>
>
>
>At 21:16 11/01/98 EST, you wrote:
>>----------------------------Original message----------------------------
>>I'm looking for a term.  The phenomenon in question is extremely
>>familiar, but I don't know of an accepted name for it.
>>
>>The phenomenon is this: a linguistic change which simplifies one
>>subsystem of a language may complicate another subsystem.
>>
>>A typical example is the history of Spanish mid vowels.  Earlier
>>Spanish had two low-mid vowels and two high-mid vowels; the low-mid
>>vowels were *automatically* diphthongized under stress, while the
>>high-mid vowels were not.  But then the two low-mid vowels merged
>>with the two higher ones.  This change simplified the phonological
>>system by removing two phonemes, but it greatly complicated the
>>morphology: the formerly automatic and transparent diphthongizations
>>became totally unpredictable and opaque, since some instances of the
>>new /e/ and /o/ diphthongized while others did not.
>>
>>Does anybody know of an accepted label for this phenomenon, which I
>>suppose we might elevate to the status of a "principle"?  If not,
>>wuld anybody like to propose one?
>>
>>Larry Trask
>>COGS
>>University of Sussex
>>Brighton BN1 9QH
>>England
>>
>>larryt at cogs.susx.ac.uk
>>
>>
>
Bh. Krishnamurti
H.No. 12-13-1233, "Bhaarati"
Street No.9, Tarnaka
Hyderabad 500 017, A.P.
India
Telephone (R)(40)701 9665
E-mail: <bhk at hd1.vsnl.net.in>
Note:Please note what follows hd is digit 1 and not letter l.
In vsnl the final character is letter l and not digit 1.


From mcv at wxs.nl  Tue Jan 13 14:32:53 1998
From: mcv at wxs.nl (Miguel Carrasquer Vidal)
Date: Tue, 13 Jan 1998 09:32:53 EST
Subject: Q: term
In-Reply-To: <m0xr1yG-0006MXC@rsunx.crn.cogs.susx.ac.uk>
Message-ID: <TUE.13.JAN.1998.093253.EST.>

----------------------------Original message----------------------------
Larry Trask <larryt at cogs.susx.ac.uk> wrote:
 
>I'm looking for a term.  The phenomenon in question is extremely
>familiar, but I don't know of an accepted name for it.
>
>The phenomenon is this: a linguistic change which simplifies one
>subsystem of a language may complicate another subsystem.
 
In the tradition of Ablaut and Lautgesetz, why not "Schlimmbesserung"?
 
=======================
Miguel Carrasquer Vidal
mcv at wxs.nl
Amsterdam


From fcosw5 at mbm1.scu.edu.tw  Tue Jan 13 14:32:26 1998
From: fcosw5 at mbm1.scu.edu.tw (Steven Schaufele)
Date: Tue, 13 Jan 1998 09:32:26 EST
Subject: diachronic compensation
Message-ID: <TUE.13.JAN.1998.093226.EST.FCOSW5@MBM1.SCU.EDU.TW>

----------------------------Original message----------------------------
Dear fellow historical & comparative linguists,
 
Wrt Larry Trask's recent question about (if i understand him correctly)
a shorthand label for the observation that diachronic `simplification'
in one area of a language's (e.g.) grammar is typically compensated for
by a relative `complication' in another, Theo Vennemann made a valuable
bibliographic recommendation and then wrote:
 
> The consequence is so self-evident that I do not really think a name is
> needed. If things were different, languages would be optimal on all
> parameters, which is impossible. Thus, the principle also follows
> from the observation that languages keep an overall identical level
> of complexity, at least as long as we do not talk about language de-
> velopment in terms of the evolution of the species.
 
It -- the diachronic compensation between different grammar-modules
mentioned earlier, as well as Theo's `observation that languages keep an
overall identical level of complexity' -- is self-evident to *us* who
spend our time and intellectual energies studying the matter; i suppose
the value of sex for rearranging genes is self-evident to geneticists,
biologists, and evolutionists.  But it's very far from evident to the
general public, including the intelligent, relatively well-educated
public.  (I'm thinking of my own parents, among others, who in spite of
being at least tetraglot if not pentaglot still tend to assume that
English as a whole is `simpler' than, e.g., Latin or Polish, and Modern
English is `simpler' than Old English or Modern German.  I'm also
thinking of my own students here in Taiwan, some of whom are astonished
at the notion that Chinese has any `grammar' at all.) This `principle'
that Larry is seeking a convenient label for is something we need to
reiterate iteratively in our dialogue with the general public, and for
that reason i'm entirely sympathetic to Larry's quest.
 
Best,
Steven
--
Steven Schaufele, Ph.D., Asst. Prof. of Linguistics, English Department
 
Soochow University, Waishuanghsi Campus, Taipei 11102, Taiwan, ROC
 
(886)(02)2881-9471 ext. 6504     fcosw5 at mbm1.scu.edu.tw
 
http://www.prairienet.org/~fcosws/homepage.html
 
 
 
        ***O syntagmata linguarum liberemini humanarum!***
 
        ***Nihil vestris privari nisi obicibus potestis!***


From mmorrison at VNET.IBM.COM  Wed Jan 14 14:15:35 1998
From: mmorrison at VNET.IBM.COM (Michael C. Morrison 8-543-4706)
Date: Wed, 14 Jan 1998 09:15:35 EST
Subject: diachronic compensation
Message-ID: <WED.14.JAN.1998.091535.EST.MMORRISON@VNET.IBM.COM>

----------------------------Original message----------------------------
*** Reply to note of Tue, 13 Jan 1998 09:32:26 -0500 (EST)
*** by fcosw5 at mbm1.scu.edu.tw
 
Steven Schaufele <fcosw5 at mbm1.scu.edu.tw> writes:
>Wrt Larry Trask's recent question about (if i understand him correctly)
>a shorthand label for the observation that diachronic `simplification'
>in one area of a language's (e.g.) grammar is typically compensated for
>by a relative `complication' in another,
 
I suggest we could borrow a term from Physics and change its meaning
appropriately:  The Equivalence Principle.  For comparative and
historical linguistics, it would mean that all languages are
_equivalently_ complex, but obviously not complex in the same ways.
And any change to one system or subsystem of a language _maintains_
the overall complexity of the language -- one part gets simpler as
another becomes more complex.
 
Alternately, we could call it the Law of Simplification ... ;)
(I'm just not sure it warrants being called a Law)
 
Michael
----------------------------------------------------------------------
Michael C. Morrison   Santa Teresa Laboratory      Phone (408)463-4706
IMS User Technology   IBM Software Solutions       Fax   (408)463-3696
Lotus Notes ID:       MCMORRIS at IBMUSM50   IBMLink:  MORRISON at TORIBM
Internet ID:          MMORRISON at VNET.IBM.COM  or  USIB47H4 at IBMMAIL.COM
IBM Mail Exchange:    USIB47H4 at IBMMAIL     or   USIB4MCM at IBMMAIL
X.400 Address:        G=mcmgm; S=morrison; P=ibmmail; A=ibmx400; C=us
----------------------------------------------------------------------


From rjanda at midway.uchicago.edu  Wed Jan 14 14:15:10 1998
From: rjanda at midway.uchicago.edu (Richard Janda)
Date: Wed, 14 Jan 1998 09:15:10 EST
Subject: Principles like trade-offs, tunnel vision,
 and local generalizations
Message-ID: <WED.14.JAN.1998.091510.EST.RJANDA@MIDWAY.UCHICAGO.EDU>

----------------------------Original message----------------------------
 
Regarding Larry Trask's query, and following up on Vit Bubenik's and Theo
Vennemann's responses (especially the latter's term "tunnel vision"):
  Certainly the phenomenon in question overlaps with the principles described
by the statement that languages practice therapy, not prophylaxis (most strong-
ly repeated by Lightfoot 1979, I believe, though the idea goes back at least
to Vendryes, doesn't it?), modulo Theo's locality restrictions, and by Sturte-
vant's Paradox about sound-change being regular but causing irregularity (in
morphology), while analogy is irregular (morpholexically sporadic) but rees-
tablishes regularity.
  In my own work, I have emphasized not only grammatical locality but also
data-related locality:  i.e., any language is truly so vast that no speaker
can keep all subsystems and their interrelated consequences in mind when
tempted to make a particular innovation or to apply a given principle which
turns out to yield what is effectively a hypercorrection.  Graphically, I
once prepared a diagram in which language structure was presented as a leaded
glass window looking out on the world, and a speaker's linguistic conscious-
ness was represented as a flashlight beam directed at something in that out-
side world, but broad enough to illuminate only a small area of the language-
structure window.  As a result, it is not surprising that a speaker may notice
competing parts of two patterns, and so be tempted to restructure part of the
window so as to yield a more consistent local pattern (within the flashlight
beam) that is actually not so consistent when viewed more globally.  (In the
diagram I used at the Linguistic Association of the Southwest meeting in Hous-
ton in 1995, the speaker was equipped with a blowtorch for redoing inconsis-
tent parts of leaded-glass windows, though this suggests more conscious tink-
ering with language than I would like to imply.)
  If correct, this factor (what Brian Joseph and I have called "a limited
window of data" restricting the amount of linguistic structure that a single
speaker/hearer can keep in mind) to which Larry Trask has indirectly pointed
in diachrony has the direct implication for synchrony that linguists probably
overestimate the degree to which speakers "consider all the relevant data"
when arriving at (tacit) analyses--hence the prevalence of what Brian and I
have called "local generalizations".  (E.g., the shift from [i] to [eI] for
the first vowel in English _academia_ _ [still not reflected in some diction-
aries] seems to reflect the local influence of _macadamia_ rather than some
general favoring of foreign-like [e] (_anemia_ in the U.S. never has [e], to
the best of my knowledge); still, there are a lot of nuts in academe.)
  With Vennemann, though, I am tempted to think that such limitations of at-
tention are so pervasive in both linguistic and non-linguistic behavior that
they may not need a special name, though "tunnel vision" and consequence-
blindness are suggestive....
     Richard Janda


From DISTERH at UNIVSCVM.SC.EDU  Thu Jan 15 20:36:38 1998
From: DISTERH at UNIVSCVM.SC.EDU (Dorothy Disterheft)
Date: Thu, 15 Jan 1998 15:36:38 EST
Subject: SLE 98 St. Andrews
Message-ID: <THU.15.JAN.1998.153638.EST.DISTERH@UNIVSCVM.SC.EDU>

Invitation
 
to Workshop during the 31st SLE Conference
at St. Andrews, 26-30 August, 1998
 
Workshop organizer: Werner Abraham, Groningen
(E-mail: ABRAHAM at LET.RUG.NL, FAX: +31-50-363 58 21)
 
Workshop topic:
`Spoken and written languages;
their structural and typological differeences'
 
The workshop aims at presentations in the following subfields to
the general title:
parsing strategies divided between spoken and written vernaculars;
typological
differences and historical changes initiated by parsing rather than
logical-structu-
ral triggers (such as, as the underwritten himself will claim in a
contribution to
the workshop, the Upper German preterite decay); typological
differences be-
tween genetically closely related languages such as Afrikaans,
Yiddish and dial-
ectal Germans (more or less strongly svo) vs. written German (sov);
what me-
diates between sov and svo other than (Charles Fries' claim)
distinctions of mor-
phological case? It will be shown in the paper referred to above
that distinctions
of a discourse-functional sort and their prerequsites in structural
terms (wide
middle field!) can contribute to the upkeeping of sov despite the
fact that case
morphology is rather weak (dialectal German and Dutch).
 
Organization: (especially younger) colleagues interested should contact
organizer
at Groningen with an abstract covering an adjacent topic no longer than
one page
(for a 20-30 minutes presentation). The intention is to submit a
collection of the-
matically unified papers to FoL as a self-contained volume.
 
Heed: To attend the conference you need to be a member of the
Societas Linguistica Europaea. Please
contact the local organizers, Dr. Christopher Beedham and Dr.
Isabel Forbes, Dept. of German, School
of Modern Languages, The University, St. Andrews, Fife KY16 9PH,
Scotland/UK, cb1 at st-and.ac.uk,
FAX (01334) 46 36 77, home page:
http://www.st-and.ac.uk/academic/modlangs/SLE98/SLE98.html
 
Mail to: Werner Abraham, Duits-Letteren-RUG,
Oude Kijk in 't Jat Straat 26, NL-9712 EK Groningen


From larryt at cogs.susx.ac.uk  Thu Jan 15 14:36:49 1998
From: larryt at cogs.susx.ac.uk (Larry Trask)
Date: Thu, 15 Jan 1998 09:36:49 EST
Subject: Sum: term
Message-ID: <THU.15.JAN.1998.093649.EST.LARRYT@COGS.SUSX.AC.UK>

----------------------------Original message----------------------------
A few days ago I posted a request for a term.  Once again, the
phenomenon I wanted a name for was this: a change that leads to
simplification in one domain often produces a simultaneous
complication in another domain.
 
The most familiar examples of this phenomenon, of course, involve
phonological simplifications and morphological complications, but
that's not the only possibility, as I perhaps should have pointed out
in my original query.  For example, syntagmatic phonological
simplifications can produce paradigmatic phonological complications,
as when palatalization in palatalizing environments produces new
marked segments, like the Czech fricative trill.  Then again,
analogical leveling (morphological simplification) can produce new
alternations in stems that formerly didn't alternate (morphological
complication), as has happened in some varieties of Serbo-Croatian (if
I'm still allowed to use that name).
 
The motive for my query was this.  As many of you know, I am compiling
a dictionary of historical and comparative linguistics.  Now, in
recent years, we have coined a rather large number of terms in the
field, and I've noticed that good names have been coined for a number
of familiar phenomena for which we formerly had no names; examples are
`actualization' (Timberlake), `metatypy' (Ross), `pandemic
irregularity' (Blust), `exaptation' (Lass), and `phonogenesis'
(Hopper), not to mention the memorable `morphanization' (Matisoff).
 
But I haven't found a recognized name for the phenomenon I'm
interested in here.  But, since the phenomenon, as Steven Schaufele
has pointed out, is such a fundamental one in our field, it seems to
me that we really ought to have a name for it.  Hence my query.
 
Fifteen people replied, and the first thing to report is that there
does indeed appear to be no recognized name for the phenomenon.
Almost everyone had one or more suggestions to make, but no two people
suggested the same term (though in one case two people came fairly
close).  A couple of people suggested terms which they themselves had
apparently used in print, but I guess those proposals haven't caught
on yet.
 
Anyway, here are the terms proposed, or most of them.  I omit a couple
of totally facetious suggestions, and one or two which were so
exceedingly long that I don't think they can be considered as terms.
A couple of people, I think, thought that I was asking specifically
for a label for the conversion of phonology into morphology, but in
fact I have in mind something more general than that.
 
BLINDNESS PRINCIPLE
CODE SHIFT
DIACHRONIC COMPENSATION
EQUILIBRIUM
HYDRA'S RAZOR
LOCAL IMPROVEMENT
LOCAL SIMPLIFICATION
MARKEDNESS CONFLICT
MORPHOLOGIZATION OF PHONOLOGICAL RULES
NATURALNESS CONFLICT
SCHLIMMBESSERUNG
SIMPLEXIFICATION
STURTEVANT'S PARADOX (unspecified variation on)
TRADE-OFF
TUNNEL VISION PRINCIPLE
 
Right.  Now what do I do?  Call for a vote?  Organize a competition
with five distinguished judges and a prize of two weeks in the PIE
homeland of your choice?  Close my eyes and stick a pin?  Ask Roger
Lass what the biologists call it?  Coin my own term and hope everybody
buys the book and believes me?  Or should I just admit defeat and not
include any term for this, on the not unreasonable ground that
dictionaries shouldn't be including words that don't exist?
 
Damned if I know.  But it *would* be nice if we had *some* name for
this.  Otherwise, how can we persuade our students it's important if
we haven't got a name for it?  I mean, I don't recall that so many
Americans go hot and bothered about visiting ever more soldiers and
bombs on the Vietnamese until somebody decided that what was happening
was `escalation', and then suddenly escalation was a hot issue.
 
Anyway, my thanks to Jacob Baltuch, Vit Bubenik, Miguel Carrasquer
Vidal, John Costello, Guy Deutscher, Hans-Olav Engel, Ralf-Stefan
Georg, Harold Koch, Bh. Krishnamurti, Paul Lloyd, Gary Miller, Steven
Schaufele, Theo Vennemann, Benji Wald, and Roger Wright.
 
(Hey -- how come no women?)
 
Larry Trask
COGS
University of Sussex
Brighton BN1 9QH
UK
 
larryt at cogs.susx.ac.uk


From hale1 at alcor.concordia.ca  Fri Jan 16 12:54:30 1998
From: hale1 at alcor.concordia.ca (MARK ROBERT HALE)
Date: Fri, 16 Jan 1998 07:54:30 EST
Subject: Sum: term
In-Reply-To: <m0xsXTY-0006MiC@rsunx.crn.cogs.susx.ac.uk>
Message-ID: <FRI.16.JAN.1998.075430.EST.HALE1@ALCOR.CONCORDIA.CA>

----------------------------Original message----------------------------
 
I apologize for responding late to this discussion. The
state of emergency has limited email opportunities up here.
 
And unfortunately I am not a woman. Anyway,...
 
Though I don't generally engage in terminological matters,
it did strike me that the discussion and terms suggested
addressed several different issues which should not, as
Trask pointed out kind of, be covered under a single
unifying term. "Sturtevant's paradox" and the "Morphologization
of Phonological Rules' refer to relationships between
phonology and morphology only. A few of the suggestions
I ddint' really understand (perhaps some of my email is
still making its way to my door), e.g.," Hydra's razor"
and "code shift".
 
"Equilibrium" seems like a label for a general diachronic
property of languages (if, indeed, it is one -- I personally
do not use the term 'simplification' because (1) I don't
have, and haven't seen, any coherent 'simplicity' metric
for grammars and (2) I can't figure out, for many e.g.
phonological changes that appear a priori to be 'simplifications'
[loss of voicing contrasts in stops, loss of voiceless nasals,
etc.] how they made anything ni the langauges in question any
more complicated -- this leads me to believe that our intuitive
sense of simplicity, if we are to retain a hypothesis of
equilibrium, can't be the right one), rather than for a
specific change of the type under discussion.
 
"Markedness" and "naturalness conflict" (aside from the fact
that I don't believe in 'markedness' or 'naturalness' -- in
a technical sense -- either) seem to describe a state ("conflict")
as well, rather than an event. [Politicians'
attempt to sanitize 'The Korean Conflict' and avoid invoking
the War Powers Act notwithstanding.]
 
On the other hand, "Tunnel Vision Principle", "Blindness
Principle", "Schlimmbesserung",  and "local improvement"
or "Local simplification" fail to capture the fact that
there was, allegedly, a 'compensating' change in another
component. Only "diachronic compensation" really does that.
 
Nevertheless, the 'Tunel Vision' terms turn out,  in my opinion,
to be the better way to state it. After all, in the cases discussed
surely the new morphological system of Spanish didn't come to
have the properties it did to 'compensate' for the changes in the
vowel system -- the morphological changes were the automatic
and epiphenomenal effects of the phonological reanalysis.
 
Surely we can all think of just as many cases in which a
phonological change gave rise to the elimination of morphological
rules or alternations.
 
Mark


From dlight at deans.umd.edu  Fri Jan 16 16:52:53 1998
From: dlight at deans.umd.edu (David Lightfoot)
Date: Fri, 16 Jan 1998 11:52:53 EST
Subject: complexity measures
Message-ID: <FRI.16.JAN.1998.115253.EST.DLIGHT@DEANS.UMD.EDU>

----------------------------Original message----------------------------
   Recent postings suggest that some people believe that languages are all
equally complex (although this is not entailed by Larry Trask's original
question, giving rise to this discussion).
   One possibility is that this belief is an empirical finding.  In which
case, there must be a way of measuring the overall complexity of a language and
somebody has found that languages all emerge with the same index.
Alternatively, it might follow from some basic principles or some
theory that languages must be equally complex. I know of no such
empirical support nor of any theoretical underpinning for such an idea.
What am I missing?
   I should have thought that if there is a simplification in some part
of a system, there doesn't necessarily have to be compensating
complexification elsewhere.


From Minkova at HUMnet.UCLA.EDU  Sat Jan 17 17:02:39 1998
From: Minkova at HUMnet.UCLA.EDU (Donka Minkova)
Date: Sat, 17 Jan 1998 12:02:39 EST
Subject: Term: a Chaucerian parallel
Message-ID: <SAT.17.JAN.1998.120239.EST.MINKOVA@HUMNET.UCLA.EDU>

----------------------------Original message----------------------------
 
>(Hey -- how come no women?)
>
>Larry Trask
 
This is not a feminist response to the terminological quest, but it seems
to me that balance in nature is so pervasive that we don't need a special
term for it. The situation is well covered by the Latin legal maxim:
 
"Qui in uno gravatur in alio debet relevari"
 
which Chaucer turns cleverly into:
 
For, John, ther is a lawe that says thus:
That gif a man in a point be agreved,
That in another he sal be releved. . .
 
The context is famous: "John" is one of the clerks in The Reeve's Tale.
Funny things are about to happen. . .
 
Donka Minkova
 
******************************
 
Donka Minkova
English Department
UCLA
Los Angeles, CA 90095
Office (310) 825-4978
Fax (310) 206-5093


From mcv at wxs.nl  Sat Jan 17 17:04:58 1998
From: mcv at wxs.nl (Miguel Carrasquer Vidal)
Date: Sat, 17 Jan 1998 12:04:58 EST
Subject: complexity measures
In-Reply-To: <199801161356.IAA17611@Ruby.deans.umd.edu>
Message-ID: <SAT.17.JAN.1998.120458.EST.>

----------------------------Original message----------------------------
David Lightfoot <dlight at deans.umd.edu> wrote:
 
>   Recent postings suggest that some people believe that languages are all
>equally complex (although this is not entailed by Larry Trask's original
>question, giving rise to this discussion).
 
We're having a parallel discussion on the Nostratic-list about this.
This was my contribution there (which somehow got lost, twice).  Maybe
I'll have better luck here.
 
 
jacob.baltuch at euronet.be (Jacob Baltuch) wrote:
 
>But how does everybody know that "all languages are equivalently
>complex"? Is this something that has been shown or is this just
>dogma?
 
There is no objective measure of "language complexity", so I don't see
how this might have been "shown".  "Dogma", on the other hand, strikes
me as too strong a term.  I'd say "heuristic".  Most languages seem to
be about equivalently complex.  But there certainly are exceptions.
 
There are even terms used to describe the processes involved:
"exoterogeny" [simplification] and "esoterogeny" [complication].
According to Malcolm Ross, in his article "Social Networks and Kinds
of Speech-Community Event" (in: Archaeology and Language I, 1997),
these terms were coined by W.R. Thurston, in the context of
Austronesian/Oceanic linguistics [the languages of New Britain, to be
exact].
 
Quoting from Ross:
 
"If a community has extensive ties with other communities and their
emblematic language is also spoken as a contact language by members of
those communities, then they will probably value their language for
its use across community boundaries.  In the terminology of Thurston
it will be an 'exoteric' lect.  Its use by a wider range of speakers
means that an exoteric lect is subject to considerable variability,
and innovations leading to greater simplicity are liable to be
preferred (and those leading to greater complexity disfavoured).  This
simplifying process Thurston calls "exoterogeny": it reduces
phonological and morphological irregularity or complexity, and makes
the language more regular, more understandable and more learnable.
The outcome of this process is what Platt, seeking a term to describe
the less "educated" forms of Singapore English, called a "creoloid".
[...]  Exoterogeny differs from koineization [discussed previously in
teh article --mcv] in an important respect.  Both koineization and
exoterogeny result in simplification, but koineization also entails
the elimination of emblematic features of its contributing lects (i.e.
levelling).  Exoterogeny does not necessarily involve more than one
lect, so that levelling need not apply.
Esoterogeny is the opposite process.  If the members of a community
have few ties with other communities and their emblematic lect is not
usually known to outsiders, then they may use it as an "in-group"
code, an "esoteric" lect from which outsiders are consciously
excluded.  Innovations leading to increased complexity and to
differences from neighbouring lects will be favoured."
 
I have the impression that bilingualism (multilingualism) is an
important (or maybe even a necessary) aspect here.  Certainly the most
spectacular examples of "esoterogeny" are found in places like
Melanesia, where *everyone* is multilingual, speaking at least their
own "emblematic lect" (subject to esoterogeny) and one or more "lingua
franca" (subject to exoterogeny).  The phenomenon is much less
conspicuous in monolingual societies.
 
I wonder: are the "Italian dialects" more "complex" than Standard
Italian?
 
=======================
Miguel Carrasquer Vidal
mcv at wxs.nl
Amsterdam


From bdbryant at mail.utexas.edu  Sat Jan 17 17:05:26 1998
From: bdbryant at mail.utexas.edu (Bobby D. Bryant)
Date: Sat, 17 Jan 1998 12:05:26 EST
Subject: Q: Minimal words?
Message-ID: <SAT.17.JAN.1998.120526.EST.BDBRYANT@MAIL.UTEXAS.EDU>

----------------------------Original message----------------------------
Does the notion of a prosodic "minimal word" find a happy home among
historical linguists?  In particular, can anyone cite a case where it
has been plausibly invoked as the driving force behind some change, or
as the endorsement for an exception to an otherwise regular change?
 
Bobby Bryant
Austin, Texas


From whiting at cc.helsinki.fi  Sat Jan 17 17:06:54 1998
From: whiting at cc.helsinki.fi (Robert Whiting)
Date: Sat, 17 Jan 1998 12:06:54 EST
Subject: Sum: term
In-Reply-To: <m0xsXTY-0006MiC@rsunx.crn.cogs.susx.ac.uk>
Message-ID: <SAT.17.JAN.1998.120654.EST.WHITING@CC.HELSINKI.FI>

----------------------------Original message----------------------------
On Thu, 15 Jan 1998, Larry Trask wrote:
 
> ----------------------------Original message----------------------------
> A few days ago I posted a request for a term.  Once again, the
> phenomenon I wanted a name for was this: a change that leads to
> simplification in one domain often produces a simultaneous
> complication in another domain.
 
Having been a systems engineer in another life, I can testify that
this is a characteristic of many types of systems, not just language.
In short, if you try to fix ("cleanup") one part of a system, you stand
a good chance of screwing up some other part of the system.  Somewhere
there is going to be an inverse relationship that is going to thwart your
best intentions.  Anyone who has done much engineering or computer
programming will know what I am talking about.
 
...
 
> Anyway, here are the terms proposed, or most of them.  I omit a couple
> of totally facetious suggestions, and one or two which were so
> exceedingly long that I don't think they can be considered as terms.
> A couple of people, I think, thought that I was asking specifically
> for a label for the conversion of phonology into morphology, but in
> fact I have in mind something more general than that.
>
> BLINDNESS PRINCIPLE
> CODE SHIFT
> DIACHRONIC COMPENSATION
> EQUILIBRIUM
> HYDRA'S RAZOR
> LOCAL IMPROVEMENT
> LOCAL SIMPLIFICATION
> MARKEDNESS CONFLICT
> MORPHOLOGIZATION OF PHONOLOGICAL RULES
> NATURALNESS CONFLICT
> SCHLIMMBESSERUNG
> SIMPLEXIFICATION
> STURTEVANT'S PARADOX (unspecified variation on)
> TRADE-OFF
> TUNNEL VISION PRINCIPLE
>
> Right.  Now what do I do?  Call for a vote?  Organize a competition
> with five distinguished judges and a prize of two weeks in the PIE
> homeland of your choice?  Close my eyes and stick a pin?  Ask Roger
> Lass what the biologists call it?  Coin my own term and hope everybody
> buys the book and believes me?  Or should I just admit defeat and not
> include any term for this, on the not unreasonable ground that
> dictionaries shouldn't be including words that don't exist?
>
 
...
 
I'd like to propose yet another term, ISENTROPIC COMPENSATION.  This would
be a mechanism that acts to keep the overall entropy (measure of disorder)
of a system more or less constant.  Languages being natural systems, one
can imagine that entropy would be fairly constant across a given language.
There must be a maximum level of entropy that a language can tolerate in
terms of complexity, learnability, and comprehensibility (it has always
been my pet theory that Sumerian died out because its entropy exceeded
this level and it became incomprehensible to its own speakers).  There is
not necessarily any minimum entropy level to a language (designed
languages can have an entropy level close to zero), but as far as I know
(somebody will correct me if I am wrong), all natuaral language have some
areas of irregularity (high entropy).  It is quite possible that native
speakers subjectively resist any overall reduction of entropy in their
language as being unnatural (just as effective medicine should taste bad
and an effective disinfectant should sting when applied to a wound).  But
in any case, a reduction of entropy in one part of a system very
frequently causes a compensatory increase in another part causing the
overall system entropy to remain at about the same level.
 
 
Bob Whiting
whiting at cc.helsinki.fi


From larryt at cogs.susx.ac.uk  Sat Jan 17 17:07:27 1998
From: larryt at cogs.susx.ac.uk (Larry Trask)
Date: Sat, 17 Jan 1998 12:07:27 EST
Subject: complexity measures
In-Reply-To: <199801161356.IAA17611@Ruby.deans.umd.edu> from "David Lightfoot"
 at Jan 16, 98 11:52:53 am
Message-ID: <SAT.17.JAN.1998.120727.EST.LARRYT@COGS.SUSX.AC.UK>

----------------------------Original message----------------------------
I have received a number of further responses to my summary of the
responses to my original posting seeking a term, and I hope to post a
comprehensive reply to all of them on HISTLING within a couple of
days.  But one point I can clear up right away.
 
David Lightfoot writes:
 
>    I should have thought that if there is a simplification in some
> part of a system, there doesn't necessarily have to be compensating
> complexification elsewhere.
 
Agreed.  It was never my intention to suggest that a simplification
must necessarily be accompanied by a complexification, and I hope I
have not given that impression.  My point was merely that this *often*
happens, and that a name for such a combination would be desirable.
 
Larry Trask
COGS
University of Sussex
Brighton BN1 9QH
UK
 
larryt at cogs.susx.ac.uk


From downingg at is2.nyu.edu  Sat Jan 17 20:17:16 1998
From: downingg at is2.nyu.edu (Gregory {Greg} Downing)
Date: Sat, 17 Jan 1998 15:17:16 EST
Subject: complexity measures
Message-ID: <SAT.17.JAN.1998.151716.EST.DOWNINGG@IS2.NYU.EDU>

----------------------------Original message----------------------------
At 11:52 AM 1/16/98 EST, you (David Lightfoot <dlight at deans.umd.edu>) wrote:
>   Recent postings suggest that some people believe that languages are all
>equally complex (although this is not entailed by Larry Trask's original
>question, giving rise to this discussion).
>   One possibility is that this belief is an empirical finding.  In which
>case, there must be a way of measuring the overall complexity of a language and
>somebody has found that languages all emerge with the same index.
>Alternatively, it might follow from some basic principles or some
>theory that languages must be equally complex. I know of no such
>empirical support nor of any theoretical underpinning for such an idea.
>What am I missing?
>   I should have thought that if there is a simplification in some part
>of a system, there doesn't necessarily have to be compensating
>complexification elsewhere.
>
 
I don't think it has been measured, and to do so in a genuine fashion would
require the most careful calibration and integration of every aspect of a
lot of languages, which I don't believe anyone has done. So it is not a
hypothesis based on all the empirical heavy lifting that would truly be
involved. I'd account for it in cultural-history or professional-history terms.
 
I'm not a linguistics professor but a professor of literature and cultural
history -- but my research area for several years has been ideas about
language in the second half of the nineteenth century and the early
twentieth century, and how those ideas might be related to literary uses of
language in the later-nineteemth and early-twentieth centuries.
 
As everyone probably knows, nineteenth-century language-authors of all
stripes tended to assert that some languages were more complex and
sophisticated, and some simpler and more primitive. When "comparative
philology" was becoming, or being replaced, by "modern linguistics" in the
very late nineteenth and early twentieth century, there was a felt
conceptual and methodological need to avoid making assumptions of this kind,
which seemed not to be based on anything empirical but instead began to be
seen as quite possibly reflections of nonscientific attitudes about which
cultures were prima facie superior and inferior. These attitudes were seen
as standing in the way of taking all languages seriously, thus hampering
sound study of the ones taken less seriously.
 
