There are 25 messages in this issue.
Topics in this digest:
1. Re: cthulhu fhtagn
From: João Ricardo de Mendonça <[EMAIL PROTECTED]>
2. Re: What's a good isolating language to look at
From: João Ricardo de Mendonça <[EMAIL PROTECTED]>
3. Re: What's a good isolating language to look at
From: Roger Mills <[EMAIL PROTECTED]>
4. Re: Thoughts on Word building
From: Jim Henry <[EMAIL PROTECTED]>
5. Re: What's a good isolating language to look at
From: Dirk Elzinga <[EMAIL PROTECTED]>
6. Re: Sources of Irregularity (was: Re: Isolating, Inflected, Word
Building, ETC.)
From: tomhchappell <[EMAIL PROTECTED]>
7. Re: cthulhu fhtagn
From: "David G. Durand" <[EMAIL PROTECTED]>
8. Re: PIE past time (was: isolating is equivalent to inflected)
From: Rob Haden <[EMAIL PROTECTED]>
9. Etymology of _insula_ (was Re: Thoughts on Word building)
From: Jörg Rhiemeier <[EMAIL PROTECTED]>
10. Re: Thoughts on Word building
From: Benct Philip Jonsson <[EMAIL PROTECTED]>
11. Re: Thoughts on Word building
From: Jörg Rhiemeier <[EMAIL PROTECTED]>
12. Re: What's a good isolating language to look at
From: tomhchappell <[EMAIL PROTECTED]>
13. Re: An unusual incorporation scheme
From: tomhchappell <[EMAIL PROTECTED]>
14. Re: What's a good isolating language to look at
From: Benct Philip Jonsson <[EMAIL PROTECTED]>
15. Re: Etymology of _insula_ (was Re: Thoughts on Word building)
From: Rob Haden <[EMAIL PROTECTED]>
16. Re: What's a good isolating language to look at
From: Dirk Elzinga <[EMAIL PROTECTED]>
17. Re: Thoughts on Word building
From: R A Brown <[EMAIL PROTECTED]>
18. Re: An unusual incorporation scheme
From: Paul Bennett <[EMAIL PROTECTED]>
19. Re: An unusual incorporation scheme
From: Paul Bennett <[EMAIL PROTECTED]>
20. OFFLIST: Re: What's a good isolating language to look at
From: tomhchappell <[EMAIL PROTECTED]>
21. Re: What's a good isolating language to look at
From: Jim Henry <[EMAIL PROTECTED]>
22. IPA/CXS questions
From: Jim Henry <[EMAIL PROTECTED]>
23. Re: Sources of Irregularity (was: Re: Isolating, Inflected, Word
Building, ETC.)
From: Jim Henry <[EMAIL PROTECTED]>
24. Re: IPA/CXS questions
From: "Mark J. Reed" <[EMAIL PROTECTED]>
25. Re: IPA/CXS questions
From: "Mark J. Reed" <[EMAIL PROTECTED]>
________________________________________________________________________
________________________________________________________________________
Message: 1
Date: Wed, 7 Dec 2005 15:22:05 -0200
From: João Ricardo de Mendonça <[EMAIL PROTECTED]>
Subject: Re: cthulhu fhtagn
On 12/5/05, Reilly Schlaier <[EMAIL PROTECTED]> wrote:
> ph'nglui mglw'nafh cthulhu r'lyeh wgah'nagl fhtagn.
> what is the exact meaning not just the translation given in the call of
> cthulhu book.
> also the pronuncation.
> thanks very much
>
The Wikipedia article on Cthulhu says (I've replaced IPA for XSAMPA):
Cthulhu's name is usually pronounced /k@'Tu:lu:/, /k@'TU:lu:/, or
/k@'t_hU:lu:/; however, according to Lovecraft, this may simply be the
closest that human vocal cords can come to reproducing the syllables
of an alien language[1]. In fact, Lovecraft speculated that
"Khlul'hloo"[2] might be closer to the actual pronunciation.
João Ricardo de Mendonça
________________________________________________________________________
________________________________________________________________________
Message: 2
Date: Wed, 7 Dec 2005 15:30:30 -0200
From: João Ricardo de Mendonça <[EMAIL PROTECTED]>
Subject: Re: What's a good isolating language to look at
On 12/6/05, R A Brown <[EMAIL PROTECTED]> wrote:
> João Ricardo de Mendonça wrote:
> > The usual examples of isolating languages are Mandarim and Vietnamese. I
> > guess Mandarim should be the easiest to find references for in the West.
>
> Yes, but Mandarin does have grammatical affixes; it is not purely
> isolating. In fact it has been argued on this on this list (more than
> once IIRC) that English is more isolating than modern Mandarin.
<SNIP>
But is there a purely isolating language? Or purely fusional, or
purely agglutinating etc? Even if Vietnamese is "(almost) 100%
isolating", it can still be claimed that it isn't purely isolating.
Maybe we should take such labels as implying observed tendencies
within a language, and not as natural categories. That's why I
mentioned the "classical examples", because this is, I believe, what
people had in mind when they created the concept of an isolating
language.
João Ricardo de Mendonça
________________________________________________________________________
________________________________________________________________________
Message: 3
Date: Wed, 7 Dec 2005 12:37:47 -0500
From: Roger Mills <[EMAIL PROTECTED]>
Subject: Re: What's a good isolating language to look at
Dirk Elzinga wrote:
> There is an easy way to find out which language is more
> analytic/synthetic. Take representative texts from each language,
> count up the number of words and then the number of morphs. You can
> then make a ratio of morphs/words--the "synthesis index"--for the
> languages in question. In a 1960 paper, Greenberg used this method and
> came to the conclusion that English had 1.68 morphs per word, on
> average. Vietnamese had 1.06. (snip etc.)
Rings a bell. Somewhere within my tenure on Conlang (Spring 2000 on), IIRC
we did this...Might be worth doing again??
I'm also getting a vague yoo-hoo from the rusty memory-banks....Seems to me
this is exactly the sort of thing Structuralists would have done even
earlier, in the 30s-40s-50s. The Greenberg paper must mention previous work
in its bibliography. (I did my intro. course work in Ling. in the middish
60s, when Chomsky was still an upstart, and Lamb was a contender :-)), and
do recall discussions of this sort of thing. One project IIRC involved
analysis of Brit. telephone conversations-- excited some controversy because
it involved recording on the sly.)
