Date: Mon, 5 Jan 1998 10:58:14 -0700 (MST)
Subject: Re: EvolLang: Gordian knot about Metaphor
Sent by: John A. Halloran
>> > Dr. John E. McLaughlin and Michelle R. Sutton wrote:
>> >
>> > For example, 'write' is from an earlier stem that meant
>> > 'scratch'. Writing with a quill on parchment was like scratching
>> > in a manipulative sense (and possibly a sensational sense as
>> > well). Same goes for the Panamint word for 'write'--it's from
>> > the old verb stem for 'paint'. Writing with ink on paper is
>> > like painting.
This reminds me of the Sumerian word mu2 meaning 'to blow' and the
parallel word mu1 meaning 'to name, to speak'.
Regards,
John Halloran
------------------------------
Date: Thu, 8 Jan 1998 17:15:25 -0700 (MST)
Subject: Re: EvolLang: Metaphor
>Sent by: Sherman Wilcox <wilcox@unm.edu>
>
[snip]
>
>For signed languages, I think it is true that the root *is* quite
>often transparent. It is also true that there does seem to be an
>erosion of transparency over time. Even so, the general claim about
>the arbitrary relation between form and meaning needs to be modified
>to reflect this. Which, of course, means that it is not a general
>claim about *language* any more, only (perhaps) spoken language.
>
>The fact that we discover erosion of transparency should remind us
>that we are observing a historical PROCESS. The end result of this
>process might very well be an arbitrary relation between form and
>meaning. But what drives this process?
>
[snip of contrast between signed and spoken languages]
A loss of sound-iconicity can be observed in the vocabulary of
Sumerian as one moves from unisyllabic forms to polysyllabic forms.
It was pretty obvious that this corresponded to early generations
forming simple words and later generations combining them in order
to express more complex concepts or to describe more recent
inventions, but it helps to explicitly state as you do that there
should be an erosion of iconicity over time, an erosion that we are
closer to in the case of deaf signed language.
>As I mentioned to Marshall Pease privately, an analogy might be
>that we have picked up some rocks (though we have not been very
>careful in ...
>
[snip of analogy]
>
>The point is not that "Rock is large, round and smooth" or that
>"Language is arbitrary in its assignment of sounds to units of
>meanings." Rather (he said tentatively), the point is that that
>*people* make correspondences of mental stuff (such as meanings)
>with physical stuff (such as flapping their tongues to make sounds,
>or flapping their hands to make signs). Until we learn to
>communicate via mental telepathy, this is the constraint we live
>with: we've gotta get physical.
>
>The physical stuff has certain qualities (though, to beat this horse
>just a little bit more, if we limit our scope of study to only one
>class of physical stuff, we'll NEVER NOTICE the variable of quality:
>how would I ever notice that different substances have different
>boiling points if I only ever looked at one substance?). And the
>operation of making correspondences has certain cognitive-social-
>historical qualities. Once I make a correspondence, and you and I
>agree on it, and we repeat it over and over again in our daily
>interactions, it becomes fairly ritualized. I don't have to utter
>the entire form for you to get the meaning. This might lead to the
>form becoming stylized -- so much so that after a while, the
>relation of the form to its meaning is quite opaque.
This process of increasingly substituting the part for the whole
answers your question as to the process that drives signs from iconic
to arbitrary.
>I find it a bit perverse that we then hold up this end product and
>say, "Here is the essence of language: it is arbitrary!" [I'm not
>saying John [McLaughlin] claims this, but hey, I'm on a roll here.]
No, but Saussure did, so you appreciate that his tenet regarding the
arbitrariness of linguistic signs is for a certain moment in time
and has obfuscated understanding the earlier time when people started
linguistically signing.
>And I'd be willing to bet that a significant part of the problem of
>discovering the origins of language is that, in doing things like
>this, we turn the historical process of "languaging" into a static
>result: "language as we know it."
That would be a good bet. But it is hard to get past the mental road
block of projecting the present onto the past, to understand the
situation of people in a different place and time on their own terms.
Their situation may have been different from what we know. The
tendency to project the present onto the past is what caused certain
modern writers to see ancient astronauts in Mayan symbolism - they
didn't care enough to investigate what that symbolism meant within
that culture at that time. In the paper on the proto-Sumerian
language invention process that is at my web site, I express how
surprised I was when I realized that the proto-Sumerians had derived
meanings for their phonemes from an iconic process of making
analogies to what the mouth and tongue did to make those sounds.
And that was because we don't do anything like that today or in any
language that I knew. But if one investigates the phonetically
simplest Sumerian morphemes in depth and on their own, without being
restrained by modern preconceptions, that is the inevitable
conclusion about how they allocated meanings to their initial set
of phonemes.
Regards,
John Halloran
------------------------------
Date: Fri, 9 Jan 1998 13:25:37 -0700 (MST)
Subject: Re: EvolLang: Metaphor
>Sent by: "Peter T. Daniels" <grammatim@worldnet.att.net>
>
>OTOH maybe you *are* claiming that the Sumerian language had its own
>separate genesis many millennia after all the other human languages
>had co-evolved with humans? Well, that would certainly explain why
>we can't find any relatives for it! (I'll have to look at your
>website next week when I have time for it.)
What I think has happened is that modern thinkers have once again
made the mistake of projecting their knowledge of the present onto
the past -- of projecting modern day speech ubiquity into the past
indefinitely. Since all humans today speak, it is radical to think of
humans not speaking. It is hard for people to accept that the past
was different in such a fundamental way from the present. It is
similar to the difficulty that we have in envisioning how six thousand
years ago most humans tattooed their bodies as a visual way of
communicating to what totem-based tribe they belonged. We don't do
that today and you find very little awareness of it even by scholars.
What I think has happened with speech is that the time of its origin
has been extended back to a time that is comfortably vague, so that
linguists don't have to deal with a time and a process that was
fundamentally different from the modern linguistic processes that
they know. To study those ancient processes might tell us something
about ourselves, but would appear to have little relevance to how
language is used today.
>Subject: Re: EvolLang: Metaphor
>
>Sent by: "Dr. John E. McLaughlin and Michelle R. Sutton"
>
[snip]
>
>between meaning and sound. If this were an innovation from an
>ancient natural pattern of all roots having sound symbolic value,
>there would be far more evidence of it in modern languages and we
>would constantly see evidence of its reoccurrence in new vocabulary
>in many languages.
Not if that manner of thinking no longer exists. Once the
relationship of meaning to sound becomes arbitrary, then the rules of
iconicity that once governed how mouth gestures should point at
objects no longer apply. New words are based primarily on adaptation
of existing lexical elements which, as Sherman says, have become
stylized to the point of being iconically opaque.
>There is no evidence that I am aware of that roots are formed by
>nonarbitrary processes. If this were a natural process, then
>nonarbitrariness would still be observable.
The single vowel words and single consonant words of Sumerian are
like a time machine reaching back into the past. Without that, we
would have little evidence for the nonarbitrary nature of early word
formation.
Regards,
John Halloran
------------------------------
Date: Sun, 11 Jan 1998 12:41:04 -0700 (MST)
Subject: Re: EvolLang: Metaphor
>Sent by: "Ronald Kephart" <rkephart@osprey.unf.edu>
>
>> ... the difficulty that we have in envisioning how six thousand
>> years ago most humans tattooed their bodies as a visual way of
>> communicating to what totem-based tribe they belonged...
>
>Are we sure about this? And, how do we know?
Claude Levi-Strauss is one scholar who discusses evidence for this
from the Americas, Australia, and Africa. The existence of this
practice in Europe is indicated by the tattoos on the 'Ice-Man' of
the Alps. In Mesopotamia, there are marks on Ubaid period clay
figures which appear to represent this practice.
>> ...The single vowel words and single consonant words of Sumerian
>> are like a time machine reaching back into the past...
>
>Are these *words* or *morphemes* that consist of only a single
>consonant?
A single consonant plus a vowel, either VC, CV, or VCV. Examples for
the /b/ and /p/ [represents aspirated /b/] phonemes are:
AB1: window; niche.
AB2: domestic cow.
UB1: corner, angle, nook; room
UB3: a drum.
UB4: cavity, hole; pitfall.
IB2, EB2: n., middle; waist; loins; thighs.
BA1: n., share, portion; rations, wages;
BA1: a shelled creature (such as a turtle or snail) (cf.,
Civil, Lexical Archives, PSD updates).
BA3: liver; omen.
PA1: sprout, bud
PA4,5,6: irrigation ditch, small canal.
PU2: well, cistern, pool.
>And how are these like a time machine, given that Sumerian itself
>is (was) a modern language? After all, English has some
>single-consonant and single-vowel *morphemes*.
These single-consonant words have common themes of meaning which
can be derived from the mouth-gestures that produce them. The analysis
for the examples above was given in a message to the list on November
25, 1997. If you blow out your cheeks like the trumpet player Dizzy
Gillespie while making the /b/ sound you will see how it could be an
icon for 'container'. Sumerian was an agglutinative, compounding
language. The words that describe Chalcolithic and Bronze Age
cultural inventions are polysyllabic words, often analyzable as
compounds. The single consonant words taken all together describe an
early Neolithic environment.
>Sent by: "Dr. John E. McLaughlin and Michelle R. Sutton"
>
>Are you actually trying to tell us that "Proto-Sumerian" is actually
>"Proto-Human"? Please say it ain't so :-). I looked at your web
>site, but I'm unconvinced, especially when you use phrases like,
>"The Sumerians Invented Their Language"!
When I see that the Sumerians had vowel-only words for important
concepts like 'water', 'food', 'sleep', 'intercourse', and 'house',
it makes me think that when they started the game of speech, they
did not have the idea of consonants. Then when they did start using
consonants, those concepts had already been mapped and did not need
to be remapped. I am not saying that the Sumerians necessarily had
the first spoken language, but I think that when they created their
language, they had not been exposed to other speakers. And I think
that what spread around the earth was not the actual elements of a
particular language, but the concept of spoken language.
[snip]
>now." I would much rather believe that modern humans and the way
>they use language is fundamentally similar to the way that they
>have used language for at least the last 50,000 years or so (because
>all human languages have some basic similarities that probably
>derive from a common ancestor, and the common ancestor must be at
>least that old to keep the Australians in the fold). I also agree
>with Peter (in another post) that you seem to be talking about
>Sumerian WRITING, not Sumerian SPEECH. These things are different.
What do you think are those basic similarities? For your argument,
should you exclude elements that are present in deaf sign-language?
Can the similarities that you find be explained by how the human mind
thinks?
It must be frustrating to work with American languages in which most
words cannot be analyzed, and then have some guy say that in his
language the words can be analyzed down to a much lower level. And
yes I do know the difference between Sumerian writing and Sumerian
speech (as far as we have been able to recover it from the writing).
Sumerologists have devoted much effort to extracting the actual
'logograms' of Sumerian speech from the signs of written cuneiform.
>Sent by: "Peter T. Daniels" <grammatim@worldnet.att.net>
>
>We have quite a bit of secure Common Semitic vocabulary (because the
>time depth is so shallow--Juris Zarins, a Gelb student shortly
>before me but an archeologist, thinks he can identify the culture
>that spoke CS in northwest Arabia ca. 6000-4000 BCE),
Proto-Afrasian (the ancestor of the Semitic family) must be older,
but like proto-Sumerian it was still an early Neolithic phenomenon.