To get rid of this attitude, the idea was formulated that no language should
be assumed prima facie to be superior or inferior to another, or more
complex or less complex than another. So it's not an empirical hypothesis
grounded in heavy lifting -- more like a methodological axiom intended by
linguists to keep themselves from making unscientific assumptions.
Obviously, for many this has evolved into "all languages are equally
complex," in that way axioms and rules have of tending toward absoluteness.
Not infrequently I heard the assumption stated in just that absolute a way
in undergrad and grad linguistics classes at the University of Michigan
fifteen years ago.
 
So my guess would be that the "all languages are equally complex" idea is an
absolutized version of a very functional methodological axiom -- but
unfortunately stated in a positive, flat-out fashion. The more
unexceptionable form of the axiom is "no language should be assumed prima
facie to be superior or inferior to another, or more complex or less complex
than another," until and unless someone has produced a solid comparison of
lg x and lg y that makes such an argument and finds general assent -- and
during the century now closing there seem to have been more pressing
projects to deal with in lx than global comparisons of lg x and lg y in all
their details.
 
Greg Downing/NYU, at greg.downing at nyu.edu or downingg at is2.nyu.edu


From alderson at netcom.com  Sun Jan 18 13:51:34 1998
From: alderson at netcom.com (Richard M. Alderson III)
Date: Sun, 18 Jan 1998 08:51:34 EST
Subject: Sum: term
In-Reply-To: <Pine.OSF.3.96.980117174208.25471D-100000@vesuri.Helsinki.FI>
 (message from Robert Whiting on Sat, 17 Jan 1998 12:06:54 EST)
Message-ID: <SUN.18.JAN.1998.085134.EST.>

----------------------------Original message----------------------------
Bob Whiting wrote:
 
>In short, if you try to fix ("cleanup") one part of a system, you stand a good
>chance of screwing up some other part of the system.
 
Or as is frequently heard around our engineering group:  "But I only made one
change, and it wasn't even in that part of the code!!!!"  (Usually, but not
always, jokingly...)
 
                                                                Rich Alderson


From delancey at darkwing.uoregon.edu  Sun Jan 18 13:52:32 1998
From: delancey at darkwing.uoregon.edu (Scott DeLancey)
Date: Sun, 18 Jan 1998 08:52:32 EST
Subject: complexity measures
In-Reply-To: <34e71708.113876326@mail.wxs.nl>
Message-ID: <SUN.18.JAN.1998.085232.EST.DELANCEY@DARKWING.UOREGON.EDU>

----------------------------Original message----------------------------
On Sat, 17 Jan 1998, Miguel Carrasquer Vidal wrote:
 
> There is no objective measure of "language complexity", so I don't see
> how this might have been "shown".  "Dogma", on the other hand, strikes
> me as too strong a term.  I'd say "heuristic".  Most languages seem to
> be about equivalently complex.  But there certainly are exceptions.
 
I'm not sure that "dogma" is too strong at all.  Absent any metric
for quantifying overall complexity--without, for that matter, any
metric for quantifying complexity of any given subsystem--any claim
that all (or even most) languages are equivalently complex is simply
meaningless.  There are lots of words one could apply to public
insistence on an empirically vacuous claim; "dogma" doesn't seem
unnecessarily harsh.
    When you say "most languages *seem* to be about equivalently
complex", can you explicate this intuition at all?  When I think
about the question, for languages I know a bit about--i.e. are
English, Hare, Tibetan, Klamath, Sunwar all about equally complex,
or not--I get no intuition at all; I don't have any sense of how
to measure or even estimate complexity so as to make the question
answerable even in principle.
 
Scott DeLancey
Department of Linguistics
University of Oregon
Eugene, OR 97403, USA
 
delancey at darkwing.uoregon.edu
http://www.uoregon.edu/~delancey/prohp.html


From manaster at umich.edu  Sun Jan 18 13:53:34 1998
From: manaster at umich.edu (manaster at umich.edu)
Date: Sun, 18 Jan 1998 08:53:34 EST
Subject: complexity measures
In-Reply-To: <199801161356.IAA17611@Ruby.deans.umd.edu>
Message-ID: <SUN.18.JAN.1998.085334.EST.>

----------------------------Original message----------------------------
I quite agree, but certainly many of us were taught this
as unshakeable dogma in grad school.  I know I was.
It was, of course, a response to the traditional
ideas about the inequality of lgs and the supposed
simplicity of "primitive lgs" as compared to
Latin, Greek, Skt, et al.  So this whole topic
has to be understood in context.  In the context
of trying to refuse 19th/early 20th century ideas
about "primitive lgs", there was something quite
concrete that was it issue adn the authors who argued that African
or Amerindian lgs were no less complex than Latin et al.
had something quite concrete in mind and were entirely
right.  Taken out of that context, the question
becomes either meaningless or requires a new
context--and THAT is how I at any rate would like
to interpret David's (if I may call you that) remarks.
 
Alexis
 
On Fri, 16 Jan 1998, David Lightfoot wrote:
 
> ----------------------------Original message----------------------------
>    Recent postings suggest that some people believe that languages are all
> equally complex (although this is not entailed by Larry Trask's original
> question, giving rise to this discussion).
>    One possibility is that this belief is an empirical finding.  In which
> case, there must be a way of measuring the overall complexity of a language
 and
> somebody has found that languages all emerge with the same index.
> Alternatively, it might follow from some basic principles or some
> theory that languages must be equally complex. I know of no such
> empirical support nor of any theoretical underpinning for such an idea.
> What am I missing?
>    I should have thought that if there is a simplification in some part
> of a system, there doesn't necessarily have to be compensating
> complexification elsewhere.
>


From mcv at wxs.nl  Sun Jan 18 13:54:22 1998
From: mcv at wxs.nl (Miguel Carrasquer Vidal)
Date: Sun, 18 Jan 1998 08:54:22 EST
Subject: complexity measures
In-Reply-To: <Pine.GSO.3.96.980117134647.28245B-100000@darkwing.uoregon.edu>
Message-ID: <SUN.18.JAN.1998.085422.EST.>

----------------------------Original message----------------------------
Scott DeLancey <delancey at darkwing.uoregon.edu> wrote:
 
>On Sat, 17 Jan 1998, Miguel Carrasquer Vidal wrote:
>
>> There is no objective measure of "language complexity", so I don't see
>> how this might have been "shown".  "Dogma", on the other hand, strikes
>> me as too strong a term.  I'd say "heuristic".  Most languages seem to
>> be about equivalently complex.  But there certainly are exceptions.
>
>    When you say "most languages *seem* to be about equivalently
>complex", can you explicate this intuition at all?
 
It's not really a primary intuition.  Everybody knows that learning a
language that is similar to your native one is subjectively "easier".
The "uneducated", "intuitive" point of view is then that some
languages are "easy" (simple) and others "hard" (complex).  Of course,
once you realize/learn that different people classify the "complexity"
of other languages differently according to their native tongue, the
result is a "corrected", "educated" intuition, summarized in "all
languages are about equally complex".  Since it's not based on any
objective measurement, it still isn't in any way a scientific fact.
"Received opinion" ("idie regue") might be a better term.
See Greg Downing's message.
 
Can we get a more objective measure of "language complexity"?  I think
that the passage I quoted from Malcolm Ross' article might hold some
promise.  It suggests that "complexity" is related in some way to the
manner in which a language is used socially.  A language used by a
small "in-group" can afford to be more "complex" than a wide-spread
"koine".  Still not an objective measure, but it might be an
indication of where to look, and what to look for...
 
=======================
Miguel Carrasquer Vidal
mcv at wxs.nl
Amsterdam


From larryt at cogs.susx.ac.uk  Sun Jan 18 13:54:54 1998
From: larryt at cogs.susx.ac.uk (Larry Trask)
Date: Sun, 18 Jan 1998 08:54:54 EST
Subject: Q: Minimal words?
In-Reply-To: <34C01FB9.1A782917@mail.utexas.edu> from "Bobby D. Bryant" at Jan
 17, 98 12:05:26 pm
Message-ID: <SUN.18.JAN.1998.085454.EST.LARRYT@COGS.SUSX.AC.UK>

----------------------------Original message----------------------------
Bobby Bryant writes:
 
 
> Does the notion of a prosodic "minimal word" find a happy home among
> historical linguists?  In particular, can anyone cite a case where
> it has been plausibly invoked as the driving force behind some
> change, or as the endorsement for an exception to an otherwise
> regular change?
 
Yes; there are a number of such cases.  M. Kenstowicz (1994),
Phonology in Generative Grammar, Blackwell, p. 640 ff., cites several
examples of historical changes which were blocked whenever they would
have produced a violation of a minimal word requirement.  Also, if I
remember correctly, Bob Dixon's grammar of Yidiny reports a conspiracy
in that language to ensure that every word-form contains an even
number of syllables.
 
Larry Trask
COGS
University of Sussex
Brighton BN1 9QH
UK
 
larryt at cogs.susx.ac.uk


From kanze at gabi-soft.fr  Sun Jan 18 22:42:19 1998
From: kanze at gabi-soft.fr (J. Kanze)
Date: Sun, 18 Jan 1998 17:42:19 EST
Subject: complexity measures
In-Reply-To: Miguel Carrasquer Vidal's message of Sat, 17 Jan 1998 12:04:58 EST
Message-ID: <SUN.18.JAN.1998.174219.EST.KANZE@GABISOFT.FR>

----------------------------Original message----------------------------
Miguel Carrasquer Vidal <mcv at wxs.nl> writes:
 
|>  I wonder: are the "Italian dialects" more "complex" than Standard
|>  Italian?
 
I'm not sure about Italian dialects, but at least in a certain sense,
phonetically, German dialects are.  I believe that this stems from the
fact that standard German is originally an artificial language, designed
(at least phonetically) to have a certain "regularity".  Thus, the
standard vowels all fit nicely in the classical trapezoid, which is
definitly not the case for Baverian or Alsatian (the two dialects I'm
most familiar with).
 
--
James Kanze    +33 (0)1 39 23 84 71    mailto: kanze at gabi-soft.fr
GABI Software, 22 rue Jacques-Lemercier, 78000 Versailles, France
Conseils en informatique orientie objet --
              -- Beratung in objektorientierter Datenverarbeitung


From kanze at gabi-soft.fr  Sun Jan 18 22:44:15 1998
From: kanze at gabi-soft.fr (J. Kanze)
Date: Sun, 18 Jan 1998 17:44:15 EST
Subject: complexity measures
In-Reply-To: Miguel Carrasquer Vidal's message of Sun, 18 Jan 1998 08:54:22 EST
Message-ID: <SUN.18.JAN.1998.174415.EST.KANZE@GABISOFT.FR>

----------------------------Original message----------------------------
Miguel Carrasquer Vidal <mcv at wxs.nl> writes:
 
|>  Can we get a more objective measure of "language complexity"?
 
Globally, I don't think so.  It's a problem of comparing apples to
oranges.  Is a language with few phonemes but a complex morphology more
complex or less complex than a language with a complex phonemic system
but simple morphology?
 
--
James Kanze    +33 (0)1 39 23 84 71    mailto: kanze at gabi-soft.fr
GABI Software, 22 rue Jacques-Lemercier, 78000 Versailles, France
Conseils en informatique orientie objet --
              -- Beratung in objektorientierter Datenverarbeitung


From reusch at uclink4.berkeley.edu  Sun Jan 18 22:51:40 1998
From: reusch at uclink4.berkeley.edu (B. Reusch)
Date: Sun, 18 Jan 1998 17:51:40 EST
Subject: complexity measures
In-Reply-To: <35039599.211827788@mail.wxs.nl>
Message-ID: <SUN.18.JAN.1998.175140.EST.REUSCH@UCLINK4.BERKELEY.EDU>

----------------------------Original message----------------------------
Miguel Carrasquer Vidal wrote:
> A language used by a
>small "in-group" can afford to be more "complex" than a wide-spread
>"koine".
 
Is what you are referring to in any sense similar to the marked vs unmarked
dichotomy?
 
 
Beatrice Reusch
University of California, Berkeley


From MFCEPRH at fs1.art.man.ac.uk  Mon Jan 19 14:07:07 1998
From: MFCEPRH at fs1.art.man.ac.uk (Richard Hogg)
Date: Mon, 19 Jan 1998 09:07:07 EST
Subject: complexity measures
Message-ID: <MON.19.JAN.1998.090707.EST.>

----------------------------Original message----------------------------
On 16 Jan 98 at 11:52, David Lightfoot wrote:
 
Original message----------------------------
>    Recent postings suggest that some people believe that languages
>    are all
> equally complex (although this is not entailed by Larry Trask's
> original question, giving rise to this discussion).
<snip>
> Alternatively, it might follow from some basic
> principles or some theory that languages must be equally complex. I
> know of no such empirical support nor of any theoretical
> underpinning for such an idea. What am I missing?
>    I should have thought that if there is a simplification in some
>    part
> of a system, there doesn't necessarily have to be compensating
> complexification elsewhere.
 
I'm sure that David's conclusion must be right, but in rejecting the
claim that all languages must be equally complex we have to be equally
careful not to embrace the claim that language change = language
evolution, i.e. that languages "improve" over time.
 
Cases where a simplification in one subsystem leads to
complexification in another subsystem are inherent in a system which
doesn't evolve but simply changes (except, perhaps, sub specie
aeternitatis).
 
Richard Hogg
 
********************************************************************************
 **********************
Richard M. Hogg            Tel: +44(0)161 275 3164
Department of English      Fax: +44(0)161 275 3256
  and American Studies     e-mail: r.m.hogg at man.ac.uk
University of Manchester   web:
 http://www.art.man.ac.uk/english/staff/rmh/home.htm
Oxford Road
Manchester M13 9PL
********************************************************************************
 **********************


From fertig at acsu.buffalo.edu  Mon Jan 19 14:06:45 1998
From: fertig at acsu.buffalo.edu (David L Fertig)
Date: Mon, 19 Jan 1998 09:06:45 EST
Subject: complexity measures
In-Reply-To: <199801161356.IAA17611@Ruby.deans.umd.edu>
Message-ID: <MON.19.JAN.1998.090645.EST.FERTIG@ACSU.BUFFALO.EDU>

----------------------------Original message----------------------------
 
I can imagine someone arguing along the following lines that under
conditions of normal transmission the overall complexity of every natural
language will tend to remain near the maximum that the human brain is
capable of handling.
 
1. Some linguistic changes are "random" from the point of view of
complexity, i.e. complexity/simplicity plays no role in their motivation.
(The side effects of changes that ARE motivated by considerations of
complexity/simplicity would be included in this category.)
 
2. Although individual changes of this type can of course either increase
or decrease complexity, on balance they are bound to increase complexity.
(This is the "entropy" that Robert Whiting discusses in his post. It's
easy to see that it has to be true, since in any system there will always
be more logically possible changes that increase complexity than that
decrease it.)
 
3. Changes that are motivated by considerations of complexity/simplicity,
on the other hand, are only activated to keep languages from exceeding the
complexity limit, i.e.  from becoming unlearnable or unusable. Their
effects still leave a language quite near maximum complexity. Together,
random and "natural" changes will thus tend to keep all normally
transmitted languages very close to the human limit for overall complexity
(and therefore roughly "equally complex"). And even after episodes of
non-normal transmission (such as pidginization), changes of the first type
will gradually restore a language to maximum complexity.
 
I'm not yet sure if I really buy this argument myself, but I'd be
interested in reactions, comments, references.  Obviously, if there's
anything to it, all the details remain to be addressed, in particular the
issue of global vs. local complexity and the significance of the kinds of
sociolinguistic factors that Miguel Carrasquer Vidal discusses in his
contributions to this discussion.
 
While I'm at it, let me throw out one more line of reasoning that arrives
at the same conclusion:
 
Languages (or speakers or learners) do not tolerate purposeless
complexity. It seems to me that this is just a generalized version of the
constraint on acquisition that Clark calls the "Principle of Contrast",
a.k.a. "no exact synonymy". When faced with any complexity, learners will
either figure out a function for it or eliminate it. This means that all
linguistic complexity serves some kind of purpose.
 
There is no logical limit to the amount of complexity that can potentially
be put to good (cognitive and/or communicative) use, and consequently
humans will tend to let their languages become as complex as their brains
can handle. In other words, up to a certain point, speakers/learners will
tend to deal with complexity by putting it to use, only when it goes
beyond that point will they deal with it by eliminating it.
 
 
Thanks in advance for any comments.
 
David Fertig


From bwald at HUMnet.UCLA.EDU  Tue Jan 20 00:27:57 1998
From: bwald at HUMnet.UCLA.EDU (bwald)
Date: Mon, 19 Jan 1998 19:27:57 EST
Subject: complexity measures
Message-ID: <MON.19.JAN.1998.192757.EST.BWALD@HUMNET.UCLA.EDU>

----------------------------Original message----------------------------
David Lightfoot wrote:
 
... it might follow from some basic principles or some
>theory that languages must be equally complex. I know of no such
>empirical support nor of any theoretical underpinning for such an idea.
>What am I missing?
 
My understanding is that linguists make the "equal complexity" assertion
because of the analytical equality of languages principle, fashioned into a
dogma for pedagogical purposes.  David surely knows that there is no agreed
upon measure and that the assertion arises from the dogmatic principle and
not from any measure of simplicity/complexity.  So, you might say that it
is precisely BECAUSE there is no simplicity matrix that this assertion is
made -- as a NULL HYPOTHESIS, if you will.  I do not know the precise
history of the principle in the present context, i.e., why it arose, but I
imagine that it was in the context of postponing (till more important
matters are settled) such ethnocentric claims/impressions as that French is
simpler than Georgian, asserted by an English speaker, or Georgian is
simpler than French or English, asserted by a Kabardian speaker, not to
mention that (written) Chinese is next to impossible to learn in middle age
(regardless of its grammar), asserted by countless Westerners, and so on
(where, obviously, a learnability principle relative to a speaker's first
language motivates judgments of the complexity of other languages).
 
Meanwhile, David is quite right to point out the mystification involved in
taking the principle too seriously.  It would mean that as some process
came to "simplify" some part of the grammar, either,
 
a.  AT THE SAME TIME, some other process would have to "complicate" some
other part of the
 
        grammar, to maintain "balance"
or,
 
b.  the grammar resulting from the simplification would be "unstable" until
compensating
        complexification took place.  (the ultimate in unprovability since
no language is globally
        stable)
 
Such ideas are not so much manifestly incorrect as they are obscure for
purposes of  serious interpretation and empirical testing.  One thing that
IS clear is that the dogma should be recognized as assigning the BURDEN OF
PROOF to being explicit and precise about how languages may differ with
respect to simplicity or complexity before claiming that they are indeed
different in this respect.
 
P.S.  One interesting manifestation of the dogma was the opinion that all
first languages are acquired in "to the same extent" in the same amount of
time by their speakers.  Presumably that meant that for any language, all
first learners will do something like give the impression of having
"mastered" the language at approximately the same age, presumably when
compared somehow with adult speakers of the same language.  Apart from the
fact that it is difficult to draw the line between not-yet-acquired and
basically-already-acquired, Dan Slobin long ago suggested that the basic
premiss is not literally true, and that child acquirers of highly inflected
languages like Russian take longer to learn such things than their
counterparts learning English.  That is an empirical finding.  But, in this
case, we know exactly what Slobin is talking about -- acquisition of
complex case paradigms, something which does not exist to be learned in
English.  I am not suggesting that we go from this to concluding that
Russian as a language is more "complex" than English, but only that its
case paradigms are more complex than English case paradigms -- something a
priori obvious by everyone's understanding of a complexity measure.  What
was not obvious before empirical observation was how long it would take
first language speakers to acquire the Russian cases and their various
forms (at least to ordinary adult mastery).
 
I suspect that there will never be a useful interpretation of such GLOBAL
claims as:
"Languages differ in over-all complexity", or, "Languages are basically the
same in over-all complexity."  But maybe this is just my opinion.-- Benji


From delancey at darkwing.uoregon.edu  Tue Jan 20 14:58:47 1998
From: delancey at darkwing.uoregon.edu (Scott DeLancey)
Date: Tue, 20 Jan 1998 09:58:47 EST
Subject: complexity measures
In-Reply-To: <Pine.GSO.3.96.980119000559.18603C-100000@xena.acsu.buffalo.edu>
Message-ID: <TUE.20.JAN.1998.095847.EST.DELANCEY@DARKWING.UOREGON.EDU>

----------------------------Original message----------------------------
On Mon, 19 Jan 1998, David L Fertig wrote:
 
> Languages (or speakers or learners) do not tolerate purposeless
> complexity. It seems to me that this is just a generalized version of the
> constraint on acquisition that Clark calls the "Principle of Contrast",
> a.k.a. "no exact synonymy". When faced with any complexity, learners will
> either figure out a function for it or eliminate it. This means that all
> linguistic complexity serves some kind of purpose.
 
This too seems like it can't be anything more than an article of faith.
Many Indo-European languages retain, not only grammatical gender,
but distinct, arbitrary declension classes within, or even to an
extent cross-cutting, gender classes.  These have been retained for
millenia.  They certainly add considerably to the overall complexity
of the language.  There's been a little work done (I remember a paper
somewhere by David Zubin) suggesting that there may be some mnemonic
function to grammatical gender, but even if so, what purpose could
possibly be served by the maintenance of significant bodies of irregular
declensions?  For that matter, what purpose is served by English
strong verbs?  True, they seem to be slowly falling out of the
language, but pretty slowly ... the language has tolerated them
for a long time, and doesn't seem to be in a hurry to clean them
up.
 
Scott DeLancey
Department of Linguistics
University of Oregon
Eugene, OR 97403, USA
 
delancey at darkwing.uoregon.edu
http://www.uoregon.edu/~delancey/prohp.html


From bwald at HUMnet.UCLA.EDU  Tue Jan 20 15:07:28 1998
From: bwald at HUMnet.UCLA.EDU (bwald)
Date: Tue, 20 Jan 1998 10:07:28 EST
Subject: complexity measures
Message-ID: <TUE.20.JAN.1998.100728.EST.BWALD@HUMNET.UCLA.EDU>

----------------------------Original message----------------------------
A few additional comments on the complexity issue.
 
First, if I had read the ensuing postings before sending my previous
message I might have more thoughtfully rephrased my use of the term
"dogma".  I wrote:
 
"My understanding is that linguists make the "equal complexity" assertion
because of the analytical equality of languages principle, fashioned into a
dogma for pedagogical purposes"
 
Various other respondents made points similar to mine about the motivation
for the principle, having to do with avoiding ethnocentrism and unfounded
prejudice about what constitutes "complexity".
(Actually, I was also saying that underlying the principle is a stance
claiming the need for the SAME linguistic resources to analyze any
language.)   The "dogma" part simply alludes to the internalization of the
principle so that it becomes dissociated from its purpose and is simply a
reflex reaction to any proposal to the contrary, no matter how reasoned.  I
would not accuse any linguist of doing this, but from my self-monitoring of
how I react to statements or questions about relative complexity  of
different languages by non-linguists, I recognize that the dogma alerts me
to the principle before I question the non-linguists further to see what
they mean.  I hasten to add that I do not dismiss everything a non-linguist
says about language, far from it.  So the dogma acts as a flag, not
necessarily a bad thing, once you get beyond it to the principle it is
intended to flag.
 
Next, I found the discussion stimulating, particularly Fertig's musings,
and Carrasquer Vidal's
citations from Malcolm Ross.  Fertig's leading suggestion is intriguing, to
the effect that languages generally work according to some principle of
entropy at near maximal brain capacity, and therefore they are all
equivalent in overall complexity.  If this is interpretable, and if that
were the case, then indeed as soon as something was simplified in the
grammar, something else would more or less instantaneously occupy the brain
"space" and maintain a constant overall complexity.  I find the idea
interesting, but wanting clarity for any kind of operationalization of the
brain mechanism or mental activity, so that it might become clear how it
could possibly be empirically tested.
 
Both Fertig and Carrasquer Vidal mentioned PIDGINS, as candidates for
"simpler" languages.  This is tricky for a number of reasons.  First,
pidgins are not first languages, and they are known to be parasitic on
first languages for their overall complexity.  For one thing, obviously the
phonological complexity of the pidgin as spoken by any particular speaker
is to some extent parasitic on other languages spoek by the speaker.  So,
it is not a forgone conclusion that pidgins can be examined as independent
languages for purposes of comparison for complexity with other kinds of
languages, most notably the languages of monolinguals or late
multilinguals.
 
Second, as far as comparison or status as independent languages, phenomena
which are called "pidgins" vary greatly in complexity and
conventionalization -- from the kind of on-the-spot makeshift forms of
communication, for which my first point applies most unproblematically, to
highly conventionalized and complex auxiliary languages, such as the
Neo-Melanesian which has a close creole counterpart in Tok Pisin, or some
quite stable varieties of West African pidgin English (which are
historically related to such creoles as Sierra Leonean Krio).  To the
extent that "learnability" can be taken as criterial of "complexity", there
have been provocative statements by some linguists, e.g., the creolist
Derek Bickerton, that creoles are maximally learnable because they are only
minimally arbitrary, with regard to grammar.  However, such a claim has
generated great controversy, to say the least.  And it has certainly not
been empirically demonstrated that they are more learnable than any other
languages.  Bickerton's arguments came from certain theoretical assumptions
he made.  They are no less suggestive than Fertig's proposal, mentioned
above, but they are no more well founded in terms of direct empirical
support.
 
Finally, in considering make-shift pidgins -- and I have been in situations
where I had to try to create one with the cooperation of interlocutors
(maybe most people have, if they have tried to communicate without a common
language), they are very stressful and hardly economical from a
production-perception point of view.  From this I detect an unclarity in
the concept "complexity".  Part of the stress has to do with trying to
communicate with a limited vocabulary, which leads to a lot of longer
circumlocutions according to some syntactic principles.  If we take
"complexity" to be a measure of abstract language knowledge, competence, or
whatever you want to call it, then the burden of using multi-word phrases
instead of semantically more complex (!?) single words is ignored, because
the same syntactic resources may be used in a fluent language as in a
make-shift pidgin.  Nevertheless, the constant appeal to multi-word phrases
in the makeshift pidgin instead of phonologically more compact single words
should count for something "complicated", shouldn't it?  After all, single
words are syntactically less complex than phrases.  How do we measure the
complexity of a language which has a limited vocabulary and constant appeal
to circumlocution involving more complex syntactic phrases against a
language which has a larger vocabulary and (for the sake of argument) the
same syntactic resources?  Should we say that it is "theoretically simpler"
but "more complex in practice"?  To be sure, no make-shift pidgin has
anywhere near the syntactic complexity of a first language; it's not even
close.  But no language belabors its limited resources more than a
makeshift pidgin.  From this at least one thing is clear, the amount of
effort involved is no measure of linguistic complexity.
 
P.S.  With respect to language acquisition, let me end with a reminder that
"language" is an abstraction to begin with.  Thus, who would deny that
child language is less complex than adult language, for the "same"
language.  OK.  Where do we stop?  Wouldn't a 40 year-old's language be
less complex than a 60 year-old's?  Why not?  Because we're not interested
in the additional complexity, say, additional vocabulary acquired in living
that long?  If we are, do we have to compare age-mates across languages to
compare language complexity (assuming we could do it, to begin with)?
Then, given individual and cultural differences, is the claim of equal
complexity one that asserts that for any speaker of a language of any age
we can find some speaker of another language of that age who has equivalent
complexity?  And, if linguists consider additional complexity, most likely
additional lexicon, trivial with regard to the problem, beyond a certain
age, as I suppose they would, then what precisely is non-trivial in
comparing languages "globally" for relative complexity?  This might have a
bearing on whether Fertig's entropy theory is as interesting as it first
appears.  I think it also has a bearing on how interesting, or not, the
notion of "global linguistic complexity" actually is.


From bwald at HUMnet.UCLA.EDU  Wed Jan 21 13:14:31 1998
From: bwald at HUMnet.UCLA.EDU (bwald)
Date: Wed, 21 Jan 1998 08:14:31 EST
Subject: complexity measures
Message-ID: <WED.21.JAN.1998.081431.EST.BWALD@HUMNET.UCLA.EDU>

----------------------------Original message----------------------------
Scott DeLancey asks:
 
For that matter, what purpose is served by English
strong verbs?  True, they seem to be slowly falling out of the
language, but pretty slowly ... the language has tolerated them
for a long time, and doesn't seem to be in a hurry to clean them
up.
 
All I can suggest about English strong verbs is that their past tenses are
certainly more salient than the mono-consonantal form of the -ed suffix,
which is not audible (or even pronounced) in some contexts.  You might also
accept, for similar reasons, that their articulation is simpler, to the
extent that they avoid the creation of consonant clusters like /vd/ in
'saved' (cf. how 'haved' > 'had', 'maked' > 'made', etc. avoids that).  In
that sense, 'brought' requires less articulatory movement and is
perceptually more different from "bring" than "bringed".  More puzzling is
the maintenance of 'double marking', as in 'tell' vs. 'told', 'toll' should
be sufficient, winning according to my criteria, over 'telled'.  A paper by
Labov & Sankoff in Language a number of years ago presented evidence that
well after childhood speakers were still learning to stablize the morpheme
boundary in such words as 'tol(#)d', 'lef(#)t', etc.  The older speakers
were most likely to treat the final segment as an -ed suffix and
omit/delete/fail-to-pronounce it at a rate comparable to the regular past
in the same phonetic environment, e.g., 'call#d', 'stuff#d'.  Younger
speakers (still adult) more often treated it at the much higher rate of
deletion found for single morphemes, as in 'col*d*', 'drif*t*', etc.  To me
this seems like evidence that vowel shifting has some advantages over the
flimsy final consonant for morpheme signalling purposes.
 
Now, wouldn't be interesting if strong verbs, or their remnants, outlive
the -ed past?  I wouldn't put it 'past' them.


From manaster at umich.edu  Wed Jan 21 17:08:50 1998
From: manaster at umich.edu (manaster at umich.edu)
Date: Wed, 21 Jan 1998 12:08:50 EST
Subject: complexity measures
In-Reply-To: <v02140b0db0e9a36fa717@[128.97.208.84]>
Message-ID: <WED.21.JAN.1998.120850.EST.>

----------------------------Original message----------------------------
I repeat: the arguments about all languages being more or less
equally complex arose within the context of early 20th cent.
(or slightly earlier) atempts by Boas, Sapir, Whorf, Jespersen,
and many others to refute earlier prejudices about "primitive"
languages--and made verygood sense indeed (a recent example
occurs in Dixon's Dyirbal grammar).  I am frankly puzled
by the discussion here which refuses to face up to that simple
fact and tries to read sthg more into the topic w/o tehre
being any semblance of atheoretical or empirical basis for
further disucssion, as I think David Lightfoot pointed out.
Is this a symptom of the long-remarked-upon reluctance to
acknowledge that science has a history?
 
AMR


From Alison.Munro at ed.ac.uk  Wed Jan 21 16:21:16 1998
From: Alison.Munro at ed.ac.uk (Alison Munro)
Date: Wed, 21 Jan 1998 11:21:16 EST
Subject: Historical Linguistics
Message-ID: <WED.21.JAN.1998.112116.EST.ALISON.MUNRO@ED.AC.UK>

----------------------------Original message----------------------------
Forthcoming books for linguists...
 
Edinburgh University Press will be publishing two new historical
linguistics titles this Spring:
 
HISTORICAL LINGUISTICS: AN INTRODUCTION by Professor Lyle Campbell of the
University of Canterbury, New Zealand   May 1998  0 7486 0775 7  £14.95
(UK)
A hands-on introduction taking students through the topics, and giving them
abundant examples and exercises to practise the principles and procedures
described.  It covers all the essential topics: grammaticalisation,
sociolinguistic contributions to linguistic change, distant genetic
relationships, areal linguistics and linguistic prehistory.
 
--- "The textbook of choice for courses in historical linguistics"
Professor William J Poser, Stanford University
--- "Absolutely the best textbook in historical linguistics"
Professor Theo Vennemann, University of Munich
 
EVERYDAY ENGLISH: 1500-1700: A READER edited by Bridget Cusack, formerly of
Edinburgh University   March 1998   0 7486 0776 5  £14.95 (UK)
An anthology of contemporary early modern language in more than 60
non-literary texts, with introductions, notes and glossaries.
 