> I just asked a corpus linguist here in the
> department, and he has not heard of anyone actually publishing on this
> question (i.e., relative morphological complexity of different
> genres).
Perhaps because (if I'm right) it was done/published in the Dark Ages before
you were born. Oh you kids....:-)))) A review of period textbooks might turn
up something; Gleason, Sapir maybe, Bloomfield??? (I'll check my old copy of
Hockett.)
________________________________________________________________________
________________________________________________________________________
Message: 4
Date: Wed, 7 Dec 2005 13:16:29 -0500
From: Jim Henry <[EMAIL PROTECTED]>
Subject: Re: Thoughts on Word building
On 12/6/05, Andreas Johansson <[EMAIL PROTECTED]> wrote:
> Quoting Jim Henry <[EMAIL PROTECTED]>:
> > On 12/6/05, caeruleancentaur <[EMAIL PROTECTED]> wrote:
> > >.......However, I'm having trouble with the
> > > word "island" itself. _Insula_ has no known etymology. The "is-"
> > In gzb I derive it from a root {vleq'tax}, "body of water surrounded by
> > land",
> > with the complement suffix: {vleq'tax-txaj}. That refers to islands
> > and continents both; I could add -ny "small" to be more specific.
> May I ask why you chose that polarity? "Land surrounded by water" seems to me
> the more basic concept.
I didn't plan it out systematically; I coined the word {vleq'tax}
pretty early on, probably for a translation I was doing,
when I decided {pwiqm-daj} (water-mass) wasn't specific
enough for this sense & I didn't want to idiomatize it.
Later I realized I could derive "island" from {vleq'tax} with
the complement suffix. But Paul Bennett's comment,
> I suspect that a lot of people encounter more lakes, ponds and puddles in
> their life than islands, not to mention rivers, creeks, brooks and streams.
sounds like a good excuse to me. Certainly it's
true of me, and since gzb is for my personal use
it makes sense.
Except, I almost never use {vleq'tax}. I use the
more general {pwiqm-daj} and let context
disambiguate it. I also have a word {siq} for
flowing water (stream, creek, river) but I might
sometimes use {pwiqm-flu} (water-flow) instead;
maybe I will eventually mark {siq} & {vleq'tax}
as archaic, coin a short word for "landmass"
& derive "enclosed body of water" from that,
for use on the rare occasions when I want
to emphasize I'm not talking about the ocean.
--
Jim Henry
http://www.pobox.com/~jimhenry
http://www.pobox.com/~jimhenry/conlang.htm
http://www.pobox.com/~jimhenry/gzb/gzb.htm
http://www.pobox.com/~jimhenry/esp.htm
...Mind the gmail Reply-to: field
________________________________________________________________________
________________________________________________________________________
Message: 5
Date: Wed, 7 Dec 2005 11:49:12 -0700
From: Dirk Elzinga <[EMAIL PROTECTED]>
Subject: Re: What's a good isolating language to look at
On 12/7/05, Roger Mills <[EMAIL PROTECTED]> wrote:
> Dirk Elzinga wrote:
> > There is an easy way to find out which language is more
> > analytic/synthetic. Take representative texts from each language,
> > count up the number of words and then the number of morphs. You can
> > then make a ratio of morphs/words--the "synthesis index"--for the
> > languages in question. In a 1960 paper, Greenberg used this method and
> > came to the conclusion that English had 1.68 morphs per word, on
> > average. Vietnamese had 1.06. (snip etc.)
>
> Rings a bell. Somewhere within my tenure on Conlang (Spring 2000 on), IIRC
> we did this...Might be worth doing again??
I did bring it up about that time (I was still at the University of
Utah; I came to BYU Fall 2001). I don't think anyone else posted
figures on their own projects, though. I think it's definitely worth
talking about.
> I'm also getting a vague yoo-hoo from the rusty memory-banks....Seems to me
> this is exactly the sort of thing Structuralists would have done even
> earlier, in the 30s-40s-50s. The Greenberg paper must mention previous work
> in its bibliography. (I did my intro. course work in Ling. in the middish
> 60s, when Chomsky was still an upstart, and Lamb was a contender :-)), and
> do recall discussions of this sort of thing.
I just looked again at the Greenberg article. The article first
appeared in 1954, in a festschrift for Wilson D. Wallis. Greenberg
also refers to a paper that Hockett presented at the 1949 LSA meeting
which has the same idea behind it, and to a talk Greenberg gave at the
Linguistics Circle of New York in January 1950. In Hockett's _Course
in Modern Linguistics_ (1958) he says: "The easiest rough measure of
morphological complexity is the average number of morphemes per word
in a representative sample." So the idea was in the air during the
50s.
One project IIRC involved
> analysis of Brit. telephone conversations-- excited some controversy because
> it involved recording on the sly.)
That'd be fun to look at ... I know there's a more recent corpus of
telephone conversations that people are using. It was created with the
speakers' knowledge and consent, though.
> > I just asked a corpus linguist here in the
> > department, and he has not heard of anyone actually publishing on this
> > question (i.e., relative morphological complexity of different
> > genres).
>
> Perhaps because (if I'm right) it was done/published in the Dark Ages before
> you were born. Oh you kids....:-)))) A review of period textbooks might turn
> up something; Gleason, Sapir maybe, Bloomfield??? (I'll check my old copy of
> Hockett.)
>
Actually, when I looked again at the Greenberg article, I found that
he did briefly address the question of style. He gives figures for
passages taken from the _Ladies' Home Journal_, R. Linton's _Study of
Man_, and O. J. Kaplan's _Mental Disorders in Later Life_; they are
1.62, 1.65, and 1.60, respectively. But that's as far as he got.
I have to say that I like the old guys! I have copies of Sapir,
Bloomfield, Hockett, Gleason, and Hall. My professional work on Native
American languages seems to resonate more strongly with them (for
obvious reasons!) than with anything I learned in grad school, though
that has proved useful as well.
Dirk
--
Gmail Warning: Watch the reply-to!
________________________________________________________________________
________________________________________________________________________
Message: 6
Date: Wed, 7 Dec 2005 19:19:08 -0000
From: tomhchappell <[EMAIL PROTECTED]>
Subject: Re: Sources of Irregularity (was: Re: Isolating, Inflected, Word
Building, ETC.)