For everyone's information, I have compared the vocabularies of
proto-Afrasian and proto-Sumerian - there is no connection between
the two, as one would expect given the geographical distance involved.
Regards,
John Halloran
------------------------------
Date: Tue, 13 Jan 1998 00:04:44 -0700 (MST)
Subject: Re: EvolLang: Metaphor
>Sent by: "Peter T. Daniels" <grammatim@worldnet.att.net>
>
>[My quote was in connection with the quantity of vocabulary
>available for doing Woerter und Sachen, for which terms relating
>specifically to agriculture, such as plow, yoke, hoe, etc., do not
>seem to be reconstructable. It has nothing to do with the alleged
>independent origin of Sumerian -- which no one has ever suggested
>was related to Semitic at any level other than proto-Human.]
There are people who speculate that Afroasiatic, Sumerian, and
Indo-European derived from a common Nostratic language family. If
one goes back to the simplest vocabularies of these languages,
especially proto-Afroasiatic and proto-Sumerian [proto-Indoeuropean
is at least two or three thousand years later] however, one finds no
congruence, so any common vocabulary that these people can attest is
either coincidence or borrowing through their having lived together
for millenia in the Fertile Crescent where we have evidence of much
cultural borrowing.
>Diakonoff's term "Afrasian" hasn't really caught on to replace
>Afroasiatic.
That may be, but I own a 1995 book by a UCLA linguist, Christopher
Ehret, called Reconstructing Proto-Afroasiatic (Proto-Afrasian):
Vowels, Tone, Consonants, and Vocabulary, in which he expresses his
own preference for the term Afrasian.
>You have not addressed my query as to why the Semitic-speakers,
>whose languages -- Akkadian and Amorite -- are fully modern, and
>whose ancestor, Afroasiatic, was maybe 10,000 years older than
>Common Semitic or Sumerian and seems also to have been fully modern,
>were able to learn and use Sumerian, an allegedly "primitive"
>language.
That Sumerian is primitive is not my claim. Indeed, it shows a
history of long evolution, if one can judge by the fact that in her
book on typology, Linguistic Diversity in Space and Time, Johanna
Nichols judged Sumerian to fall in the highest complexity level.
(page 193). My claim is that if one groups the Sumerian vocabulary
according to the phonetic structure of the words, the words with the
simplest structures are remnants of early stages in the language's
invention.
I don't know anyone who has studied the vocabulary of
proto-Afroasiatic who would agree with a date as early as 13,000 BCE.
Regards,
John Halloran
------------------------------
Date: Tue, 13 Jan 1998 00:04:44 -0700 (MST)
Subject: Re: EvolLang: Metaphor
>Sent by: "Dr. John E. McLaughlin and Michelle R. Sutton"
>
>> A single consonant plus a vowel, either VC, CV, or VCV. Examples
>> for the /b/ and /p/ [represents aspirated /b/] phonemes are:
>>
>> AB1: window; niche.
>> AB2: domestic cow.
>> UB1: corner, angle, nook; room
>> UB3: a drum.
>> UB4: cavity, hole; pitfall.
>> IB2, EB2: n., middle; waist; loins; thighs.
>>
>> BA1: n., share, portion; rations, wages;
>> BA1: a shelled creature (such as a turtle or snail) (cf.,
>> Civil, Lexical Archives, PSD updates).
>> BA3: liver; omen.
>>
>> PA1: sprout, bud
>> PA4,5,6: irrigation ditch, small canal.
>> PU2: well, cistern, pool.
>
>These are not "single-consonant" words, then. They are monosyllabic
>words.
I use the term single-consonant to distinguish these words from the
monosyllabic CVC type of word, a common type which makes up about
30% of Sumerian logograms. The referents of the CVC words are more
complex, specific, and sophisticated than are the referents of the
words with simpler phonetic structures.
The Sumerian vocabulary is interesting because of the way that they
appear to have exhausted the phonetic combinations possible with one
structure before switching to a more complex structure for additional
lexemes. Part of the reason for this pattern in Sumerian, and its
evident lack in other languages [proto-Afroasiatic is somewhat of an
early exception with its regular CVC pattern], may be that the
inventors of other languages already had the examples of different
word structures from exposure to already developed seed languages, so
they started from higher levels.
[snip]
>'eat' and 'drink'? It really isn't at all unusual to find
>monosyllabic morphemes for some of these notions in many languages,
>and even vowel-only morphemes. High frequency words tend to be
>monosyllables in most languages. Statistically, there is probably a
>high ratio of the world's languages where these high-frequency
>morphemes are vowel-only.
Probably? Which language that you know presents the closest
approximation to Sumerian's use of vowel-only words to represent
basic noun concepts, preferably involving food, drink, shelter, rest,
and sex?
I don't think that the phonetically simple words of Sumerian got to
be that way because of high use - they are constituent elements in
more complex compounds that would have been formed not long after
their own constituents' creations.
[snip]
>Try the basic functional categories of noun, verb, adjective, adverb.
>All human languages have these four elements in their language. The
>basic inverse relationship between morphology and syntax. The same
>basic set of relationships vis a vis time and completion of action
>- -- completed versus uncompleted and real versus unreal to name a
>couple. These are just off the top of my head at the moment.
>
>> For your argument, should you exclude elements that are present
>> in deaf sign-language?
>
> From my understanding, most signed languages have these features as well.
Then how can you use these common features as evidence that all
spoken languages must be descended from an Ur-language that existed
50,000 years ago? Do sign languages only make these distinctions
because they were created by users of spoken language? Can you think
of a useful language that makes different distinctions? If not, then
are we indebted to a genius the like of whom the earth has not seen
since for the first speech complete with all of these distinctions?
That must have been hard, teaching such a fully-formed language to
people who had never spoken before. Or would you admit that any
population starting to speak could gradually evolve these categories?
>For the record, I've studied Indo-European, Finno-Ugric, Semitic,
>Bantu, Uto-Aztecan, Caddoan, Penutian, and Wakashan languages and
>they all exhibit a majority of unanalyzable bi- or polysyllabic
>roots.
That was my impression also. Sumerian is different in that many of
its words are analyzable compounds. Even the simple noncompounds can
be analyzed as iconic mouth gestures.
John Halloran
------------------------------
Date: Wed, 14 Jan 1998 20:19:00 -0700 (MST)
Subject: Re: Re: EvolLang: Metaphor
>Sent by: "Ronald Kephart" <rkephart@osprey.unf.edu>
>
>> ...If you look at a survey like Anatole V. Lyovin, An
>> Introduction to the Languages of the World (1997), it is clear
>> that there are many language families that simply are not
>> genetically related...
>
[snip]
>
>I suppose we *could* argue that different language families,
>although having distinct, unique origins, came up with similar
>categories due to the similar contextual experiences of the peoples
>who created them. It seems more likely to me that, since modern
>Homo sapiens appear to share a unique origin, the human languages
>they speak do also.
The early Jomon culture invented pottery in southwestern Japan
around 12,500 B.P.
The earliest occurrence of pottery found in the Near East is at
Ganj Dareh in Persia/Iran around 10,000 B.P.
A reasonable person would say that the modern Homo sapiens at these
far-removed sites invented pottery independently, that pottery does
not 'share a unique origin'. Why must language be different?
[snip]
>>
>> ...I am not saying that the Sumerians necessarily had the first
>> spoken language, but I think that when they created their
>> language, they had not been exposed to other speakers...
>
>I don't think people could create a language ex nihilo, and in any
>case the Sumerians did not exist totally cut off from speakers of
>other languages (did they?). This goes back to debates we've had
>on Anthro-L about 'language as a human invention'. My argument is
>that to invent something like language, people would have already
>had to have a language to do it in. How else could they do it?
So are you saying that humans did not evolve from nonspeaking
primates? Is that your position?
The evolution of symbolic counting tokens in the Near East,
including at Ganj Dareh, from counting pebbles, may have provided
the jumping off point that humans needed in order to conceive of a
collection of symbolic icons to represent significant objects in
their world. You are right that in studying the ancient world one
never finds anything to arise 'ex nihilo' - if one searches hard
enough, one always finds that a pre-existing form led to the new
form. There are pictures of some of these tokens together with a
commentary at my web site.
Regards,
John Halloran
------------------------------
Date: Thu, 15 Jan 1998 18:13:22 -0700 (MST)
Subject: Re: EvolLang: Metaphor
> Sent by: "Peter T. Daniels" <grammatim@worldnet.att.net>
>
>> A reasonable person would say that the modern Homo sapiens at
>> these far-removed sites invented pottery independently, that
>> pottery does not 'share a unique origin'. Why must language be
>> different?
>
>Because language is the product of evolution, not of ingenuity.
That sounds like an article of faith. If language is not the
product of intelligence, why have less intelligent animals not
evolved language?
Language cannot influence the evolution of the brain before
language exists. Language can only be a factor in evolutionary
selection after speakers have already invented a rudimentary form
of language. And we are not talking about hard-coded vocalizations,
but a set of symbols that varies widely and requires plastic
intelligence in order to learn.
>Are you now claiming that people used pebbles and then symbolic
>tokens to count ... before they had number words?
Think about it. Pebbles are more concrete than spoken words.
Counting with words is relatively abstract compared to counting
with pebbles. And the archaeological record does show that humans
throughout the Near East started counting with clay pebbles shaped
in different ways to symbolize different goods.
Regards,
John Halloran
------------------------------
Date: Thu, 15 Jan 1998 18:13:32 -0700 (MST)
Subject: Re: EvolLang: Metaphor
>Sent by: Greg Laden <gladen@earthlink.net>
>
>1) Sumerian is a creole. What would a creole look like in writing?
>It would look "new" in some ways, right? But in what ways?
This is the only real possibility of the ones that you mention.
There is a recent monograph by Jens Hoyrup which proposes this to
be the case, but I remember that I had a number of problems with his
facts and conclusions when I read it. If someone here wanted to
propose evidence for this, I could evaluate it and respond. To
start, I will just express my doubt that the phonetically simplest
words of a creole language would show clusters of meanings which are
iconically related to the phonemes' manner of articulation. Since
Jess Tauber has studied this in so many languages, perhaps he could
comment on the phonemic symbolisms found in creoles.
Don't focus too much on the writing. There is a perfectly good
language behind the writing, which one can learn by studying with
one of the twenty or so working Sumerologists in the world.
Alternately, there is a perfectly good grammar of the language in
the book by Marie-Louise Thomsen, referenced in the Lexicon
Bibliography at my web site.
Regards,
John Halloran
------------------------------
Date: Fri, 16 Jan 1998 11:57:28 -0700 (MST)
Subject: Re: EvolLang: Gene Trees, Lang Trees
>Sent by: "Peter T. Daniels" <grammatim@worldnet.att.net>
>
>And those are exactly the most illegitimate ones. There IS NO
>Amerind, much less a Eurasiatic incorporating it; there IS NO
>Dene-Caucasian. There PROBABLY IS NO Austric (because the
>materials are even less amenable to comparison than those in
>other phyla). There IS NO Indo-Pacific. (I don't have the chart
>in front of me and it's distasteful even to open Ruhlen's book
>to see the other Macros.)
>
[speaking of Cavalli-Sforza and Merritt Ruhlen]
>
>Fine ... but that's not the point. They are claiming WORLD-WIDE
>agreement, and the right side of the chart (the language side)
>is entirely fantasy.