Edinburgh University Press is offering a limited number of inspection
copies to academics teaching courses in the above areas who feel that one
or both of these books would be appropriate for their classes.  Please
contact Alison Munro at Edinburgh University Press with details of your
class name, size and level, your institutional postal address, and the
reason why you think your requested book is suitable.  Email:
Alison.Munro at ed.ac.uk
 
 
Alison Munro
 
Marketing Director, Edinburgh University Press,
22 George Square, Edinburgh EH8 9LF   http://www.eup.ed.ac.uk/


From mc at ai.uga.edu  Thu Jan 22 02:03:20 1998
From: mc at ai.uga.edu (Michael A. Covington)
Date: Wed, 21 Jan 1998 21:03:20 EST
Subject: Computers and mathematics in HL
Message-ID: <WED.21.JAN.1998.210320.EST.MC@AI.UGA.EDU>

----------------------------Original message----------------------------
I'm just wondering... would there be much interest in a special session
on "Computers and Mathematics in Historical Linguistics" at next
year's LSA (Jan. 1999, Los Angeles)?
 
Michael A. Covington                       http://www.ai.uga.edu/~mc
Chairman, Computer Security Team, and
Associate Director, Artificial Intelligence Center
The University of Georgia, Athens, GA 30602-7415 U.S.A.


From manaster at umich.edu  Fri Jan 23 22:35:43 1998
From: manaster at umich.edu (manaster at umich.edu)
Date: Fri, 23 Jan 1998 17:35:43 EST
Subject: No subject
In-Reply-To: <Pine.SOL.3.95.980123123351.3844B-100000@qbert.rs.itd.umich.edu>
Message-ID: <FRI.23.JAN.1998.173543.EST.>

----------------------------Original message----------------------------
There is another issue I would like to broach. As I argued in
my paper on Pakawan and Coahuiltecan in Anthro Lx last year
(and in my survey of Nsotratic in Studies in Lg in 1993),
in recent (and not-so-recent) debates about lg classification,
people too often reason as though refuting a part of somebody
else's argument FOR a proposed classification constitues an
argyument AGAINST that clasifcation.  This seems plain bad
logic to me, and yet it is seems to be prevalent.  Any
comments?  AMR


From delancey at darkwing.uoregon.edu  Fri Jan 23 22:39:34 1998
From: delancey at darkwing.uoregon.edu (Scott DeLancey)
Date: Fri, 23 Jan 1998 17:39:34 EST
Subject: your mail
In-Reply-To: <Pine.SOL.3.95.980123125414.9924B-100000@qbert.rs.itd.umich.edu>
Message-ID: <FRI.23.JAN.1998.173934.EST.DELANCEY@DARKWING.UOREGON.EDU>

----------------------------Original message----------------------------
On Fri, 23 Jan 1998 manaster at umich.edu wrote:
 
> There is another issue I would like to broach. As I argued in
> my paper on Pakawan and Coahuiltecan in Anthro Lx last year
> (and in my survey of Nsotratic in Studies in Lg in 1993),
> in recent (and not-so-recent) debates about lg classification,
> people too often reason as though refuting a part of somebody
> else's argument FOR a proposed classification constitues an
> argyument AGAINST that clasifcation.  This seems plain bad
> logic to me, and yet it is seems to be prevalent.  Any
> comments?  AMR
 
In principle, it isn't necessarily bad logic.  Lacking any evidence
for a relationship between two languages, the null hypothesis must
be that there is no relationship.  (Or, more properly, no relationship
at whatever level the discussion is concerned with--I might sloppily
express disbelief in the relationship of Takelma and Kalapuya, and
mean not that I don't believe that they are both Penutian, but only
that they have no closer relationship than that).  Therefore, if
only a small amount of evidence has been presented for the relationship
of X and Y, an argument which offers an alternative explanation for
that evidence is legitimately an argument against the claimed
relationship.
     Where this does get illogical is where refutation of some *part*
of the evidence which has been presented for a relationship is taken
as an argument against it.  This of course makes no sense at all.
Unfortunately I have to agree with you that this is pretty common.
 
Scott DeLancey
Department of Linguistics
University of Oregon
Eugene, OR 97403, USA
 
delancey at darkwing.uoregon.edu
http://www.uoregon.edu/~delancey/prohp.html


From larryt at cogs.susx.ac.uk  Sat Jan 24 17:45:20 1998
From: larryt at cogs.susx.ac.uk (Larry Trask)
Date: Sat, 24 Jan 1998 12:45:20 EST
Subject: Alexis on classification
In-Reply-To: <Pine.SOL.3.95.980123125414.9924B-100000@qbert.rs.itd.umich.edu>
 from "manaster@umich.edu" at Jan 23, 98 05:35:43 pm
Message-ID: <SAT.24.JAN.1998.124520.EST.LARRYT@COGS.SUSX.AC.UK>

----------------------------Original message----------------------------
Alexis M R writes:
 
> There is another issue I would like to broach. As I argued in
> my paper on Pakawan and Coahuiltecan in Anthro Lx last year
> (and in my survey of Nsotratic in Studies in Lg in 1993),
> in recent (and not-so-recent) debates about lg classification,
> people too often reason as though refuting a part of somebody
> else's argument FOR a proposed classification constitues an
> argyument AGAINST that clasifcation.  This seems plain bad
> logic to me, and yet it is seems to be prevalent.  Any
> comments?  AMR
 
I agree that such reasoning is in principle illogical.  In practice,
though, the outcome is sometimes different.
 
In my own various critiques of comparisons of Basque with language L,
I have invariably concluded that the overwhelming majority of the
individual comparisons adduced fail to stand up, even on the Basque
side alone, before anyone has considered those comparisons from the
other side, and I have further concluded that what is left
undemolished is insufficient to constitute interesting evidence: it's
nothing more than a handful of arbitrary and miscellaneous
resemblances.
 
I therefore see no great obstacle to reasoning as follows: "Look --
you've spent years scouring these languages for evidence of
relatedness, and all you've been able to come up with is this
miserable collection of junk.  Therefore, there are good grounds for
concluding that the languages are not discoverably related at all."
 
So I attach importance to the volume of evidence.  Destroying just one
or two pieces of a more sizeable case does not demolish the case, but
destroying almost all of the case presented leads me to conclude that
the case itself is not there.
 
Larry Trask
COGS
University of Sussex
Brighton BN1 9QH
UK
 
larryt at cogs.susx.ac.uk


From manaster at umich.edu  Sat Jan 24 17:46:25 1998
From: manaster at umich.edu (manaster at umich.edu)
Date: Sat, 24 Jan 1998 12:46:25 EST
Subject: Alexis on classification
In-Reply-To: <m0xw4JN-0006MXC@rsunx.crn.cogs.susx.ac.uk>
Message-ID: <SAT.24.JAN.1998.124625.EST.>

----------------------------Original message----------------------------
 
Larry Trask writes:
 
>
> I therefore see no great obstacle to reasoning as follows: "Look --
> you've spent years scouring these languages for evidence of
> relatedness, and all you've been able to come up with is this
> miserable collection of junk.  Therefore, there are good grounds for
> concluding that the languages are not discoverably related at all."
>
 
There is in my view a subtle fallacy here (and also I think in
Scott Delancey's reply to my posting).  It is true that if someone's
arguments/data are exploded, they have no case left for their theory.
But this does not justify concluding that there IS no "discoverable"
case for the same theory.  For example, I showed that Sapir did not
so much have bad arguments as had no arguments at all for relating the
"Coahuiltecan" languages to each other, only for relating one of them
(Tonkawa) to Hokan (I am oversimplifying slightly).  But there is
lots of other evidence for relating some of tehse languages (esp.
what I call the Pakawan group) to each other. I dont know how many
of you have read my work (in Anthro Lg) or agree with it, but that
is not germane to the point at issue here, namely, that even the
complete failure of someone's arguments for a given lg family
only means that THAT case has failed.  I would of course argue that
Altaic is a good example of this too.  The early work on relating
these languages (e.g., Halevy) was often a joke, but I would argue
that the more recent work establishes their relatedness.  An even
clearer example of the same is Uto-Aztecan, which is clearer simply
because no one questions it now (not even the people like Doerfer or
Campbell who deny Altaic or Pakawan).
 
AMR


From larryt at cogs.susx.ac.uk  Sun Jan 25 22:04:51 1998
From: larryt at cogs.susx.ac.uk (Larry Trask)
Date: Sun, 25 Jan 1998 17:04:51 EST
Subject: Alexis on classification
In-Reply-To: <Pine.SOL.3.95.980124085800.10219C-100000@pacman.rs.itd.umich.edu> from "manaster@umich.edu" at Jan 24,
 98 09:07:50 am
Message-ID: <SUN.25.JAN.1998.170451.EST.LARRYT@COGS.SUSX.AC.UK>

----------------------------Original message----------------------------
Alexis writes:
 
> There is in my view a subtle fallacy here (and also I think in
> Scott Delancey's reply to my posting).  It is true that if someone's
> arguments/data are exploded, they have no case left for their theory.
> But this does not justify concluding that there IS no "discoverable"
> case for the same theory.
 
And I have not said so.  I have only said that prolonged failure to
make a persuasive case leads to a more than reasonable conclusion that
there is probably no case to be made.
 
> For example, I showed that Sapir did not
> so much have bad arguments as had no arguments at all for relating the
> "Coahuiltecan" languages to each other, only for relating one of them
> (Tonkawa) to Hokan (I am oversimplifying slightly).  But there is
> lots of other evidence for relating some of tehse languages (esp.
> what I call the Pakawan group) to each other.
 
I think one might reasonably ask why Sapir thought it was worth while
putting forward a proposal on the basis of no evudence at all.  Even
if someone else later presents evidence for the same proposal, I think
Sapir can be credited with no more than a lucky guess, or at best
perhaps a sixth sense, if you believe in such stuff.  Credit should go
only to the person who comes up with real evidence.
 
I call this the "Democritus fallacy".  A number of ancient Greek
philosophers speculated wildly about the nature of the world, all of
them on the basis of no evidence at all.  Most of their speculations
are dismissed today as empty and worthless.  However, quite by chance,
Democritus's speculation turned out to look something like the atomic
theory settled on by chemists over 2000 years later, on the basis of
evidence.  Consequently, chemistry textbooks often give Democritus
credit for being the founder of the atomic theory.  But this is
absurd: Democritus had no more basis for his speculations than any
other Greek; he just got lucky.
 
The first person to predict that the surface of Venus would prove to
be exceedingly hot was the crackpot Velikovsky.  But he made his guess
on the basis of mad ideas of his own and no evidence -- hence he does
*not* get credit for a prediction that has proved to be true.
 
> I dont know how many
> of you have read my work (in Anthro Lg) or agree with it, but that
> is not germane to the point at issue here, namely, that even the
> complete failure of someone's arguments for a given lg family
> only means that THAT case has failed.
 
Agreed, but surely there must come a time...
 
[snip]
 
Larry Trask
COGS
University of Sussex
Brighton BN1 9QH
UK
 
larryt at cogs.susx.ac.uk


From manaster at umich.edu  Mon Jan 26 22:16:34 1998
From: manaster at umich.edu (manaster at umich.edu)
Date: Mon, 26 Jan 1998 17:16:34 EST
Subject: Alexis on classification
In-Reply-To: <m0xwTTo-0006LJC@rsunx.crn.cogs.susx.ac.uk>
Message-ID: <MON.26.JAN.1998.171634.EST.>

----------------------------Original message----------------------------
Larry Trask writes:
 
I call this the "Democritus fallacy".  A number of ancient Greek
philosophers speculated wildly about the nature of the world, all of
them on the basis of no evidence at all.  Most of their speculations
are dismissed today as empty and worthless.  However, quite by chance,
Democritus's speculation turned out to look something like the atomic
theory settled on by chemists over 2000 years later, on the basis of
evidence.  Consequently, chemistry textbooks often give Democritus
credit for being the founder of the atomic theory.  But this is
absurd: Democritus had no more basis for his speculations than any
other Greek; he just got lucky.
 
**End of quote**
 
But I am not trying to argue that Sapir for example should
get credit for recognizing Pakawan despite the lack of data
in his published work on this.  In fact, I never said anything
about credit.  I am concerned about the fact that a lot of
linguists, yourself among them, have been publishing all kinds
of statements about the validity or otherwise of work in
linguistic classification while staunchly avoiding getting
their hands dirty at all or only doing so in cases where one
is shooting ducks in a barrel.  For example, there has been
a lot of often quite uninformed discussion of Nostratic
but no real attempts to evaluate the actual claims of the
theory (the best-informed work I know, by Brent Vine, does
not even venture outside IE and hence by definition does not
addres any of the comparisons among lg families which are the
whole point of the theory!).  I have seen no substantive
reactions to Vovin's Ainu-Austroasiatic.  Swadesh and Hamp's
comparison of Eskimo-Aluet and Chukchee-Kamchatkan has been
sitting around for decades with almost no response.  My
recent defense of Nadene has elicited no response, and
my work on Pakawan has only resulted in a predictable
response from Campbell (a reply to which will be out soon),
but neither he nor anyone else has looked at the actual
arguments for Pakawan.  Recent attacks on Altaic by
Nichols are based on third-hands sources.  ALmost all
critics of Greenberg's Amerind, myself excepted, have had
nothing more to say that that he got some forms wrong,
but do not address the comparisons he makes at all.
And so on.
 
Of course, I find the uncritical acceptance of untested
or even provably wrong proposals on the other side (whether
in the case of Nostratic, Dene-Caucasian, Amerind, or
any other such proposal) equally distressing.
 
AMR


From alderson at netcom.com  Tue Jan 27 19:04:28 1998
From: alderson at netcom.com (Richard M. Alderson III)
Date: Tue, 27 Jan 1998 14:04:28 EST
Subject: Alexis on classification (again) (fwd)
In-Reply-To: <Pine.SOL.3.95.980127084653.750E-100000@joust.rs.itd.umich.edu>
 (manaster@umich.edu)
Message-ID: <TUE.27.JAN.1998.140428.EST.>

----------------------------Original message----------------------------
Alexis wrote:
 
>Where then are the competent critics of controversial classificatory proposals
>in linguistics?
 
Unborn as yet?  At least, not yet studying linguistics?
 
Nostratic (to take the example of one of the largest families proposed) has had
its Jones (Pedersen) and its Bopps & Rasks (Cuny, Illich-Svitych, et al.), but
it hasn't seen its Schleicher yet, much less its Neogrammarians.
 
It takes a very long time to become a competent Indo-Europeanist, and there at
least the student has the advantage of 200 years of publications and pedagogy.
In order to further Nostratic studies, we need to do with Indo-European what
the Indo-Europeanists have always done with Latin and Greek:  Assume that the
student simply knows those languages, needs to learn Sanskrit and half a dozen
others, and can be taught the rudiments of Indo-European at the same time.
 
Thus, to produce competent Nostraticists, we need to assume a background in
Indo-European, and teach Uralic, Altaic, Kartvelian, and so on, while teaching
the rudiments of Nostratic.  But to do that, we have to have decent historical
work in all those fields that can simply be treated as correct, much as we
treat Greek and Latin and Hittite and Gothic and Vedic and all the others when
we teach (or learn) Indo-European--and we certainly aren't there, yet.  (We
also know, of course, that new discoveries about any language may change every-
thing we thought we knew.)
 
That covers only one "superfamily", of course, and that won't produce those who
might be competent to judge any other "controversial classificatory proposals",
if those who are to be considered competent must be expert in more than one of
the sub-families in the proposal.
 
Alternatively, we might expect that competent critics could be found among
those trained as historical linguists no matter what language area they have
chosen to concentrate their work in, and ask that these judge proposals by the
rigour of the methodology applied to produce them.  Why is this unsatisfactory?
 
                                                                Rich Alderson


From larryt at cogs.susx.ac.uk  Tue Jan 27 14:50:59 1998
From: larryt at cogs.susx.ac.uk (Larry Trask)
Date: Tue, 27 Jan 1998 09:50:59 EST
Subject: Alexis on classification (again)
In-Reply-To: <Pine.SOL.3.95.980127084459.750D-100000@joust.rs.itd.umich.edu>
 from "manaster@umich.edu" at Jan 27, 98 08:46:35 am
Message-ID: <TUE.27.JAN.1998.095059.EST.LARRYT@COGS.SUSX.AC.UK>

----------------------------Original message----------------------------
Alexis writes:
 
> Where then are the competent critics of controversial
> classificatory proposals in linguistics?
 
A very reasonable question, and one I'm not sure has a good answer.
 
You mentioned someone who had criticized Nostratic purely from the IE
point of view.  Well, I suppose he just didn't feel competent to
evaluate the data from the other five families.  So, what can anybody
do except to comment on the data he feels knowledgeable about?  If
specialists from all six families are prepared to comment on the
Nostratic use of data from their specialist families, then that's a
start.  Maybe then someone will feel himself in a position to try to
tie up the various comments.
 
Look: comparative linguistics is hard, and, the vaster the proposal on
the table, the harder it gets either to formulate a decent proposal or
to evaluate the result.  That's just the way it is, I think.  I don't
think these difficulties arise from the shortcomings of linguists; I
think they derive from the nearly intractable nature of the problem.
 
Larry Trask
COGS
University of Sussex
Brighton BN1 9QH
UK
 
larryt at cogs.susx.ac.uk


From manaster at umich.edu  Tue Jan 27 14:50:44 1998
From: manaster at umich.edu (manaster at umich.edu)
Date: Tue, 27 Jan 1998 09:50:44 EST
Subject: Alexis on classification (again)
In-Reply-To: <m0xx80q-0006MdC@rsunx.crn.cogs.susx.ac.uk>
Message-ID: <TUE.27.JAN.1998.095044.EST.>

----------------------------Original message----------------------------
Then just one last question, directed to anyone who cares to
respond, not necessarily Larry, although I would like to hear
from him of course:
 
Where then are the competent critics of controversial
classificatory proposals in linguistics?
 
Alexis MR
 
On Tue, 27 Jan 1998, Larry Trask wrote:
 
> Alexis writes:
>
> > But I am not trying to argue that Sapir for example should
> > get credit for recognizing Pakawan despite the lack of data
> > in his published work on this.  In fact, I never said anything
> > about credit.
>
> OK; my apologies, then, but you did give me the impression that you
> thought Sapir had done something interesting.
>
> > I am concerned about the fact that a lot of linguists, yourself
> > among them, have been publishing all kinds of statements about the
> > validity or otherwise of work in linguistic classification while
> > staunchly avoiding getting their hands dirty at all or only doing so
> > in cases where one is shooting ducks in a barrel.
>
> I have indeed blasted a few encupated anatids, but it's not my fault
> that my specialist language seems to attract nothing else.
>
> As for getting my hands dirty, I have already explained that I do not
> choose to undertake comparative work, because I believe I am better
> qualified to do other things.  We can't *all* be comparativists,
> Alexis -- somebody has to do the work ;-)
>
> [on the absence of informed evaluations of Nostratic and other
> proposals]
>
> It may just be that few of us feel competent to undertake a
> magisterial scrutiny of proposals embracing vast numbers of languages
> and families.
>
> > Of course, I find the uncritical acceptance of untested
> > or even provably wrong proposals on the other side (whether
> > in the case of Nostratic, Dene-Caucasian, Amerind, or
> > any other such proposal) equally distressing.
>
> Good.
>
> Larry Trask
> COGS
> University of Sussex
> Brighton BN1 9QH
> UK
>
> larryt at cogs.susx.ac.uk
>
>
>
>
>


From larryt at cogs.susx.ac.uk  Tue Jan 27 14:44:24 1998
From: larryt at cogs.susx.ac.uk (Larry Trask)
Date: Tue, 27 Jan 1998 09:44:24 EST
Subject: Alexis on classification (again)
In-Reply-To: <Pine.SOL.3.95.980126115328.18961C-100000@joust.rs.itd.umich.edu>
 from "manaster@umich.edu" at Jan 26, 98 05:16:34 pm
Message-ID: <TUE.27.JAN.1998.094424.EST.LARRYT@COGS.SUSX.AC.UK>

----------------------------Original message----------------------------
Alexis writes:
 
> But I am not trying to argue that Sapir for example should
> get credit for recognizing Pakawan despite the lack of data
> in his published work on this.  In fact, I never said anything
> about credit.
 
OK; my apologies, then, but you did give me the impression that you
thought Sapir had done something interesting.
 
> I am concerned about the fact that a lot of linguists, yourself
> among them, have been publishing all kinds of statements about the
> validity or otherwise of work in linguistic classification while
> staunchly avoiding getting their hands dirty at all or only doing so
> in cases where one is shooting ducks in a barrel.
 
I have indeed blasted a few encupated anatids, but it's not my fault
that my specialist language seems to attract nothing else.
 
As for getting my hands dirty, I have already explained that I do not
choose to undertake comparative work, because I believe I am better
qualified to do other things.  We can't *all* be comparativists,
Alexis -- somebody has to do the work ;-)
 
[on the absence of informed evaluations of Nostratic and other
proposals]
 
It may just be that few of us feel competent to undertake a
magisterial scrutiny of proposals embracing vast numbers of languages
and families.
 
> Of course, I find the uncritical acceptance of untested
> or even provably wrong proposals on the other side (whether
> in the case of Nostratic, Dene-Caucasian, Amerind, or
> any other such proposal) equally distressing.
 
Good.
 
Larry Trask
COGS
University of Sussex
Brighton BN1 9QH
UK
 
larryt at cogs.susx.ac.uk


From johanna at uclink.berkeley.edu  Wed Jan 28 03:02:59 1998
From: johanna at uclink.berkeley.edu (Johanna Nichols)
Date: Tue, 27 Jan 1998 22:02:59 EST
Subject: Alexis on classification
In-Reply-To: <Pine.SOL.3.95.980126115328.18961C-100000@joust.rs.itd.umich.edu>
Message-ID: <TUE.27.JAN.1998.220259.EST.JOHANNA@UCLINK.BERKELEY.EDU>

----------------------------Original message----------------------------
>----------------------------Original message----------------------------
Alexis writes:
 
 
>  Recent attacks on Altaic by
>Nichols are based on third-hands sources.
 
Could you supply some references?  I try to keep abreast of this
literature, but I have seen nothing on Altaic by Nichols that I would
describe as either recent or an attack.
 
ALmost all
>critics of Greenberg's Amerind, myself excepted, have had
>nothing more to say that that he got some forms wrong,
>but do not address the comparisons he makes at all.
 
This isn't an accurate characterization of the reviews and critiques I've
read, and I'm sure I've read most of them.  They do criticize the quality
of the data; they also discuss explicitly whether errors in the data
vitiate the comparisons; they  discuss the number and the quality of the
putative cognate sets and the density of language representation within
them; they discuss method more generally; and they ask whether, all in all,
Greenberg's material makes a case for Amerind or its subgroups.  I would
call this addressing the comparisons.
 
 
Johanna Nichols


From delancey at darkwing.uoregon.edu  Wed Jan 28 20:34:58 1998
From: delancey at darkwing.uoregon.edu (Scott DeLancey)
Date: Wed, 28 Jan 1998 15:34:58 EST
Subject: your mail
In-Reply-To: <Pine.SOL.3.95.980127211111.5112D-100000@qbert.rs.itd.umich.edu>
Message-ID: <WED.28.JAN.1998.153458.EST.DELANCEY@DARKWING.UOREGON.EDU>

----------------------------Original message----------------------------
On Tue, 27 Jan 1998 manaster at umich.edu wrote:
 
> Even if ALL the evidence presented by someone for
> theory X is refuted (or if no evidence was ever
> presented in the first place, as happened with Sapir
> several times), that does not mean that X is false
> or incapable of being shown true or even unworthy of
> further effort.
 
I wouldn't dream of disagreeing with this.  I think we're
actually mixing up two separate debates here.
 
I do know the mind-set that you are objecting too; I have a
little experience with this myself.  (Many folks of that temperament
don't like Penutian any bettre than they do Altaic).  And indeed,
the attitude that, if someone can demolish the evidence that's
been put forward for a relationship, then that hypothesis of
relationship can be considered refuted from then on, is irrational
and absurd.
 
But if we ignore those people--i.e. if we argue simply the
logical merits of the issue, rather than the actual ideological
climate of contemporary historical linguistics--then I stand
by my earlier posting.  Myself, I'm a confirmed monogeneticist;
on simple Darwinian grounds monogenesis has to be correct.  So
I assume that all languages are related at some level.  The issue
facing any comparativist working on a relationship that is not yet
established is, are these languages demonstrably related at the
level that I am working at.  As I said, the default hypothesis
has to be no--we cannot claim relationship without some evidence.
Therefore, if all the putative evidence for a relationship has
been refuted, we are back to square one--no evidence, therefore
no assumption of relationship.  (As we all know, there's a lot
of room for disagreement about what kinds and amounts of evidence
are necessary to license a working hypothesis about relationship,
but that's a methodological issue, not a logical one).  No assumption
of relationship means just that--it most definitely does not mean
that investigating the possibility further is an irresponsible
waste of time, as some of our colleagues do seem to assume.
 
Scott DeLancey
Department of Linguistics
University of Oregon
Eugene, OR 97403, USA
 
delancey at darkwing.uoregon.edu
http://www.uoregon.edu/~delancey/prohp.html


From Roger.Wright at liverpool.ac.uk  Wed Jan 28 14:03:54 1998
From: Roger.Wright at liverpool.ac.uk (Roger Wright)
Date: Wed, 28 Jan 1998 09:03:54 EST
Subject: Alexis on classification (again) (fwd)
In-Reply-To: <199801271829.KAA18879@netcom16.netcom.com>
Message-ID: <WED.28.JAN.1998.090354.EST.ROGER.WRIGHT@LIVERPOOL.AC.UK>

----------------------------Original message----------------------------
>
>Alternatively, we might expect that competent critics could be found among
>those trained as historical linguists no matter what language area they have
>chosen to concentrate their work in, and ask that these judge proposals by the
>rigour of the methodology applied to produce them.  Why is this unsatisfactory?
>
 
 
Because whether a hypothesis is right or not has no connection with the
methodological rigour used by its advocates. For example, Ptolemaic
astronomy was very methodologically rigorous, and at first, at least,
Copernicus's ideas were just bright ideas, not worked out with any
precision. This is the converse of what Larry Trask called the
Democritus fallacy; lack of rigour does not imply that a theory is
wrong, merely that it hasn't yet been shown to be right.
        I find rigour off-putting, myself. What is all this geometry
trying to hide?
                                                        Roger Wright


From manaster at umich.edu  Wed Jan 28 14:03:11 1998
From: manaster at umich.edu (manaster at umich.edu)
Date: Wed, 28 Jan 1998 09:03:11 EST
Subject: your mail
In-Reply-To: <199801280305.WAA29457@stayhungry.rs.itd.umich.edu>
Message-ID: <WED.28.JAN.1998.090311.EST.>

----------------------------Original message----------------------------
 
> >I would like to hear from others on this point, but
> >it seems to me that we should add this to Scott's
> >remarks:
> >
> >Even if ALL the evidence presented by someone for
> >theory X is refuted (or if no evidence was ever
> >presented in the first place, as happened with Sapir
> >several times), that does not mean that X is false
> >or incapable of being shown true or even unworthy of
> >further effort.  There is of course no way of knowing
> >whether the effort will be rewarded or not, but
> >SOMEBODY has to keep trying to classify the languages
> >of the world.  I think that perhaps the current
> >crisis in this field has to do in part with false
> >perceptions of the history of the field.  If more people
> >knew about the story of how various classifications
> >were worked out in the course of this century (e.g.,
> >Vietnamese as Mon-Khmer, Pama-Nyungan as Australian,
> >Algic, Anatolian as Indo-European, Eskimo-Aleut, and so on),
> >perhaps there would be more appreciation of just how
> >much of an evolving, progressing field this is and how
> >much realistic work remains to be done--instead of the
> >caricature which we are daily fed by the extremists
> >on both sides.
> >
> >AMR
> >
> >On Fri, 23 Jan 1998, Scott DeLancey wrote:
> >>
> >> ....In principle, it isn't necessarily bad logic.  Lacking any
> >> evidence
> >> for a relationship between two languages, the null hypothesis must
> >> be that there is no relationship....


From manaster at umich.edu  Wed Jan 28 14:02:20 1998
From: manaster at umich.edu (manaster at umich.edu)
Date: Wed, 28 Jan 1998 09:02:20 EST
Subject: Alexis on classification
In-Reply-To: <v03020901b0f41f47b45a@[128.32.100.155]>
Message-ID: <WED.28.JAN.1998.090220.EST.>

----------------------------Original message----------------------------
Johanna Nichols professes not to know what I am talking about
when I say "Recent attacks on Altaic by Nichols are based on third-hands
sources" and asks for references.  Here they are:
 
 Nichols (1992: 4), in an otherwise well-informed
book,  claims that, after eliminating purely typological similarities
between Turkic, Mongolian, and Tungusic,
 
the evidence [sc. for Altaic] was reduced to the pronominal root
resemblances and a set of putative cognates. When the cognates proved not
to be valid, Altaic was abandoned, and the received view now is that
Turkic, Mongolian, and Tungusic are unrelated (see Unger 1990 [=our Unger
1990b]).
 
Nichols, J. 1992. Linguistic Diversity in Space and Time.
Chicago--London: University of Chicago Press.
 
Unger, J. M. 1990b.  Summary report of the Altaic panel.   In Baldi, P.
(ed.), Linguistic change and reconstruction methodology.Berlin--New York:
Mouton de   Gruyter. 479-482.
 
This is from a forthcoming paper by R-S Georg, R. Michalove, P. Sidwell,
and myself, to appear in JL, and dealing with the history and current
state of Altaic studies, in which we discuss in some detail several
recent misrepresentations of the same, by Nichols and others.
 
Of course, one can quibble as to whether 1992 is 'recent'.
 
Alexis
 
PS. As for our conflicting opinions of the mass of work
dealing with Greenberg's Amerind hypothesis, I think it
is more appropriate if I leave it to others to decide whether
my critique is or is not of an entirely different order
(as I believe) than those of most other authors.


From rjanda at midway.uchicago.edu  Thu Jan 29 13:45:52 1998
From: rjanda at midway.uchicago.edu (Richard Janda)
Date: Thu, 29 Jan 1998 08:45:52 EST
Subject: Monogenesis and "simple Darwinian grounds"
Message-ID: <THU.29.JAN.1998.084552.EST.RJANDA@MIDWAY.UCHICAGO.EDU>

----------------------------Original message----------------------------
 
  While agreeing with most of Scott DeLancey's last posting, I believe that
it is a demonstrable fallacy to claim that, "on simple Darwinian grounds",
"monogenesis has to be correct".  I'm no geneticist, but it seems clear to
me that there's a major confusion here between biological prerequisites for
language and the initial set of arbitrary sound/meaning associations charac-
terizing the ultimate ancestor of some language family.  That is, even as-
suming that, due to selection, some mutation which provided the prerequisites
for language came to characterize all living humans at some point, it does
not follow that there was only one occasion on which sound/meaning associa-
tions were arrived at and then passed on to succeeding generations.  Rather,
a language-facilitating mutation could have been selected for but then not
immediately acted on, as it were (and this does not strike me as unlikely).
It hardly strains credulity to imagine that two groups of early humans living
widely separated from each other could independently have stumbled onto the
use of sound/meaning associations for communication in specifically linguistic
ways.  Even if their syntaxes and morphologies were extremely similar or
identical, we again must recognize the possibility that their lexicons (lexi-
ca?) were virtually non-overlapping.  A number of writers have suggested that
pre-existing ritualized behaviors (particularly vocalizations) could have
provided the impetus or at least the model for the development of a lexicon,
and these behaviors might well have been pan-human, but the factor of arbi-
trariness leads us to the conclusion that their specific extensions to new
instances of sound/meaning association need not have been universal.  In that
case, though, we would not have monogenesis of the actual substance of lan-
guages--at least not necessarily.  After all, arbitrariness is a historical
linguist's best friend in arguments intending to sh?w that extensive sound/
meaning correspondences (of certain types) between languages require the as-
sumption of a shared ancestral linguistic stage, rather than borrowing or ac-
cidental convergence.  So doesn't one basic tool of our trade that is con-
stantly used in establishing convergences projected backward from the present
force us to allow for the possibility that primeval arbitrariness permitted
and perhaps even favored polygenesis (as long as we distinguish the monogene-
sis of ability from possibly multiple, independent exercises of that ability)?
     Richard Janda


From sally at isp.pitt.edu  Thu Jan 29 13:53:23 1998
From: sally at isp.pitt.edu (Sarah G. Thomason)
Date: Thu, 29 Jan 1998 08:53:23 EST
Subject: Alexis on classification
In-Reply-To: Your message of "Wed, 28 Jan 1998 09:02:20 EST."
 <Pine.SOL.3.95.980127222724.17496D-100000@qbert.rs.itd.umich.edu>
Message-ID: <THU.29.JAN.1998.085323.EST.SALLY@ISP.PITT.EDU>

----------------------------Original message----------------------------
 
 
Let's see: Alexis Manaster Ramer accuses Johanna Nichols of making
"attacks on Altaic...based on third-hands [sic] sources".  Nichols
asks for references.  Manaster Ramer gives one reference to a work
by Nichols (therefore at most one attack).  As evidence for his
assertion that she attacks Altaic, he quotes her negative
assessment of the evidence in favor of the Altaic hypothesis.  Perhaps
Manaster Ramer views any argument against a proposal that he favors as
an attack on the proposal; I hope and believe that most linguists are
more cautious in using such inflammatory terms.  And it isn't
clear, to me at least, why he says that her assessment is based on
third-hand sources: Unger is a specialist, and the panel he's
reporting on was composed of himself and other specialists.  They
may have been wrong, but even if they were, that wouldn't justify
Manaster Ramer's characterization of their, or of Nichols', comments
on Altaic.  Polemics won't help convince people, so one hopes that
the editors of JL will check Manaster Ramer et al.'s paper carefully
for accuracy and tone before publishing it.
 