--- In [email protected], Roger Mills <[EMAIL PROTECTED]> wrote:
> [snip]
> As the old saying goes, "sound change produces irregularity,
> analogy restores regularity". And borrowing, esp. between related
> languages, can introduce irregularity.
But that's not all, right?
Used to be English had "eode" as the past tense of "go". (Suppletive
and irregular already, I know, but that doesn't vitiate my point.)
Used to be English had "went" as the past tense of "wend";
regular, like "bend-bent, lend-lent, pend-pent, rend-rent, send-sent".
(Forgive me if some of the above are errors; the point is, there was
such a paradigm and it had several verbs in it.)
But people got used to wending their way home instead of just going
there, at least in the past tense; so "I home went" instead of "I
home eode". (Or something like that.)
Right or wrong?
--
Is it not also true that "to be" was put together out of three
different verbs? "was" from "wax", etc.?
-----
Tom H.C. in MI
________________________________________________________________________
________________________________________________________________________
Message: 7
Date: Wed, 7 Dec 2005 19:42:50 +0000
From: "David G. Durand" <[EMAIL PROTECTED]>
Subject: Re: cthulhu fhtagn
On 12/7/05, Aaron Grahn <[EMAIL PROTECTED]> wrote:
> My belief (with not much basis, except that I don't think Lovecraft was
> a conlanger) is that it has no precise meaning. I would render fhtagn as
> to wait, third person singular present indicative, simply because
> "Cthulhu fhtagn" is repeated as a single unit, and "Cthulhu waits" makes
> sense in an eerie kind of way.
I've always tended to go with this sub-phrase theory as well. I've got
a low-energy project going to create a R'lyehian language, but for the
moment I just have a script. The Truetype font is available in the
files area of the neographies group at Yahoo. It's a very
compositional script with a squiddy feel with squarish glyph blocks
ala Mayan. I've gone far enough to determine that the written grammar
is two-dimensional/tree-shaped, and that the linear form of the
language is an artifact of our brains trying to comprehend
communication forms tailored to a consciousness with direct experience
of higher dimensions.
I think Lovecraft's letters show that he was not a conlanger, but a
writer striving for an unearthly effect and consonant/vowel
combinations that would be unsettling as well as alien.
I think if we remember that spatial inflections (or audible mutations)
can represent the relational and connective tissue of the sentence,
that there is plenty of room for the needed meanings to attach to the
sounds left over after removing the words that we "know".
I've taken my own non-standard approach to pronouncing the language of
the "Great Old Ones", but one that is based in a combination of
english convention and IPA-like indication of voicing and aspiration.
So cth is the cluster kT as you would think, but u is U lh is a
voiceless lateral fricative. The final 'h' in R'lyeh is pronounced
(perhaps pharyngeally (a word I'm sure Lovecraft would have liked)),
and so forth. Some of it's based on the idea that violating English
phonotactics and sound inventory is not that hard, and makes it more
fun.
--
-- David
(Providence born and still resident!)
________________________________________________________________________
________________________________________________________________________
Message: 8
Date: Wed, 7 Dec 2005 14:54:51 -0500
From: Rob Haden <[EMAIL PROTECTED]>
Subject: Re: PIE past time (was: isolating is equivalent to inflected)
On Tue, 6 Dec 2005 14:06:15 +0000, R A Brown <[EMAIL PROTECTED]>
wrote:
>_laudatus_ just is _not_ past. It shows perfect aspect, but is
>indifferent as regards tense.
It is also (generally, at least) passive. That is, _lauda:tus_
meant 'having been praised', not 'having praised'.
>>The imperfect is _laudabam_ and the perfect is _laudavi_, without a
>>dental ending.
>
>The imperfect is certainly past and, as Andreas says, has no dental ending.
Current speculation has the Latin imperfect and future tenses arising from
forms of the IE verb *bheux- 'be(come)' (here I use <x> for 'h2').
Corroborating evidence for this hypothesis comes from related Italic
languages, such as the closely related Faliscan: cf. Faliscan _carefo_ 'I
will be without' vs. Latin _care:bo:_ (here Faliscan has */bh/ > /f/ in
medial position, while Latin does not).
>As for _laudavi_ that could be either _present_ perfect (I have praised)
>or simple past (I praised); the difference between the two meanings was
>felt by Classical writers in that if the so-called 'perfect tense' had a
>present perfect meaning, 'primary tenses' of the subjunctive were used
>in subordinate clauses; but if it had the simple past meaning, then
>'historic tenses' were used. (Of course 'tenses' when referring to the
>subjunctive forms did not have the strict meaning of "time reference").
The two-way usage here probably comes from the fact that the Latin "perfect
tense" had its origins in the IE stative conjugation. This IE verb form
was tenseless and stood as an aspect apart from that of the eventive
conjugation, which took a different set of endings in the singular.
However, the _v_-element present in the Latin 1st-conjugation perfect forms
has an unknown origin, unless it too somehow stems from the IE verb *bheux-
. In any case, the Latin "perfect tense" is the result of combining the
earlier stative conjugation with other IE elements (including some eventive
endings, such as the present-indicative marker *-i in 1/2sg, 3sg *-t, and
the sigmatic aorist suffix *-s).
>The only two tenses that are unambiguously past in Latin are the
>imperfect (laudabam) and the pluperfect (laudaveram) - neither has
>dental endings.
>
>> Everything I've read would indicate that the dental morpheme in Germanic
>> pasts and imperfects are indeed derived from a form of the
>> auxillary "do".
>
>And everything I've read indicates that also. As for dentals showing
>past time, one may, perhaps, think of the 'weak aorist' of ancient Greek
>being formed withe dental fricative [s]. But, as the subjunctive,
>optative, participle & infinitives forms of the aorist clearly show,
>this was an _aspectual_ marking, not one of past time. Past time per_se
>in ancient Greek was marked by the prefix e- (or, if verb with a vowel,
>by lengthening the vowel): the so-called 'augment'.
>
>I am not aware of dental being a mark of past time in PIE; but I must
>confess I have not kept up to date with the latest thinking on PIE. I
>should welcome enlightenment.
What's interesting about the IE verb system is that it must have been a
system in flux. Namely, at the time of "breakup" (i.e. earliest dialectal
divergence to interfere with intelligibility), IE's verb system was
shifting from a primarily aspect-based scheme to a tense-based one.