Listen to Peter. He is saying what the specialists know.
Regards,
John Halloran
------------------------------
Date: Fri, 16 Jan 1998 11:57:27 -0700 (MST)
Subject: EvolLang: Re: Unspeaking tribes
>Sent by: Greg Laden <gladen@earthlink.net>
>
>>The Least Moves Hypothesis must be considered in context. The
>>context in this case was that the Fertile Crescent regions to the
>>north and east of Palestine were inhabited by peoples with equally
>>developed cultures into which the PAA language could expand only
>>with difficulty. By contrast, the regions of northeast Africa were
>>culturally undeveloped and, dare I say it, may have been inhabited
>>by unspeaking tribes eager to learn bits and pieces of the PAA way
>>of communicating, but having to invent many lexical items on their
>>own. As far as mass migration goes, the archaeological record does
>>show that Palestine became deserted of settlements at this time.
>
>Unspeaking tribes? Please elaborate!
What I will say is that comparative linguists are confronted in many
areas with confusing situations where languages appear to be related,
but only imperfectly, such as when they share considerable
morphology, but not basic lexical items. This is the case within the
Altaic 'family' and even between Uralic and 'Altaic'. If these
languages share morphology because they are descended from a common
ancestor, their lexicons should also show a common origin, but they
do not. It is hard for comparative linguists to understand how
speakers of neighboring languages, each with fully developed
morphologies, could abandon their own morphology in favor of that
of their neighbor, especially since the morphology of a language is
seen as more fundamental than the items of its lexicon.
When morphological and lexical similarities are sufficient to group
languages into larger language families, we should not take the
biological metaphor of a family literally and think that those
languages are necessarily descended from a common ancestor. Rather,
we need to be open to the possibility that an unspeaking population
acquired language elements from a genetically unrelated speaking
population and innovated the rest of a new language on their own.
Regards,
John Halloran
------------------------------
Date: Fri, 16 Jan 1998 13:32:37 -0700 (MST)
Subject: Re: EvolLang: Punctuated equilibrium
>Sent by: "Peter T. Daniels" <grammatim@worldnet.att.net>
>
[snip]
>
>Of course the tree-structure par excellence is Austronesian, where
>there were successive breakups with virtually no subsequent contact.
>You can actually see shared innovations tracking successively all
>the way across the Pacific!
The tree-structure conforms so nicely to the genetic descent model
in this case because of the late date of human spread across the
Pacific, so that in this case agriculture and language traveled
together with the human colonists, instead of diffusing later as a
cultural wave.
I quote from the long article in the Encyclopedia Brittanica CD by
Andrew K. Pawley.
"it is clear that the Proto-Austronesian speech community possessed
agriculture and may have been responsible for its introduction --
along with that of several other important cultural innovations --
into the Pacific"
"The dispersal of Austronesian languages in Oceania cannot have
begun later than around 2000 BC, with 3000 BC appearing to be a more
realistic estimate."
>From the Grolier electronic encyclopedia there is the following
statement under Diffusion, Cultural, which describes how the
concepts, but not the details, of cultural inventions such as
language can spread as a stimulus wave.
"Many examples of cultural diffusion result from what is called
stimulus diffusion, whereby the idea or principle behind a
particular culture trait is diffused even though the culture trait
itself is not adopted. Stimulus diffusion must have played an
important role in the spread of agriculture and animal domestication
throughout the prehistoric world. It is likely that nomadic Siberian
tribes knew of domestic horses and cattle before they
domesticated the reindeer."
Regards,
John Halloran
------------------------------
Date: Fri, 16 Jan 1998 23:51:51 -0700 (MST)
Subject: Re: EvolLang: Re: Unspeaking tribes
>Sent by: Greg Laden <gladen@earthlink.net>
>
>>What I will say is that comparative linguists are confronted in
>>many areas with confusing situations where languages appear to
>>be related, but only imperfectly, such as when they share
>>considerable morphology, but not basic lexical items. ...
>>It is hard for comparative linguists to understand how
>>speakers of neighboring languages, each with fully developed
>>morphologies, could abandon their own morphology in favor of that
>>of their neighbor, especially since the morphology of a language
>>is seen as more fundamental than the items of its lexicon.
>
>Is this like the "Bantuizatin" of Central Sudanic languages? For
>example, Bila (or Bira), a language spoken in the Epulu area of
>the Ituri Forest has a Bantu morphology, but much of the lexicon
>is Central Sudanic, sharing many terms with neighboring Central
>Sudanic speakers. The Bila themselves are culturally more Central
>Sudanic than they are Bantu (maybe physically as well). Nearby,
>other groups have a Central Sudanic morphology, but there are many
>Bantu features found among the morphology. According to a Belgian
>linguist I once met who was studying this, this has been a
>progressive pattern over time whereby the Central Sudanic
>languages have changed into Bantu languages for the last couple of
>hundred years or so.
>
>Perhaps sadly, my understanding is that Central Sudanic language
>structure is a bit unusual and rich in many features that make
>linguists have wet dreams. I'm told this includes a large number
>of tenses, including some highly specialized (like, specific time
>periods including the first few days of the universe, and the last
>day of the universe) and odd things like plural verbs. The Bantu
>structure is, of course, what is found in the main trade languages
>in the area (KiNgana, a dialect of KiSwahili and Lingala).
Thank you for adding this last part. There are probably many reasons
for adaptations that look like language convergence, but one of the
strongest will always be to increase the efficiency of a language.
Languages do evolve over time in order to become more efficient tools
for their users. In this case, a language with an inefficient
morphology is gradually adopting the more efficient morphological
features of a neighboring language.
>>... Rather, we need to be open to the possibility that an
>>unspeaking population acquired language elements from a genetically
>>unrelated speaking population and innovated the rest of a new
>>language on their own.
>
>But, what is an unspeaking population? I want to be sure I
>understand. Do you mean groups (historically, I assume) of people
>sans language? Or is this a linguistic term I need to look up in
>my trusty Cambridge Encyclopedia of Language?
I certainly don't mean without vocalizations. I do mean without
spoken language as a set of culturally learned symbolic mappings.
As an anthropologist, you may like to comment on my belief that a
set of communally agreed symbolic vocalizations became especially
important when people began living in settled communities which
exceeded the under 40 count that I understand characterized humans
still living an unsettled tribal existence. I understand that
skeletons from the first cities show a lot of skull and other
injuries from interpersonal disputes. People who know each
other intimately know each other's idiosyncratic signals, but
relative strangers need a societally imposed standard symbol system
in order to communicate and avoid misunderstandings. Also, as the
technological horizon advances, there are more objects about which
to communicate.
So what do you think specifically about my belief that there was a
greater need for a culturally learned spoken language as groups
exceeded 40 individuals, as was the case in the first cities?
Regards,
John Halloran
------------------------------
Date: Sat, 17 Jan 1998 21:58:19 -0700 (MST)
Subject: Re: EvolLang: Metaphor
>Sent by: Dan Moonhawk Alford <dalford@haywire.csuhayward.edu>
>
>> >And how are these like a time machine, given that Sumerian
>> >itself is (was) a modern language? After all, English has some
>> >single-consonant and single-vowel *morphemes*.
>>
>> These single-consonant words have common themes of meaning which
>> can be derived from the mouth-gestures that produce them.
>
>I'd like to slow this sentence down, because I seem to remember it
>coming up regarding Sanskrit -- *another* language that claims to
>begin with a writing system rather than speaking.
Sanskrit may be a written language today, but originally it was a
spoken language descended from the Indo-Iranian branch of
Proto-Indo-European. Likewise, Sumerian may have become the classical
written language of Babylonia, but it originated in my opinion
between 10,000 and 9,000 years BP as a spoken language until it
ceased to be spoken around 3800 BP. The Sumerians did not invent
writing until around 5400 BP. As far as the mouth gestures go, I
have already discussed on the list the symbolism of the phoneme /b/.
I am talking about the kind of symbolism that makes the Sumerian
word /ig/ mean 'door, entrance' - one closes the entrance to the
throat in order to make this sound, so the physical process of
saying the word is an icon for its meaning. Another example is the
Sumerian word /an/ which means 'sky; to be high' in which one points
to the sky by touching one's tongue up to the alveolar ridge of a
wide open mouth.
>> When I see that the Sumerians had vowel-only words for important
>> concepts like 'water', 'food', 'sleep', 'intercourse', and
>> 'house', it makes me think that when they started the game of
>> speech, they did not have the idea of consonants. Then when they
>> did start using consonants, those concepts had already been mapped
>> and did not need to be remapped. I am not saying that the
>> Sumerians necessarily had the first spoken language, but I think
>> that when they created their language, they had not been exposed
>> to other speakers. And I think that what spread around the earth
>> was not the actual elements of a particular language, but the
>> concept of spoken language.
>
>Interesting hypothesis, perhaps going back to the larynx dropping
>into place so we could make consonants?
I would not make that connection. The larynx dropped much earlier
in human evolution. I connect the dropping of the larynx to the
way that it created a larger resonating chamber for long-distance
vocalizations necessitated by the widening range of human foraging
and hunting compared to that of other primates.
>> It must be frustrating to work with American languages in which
>> most words cannot be analyzed, and then have some guy say that in
>> his language the words can be analyzed down to a much lower level.
>
>This is the position I feel in when I talk about how words in
>entirely different languages in Native America are made up from the
>same exact morphemes said differently in different language, when I
>find a common factor to be Native American Sign Language, which
>contains the same exact morphemes in sign. (E.g., "God" from "big"
>+ "medicine" + "upward".)
Does that indicate to you that the universal sign language preceded
the spoken words in those languages that are compounds of those
morphemes? How basic do such words get? If the answer is 'very
basic', meaning that even primitive people would need those words to
communicate, then do the languages have alternative expressions for
those concepts? What do scholars know of the history of Native
American Sign Language?
Regards,
John Halloran
------------------------------
Date: Sun, 18 Jan 1998 18:28:39 -0700 (MST)
Subject: Re: EvolLang: Re: Unspeaking tribes
>Sent by: "Ronald Kephart" <rkephart@osprey.unf.edu>
>
>>In message <199801170651.XAA01882@smtp01.primenet.com> writes:
>>
>> ...Languages do evolve over time in order to become more
>> efficient tools for their users. In this case, a language with an
>> inefficient morphology is gradually adopting the more efficient
>> morphological features of a neighboring language...
>
>Is there evidence that increased efficiency is, in fact, a driving
>force in language change? What would an "inefficient morphology"
>look like?
[snip]
Here we are really talking about the subject of this list, Language
Evolution, if one means by evolution Selection of the Fittest. As
far as I know, Bernard H. Bichakjian has done the most research into
diachronic evolution/selection of features in specific languages. He
has also researched a gradual process whereby there is selection for
phonemes that are produced at an earlier age in human ontogenesis. I
don't think he would mind if I quoted the abstract of a recent paper
entitled, Evolution: From Biology to Language from Darmst dter Beitr
ge zur Naturgeschichte, Heft 6, 153-163, Darmstadt, 15. Oktober 1996.
"The main features of the hominid evolution are known. It was
essentially produced by a neotenous process, whereby the juvenile
features, which on the whole are advantageous, were preserved and
expanded, while the corresponding senile attributes were gradually
eliminated.
In mainstream linguistics, evolution is taboo, and yet language
is not a discrete instrument that sprung full-blown and remained
frozen in its alleged plenitude. Language is a technique that has
developed and goes on expanding and improving.