  --  Sally Thomason
      sally at isp.pitt.edu
 
P.S. I, for one, disagree with Manaster Ramer's view of the
     value of his own critique of Greenberg's Amerind vs. other
     criticismss of Greenberg's proposals.  Rankins' IJAL review
     is outstanding, as is Poser's IJAL article.  And there are others, too.


From manaster at umich.edu  Thu Jan 29 13:53:52 1998
From: manaster at umich.edu (manaster at umich.edu)
Date: Thu, 29 Jan 1998 08:53:52 EST
Subject: Alexis on classification
In-Reply-To: <1450.886038014@pogo.isp.pitt.edu>
Message-ID: <THU.29.JAN.1998.085352.EST.>

----------------------------Original message----------------------------
Dr. Thomason errs: Unger is not a specialist on the ALtaic
problem and does not claim to be, what Nichols refers to
is Unger's summary of an unpublished paper by Larry Clark
which in turn is supposed to have been a summary for the
panel of the views of Doerfer and Rona-Tas, two well-known
experts on the Altaic question.  This is what makes Nichols'
entirely erroneous statments third-hand, and as to why they
are erroneous, there is no better argument than that Rona-Tas
in fact has consistently refused to reject Altaic (as opposed
to Doerfer). Anyone who knows anything about Altaic studies
would know that, and would also know that Dr. Nichols'
statements are entirely unfounded.  As for Dr. Thomason's
further implications, I would add that (a) one of the coauthors
of my paper in JL is R-S Georg, himself a well-regarded
specialist on the Altaic languages AND an opponent of the
Altaic theory (yes, OPPONENT but a well-informed and honest
one), and (b) I dont think the editors of JL need any lessons
from someone with Dr. Thomason's record.
 
AMR
 
On Wed, 28 Jan 1998, Sarah G. Thomason wrote:
 
>
>
> Let's see: Alexis Manaster Ramer accuses Johanna Nichols of making
> "attacks on Altaic...based on third-hands [sic] sources".  Nichols
> asks for references.  Manaster Ramer gives one reference to a work
> by Nichols (therefore at most one attack).  As evidence for his
> assertion that she attacks Altaic, he quotes her negative
> assessment of the evidence in favor of the Altaic hypothesis.  Perhaps
> Manaster Ramer views any argument against a proposal that he favors as
> an attack on the proposal; I hope and believe that most linguists are
> more cautious in using such inflammatory terms.  And it isn't
> clear, to me at least, why he says that her assessment is based on
> third-hand sources: Unger is a specialist, and the panel he's
> reporting on was composed of himself and other specialists.  They
> may have been wrong, but even if they were, that wouldn't justify
> Manaster Ramer's characterization of their, or of Nichols', comments
> on Altaic.  Polemics won't help convince people, so one hopes that
> the editors of JL will check Manaster Ramer et al.'s paper carefully
> for accuracy and tone before publishing it.
>
>   --  Sally Thomason
>       sally at isp.pitt.edu
>
> P.S. I, for one, disagree with Manaster Ramer's view of the
>      value of his own critique of Greenberg's Amerind vs. other
>      criticismss of Greenberg's proposals.  Rankins' IJAL review
>      is outstanding, as is Poser's IJAL article.  And there are others, too.
>


From johanna at uclink.berkeley.edu  Thu Jan 29 13:58:05 1998
From: johanna at uclink.berkeley.edu (Johanna Nichols)
Date: Thu, 29 Jan 1998 08:58:05 EST
Subject: various posts on classification
Message-ID: <THU.29.JAN.1998.085805.EST.JOHANNA@UCLINK.BERKELEY.EDU>

----------------------------Original message----------------------------
Alexis writes:
 
> Even if ALL the evidence presented by someone for
> theory X is refuted (or if no evidence was ever
> presented in the first place, as happened with Sapir
> several times), that does not mean that X is false
> or incapable of being shown true or even unworthy of
> further effort.
 
and Scott writes:
 
No assumption
of relationship means just that--it most definitely does not mean
that investigating the possibility further is an irresponsible
waste of time, as some of our colleagues do seem to assume.
 
 
I would like to insert the perspective of someone who not only pursues
demonstration of genetic relatedness but also needs to *apply* the results
of historical scholarship in designing samples and doing comparison.  Using
only proven families, evaluating the evidence proposed for relatedness, and
rejecting (as a basis for comparison and sampling) the many proposed
relationships for which no probative evidence has been presented, are
crucial to proper comparison.  Using a consistently designed sample and
reporting the basis for the design to one's readers are responsible
science.  Elsewhere in their posts Alexis mentions 'extremists' and Scott
refers to the 'ideological
climate of contemporary historical linguistics'.  Would that we had, across
most of the historical linguistics literature, a position -- any position
-- as consistent as extremism, or any reliably identifiable ideological
climate.  Then applications could be done with confidence that apples and
apples were being compared.  Note that a wish to compare apples and apples,
or oranges and oranges, or whatever and its ilk, is not the same thing as
liking apples, destroying orange trees, or declaring apples true and
oranges false.  It's just sorting things so you know what you're counting.
 
It is fiendishly difficult even for me, a professional historical linguist,
to sort out all and only apples in the literature.  Little wonder that
geneticists, archeologists, and others can't sort things out properly.  If
my experience is typical, putting the probative evidence up front in more
publications and reviewing it in surveys would foster communication in our
field and between fields and would go far to create consensus about
standards, relatedness, and the worth of different kinds of historical
comparison.


From johanna at uclink.berkeley.edu  Thu Jan 29 21:26:12 1998
From: johanna at uclink.berkeley.edu (Johanna Nichols)
Date: Thu, 29 Jan 1998 16:26:12 EST
Subject: Alexis on classification
In-Reply-To: <Pine.SOL.3.95.980128220716.16857D-100000@qbert.rs.itd.umich.edu>
Message-ID: <THU.29.JAN.1998.162612.EST.JOHANNA@UCLINK.BERKELEY.EDU>

----------------------------Original message----------------------------
Alexis Manaster Ramer writes:
 
  This is what makes Nichols'
>entirely erroneous statments third-hand, and as to why they
>are erroneous, there is no better argument than that Rona-Tas
>in fact has consistently refused to reject Altaic (as opposed
>to Doerfer).
 
        This presupposes that the mere fact that someone believes in it
means that a genetic relationship is real.  It doesn't.  Logically, as
Scott DeLancey has pointed out, the null hypothesis of no relationship
holds until relationship is demonstrated.  Opinion is not the same thing as
demonstrated fact.
 
 
and he writes:
 
Anyone who knows anything about Altaic studies
>would know that, and would also know that Dr. Nichols'
>statements are entirely unfounded.
 
        It should not be necessary to be a specialist in, or even well
acquainted with, a language group or the work on it in order to retrieve
from the literature an accurate assessment of whether the group is a
family.  It's up to the Altaicists to lay out the evidence they regard as
probative and why they regard it as probative.  I've seen nothing I find
probative.  So, though I actually believe that at least Tungusic and
Mongolian are related, I can't treat Altaic as a family for purposes of
sample design, language family censuses, etc.
        I hope the paper by Manaster Ramer et al. reviews the evidence and
not just the history of opinions.
        For what it is worth, in deciding how to regard the genetic status
of Altaic in drawing up the sample for the 1992 book in which I referred to
Unger's 1991 article, I conferred with various people who work on Altaic,
sought information in reference works and overviews, looked over
grammatical paradigms myself, and came to the conclusion that relatedness
wasn't proved and wasn't assumed by enough people in the field to be
regarded as received view.  The 1987 panel discussion reported by Unger,
which I attended, seemed to me to be surveying and presenting received
view, and it reviewed and rejected some of the evidence previously regarded
as probative.  My notes on the panel are not very detailed, and Unger's
piece presents less detail than I recall the actual panel presenting, but I
had taken the panel to be authoritative and therefore took the report of it
to be authoritative.  In any event I came to the panel meeting believing
that Turkic, Tungusic, and Mongolian are a family and that that was
received view, and left the meeting believing that genetic relatedness of
Altaic was not received view among Altaicists and that indeed much of the
evidence was less sound than I had thought.
        I believe this little history underscores the importance of
specialists presenting the evidence (and not just the opinions) in surveys
in the field.  It should be possible for any trained linguist to find, in
reference works, an objective statement of the current received view in the
field and the evidence and standards on which received view is based.
 
        My reading of Alexis Manaster Ramer's posting of Jan. 27 is that it
concedes my two points, namely (1) he knows of no work by me that is either
recent or an attack on Altaic [much less both] and (2) the reviews and
other  critiques of Greenberg's Language in the Americas (1987) do discuss
Greenberg's comparisons.


From delancey at darkwing.uoregon.edu  Thu Jan 29 19:22:01 1998
From: delancey at darkwing.uoregon.edu (Scott DeLancey)
Date: Thu, 29 Jan 1998 14:22:01 EST
Subject: Monogenesis and "simple Darwinian grounds"
In-Reply-To: <199801290434.WAA26863@harper.uchicago.edu>
Message-ID: <THU.29.JAN.1998.142201.EST.DELANCEY@DARKWING.UOREGON.EDU>

----------------------------Original message----------------------------
On Thu, 29 Jan 1998, Richard Janda wrote:
 
>   While agreeing with most of Scott DeLancey's last posting, I believe that
> it is a demonstrable fallacy to claim that, "on simple Darwinian grounds",
> "monogenesis has to be correct".
 
Well, I can't agree with some of your points.  In particular:
 
> terizing the ultimate ancestor of some language family.  That is, even as-
> suming that, due to selection, some mutation which provided the prerequisites
> for language came to characterize all living humans at some point, it does
> not follow that there was only one occasion on which sound/meaning associa-
> tions were arrived at and then passed on to succeeding generations.  Rather,
> a language-facilitating mutation could have been selected for but then not
> immediately acted on, as it were (and this does not strike me as unlikely).
 
This is not only unlikely, it's not even coherent.  How can something be
selected for without being "acted on"?  Selection isn't an abstract
intellectual exercise; it's a label for the fact that some individuals
reproduce more successfully than others, so that their genetic endowment
is disproportionately represented in succeeding generations.  Selection
in connection with a particular trait occurs because that trait leads
to differential reproductive success.  Thus a trait can only be selected
for (or against) if it is manifesting itself in the individuals carrying
it.
     But I had in mind a different Darwinian argument.  Richard says:
 
> It hardly strains credulity to imagine that two groups of early humans living
> widely separated from each other could independently have stumbled onto the
> use of sound/meaning associations for communication in specifically linguistic
> ways.
 
It may not strain credulity, but it's vanishingly unlikely.  Once one
population has developed an advantageous trait, it will then out-reproduce
competing populations which lack that trait, and eventually replace them.
Once one population has developed language, it's not likely that anyone
else will get the chance.  I'm no archeologist, but I understand that
the archeological record suggests an explosively fast expansion of
modern humans (at the expense of Neanderthals).  The obvious inference
is that a single population developed language, giving it a selectional
advantage which allowed it to overspread the world very quickly.
 
Scott DeLancey
Department of Linguistics
University of Oregon
Eugene, OR 97403, USA
 
delancey at darkwing.uoregon.edu
http://www.uoregon.edu/~delancey/prohp.html


From m.cysouw at let.kun.nl  Thu Jan 29 19:20:56 1998
From: m.cysouw at let.kun.nl (Michael Cysouw)
Date: Thu, 29 Jan 1998 14:20:56 EST
Subject: Monogenesis and "simple Darwinian grounds"
In-Reply-To: <199801290434.WAA26863@harper.uchicago.edu>
Message-ID: <THU.29.JAN.1998.142056.EST.M.CYSOUW@LET.KUN.NL>

----------------------------Original message----------------------------
The debate on the monogenesis vs. polygenesis of language normally suffers
from the assumption that we know what we mean when we say 'language' or
'human language'. The conviction people have about monogenesis/polygenesis
seems strongly correlated with their conviction about the specificity of
linguistic behaviour.
 
In it's most general reading, 'language' is a kind of human behaviour with
some language-specific aspects and some general behavioural aspects. One
would want to decide the monogenesis vs. polygenesis debate on the ground
of the origin of the language-specific aspects of humanity (whether
biological or anthropological): were they invented once or more than once?
But as it is still rather unclear where to place the dividing line between
language-specific aspects of human behaviour from language-inspecific
aspects that question is premature.
 
People who believe in monogenesis would normally stress the fact that there
isn't much in linguistic behaviour that we do not find elsewhere in the
human behaviour. They define 'language' so broadly that the question is not
*whether* there has been monogenesis, but rather *which aspects* of this
broadly defined cluster of behavioural traits called language arose before
the splitting of humanity. This seems to be what Scott DeLancey's is
arguing for:
 
> I assume that all languages are related at some level.  The issue
> facing any comparativist working on a relationship that is not yet
> established is, are these languages demonstrably related at the
> level that I am working at.
 
People who believe in (the possibility of) polygenesis normally stress the
fact that there should be some specific universal characteristic of
'language'. The question then remains open when this specific
characteristic arose; it could be monogenetic, but it could also be
polygenetic. This seems to be the possition of Richard Janda. His scenario
needs a strict distinction between the biological preconditions for
language (which are necessary monogenetic) and anthropological use of those
preconditions, which could be imagined to be polygenesic, convergently
leading to one sort of language. The problem of this position always is to
explicate the defining characteristic of 'language' where to decide the
question on.
 
Michael Cysouw
University of Nijmegen, The Netherlands


From sgbrady at ucdavis.edu  Fri Jan 30 00:33:15 1998
From: sgbrady at ucdavis.edu (Sean Brady)
Date: Thu, 29 Jan 1998 19:33:15 EST
Subject: Darwin and language
Message-ID: <THU.29.JAN.1998.193315.EST.SGBRADY@UCDAVIS.EDU>

----------------------------Original message----------------------------
 
Some evolutionary ideas are being discussed in a misinformed fashion.  I
take it that "Darwinian" means something like "due to biological natural
selection on indivduals" (I missed the first message). If so, a few things
need to be cleared up.
 
          The scenario described by Janda:
 
   "That is, even as-
 suming that, due to selection, some mutation which provided the prerequisites
for language came to characterize all living humans at some point, it does
 not follow that there was only one occasion on which sound/meaning associa-
 tions were arrived at and then passed on to succeeding generations.  Rather,
 a language-facilitating mutation could have been selected for but then not
 immediately acted on, as it were (and this does not strike me as unlikely)."
 
is quite reasonable, contrary to the counterargument presented.  This
process, whereby a character that evolved for other functions, or for no
function at all, but which has been co-opted for a new use, is called
exaptation.  Many biological features are thought to have evolved in this
way, perhaps including such dramatic ones as flight in birds.
 
        Also, the process described by DeLancey:
 
  " But I had in mind a different Darwinian argument....  Once one
population has developed an advantageous trait, it will then out-reproduce
competing populations which lack that trait, and eventually replace them.
Once one population has developed language, it's not likely that anyone
else will get the chance.  I'm no archeologist, but I understand that
the archeological record suggests an explosively fast expansion of
modern humans (at the expense of Neanderthals).  The obvious inference
is that a single population developed language, giving it a selectional
advantage which allowed it to overspread the world very quickly"
 
is usually not thought of as "Darwinian" because it defines group
selection, in which populations, and not individuals, are the unit of
selection.  Although group selection has been out of favor for the past few
decades, it is enjoying a comeback lately, and seems a perfectly plausible
mechanism for evolution in organisms such as humans.  But without adequate
information, we can never assume that any population with an advantageous
trait will ALWAYS drive all other population to extinction; there are
countless counterexamples to this assertion in the biological literature.
But in this case, the idea certainly has some merit to it.
 
        Finally, and most importantly, nothing in modern evolutionary
theory that I know of has anything to say about whether a trait can evolve
multiple times or not, which is really what we are debating.  Both
scenarios are perfectly plausible.  This is analagous to the Out-of-Africa
vs. multiregionalism debate in anthropology, in which neither side can say
the other is wrong in theory; evolutionary facts are required to test
alternative hypothesis, each of which are very reasonable.  Until we have
similar types of data concerning langauge evolution (admittedly a lot
harder to get!), I do not think that evoluationary theory has much to
contribue to this debate.
 
 
Sean Brady
Center for Population Biology
University of California, Davis
sgbrady at ucdavis.edu
(916) 752-9977


From johanna at uclink.berkeley.edu  Fri Jan 30 13:15:06 1998
From: johanna at uclink.berkeley.edu (Johanna Nichols)
Date: Fri, 30 Jan 1998 08:15:06 EST
Subject: Monogenesis and "simple Darwinian grounds"
In-Reply-To: <Pine.GSO.3.96.980129083519.13584A-100000@darkwing.uoregon.edu>
Message-ID: <FRI.30.JAN.1998.081506.EST.JOHANNA@UCLINK.BERKELEY.EDU>

----------------------------Original message----------------------------
This has nothing to do with selection, but note that estimates of the size
of the earliest modern human population (ca. 100,000 years ago) range from
a few tens of thousands to a million or more.  The range within which
modern humans evolved, and within which homo erectus evolved, and so on was
comparable in dimensions and shape to New Guinea, and the ecological
context was riverine and lacustrine tropical.  Under comparable ecological
conditions, modern languages of the least complex societies show tremendous
diversity of language families and languages, as well as small
speech-community sizes for languages (a few hundred individuals per
language in many cases).  There would have been room for a hundred or so
languages and a dozen or so distinct language families in the early human
range.  It's hard to imagine that a species could have been viable if its
entire population was comparable to the modal number of speakers per
language in New Guinea and Australia.
 
This makes it hard for me to believe that there was ever just one language.
There might well be a secondary kind of monogenesis in that all but one
line has gone extinct, not for reasons having to do with selection but
accidentally, by ordinary drift (just as, over enough time in a smallish
population, sooner or later all but one of the last names daughter out and
everybody has the same last name, though that doesn't mean they had only
one ancestor).  Unless dialect differentiation and sound change happened at
very different rates or in very different ways or not at all in the early
stages of language development (which I doubt profoundly, since as I
understand it even bird species that learn their songs have song dialects,
which indicates that differentiation with learned transmission isn't unique
to language), I think language as we know it evolved very gradually from
something more primitive and the entire evolution was as a number of
dialects or languages or indeed even language families.
 
If it weren't for the population size and range factors, I'd vote for
monogenesis.
 
Johanna Nichols


From mcv at wxs.nl  Fri Jan 30 18:24:02 1998
From: mcv at wxs.nl (Miguel Carrasquer Vidal)
Date: Fri, 30 Jan 1998 13:24:02 EST
Subject: Monogenesis and "simple Darwinian grounds"
In-Reply-To: <v03020900b0f6dc0246ee@[128.32.100.155]>
Message-ID: <FRI.30.JAN.1998.132402.EST.>

----------------------------Original message----------------------------
Johanna Nichols <johanna at uclink.berkeley.edu> wrote:
 
>This has nothing to do with selection, but note that estimates of the size
>of the earliest modern human population (ca. 100,000 years ago) range from
>a few tens of thousands to a million or more.  The range within which
>modern humans evolved, and within which homo erectus evolved, and so on was
>comparable in dimensions and shape to New Guinea, and the ecological
>context was riverine and lacustrine tropical.  Under comparable ecological
>conditions, modern languages of the least complex societies show tremendous
>diversity of language families and languages, as well as small
>speech-community sizes for languages (a few hundred individuals per
>language in many cases).
 
This whole subject is of course plagued with uncertainties and
assumptions.  I tend to like Eldredge/Gould's theory of "punctuated
equilibrium", which states that speciation generally takes place in
small isolated populations.  I also tend to think that the origin of
"language as we know it" and the origin of our species are related
events, which would indeed take us back to Africa some 100,000 years
ago.  As to population size and area, my guess would be probably less
than a million and somewhat smaller than New Guinea, but somewhere in
that general range.
 
What I don't know is if it's valid to compare the situation in New
Guinea now (or rather, before the colonial era) to the situation in
Africa then.  The linguistic situation in New Guinea is the result of
many thousands of years of human settlement and, presumably,
linguistic differentiation (with little in the way of external
influences outside of the litoral areas where Austronesian languages
are spoken).  The situation at the time of language origin/human
speciation [IF we can equate the two] has no linguistic history, by
definition.
 
>This makes it hard for me to believe that there was ever just one language.
>There might well be a secondary kind of monogenesis in that all but one
>line has gone extinct
 
That idea has always appealed to me.  For one thing, it allows us not
worry about whether homo erectus or h. neanderthaliensis had language.
If they had, those languages are extinct and only the branch(es)
spoken by the earliest h.s.[s.] can survive.  Another situation which
might (but certainly need not) lead to monogenesis of this secondary
kind is the question of early human dispersal.  Even assuming that the
human "homeland" supported several (unrelated?) languages, it may well
be that only one group migrated out, their language becoming the
ancestral language of all varieties spoken outside of a little corner
of Africa.  A small reflux movement would then provide "instant
monogenesis".
 
But let me state again the many uncertainties.  We don't know exactly
how human speciation took place.  We don't know if language predates,
postdates or coincides with the speciation event.  Even assuming a
small isolated "original" population, possessing language, we don't
know how small the original population and area were exactly, making
it hard to say whether a single or multiple dialects/languages may
have been in existence.  We don't know how humans spread out from
their original habitat.  We don't know if secondary extinctions may
have led to "monogenesis of the second kind" at a (much) later date.
 
The odds are that all human languages share a common origin, if not by
primary, then by secondary monogenesis, but there's no way to know for
sure.
 
=======================
Miguel Carrasquer Vidal
mcv at wxs.nl
Amsterdam


From mmorrison at VNET.IBM.COM  Sat Jan 31 16:50:18 1998
From: mmorrison at VNET.IBM.COM (Michael C. Morrison 8-543-4706)
Date: Sat, 31 Jan 1998 11:50:18 EST
Subject: Monogenesis and "simple Darwinian grounds"
Message-ID: <SAT.31.JAN.1998.115018.EST.MMORRISON@VNET.IBM.COM>

----------------------------Original message----------------------------
*** Reply to note of Fri, 30 Jan 1998 13:24:02 -0500 (EST)
*** by mcv at wxs.nl
 
I have to say that this subject is endlessly fascinating, but I also
have to ask, what's the point?  Almost by definition, we will not
know the answers to the questions it raises or the uncertainties
Miguel lists.  Consider current historical linguistic knowledge
and supposition:
 
 - All spoken languages extant today are descended from some prior
   language.  Many of these prior languages are no longer represented
   by any other current usage.  For example, modern Italian can be said
   to have descended from either Latin or a cousin to Latin; neither
   Latin nor a reputed cousin is in use today by native speakers.
 
 - Most (if not all) historical data we have about modern languages
   comes from the written record as presented to us through history
   or archaeology.  At most, this written record gives us insight
   into language as used maybe 8000 years ago.
 
 - Comparative work has given us much data about languages that
   predate the written record.  This data is by no means certain,
   but for some language families, we have some consensus about the
   proto language, for example, proto-Indo-European.  But even these
   proto languages only extend our view of the past by another few
   thousand years or so.
 
So at the best, our current knowledge of our earliest languages only
takes us back 10, maybe 15, thousand years.  What about the other
85,000 years of human language (assuming humans developed language
about 100,000 years ago, as seems probable -- unproven, but likely)?
Can we ever know about these truly ancient languages?  If the answer
is 'no', then the whole monogenesis v. polygenesis argument becomes an
exercise in philosophy, not in linguistics.  If the answer is 'yes',
then we can solve the problem because we will have found either one
superfamily for all languages or several superfamilies.
 
Personally, I don't believe we will be able to reconcile the many
language families of today to a degree that will answer the genesis
question.  Based on the data with which we have to work, I doubt
any overarching superfamily will be accepted by the field as a
whole, and so the debate will continue.  Sigh
 
Michael
----------------------------------------------------------------------
Michael C. Morrison   Santa Teresa Laboratory      Phone (408)463-4706
IMS User Technology   IBM Software Solutions       Fax   (408)463-3696
Lotus Notes ID:       MCMORRIS at IBMUSM50   IBMLink:  MORRISON at TORIBM
Internet ID:          MMORRISON at VNET.IBM.COM  or  USIB47H4 at IBMMAIL.COM
IBM Mail Exchange:    USIB47H4 at IBMMAIL     or   USIB4MCM at IBMMAIL
X.400 Address:        G=mcmgm; S=morrison; P=ibmmail; A=ibmx400; C=us
----------------------------------------------------------------------


From gonzalor at jhu.edu  Mon Jan 12 02:14:06 1998
From: gonzalor at jhu.edu (Gonzalo Rubio)
Date: Sun, 11 Jan 1998 21:14:06 EST
Subject: Quechua en UPenn
Message-ID: <SUN.11.JAN.1998.211406.EST.GONZALOR@JHUNIX.HCF.JHU.EDU>

----------------------------Original message----------------------------
 
I got this from a friend. I think some of you may find it interesting,
____________________________
 
Gonzalo Rubio
Near Eastern Studies
The Johns Hopkins University
gonzalor at jhu.edu
____________________________
 
---------- Forwarded message ----------
Date: Wed, 7 Jan 1998 14:09:15 -0500
From: Ulises Zevallos <uzevallo at NIMBUS.OCIS.TEMPLE.EDU>
Subject: Quechua en UPenn
 
*Elementary Quechua I and II will be offered at The University  of
Pennsylvania in Summer  1998.*  These courses are open to graduate
students, advanced undergraduates, professionals and other interested
individuals.  They will be taught by Serafin Coronel-Molina, a native
speaker of Quechua, using a combination of traditional and multimedia
texts.  The course has been developed in close consultation with Dr.
Nancy H. Hornberger of the Graduate School of Education at the University of
Pennsylvania.
 
Elementary Quechua I (LING 140 for undergraduates; LING 508 for
graduates) will be offered in the first summer session (May 19 to June
26, 1998), and will meet five days a week from 10:00 am to 12:00 noon.
Elementary Quechua II (LING 141 for undergraduates; LING 509 for
graduates) will be offered in the second summer session (June 29 to
August 7, 1998), and will also meet five days a week from 10:00 am to
12:00 noon.
 
Elementary Quechua I is intended for students with no previous study
experience in Quechua.  It introduces students to the language and
culture of the Quechua people.  This is the language that was spoken by
the ancient Incas and is still spoken today by more than 10 million
speakers throughout the Andean countries of South America. The variety
taught will be from the Southern Quechua family spoken in Peru. The
course will promote the development of the four language skills:
speaking, listening, reading and writing, providing a good practical
command of oral and written skills appropriate for everyday situations.
In addition, classwork will include discussion of native Andean culture,
as well as the changing face of Quechua culture in light of recent
migration trends.
 
Elementary Quechua II is a continuation of Elementary Quechua I for
students who have taken the first session course or who have previously
studied Southern Peruvian Quechua at the beginning level.  The format
will be the same as for Elementary Quechua I, with continued building of
the four essential language skills:  speaking, listening, reading and
writing.  Grammatical structures will be continually reviewed throughout
this course, while a rich input of material in Quechua is provided with
the goal of increasing the range of vocabulary and linguistic structures
as well as knowledge of the culture.
 
For further information and registration, contact the Penn Language
Center, 401 Lauder-Fischer Hall, Philadelphia, PA  19104-6330.  Telephone
(215) 898-6039, fax (215) 573-2139.  E-mail <vassilie at sas.upenn.edu> (Ms.
Lada Vassilieva).


From larryt at cogs.susx.ac.uk  Mon Jan 12 02:16:29 1998
From: larryt at cogs.susx.ac.uk (Larry Trask)
Date: Sun, 11 Jan 1998 21:16:29 EST
Subject: Q: term
Message-ID: <SUN.11.JAN.1998.211629.EST.LARRYT@COGS.SUSX.AC.UK>

----------------------------Original message----------------------------
I'm looking for a term.  The phenomenon in question is extremely
familiar, but I don't know of an accepted name for it.
 
The phenomenon is this: a linguistic change which simplifies one
subsystem of a language may complicate another subsystem.
 
A typical example is the history of Spanish mid vowels.  Earlier
Spanish had two low-mid vowels and two high-mid vowels; the low-mid
vowels were *automatically* diphthongized under stress, while the
high-mid vowels were not.  But then the two low-mid vowels merged
with the two higher ones.  This change simplified the phonological
system by removing two phonemes, but it greatly complicated the
morphology: the formerly automatic and transparent diphthongizations
became totally unpredictable and opaque, since some instances of the
new /e/ and /o/ diphthongized while others did not.
 
Does anybody know of an accepted label for this phenomenon, which I
suppose we might elevate to the status of a "principle"?  If not,
wuld anybody like to propose one?
 
Larry Trask
COGS
University of Sussex
Brighton BN1 9QH
England
 
larryt at cogs.susx.ac.uk


From jharvey at ucla.edu  Mon Jan 12 22:04:50 1998
From: jharvey at ucla.edu (Jasmin Harvey)
Date: Mon, 12 Jan 1998 17:04:50 EST
Subject: Quechua en UCLA
Message-ID: <MON.12.JAN.1998.170450.EST.JHARVEY@UCLA.EDU>

----------------------------Original message----------------------------
For those on the west coast who are interested in studying Quechua, it is
also offered fairly often at UCLA as part of the Indigenous Languages of the
Americas section of the Linguistics department.  I don't know whether it
will be offered this following quarter, but the online schedule of classes,
accessible from their web page at http://www.registrar.ucla.edu/schedule/
will allow you to search, and the course description may be found in the
catalogue at http://www.registrar.ucla.edu/catalog/catalog-Indigeno.html
 
Jasmin Harvey
Germanic Linguistics, UCLA
jharvey at ucla.edu


From vbubenik at morgan.ucs.mun.ca  Mon Jan 12 20:18:37 1998
From: vbubenik at morgan.ucs.mun.ca (Vit Bubenik)
Date: Mon, 12 Jan 1998 15:18:37 EST
Subject: Q: term
In-Reply-To: <m0xr1yG-0006MXC@rsunx.crn.cogs.susx.ac.uk>
Message-ID: <MON.12.JAN.1998.151837.EST.VBUBENIK@MORGAN.UCS.MUN.CA>

----------------------------Original message----------------------------
I am using the term "trade-off" between phonology and morphology (or
between morphology and syntax) for this phenomenon.
Vit Bubenik, Linguistics, Memorial University of Newfoundland, St.John's,
Canada.
 