Originally, IE did not mark tense morphologically. The primary aspectual
contrast in IE verbs was stative vs. eventive aspect. Eventive verbs were
split further into aorist vs. imperfect aspects (or, if you prefer,
punctual vs. durative). The distinctions within eventives were lexical,
and there were means by which a verb in one category could be derived from
a verb in the other. By and large, however, the stative verbs seem to have
been derived from eventives, since almost all of the verbs in the stative
class involve eventive roots plus reduplication (of the Ce-CoC variety) and
different endings in 1-3sg. Statives also could not inflect for mood or
voice, while eventives could. IE had active and middle voices and
indicative, imperative, subjunctive, and optative moods. Interestingly,
the middle voice took different endings from both the active voice and
stative verbs, although they seem to have been more similar to the latter.
It is this similarity, and the nature of the stative conjugation, that have
led me to suspect that the middle voice and stative verbs had a common
origin, with the former becoming subsumed under the eventive category.
Such a state of affairs would explain the rather large number of deponent
middle verbs in IE.
Anyways, around the time IE broke up, the above verb system was shifting to
a more tense-based one. In particular, a particle *i, which was used to
denote present time, became enclitic and then bound to the verb, resulting
in a past/non-past distinction. Greek, Indo-Iranian, and Armenian also
bounded another particle, *e, to the verb, first proclitically and then as
a prefix, forming the "temporal augment" in those languages. Whether this
was a shared innovation among those languages is unknown right now. At
some point around this time, the stative conjugation came to be
reinterpreted as a conjugation for perfect aspect. Further changes then
occurred, involving the absorption of some derived verb forms (mainly the n-
infix "presents" and the sigmatic "aorists") into the tense system.
The "classical IE" verb system shows itself at this point, with present,
imperfect, perfect, and aorist "tenses" (really combinations of tense and
aspect). Later on, this system was in turn changed by the various
descendant languages.
If you have any questions, feel free to ask. :) Sorry for the dry
description, but it's hard to make this subject sound exciting. :P
- Rob
________________________________________________________________________
________________________________________________________________________
Message: 9
Date: Wed, 7 Dec 2005 21:20:17 +0100
From: Jörg Rhiemeier <[EMAIL PROTECTED]>
Subject: Etymology of _insula_ (was Re: Thoughts on Word building)
Hallo!
caeruleancentaur wrote:
> --- In [email protected], Andreas Johansson <[EMAIL PROTECTED]> wrote:
>
> >Also, in what sense does a European language like German not have
> >a "compounding system"? It's brimful of compounds!
>
> I use German a lot for finding compound nouns. It is, indeed, brimful.
>
> Someone recently gave a compound from the Chinese for "peninsula":
> half-island. I had wanted to use the literal translation from the
> Latin: almost-island. However, I'm having trouble with the
> word "island" itself. _Insula_ has no known etymology.
Irish has _inis_, Welsh has _ynys_, and Latin _insula_ looks
like a diminutive of a similar word (perhaps _insula_ < *inis-ula?).
So this looks like an Italo-Celtic etymology, but there may be
problems with this I do not see. Ray?
Greetings,
Jörg.
________________________________________________________________________
________________________________________________________________________
Message: 10
Date: Wed, 7 Dec 2005 20:55:09 +0100
From: Benct Philip Jonsson <[EMAIL PROTECTED]>
Subject: Re: Thoughts on Word building
Andreas Johansson skrev:
> That would mean that the etymological meaning of Scandinavian _öy, _ö_, _ey_
> "island" is simply "water", since it, I'm told, is cognate with the first half
> of "island". Seems like an odd shift in meaning to me.
No, _øy_ is derived from an adjective *a:k_wjo- "related to water".
In German the same stem becomes _Aue_ "Floodplain, meadow".
--
/BP 8^)>
--
Benct Philip Jonsson -- melroch at melroch dot se
Solitudinem faciunt pacem appellant!
(Tacitus)
________________________________________________________________________
________________________________________________________________________
Message: 11
Date: Wed, 7 Dec 2005 21:36:21 +0100
From: Jörg Rhiemeier <[EMAIL PROTECTED]>
Subject: Re: Thoughts on Word building
Hallo!
Andreas Johansson wrote:
> Quoting Taka Tunu <[EMAIL PROTECTED]>:
>
> > I respond to your mails because they may give people who don't practice
> > kanjis
> > and compounding systems the feeling that they are less "semantically
> > efficient"
> > than the European derivational systems and this is just wrong.
>
> I'd think someone who'd compare the "semantical efficience" of kanji (a
> writing
> system) and European systems of derivation (an entity vague enough to approach
> meaninglessness, but certainly not a writing system) is a lost case anyway.
Well put. How do you compare writing systems with (subsystems of)
languages? That's meaningless. But there are countless crackpots who
cannot tell one from the other, uttering such claims as "Sumerian is
the world's oldest language". Beep beep. And "European language" is a
meaninglessly vague concept anyway.
Taka Tunu (who is it, by the way, hiding behind that sobriquet?),
judging from his flamewar with Henrik, seems to have more opinion than
sense anyway, and tends to comment on posts he hasn't read or at least
hasn't understood. (I have observed that people hiding behind
pseudonyms often post more bull of all sorts than people using their
real name, by the way.) Just my impression.
> Also, in what sense does a European language like German not have a
> "compounding
> system"? It's brimful of compounds!
As a native speaker, I can only confirm this.
Greetings,
Jörg.
________________________________________________________________________
________________________________________________________________________
Message: 12
Date: Wed, 7 Dec 2005 20:00:52 -0000
From: tomhchappell <[EMAIL PROTECTED]>
Subject: Re: What's a good isolating language to look at
--- In [email protected], João Ricardo de Mendonça
<[EMAIL PROTECTED]> wrote:
>
> On 12/6/05, R A Brown <[EMAIL PROTECTED]> wrote:
>> João Ricardo de Mendonça wrote:
>>> The usual examples of isolating languages are Mandarim and
>>> Vietnamese. I
>>> guess Mandarim should be the easiest to find references for in
>>> the West.
>>
>> Yes, but Mandarin does have grammatical affixes; it is not purely
>> isolating. In fact it has been argued on this on this list (more
>> than
>> once IIRC) that English is more isolating than modern Mandarin.
>
> <SNIP>
>
> But is there a purely isolating language? Or purely fusional, or
> purely agglutinating etc? Even if Vietnamese is "(almost) 100%
> isolating", it can still be claimed that it isn't purely isolating.