This paper presents evidence showing the existence of
developmental patterns in linguistics and argues that the
evolutionary process can only be uncovered and understood when the
observation includes the neurological events that underlie the
acquisition of linguistic features. The combined observation of the
linguistic data and their neurological interface suggests that
language evolution is also a neotenous process that provides its
speakers with important selective advantages."
While acknowledging that not all linguistic change is developmental,
he points out in his introduction that linguists always insisting on
using the word change instead of the word evolution is like always
saying that President Kennedy died in Dallas and never saying that
he was murdered in Dallas. One loses important information.
Regards,
John Halloran
------------------------------
Date: Mon, 19 Jan 1998 09:49:12 -0800 (PST)
Subject: Re: Re: EvolLang: Metaphor
> On Wed, 14 Jan 1998, John A. Halloran wrote:
>
> So are you saying that humans did not evolve from nonspeaking
> primates? Is that your position?
>
> The evolution of symbolic counting tokens in the Near East,
> including at Ganj Dareh, from counting pebbles, may have provided
> the jumping off point that humans needed in order to conceive of a
> collection of symbolic icons to represent significant objects in
> their world.
I don't know if y'all've seen it or not, but the Nova program
entitled "Can Chimps Talk?" has a FASCINATING bit about numeracy
which I believe goes to your point.
In it, the chimps are "playing a game" whose single rule seems to be
beyond them at first. A chimp is presented with two plates with,
say, 5 M&Ms and 3 M&Ms, and allowed to pick one; whatever the chimp
picks is given to the other chimp (that's the one rule), and the
picking chimp gets the plate he didn't pick. They always pick the
one with the most, which then goes to the other chimp, to his/her
dismay.
Then someone got a bright idea and substituted numbers on a card
in the plates for the real stuff; they had already learned numbers
and counting. When presented with the new task, something new
happened: the chimp had a kind of AHA experience and picked the
lower number, getting therefore the higher number of M&Ms! All of
a sudden, the cultural rule was clear!
But the moment they switched the task back to the old system, with
actual M&Ms, the physical appetite again replaced the cultural rule,
and the chimps invariably picked the larger amount again, which was
given away, and therefore got the smaller amount.
They switched back & forth between M&Ms and numbers, and the results
were totally consistent each time: actual candy, higher; numbers,
lower. For whatever reason, the cultural rule could not be held "in
mind" when presented with the food, and could only be accessed when
numbers were used.
This episode taught me something very important about the role of
numeracy in culture. It might have been the "kickstart" you're
looking for, where cultural games and rules could come into play
for the first time.
warm regards, moonhawk
------------------------------
>> From: John A. Halloran
>>
>> I don't think there is anything wrong with acknowledging that
>> language evolves to become a more efficient communication medium.
>> If that is controversial, then additional scholars will simply
>> have to ask if greater efficiency or other causes account for the
>> individual changes that they witness. Generalizing to claim that
>> certain language features or types are more efficient than are
>> other types might be controversial because it would then brand
>> languages that lacked those features as primitive or somehow less
>> evolved.
>
>What is efficiency in language? Brevity? Precision? My intuition
>is that we bring in new expressions for impact, for charm, for
>fun and variety, to refer to new things... We're just as liable
>to introduce a verbose expression as a terse one, just as liable
>to push the limits of grammatical propriety as to respect and
>refine it. Should we expect speakers of other languages to be
>different?
Efficiency has already won out almost hands down.
There is high degree of inverse correlation between frequency of
usage of a word and its length. It's called Zipf's law and
quantitative results relating to it can be seen in the last few
issues of the Journal of Quantitative Linguistics as well as in
other places.
Regards,
Aslan Kaplanov
------------------------------
Date: Thu, 22 Jan 1998 12:13:40 -0700 (MST)
Subject: Re: EvolLang: Metaphor
>Sent by: "Dr. John E. McLaughlin and Michelle R. Sutton"
>
>> <snip> I am talking about the kind of symbolism that makes the
>> Sumerian word /ig/ mean 'door, entrance' - one closes the
>> entrance to the throat in order to make this sound, so the
>> physical process of saying the word is an icon for its meaning.
>
>What is so special about [g] in this regard? What about [d] in
>English 'door', closing the entrance to the oral cavity? What
>about [m] in Comanche muhju, closing the entrance to the mouth?
These are structurally complex words which date from a time when
iconicity between mouth-gesture and referent was no longer a
conscious process.
[snip]
>I can't see anyone's tongue operating when I'm talking to them,
>so the idea that the tongue movements somehow correlate with
>meaning (except in a few well-documented cases like blow and puff
>and [b] or [p]), is fishy.
Marshall Pease already responded to this, saying " There's a
theory that we *interpret* gesture by "feeling" it on our
proprioceptive center, "as if" on our own body."
>(Take 'steal' for example, if you can describe in reasonable
>metaphor or analogy how the Sumerian written "phonetics" correlate
>to the meaning, then you might have even an outside chance of
>convincing me that the sound-meaning relationship is not 99%
>arbitrary.)
That concept does not appear to be central or basic enough to be
included in the 188 V, VC, and CV words of the proto-Sumerian
vocabulary. The words that mean 'to steal', like ZUH, KAR, and
LILIB are derivative, with ZEH meaning 'female goat-kid', something
valuable that was easily stolen, KAR also meaning 'to abduct' and
'quay, harbor', and LUL meaning 'to lie, deceive'.
The word DAB meaning 'to take, hold, receive' can be analyzed
similar to the cognate word DUB, 'to store', in that the component
elements mean 'to go' + 'container' (cf., BAD, 'to open, release'
from 'container' + 'to go').
Regards,
John Halloran
------------------------------
Date: Thu, 22 Jan 1998 12:13:39 -0700 (MST)
Subject: Re: EvolLang: Re: Unspeaking tribes
>Sent by: kaplanov@juno.com (Aslan Kaplanov)
>
>>Sent by: mcv@wxs.nl (Miguel Carrasquer Vidal)
>>
>>I have not read this paper, but I can offer some comments on
>>Bernard H. Bichakjian's contribution to the recent collection
>>"Archaeology and Language I" (eds. Roger Blench and Matthew
>>Spriggs; Routledge, 1997), entitled "Evolution and the biological
>>correlates of linguistic features". In this paper, Bichakjian
>>sets out his thoughts on the "evolution" of language, as it
>>manifests itself through what he describes as "unidirectional
>>changes", providing "selective advantage", in the linguistical
>>historical record. Some of these changes are:
>
[snip]
>
>>1. Laryngeals replaced by long vowels, in turn replaced by
>>different vowel qualities.
>
>How did humanity get laryngeals if long vowels are the greatest
>thing in the past few millenia?
>
>>2. Complex stop consonsants (e.g. aspirates, ejectives) replaced
>>by fricatives.
>
>Why would humans invent complex consonants at all if it is so much
>easier to invent fricatives?
Mary LeCron Foster could answer those questions with her theory of
meaningful phonemes (or 'phememes') in early Indo-European. These
are explained in most detail in "The Symbolic Structure of
Primordial Language," in Human Evolution: Biosocial Perspectives,
ed. S.L. Washburn and E.R. McCown (Perspectives on Human Evolution,
vol. IV, Menlo Park, 1978), pp. 77-121. When one tries to map
different types of motions/actions to phonemic gestures, one ends
up with some pretty complicated phonemes to express a range of
complicated motions. The phonetic simplification process on which
Bichakjian commented for Indo-European is actually a by-product of
the shift from iconic mouth-gestures to words as phonetically
arbitrary symbols.
Edwin G. Pulleyblank has also discussed how in pre-language, sound
units were originally also units of sense ("The Beginnings of
Duality of Patterning in Language" in Eric de Grolier (ed.),
Glossogenetics: The Origin and Evolution of Language (1983), pp.
369-410.) Regarding Kabardian, he says that it has 48 consonantal
phonemes, 52 biconsonantal clusters, and 7 triconsonantal clusters,
which with the addition of /a/ gives 193 monosyllabic units or
segments, almost all of which are morphemes in their vocabulary.
He says, "As Kuipers notes, the inaccessibility of the region in
which they have lived since time immemorial has provided conditions
that could have led to unusual conservatism." He says that when
one reconstructs both Indo-European and Sino-Tibetan, their
earliest forms show similar structural features.
Regards,
John Halloran
------------------------------
Date: Thu, 22 Jan 1998 12:13:40 -0700 (MST)
Subject: Re: EvolLang: Re: organic & linguistic evolution
>Sent by: "Larry Trask" <larryt@cogs.susx.ac.uk>
>
>> If a larger phonemic repertoire allows one to communicate with a
>> wider set of structurally short words because it increases the
>> possible number of CVC phoneme combinations, that could also be
>> said to make language a more efficient tool.
>
>Hardly. Hawaiian has eight consonants and no CVC sequences at all.
>One of the Khoisan languages has about 114 consonants, many of them
>very highly marked and difficult for outsiders to learn, and many
>of them extremely similar to one another and distinguished only by
>subtle and difficult articulatory maneuvers. Even if this permits
>a huge number of CVC sequences, how does that make the Khoisan
>language "more efficient"? Hawaiian is certainly vastly easier to
>learn.
Good contrast. The history of these peoples is relevant. Hawaiian
comes at the tail end of an evolutionary process of simplification
occasioned by the island hopping and spread through the Pacific.
The children of a small boatload of travelers can be expected to
simplify their parent's speech. As far as I know, the speakers of
Khoisan have always lived in the same place. My knowledge of the
special phonemes of Khoisan is limited to secondary literature,
such as the comments of Roman Stopa that different clicks serve
as determinatives or adjectives marking words, such as the dental
click /for long, extended objects or the labial click O
accompanying agreeable emotions ("Supposed First Words of Apeman"
in Eric de Grolier (ed.), Glossogenetics: The Origin and Evolution
of Language (1983), pp. 491-512.) So Hawaiian is much easier to
learn, but Khoisan communicates more information. In this case,
I am afraid that Bichakjian is right about the evolutionary
direction of language phonology in that Khoisan is a language
maintained by a stable group of mature adults and Hawaiian as a
language is a product of young people. So the degree of
instability in the history of the speakers can be correlated with
how complex a language they are able to maintain, and apparently
whether the language has 'neotenous' characteristics. Instability
and loss through death is also important to the pace of biological
evolution.
Regards,
John Halloran
------------------------------
Date: Thu, 22 Jan 1998 19:58:33 -0700 (MST)
Subject: Re: EvolLang: efficiency
>Sent by: kaplanov@juno.com (Aslan Kaplanov)
>
>>Sent by: Cliff Lundberg <cliff@noevalley.com>
>>
>>What is efficiency in language? Brevity? Precision? My intuition
>>is that we bring in new expressions for impact, for charm, for
>>fun and variety, to refer to new things... We're just as liable
>>to introduce a verbose expression as a terse one, just as liable
>>to push the limits of grammatical propriety as to respect and
>>refine it. Should we expect speakers of other languages to be
>>different?
>
>Efficiency has already won out almost hands down.
>
[snip]
A form of efficiency can explain choices as to which sets of sounds
or sets of compounds become accepted into the lexicon of a language.
If the holistic prosody or right-brain characteristics of a new word
support the analytical or left-brain semantics of a new word, that
word will be more easily recalled to the mind, which makes that word
more efficient than other choices would be. In the U.S. a new term
has recently entered the lexicon, 'road rage', to describe a
psychological state of anger that can grip a driver behind the wheel.