On Sun, 11 Jan 1998, Larry Trask wrote:
 
> ----------------------------Original message----------------------------
> I'm looking for a term.  The phenomenon in question is extremely
> familiar, but I don't know of an accepted name for it.
>
> The phenomenon is this: a linguistic change which simplifies one
> subsystem of a language may complicate another subsystem.
>
> A typical example is the history of Spanish mid vowels.  Earlier
> Spanish had two low-mid vowels and two high-mid vowels; the low-mid
> vowels were *automatically* diphthongized under stress, while the
> high-mid vowels were not.  But then the two low-mid vowels merged
> with the two higher ones.  This change simplified the phonological
> system by removing two phonemes, but it greatly complicated the
> morphology: the formerly automatic and transparent diphthongizations
> became totally unpredictable and opaque, since some instances of the
> new /e/ and /o/ diphthongized while others did not.
>
> Does anybody know of an accepted label for this phenomenon, which I
> suppose we might elevate to the status of a "principle"?  If not,
> wuld anybody like to propose one?
>
> Larry Trask
> COGS
> University of Sussex
> Brighton BN1 9QH
> England
>
> larryt at cogs.susx.ac.uk
>


From tvn at cis.uni-muenchen.de  Mon Jan 12 20:17:09 1998
From: tvn at cis.uni-muenchen.de (Theo Vennemann)
Date: Mon, 12 Jan 1998 15:17:09 EST
Subject: TUNNEL VISION
Message-ID: <MON.12.JAN.1998.151709.EST.TVN@CIS.UNIMUENCHEN.DE>

----------------------------Original message----------------------------
Today I opened the following message signed by Larry Trask.
I would like to respond.
 
>----------------------------Original message----------------------------
>I'm looking for a term.  The phenomenon in question is extremely
>familiar, but I don't know of an accepted name for it.
>
>The phenomenon is this: a linguistic change which simplifies one
>subsystem of a language may complicate another subsystem.
>
>A typical example is the history of Spanish mid vowels.  Earlier
>Spanish had two low-mid vowels and two high-mid vowels; the low-mid
>vowels were *automatically* diphthongized under stress, while the
>high-mid vowels were not.  But then the two low-mid vowels merged
>with the two higher ones.  This change simplified the phonological
>system by removing two phonemes, but it greatly complicated the
>morphology: the formerly automatic and transparent diphthongizations
>became totally unpredictable and opaque, since some instances of the
>new /e/ and /o/ diphthongized while others did not.
>
>Does anybody know of an accepted label for this phenomenon, which I
>suppose we might elevate to the status of a "principle"?  If not,
>wuld anybody like to propose one?
 
Doesn't the example given above fall under the principle that all lan-
guage change is LOCAL improvement, i.e. change on the parameter
it is working on (and may thus cause complications on other parame-
ters)?
 
See "Language change as language improvement", in Vincenzo Orioles,
ed., Modelli esplicativi della diacronia linguistica: Atti del Convegno
della Societ? Italiana di Glottologia, Pavia, 15-17 settembre 1988. Pisa
(Giardini Editori e Stampatori), 1989, 11-35. Reprinted in: Language
change as language improvement", in Charles Jones, ed., Historical
linguistics: Problems and perspectives. London (Longman), 1993,
319-344.
 
The consequence is so self-evident that I do not really think a name is
needed. If things were different, languages would be optimal on all
parameters, which is impossible. Thus, the principle also follows
from the observation that languages keep an overall identical level
of complexity, at least as long as we do not talk about language de-
velopment in terms of the evolution of the species.
 
One could say that language change is, in general, blind to its own side
effects.Thus, if a name were needed, one could call it the blindness
principle. Or if it is to be emphasised that the principle only looks in
its own direction, but neither left nor right, one could call it the
principle of tunnel vision.
 
Theo Vennemann,
12 January 1998.


From bhk at hd1.vsnl.net.in  Tue Jan 13 14:33:32 1998
From: bhk at hd1.vsnl.net.in (Bh. Krishnamurti)
Date: Tue, 13 Jan 1998 09:33:32 EST
Subject: Q: term
Message-ID: <TUE.13.JAN.1998.093332.EST.BHK@HD1.VSNL.NET.IN>

----------------------------Original message----------------------------
>Date: Mon, 12 Jan 1998 22:10:13
>To: Larry Trask <larryt at cogs.susx.ac.uk>
>From: "Bh. Krishnamurti" <bhk at hd1.vsnl.net.in>
>Subject: Re: Q: term
>
>I noticed a similar phenomenon in the history of the Telugu language. I
called the process involved 'morphologization of sound change'. In short,
the process was as follows: Verb roots of the type (C)Vn- (some 7 in number)
when followed by a dental stop in modern Telugu change it to a retroflex,
e.g. an- 'to say' +tu: --> aN-Tu:'saying' (caps for retroflexes). There is
no phonetic motivativation for a dental stop to become a retroflex after a
non-retroflex nasal (dental-alveolar). Historically, the final n represents
a merger of an alveolar n and a retroflex N. The intermediate rules were (1)
n+tt-->n+t't' (a dental t following an an alveolar n became an alveolar t';
(2) some roots ended in N, so N+tt--> N+TT ( a dental t became aa retroflex
T  after a retroflex N): (3) Alveolar t't' merged with retroflex TT
unconditionally; (4)N-->n in all environments except before retrolexes
[phonetically]. So we have roots ending in n (< n and N); t't'-->TT; these
two sound changes are followed by another regular sound change,(5)Geminate
reduction: CC-->C/C+__. Rules (1) to (5)are all automatic. As a consequence
of these we have a complicated morphophonemics in modern Telugu, e.g. an+tu:
--> aN-Tu: 'saying' (older *aN-TTu: < *an-t't'u:); kon-Tu: ('buying' from
*koN+TTu:; older root was *koN-). The problem here is that we have to shift
the cause (conditioning factor) of retroflexion of t to T to the canonical
form of the root (i.e. (C)Vn-) which is not either phonetically or
phonologically a natural cause for a dental to become a retroflex.
>
>In the Spanish case, one solution is to treat diphthongized e and o as a
separate class of stems from those with e and o  which do not diphthongize.
This classification reflects their history. This is also a case of a regular
sound change creating morphological conditioning; simplification achieved in
phonology has a cost in morphology. This phenomenon is common in many languages.
>
>With regards, Bh.K.
>
>
>
>At 21:16 11/01/98 EST, you wrote:
>>----------------------------Original message----------------------------
>>I'm looking for a term.  The phenomenon in question is extremely
>>familiar, but I don't know of an accepted name for it.
>>
>>The phenomenon is this: a linguistic change which simplifies one
>>subsystem of a language may complicate another subsystem.
>>
>>A typical example is the history of Spanish mid vowels.  Earlier
>>Spanish had two low-mid vowels and two high-mid vowels; the low-mid
>>vowels were *automatically* diphthongized under stress, while the
>>high-mid vowels were not.  But then the two low-mid vowels merged
>>with the two higher ones.  This change simplified the phonological
>>system by removing two phonemes, but it greatly complicated the
>>morphology: the formerly automatic and transparent diphthongizations
>>became totally unpredictable and opaque, since some instances of the
>>new /e/ and /o/ diphthongized while others did not.
>>
>>Does anybody know of an accepted label for this phenomenon, which I
>>suppose we might elevate to the status of a "principle"?  If not,
>>wuld anybody like to propose one?
>>
>>Larry Trask
>>COGS
>>University of Sussex
>>Brighton BN1 9QH
>>England
>>
>>larryt at cogs.susx.ac.uk
>>
>>
>
Bh. Krishnamurti
H.No. 12-13-1233, "Bhaarati"
Street No.9, Tarnaka
Hyderabad 500 017, A.P.
India
Telephone (R)(40)701 9665
E-mail: <bhk at hd1.vsnl.net.in>
Note:Please note what follows hd is digit 1 and not letter l.
In vsnl the final character is letter l and not digit 1.


From mcv at wxs.nl  Tue Jan 13 14:32:53 1998
From: mcv at wxs.nl (Miguel Carrasquer Vidal)
Date: Tue, 13 Jan 1998 09:32:53 EST
Subject: Q: term
In-Reply-To: <m0xr1yG-0006MXC@rsunx.crn.cogs.susx.ac.uk>
Message-ID: <TUE.13.JAN.1998.093253.EST.>

----------------------------Original message----------------------------
Larry Trask <larryt at cogs.susx.ac.uk> wrote:
 
>I'm looking for a term.  The phenomenon in question is extremely
>familiar, but I don't know of an accepted name for it.
>
>The phenomenon is this: a linguistic change which simplifies one
>subsystem of a language may complicate another subsystem.
 
In the tradition of Ablaut and Lautgesetz, why not "Schlimmbesserung"?
 
=======================
Miguel Carrasquer Vidal
mcv at wxs.nl
Amsterdam


From fcosw5 at mbm1.scu.edu.tw  Tue Jan 13 14:32:26 1998
From: fcosw5 at mbm1.scu.edu.tw (Steven Schaufele)
Date: Tue, 13 Jan 1998 09:32:26 EST
Subject: diachronic compensation
Message-ID: <TUE.13.JAN.1998.093226.EST.FCOSW5@MBM1.SCU.EDU.TW>

----------------------------Original message----------------------------
Dear fellow historical & comparative linguists,
 
Wrt Larry Trask's recent question about (if i understand him correctly)
a shorthand label for the observation that diachronic `simplification'
in one area of a language's (e.g.) grammar is typically compensated for
by a relative `complication' in another, Theo Vennemann made a valuable
bibliographic recommendation and then wrote:
 
> The consequence is so self-evident that I do not really think a name is
> needed. If things were different, languages would be optimal on all
> parameters, which is impossible. Thus, the principle also follows
> from the observation that languages keep an overall identical level
> of complexity, at least as long as we do not talk about language de-
> velopment in terms of the evolution of the species.
 
It -- the diachronic compensation between different grammar-modules
mentioned earlier, as well as Theo's `observation that languages keep an
overall identical level of complexity' -- is self-evident to *us* who
spend our time and intellectual energies studying the matter; i suppose
the value of sex for rearranging genes is self-evident to geneticists,
biologists, and evolutionists.  But it's very far from evident to the
general public, including the intelligent, relatively well-educated
public.  (I'm thinking of my own parents, among others, who in spite of
being at least tetraglot if not pentaglot still tend to assume that
English as a whole is `simpler' than, e.g., Latin or Polish, and Modern
English is `simpler' than Old English or Modern German.  I'm also
thinking of my own students here in Taiwan, some of whom are astonished
at the notion that Chinese has any `grammar' at all.) This `principle'
that Larry is seeking a convenient label for is something we need to
reiterate iteratively in our dialogue with the general public, and for
that reason i'm entirely sympathetic to Larry's quest.
 
Best,
Steven
--
Steven Schaufele, Ph.D., Asst. Prof. of Linguistics, English Department
 
Soochow University, Waishuanghsi Campus, Taipei 11102, Taiwan, ROC
 
(886)(02)2881-9471 ext. 6504     fcosw5 at mbm1.scu.edu.tw
 
http://www.prairienet.org/~fcosws/homepage.html
 
 
 
        ***O syntagmata linguarum liberemini humanarum!***
 
        ***Nihil vestris privari nisi obicibus potestis!***


From mmorrison at VNET.IBM.COM  Wed Jan 14 14:15:35 1998
From: mmorrison at VNET.IBM.COM (Michael C. Morrison 8-543-4706)
Date: Wed, 14 Jan 1998 09:15:35 EST
Subject: diachronic compensation
Message-ID: <WED.14.JAN.1998.091535.EST.MMORRISON@VNET.IBM.COM>

----------------------------Original message----------------------------
*** Reply to note of Tue, 13 Jan 1998 09:32:26 -0500 (EST)
*** by fcosw5 at mbm1.scu.edu.tw
 
Steven Schaufele <fcosw5 at mbm1.scu.edu.tw> writes:
>Wrt Larry Trask's recent question about (if i understand him correctly)
>a shorthand label for the observation that diachronic `simplification'
>in one area of a language's (e.g.) grammar is typically compensated for
>by a relative `complication' in another,
 
I suggest we could borrow a term from Physics and change its meaning
appropriately:  The Equivalence Principle.  For comparative and
historical linguistics, it would mean that all languages are
_equivalently_ complex, but obviously not complex in the same ways.
And any change to one system or subsystem of a language _maintains_
the overall complexity of the language -- one part gets simpler as
another becomes more complex.
 
Alternately, we could call it the Law of Simplification ... ;)
(I'm just not sure it warrants being called a Law)
 
Michael
----------------------------------------------------------------------
Michael C. Morrison   Santa Teresa Laboratory      Phone (408)463-4706
IMS User Technology   IBM Software Solutions       Fax   (408)463-3696
Lotus Notes ID:       MCMORRIS at IBMUSM50   IBMLink:  MORRISON at TORIBM
Internet ID:          MMORRISON at VNET.IBM.COM  or  USIB47H4 at IBMMAIL.COM
IBM Mail Exchange:    USIB47H4 at IBMMAIL     or   USIB4MCM at IBMMAIL
X.400 Address:        G=mcmgm; S=morrison; P=ibmmail; A=ibmx400; C=us
----------------------------------------------------------------------


From rjanda at midway.uchicago.edu  Wed Jan 14 14:15:10 1998
From: rjanda at midway.uchicago.edu (Richard Janda)
Date: Wed, 14 Jan 1998 09:15:10 EST
Subject: Principles like trade-offs, tunnel vision,
 and local generalizations
Message-ID: <WED.14.JAN.1998.091510.EST.RJANDA@MIDWAY.UCHICAGO.EDU>

----------------------------Original message----------------------------
 
Regarding Larry Trask's query, and following up on Vit Bubenik's and Theo
Vennemann's responses (especially the latter's term "tunnel vision"):
  Certainly the phenomenon in question overlaps with the principles described
by the statement that languages practice therapy, not prophylaxis (most strong-
ly repeated by Lightfoot 1979, I believe, though the idea goes back at least
to Vendryes, doesn't it?), modulo Theo's locality restrictions, and by Sturte-
vant's Paradox about sound-change being regular but causing irregularity (in
morphology), while analogy is irregular (morpholexically sporadic) but rees-
tablishes regularity.
  In my own work, I have emphasized not only grammatical locality but also
data-related locality:  i.e., any language is truly so vast that no speaker
can keep all subsystems and their interrelated consequences in mind when
tempted to make a particular innovation or to apply a given principle which
turns out to yield what is effectively a hypercorrection.  Graphically, I
once prepared a diagram in which language structure was presented as a leaded
glass window looking out on the world, and a speaker's linguistic conscious-
ness was represented as a flashlight beam directed at something in that out-
side world, but broad enough to illuminate only a small area of the language-
structure window.  As a result, it is not surprising that a speaker may notice
competing parts of two patterns, and so be tempted to restructure part of the
window so as to yield a more consistent local pattern (within the flashlight
beam) that is actually not so consistent when viewed more globally.  (In the
diagram I used at the Linguistic Association of the Southwest meeting in Hous-
ton in 1995, the speaker was equipped with a blowtorch for redoing inconsis-
tent parts of leaded-glass windows, though this suggests more conscious tink-
ering with language than I would like to imply.)
  If correct, this factor (what Brian Joseph and I have called "a limited
window of data" restricting the amount of linguistic structure that a single
speaker/hearer can keep in mind) to which Larry Trask has indirectly pointed
in diachrony has the direct implication for synchrony that linguists probably
overestimate the degree to which speakers "consider all the relevant data"
when arriving at (tacit) analyses--hence the prevalence of what Brian and I
have called "local generalizations".  (E.g., the shift from [i] to [eI] for
the first vowel in English _academia_ _ [still not reflected in some diction-
aries] seems to reflect the local influence of _macadamia_ rather than some
general favoring of foreign-like [e] (_anemia_ in the U.S. never has [e], to
the best of my knowledge); still, there are a lot of nuts in academe.)
  With Vennemann, though, I am tempted to think that such limitations of at-
tention are so pervasive in both linguistic and non-linguistic behavior that
they may not need a special name, though "tunnel vision" and consequence-
blindness are suggestive....
     Richard Janda


From DISTERH at UNIVSCVM.SC.EDU  Thu Jan 15 20:36:38 1998
From: DISTERH at UNIVSCVM.SC.EDU (Dorothy Disterheft)
Date: Thu, 15 Jan 1998 15:36:38 EST
Subject: SLE 98 St. Andrews
Message-ID: <THU.15.JAN.1998.153638.EST.DISTERH@UNIVSCVM.SC.EDU>

Invitation
 
to Workshop during the 31st SLE Conference
at St. Andrews, 26-30 August, 1998
 
Workshop organizer: Werner Abraham, Groningen
(E-mail: ABRAHAM at LET.RUG.NL, FAX: +31-50-363 58 21)
 
Workshop topic:
`Spoken and written languages;
their structural and typological differeences'
 
The workshop aims at presentations in the following subfields to
the general title:
parsing strategies divided between spoken and written vernaculars;
typological
differences and historical changes initiated by parsing rather than
logical-structu-
ral triggers (such as, as the underwritten himself will claim in a
contribution to
the workshop, the Upper German preterite decay); typological
differences be-
tween genetically closely related languages such as Afrikaans,
Yiddish and dial-
ectal Germans (more or less strongly svo) vs. written German (sov);
what me-
diates between sov and svo other than (Charles Fries' claim)
distinctions of mor-
phological case? It will be shown in the paper referred to above
that distinctions
of a discourse-functional sort and their prerequsites in structural
terms (wide
middle field!) can contribute to the upkeeping of sov despite the
fact that case
morphology is rather weak (dialectal German and Dutch).
 
Organization: (especially younger) colleagues interested should contact
organizer
at Groningen with an abstract covering an adjacent topic no longer than
one page
(for a 20-30 minutes presentation). The intention is to submit a
collection of the-
matically unified papers to FoL as a self-contained volume.
 
Heed: To attend the conference you need to be a member of the
Societas Linguistica Europaea. Please
contact the local organizers, Dr. Christopher Beedham and Dr.
Isabel Forbes, Dept. of German, School
of Modern Languages, The University, St. Andrews, Fife KY16 9PH,
Scotland/UK, cb1 at st-and.ac.uk,
FAX (01334) 46 36 77, home page:
http://www.st-and.ac.uk/academic/modlangs/SLE98/SLE98.html
 
Mail to: Werner Abraham, Duits-Letteren-RUG,
Oude Kijk in 't Jat Straat 26, NL-9712 EK Groningen


From larryt at cogs.susx.ac.uk  Thu Jan 15 14:36:49 1998
From: larryt at cogs.susx.ac.uk (Larry Trask)
Date: Thu, 15 Jan 1998 09:36:49 EST
Subject: Sum: term
Message-ID: <THU.15.JAN.1998.093649.EST.LARRYT@COGS.SUSX.AC.UK>

----------------------------Original message----------------------------
A few days ago I posted a request for a term.  Once again, the
phenomenon I wanted a name for was this: a change that leads to
simplification in one domain often produces a simultaneous
complication in another domain.
 
The most familiar examples of this phenomenon, of course, involve
phonological simplifications and morphological complications, but
that's not the only possibility, as I perhaps should have pointed out
in my original query.  For example, syntagmatic phonological
simplifications can produce paradigmatic phonological complications,
as when palatalization in palatalizing environments produces new
marked segments, like the Czech fricative trill.  Then again,
analogical leveling (morphological simplification) can produce new
alternations in stems that formerly didn't alternate (morphological
complication), as has happened in some varieties of Serbo-Croatian (if
I'm still allowed to use that name).
 
The motive for my query was this.  As many of you know, I am compiling
a dictionary of historical and comparative linguistics.  Now, in
recent years, we have coined a rather large number of terms in the
field, and I've noticed that good names have been coined for a number
of familiar phenomena for which we formerly had no names; examples are
`actualization' (Timberlake), `metatypy' (Ross), `pandemic
irregularity' (Blust), `exaptation' (Lass), and `phonogenesis'
(Hopper), not to mention the memorable `morphanization' (Matisoff).
 
But I haven't found a recognized name for the phenomenon I'm
interested in here.  But, since the phenomenon, as Steven Schaufele
has pointed out, is such a fundamental one in our field, it seems to
me that we really ought to have a name for it.  Hence my query.
 
Fifteen people replied, and the first thing to report is that there
does indeed appear to be no recognized name for the phenomenon.
Almost everyone had one or more suggestions to make, but no two people
suggested the same term (though in one case two people came fairly
close).  A couple of people suggested terms which they themselves had
apparently used in print, but I guess those proposals haven't caught
on yet.
 
Anyway, here are the terms proposed, or most of them.  I omit a couple
of totally facetious suggestions, and one or two which were so
exceedingly long that I don't think they can be considered as terms.
A couple of people, I think, thought that I was asking specifically
for a label for the conversion of phonology into morphology, but in
fact I have in mind something more general than that.
 
BLINDNESS PRINCIPLE
CODE SHIFT
DIACHRONIC COMPENSATION
EQUILIBRIUM
HYDRA'S RAZOR
LOCAL IMPROVEMENT
LOCAL SIMPLIFICATION
MARKEDNESS CONFLICT
MORPHOLOGIZATION OF PHONOLOGICAL RULES
NATURALNESS CONFLICT
SCHLIMMBESSERUNG
SIMPLEXIFICATION
STURTEVANT'S PARADOX (unspecified variation on)
TRADE-OFF
TUNNEL VISION PRINCIPLE
 
Right.  Now what do I do?  Call for a vote?  Organize a competition
with five distinguished judges and a prize of two weeks in the PIE
homeland of your choice?  Close my eyes and stick a pin?  Ask Roger
Lass what the biologists call it?  Coin my own term and hope everybody
buys the book and believes me?  Or should I just admit defeat and not
include any term for this, on the not unreasonable ground that
dictionaries shouldn't be including words that don't exist?
 
Damned if I know.  But it *would* be nice if we had *some* name for
this.  Otherwise, how can we persuade our students it's important if
we haven't got a name for it?  I mean, I don't recall that so many
Americans go hot and bothered about visiting ever more soldiers and
bombs on the Vietnamese until somebody decided that what was happening
was `escalation', and then suddenly escalation was a hot issue.
 
Anyway, my thanks to Jacob Baltuch, Vit Bubenik, Miguel Carrasquer
Vidal, John Costello, Guy Deutscher, Hans-Olav Engel, Ralf-Stefan
Georg, Harold Koch, Bh. Krishnamurti, Paul Lloyd, Gary Miller, Steven
Schaufele, Theo Vennemann, Benji Wald, and Roger Wright.
 
(Hey -- how come no women?)
 
Larry Trask
COGS
University of Sussex
Brighton BN1 9QH
UK
 
larryt at cogs.susx.ac.uk


From hale1 at alcor.concordia.ca  Fri Jan 16 12:54:30 1998
From: hale1 at alcor.concordia.ca (MARK ROBERT HALE)
Date: Fri, 16 Jan 1998 07:54:30 EST
Subject: Sum: term
In-Reply-To: <m0xsXTY-0006MiC@rsunx.crn.cogs.susx.ac.uk>
Message-ID: <FRI.16.JAN.1998.075430.EST.HALE1@ALCOR.CONCORDIA.CA>

----------------------------Original message----------------------------
 
I apologize for responding late to this discussion. The
state of emergency has limited email opportunities up here.
 
And unfortunately I am not a woman. Anyway,...
 
Though I don't generally engage in terminological matters,
it did strike me that the discussion and terms suggested
addressed several different issues which should not, as
Trask pointed out kind of, be covered under a single
unifying term. "Sturtevant's paradox" and the "Morphologization
of Phonological Rules' refer to relationships between
phonology and morphology only. A few of the suggestions
I ddint' really understand (perhaps some of my email is
still making its way to my door), e.g.," Hydra's razor"
and "code shift".
 
"Equilibrium" seems like a label for a general diachronic
property of languages (if, indeed, it is one -- I personally
do not use the term 'simplification' because (1) I don't
have, and haven't seen, any coherent 'simplicity' metric
for grammars and (2) I can't figure out, for many e.g.
phonological changes that appear a priori to be 'simplifications'
[loss of voicing contrasts in stops, loss of voiceless nasals,
etc.] how they made anything ni the langauges in question any
more complicated -- this leads me to believe that our intuitive
sense of simplicity, if we are to retain a hypothesis of
equilibrium, can't be the right one), rather than for a
specific change of the type under discussion.
 
"Markedness" and "naturalness conflict" (aside from the fact
that I don't believe in 'markedness' or 'naturalness' -- in
a technical sense -- either) seem to describe a state ("conflict")
as well, rather than an event. [Politicians'
attempt to sanitize 'The Korean Conflict' and avoid invoking
the War Powers Act notwithstanding.]
 
On the other hand, "Tunnel Vision Principle", "Blindness
Principle", "Schlimmbesserung",  and "local improvement"
or "Local simplification" fail to capture the fact that
there was, allegedly, a 'compensating' change in another
component. Only "diachronic compensation" really does that.
 
Nevertheless, the 'Tunel Vision' terms turn out,  in my opinion,
to be the better way to state it. After all, in the cases discussed
surely the new morphological system of Spanish didn't come to
have the properties it did to 'compensate' for the changes in the
vowel system -- the morphological changes were the automatic
and epiphenomenal effects of the phonological reanalysis.
 
Surely we can all think of just as many cases in which a
phonological change gave rise to the elimination of morphological
rules or alternations.
 
Mark


From dlight at deans.umd.edu  Fri Jan 16 16:52:53 1998
From: dlight at deans.umd.edu (David Lightfoot)
Date: Fri, 16 Jan 1998 11:52:53 EST
Subject: complexity measures
Message-ID: <FRI.16.JAN.1998.115253.EST.DLIGHT@DEANS.UMD.EDU>

----------------------------Original message----------------------------
   Recent postings suggest that some people believe that languages are all
equally complex (although this is not entailed by Larry Trask's original
question, giving rise to this discussion).
   One possibility is that this belief is an empirical finding.  In which
case, there must be a way of measuring the overall complexity of a language and
somebody has found that languages all emerge with the same index.
Alternatively, it might follow from some basic principles or some
theory that languages must be equally complex. I know of no such
empirical support nor of any theoretical underpinning for such an idea.
What am I missing?
   I should have thought that if there is a simplification in some part
of a system, there doesn't necessarily have to be compensating
complexification elsewhere.


From Minkova at HUMnet.UCLA.EDU  Sat Jan 17 17:02:39 1998
From: Minkova at HUMnet.UCLA.EDU (Donka Minkova)
Date: Sat, 17 Jan 1998 12:02:39 EST
Subject: Term: a Chaucerian parallel
Message-ID: <SAT.17.JAN.1998.120239.EST.MINKOVA@HUMNET.UCLA.EDU>

----------------------------Original message----------------------------
 
>(Hey -- how come no women?)
>
>Larry Trask
 
This is not a feminist response to the terminological quest, but it seems
to me that balance in nature is so pervasive that we don't need a special
term for it. The situation is well covered by the Latin legal maxim:
 
"Qui in uno gravatur in alio debet relevari"
 
which Chaucer turns cleverly into:
 
For, John, ther is a lawe that says thus:
That gif a man in a point be agreved,
That in another he sal be releved. . .
 
The context is famous: "John" is one of the clerks in The Reeve's Tale.
Funny things are about to happen. . .
 
Donka Minkova
 
******************************
 
Donka Minkova
English Department
UCLA
Los Angeles, CA 90095
Office (310) 825-4978
Fax (310) 206-5093


From mcv at wxs.nl  Sat Jan 17 17:04:58 1998
From: mcv at wxs.nl (Miguel Carrasquer Vidal)
Date: Sat, 17 Jan 1998 12:04:58 EST
Subject: complexity measures
In-Reply-To: <199801161356.IAA17611@Ruby.deans.umd.edu>
Message-ID: <SAT.17.JAN.1998.120458.EST.>

----------------------------Original message----------------------------
David Lightfoot <dlight at deans.umd.edu> wrote:
 
>   Recent postings suggest that some people believe that languages are all
>equally complex (although this is not entailed by Larry Trask's original
>question, giving rise to this discussion).
 
We're having a parallel discussion on the Nostratic-list about this.
This was my contribution there (which somehow got lost, twice).  Maybe
I'll have better luck here.
 
 
jacob.baltuch at euronet.be (Jacob Baltuch) wrote:
 
>But how does everybody know that "all languages are equivalently
>complex"? Is this something that has been shown or is this just
>dogma?
 
There is no objective measure of "language complexity", so I don't see
how this might have been "shown".  "Dogma", on the other hand, strikes
me as too strong a term.  I'd say "heuristic".  Most languages seem to
be about equivalently complex.  But there certainly are exceptions.
 
There are even terms used to describe the processes involved:
"exoterogeny" [simplification] and "esoterogeny" [complication].
According to Malcolm Ross, in his article "Social Networks and Kinds
of Speech-Community Event" (in: Archaeology and Language I, 1997),
these terms were coined by W.R. Thurston, in the context of
Austronesian/Oceanic linguistics [the languages of New Britain, to be
exact].
 
Quoting from Ross:
 
"If a community has extensive ties with other communities and their
emblematic language is also spoken as a contact language by members of
those communities, then they will probably value their language for
its use across community boundaries.  In the terminology of Thurston
it will be an 'exoteric' lect.  Its use by a wider range of speakers
means that an exoteric lect is subject to considerable variability,
and innovations leading to greater simplicity are liable to be
preferred (and those leading to greater complexity disfavoured).  This
simplifying process Thurston calls "exoterogeny": it reduces
phonological and morphological irregularity or complexity, and makes
the language more regular, more understandable and more learnable.
The outcome of this process is what Platt, seeking a term to describe
the less "educated" forms of Singapore English, called a "creoloid".
[...]  Exoterogeny differs from koineization [discussed previously in
teh article --mcv] in an important respect.  Both koineization and
exoterogeny result in simplification, but koineization also entails
the elimination of emblematic features of its contributing lects (i.e.
levelling).  Exoterogeny does not necessarily involve more than one
lect, so that levelling need not apply.
Esoterogeny is the opposite process.  If the members of a community
have few ties with other communities and their emblematic lect is not
usually known to outsiders, then they may use it as an "in-group"
code, an "esoteric" lect from which outsiders are consciously
excluded.  Innovations leading to increased complexity and to
differences from neighbouring lects will be favoured."
 
I have the impression that bilingualism (multilingualism) is an
important (or maybe even a necessary) aspect here.  Certainly the most
spectacular examples of "esoterogeny" are found in places like
Melanesia, where *everyone* is multilingual, speaking at least their
own "emblematic lect" (subject to esoterogeny) and one or more "lingua
franca" (subject to exoterogeny).  The phenomenon is much less
conspicuous in monolingual societies.
 
I wonder: are the "Italian dialects" more "complex" than Standard
Italian?
 
=======================
Miguel Carrasquer Vidal
mcv at wxs.nl
Amsterdam


From bdbryant at mail.utexas.edu  Sat Jan 17 17:05:26 1998
From: bdbryant at mail.utexas.edu (Bobby D. Bryant)
Date: Sat, 17 Jan 1998 12:05:26 EST
Subject: Q: Minimal words?
Message-ID: <SAT.17.JAN.1998.120526.EST.BDBRYANT@MAIL.UTEXAS.EDU>

----------------------------Original message----------------------------
Does the notion of a prosodic "minimal word" find a happy home among
historical linguists?  In particular, can anyone cite a case where it
has been plausibly invoked as the driving force behind some change, or
as the endorsement for an exception to an otherwise regular change?
 
Bobby Bryant
Austin, Texas


From whiting at cc.helsinki.fi  Sat Jan 17 17:06:54 1998
From: whiting at cc.helsinki.fi (Robert Whiting)
Date: Sat, 17 Jan 1998 12:06:54 EST
Subject: Sum: term
In-Reply-To: <m0xsXTY-0006MiC@rsunx.crn.cogs.susx.ac.uk>
Message-ID: <SAT.17.JAN.1998.120654.EST.WHITING@CC.HELSINKI.FI>

----------------------------Original message----------------------------
On Thu, 15 Jan 1998, Larry Trask wrote:
 
> ----------------------------Original message----------------------------
> A few days ago I posted a request for a term.  Once again, the
> phenomenon I wanted a name for was this: a change that leads to
> simplification in one domain often produces a simultaneous
> complication in another domain.
 