> Maybe we should take such labels as implying observed tendencies
> within a language, and not as natural categories. That's why I
> mentioned the "classical examples", because this is, I believe, what
> people had in mind when they created the concept of an isolating
> language.
>
> João Ricardo de Mendonça
>
--- In [email protected], Roger Mills <[EMAIL PROTECTED]> wrote:
>
> Dirk Elzinga wrote:
>> There is an easy way to find out which language is more
>> analytic/synthetic. Take representative texts from each language,
>> count up the number of words and then the number of morphs. You can
>> then make a ratio of morphs/words--the "synthesis index"--for the
>> languages in question. In a 1960 paper, Greenberg used this method
>> and
>> came to the conclusion that English had 1.68 morphs per word, on
>> average. Vietnamese had 1.06. (snip etc.)
>
> Rings a bell. Somewhere within my tenure on Conlang (Spring 2000
> on), IIRC we did this...Might be worth doing again??
>
> I'm also getting a vague yoo-hoo from the rusty memory-
> banks....Seems to me
> this is exactly the sort of thing Structuralists would have done
> even
> earlier, in the 30s-40s-50s. The Greenberg paper must mention
> previous work
> in its bibliography.
> (I did my intro. course work in Ling. in the middish
> 60s, when Chomsky was still an upstart,
> and Lamb was a contender :-)), and
> do recall discussions of this sort of thing. One project IIRC
> involved
> analysis of Brit. telephone conversations-- excited some
> controversy because
> it involved recording on the sly.)
>
>> I just asked a corpus linguist here in the
>> department, and he has not heard of anyone actually publishing on
>> this
>> question (i.e., relative morphological complexity of different
>> genres).
>
> Perhaps because (if I'm right) it was done/published in the Dark
> Ages before
> you were born. Oh you kids....:-)))) A review of period textbooks
> might turn
> up something; Gleason, Sapir maybe, Bloomfield??? (I'll check my
> old copy of Hockett.)
>
Thorndike, too, apparently: -- this struck me because I read one of
his Depression-era word-counting reports, and he has a dictionary
named after him.
Joan L. Bybee answered the question João Ricardo de Mendonça asks of
Ray A. Brown "No" in her "Morphology" book (1985). (Incidentally,
she's at the University of New Mexico now; and she is the same author
as the Joan B. Hooper who published from around 1972 (or before) to
around 1981 (or after), tho' she apparently started using Joan L.
Bybee (again?) in 1980 or before.)
Bybee quotes Sapir as having also decided that there weren't any
purely isolating, nor purely polysynthetic, nor purely agglutinating,
nor purely fusing languages, IIRC; and quotes counts collected by
Thorndike supporting such a conclusion.
Bybee makes her own point, that, the derivational morphology of a
language is likely to be more towards the fusing (as opposed to the
agglutinating) end than the inflectional morphology of the same
language. (Of course, she also makes the point -- and I know she was
not the first author I read to make it -- that, in any given
language, the dividing line between "derivation" and "inflection"
is "fuzzy".) She backs it up with several examples; Slavic aspectual
affixes that have more semantic content than just aspect; patient-
number-agreement affixes in various (mostly ergative) languages that
have more semantic content than just grammatical number of patients;
etc.
--
Tom H.C. in MI
________________________________________________________________________
________________________________________________________________________
Message: 13
Date: Wed, 7 Dec 2005 20:33:41 -0000
From: tomhchappell <[EMAIL PROTECTED]>
Subject: Re: An unusual incorporation scheme
--- In [email protected], Peter Bleackley
<[EMAIL PROTECTED]> wrote:
>
> Suppose we start off with a simple isolating language, with basic
> word order
>
> S V I O
>
> where I is the indirect object. Now suppose that the language
> evolves to
> start incorporating the least active argument into the verb.
The least referential, most generic element is incorporated first.
A noun, when incorporated, is never referential and never specific;
if a referential and/or specific and/or definite meaning is intended,
a free form is always used, not an incorporated form.
> We
> therefore
> have the following possible sentence structures
>
> Intransitive S.V
> Transitive S V.O
> Ditransitive S V.I O
A noun, when incorporated, is never referential and never specific;
if a referential and/or specific and/or definite meaning is intended,
a free form is always used, not an incorporated form.
This is one reason proper nouns, for instance, aren't ever
incorporated. (Except as in-jokes.)
The product of incorporation has to be a characteristic or customary
or frequent or habitual or culturally salient activity.
That's why anyone can be a dog-minder, but only people in my house
can be Harley-minders. (Or even know what Harley-minding is, as
meant when spoken by a member of my household.)
In prototypical ditransitive sentences, the primary or indirect
object -- the recipient or beneficiary -- is a good deal likelier to
be human or animate than the secondary or direct object (the patient
or theme). It is, therefore (or, perhaps, merely, "also" instead
of "therefore"), much more likely to be specific and referential than
the secondary or direct object.
In my perhaps-insufficiently-humble opinion, most ditransitive
clauses will not have the primary or indirect object -- certainly not
the recipient or beneficiary -- a better candidate for incorporation
than the secondary or direct object -- certainly not better than the
patient or theme.
As examples; "I gave the toy to the child" -- all definite.
"I gave toys to children" -- neither object specific.
"I toy-gave children" is a lot likelier than "I child-gave toys".
"I went all over town, and spent all night, toy-giving" is a lot
likelier than "I went all over town, and spent all night, child-
giving".
But, I don't think it's impossible; I asked earlier this year about
incorporating subjects in English. It can happen; it's just not
likely to be the usual thing.
(BTW:
Also, I'm told, the incorporated nouns are "always" (?) at the most
common level of folk-taxonomy. One can go tree-chopping, but not, as
I understand it, pine-chopping or oak-chopping; although, around this
time of year, Christmas-tree-chopping is a distinct possibility
(because of the "customary, culturally salient" thing, I guess.))
> This would be an unusual tripartite system, with a
> dechticaetiative
> element
> (mind you S V I O could be said to be
> dechticaetiative
> to start with).
A new word for me!
Is "dechticaetiative" a synonym for "primative/secundative" and the
opposite of "direct/indirect", or does it apply strictly to passive
constructions promoting the primary object instead of the direct
object?
If it is a synonym, it's shorter than "primative/secundative";
(although harder to memorize the spelling, and harder -- for me -- to
pronounce). I'll try to learn it.