This is actually a technical term for a specific condition. But
note the symmetry of the phrase's prosodic characteristics. They
allow the phrase to be more easily recalled to the mind, which makes
the compound a more efficient choice than are other combinations
which might have been more descriptive from a purely semantic point
of view.
Similarly, if there are already other words using a phoneme or set
of phonemes in a particular word position then efficiency of recall
and association will maintain a cluster of such words having similar
meanings, although their origins may have been different. Efficiency
of recall is aided by unconscious sound symbolism of the type that
David Crystal's The Cambridge Encyclopedia of Language, chapter 30,
on sound symbolism documents for words in English starting with sl-,
slime, slither, slug, sloppy, etc. This example combines both sound
symbolic and phoneme association clustering to achieve greater
efficiency of recall and association.
Language features do not exist for themselves; they exist to serve
human beings. If they do not serve that purpose, they are out the
door.
Regards,
John Halloran
------------------------------
Date: Fri, 23 Jan 1998 11:43:09 -0700 (MST)
Subject: Re: EvolLang: Re: Metaphor
>Sent by: Sherman Wilcox <wilcox@unm.edu>
>
[snip]
>
>A rough approximation at iconicity or sound-symbolism that I would
>suggest, based on these observations (on which I would certainly
>welcome feedback or criticism), is: iconicity is a special type of
>cross-domain mapping in which we map features of our conception of
>articulators onto features of our conception of some aspect of
>meaning.
That is a very precise definition. The element that remains to be
further elucidated is that of 'cross-domain mapping'. Do you have
more information about that? It would be valuable to collect other
instances of cross-domain mapping. Would the traffic lights of red,
yellow, and green mapping to stop, caution, and go be a good
example? Are the elements of this mapping completely arbitrary, or
are they based on some natural or cultural analogies?
Cross-domain mappings should be distinguished from representing a
thing or person with a photograph, recording, or painting as one is
then doing a one-to-one mapping in the same perceptual modality.
The little clay cones, spheres, and disks that have been found
throughout the Near East starting from about 9,000 BCE (calibrated)
were evidently examples of cross-domain mapping as they certainly
were not counters for big cones, spheres, and disks!
Regards,
John Halloran
------------------------------
Date: Fri, 23 Jan 1998 19:06:04 -0700 (MST)
Subject: Re: EvolLang: Re: organic & linguistic evolution
>Sent by: "Ronald Kephart" <rkephart@osprey.unf.edu>
>
>> ...Khoisan is a language maintained by a stable group of mature
>> adults and Hawaiian as a language is a product of young people.
>
>Are you (or Bichakjian) saying that there are no Khoisan young
>people? Are you saying that there are no mature adult speakers of
>Hawaiian?
Why are you asking that question?
>More crucially, are you (or Bichakjian) denying that both these
>languages pass through the filter of child language acquisition???
The issue with Hawaiian is continuity of language and whether it
was fully maintained by a small boatload of travelers. If a mother
died, her child would learn speech from other children without
correction because there would be no adult community on the new
island. The issue here is not conditions in the present, but
conditions during and after a voyage in the past. Hawaiian is
famous for its small number of consonants, in which, for example,
Proto-Polynesian / *s/ and /*f/ merged into /h/, so one can expect
its historical conditions to have been unusual. Although Hawaiian
has only 8 consonants, its large inventory of vowel phonemes (short
and long monophthongs and diphthongs) compensates to make a
reasonable inventory of 33 total phonemes. The bottom line is that
due to its unusual historical conditions, Hawaiian speech became
simplified which accounts for why it is easier to learn than is
Khoisan, Kabardian, or other languages that have not 'evolved' or
strayed significantly away from their origins. I understand that
Spanish spoken in the Americas seems childish to Spanish speakers
in Spain, who feel that they possess more precise forms of
expression. There must be many factors that can cause a language to
simplify, such as acquisition by non-native speakers. Historical
instability can accelerate the pace of language simplification.
If a linguist believes for personal reasons that language
comparable to modern speech had already been perfected by the
Palaeolithic, that linguist will find discussion of evolution of
language since Neolithic times to be distasteful. For such
linguists, the Proto-Indo-European that existed as of 4000 BCE was
a fully modern language descended from many millenia of earlier
fully modern languages. In such conditions, one can only speak of
language change, of oscillation around a center as far as we can
go into the retrievable past, and not of language evolution.
There would be no reason to have the kinds of trends that
Bichakjian's articles describe, such as a shift from SOV to SVO
and SVO to VSO phrases [a shift confirmed by the Campbell world
languages book], during historical times. So this linguistic
discussion is not occurring in a vacuum.
Regards,
John Halloran
------------------------------
Date: Sun, 25 Jan 1998 02:13:38 -0700 (MST)
Subject: Re: EvolLang: Metaphor
>Sent by: "Dr. John E. McLaughlin <mclasutt@brigham.net>
>
>> >(Take 'steal' for example, if you can describe in reasonable
>> >metaphor or analogy how the Sumerian written "phonetics"
>> >correlate to the meaning, then you might have even an outside
>> >chance of convincing me that the sound-meaning relationship is
>> >not 99% arbitrary.)
>>
>> That concept does not appear to be central or basic enough to be
>> included in the 188 V, VC, and CV words of the proto-Sumerian
>> vocabulary. The words that mean 'to steal', like ZUH, KAR, and
>> LILIB are derivative, with ZEH meaning 'female goat-kid',
>> something valuable that was easily stolen, KAR also meaning 'to
>> abduct' and 'quay, harbor', and LUL meaning 'to lie, deceive'.
>
>So how does 'to steal' derive from 'female goat-kid'?
Let me first do an analysis of ZEH, 'female goat-kid'.
Here are some Sumerian monoconsonantal words from which I derive
the morphemic meaning of the consonantal phonemes in ZEH.
za(2): stone, rock; precious or semiprecious stone;
hailstone; pit; kernel.
zu2, su11: tooth, teeth; ivory; flint, chert; obsidian;
natural glass.
ze2: to cut; to shear.
uz3, ud5, ut5: she-goat.
izi: fire.
uzu: flesh; cut of meat; omen.
Looking at these words based on the /z/ phoneme, I conclude that
the position of the tongue against the teeth for this
dental-alveolar voiced fricative caused it to mean 'tooth, teeth'
and 'to cut'. The resemblance of the spoken phoneme to the sound
of sizzling meat led to the associations 'to cook, roast' and 'meat
(animal)'. This is iconicity in proto-Sumerian. It is a matter of
primitive, concrete analogies which have not yet become organized
into a cohesive system of conceptual polarities, such as Foster
found for PIE and Jess Tauber has described on this list for other
languages.
uh: lice; insect, parasite, vermin.
ha2, hi-a: numerous; diverse; assorted; mixed.
ha3, hu3, a6, u: ten (usually written: u).
hu: bird (earlier word than mu$en).
he, hi: to mix.
he(2): abundance; abundant.
The phoneme here represented by /h/ is a velar voiceless fricative
in which the back of the tongue is held close to the velum so that
when air is forced past the constriction the result is a series of
small explosions. The repetitive small explosions led to the
meaning 'numerous small things'.
The image that one gets by combining these two phememes [Mary
Foster's term] is that of small, numerous meat animals. James
Mellaart has written that, in the Near East, goats were kept as
a meat supply because they "have just enough meat on them to feed
a family."
There is a complex of other words in Sumerian that are cognate to
ZUH, 'to steal', such as SUH, 'to uproot; to relocate; to become
confused', and ZAH2,3, 'to flee; to hide; to abandon; to be lost;
to perish'. Besides ZEH meaning 'goat kid', $AH means 'domestic
pig' [$ or 'sh' as a phememe has associations such as 'quantity;
much; portion; grain; excrement']. One could theoretically derive
ZUH, 'to steal', from the 'to cut' meaning of the /z/ phememe,
but the similar concept words are clearly related, and the animal
words derive solidly from the meanings abstracted for their
phememes.
>Do you have a similar example from another language where the
>name for a domestic animal is now the term for pilfery? The
>derivation seems far-fetched, to say the least.
I am not a comparative linguist. My response is that a small
animal is the one thing that was most likely to go missing among
an early agricultural and pastoral population. In many cases a
predatory animal would have been responsible for the deed.
>'Abduct' and 'steal' I'll buy, but what do they have to do with
>'harbor'? Do you have cross-linguistic examples?
Sumerian literature has many references to bandits and robbers
who plied the canals and waterways in their long boats, from which
they ventured to steal property and abduct women. It was a real
threat which degraded the quality of life there.
>'Lie, deceive' is also close to 'steal'. I assume that LUL is not
>derivable from something else. How is that sound sequence
>reminiscent of 'lie' or 'deceive'?
The basic phememic meaning of /l/ in Sumerian was 'easy, light,
abundant', which comes from the ease of making this sound. There
are some words with a structure similar to LUL, words whose
semantics are enlightening, even if the words do not meet a
particular linguist's definition of 'cognate'.
lal, la2: light, deficient; minus.
lal3: honey; date-syrup.
lul, lu5: n., liar; lie.
v., to lie, deceive.
adj., false; treacherous.
lil: fool, moron.
lil2: n., breath; wind; infection; spirit (of a place).
v., to infect.
>What you seem to be saying is that since you cannot find links
>between the sounds and the meanings, they must be derived words.
>I'm not following your argument here, or are you saying that only
>the V, CV, and VC configurations can be derived from a
>relationship between mouth gestures and meaning?
It is easier to deduce the meanings of the consonants in the
monoconsonantal words. I can find links between the sounds and the
meanings in many diconsonantal words, as demonstrated above, but it
is harder work to do so. I have not tried hard to do so in the past
because I assumed then, apparently mistakenly, that these words
derive from a time when phoneme iconicity was becoming opaque to the
language speakers.
>> The word DAB meaning 'to take, hold, receive' can be analyzed
>> similar to the cognate word DUB, 'to store', in that the
>> component elements mean 'to go' + 'container' (cf., BAD, 'to
>> open, release' from 'container' + 'to go').
>
>What does this have to do with 'steal'?
You wanted me to analyze your word, but 'to take' was the closest
analyzable concept I could find. I doubt that people placed a big
emphasis on personal property or the concept of 'stealing' 10,000
years ago. I gather from the lack of response that no one
disputes my analysis of the example that I chose.
Regards,
John Halloran
------------------------------
Date: Mon, 26 Jan 1998 13:42:59 -0700 (MST)
Subject: Re: EvolLang: organic & linguistic evolution
>Sent by: "Ronald Kephart" <rkephart@osprey.unf.edu>
>
>> ...Although Hawaiian has only 8 consonants, its large inventory
>> of vowel phonemes (short and long monophthongs and diphthongs)
>> compensates to make a reasonable inventory of 33 total phonemes.
>
>Please list the phonemes of Hawaiian. Normally, I don't think
>diphthongs are treated as separate phonemes. My info has:
>
>Consonants: p k ?
> h
> m n
> l w
>
>Vowels: i a e o u; plus vowel length (a:, e:, etc.) which
>would count as one phonemic contrast (right?).
See An Introduction to the Languages of the World (1997), Anatole V.
Lyovin, which devotes pp. 257 - 282 to a sketch of Hawaiian. The
author is Associate Professor of Linguistics at the University of
Hawaii, Manoa.