Having been a systems engineer in another life, I can testify that
this is a characteristic of many types of systems, not just language.
In short, if you try to fix ("cleanup") one part of a system, you stand
a good chance of screwing up some other part of the system.  Somewhere
there is going to be an inverse relationship that is going to thwart your
best intentions.  Anyone who has done much engineering or computer
programming will know what I am talking about.
 
...
 
> Anyway, here are the terms proposed, or most of them.  I omit a couple
> of totally facetious suggestions, and one or two which were so
> exceedingly long that I don't think they can be considered as terms.
> A couple of people, I think, thought that I was asking specifically
> for a label for the conversion of phonology into morphology, but in
> fact I have in mind something more general than that.
>
> BLINDNESS PRINCIPLE
> CODE SHIFT
> DIACHRONIC COMPENSATION
> EQUILIBRIUM
> HYDRA'S RAZOR
> LOCAL IMPROVEMENT
> LOCAL SIMPLIFICATION
> MARKEDNESS CONFLICT
> MORPHOLOGIZATION OF PHONOLOGICAL RULES
> NATURALNESS CONFLICT
> SCHLIMMBESSERUNG
> SIMPLEXIFICATION
> STURTEVANT'S PARADOX (unspecified variation on)
> TRADE-OFF
> TUNNEL VISION PRINCIPLE
>
> Right.  Now what do I do?  Call for a vote?  Organize a competition
> with five distinguished judges and a prize of two weeks in the PIE
> homeland of your choice?  Close my eyes and stick a pin?  Ask Roger
> Lass what the biologists call it?  Coin my own term and hope everybody
> buys the book and believes me?  Or should I just admit defeat and not
> include any term for this, on the not unreasonable ground that
> dictionaries shouldn't be including words that don't exist?
>
 
...
 
I'd like to propose yet another term, ISENTROPIC COMPENSATION.  This would
be a mechanism that acts to keep the overall entropy (measure of disorder)
of a system more or less constant.  Languages being natural systems, one
can imagine that entropy would be fairly constant across a given language.
There must be a maximum level of entropy that a language can tolerate in
terms of complexity, learnability, and comprehensibility (it has always
been my pet theory that Sumerian died out because its entropy exceeded
this level and it became incomprehensible to its own speakers).  There is
not necessarily any minimum entropy level to a language (designed
languages can have an entropy level close to zero), but as far as I know
(somebody will correct me if I am wrong), all natuaral language have some
areas of irregularity (high entropy).  It is quite possible that native
speakers subjectively resist any overall reduction of entropy in their
language as being unnatural (just as effective medicine should taste bad
and an effective disinfectant should sting when applied to a wound).  But
in any case, a reduction of entropy in one part of a system very
frequently causes a compensatory increase in another part causing the
overall system entropy to remain at about the same level.
 
 
Bob Whiting
whiting at cc.helsinki.fi


From larryt at cogs.susx.ac.uk  Sat Jan 17 17:07:27 1998
From: larryt at cogs.susx.ac.uk (Larry Trask)
Date: Sat, 17 Jan 1998 12:07:27 EST
Subject: complexity measures
In-Reply-To: <199801161356.IAA17611@Ruby.deans.umd.edu> from "David Lightfoot"
 at Jan 16, 98 11:52:53 am
Message-ID: <SAT.17.JAN.1998.120727.EST.LARRYT@COGS.SUSX.AC.UK>

----------------------------Original message----------------------------
I have received a number of further responses to my summary of the
responses to my original posting seeking a term, and I hope to post a
comprehensive reply to all of them on HISTLING within a couple of
days.  But one point I can clear up right away.
 
David Lightfoot writes:
 
>    I should have thought that if there is a simplification in some
> part of a system, there doesn't necessarily have to be compensating
> complexification elsewhere.
 
Agreed.  It was never my intention to suggest that a simplification
must necessarily be accompanied by a complexification, and I hope I
have not given that impression.  My point was merely that this *often*
happens, and that a name for such a combination would be desirable.
 
Larry Trask
COGS
University of Sussex
Brighton BN1 9QH
UK
 
larryt at cogs.susx.ac.uk


From downingg at is2.nyu.edu  Sat Jan 17 20:17:16 1998
From: downingg at is2.nyu.edu (Gregory {Greg} Downing)
Date: Sat, 17 Jan 1998 15:17:16 EST
Subject: complexity measures
Message-ID: <SAT.17.JAN.1998.151716.EST.DOWNINGG@IS2.NYU.EDU>

----------------------------Original message----------------------------
At 11:52 AM 1/16/98 EST, you (David Lightfoot <dlight at deans.umd.edu>) wrote:
>   Recent postings suggest that some people believe that languages are all
>equally complex (although this is not entailed by Larry Trask's original
>question, giving rise to this discussion).
>   One possibility is that this belief is an empirical finding.  In which
>case, there must be a way of measuring the overall complexity of a language and
>somebody has found that languages all emerge with the same index.
>Alternatively, it might follow from some basic principles or some
>theory that languages must be equally complex. I know of no such
>empirical support nor of any theoretical underpinning for such an idea.
>What am I missing?
>   I should have thought that if there is a simplification in some part
>of a system, there doesn't necessarily have to be compensating
>complexification elsewhere.
>
 
I don't think it has been measured, and to do so in a genuine fashion would
require the most careful calibration and integration of every aspect of a
lot of languages, which I don't believe anyone has done. So it is not a
hypothesis based on all the empirical heavy lifting that would truly be
involved. I'd account for it in cultural-history or professional-history terms.
 
I'm not a linguistics professor but a professor of literature and cultural
history -- but my research area for several years has been ideas about
language in the second half of the nineteenth century and the early
twentieth century, and how those ideas might be related to literary uses of
language in the later-nineteemth and early-twentieth centuries.
 
As everyone probably knows, nineteenth-century language-authors of all
stripes tended to assert that some languages were more complex and
sophisticated, and some simpler and more primitive. When "comparative
philology" was becoming, or being replaced, by "modern linguistics" in the
very late nineteenth and early twentieth century, there was a felt
conceptual and methodological need to avoid making assumptions of this kind,
which seemed not to be based on anything empirical but instead began to be
seen as quite possibly reflections of nonscientific attitudes about which
cultures were prima facie superior and inferior. These attitudes were seen
as standing in the way of taking all languages seriously, thus hampering
sound study of the ones taken less seriously.
 
To get rid of this attitude, the idea was formulated that no language should
be assumed prima facie to be superior or inferior to another, or more
complex or less complex than another. So it's not an empirical hypothesis
grounded in heavy lifting -- more like a methodological axiom intended by
linguists to keep themselves from making unscientific assumptions.
Obviously, for many this has evolved into "all languages are equally
complex," in that way axioms and rules have of tending toward absoluteness.
Not infrequently I heard the assumption stated in just that absolute a way
in undergrad and grad linguistics classes at the University of Michigan
fifteen years ago.
 
So my guess would be that the "all languages are equally complex" idea is an
absolutized version of a very functional methodological axiom -- but
unfortunately stated in a positive, flat-out fashion. The more
unexceptionable form of the axiom is "no language should be assumed prima
facie to be superior or inferior to another, or more complex or less complex
than another," until and unless someone has produced a solid comparison of
lg x and lg y that makes such an argument and finds general assent -- and
during the century now closing there seem to have been more pressing
projects to deal with in lx than global comparisons of lg x and lg y in all
their details.
 
Greg Downing/NYU, at greg.downing at nyu.edu or downingg at is2.nyu.edu


From alderson at netcom.com  Sun Jan 18 13:51:34 1998
From: alderson at netcom.com (Richard M. Alderson III)
Date: Sun, 18 Jan 1998 08:51:34 EST
Subject: Sum: term
In-Reply-To: <Pine.OSF.3.96.980117174208.25471D-100000@vesuri.Helsinki.FI>
 (message from Robert Whiting on Sat, 17 Jan 1998 12:06:54 EST)
Message-ID: <SUN.18.JAN.1998.085134.EST.>

----------------------------Original message----------------------------
Bob Whiting wrote:
 
>In short, if you try to fix ("cleanup") one part of a system, you stand a good
>chance of screwing up some other part of the system.
 
Or as is frequently heard around our engineering group:  "But I only made one
change, and it wasn't even in that part of the code!!!!"  (Usually, but not
always, jokingly...)
 
                                                                Rich Alderson


From delancey at darkwing.uoregon.edu  Sun Jan 18 13:52:32 1998
From: delancey at darkwing.uoregon.edu (Scott DeLancey)
Date: Sun, 18 Jan 1998 08:52:32 EST
Subject: complexity measures
In-Reply-To: <34e71708.113876326@mail.wxs.nl>
Message-ID: <SUN.18.JAN.1998.085232.EST.DELANCEY@DARKWING.UOREGON.EDU>

----------------------------Original message----------------------------
On Sat, 17 Jan 1998, Miguel Carrasquer Vidal wrote:
 
> There is no objective measure of "language complexity", so I don't see
> how this might have been "shown".  "Dogma", on the other hand, strikes
> me as too strong a term.  I'd say "heuristic".  Most languages seem to
> be about equivalently complex.  But there certainly are exceptions.
 
I'm not sure that "dogma" is too strong at all.  Absent any metric
for quantifying overall complexity--without, for that matter, any
metric for quantifying complexity of any given subsystem--any claim
that all (or even most) languages are equivalently complex is simply
meaningless.  There are lots of words one could apply to public
insistence on an empirically vacuous claim; "dogma" doesn't seem
unnecessarily harsh.
    When you say "most languages *seem* to be about equivalently
complex", can you explicate this intuition at all?  When I think
about the question, for languages I know a bit about--i.e. are
English, Hare, Tibetan, Klamath, Sunwar all about equally complex,
or not--I get no intuition at all; I don't have any sense of how
to measure or even estimate complexity so as to make the question
answerable even in principle.
 
Scott DeLancey
Department of Linguistics
University of Oregon
Eugene, OR 97403, USA
 
delancey at darkwing.uoregon.edu
http://www.uoregon.edu/~delancey/prohp.html


From manaster at umich.edu  Sun Jan 18 13:53:34 1998
From: manaster at umich.edu (manaster at umich.edu)
Date: Sun, 18 Jan 1998 08:53:34 EST
Subject: complexity measures
In-Reply-To: <199801161356.IAA17611@Ruby.deans.umd.edu>
Message-ID: <SUN.18.JAN.1998.085334.EST.>

----------------------------Original message----------------------------
I quite agree, but certainly many of us were taught this
as unshakeable dogma in grad school.  I know I was.
It was, of course, a response to the traditional
ideas about the inequality of lgs and the supposed
simplicity of "primitive lgs" as compared to
Latin, Greek, Skt, et al.  So this whole topic
has to be understood in context.  In the context
of trying to refuse 19th/early 20th century ideas
about "primitive lgs", there was something quite
concrete that was it issue adn the authors who argued that African
or Amerindian lgs were no less complex than Latin et al.
had something quite concrete in mind and were entirely
right.  Taken out of that context, the question
becomes either meaningless or requires a new
context--and THAT is how I at any rate would like
to interpret David's (if I may call you that) remarks.
 
Alexis
 
On Fri, 16 Jan 1998, David Lightfoot wrote:
 
> ----------------------------Original message----------------------------
>    Recent postings suggest that some people believe that languages are all
> equally complex (although this is not entailed by Larry Trask's original
> question, giving rise to this discussion).
>    One possibility is that this belief is an empirical finding.  In which
> case, there must be a way of measuring the overall complexity of a language
 and
> somebody has found that languages all emerge with the same index.
> Alternatively, it might follow from some basic principles or some
> theory that languages must be equally complex. I know of no such
> empirical support nor of any theoretical underpinning for such an idea.
> What am I missing?
>    I should have thought that if there is a simplification in some part
> of a system, there doesn't necessarily have to be compensating
> complexification elsewhere.
>


From mcv at wxs.nl  Sun Jan 18 13:54:22 1998
From: mcv at wxs.nl (Miguel Carrasquer Vidal)
Date: Sun, 18 Jan 1998 08:54:22 EST
Subject: complexity measures
In-Reply-To: <Pine.GSO.3.96.980117134647.28245B-100000@darkwing.uoregon.edu>
Message-ID: <SUN.18.JAN.1998.085422.EST.>

----------------------------Original message----------------------------
Scott DeLancey <delancey at darkwing.uoregon.edu> wrote:
 
>On Sat, 17 Jan 1998, Miguel Carrasquer Vidal wrote:
>
>> There is no objective measure of "language complexity", so I don't see
>> how this might have been "shown".  "Dogma", on the other hand, strikes
>> me as too strong a term.  I'd say "heuristic".  Most languages seem to
>> be about equivalently complex.  But there certainly are exceptions.
>
>    When you say "most languages *seem* to be about equivalently
>complex", can you explicate this intuition at all?
 
It's not really a primary intuition.  Everybody knows that learning a
language that is similar to your native one is subjectively "easier".
The "uneducated", "intuitive" point of view is then that some
languages are "easy" (simple) and others "hard" (complex).  Of course,
once you realize/learn that different people classify the "complexity"
of other languages differently according to their native tongue, the
result is a "corrected", "educated" intuition, summarized in "all
languages are about equally complex".  Since it's not based on any
objective measurement, it still isn't in any way a scientific fact.
"Received opinion" ("idie regue") might be a better term.
See Greg Downing's message.
 
Can we get a more objective measure of "language complexity"?  I think
that the passage I quoted from Malcolm Ross' article might hold some
promise.  It suggests that "complexity" is related in some way to the
manner in which a language is used socially.  A language used by a
small "in-group" can afford to be more "complex" than a wide-spread
"koine".  Still not an objective measure, but it might be an
indication of where to look, and what to look for...
 
=======================
Miguel Carrasquer Vidal
mcv at wxs.nl
Amsterdam


From larryt at cogs.susx.ac.uk  Sun Jan 18 13:54:54 1998
From: larryt at cogs.susx.ac.uk (Larry Trask)
Date: Sun, 18 Jan 1998 08:54:54 EST
Subject: Q: Minimal words?
In-Reply-To: <34C01FB9.1A782917@mail.utexas.edu> from "Bobby D. Bryant" at Jan
 17, 98 12:05:26 pm
Message-ID: <SUN.18.JAN.1998.085454.EST.LARRYT@COGS.SUSX.AC.UK>

----------------------------Original message----------------------------
Bobby Bryant writes:
 
 
> Does the notion of a prosodic "minimal word" find a happy home among
> historical linguists?  In particular, can anyone cite a case where
> it has been plausibly invoked as the driving force behind some
> change, or as the endorsement for an exception to an otherwise
> regular change?
 
Yes; there are a number of such cases.  M. Kenstowicz (1994),
Phonology in Generative Grammar, Blackwell, p. 640 ff., cites several
examples of historical changes which were blocked whenever they would
have produced a violation of a minimal word requirement.  Also, if I
remember correctly, Bob Dixon's grammar of Yidiny reports a conspiracy
in that language to ensure that every word-form contains an even
number of syllables.
 
Larry Trask
COGS
University of Sussex
Brighton BN1 9QH
UK
 
larryt at cogs.susx.ac.uk


From kanze at gabi-soft.fr  Sun Jan 18 22:42:19 1998
From: kanze at gabi-soft.fr (J. Kanze)
Date: Sun, 18 Jan 1998 17:42:19 EST
Subject: complexity measures
In-Reply-To: Miguel Carrasquer Vidal's message of Sat, 17 Jan 1998 12:04:58 EST
Message-ID: <SUN.18.JAN.1998.174219.EST.KANZE@GABISOFT.FR>

----------------------------Original message----------------------------
Miguel Carrasquer Vidal <mcv at wxs.nl> writes:
 
|>  I wonder: are the "Italian dialects" more "complex" than Standard
|>  Italian?
 
I'm not sure about Italian dialects, but at least in a certain sense,
phonetically, German dialects are.  I believe that this stems from the
fact that standard German is originally an artificial language, designed
(at least phonetically) to have a certain "regularity".  Thus, the
standard vowels all fit nicely in the classical trapezoid, which is
definitly not the case for Baverian or Alsatian (the two dialects I'm
most familiar with).
 
--
James Kanze    +33 (0)1 39 23 84 71    mailto: kanze at gabi-soft.fr
GABI Software, 22 rue Jacques-Lemercier, 78000 Versailles, France
Conseils en informatique orientie objet --
              -- Beratung in objektorientierter Datenverarbeitung


From kanze at gabi-soft.fr  Sun Jan 18 22:44:15 1998
From: kanze at gabi-soft.fr (J. Kanze)
Date: Sun, 18 Jan 1998 17:44:15 EST
Subject: complexity measures
In-Reply-To: Miguel Carrasquer Vidal's message of Sun, 18 Jan 1998 08:54:22 EST
Message-ID: <SUN.18.JAN.1998.174415.EST.KANZE@GABISOFT.FR>

----------------------------Original message----------------------------
Miguel Carrasquer Vidal <mcv at wxs.nl> writes:
 
|>  Can we get a more objective measure of "language complexity"?
 
Globally, I don't think so.  It's a problem of comparing apples to
oranges.  Is a language with few phonemes but a complex morphology more
complex or less complex than a language with a complex phonemic system
but simple morphology?
 
--
James Kanze    +33 (0)1 39 23 84 71    mailto: kanze at gabi-soft.fr
GABI Software, 22 rue Jacques-Lemercier, 78000 Versailles, France
Conseils en informatique orientie objet --
              -- Beratung in objektorientierter Datenverarbeitung


From reusch at uclink4.berkeley.edu  Sun Jan 18 22:51:40 1998
From: reusch at uclink4.berkeley.edu (B. Reusch)
Date: Sun, 18 Jan 1998 17:51:40 EST
Subject: complexity measures
In-Reply-To: <35039599.211827788@mail.wxs.nl>
Message-ID: <SUN.18.JAN.1998.175140.EST.REUSCH@UCLINK4.BERKELEY.EDU>

----------------------------Original message----------------------------
Miguel Carrasquer Vidal wrote:
> A language used by a
>small "in-group" can afford to be more "complex" than a wide-spread
>"koine".
 
Is what you are referring to in any sense similar to the marked vs unmarked
dichotomy?
 
 
Beatrice Reusch
University of California, Berkeley


From MFCEPRH at fs1.art.man.ac.uk  Mon Jan 19 14:07:07 1998
From: MFCEPRH at fs1.art.man.ac.uk (Richard Hogg)
Date: Mon, 19 Jan 1998 09:07:07 EST
Subject: complexity measures
Message-ID: <MON.19.JAN.1998.090707.EST.>

----------------------------Original message----------------------------
On 16 Jan 98 at 11:52, David Lightfoot wrote:
 
Original message----------------------------
>    Recent postings suggest that some people believe that languages
>    are all
> equally complex (although this is not entailed by Larry Trask's
> original question, giving rise to this discussion).
<snip>
> Alternatively, it might follow from some basic
> principles or some theory that languages must be equally complex. I
> know of no such empirical support nor of any theoretical
> underpinning for such an idea. What am I missing?
>    I should have thought that if there is a simplification in some
>    part
> of a system, there doesn't necessarily have to be compensating
> complexification elsewhere.
 
I'm sure that David's conclusion must be right, but in rejecting the
claim that all languages must be equally complex we have to be equally
careful not to embrace the claim that language change = language
evolution, i.e. that languages "improve" over time.
 
Cases where a simplification in one subsystem leads to
complexification in another subsystem are inherent in a system which
doesn't evolve but simply changes (except, perhaps, sub specie
aeternitatis).
 
Richard Hogg
 
********************************************************************************
 **********************
Richard M. Hogg            Tel: +44(0)161 275 3164
Department of English      Fax: +44(0)161 275 3256
  and American Studies     e-mail: r.m.hogg at man.ac.uk
University of Manchester   web:
 http://www.art.man.ac.uk/english/staff/rmh/home.htm
Oxford Road
Manchester M13 9PL
********************************************************************************
 **********************


From fertig at acsu.buffalo.edu  Mon Jan 19 14:06:45 1998
From: fertig at acsu.buffalo.edu (David L Fertig)
Date: Mon, 19 Jan 1998 09:06:45 EST
Subject: complexity measures
In-Reply-To: <199801161356.IAA17611@Ruby.deans.umd.edu>
Message-ID: <MON.19.JAN.1998.090645.EST.FERTIG@ACSU.BUFFALO.EDU>

----------------------------Original message----------------------------
 
I can imagine someone arguing along the following lines that under
conditions of normal transmission the overall complexity of every natural
language will tend to remain near the maximum that the human brain is
capable of handling.
 
1. Some linguistic changes are "random" from the point of view of
complexity, i.e. complexity/simplicity plays no role in their motivation.
(The side effects of changes that ARE motivated by considerations of
complexity/simplicity would be included in this category.)
 
2. Although individual changes of this type can of course either increase
or decrease complexity, on balance they are bound to increase complexity.
(This is the "entropy" that Robert Whiting discusses in his post. It's
easy to see that it has to be true, since in any system there will always
be more logically possible changes that increase complexity than that
decrease it.)
 
3. Changes that are motivated by considerations of complexity/simplicity,
on the other hand, are only activated to keep languages from exceeding the
complexity limit, i.e.  from becoming unlearnable or unusable. Their
effects still leave a language quite near maximum complexity. Together,
random and "natural" changes will thus tend to keep all normally
transmitted languages very close to the human limit for overall complexity
(and therefore roughly "equally complex"). And even after episodes of
non-normal transmission (such as pidginization), changes of the first type
will gradually restore a language to maximum complexity.
 
I'm not yet sure if I really buy this argument myself, but I'd be
interested in reactions, comments, references.  Obviously, if there's
anything to it, all the details remain to be addressed, in particular the
issue of global vs. local complexity and the significance of the kinds of
sociolinguistic factors that Miguel Carrasquer Vidal discusses in his
contributions to this discussion.
 
While I'm at it, let me throw out one more line of reasoning that arrives
at the same conclusion:
 
Languages (or speakers or learners) do not tolerate purposeless
complexity. It seems to me that this is just a generalized version of the
constraint on acquisition that Clark calls the "Principle of Contrast",
a.k.a. "no exact synonymy". When faced with any complexity, learners will
either figure out a function for it or eliminate it. This means that all
linguistic complexity serves some kind of purpose.
 
There is no logical limit to the amount of complexity that can potentially
be put to good (cognitive and/or communicative) use, and consequently
humans will tend to let their languages become as complex as their brains
can handle. In other words, up to a certain point, speakers/learners will
tend to deal with complexity by putting it to use, only when it goes
beyond that point will they deal with it by eliminating it.
 
 
Thanks in advance for any comments.
 
David Fertig


From bwald at HUMnet.UCLA.EDU  Tue Jan 20 00:27:57 1998
From: bwald at HUMnet.UCLA.EDU (bwald)
Date: Mon, 19 Jan 1998 19:27:57 EST
Subject: complexity measures
Message-ID: <MON.19.JAN.1998.192757.EST.BWALD@HUMNET.UCLA.EDU>

----------------------------Original message----------------------------
David Lightfoot wrote:
 
... it might follow from some basic principles or some
>theory that languages must be equally complex. I know of no such
>empirical support nor of any theoretical underpinning for such an idea.
>What am I missing?
 
My understanding is that linguists make the "equal complexity" assertion
because of the analytical equality of languages principle, fashioned into a
dogma for pedagogical purposes.  David surely knows that there is no agreed
upon measure and that the assertion arises from the dogmatic principle and
not from any measure of simplicity/complexity.  So, you might say that it
is precisely BECAUSE there is no simplicity matrix that this assertion is
made -- as a NULL HYPOTHESIS, if you will.  I do not know the precise
history of the principle in the present context, i.e., why it arose, but I
imagine that it was in the context of postponing (till more important
matters are settled) such ethnocentric claims/impressions as that French is
simpler than Georgian, asserted by an English speaker, or Georgian is
simpler than French or English, asserted by a Kabardian speaker, not to
mention that (written) Chinese is next to impossible to learn in middle age
(regardless of its grammar), asserted by countless Westerners, and so on
(where, obviously, a learnability principle relative to a speaker's first
language motivates judgments of the complexity of other languages).
 
Meanwhile, David is quite right to point out the mystification involved in
taking the principle too seriously.  It would mean that as some process
came to "simplify" some part of the grammar, either,
 
a.  AT THE SAME TIME, some other process would have to "complicate" some
other part of the
 
        grammar, to maintain "balance"
or,
 
b.  the grammar resulting from the simplification would be "unstable" until
compensating
        complexification took place.  (the ultimate in unprovability since
no language is globally
        stable)
 
Such ideas are not so much manifestly incorrect as they are obscure for
purposes of  serious interpretation and empirical testing.  One thing that
IS clear is that the dogma should be recognized as assigning the BURDEN OF
PROOF to being explicit and precise about how languages may differ with
respect to simplicity or complexity before claiming that they are indeed
different in this respect.
 
P.S.  One interesting manifestation of the dogma was the opinion that all
first languages are acquired in "to the same extent" in the same amount of
time by their speakers.  Presumably that meant that for any language, all
first learners will do something like give the impression of having
"mastered" the language at approximately the same age, presumably when
compared somehow with adult speakers of the same language.  Apart from the
fact that it is difficult to draw the line between not-yet-acquired and
basically-already-acquired, Dan Slobin long ago suggested that the basic
premiss is not literally true, and that child acquirers of highly inflected
languages like Russian take longer to learn such things than their
counterparts learning English.  That is an empirical finding.  But, in this
case, we know exactly what Slobin is talking about -- acquisition of
complex case paradigms, something which does not exist to be learned in
English.  I am not suggesting that we go from this to concluding that
Russian as a language is more "complex" than English, but only that its
case paradigms are more complex than English case paradigms -- something a
priori obvious by everyone's understanding of a complexity measure.  What
was not obvious before empirical observation was how long it would take
first language speakers to acquire the Russian cases and their various
forms (at least to ordinary adult mastery).
 
I suspect that there will never be a useful interpretation of such GLOBAL
claims as:
"Languages differ in over-all complexity", or, "Languages are basically the
same in over-all complexity."  But maybe this is just my opinion.-- Benji


From delancey at darkwing.uoregon.edu  Tue Jan 20 14:58:47 1998
From: delancey at darkwing.uoregon.edu (Scott DeLancey)
Date: Tue, 20 Jan 1998 09:58:47 EST
Subject: complexity measures
In-Reply-To: <Pine.GSO.3.96.980119000559.18603C-100000@xena.acsu.buffalo.edu>
Message-ID: <TUE.20.JAN.1998.095847.EST.DELANCEY@DARKWING.UOREGON.EDU>

----------------------------Original message----------------------------
On Mon, 19 Jan 1998, David L Fertig wrote:
 
> Languages (or speakers or learners) do not tolerate purposeless
> complexity. It seems to me that this is just a generalized version of the
> constraint on acquisition that Clark calls the "Principle of Contrast",
> a.k.a. "no exact synonymy". When faced with any complexity, learners will
> either figure out a function for it or eliminate it. This means that all
> linguistic complexity serves some kind of purpose.
 
This too seems like it can't be anything more than an article of faith.
Many Indo-European languages retain, not only grammatical gender,
but distinct, arbitrary declension classes within, or even to an
extent cross-cutting, gender classes.  These have been retained for
millenia.  They certainly add considerably to the overall complexity
of the language.  There's been a little work done (I remember a paper
somewhere by David Zubin) suggesting that there may be some mnemonic
function to grammatical gender, but even if so, what purpose could
possibly be served by the maintenance of significant bodies of irregular
declensions?  For that matter, what purpose is served by English
strong verbs?  True, they seem to be slowly falling out of the
language, but pretty slowly ... the language has tolerated them
for a long time, and doesn't seem to be in a hurry to clean them
up.
 
Scott DeLancey
Department of Linguistics
University of Oregon
Eugene, OR 97403, USA
 
delancey at darkwing.uoregon.edu
http://www.uoregon.edu/~delancey/prohp.html


From bwald at HUMnet.UCLA.EDU  Tue Jan 20 15:07:28 1998
From: bwald at HUMnet.UCLA.EDU (bwald)
Date: Tue, 20 Jan 1998 10:07:28 EST
Subject: complexity measures
Message-ID: <TUE.20.JAN.1998.100728.EST.BWALD@HUMNET.UCLA.EDU>

----------------------------Original message----------------------------
A few additional comments on the complexity issue.
 
First, if I had read the ensuing postings before sending my previous
message I might have more thoughtfully rephrased my use of the term
"dogma".  I wrote:
 
"My understanding is that linguists make the "equal complexity" assertion
because of the analytical equality of languages principle, fashioned into a
dogma for pedagogical purposes"
 
Various other respondents made points similar to mine about the motivation
for the principle, having to do with avoiding ethnocentrism and unfounded
prejudice about what constitutes "complexity".
(Actually, I was also saying that underlying the principle is a stance
claiming the need for the SAME linguistic resources to analyze any
language.)   The "dogma" part simply alludes to the internalization of the
principle so that it becomes dissociated from its purpose and is simply a
reflex reaction to any proposal to the contrary, no matter how reasoned.  I
would not accuse any linguist of doing this, but from my self-monitoring of
how I react to statements or questions about relative complexity  of
different languages by non-linguists, I recognize that the dogma alerts me
to the principle before I question the non-linguists further to see what
they mean.  I hasten to add that I do not dismiss everything a non-linguist
says about language, far from it.  So the dogma acts as a flag, not
necessarily a bad thing, once you get beyond it to the principle it is
intended to flag.
 
Next, I found the discussion stimulating, particularly Fertig's musings,
and Carrasquer Vidal's
citations from Malcolm Ross.  Fertig's leading suggestion is intriguing, to
the effect that languages generally work according to some principle of
entropy at near maximal brain capacity, and therefore they are all
equivalent in overall complexity.  If this is interpretable, and if that
were the case, then indeed as soon as something was simplified in the
grammar, something else would more or less instantaneously occupy the brain
"space" and maintain a constant overall complexity.  I find the idea
interesting, but wanting clarity for any kind of operationalization of the
brain mechanism or mental activity, so that it might become clear how it
could possibly be empirically tested.
 
Both Fertig and Carrasquer Vidal mentioned PIDGINS, as candidates for
"simpler" languages.  This is tricky for a number of reasons.  First,
pidgins are not first languages, and they are known to be parasitic on
first languages for their overall complexity.  For one thing, obviously the
phonological complexity of the pidgin as spoken by any particular speaker
is to some extent parasitic on other languages spoek by the speaker.  So,
it is not a forgone conclusion that pidgins can be examined as independent
languages for purposes of comparison for complexity with other kinds of
languages, most notably the languages of monolinguals or late
multilinguals.
 
Second, as far as comparison or status as independent languages, phenomena
which are called "pidgins" vary greatly in complexity and
conventionalization -- from the kind of on-the-spot makeshift forms of
communication, for which my first point applies most unproblematically, to
highly conventionalized and complex auxiliary languages, such as the
Neo-Melanesian which has a close creole counterpart in Tok Pisin, or some
quite stable varieties of West African pidgin English (which are
historically related to such creoles as Sierra Leonean Krio).  To the
extent that "learnability" can be taken as criterial of "complexity", there
have been provocative statements by some linguists, e.g., the creolist
Derek Bickerton, that creoles are maximally learnable because they are only
minimally arbitrary, with regard to grammar.  However, such a claim has
generated great controversy, to say the least.  And it has certainly not
been empirically demonstrated that they are more learnable than any other
languages.  Bickerton's arguments came from certain theoretical assumptions
he made.  They are no less suggestive than Fertig's proposal, mentioned
above, but they are no more well founded in terms of direct empirical
support.
 
Finally, in considering make-shift pidgins -- and I have been in situations
where I had to try to create one with the cooperation of interlocutors
(maybe most people have, if they have tried to communicate without a common
language), they are very stressful and hardly economical from a
production-perception point of view.  From this I detect an unclarity in
the concept "complexity".  Part of the stress has to do with trying to
communicate with a limited vocabulary, which leads to a lot of longer
circumlocutions according to some syntactic principles.  If we take
"complexity" to be a measure of abstract language knowledge, competence, or
whatever you want to call it, then the burden of using multi-word phrases
instead of semantically more complex (!?) single words is ignored, because
the same syntactic resources may be used in a fluent language as in a
make-shift pidgin.  Nevertheless, the constant appeal to multi-word phrases
in the makeshift pidgin instead of phonologically more compact single words
should count for something "complicated", shouldn't it?  After all, single
words are syntactically less complex than phrases.  How do we measure the
complexity of a language which has a limited vocabulary and constant appeal
to circumlocution involving more complex syntactic phrases against a
language which has a larger vocabulary and (for the sake of argument) the
same syntactic resources?  Should we say that it is "theoretically simpler"
but "more complex in practice"?  To be sure, no make-shift pidgin has
anywhere near the syntactic complexity of a first language; it's not even
close.  But no language belabors its limited resources more than a
makeshift pidgin.  From this at least one thing is clear, the amount of
effort involved is no measure of linguistic complexity.
 