> I came up with this while playing around with the idea that an
> ergative
> language with incorporation might well incorporate absolutes,
> rather than
> objects.
I will bet you a Confederate sou* that in fact ergative incorporating
languages _do_, for the most part, incorporate the absolutives as
opposed to the ergatives. In effect, "the same" is true for
accusative incorporating languages as well; intransitives would have
to incorporate the only argument available, if they incorporate
anything at all, while monotransitives usually incorporate "the"
object, and ditransitives usually incorporate "an" object, much more
easily than "the subject".
In most incorporating languages, the first participant incorporated,
(if one is incorporated), is the "patient" (if there is a patient),
regardless of alignment.
*To the best of my knowledge the Confederate States of America never
coined nor issued any "sou"s. Even if they did, they wouldn't be
worth anything now. So, if I lose my bet, you probably won't be too
inconvenienced by the fact that I don't have one.
>
> What do you think?
I liked your question.
>
> Pete
>
Tom H.C. in MI
P.S. I think English is primative/secundative, too; I suppose that
makes me one of those people who think English is dechticaetiative.
--T
________________________________________________________________________
________________________________________________________________________
Message: 14
Date: Wed, 7 Dec 2005 21:01:00 +0100
From: Benct Philip Jonsson <[EMAIL PROTECTED]>
Subject: Re: What's a good isolating language to look at
Roger Mills skrev:
> Perhaps because (if I'm right) it was done/published in the Dark Ages before
> you were born. Oh you kids....:-)))) A review of period textbooks might turn
> up something; Gleason, Sapir maybe, Bloomfield??? (I'll check my old copy of
> Hockett.)
>
>
I was born around the time you were getting out of your Ling. intro
courses, yet I am the proud owner of a copy of Gleason, bought in nearly
mint condition at my friend's used books shop here in Gothenburg.
Oh wonder!
--
/BP 8^)>
--
Benct Philip Jonsson -- melroch at melroch dot se
Solitudinem faciunt pacem appellant!
(Tacitus)
________________________________________________________________________
________________________________________________________________________
Message: 15
Date: Wed, 7 Dec 2005 15:47:38 -0500
From: Rob Haden <[EMAIL PROTECTED]>
Subject: Re: Etymology of _insula_ (was Re: Thoughts on Word building)
On Wed, 7 Dec 2005 21:20:17 +0100, =?iso-8859-1?Q?J=F6rg?= Rhiemeier
<[EMAIL PROTECTED]> wrote:
>Hallo!
>
>caeruleancentaur wrote:
>
>> --- In [email protected], Andreas Johansson <[EMAIL PROTECTED]> wrote:
>>
>> >Also, in what sense does a European language like German not have
>> >a "compounding system"? It's brimful of compounds!
>>
>> I use German a lot for finding compound nouns. It is, indeed, brimful.
>>
>> Someone recently gave a compound from the Chinese for "peninsula":
>> half-island. I had wanted to use the literal translation from the
>> Latin: almost-island. However, I'm having trouble with the
>> word "island" itself. _Insula_ has no known etymology.
>
>Irish has _inis_, Welsh has _ynys_, and Latin _insula_ looks
>like a diminutive of a similar word (perhaps _insula_ < *inis-ula?).
>So this looks like an Italo-Celtic etymology, but there may be
>problems with this I do not see. Ray?
>
>Greetings,
>
>Jörg.
>=========================================================================
AFAIK, *inis-ula would have given Latin *inirula or *inilla (< *inirla).
Maybe the ancestral form was *inis-la:, but that probably would have given
Latin *ini:la. However, *insla: would have regularly given _insula_, I
think. If syncope operated sufficiently early, it could have prevented the
disappearance of */s/ before resonants with compensatory lengthening (a
regular process in early Latin). Hence, we could have *inis-la: 'little
_inis_' (*-la was a diminutive suffix) first becoming *insla: (syncope) and
then _insula_ (regular addition of /u/ between a medial */Cl/ cluster; cf.
medial */kl/ > */kul/ <-cul->).
- Rob
________________________________________________________________________
________________________________________________________________________
Message: 16
Date: Wed, 7 Dec 2005 13:49:45 -0700
From: Dirk Elzinga <[EMAIL PROTECTED]>
Subject: Re: What's a good isolating language to look at
On 12/7/05, tomhchappell <[EMAIL PROTECTED]> wrote:
>
> --- In [email protected], Roger Mills <[EMAIL PROTECTED]> wrote:
> >
> > Perhaps because (if I'm right) it was done/published in the Dark
> > Ages before
> > you were born. Oh you kids....:-)))) A review of period textbooks
> > might turn
> > up something; Gleason, Sapir maybe, Bloomfield??? (I'll check my
> > old copy of Hockett.)
> >
>
> Thorndike, too, apparently: -- this struck me because I read one of
> his Depression-era word-counting reports, and he has a dictionary
> named after him.
Do you have a reference for the Thorndike report? All I can find in
our library that seems to be related are the 1944 and 1921 versions of
his Teacher's Word Book.
Thanks.
Dirk
--
Gmail Warning: Watch the reply-to!
________________________________________________________________________
________________________________________________________________________
Message: 17
Date: Wed, 7 Dec 2005 21:00:49 +0000
From: R A Brown <[EMAIL PROTECTED]>
Subject: Re: Thoughts on Word building
Roger Mills wrote:
> Charlie wrote:
[snip]
> However, I'm having trouble with the
>
>>word "island" itself. _Insula_ has no known etymology.
>
>
> Speculation: possibly related to solus 'alone'?????
No, no - the 'o' of _solus_ is long, the 'u' of _insula_ is short. They
cannot possibly be connected.
The Latin is almost certainly diminutive of *ins-, which is cognate with
word for 'island' in Insular Celtic langs, cf.
Breton: enez
Cornish: enys
Welsh: ynys
Gaelic: innis
Irish: inis
Some people connect the Celtic & Latin forms also with the ancient Greek
/na:sos/ or /nE:sos / (according to dialect; the modern Greek is /nisos/).
FWIW 'island' in Speedwords is _tersol_ <-- ter (earth, land) + sol
(alone, single, sole). The 1951 dictionary gives no entry for 'peninsula'.