He agrees with your list of consonants and short and long
monophthong vowels. But then he says, "In addition to the
monophthongs there are also the following short diphthong phonemes:
iu, ei, eu, ai, ae, ao, au, oi, ou
"And the following long diphthong phonemes:
ai, au, ae, ao, ei, ou [where there are macrons over each a]
"All other combinations of vowels act as sequences of separate vowel
phonemes, not diphthongs."
Referencing a 1980 article by Albert J. Schuetz which reanalyzed the
Hawaiian vowel system, he says, "the total number of phonemes adds
up to 33 units: 5 short monophthongs + 5 long monophthongs + 9 short
diphthongs + 6 long diphthongs + 8 consonants."
>> ... Hawaiian speech became simplified which accounts for why it
>> is easier to learn than is Khoisan, Kabardian, or other languages
>> that have not 'evolved' or strayed significantly away from their
>> origins...
>
>It might well be easier for a Samoan speaker to *learn* Hawaiian
>than Khoisan. Is there good evidence for differential ease of
>*acquisition* of these languages by children? Do Khoisan speaking
>children take *longer* to acquire their language than Hawaiian
>speaking children?
That is a good question which could be asked about many languages,
and an attempt made to correlate that ease with the evolutionary
history of the languages. Are creole languages easier or more
difficult to learn? Larry Trask is the one who said, "Hawaiian is
certainly vastly easier to learn" in the original post. Perhaps
he was speaking only of learning it as a second language.
>I would argue (being a monogenesist) that *all* languages have
>strayed equally from their origin.
>> ...I understand that Spanish spoken in the Americas seems
>> childish to Spanish speakers in Spain, who feel that they possess
>> more precise forms of expression...
>
>Of course, we as linguists cannot accept folk ideologies about
>language for scientific fact. Note that speakers of standard
>average English make similar claims about Black English, English
>Creoles, and so on, as do metropolitan French speakers about
>Haitian Kreyol. These statements reflect attitudes about *speakers*
>of languages, converted (unconsciously?) to statements about their
>*languages* so as to hide ethnocentism, racism, etc.
Those are good points. I have corresponded with a translator in
Spain who traces the offensive sounding forms to proximity to
American English. He says that in Spain they have to dub
Mexican soap operas into *Spanish*. Argentine Spanish sounds much
better to them.
>I think that we need to follow the advice of Jose-Luis Mendivil
>Giro who pointed out a couple of days ago that we should make a
>distinction between evolution of Language (Spanish "Lenguaje") in
>the species and the changes that occur between and among languages
>(Spanish "lenguas" or "idiomas") that get us, say, Spanish from
>Latin, which may or may not be anything like the development of
>Language out of Proto-Language in the species.
>
>In fact, as I write it occurs to me that the distinction between
>homology and analogy might also be important here, tho I haven't
>had enough coffee yet this morning to take it further.
As I understand it, homologies or homologues in biology and
cladistics refers to equivalent functioning forms in different
species, which may or may not result from evolutionary descent from
an earlier ancestor. Isn't it the case in biology that different
animal species can develop very similar forms in response to similar
ecological niches? Wouldn't this suggest that humans could also
develop homologous language features without requiring that the
features descend in an evolutionary way from an earlier ancestor?
Regards,
John Halloran
------------------------------
Date: Mon, 26 Jan 1998 17:45:53 -0700 (MST)
Subject: Re: EvolLang: vowels and consonants
>Sent by: mcv@wxs.nl (Miguel Carrasquer Vidal)
>
>>Zylogy <Zylogy@aol.com> wrote:
>>
>>The semantics of vowels seem to cluster around time and space- for
>>time, aspectual notions (usually shwa for potential or irrealis, a
>>for actual- in a variety of aspectual subsenses). Spatially, i,e
>>usually refer to spatial minimization, with materials contained
>>within a cavity being squeezed out, or for the external surface of
>>an object, a long, thin spatial distribution. Low, a-type vowels
>>make reference to broad, flat surfaces and maximal contact of
>>objects with them. u,o refer to enclosing spaces which are
>>spatially maximizing- in other words, expanding volumetrically,
>>sucking materials into them, etc. Finally, shwa-type vowels refer
>>to objects which have no moorings, materials divorced from
>>containment, etc.
>
>Isn't it just bloody typical that no attempt is made to back any
>of this up with any kind of data, from any language?
Jess,
Don't let Miguel's abrasive manner get to you. I appreciate your
high level point of view and am learning new things from you. I
have also asked if you could put some of your data up on a web site,
and Matt has offered to provide space for that. But I will be
surprised if determined skeptics are convinced by your data
supporting the conceptual polarities and frameworks that you have
been describing. Your mind is evidently good at parallel processing.
There are linguists who think in the way that you do - Foster,
Pulleyblank, Hewes, von Raffler-Engel, Fonagy, and de Grolier all
published papers devoted to exploring early isomorphisms between
sound and meaning in the 1983 Glossogenetics volume edited by Eric
de Grolier.
If the following quote from Miguel is any indication, however, you
will never convince him.
mcv@pi.net (Miguel Carrasquer Vidal) writes:
|> Let me state a basic rule of linguistics: apart from a very few
|> onomatopoeic words, there is *no relation at all between sound
|> and meaning in language*.
|> Let me repeat: *no relation at all*
|> This is not negotiable.
In contrast, Pulleyblank writes, "Chinese also has many words meaning
"turn", "bend", "round", etc., that seem to have had a uniconsonantal
root w. Many words meaning "dark" either begin or end in m, and so
on." ibid, p. 382. On page 404 he writes, "A side result of our
present investigation, however, is that the earliest recoverable
forms of Sino-Tibetan and Indo-European do show a transparently
consonantal structure that is found in only a few remote and isolated
languages at the present day but is closer to what spoken language
was probably like when it first developed."
Regards,
John Halloran
------------------------------
Date: Tue, 27 Jan 1998 01:27:24 -0700 (MST)
Subject: Re: EvolLang: Metaphor
>Sent by: Michelle Sutton and John McLaughlin
>
>Problem #1: Big, nearly fatal problem: How do you know what any of
>the SOUNDS were in Sumerian? We know the writing system and we
>know what the Assyrians (or whoever those first Semitic writers
>were) thought Sumerian sounded like, but that is like asking a New
>Yorker to pronounce Arma virumque cano, Troiae qui primus ab oris.
>Do we then know what Virgilian Latin sounded like? How do you know
>that Sumerian /z/ was dental OR alveolar (or even voiced). You
>cannot make an argument for meaning if it is based on the phonetics
>of Sumerian. Unless you can produce a tape of Gilgamesh telling his
>own story, in language, then you CANNOT argue about Sumerian based
>on phonetics.
That is a big assertion for a non-Sumerologist to make. Sumerian
loaned at least 750 words into Akkadian, a Semitic language whose
native words are related by a series of phonological rules to
vocables in other Semitic languages. Akkadian had the distinctions
of unvoiced, emphatic, and voiced sibilants; Assyriologists know
which written cuneiform syllable graphemes are used for which
phonemes. Different syllable graphemes are used for /emphatic s/,
for /s, sh/, for /s/, for /sh/, and for /z, s, emphatic s/. The
latter syllable graphemes are the ones that Sumerologists render
as /z/. Certain Sumerian loanwords show up in Akkadian with the /z/
phoneme and others show up with the /s/ phoneme. These two phonemes
are written differently in Sumerian, so Sumerologists think that
they know which words have /z/ and which have /s/. As to whether
the phoneme was dental or alveolar, I would think the fact that
'tooth, teeth' is one of the meanings gives us a clue.
>> The phoneme here represented by /h/ is a velar voiceless fricative
>> in which the back of the tongue is held close to the velum so
>> that when air is forced past the constriction the result is a
>> series of small explosions.
>
>Whoa there! Once again, how do you know that /h/ in Sumerian (or
>"Proto-Sumerian") was voiceless or a fricative or velar? Where are
>your tape recordings? Was Virgil's virumque pronounced with an
>initial [w] or initial [v] or initial something else? WE DON'T
>KNOW! YOU don't know how /h/ in Sumerian was pronounced other than
>trusting in how the Assyrians wrote it in their writing system.
>Your use of Assyrian spelling to represent Sumerian phonetics is
>like using Pope John Paul II's Latin pronounciation to make
>assertions about how Cicero spoke.
You're shouting about an area that is not your specialty. Sumerian
did not have the normal /h/. Because of this, there is normally no
ambiguity in using /h/ to represent the /h with the 'dish'/ that a
good Sumerian transliteration font uses. Regarding the /h with the
'dish'/, I quote, "This represents a voiceless velar fricative, i.e.,
I.P.A. /x/, in accord with normal Assyriological and Semitistic
convention." Stephen J. Lieberman, The Sumerian Loanwords in
Old-Babylonian Akkadian (1977), p. 66. In Pullum and Ladusaw's
Phonetic Symbol Guide, IPA /x/ is described as a "voiceless velar
median fricative". I think that after more than one hundred years
of work, Near Eastern scholars know what this phoneme was. I will
just say that the 'hardness' of this /x/ phoneme is indicated by the
fact that for some words in some areas it interchanged with /g/ and
/k/.
>> >'Abduct' and 'steal' I'll buy, but what do they have to do with
>> >'harbor'? Do you have cross-linguistic examples?
>>
>> Sumerian literature has many references to bandits and robbers
>> who plied the canals and waterways in their long boats, from which
>> they ventured to steal property and abduct women. It was a real
>> threat which degraded the quality of life there.
>
>Has anyone ever measured your arm length? :-) This stretch is
>utterly incredible.
No, sir. The Sumerian word KAR has several meanings.
kar: n., embankment; quay-wall; mooring-place; harbor; marketplace;
port authority.
v., to take away; to steal; to raid, capture, pillage; to
escape; to avoid.
Your comment about the importance of private property in the Near
East appears to overlook the 6,000 year time-gap between the start
of the Neolithic and Hammurapi's Laws. When I talk about
proto-Sumerian, I am talking about dwellers in small Neolithic
villages in western Iran, such as Ganj Dareh, not about the later
sophisticated city-states of the Mesopotamian flood plain.
Regards,
John Halloran
------------------------------
Date: Thu, 29 Jan 1998 01:24:34 -0700 (MST)
Subject: Re: EvolLang: Sumerian transliterations
>Sent by: "Peter T. Daniels" <grammatim@worldnet.att.net>
>
>Mr. Halloran, who previously has confused Sumerian writing with
>Sumerian language,
You may have thought that at one time.
>now confuses roman-letter transliterations of Sumerian sounds
>(presumably phonemes) with the realizations of those phonemes.
Would you care to explain what you mean?
>Would he please enlighten us as to where he studied Sumerian? Does
>he operate within the Jacobsen school, or the Kramer school, or
>the Hallo school, or the Landsberger/Civil school, or is his mentor
>one of the distinguished European scholars who have devoted their
>lives to this pursuit?
The resources that I used to compile the Lexicon of Sumerian
Logograms are listed in its Bibliography. I have studied Sumerian
at UCLA over the last year with Robert Englund, including a graduate
class last fall.
When I was an undergraduate at UC Santa Barbara I took 48 units of
Hebrew, mostly independent studies, with Robert Hetzron, the
distinguished Semitic and Afroasiatic linguist. You will recall
that you and I corresponded over the summer regarding the late Dr.