P.S.  With respect to language acquisition, let me end with a reminder that
"language" is an abstraction to begin with.  Thus, who would deny that
child language is less complex than adult language, for the "same"
language.  OK.  Where do we stop?  Wouldn't a 40 year-old's language be
less complex than a 60 year-old's?  Why not?  Because we're not interested
in the additional complexity, say, additional vocabulary acquired in living
that long?  If we are, do we have to compare age-mates across languages to
compare language complexity (assuming we could do it, to begin with)?
Then, given individual and cultural differences, is the claim of equal
complexity one that asserts that for any speaker of a language of any age
we can find some speaker of another language of that age who has equivalent
complexity?  And, if linguists consider additional complexity, most likely
additional lexicon, trivial with regard to the problem, beyond a certain
age, as I suppose they would, then what precisely is non-trivial in
comparing languages "globally" for relative complexity?  This might have a
bearing on whether Fertig's entropy theory is as interesting as it first
appears.  I think it also has a bearing on how interesting, or not, the
notion of "global linguistic complexity" actually is.


From bwald at HUMnet.UCLA.EDU  Wed Jan 21 13:14:31 1998
From: bwald at HUMnet.UCLA.EDU (bwald)
Date: Wed, 21 Jan 1998 08:14:31 EST
Subject: complexity measures
Message-ID: <WED.21.JAN.1998.081431.EST.BWALD@HUMNET.UCLA.EDU>

----------------------------Original message----------------------------
Scott DeLancey asks:
 
For that matter, what purpose is served by English
strong verbs?  True, they seem to be slowly falling out of the
language, but pretty slowly ... the language has tolerated them
for a long time, and doesn't seem to be in a hurry to clean them
up.
 
All I can suggest about English strong verbs is that their past tenses are
certainly more salient than the mono-consonantal form of the -ed suffix,
which is not audible (or even pronounced) in some contexts.  You might also
accept, for similar reasons, that their articulation is simpler, to the
extent that they avoid the creation of consonant clusters like /vd/ in
'saved' (cf. how 'haved' > 'had', 'maked' > 'made', etc. avoids that).  In
that sense, 'brought' requires less articulatory movement and is
perceptually more different from "bring" than "bringed".  More puzzling is
the maintenance of 'double marking', as in 'tell' vs. 'told', 'toll' should
be sufficient, winning according to my criteria, over 'telled'.  A paper by
Labov & Sankoff in Language a number of years ago presented evidence that
well after childhood speakers were still learning to stablize the morpheme
boundary in such words as 'tol(#)d', 'lef(#)t', etc.  The older speakers
were most likely to treat the final segment as an -ed suffix and
omit/delete/fail-to-pronounce it at a rate comparable to the regular past
in the same phonetic environment, e.g., 'call#d', 'stuff#d'.  Younger
speakers (still adult) more often treated it at the much higher rate of
deletion found for single morphemes, as in 'col*d*', 'drif*t*', etc.  To me
this seems like evidence that vowel shifting has some advantages over the
flimsy final consonant for morpheme signalling purposes.
 
Now, wouldn't be interesting if strong verbs, or their remnants, outlive
the -ed past?  I wouldn't put it 'past' them.


From manaster at umich.edu  Wed Jan 21 17:08:50 1998
From: manaster at umich.edu (manaster at umich.edu)
Date: Wed, 21 Jan 1998 12:08:50 EST
Subject: complexity measures
In-Reply-To: <v02140b0db0e9a36fa717@[128.97.208.84]>
Message-ID: <WED.21.JAN.1998.120850.EST.>

----------------------------Original message----------------------------
I repeat: the arguments about all languages being more or less
equally complex arose within the context of early 20th cent.
(or slightly earlier) atempts by Boas, Sapir, Whorf, Jespersen,
and many others to refute earlier prejudices about "primitive"
languages--and made verygood sense indeed (a recent example
occurs in Dixon's Dyirbal grammar).  I am frankly puzled
by the discussion here which refuses to face up to that simple
fact and tries to read sthg more into the topic w/o tehre
being any semblance of atheoretical or empirical basis for
further disucssion, as I think David Lightfoot pointed out.
Is this a symptom of the long-remarked-upon reluctance to
acknowledge that science has a history?
 
AMR


From Alison.Munro at ed.ac.uk  Wed Jan 21 16:21:16 1998
From: Alison.Munro at ed.ac.uk (Alison Munro)
Date: Wed, 21 Jan 1998 11:21:16 EST
Subject: Historical Linguistics
Message-ID: <WED.21.JAN.1998.112116.EST.ALISON.MUNRO@ED.AC.UK>

----------------------------Original message----------------------------
Forthcoming books for linguists...
 
Edinburgh University Press will be publishing two new historical
linguistics titles this Spring:
 
HISTORICAL LINGUISTICS: AN INTRODUCTION by Professor Lyle Campbell of the
University of Canterbury, New Zealand   May 1998  0 7486 0775 7  ?14.95
(UK)
A hands-on introduction taking students through the topics, and giving them
abundant examples and exercises to practise the principles and procedures
described.  It covers all the essential topics: grammaticalisation,
sociolinguistic contributions to linguistic change, distant genetic
relationships, areal linguistics and linguistic prehistory.
 
--- "The textbook of choice for courses in historical linguistics"
Professor William J Poser, Stanford University
--- "Absolutely the best textbook in historical linguistics"
Professor Theo Vennemann, University of Munich
 
EVERYDAY ENGLISH: 1500-1700: A READER edited by Bridget Cusack, formerly of
Edinburgh University   March 1998   0 7486 0776 5  ?14.95 (UK)
An anthology of contemporary early modern language in more than 60
non-literary texts, with introductions, notes and glossaries.
 
Edinburgh University Press is offering a limited number of inspection
copies to academics teaching courses in the above areas who feel that one
or both of these books would be appropriate for their classes.  Please
contact Alison Munro at Edinburgh University Press with details of your
class name, size and level, your institutional postal address, and the
reason why you think your requested book is suitable.  Email:
Alison.Munro at ed.ac.uk
 
 
Alison Munro
 
Marketing Director, Edinburgh University Press,
22 George Square, Edinburgh EH8 9LF   http://www.eup.ed.ac.uk/


From mc at ai.uga.edu  Thu Jan 22 02:03:20 1998
From: mc at ai.uga.edu (Michael A. Covington)
Date: Wed, 21 Jan 1998 21:03:20 EST
Subject: Computers and mathematics in HL
Message-ID: <WED.21.JAN.1998.210320.EST.MC@AI.UGA.EDU>

----------------------------Original message----------------------------
I'm just wondering... would there be much interest in a special session
on "Computers and Mathematics in Historical Linguistics" at next
year's LSA (Jan. 1999, Los Angeles)?
 
Michael A. Covington                       http://www.ai.uga.edu/~mc
Chairman, Computer Security Team, and
Associate Director, Artificial Intelligence Center
The University of Georgia, Athens, GA 30602-7415 U.S.A.


From manaster at umich.edu  Fri Jan 23 22:35:43 1998
From: manaster at umich.edu (manaster at umich.edu)
Date: Fri, 23 Jan 1998 17:35:43 EST
Subject: No subject
In-Reply-To: <Pine.SOL.3.95.980123123351.3844B-100000@qbert.rs.itd.umich.edu>
Message-ID: <FRI.23.JAN.1998.173543.EST.>

----------------------------Original message----------------------------
There is another issue I would like to broach. As I argued in
my paper on Pakawan and Coahuiltecan in Anthro Lx last year
(and in my survey of Nsotratic in Studies in Lg in 1993),
in recent (and not-so-recent) debates about lg classification,
people too often reason as though refuting a part of somebody
else's argument FOR a proposed classification constitues an
argyument AGAINST that clasifcation.  This seems plain bad
logic to me, and yet it is seems to be prevalent.  Any
comments?  AMR


From delancey at darkwing.uoregon.edu  Fri Jan 23 22:39:34 1998
From: delancey at darkwing.uoregon.edu (Scott DeLancey)
Date: Fri, 23 Jan 1998 17:39:34 EST
Subject: your mail
In-Reply-To: <Pine.SOL.3.95.980123125414.9924B-100000@qbert.rs.itd.umich.edu>
Message-ID: <FRI.23.JAN.1998.173934.EST.DELANCEY@DARKWING.UOREGON.EDU>

----------------------------Original message----------------------------
On Fri, 23 Jan 1998 manaster at umich.edu wrote:
 
> There is another issue I would like to broach. As I argued in
> my paper on Pakawan and Coahuiltecan in Anthro Lx last year
> (and in my survey of Nsotratic in Studies in Lg in 1993),
> in recent (and not-so-recent) debates about lg classification,
> people too often reason as though refuting a part of somebody
> else's argument FOR a proposed classification constitues an
> argyument AGAINST that clasifcation.  This seems plain bad
> logic to me, and yet it is seems to be prevalent.  Any
> comments?  AMR
 
In principle, it isn't necessarily bad logic.  Lacking any evidence
for a relationship between two languages, the null hypothesis must
be that there is no relationship.  (Or, more properly, no relationship
at whatever level the discussion is concerned with--I might sloppily
express disbelief in the relationship of Takelma and Kalapuya, and
mean not that I don't believe that they are both Penutian, but only
that they have no closer relationship than that).  Therefore, if
only a small amount of evidence has been presented for the relationship
of X and Y, an argument which offers an alternative explanation for
that evidence is legitimately an argument against the claimed
relationship.
     Where this does get illogical is where refutation of some *part*
of the evidence which has been presented for a relationship is taken
as an argument against it.  This of course makes no sense at all.
Unfortunately I have to agree with you that this is pretty common.
 
Scott DeLancey
Department of Linguistics
University of Oregon
Eugene, OR 97403, USA
 
delancey at darkwing.uoregon.edu
http://www.uoregon.edu/~delancey/prohp.html


From larryt at cogs.susx.ac.uk  Sat Jan 24 17:45:20 1998
From: larryt at cogs.susx.ac.uk (Larry Trask)
Date: Sat, 24 Jan 1998 12:45:20 EST
Subject: Alexis on classification
In-Reply-To: <Pine.SOL.3.95.980123125414.9924B-100000@qbert.rs.itd.umich.edu>
 from "manaster@umich.edu" at Jan 23, 98 05:35:43 pm
Message-ID: <SAT.24.JAN.1998.124520.EST.LARRYT@COGS.SUSX.AC.UK>

----------------------------Original message----------------------------
Alexis M R writes:
 
> There is another issue I would like to broach. As I argued in
> my paper on Pakawan and Coahuiltecan in Anthro Lx last year
> (and in my survey of Nsotratic in Studies in Lg in 1993),
> in recent (and not-so-recent) debates about lg classification,
> people too often reason as though refuting a part of somebody
> else's argument FOR a proposed classification constitues an
> argyument AGAINST that clasifcation.  This seems plain bad
> logic to me, and yet it is seems to be prevalent.  Any
> comments?  AMR
 
I agree that such reasoning is in principle illogical.  In practice,
though, the outcome is sometimes different.
 
In my own various critiques of comparisons of Basque with language L,
I have invariably concluded that the overwhelming majority of the
individual comparisons adduced fail to stand up, even on the Basque
side alone, before anyone has considered those comparisons from the
other side, and I have further concluded that what is left
undemolished is insufficient to constitute interesting evidence: it's
nothing more than a handful of arbitrary and miscellaneous
resemblances.
 
I therefore see no great obstacle to reasoning as follows: "Look --
you've spent years scouring these languages for evidence of
relatedness, and all you've been able to come up with is this
miserable collection of junk.  Therefore, there are good grounds for
concluding that the languages are not discoverably related at all."
 
So I attach importance to the volume of evidence.  Destroying just one
or two pieces of a more sizeable case does not demolish the case, but
destroying almost all of the case presented leads me to conclude that
the case itself is not there.
 
Larry Trask
COGS
University of Sussex
Brighton BN1 9QH
UK
 
larryt at cogs.susx.ac.uk


From manaster at umich.edu  Sat Jan 24 17:46:25 1998
From: manaster at umich.edu (manaster at umich.edu)
Date: Sat, 24 Jan 1998 12:46:25 EST
Subject: Alexis on classification
In-Reply-To: <m0xw4JN-0006MXC@rsunx.crn.cogs.susx.ac.uk>
Message-ID: <SAT.24.JAN.1998.124625.EST.>

----------------------------Original message----------------------------
 
Larry Trask writes:
 
>
> I therefore see no great obstacle to reasoning as follows: "Look --
> you've spent years scouring these languages for evidence of
> relatedness, and all you've been able to come up with is this
> miserable collection of junk.  Therefore, there are good grounds for
> concluding that the languages are not discoverably related at all."
>
 
There is in my view a subtle fallacy here (and also I think in
Scott Delancey's reply to my posting).  It is true that if someone's
arguments/data are exploded, they have no case left for their theory.
But this does not justify concluding that there IS no "discoverable"
case for the same theory.  For example, I showed that Sapir did not
so much have bad arguments as had no arguments at all for relating the
"Coahuiltecan" languages to each other, only for relating one of them
(Tonkawa) to Hokan (I am oversimplifying slightly).  But there is
lots of other evidence for relating some of tehse languages (esp.
what I call the Pakawan group) to each other. I dont know how many
of you have read my work (in Anthro Lg) or agree with it, but that
is not germane to the point at issue here, namely, that even the
complete failure of someone's arguments for a given lg family
only means that THAT case has failed.  I would of course argue that
Altaic is a good example of this too.  The early work on relating
these languages (e.g., Halevy) was often a joke, but I would argue
that the more recent work establishes their relatedness.  An even
clearer example of the same is Uto-Aztecan, which is clearer simply
because no one questions it now (not even the people like Doerfer or
Campbell who deny Altaic or Pakawan).
 
AMR


From larryt at cogs.susx.ac.uk  Sun Jan 25 22:04:51 1998
From: larryt at cogs.susx.ac.uk (Larry Trask)
Date: Sun, 25 Jan 1998 17:04:51 EST
Subject: Alexis on classification
In-Reply-To: <Pine.SOL.3.95.980124085800.10219C-100000@pacman.rs.itd.umich.edu> from "manaster@umich.edu" at Jan 24,
 98 09:07:50 am
Message-ID: <SUN.25.JAN.1998.170451.EST.LARRYT@COGS.SUSX.AC.UK>

----------------------------Original message----------------------------
Alexis writes:
 
> There is in my view a subtle fallacy here (and also I think in
> Scott Delancey's reply to my posting).  It is true that if someone's
> arguments/data are exploded, they have no case left for their theory.
> But this does not justify concluding that there IS no "discoverable"
> case for the same theory.
 
And I have not said so.  I have only said that prolonged failure to
make a persuasive case leads to a more than reasonable conclusion that
there is probably no case to be made.
 
> For example, I showed that Sapir did not
> so much have bad arguments as had no arguments at all for relating the
> "Coahuiltecan" languages to each other, only for relating one of them
> (Tonkawa) to Hokan (I am oversimplifying slightly).  But there is
> lots of other evidence for relating some of tehse languages (esp.
> what I call the Pakawan group) to each other.
 
I think one might reasonably ask why Sapir thought it was worth while
putting forward a proposal on the basis of no evudence at all.  Even
if someone else later presents evidence for the same proposal, I think
Sapir can be credited with no more than a lucky guess, or at best
perhaps a sixth sense, if you believe in such stuff.  Credit should go
only to the person who comes up with real evidence.
 
I call this the "Democritus fallacy".  A number of ancient Greek
philosophers speculated wildly about the nature of the world, all of
them on the basis of no evidence at all.  Most of their speculations
are dismissed today as empty and worthless.  However, quite by chance,
Democritus's speculation turned out to look something like the atomic
theory settled on by chemists over 2000 years later, on the basis of
evidence.  Consequently, chemistry textbooks often give Democritus
credit for being the founder of the atomic theory.  But this is
absurd: Democritus had no more basis for his speculations than any
other Greek; he just got lucky.
 
The first person to predict that the surface of Venus would prove to
be exceedingly hot was the crackpot Velikovsky.  But he made his guess
on the basis of mad ideas of his own and no evidence -- hence he does
*not* get credit for a prediction that has proved to be true.
 
> I dont know how many
> of you have read my work (in Anthro Lg) or agree with it, but that
> is not germane to the point at issue here, namely, that even the
> complete failure of someone's arguments for a given lg family
> only means that THAT case has failed.
 
Agreed, but surely there must come a time...
 
[snip]
 
Larry Trask
COGS
University of Sussex
Brighton BN1 9QH
UK
 
larryt at cogs.susx.ac.uk


From manaster at umich.edu  Mon Jan 26 22:16:34 1998
From: manaster at umich.edu (manaster at umich.edu)
Date: Mon, 26 Jan 1998 17:16:34 EST
Subject: Alexis on classification
In-Reply-To: <m0xwTTo-0006LJC@rsunx.crn.cogs.susx.ac.uk>
Message-ID: <MON.26.JAN.1998.171634.EST.>

----------------------------Original message----------------------------
Larry Trask writes:
 
I call this the "Democritus fallacy".  A number of ancient Greek
philosophers speculated wildly about the nature of the world, all of
them on the basis of no evidence at all.  Most of their speculations
are dismissed today as empty and worthless.  However, quite by chance,
Democritus's speculation turned out to look something like the atomic
theory settled on by chemists over 2000 years later, on the basis of
evidence.  Consequently, chemistry textbooks often give Democritus
credit for being the founder of the atomic theory.  But this is
absurd: Democritus had no more basis for his speculations than any
other Greek; he just got lucky.
 
**End of quote**
 
But I am not trying to argue that Sapir for example should
get credit for recognizing Pakawan despite the lack of data
in his published work on this.  In fact, I never said anything
about credit.  I am concerned about the fact that a lot of
linguists, yourself among them, have been publishing all kinds
of statements about the validity or otherwise of work in
linguistic classification while staunchly avoiding getting
their hands dirty at all or only doing so in cases where one
is shooting ducks in a barrel.  For example, there has been
a lot of often quite uninformed discussion of Nostratic
but no real attempts to evaluate the actual claims of the
theory (the best-informed work I know, by Brent Vine, does
not even venture outside IE and hence by definition does not
addres any of the comparisons among lg families which are the
whole point of the theory!).  I have seen no substantive
reactions to Vovin's Ainu-Austroasiatic.  Swadesh and Hamp's
comparison of Eskimo-Aluet and Chukchee-Kamchatkan has been
sitting around for decades with almost no response.  My
recent defense of Nadene has elicited no response, and
my work on Pakawan has only resulted in a predictable
response from Campbell (a reply to which will be out soon),
but neither he nor anyone else has looked at the actual
arguments for Pakawan.  Recent attacks on Altaic by
Nichols are based on third-hands sources.  ALmost all
critics of Greenberg's Amerind, myself excepted, have had
nothing more to say that that he got some forms wrong,
but do not address the comparisons he makes at all.
And so on.
 
Of course, I find the uncritical acceptance of untested
or even provably wrong proposals on the other side (whether
in the case of Nostratic, Dene-Caucasian, Amerind, or
any other such proposal) equally distressing.
 
AMR


From alderson at netcom.com  Tue Jan 27 19:04:28 1998
From: alderson at netcom.com (Richard M. Alderson III)
Date: Tue, 27 Jan 1998 14:04:28 EST
Subject: Alexis on classification (again) (fwd)
In-Reply-To: <Pine.SOL.3.95.980127084653.750E-100000@joust.rs.itd.umich.edu>
 (manaster@umich.edu)
Message-ID: <TUE.27.JAN.1998.140428.EST.>

----------------------------Original message----------------------------
Alexis wrote:
 
>Where then are the competent critics of controversial classificatory proposals
>in linguistics?
 
Unborn as yet?  At least, not yet studying linguistics?
 
Nostratic (to take the example of one of the largest families proposed) has had
its Jones (Pedersen) and its Bopps & Rasks (Cuny, Illich-Svitych, et al.), but
it hasn't seen its Schleicher yet, much less its Neogrammarians.
 
It takes a very long time to become a competent Indo-Europeanist, and there at
least the student has the advantage of 200 years of publications and pedagogy.
In order to further Nostratic studies, we need to do with Indo-European what
the Indo-Europeanists have always done with Latin and Greek:  Assume that the
student simply knows those languages, needs to learn Sanskrit and half a dozen
others, and can be taught the rudiments of Indo-European at the same time.
 
Thus, to produce competent Nostraticists, we need to assume a background in
Indo-European, and teach Uralic, Altaic, Kartvelian, and so on, while teaching
the rudiments of Nostratic.  But to do that, we have to have decent historical
work in all those fields that can simply be treated as correct, much as we
treat Greek and Latin and Hittite and Gothic and Vedic and all the others when
we teach (or learn) Indo-European--and we certainly aren't there, yet.  (We
also know, of course, that new discoveries about any language may change every-
thing we thought we knew.)
 
That covers only one "superfamily", of course, and that won't produce those who
might be competent to judge any other "controversial classificatory proposals",
if those who are to be considered competent must be expert in more than one of
the sub-families in the proposal.
 
Alternatively, we might expect that competent critics could be found among
those trained as historical linguists no matter what language area they have
chosen to concentrate their work in, and ask that these judge proposals by the
rigour of the methodology applied to produce them.  Why is this unsatisfactory?
 
                                                                Rich Alderson


From larryt at cogs.susx.ac.uk  Tue Jan 27 14:50:59 1998
From: larryt at cogs.susx.ac.uk (Larry Trask)
Date: Tue, 27 Jan 1998 09:50:59 EST
Subject: Alexis on classification (again)
In-Reply-To: <Pine.SOL.3.95.980127084459.750D-100000@joust.rs.itd.umich.edu>
 from "manaster@umich.edu" at Jan 27, 98 08:46:35 am
Message-ID: <TUE.27.JAN.1998.095059.EST.LARRYT@COGS.SUSX.AC.UK>

----------------------------Original message----------------------------
Alexis writes:
 
> Where then are the competent critics of controversial
> classificatory proposals in linguistics?
 
A very reasonable question, and one I'm not sure has a good answer.
 
You mentioned someone who had criticized Nostratic purely from the IE
point of view.  Well, I suppose he just didn't feel competent to
evaluate the data from the other five families.  So, what can anybody
do except to comment on the data he feels knowledgeable about?  If
specialists from all six families are prepared to comment on the
Nostratic use of data from their specialist families, then that's a
start.  Maybe then someone will feel himself in a position to try to
tie up the various comments.
 
Look: comparative linguistics is hard, and, the vaster the proposal on
the table, the harder it gets either to formulate a decent proposal or
to evaluate the result.  That's just the way it is, I think.  I don't
think these difficulties arise from the shortcomings of linguists; I
think they derive from the nearly intractable nature of the problem.
 
Larry Trask
COGS
University of Sussex
Brighton BN1 9QH
UK
 
larryt at cogs.susx.ac.uk


From manaster at umich.edu  Tue Jan 27 14:50:44 1998
From: manaster at umich.edu (manaster at umich.edu)
Date: Tue, 27 Jan 1998 09:50:44 EST
Subject: Alexis on classification (again)
In-Reply-To: <m0xx80q-0006MdC@rsunx.crn.cogs.susx.ac.uk>
Message-ID: <TUE.27.JAN.1998.095044.EST.>

----------------------------Original message----------------------------
Then just one last question, directed to anyone who cares to
respond, not necessarily Larry, although I would like to hear
from him of course:
 
Where then are the competent critics of controversial
classificatory proposals in linguistics?
 
Alexis MR
 
On Tue, 27 Jan 1998, Larry Trask wrote:
 
> Alexis writes:
>
> > But I am not trying to argue that Sapir for example should
> > get credit for recognizing Pakawan despite the lack of data
> > in his published work on this.  In fact, I never said anything
> > about credit.
>
> OK; my apologies, then, but you did give me the impression that you
> thought Sapir had done something interesting.
>
> > I am concerned about the fact that a lot of linguists, yourself
> > among them, have been publishing all kinds of statements about the
> > validity or otherwise of work in linguistic classification while
> > staunchly avoiding getting their hands dirty at all or only doing so
> > in cases where one is shooting ducks in a barrel.
>
> I have indeed blasted a few encupated anatids, but it's not my fault
> that my specialist language seems to attract nothing else.
>
> As for getting my hands dirty, I have already explained that I do not
> choose to undertake comparative work, because I believe I am better
> qualified to do other things.  We can't *all* be comparativists,
> Alexis -- somebody has to do the work ;-)
>
> [on the absence of informed evaluations of Nostratic and other
> proposals]
>
> It may just be that few of us feel competent to undertake a
> magisterial scrutiny of proposals embracing vast numbers of languages
> and families.
>
> > Of course, I find the uncritical acceptance of untested
> > or even provably wrong proposals on the other side (whether
> > in the case of Nostratic, Dene-Caucasian, Amerind, or
> > any other such proposal) equally distressing.
>
> Good.
>
> Larry Trask
> COGS
> University of Sussex
> Brighton BN1 9QH
> UK
>
> larryt at cogs.susx.ac.uk
>
>
>
>
>


From larryt at cogs.susx.ac.uk  Tue Jan 27 14:44:24 1998
From: larryt at cogs.susx.ac.uk (Larry Trask)
Date: Tue, 27 Jan 1998 09:44:24 EST
Subject: Alexis on classification (again)
In-Reply-To: <Pine.SOL.3.95.980126115328.18961C-100000@joust.rs.itd.umich.edu>
 from "manaster@umich.edu" at Jan 26, 98 05:16:34 pm
Message-ID: <TUE.27.JAN.1998.094424.EST.LARRYT@COGS.SUSX.AC.UK>

----------------------------Original message----------------------------
Alexis writes:
 
> But I am not trying to argue that Sapir for example should
> get credit for recognizing Pakawan despite the lack of data
> in his published work on this.  In fact, I never said anything
> about credit.
 
OK; my apologies, then, but you did give me the impression that you
thought Sapir had done something interesting.
 
> I am concerned about the fact that a lot of linguists, yourself
> among them, have been publishing all kinds of statements about the
> validity or otherwise of work in linguistic classification while
> staunchly avoiding getting their hands dirty at all or only doing so
> in cases where one is shooting ducks in a barrel.
 
I have indeed blasted a few encupated anatids, but it's not my fault
that my specialist language seems to attract nothing else.
 
As for getting my hands dirty, I have already explained that I do not
choose to undertake comparative work, because I believe I am better
qualified to do other things.  We can't *all* be comparativists,
Alexis -- somebody has to do the work ;-)
 
[on the absence of informed evaluations of Nostratic and other
proposals]
 
It may just be that few of us feel competent to undertake a
magisterial scrutiny of proposals embracing vast numbers of languages
and families.
 
> Of course, I find the uncritical acceptance of untested
> or even provably wrong proposals on the other side (whether
> in the case of Nostratic, Dene-Caucasian, Amerind, or
> any other such proposal) equally distressing.
 
Good.
 
Larry Trask
COGS
University of Sussex
Brighton BN1 9QH
UK
 
larryt at cogs.susx.ac.uk


From johanna at uclink.berkeley.edu  Wed Jan 28 03:02:59 1998
From: johanna at uclink.berkeley.edu (Johanna Nichols)
Date: Tue, 27 Jan 1998 22:02:59 EST
Subject: Alexis on classification
In-Reply-To: <Pine.SOL.3.95.980126115328.18961C-100000@joust.rs.itd.umich.edu>
Message-ID: <TUE.27.JAN.1998.220259.EST.JOHANNA@UCLINK.BERKELEY.EDU>

----------------------------Original message----------------------------
>----------------------------Original message----------------------------
Alexis writes:
 
 
>  Recent attacks on Altaic by
>Nichols are based on third-hands sources.
 
Could you supply some references?  I try to keep abreast of this
literature, but I have seen nothing on Altaic by Nichols that I would
describe as either recent or an attack.
 
ALmost all
>critics of Greenberg's Amerind, myself excepted, have had
>nothing more to say that that he got some forms wrong,
>but do not address the comparisons he makes at all.
 
This isn't an accurate characterization of the reviews and critiques I've
read, and I'm sure I've read most of them.  They do criticize the quality
of the data; they also discuss explicitly whether errors in the data
vitiate the comparisons; they  discuss the number and the quality of the
putative cognate sets and the density of language representation within
them; they discuss method more generally; and they ask whether, all in all,
Greenberg's material makes a case for Amerind or its subgroups.  I would
call this addressing the comparisons.
 
 
Johanna Nichols


From delancey at darkwing.uoregon.edu  Wed Jan 28 20:34:58 1998
From: delancey at darkwing.uoregon.edu (Scott DeLancey)
Date: Wed, 28 Jan 1998 15:34:58 EST
Subject: your mail
In-Reply-To: <Pine.SOL.3.95.980127211111.5112D-100000@qbert.rs.itd.umich.edu>
Message-ID: <WED.28.JAN.1998.153458.EST.DELANCEY@DARKWING.UOREGON.EDU>

----------------------------Original message----------------------------
On Tue, 27 Jan 1998 manaster at umich.edu wrote:
 
> Even if ALL the evidence presented by someone for
> theory X is refuted (or if no evidence was ever
> presented in the first place, as happened with Sapir
> several times), that does not mean that X is false
> or incapable of being shown true or even unworthy of
> further effort.
 
I wouldn't dream of disagreeing with this.  I think we're
actually mixing up two separate debates here.
 
I do know the mind-set that you are objecting too; I have a
little experience with this myself.  (Many folks of that temperament
don't like Penutian any bettre than they do Altaic).  And indeed,
the attitude that, if someone can demolish the evidence that's
been put forward for a relationship, then that hypothesis of
relationship can be considered refuted from then on, is irrational
and absurd.
 
But if we ignore those people--i.e. if we argue simply the
logical merits of the issue, rather than the actual ideological
climate of contemporary historical linguistics--then I stand
by my earlier posting.  Myself, I'm a confirmed monogeneticist;
on simple Darwinian grounds monogenesis has to be correct.  So
I assume that all languages are related at some level.  The issue
facing any comparativist working on a relationship that is not yet
established is, are these languages demonstrably related at the
level that I am working at.  As I said, the default hypothesis
has to be no--we cannot claim relationship without some evidence.
Therefore, if all the putative evidence for a relationship has
been refuted, we are back to square one--no evidence, therefore
no assumption of relationship.  (As we all know, there's a lot
of room for disagreement about what kinds and amounts of evidence
are necessary to license a working hypothesis about relationship,
but that's a methodological issue, not a logical one).  No assumption
of relationship means just that--it most definitely does not mean
that investigating the possibility further is an irresponsible
waste of time, as some of our colleagues do seem to assume.
 