--
Ray
==================================
[EMAIL PROTECTED]
http://www.carolandray.plus.com
==================================
MAKE POVERTY HISTORY
________________________________________________________________________
________________________________________________________________________
Message: 18
Date: Wed, 7 Dec 2005 16:28:24 -0500
From: Paul Bennett <[EMAIL PROTECTED]>
Subject: Re: An unusual incorporation scheme
On Wed, 07 Dec 2005 07:47:29 -0500, Peter Bleackley
<[EMAIL PROTECTED]> wrote:
> Suppose we had
>
> Intransitive: Iswim
> Transitive: I readbook
> Ditransitive: I gaveMary cake
Br'ga incorporates tools and paths. I suppose that's in the same league,
isn't it, depending on how I treat certain types of verbs and nouns?
Paul
________________________________________________________________________
________________________________________________________________________
Message: 19
Date: Wed, 7 Dec 2005 16:34:18 -0500
From: Paul Bennett <[EMAIL PROTECTED]>
Subject: Re: An unusual incorporation scheme
On Wed, 07 Dec 2005 15:33:41 -0500, tomhchappell <[EMAIL PROTECTED]>
wrote:
> A new word for me!
> Is "dechticaetiative" a synonym for "primative/secundative" and the
> opposite of "direct/indirect",
Yes.
> or does it apply strictly to passive
> constructions promoting the primary object instead of the direct
> object?
No.
> If it is a synonym, it's shorter than "primative/secundative";
> (although harder to memorize the spelling, and harder -- for me -- to
> pronounce). I'll try to learn it.
/dek.t@'[EMAIL PROTECTED]/ as far as I recall. I can't find my Trask right now,
and that's a pain, because the description given there makes things really
clear.
> P.S. I think English is primative/secundative, too; I suppose that
> makes me one of those people who think English is dechticaetiative.
I think English is a language capable of expressing dative or
dechticaetiative phrases, thus both terms are IMO meaningless. What the
deep syntax is, is I rather fear, a question for the ages.
Paul
________________________________________________________________________
________________________________________________________________________
Message: 20
Date: Wed, 7 Dec 2005 21:19:52 -0000
From: tomhchappell <[EMAIL PROTECTED]>
Subject: OFFLIST: Re: What's a good isolating language to look at
--- In [email protected], Dirk Elzinga <[EMAIL PROTECTED]>
wrote:
>
> On 12/7/05, tomhchappell <[EMAIL PROTECTED]> wrote:
>> [snip]
>>
>> Thorndike, too, apparently: -- this struck me because I read one of
>> his Depression-era word-counting reports, and he has a dictionary
>> named after him.
>
> Do you have a reference for the Thorndike report? All I can find in
> our library that seems to be related are the 1944 and 1921 versions
> of his Teacher's Word Book.
The only reference in Bybee's 1985 "Morphology" of which Thorndike is
the primary author is the 1944 Thorndike & Lorge "Teacher's Word Book
of 30,000 Words".
I am no longer in even the same state, much less same town or same
university, as the library where I read the report to which I
referred in my first paragraph; it may have been the "Teacher's Word
Book", or maybe something else.
A quick look through Bybee's index at
the "inflect.." "deriv..", "fusi..", "isola..", "agglu..", "synthe.."
pages doesn't get me to the part where she mentions Thorndike. Of
course I couldn't look at all the "infle.." and "deriv.." ones -- not
enough time -- so those must be where it is. Chapter titles make me
think it might be in Chapter 4 "Lexical/Derivational/Inflectional
Continuum", and a Sapir quote is in 4.5 "difference between deri. and
infl. meaning" on p.98 (another in 4.6 "compounding & incorporation"
on p. 108), so I suspect it's in chapter 4; maybe it's before.
Too bad the habits of an "author index" and a "language index" hadn't
gained currency by 1985.
>
> Thanks.
You're welcome. Sorry that the best I could do was "the book you're
thinking of is probably it". If I find out more I'll let you know.
>
> Dirk
> --
> Gmail Warning: Watch the reply-to!
This was sent to you personally on purpose; it is not a Gmail-caused
error.
________________________________________________________________________
________________________________________________________________________
Message: 21
Date: Wed, 7 Dec 2005 17:06:46 -0500
From: Jim Henry <[EMAIL PROTECTED]>
Subject: Re: What's a good isolating language to look at
On 12/7/05, Dirk Elzinga <[EMAIL PROTECTED]> wrote:
> There is an easy way to find out which language is more
> analytic/synthetic. Take representative texts from each language,
> count up the number of words and then the number of morphs. You can
> then make a ratio of morphs/words--the "synthesis index"--for the
> languages in question. In a 1960 paper, Greenberg used this method and
> came to the conclusion that English had 1.68 morphs per word, on
> average. Vietnamese had 1.06. (Other languages that Greenberg looked
> at were Sanskrit, 2.59; Anglo-Saxon, 2.12; Persian 1.52; Yakut, 2.17;
> Swahili, 2.55; and Eskimo, 3.72. I looked at Shoshoni and my own
> conlang, Miapimoquitch, and came to 1.57 and 2.43, respectively.)
I checked the gjâ-zym-byn version of the Babel text
(not on my web site yet; I finished the second draft
translation awhile ago, but got busy with other things
before I had time to do an interlinear gloss). It has 208 words
and 359 morphemes*, so an index of synthesis of 1.726.
Like many engelangs** it's purely agglutinative; I don't have to
count to see that the agglutinativity index is 1.0.
John C. Wells, in _Lingvistikaj Aspektoj de Esperanto_,***
quotes the Greenberg article and calculates indexes
of synthesis and agglutinativity for Esperanto.
With texts from various authors he came up with
indexes of synthesis from 1.80 to 2.05 -- a lot more
variation than in the three different English texts
Greenberg analyzed. Wells found indexes of
agglutinativity of 1.0 in the texts he analyzed,
but guessed that the actual index in the corpus
as a whole is probably around 0.9999. (There
are two affixes, -cxj- and -nj-, that act fusionally
rather than agglutinatively.)
[*] - I counted each phoneme in the spacetime
postpositions as a separate morpheme -- I'm
sure that was right -- but I also counted each
conjunction as two morphemes, and
I'm not as confident of that analysis. In theory
the truth-value conjunctions are formed from
individual one-phoneme morphemes corresponding
to the rows of a truth table, but in practice only
a few of the 16 possible truth-value conjunctions
are used often enough that they become part
of the language-in-my-head as opposed to
the language-on-paper, and I don't
use that derivation subsystem productively.