Hetzron.
My ideas and observations are my own and should not be blamed on
anyone else. I believe that some of my ideas are unique because I
based them on an analysis of the Sumerian vocabulary as an object in
itself, instead of just as a tool for translating texts, which is I
believe how most Sumerologists and Assyriologists learn the language.
As should be evident from the new version 2.1 of the Lexicon that is
up at my web site, I have spent a great deal of time with the
Sumerian vocabulary, enough to have developed a love for it and to
feel that I understand certain of its principles.
I don't have a problem with anyone asking if I have considered this
point or that point when it comes to proto-Sumerian. I welcome that.
But please don't lecture me without having yourself tried to study
the subject in depth.
Regards,
John Halloran
------------------------------
Date: Thu, 29 Jan 1998 03:14:34 -0700 (MST)
Subject: Re: EvolLang: Metaphor
>Sent by: Michelle Sutton and John McLaughlin
>
[snip]
>
>First, Sumerian is not related to Semitic, so nothing in the
>phonology of Semitic gives us any evidence about the phonetics of
>Sumerian.
Thank you for the evident thought given to this new post. But I
must say that your use of the words 'nothing' and 'any' in the above
invalidate the statement.
>Borrowings are almost always adapted to the phonology of the
>recepient language and very rarely preserve the phonetics of the
>original language. In English, we say [bif] and not [b@f] for the
>meat of cattle (I'm using the @ for a mid front rounded vowel).
It looks to me like English preserved 2 out of the 3 phonemes, which
is not "very rarely".
>Second, we don't know exactly how Akkadian was pronounced.
>
[snip]
>
>Third, when you think about the long chain of only written
>transmission from Sumerian to Ancient Semitic to Modern Semitic,
>you have lost any ability to be confident about the sounds of an
>ancient language. Yes, you may be able to tell whether a particular
>sound was alveolar or labial, but you cannot determine laminal
>versus apical, dental versus alveolar, even voiced versus voiceless.
>If you have any question about determining the phonetic values of a
>given written letter, just look at the history of the q, k, c, g
>complex going from Semitic through Greek through Etruscan through
>Latin to English and French (and thence to English again).
We know that the Sumerians had phonemes in certain approximate areas.
We also know which words used these phonemes. As in the case of
Sumerian words in which the only consonant was the phoneme that
Sumerologists believe to have been IPA /x/, which cluster in meaning
around the idea of 'small, numerous things', does it matter where
precisely it was pronounced? Semantically, the /x/ words oppose the
/sh/ words. Perhaps I should just speak of a 'back fricative' and a
'front fricative'. To agree with the iconicity, that is all I need.
[snip]
>
>was a set of related sounds in that language. And, once again, how
>much did Akkadian phonology influence the recording of the phonemes
>of Sumerian?
Sumerologists have succeeded in identifying a velar nasal in Sumerian
words. Akkadian did not have that sound.
>> kar: n., embankment; quay-wall; mooring-place; harbor; marketplace;
>> port authority.
>> v., to take away; to steal; to raid, capture, pillage; to
>> escape; to avoid.
>
>OK, these two look alike, but aren't they just homophones? Even the
>secondary meanings don't show any relationship. Usually there's
>something there. There's nothing here except the shape of the root
>to relate these two meanings. Maybe Sumerian had two 'k'-like sounds
>and Akkadian had to collapse them into one because they weren't in
>Akkadian?
The word is a Sumerian logogram, usually represented by a particular
ideographic sign. So the identity is not in question. The first
entry collects all the noun meanings of the logogram. The second
entry collects all the verb meanings of the logogram. There is a lot
of polysemy in Sumerian [this is not to be confused with how a sign
sometimes has several logographic readings - what I discuss here are
true logograms or words in the spoken language, not written signs].
To me the semantics simply follow from the culture, history, and
environment of the Sumerians, including the water-going bandits that
I mentioned before.
If you think the polysemy of that word was extreme, consider this
one:
ul: n., bud; flower; ornament; joy, pleasure, satisfaction; a
unit of measurement.
v., to glitter, shine.
adj., remote, distant (in time); ancient, enduring.
>"Proto-Sumerian" is a misnomer. "Proto-" implies a carefully
>reconstructed common ancestor of a group of related languages (e.g.,
>Proto-Indo-European, Proto-Germanic, Proto-Uto-Aztecan, etc.).
>"Pre-" implies internal reconstruction and involves a set of things
>of varying time depth before the moment the language is recorded
>(e.g., Pre-Comanche means a change that happened sometime before the
>present, Pre-Anglo-Saxon means a change that happened sometime
>before the Old English records). If you are doing internal
>reconstruction of Sumerian, you must use the term Pre-Sumerian, but
>you cannot assign a date to anything because you have no
>corroborating evidence.
If I am using the English language, that is the correct term for the
V, VC, and CV vocabulary, which together describe a culture and an
environment that is early Neolithic. Prot- or proto- mean "1. a.
first in time"; "2. a. first or lowest of a series"; "3. a. first
formed"; "4. *cap* relating to or constituting the recorded or
assumed language that is ancestral to a language...." These
definitions are all correct descriptions of what I mean.
>You stated in another post that you cannot give cross-linguistic
>evidence because you haven't investigated any other languages.
I said that I am not a comparative linguist, not that I haven't
learned other languages. I don't trust myself to do comparative
linguistics. I am good at analysis, at breaking things down.
Joining together words whose phonology and semantics only vaguely
resemble each other is not my forte.
Regards,
John Halloran
------------------------------
Date: Fri, 30 Jan 1998 12:14:27 -0700 (MST)
Subject: Re: EvolLang: history
>Sent by: "Peter T. Daniels" <grammatim@worldnet.att.net>
>
>> Zylogy wrote:
>>
>> Anyway, what did you folks get out of the -ag set?
>> Here's another, just to be cantankerous:
>>
>> to cam, clam, cram, dam, jam, ram, slam
>
>
>Well, two of 'em aren't verbs ...
Jess, you deserve more of a response than this. The symbolism of
the -ag set seemed to relate to 'going down', in strength, value or
literally, appropriately represented by an articulation sequence
that ends down in the throat. In contrast, clam, dam, and slam end
with one closing the mouth at the front with on-going finality.
These gestalt word characteristics increase the utility of these
words to humans, enabling appropriate processing by both the left
brain hemisphere and the right brain hemisphere. Has any study been
done on which parts of the brain process the speech features subsumed
under the term 'prosody'?
Looking in David Crystal's Dictionary of Language and Languages the
above word characteristics fall under synesthesia and phonesthetics,
with the addition that our non-verbal hemispheres probably look at
the articulation of the word as a whole.
I know that you believe many of these word characteristics to be
represented in a particular language's vocabulary as a result of
natural selection, with proposed words that lack appropriate gestalt
characteristics dying out to be replaced by more expressive words.
This should not be confused with the more conscious process of iconic
symbolism that I and others believe to have been used by many
language inventors.
So the former process should be perfectly acceptable to the
'Long-Rangers', as it does not say anything about when and how
speech initially arose.
In regard to the latter process, I don't have evidence that iconicity
has been a conscious feature of any language for thousands of years
(although Jess did say something about some indigenous cultures
having recent traditions of a person who was responsible for new word
formation). And different language cultures saw different symbolism
in the same articulations - where the Chinese saw a cave in /m/
signifying 'dark', the Sumerians made female and generative
associations, suggesting that they saw a womb in /m/. I don't think
anyone is claiming universal symbolism for any articulatory icons.
Regards,
John Halloran
------------------------------
Date: Sun, 1 Feb 1998 01:41:34 -0700 (MST)
Subject: Re: EvolLang: Metaphor
>Sent by: John McLaughlin <mclasutt@brigham.net>
>
>> Sent by: kaplanov@juno.com
>>
>> > <mclasutt@brigham.net> writes:
>> >
>> >Here is the root of your problem. You are trying to analyze
>> >everything as polysemy and nothing as homophony. (For the
>>
>> Maybe Sumerologists are taking the Akkadian evidence too
>> strongly. Some languages have particular characteristics
>> others don't have. Semitic with its emphatics, Chinese with
>> its tones, IE with its consonant-clusters, etc.
>>
>> Maybe Sumerians had more vowels than Semitic /aiu/
>> or the common five-vowels system (add /oe/) and caused
>> both polysemy and homophony in Akkadian.
>
>This is a very good observation. I had been thinking about
>Sumerian consonants vis a vis Akkadian and Ancient Semitic, but I
>agree with you that it is far more likely that the vowel systems
>were different, leading to all kinds of problems for the Akkadian
>transliteraters.
Sorry to dash cold water on this speculation, but we are talking
about polysemy or homophony by the Sumerians, not the Akkadians. The
reason is because we do not just know these words in transliteration
by the Akkadians, but in thousands of texts from the Sumerians
themselves in which a single sign represents the word /ul/ or the
word /kar/. If separate words were merged together in a homophony
process, it occurred earlier when the Sumerians invented the
logographic signs of their writing system. The most commonly cited
example is using the sign for the logogram /ti/, 'arrow', as the sign
for the logogram /til/ or /ti/, 'to live; life'. These conflations
are actually pretty rare and are easily distinguished. If there is
not a logical connection between 'ul' meaning 'flower, ornament,
joy; to glitter, shine' and 'ancient, enduring', it may just be
because my lexicon does not yet include the meaning 'star' that is
given in Labat's Manuel, a meaning that is able to bridge the two
concepts.
Regards,
John Halloran
------------------------------
Date: Mon, 2 Feb 1998 02:27:54 -0700 (MST)
Subject: Re: EvolLang: Language origins: a linguist's view
>Sent by: "Larry Trask" <larryt@cogs.susx.ac.uk>
>
>> Steven Wise writes:
>>
>> One of the most important aspects of the work of Dr. Sue
>> Savage-Rumbaugh with the bonobo, Kanzi, for example is precisely
>> that Kanzi began to learn language, or language-like skills, with
>> no training at all. Indeed, it was Kanzi's adopted mother who was
>> being trained and not very successfully. Kanzi, who was an infant,
>> learned by observation alone. Dr. Savage-Rumbaugh has since
>> replicated this with other bonobos.
>
>A fair point, and I would have mentioned it if I'd had more space,
>but I didn't, and so I was forced to oversimplify on a number of
>points. Still and all, however impressive Dr Savage-Rumbaugh's
>bonobos are, they are reacting to laboratory conditions, in which
>they appear to exhibit behavior they do not exhibit in their natural
>state. And I consider this significant.
The ability of a nonhuman infant to acquire language just from
exposure to it is an argument that human children all learn language
not because they have a language faculty but because they are all
normally exposed to language.
>Sent by: "Larry Trask" <larryt@cogs.susx.ac.uk>
>
>> Arslan Kaplanov writes:
Kudos to both Larry and Arslan for a very stimulating discussion.
>I am not denying that our language faculty has close links with
>other parts of the brain. But using abstractions like `absence' and
>`non-identity' is definitely *not* just "naming". Using `Daddy' is
>naming; using `dog' is a much more sophisticated type of naming; but
>using `absence' simply does not qualify as any kind of naming I can
>think of. Little Johnny sees Daddy and says "Daddy"; little Johnny
>sees a dog and says "dog"; little Johnny sees an absence and says
>"absence". Are you kidding?