Scott DeLancey
Department of Linguistics
University of Oregon
Eugene, OR 97403, USA
 
delancey at darkwing.uoregon.edu
http://www.uoregon.edu/~delancey/prohp.html


From Roger.Wright at liverpool.ac.uk  Wed Jan 28 14:03:54 1998
From: Roger.Wright at liverpool.ac.uk (Roger Wright)
Date: Wed, 28 Jan 1998 09:03:54 EST
Subject: Alexis on classification (again) (fwd)
In-Reply-To: <199801271829.KAA18879@netcom16.netcom.com>
Message-ID: <WED.28.JAN.1998.090354.EST.ROGER.WRIGHT@LIVERPOOL.AC.UK>

----------------------------Original message----------------------------
>
>Alternatively, we might expect that competent critics could be found among
>those trained as historical linguists no matter what language area they have
>chosen to concentrate their work in, and ask that these judge proposals by the
>rigour of the methodology applied to produce them.  Why is this unsatisfactory?
>
 
 
Because whether a hypothesis is right or not has no connection with the
methodological rigour used by its advocates. For example, Ptolemaic
astronomy was very methodologically rigorous, and at first, at least,
Copernicus's ideas were just bright ideas, not worked out with any
precision. This is the converse of what Larry Trask called the
Democritus fallacy; lack of rigour does not imply that a theory is
wrong, merely that it hasn't yet been shown to be right.
        I find rigour off-putting, myself. What is all this geometry
trying to hide?
                                                        Roger Wright


From manaster at umich.edu  Wed Jan 28 14:03:11 1998
From: manaster at umich.edu (manaster at umich.edu)
Date: Wed, 28 Jan 1998 09:03:11 EST
Subject: your mail
In-Reply-To: <199801280305.WAA29457@stayhungry.rs.itd.umich.edu>
Message-ID: <WED.28.JAN.1998.090311.EST.>

----------------------------Original message----------------------------
 
> >I would like to hear from others on this point, but
> >it seems to me that we should add this to Scott's
> >remarks:
> >
> >Even if ALL the evidence presented by someone for
> >theory X is refuted (or if no evidence was ever
> >presented in the first place, as happened with Sapir
> >several times), that does not mean that X is false
> >or incapable of being shown true or even unworthy of
> >further effort.  There is of course no way of knowing
> >whether the effort will be rewarded or not, but
> >SOMEBODY has to keep trying to classify the languages
> >of the world.  I think that perhaps the current
> >crisis in this field has to do in part with false
> >perceptions of the history of the field.  If more people
> >knew about the story of how various classifications
> >were worked out in the course of this century (e.g.,
> >Vietnamese as Mon-Khmer, Pama-Nyungan as Australian,
> >Algic, Anatolian as Indo-European, Eskimo-Aleut, and so on),
> >perhaps there would be more appreciation of just how
> >much of an evolving, progressing field this is and how
> >much realistic work remains to be done--instead of the
> >caricature which we are daily fed by the extremists
> >on both sides.
> >
> >AMR
> >
> >On Fri, 23 Jan 1998, Scott DeLancey wrote:
> >>
> >> ....In principle, it isn't necessarily bad logic.  Lacking any
> >> evidence
> >> for a relationship between two languages, the null hypothesis must
> >> be that there is no relationship....


From manaster at umich.edu  Wed Jan 28 14:02:20 1998
From: manaster at umich.edu (manaster at umich.edu)
Date: Wed, 28 Jan 1998 09:02:20 EST
Subject: Alexis on classification
In-Reply-To: <v03020901b0f41f47b45a@[128.32.100.155]>
Message-ID: <WED.28.JAN.1998.090220.EST.>

----------------------------Original message----------------------------
Johanna Nichols professes not to know what I am talking about
when I say "Recent attacks on Altaic by Nichols are based on third-hands
sources" and asks for references.  Here they are:
 
 Nichols (1992: 4), in an otherwise well-informed
book,  claims that, after eliminating purely typological similarities
between Turkic, Mongolian, and Tungusic,
 
the evidence [sc. for Altaic] was reduced to the pronominal root
resemblances and a set of putative cognates. When the cognates proved not
to be valid, Altaic was abandoned, and the received view now is that
Turkic, Mongolian, and Tungusic are unrelated (see Unger 1990 [=our Unger
1990b]).
 
Nichols, J. 1992. Linguistic Diversity in Space and Time.
Chicago--London: University of Chicago Press.
 
Unger, J. M. 1990b.  Summary report of the Altaic panel.   In Baldi, P.
(ed.), Linguistic change and reconstruction methodology.Berlin--New York:
Mouton de   Gruyter. 479-482.
 
This is from a forthcoming paper by R-S Georg, R. Michalove, P. Sidwell,
and myself, to appear in JL, and dealing with the history and current
state of Altaic studies, in which we discuss in some detail several
recent misrepresentations of the same, by Nichols and others.
 
Of course, one can quibble as to whether 1992 is 'recent'.
 
Alexis
 
PS. As for our conflicting opinions of the mass of work
dealing with Greenberg's Amerind hypothesis, I think it
is more appropriate if I leave it to others to decide whether
my critique is or is not of an entirely different order
(as I believe) than those of most other authors.


From rjanda at midway.uchicago.edu  Thu Jan 29 13:45:52 1998
From: rjanda at midway.uchicago.edu (Richard Janda)
Date: Thu, 29 Jan 1998 08:45:52 EST
Subject: Monogenesis and "simple Darwinian grounds"
Message-ID: <THU.29.JAN.1998.084552.EST.RJANDA@MIDWAY.UCHICAGO.EDU>

----------------------------Original message----------------------------
 
  While agreeing with most of Scott DeLancey's last posting, I believe that
it is a demonstrable fallacy to claim that, "on simple Darwinian grounds",
"monogenesis has to be correct".  I'm no geneticist, but it seems clear to
me that there's a major confusion here between biological prerequisites for
language and the initial set of arbitrary sound/meaning associations charac-
terizing the ultimate ancestor of some language family.  That is, even as-
suming that, due to selection, some mutation which provided the prerequisites
for language came to characterize all living humans at some point, it does
not follow that there was only one occasion on which sound/meaning associa-
tions were arrived at and then passed on to succeeding generations.  Rather,
a language-facilitating mutation could have been selected for but then not
immediately acted on, as it were (and this does not strike me as unlikely).
It hardly strains credulity to imagine that two groups of early humans living
widely separated from each other could independently have stumbled onto the
use of sound/meaning associations for communication in specifically linguistic
ways.  Even if their syntaxes and morphologies were extremely similar or
identical, we again must recognize the possibility that their lexicons (lexi-
ca?) were virtually non-overlapping.  A number of writers have suggested that
pre-existing ritualized behaviors (particularly vocalizations) could have
provided the impetus or at least the model for the development of a lexicon,
and these behaviors might well have been pan-human, but the factor of arbi-
trariness leads us to the conclusion that their specific extensions to new
instances of sound/meaning association need not have been universal.  In that
case, though, we would not have monogenesis of the actual substance of lan-
guages--at least not necessarily.  After all, arbitrariness is a historical
linguist's best friend in arguments intending to sh?w that extensive sound/
meaning correspondences (of certain types) between languages require the as-
sumption of a shared ancestral linguistic stage, rather than borrowing or ac-
cidental convergence.  So doesn't one basic tool of our trade that is con-
stantly used in establishing convergences projected backward from the present
force us to allow for the possibility that primeval arbitrariness permitted
and perhaps even favored polygenesis (as long as we distinguish the monogene-
sis of ability from possibly multiple, independent exercises of that ability)?
     Richard Janda


From sally at isp.pitt.edu  Thu Jan 29 13:53:23 1998
From: sally at isp.pitt.edu (Sarah G. Thomason)
Date: Thu, 29 Jan 1998 08:53:23 EST
Subject: Alexis on classification
In-Reply-To: Your message of "Wed, 28 Jan 1998 09:02:20 EST."
 <Pine.SOL.3.95.980127222724.17496D-100000@qbert.rs.itd.umich.edu>
Message-ID: <THU.29.JAN.1998.085323.EST.SALLY@ISP.PITT.EDU>

----------------------------Original message----------------------------
 
 
Let's see: Alexis Manaster Ramer accuses Johanna Nichols of making
"attacks on Altaic...based on third-hands [sic] sources".  Nichols
asks for references.  Manaster Ramer gives one reference to a work
by Nichols (therefore at most one attack).  As evidence for his
assertion that she attacks Altaic, he quotes her negative
assessment of the evidence in favor of the Altaic hypothesis.  Perhaps
Manaster Ramer views any argument against a proposal that he favors as
an attack on the proposal; I hope and believe that most linguists are
more cautious in using such inflammatory terms.  And it isn't
clear, to me at least, why he says that her assessment is based on
third-hand sources: Unger is a specialist, and the panel he's
reporting on was composed of himself and other specialists.  They
may have been wrong, but even if they were, that wouldn't justify
Manaster Ramer's characterization of their, or of Nichols', comments
on Altaic.  Polemics won't help convince people, so one hopes that
the editors of JL will check Manaster Ramer et al.'s paper carefully
for accuracy and tone before publishing it.
 
  --  Sally Thomason
      sally at isp.pitt.edu
 
P.S. I, for one, disagree with Manaster Ramer's view of the
     value of his own critique of Greenberg's Amerind vs. other
     criticismss of Greenberg's proposals.  Rankins' IJAL review
     is outstanding, as is Poser's IJAL article.  And there are others, too.


From manaster at umich.edu  Thu Jan 29 13:53:52 1998
From: manaster at umich.edu (manaster at umich.edu)
Date: Thu, 29 Jan 1998 08:53:52 EST
Subject: Alexis on classification
In-Reply-To: <1450.886038014@pogo.isp.pitt.edu>
Message-ID: <THU.29.JAN.1998.085352.EST.>

----------------------------Original message----------------------------
Dr. Thomason errs: Unger is not a specialist on the ALtaic
problem and does not claim to be, what Nichols refers to
is Unger's summary of an unpublished paper by Larry Clark
which in turn is supposed to have been a summary for the
panel of the views of Doerfer and Rona-Tas, two well-known
experts on the Altaic question.  This is what makes Nichols'
entirely erroneous statments third-hand, and as to why they
are erroneous, there is no better argument than that Rona-Tas
in fact has consistently refused to reject Altaic (as opposed
to Doerfer). Anyone who knows anything about Altaic studies
would know that, and would also know that Dr. Nichols'
statements are entirely unfounded.  As for Dr. Thomason's
further implications, I would add that (a) one of the coauthors
of my paper in JL is R-S Georg, himself a well-regarded
specialist on the Altaic languages AND an opponent of the
Altaic theory (yes, OPPONENT but a well-informed and honest
one), and (b) I dont think the editors of JL need any lessons
from someone with Dr. Thomason's record.
 
AMR
 
On Wed, 28 Jan 1998, Sarah G. Thomason wrote:
 
>
>
> Let's see: Alexis Manaster Ramer accuses Johanna Nichols of making
> "attacks on Altaic...based on third-hands [sic] sources".  Nichols
> asks for references.  Manaster Ramer gives one reference to a work
> by Nichols (therefore at most one attack).  As evidence for his
> assertion that she attacks Altaic, he quotes her negative
> assessment of the evidence in favor of the Altaic hypothesis.  Perhaps
> Manaster Ramer views any argument against a proposal that he favors as
> an attack on the proposal; I hope and believe that most linguists are
> more cautious in using such inflammatory terms.  And it isn't
> clear, to me at least, why he says that her assessment is based on
> third-hand sources: Unger is a specialist, and the panel he's
> reporting on was composed of himself and other specialists.  They
> may have been wrong, but even if they were, that wouldn't justify
> Manaster Ramer's characterization of their, or of Nichols', comments
> on Altaic.  Polemics won't help convince people, so one hopes that
> the editors of JL will check Manaster Ramer et al.'s paper carefully
> for accuracy and tone before publishing it.
>
>   --  Sally Thomason
>       sally at isp.pitt.edu
>
> P.S. I, for one, disagree with Manaster Ramer's view of the
>      value of his own critique of Greenberg's Amerind vs. other
>      criticismss of Greenberg's proposals.  Rankins' IJAL review
>      is outstanding, as is Poser's IJAL article.  And there are others, too.
>


From johanna at uclink.berkeley.edu  Thu Jan 29 13:58:05 1998
From: johanna at uclink.berkeley.edu (Johanna Nichols)
Date: Thu, 29 Jan 1998 08:58:05 EST
Subject: various posts on classification
Message-ID: <THU.29.JAN.1998.085805.EST.JOHANNA@UCLINK.BERKELEY.EDU>

----------------------------Original message----------------------------
Alexis writes:
 
> Even if ALL the evidence presented by someone for
> theory X is refuted (or if no evidence was ever
> presented in the first place, as happened with Sapir
> several times), that does not mean that X is false
> or incapable of being shown true or even unworthy of
> further effort.
 
and Scott writes:
 
No assumption
of relationship means just that--it most definitely does not mean
that investigating the possibility further is an irresponsible
waste of time, as some of our colleagues do seem to assume.
 
 
I would like to insert the perspective of someone who not only pursues
demonstration of genetic relatedness but also needs to *apply* the results
of historical scholarship in designing samples and doing comparison.  Using
only proven families, evaluating the evidence proposed for relatedness, and
rejecting (as a basis for comparison and sampling) the many proposed
relationships for which no probative evidence has been presented, are
crucial to proper comparison.  Using a consistently designed sample and
reporting the basis for the design to one's readers are responsible
science.  Elsewhere in their posts Alexis mentions 'extremists' and Scott
refers to the 'ideological
climate of contemporary historical linguistics'.  Would that we had, across
most of the historical linguistics literature, a position -- any position
-- as consistent as extremism, or any reliably identifiable ideological
climate.  Then applications could be done with confidence that apples and
apples were being compared.  Note that a wish to compare apples and apples,
or oranges and oranges, or whatever and its ilk, is not the same thing as
liking apples, destroying orange trees, or declaring apples true and
oranges false.  It's just sorting things so you know what you're counting.
 
It is fiendishly difficult even for me, a professional historical linguist,
to sort out all and only apples in the literature.  Little wonder that
geneticists, archeologists, and others can't sort things out properly.  If
my experience is typical, putting the probative evidence up front in more
publications and reviewing it in surveys would foster communication in our
field and between fields and would go far to create consensus about
standards, relatedness, and the worth of different kinds of historical
comparison.


From johanna at uclink.berkeley.edu  Thu Jan 29 21:26:12 1998
From: johanna at uclink.berkeley.edu (Johanna Nichols)
Date: Thu, 29 Jan 1998 16:26:12 EST
Subject: Alexis on classification
In-Reply-To: <Pine.SOL.3.95.980128220716.16857D-100000@qbert.rs.itd.umich.edu>
Message-ID: <THU.29.JAN.1998.162612.EST.JOHANNA@UCLINK.BERKELEY.EDU>

----------------------------Original message----------------------------
Alexis Manaster Ramer writes:
 
  This is what makes Nichols'
>entirely erroneous statments third-hand, and as to why they
>are erroneous, there is no better argument than that Rona-Tas
>in fact has consistently refused to reject Altaic (as opposed
>to Doerfer).
 
        This presupposes that the mere fact that someone believes in it
means that a genetic relationship is real.  It doesn't.  Logically, as
Scott DeLancey has pointed out, the null hypothesis of no relationship
holds until relationship is demonstrated.  Opinion is not the same thing as
demonstrated fact.
 
 
and he writes:
 
Anyone who knows anything about Altaic studies
>would know that, and would also know that Dr. Nichols'
>statements are entirely unfounded.
 
        It should not be necessary to be a specialist in, or even well
acquainted with, a language group or the work on it in order to retrieve
from the literature an accurate assessment of whether the group is a
family.  It's up to the Altaicists to lay out the evidence they regard as
probative and why they regard it as probative.  I've seen nothing I find
probative.  So, though I actually believe that at least Tungusic and
Mongolian are related, I can't treat Altaic as a family for purposes of
sample design, language family censuses, etc.
        I hope the paper by Manaster Ramer et al. reviews the evidence and
not just the history of opinions.
        For what it is worth, in deciding how to regard the genetic status
of Altaic in drawing up the sample for the 1992 book in which I referred to
Unger's 1991 article, I conferred with various people who work on Altaic,
sought information in reference works and overviews, looked over
grammatical paradigms myself, and came to the conclusion that relatedness
wasn't proved and wasn't assumed by enough people in the field to be
regarded as received view.  The 1987 panel discussion reported by Unger,
which I attended, seemed to me to be surveying and presenting received
view, and it reviewed and rejected some of the evidence previously regarded
as probative.  My notes on the panel are not very detailed, and Unger's
piece presents less detail than I recall the actual panel presenting, but I
had taken the panel to be authoritative and therefore took the report of it
to be authoritative.  In any event I came to the panel meeting believing
that Turkic, Tungusic, and Mongolian are a family and that that was
received view, and left the meeting believing that genetic relatedness of
Altaic was not received view among Altaicists and that indeed much of the
evidence was less sound than I had thought.
        I believe this little history underscores the importance of
specialists presenting the evidence (and not just the opinions) in surveys
in the field.  It should be possible for any trained linguist to find, in
reference works, an objective statement of the current received view in the
field and the evidence and standards on which received view is based.
 
        My reading of Alexis Manaster Ramer's posting of Jan. 27 is that it
concedes my two points, namely (1) he knows of no work by me that is either
recent or an attack on Altaic [much less both] and (2) the reviews and
other  critiques of Greenberg's Language in the Americas (1987) do discuss
Greenberg's comparisons.


From delancey at darkwing.uoregon.edu  Thu Jan 29 19:22:01 1998
From: delancey at darkwing.uoregon.edu (Scott DeLancey)
Date: Thu, 29 Jan 1998 14:22:01 EST
Subject: Monogenesis and "simple Darwinian grounds"
In-Reply-To: <199801290434.WAA26863@harper.uchicago.edu>
Message-ID: <THU.29.JAN.1998.142201.EST.DELANCEY@DARKWING.UOREGON.EDU>

----------------------------Original message----------------------------
On Thu, 29 Jan 1998, Richard Janda wrote:
 
>   While agreeing with most of Scott DeLancey's last posting, I believe that
> it is a demonstrable fallacy to claim that, "on simple Darwinian grounds",
> "monogenesis has to be correct".
 
Well, I can't agree with some of your points.  In particular:
 
> terizing the ultimate ancestor of some language family.  That is, even as-
> suming that, due to selection, some mutation which provided the prerequisites
> for language came to characterize all living humans at some point, it does
> not follow that there was only one occasion on which sound/meaning associa-
> tions were arrived at and then passed on to succeeding generations.  Rather,
> a language-facilitating mutation could have been selected for but then not
> immediately acted on, as it were (and this does not strike me as unlikely).
 
This is not only unlikely, it's not even coherent.  How can something be
selected for without being "acted on"?  Selection isn't an abstract
intellectual exercise; it's a label for the fact that some individuals
reproduce more successfully than others, so that their genetic endowment
is disproportionately represented in succeeding generations.  Selection
in connection with a particular trait occurs because that trait leads
to differential reproductive success.  Thus a trait can only be selected
for (or against) if it is manifesting itself in the individuals carrying
it.
     But I had in mind a different Darwinian argument.  Richard says:
 
> It hardly strains credulity to imagine that two groups of early humans living
> widely separated from each other could independently have stumbled onto the
> use of sound/meaning associations for communication in specifically linguistic
> ways.
 
It may not strain credulity, but it's vanishingly unlikely.  Once one
population has developed an advantageous trait, it will then out-reproduce
competing populations which lack that trait, and eventually replace them.
Once one population has developed language, it's not likely that anyone
else will get the chance.  I'm no archeologist, but I understand that
the archeological record suggests an explosively fast expansion of
modern humans (at the expense of Neanderthals).  The obvious inference
is that a single population developed language, giving it a selectional
advantage which allowed it to overspread the world very quickly.
 
Scott DeLancey
Department of Linguistics
University of Oregon
Eugene, OR 97403, USA
 
delancey at darkwing.uoregon.edu
http://www.uoregon.edu/~delancey/prohp.html


From m.cysouw at let.kun.nl  Thu Jan 29 19:20:56 1998
From: m.cysouw at let.kun.nl (Michael Cysouw)
Date: Thu, 29 Jan 1998 14:20:56 EST
Subject: Monogenesis and "simple Darwinian grounds"
In-Reply-To: <199801290434.WAA26863@harper.uchicago.edu>
Message-ID: <THU.29.JAN.1998.142056.EST.M.CYSOUW@LET.KUN.NL>

----------------------------Original message----------------------------
The debate on the monogenesis vs. polygenesis of language normally suffers
from the assumption that we know what we mean when we say 'language' or
'human language'. The conviction people have about monogenesis/polygenesis
seems strongly correlated with their conviction about the specificity of
linguistic behaviour.
 
In it's most general reading, 'language' is a kind of human behaviour with
some language-specific aspects and some general behavioural aspects. One
would want to decide the monogenesis vs. polygenesis debate on the ground
of the origin of the language-specific aspects of humanity (whether
biological or anthropological): were they invented once or more than once?
But as it is still rather unclear where to place the dividing line between
language-specific aspects of human behaviour from language-inspecific
aspects that question is premature.
 
People who believe in monogenesis would normally stress the fact that there
isn't much in linguistic behaviour that we do not find elsewhere in the
human behaviour. They define 'language' so broadly that the question is not
*whether* there has been monogenesis, but rather *which aspects* of this
broadly defined cluster of behavioural traits called language arose before
the splitting of humanity. This seems to be what Scott DeLancey's is
arguing for:
 
> I assume that all languages are related at some level.  The issue
> facing any comparativist working on a relationship that is not yet
> established is, are these languages demonstrably related at the
> level that I am working at.
 
People who believe in (the possibility of) polygenesis normally stress the
fact that there should be some specific universal characteristic of
'language'. The question then remains open when this specific
characteristic arose; it could be monogenetic, but it could also be
polygenetic. This seems to be the possition of Richard Janda. His scenario
needs a strict distinction between the biological preconditions for
language (which are necessary monogenetic) and anthropological use of those
preconditions, which could be imagined to be polygenesic, convergently
leading to one sort of language. The problem of this position always is to
explicate the defining characteristic of 'language' where to decide the
question on.
 
Michael Cysouw
University of Nijmegen, The Netherlands


From sgbrady at ucdavis.edu  Fri Jan 30 00:33:15 1998
From: sgbrady at ucdavis.edu (Sean Brady)
Date: Thu, 29 Jan 1998 19:33:15 EST
Subject: Darwin and language
Message-ID: <THU.29.JAN.1998.193315.EST.SGBRADY@UCDAVIS.EDU>

----------------------------Original message----------------------------
 
Some evolutionary ideas are being discussed in a misinformed fashion.  I
take it that "Darwinian" means something like "due to biological natural
selection on indivduals" (I missed the first message). If so, a few things
need to be cleared up.
 
          The scenario described by Janda:
 
   "That is, even as-
 suming that, due to selection, some mutation which provided the prerequisites
for language came to characterize all living humans at some point, it does
 not follow that there was only one occasion on which sound/meaning associa-
 tions were arrived at and then passed on to succeeding generations.  Rather,
 a language-facilitating mutation could have been selected for but then not
 immediately acted on, as it were (and this does not strike me as unlikely)."
 
is quite reasonable, contrary to the counterargument presented.  This
process, whereby a character that evolved for other functions, or for no
function at all, but which has been co-opted for a new use, is called
exaptation.  Many biological features are thought to have evolved in this
way, perhaps including such dramatic ones as flight in birds.
 
        Also, the process described by DeLancey:
 
  " But I had in mind a different Darwinian argument....  Once one
population has developed an advantageous trait, it will then out-reproduce
competing populations which lack that trait, and eventually replace them.
Once one population has developed language, it's not likely that anyone
else will get the chance.  I'm no archeologist, but I understand that
the archeological record suggests an explosively fast expansion of
modern humans (at the expense of Neanderthals).  The obvious inference
is that a single population developed language, giving it a selectional
advantage which allowed it to overspread the world very quickly"
 
is usually not thought of as "Darwinian" because it defines group
selection, in which populations, and not individuals, are the unit of
selection.  Although group selection has been out of favor for the past few
decades, it is enjoying a comeback lately, and seems a perfectly plausible
mechanism for evolution in organisms such as humans.  But without adequate
information, we can never assume that any population with an advantageous
trait will ALWAYS drive all other population to extinction; there are
countless counterexamples to this assertion in the biological literature.
But in this case, the idea certainly has some merit to it.
 
        Finally, and most importantly, nothing in modern evolutionary
theory that I know of has anything to say about whether a trait can evolve
multiple times or not, which is really what we are debating.  Both
scenarios are perfectly plausible.  This is analagous to the Out-of-Africa
vs. multiregionalism debate in anthropology, in which neither side can say
the other is wrong in theory; evolutionary facts are required to test
alternative hypothesis, each of which are very reasonable.  Until we have
similar types of data concerning langauge evolution (admittedly a lot
harder to get!), I do not think that evoluationary theory has much to
contribue to this debate.
 
 
Sean Brady
Center for Population Biology
University of California, Davis
sgbrady at ucdavis.edu
(916) 752-9977


From johanna at uclink.berkeley.edu  Fri Jan 30 13:15:06 1998
From: johanna at uclink.berkeley.edu (Johanna Nichols)
Date: Fri, 30 Jan 1998 08:15:06 EST
Subject: Monogenesis and "simple Darwinian grounds"
In-Reply-To: <Pine.GSO.3.96.980129083519.13584A-100000@darkwing.uoregon.edu>
Message-ID: <FRI.30.JAN.1998.081506.EST.JOHANNA@UCLINK.BERKELEY.EDU>

----------------------------Original message----------------------------
This has nothing to do with selection, but note that estimates of the size
of the earliest modern human population (ca. 100,000 years ago) range from
a few tens of thousands to a million or more.  The range within which
modern humans evolved, and within which homo erectus evolved, and so on was
comparable in dimensions and shape to New Guinea, and the ecological
context was riverine and lacustrine tropical.  Under comparable ecological
conditions, modern languages of the least complex societies show tremendous
diversity of language families and languages, as well as small
speech-community sizes for languages (a few hundred individuals per
language in many cases).  There would have been room for a hundred or so
languages and a dozen or so distinct language families in the early human
range.  It's hard to imagine that a species could have been viable if its
entire population was comparable to the modal number of speakers per
language in New Guinea and Australia.
 
This makes it hard for me to believe that there was ever just one language.
There might well be a secondary kind of monogenesis in that all but one
line has gone extinct, not for reasons having to do with selection but
accidentally, by ordinary drift (just as, over enough time in a smallish
population, sooner or later all but one of the last names daughter out and
everybody has the same last name, though that doesn't mean they had only
one ancestor).  Unless dialect differentiation and sound change happened at
very different rates or in very different ways or not at all in the early
stages of language development (which I doubt profoundly, since as I
understand it even bird species that learn their songs have song dialects,
which indicates that differentiation with learned transmission isn't unique
to language), I think language as we know it evolved very gradually from
something more primitive and the entire evolution was as a number of
dialects or languages or indeed even language families.
 
If it weren't for the population size and range factors, I'd vote for
monogenesis.
 
Johanna Nichols


From mcv at wxs.nl  Fri Jan 30 18:24:02 1998
From: mcv at wxs.nl (Miguel Carrasquer Vidal)
Date: Fri, 30 Jan 1998 13:24:02 EST
Subject: Monogenesis and "simple Darwinian grounds"
In-Reply-To: <v03020900b0f6dc0246ee@[128.32.100.155]>
Message-ID: <FRI.30.JAN.1998.132402.EST.>

----------------------------Original message----------------------------
Johanna Nichols <johanna at uclink.berkeley.edu> wrote:
 
>This has nothing to do with selection, but note that estimates of the size
>of the earliest modern human population (ca. 100,000 years ago) range from
>a few tens of thousands to a million or more.  The range within which
>modern humans evolved, and within which homo erectus evolved, and so on was
>comparable in dimensions and shape to New Guinea, and the ecological
>context was riverine and lacustrine tropical.  Under comparable ecological
>conditions, modern languages of the least complex societies show tremendous
>diversity of language families and languages, as well as small
>speech-community sizes for languages (a few hundred individuals per
>language in many cases).
 
This whole subject is of course plagued with uncertainties and
assumptions.  I tend to like Eldredge/Gould's theory of "punctuated
equilibrium", which states that speciation generally takes place in
small isolated populations.  I also tend to think that the origin of
"language as we know it" and the origin of our species are related
events, which would indeed take us back to Africa some 100,000 years
ago.  As to population size and area, my guess would be probably less
than a million and somewhat smaller than New Guinea, but somewhere in
that general range.
 
What I don't know is if it's valid to compare the situation in New
Guinea now (or rather, before the colonial era) to the situation in
Africa then.  The linguistic situation in New Guinea is the result of
many thousands of years of human settlement and, presumably,
linguistic differentiation (with little in the way of external
influences outside of the litoral areas where Austronesian languages
are spoken).  The situation at the time of language origin/human
speciation [IF we can equate the two] has no linguistic history, by
definition.
 
>This makes it hard for me to believe that there was ever just one language.
>There might well be a secondary kind of monogenesis in that all but one
>line has gone extinct
 
That idea has always appealed to me.  For one thing, it allows us not
worry about whether homo erectus or h. neanderthaliensis had language.
If they had, those languages are extinct and only the branch(es)
spoken by the earliest h.s.[s.] can survive.  Another situation which
might (but certainly need not) lead to monogenesis of this secondary
kind is the question of early human dispersal.  Even assuming that the
human "homeland" supported several (unrelated?) languages, it may well
be that only one group migrated out, their language becoming the
ancestral language of all varieties spoken outside of a little corner
of Africa.  A small reflux movement would then provide "instant
monogenesis".
 
But let me state again the many uncertainties.  We don't know exactly
how human speciation took place.  We don't know if language predates,
postdates or coincides with the speciation event.  Even assuming a
small isolated "original" population, possessing language, we don't
know how small the original population and area were exactly, making
it hard to say whether a single or multiple dialects/languages may
have been in existence.  We don't know how humans spread out from
their original habitat.  We don't know if secondary extinctions may
have led to "monogenesis of the second kind" at a (much) later date.
 
The odds are that all human languages share a common origin, if not by
primary, then by secondary monogenesis, but there's no way to know for
sure.
 
=======================
Miguel Carrasquer Vidal
mcv at wxs.nl
Amsterdam


From mmorrison at VNET.IBM.COM  Sat Jan 31 16:50:18 1998
From: mmorrison at VNET.IBM.COM (Michael C. Morrison 8-543-4706)
Date: Sat, 31 Jan 1998 11:50:18 EST
Subject: Monogenesis and "simple Darwinian grounds"
Message-ID: <SAT.31.JAN.1998.115018.EST.MMORRISON@VNET.IBM.COM>

----------------------------Original message----------------------------
*** Reply to note of Fri, 30 Jan 1998 13:24:02 -0500 (EST)
*** by mcv at wxs.nl
 
I have to say that this subject is endlessly fascinating, but I also
have to ask, what's the point?  Almost by definition, we will not
know the answers to the questions it raises or the uncertainties
Miguel lists.  Consider current historical linguistic knowledge
and supposition:
 
 - All spoken languages extant today are descended from some prior
   language.  Many of these prior languages are no longer represented
   by any other current usage.  For example, modern Italian can be said
   to have descended from either Latin or a cousin to Latin; neither
   Latin nor a reputed cousin is in use today by native speakers.
 
 - Most (if not all) historical data we have about modern languages
   comes from the written record as presented to us through history
   or archaeology.  At most, this written record gives us insight
   into language as used maybe 8000 years ago.
 
 - Comparative work has given us much data about languages that
   predate the written record.  This data is by no means certain,
   but for some language families, we have some consensus about the
   proto language, for example, proto-Indo-European.  But even these
   proto languages only extend our view of the past by another few
   thousand years or so.
 
So at the best, our current knowledge of our earliest languages only
takes us back 10, maybe 15, thousand years.  What about the other
85,000 years of human language (assuming humans developed language
about 100,000 years ago, as seems probable -- unproven, but likely)?
Can we ever know about these truly ancient languages?  If the answer
is 'no', then the whole monogenesis v. polygenesis argument becomes an
exercise in philosophy, not in linguistics.  If the answer is 'yes',
then we can solve the problem because we will have found either one
superfamily for all languages or several superfamilies.
 
Personally, I don't believe we will be able to reconcile the many
language families of today to a degree that will answer the genesis
question.  Based on the data with which we have to work, I doubt
any overarching superfamily will be accepted by the field as a
whole, and so the debate will continue.  Sigh
 
Michael
----------------------------------------------------------------------
Michael C. Morrison   Santa Teresa Laboratory      Phone (408)463-4706
IMS User Technology   IBM Software Solutions       Fax   (408)463-3696
Lotus Notes ID:       MCMORRIS at IBMUSM50   IBMLink:  MORRISON at TORIBM
Internet ID:          MMORRISON at VNET.IBM.COM  or  USIB47H4 at IBMMAIL.COM
IBM Mail Exchange:    USIB47H4 at IBMMAIL     or   USIB4MCM at IBMMAIL
X.400 Address:        G=mcmgm; S=morrison; P=ibmmail; A=ibmx400; C=us
----------------------------------------------------------------------