I'm more confident about counting the causal
relation conjunctions as two morphemes
each, because I regularly use most of the
eight possible combinations of four prefixes
and two suffixes in that group. If I don't
count each instance of "kinq" (and) as
two morphemes but one, I get 340 morphemes
and an index of 1.63.
[**] - Or all engelangs? Ithkuil may be partly
fusional, but I haven't studied it in over a year
so I may be misremembering. I can't
think of any others that aren't agglutinating
or isolating.
[***] - UEA, 1978; second edition 1989.
--
Jim Henry
http://www.pobox.com/~jimhenry/gzb/gzb.htm
...Mind the gmail Reply-to: field
________________________________________________________________________
________________________________________________________________________
Message: 22
Date: Wed, 7 Dec 2005 17:28:46 -0500
From: Jim Henry <[EMAIL PROTECTED]>
Subject: IPA/CXS questions
I'm finally getting around to adding IPA and CXS columns to
the tables on my gzb phonology page (about time, I hear you say).
I'm stymied by two seemingly simple questions that I can't seem
to answer from the Wikipedia IPA article or Henrik's CXS page:
1. What is the correct IPA/CXS for the bilabial semivowel
that's syllable-initial in English "wait", "went", etc?
I know that or a similar sound in a falling diphthong like
is witten /U/ in CXS, ʊ in Unicode IPA; but is
it transcribed the same way initially?
2. How to transcribe the retroflex vowel (or rhoticized schwa?)
that's final in English "finger", initial in "earn"? Some transcriptions
I've seen just use schwa, which I reckon must be based
on non-rhotic dialects that lack this retroflex vowel.
--
Jim Henry
http://www.pobox.com/~jimhenry/gzb/gzb.htm
...Mind the gmail Reply-to: field
________________________________________________________________________
________________________________________________________________________
Message: 23
Date: Wed, 7 Dec 2005 17:15:27 -0500
From: Jim Henry <[EMAIL PROTECTED]>
Subject: Re: Sources of Irregularity (was: Re: Isolating, Inflected, Word
Building, ETC.)
On 12/7/05, tomhchappell <[EMAIL PROTECTED]> wrote:
> Used to be English had "eode" as the past tense of "go". (Suppletive
> and irregular already, I know, but that doesn't vitiate my point.)
>
> Used to be English had "went" as the past tense of "wend";
> regular, like "bend-bent, lend-lent, pend-pent, rend-rent, send-sent".
> (Forgive me if some of the above are errors; the point is, there was
> such a paradigm and it had several verbs in it.)
>
> But people got used to wending their way home instead of just going
> there, at least in the past tense; so "I home went" instead of "I
> home eode". (Or something like that.)
Something similar happened in French with three or more
different verbs combining to be the different stems
of "être" in various persons, tenses and moods. But I can't
recall what the specific Latin or Old French verbs were.
I think one was related to "suivre" (to follow); indeed
the first person singular present indicative active
is the same for both "être" and "suive", "je suis".
A quick Google search didn't turn up anything.
Ancient Greek used a perfect tense verb originally
meaning "to have seen", as a present tense verb
meaning "to know": "oida".
--
Jim Henry
http://www.pobox.com/~jimhenry
...Mind the gmail Reply-to: field
________________________________________________________________________
________________________________________________________________________
Message: 24
Date: Wed, 7 Dec 2005 17:43:54 -0500
From: "Mark J. Reed" <[EMAIL PROTECTED]>
Subject: Re: IPA/CXS questions
On 12/7/05, Mark J. Reed <[EMAIL PROTECTED]> wrote:
>
> 2. How to transcribe the retroflex vowel (or rhoticized schwa?)
> > that's final in English "finger", initial in "earn"? Some
> > transcriptions
> > I've seen just use schwa, which I reckon must be based
> > on non-rhotic dialects that lack this retroflex vowel.
>
>
> Correct. In rhotic dialects of English, that's regarded as a syllabic
> (vocalic) version of the consonantal "r" phoneme, so it's represented by the
> symbol for that sound with the "syllabic" diacritic - an underdot in IPA, a
> trailing = in CXS. The specific sound varies by dialect but r\= (IPA
> ɹ plus an underdot) is a popular choice.
>
Oh, I should mention, that representation alternates pretty much arbitrarily
with the use of a rhoticized schwa. That'd be CXS |@`|, IPA ɚ LATIN
SMALL LETTER SCHWA WITH HOOK.
---
Mark J. Reed <[EMAIL PROTECTED]>
[This message contained attachments]
________________________________________________________________________
________________________________________________________________________
Message: 25
Date: Wed, 7 Dec 2005 17:41:51 -0500
From: "Mark J. Reed" <[EMAIL PROTECTED]>
Subject: Re: IPA/CXS questions
On 12/7/05, Jim Henry <[EMAIL PROTECTED]> wrote:
>
>
> 1. What is the correct IPA/CXS for the bilabial semivowel
> that's syllable-initial in English "wait", "went", etc?
It's |w|.
I know that or a similar sound in a falling diphthong like
> is witten /U/ in CXS, ʊ in Unicode IPA; but is
> it transcribed the same way initially?
No, |U| is the vowel in "put", not a semivowel of any type.
2. How to transcribe the retroflex vowel (or rhoticized schwa?)
> that's final in English "finger", initial in "earn"? Some transcriptions
> I've seen just use schwa, which I reckon must be based
> on non-rhotic dialects that lack this retroflex vowel.
Correct. In rhotic dialects of English, that's regarded as a syllabic
(vocalic) version of the consonantal "r" phoneme, so it's represented by the
symbol for that sound with the "syllabic" diacritic - an underdot in IPA, a
trailing = in CXS. The specific sound varies by dialect but r\= (IPA
ɹ plus an underdot) is a popular choice.
--
> Jim Henry
> http://www.pobox.com/~jimhenry/gzb/gzb.htm
> ...Mind the gmail Reply-to: field
>
--
Mark J. Reed <[EMAIL PROTECTED]>
[This message contained attachments]
________________________________________________________________________
________________________________________________________________________
------------------------------------------------------------------------
Yahoo! Groups Links
<*> To visit your group on the web, go to:
http://groups.yahoo.com/group/conlang/
<*> To unsubscribe from this group, send an email to:
[EMAIL PROTECTED]
<*> Your use of Yahoo! Groups is subject to:
http://docs.yahoo.com/info/terms/
------------------------------------------------------------------------