That's an interesting point. Sumerian does not have words for
'absence' or 'zero'. Sumerian just has a negation word, /nu/. But
then the Sumerians always seemed to me like pretty concrete-minded
people. I wonder if we are not here talking about conceptual
sophistication, about an evolving ability on the part of people to
handle not just the concrete but the abstract. A book that I once
read about thinking in born-deaf people pointed out how helpful
learning a language can be in acquiring abstract concepts. But you
have to admit that the concepts must come first, to then be promoted
and carried forward by the language.
A postscript to this point is that one probably would not find a
word for 'absence' in the Hebrew Old Testament either. Biblical
Hebrew was fairly easy for me to learn because the vocabulary and
therefore the concepts were more limited than in modern language.
Scientific or scholarly German has many more abstract concept words
than does Biblical Hebrew.
>> > We have yet to see any persuasive evidence that any other
>> >creature on earth possesses anything even remotely comparable in
>> >even a single one of these respects. Human language is truly
>> >unique, a thing apart. It does not resemble what vervet monkeys
>> >do.
>
>> Yes it does. It is a matter of degree, not of kind. It is the same
>> kind but of different degrees. By this type of reasoning you'd
>> have to claim that the type of heat that sets papers on fire is
>> not the same type of heat that warms us during the day. Of course
>> the only difference is in the temperature, clearly a matter of
>> degree.
>
>I'm sorry, but statements of this kind make me despair of conducting
>a rational dialogue. The statement that vervet signals are similar
>in nature to human language and different only in degree is, so far
>as I can see, an article of faith, supported by no evidence at all,
>and flagrantly in conflict with a huge body of evidence.
I have to agree with Larry's last statement here. But Arslan is
correct to indicate that there is continuity between the mental
abilities of animals and humans, that the difference in brain power
is mainly a matter of degree.
>> > *Proposition 4: Human beings are born with a powerful instinct
>> > to learn language.*
>
>> This is the "innateness" hypothesis again in different clothing.
>> If it is really going to mean anything it should be clearly
>> stated. What is innateness and what is instinct. Merely making up
>> names for things is not an explanation.
>
>I agree with your last statement, of course. Now, I am not a
>supporter of Chomsky's version of innateness, but, like most
>linguists, I do believe that any healthy human child *will* learn a
>first language, come what may, unless it is actively prevented from
>doing so. And this observation needs a good explanation.
See the discussion above regarding the bonobo infant learning
language through exposure to language.
<snip>
>> But I think you want to say more because at least some linguists
>> want to claim more. But what exactly?
>
>I want to claim that we have a language faculty that is absent from
>other living species, that's all.
What do you mean by faculty? Do you mean that we speak because we
have that faculty? If it is so necessary, how did we start speaking,
and how does the bonobo infant acquire language without human
instruction? Or do we speak because somewhere along the line our
parents came up with the idea of communicating with symbolic
vocalizations, and their children learned and elaborated on that
process?
Regards,
John Halloran
------------------------------
Date: Fri, 6 Feb 1998 00:45:52 -0700 (MST)
Subject: EvolLang: Proto-Language - Characteristics
>Sent by: Matt Fraser <mattf+@pitt.edu>
>
[snip]
>
>What are the minimum criteria for a Proto-Language? Clearly (IMO),
>chimps and gorillas learning sign language (particularly Kanji's
>achievments), the recent neurological study on chimp brains and
>the close proximaty of chimps and humans evolutionarilly together
>suggest that chimps have the functional hard wiring for limited
>language. Limited HUMAN language. Are they Proto-Language users
>in the wild - is it under our noses and we just aren't seeing it?
>
[snip]
>
>Dunno, seems to me that we have to define Human Language as what is
>the LEAST (I know, I know - keep it down, it's late and people are
>trying to sleep - just got excited - sorry) common denominator of
>what we have now in modern humans that would still be considered
>human language, and if any of that is taken away - whaddaya got?
>Proto-Language? Any non-human critters doing it?
Creatures such as the vervet monkeys are capable of making
appropriate referential communication. But they have not graduated
to the idea of using unique signs for a wide range of objects,
actions, and concepts. One does not see a monkey teaching his fellow
monkeys signs for earth, sky, tree, water, hand, grasp, release, go,
come, etc. I am sure that they do have ways of communicating good
and bad and information that is important to their survival. One
could say that the difference between animals' use of referent signs
and our own use of referent signs is a matter of degree, but that
does not have to mean that human use of referent signs evolved as an
unbroken amplification of animal use of referent signs. The way in
which humans make their signs cannot be traced to the way in which
existing animals make signs. One cannot even trace an evolutionary
path by which the different animal and insect species that do have
some form of referential signing system relate to each other, so
there should be no problem with saying that the human signing system
also evolved independently.
Another crucial difference is that animal and insect signs appear to
be genetically mediated, so that, like the songs of particular bird
species, there are phylogenetic channels of referential meaning which
are waiting to be refined by ontogenetic experience to reach full
realization. This contrasts to the large plastic net of the infant
human brain, waiting to receive a culture transfusion whose
characteristics are unpredictable. When the culture dump has been
completed, the unused connections wither away. The plastic
adaptability of the young human brain is what enables humans to live
across so many ecological niches. To the extent that bonobos and
other young animals have surplus cortical connections for learning
what is unique about the particular environment into which they have
landed, as opposed to being born with all needed knowledge and
instincts programmed into them by their genes, they should be able to
master a few culturally determined referential signs. Whether they
do that would depend on whether they live in a community that has a
culture of signs that is passed down through the minds of succeeding
generations.
Regards,
John Halloran
------------------------------
Date: Fri, 13 Feb 1998 12:20:07 -0700 (MST)
Subject: Re: EvolLang: On defining `language' and `communication'
>Sent by: mcv@wxs.nl (Miguel Carrasquer Vidal)
>
[snip]
>
>Of course, not all deaf children invent their own language. For
>one thing, it takes at least two individuals (e.g. siblings creating
>their own "home signs"). And not all deaf siblings inevitably
>create those "home signs", although many do. In that sense, one
>cannot say that the language instinct is "innate". But it doesn't
>seem to take exceptional genius to come up with the spark of
>invention needed to create a pidgin. It seems to be well within
>the reach of an averagely (or maybe only slightly above-averagely)
>intelligent group of children. One might compare it to cats'
>ability to open doors. Not all cats figure out how to do it, but
>many do [quite independently of each other].
>
>What *is* innate in human beings is the ability to learn language
>once one is exposed to it, barring cerebral damage of course. And
>the case of the deaf communities demonstrates that there are
>typically enough groups of individuals (two is enough) to create
>langauge "ex nihilo" (two generations is enough) again and again.
>
>To me this implies that the question "How did language emerge?" has
>already been answered. It just happened again in Nicaragua...
>
>It also seems to imply that language is potentially as old as our
>brain, i.e. as our species
>
[snip]
I don't think one can extrapolate from the ease with which modern
children create symbolic language that it was just as easy for the
first speaking populations to create and use symbolic language.
Don't you agree that natural and sexual selection have selected for
brains that are good at language during the time that populations
have been speaking?
I have discussed before on this list the fallacy of projecting the
conditions that we know in the present back into the past.
Since you are studying the deaf, pay attention to how they are quite
capable of concrete thought even when they lack symbolic language.
The cognitive abilities of the uneducated born-deaf make me
comfortable with the idea that humans could achieve paleolithic and
early neolithic culture without having modern symbolic language.
Regards,
John Halloran
------------------------------
Date: Fri, 13 Feb 1998 12:33:09 -0700 (MST)
Subject: EvolLang: Refuting The Language Instinct
With my belief that modern symbolic language is a relatively recent
invention, I oppose the view that language use has been hard-wired
into our brains by hundreds of thousands of years of language use.
The following book announcement on the Linguist List comes as
welcome news. Matt is free to snip whatever parts are inappropriate
for our list.
*****
Geoffrey Sampson
School of Cognitive & Computing Sciences
University of Sussex
Falmer, Brighton BN1 9QH, GB
e-mail geoffs@cogs.susx.ac.uk
tel. +44 1273 678525
fax +44 1273 671320
Web site http://www.grs.u-net.com
-----------------------------------------------
Title: Educating Eve
Author: Geoffrey Sampson
Author Affiliation: Lecturer at the School of Computer Science and
Artificial Intelligence at the University of Sussex, UK
Publisher: Cassell
Field: Linguistics,
Format: hardback
Price #45.00/$67.95
Order details: UK and rest of world please call +44 1202 665432
USA: please contact Books International, PO Box 605, Herndon,
Virginia 20172-0605 on tel 703 661 1589, fax 703 661 1501
Synopsis of the book:
Are we creatures who learn new things? Or does human mental development
consist of awakening instinctive structures of thought?
A view has gained ground - powerfully advocated, for example, by
Steven Pinker's book The Language Instinct - that language in much of
its detail is hard-wired in our genes. Others add that this also
holds true for much of the specific knowledge and understanding
expressed in language. When the first human Eve evolved from pre-human
apes (it is claimed), her biological inheritance comprised not just a
distinctive anatomy but a rich structure of cognition.
Despite the impressive roll of converts which these ideas have gained,
there is no good reason to believe them. The arguments of Pinker and
others depend on earlier and more technical contributions, by writers
such as Noam Chomsky. Many readers take these foundations on trust,
not realizing how weak they are. This book examines the various
arguments for instinctive knowledge, and finds that each one rests on
false premises or embodies a logical fallacy.
A different picture of learning is suggested by Karl Popper's account
of knowledge growing through 'conjectures and refutations'. The facts
of human language are best explained by taking language acquisition to
be a case of Popperian learning. Eve was not born a know-all. She was
born knowing nothing, but able to learn anything. That is why we can
find ways to think and talk about a world that goes on changing today.
*****
Regards,
John Halloran
------------------------------
Date: Sat, 14 Feb 1998 15:44:19 -0700 (MST)
Subject: Re: EvolLang: On defining `language' and `communication'
>Sent by: mcv@wxs.nl (Miguel Carrasquer Vidal)
>
>>John A. Halloran wrote:
>
>>I don't think one can extrapolate from the ease with which modern
>>children create symbolic language
>
>What does "symbolic language" mean?
Communication which divides up the world into discrete objects and
actions and uses symbols to refer to these discrete parts.
>>that it was just as easy for the first speaking populations to
>>create and use symbolic language.
>>
>>[...]
>>
>>Since you are studying the deaf, pay attention to how they are
>>quite capable of concrete thought even when they lack symbolic
>>language.
>
>Isn't sign language "symbolic language" then?
Sorry. I thought your book on the deaf would extend into discussion
of the uneducated born-deaf, who were never taught sign language.
Quoting from an excellent book on the subject, "The evidence for
conceptual thinking in the linguistically deficient deaf has been
presented and leads to the direct conclusion that thinking develops
through living contact with the environment regardless of the
presence or absence of a ready-made linguistic symbol system." Hans
G. Furth, Thinking Without Language: Psychological Implications of
Deafness (1966). pp. 198-99. He says that any theory that would
explain thinking in terms of verbal or other symbolic units is flawed
and rests on false assumptions.
Earlier, Sir Richard Paget made a similar statement, "As to the
uneducated born-deaf, the absence of any training in the art of
giving a *name* to each element of the events which they experience
(such as all hearing children obtain in their early days) has this
effect: viz., that the born-deaf naturally appreciate events as a
whole. The consequence is that they cannot direct their thoughts
to the separate items of which (to our minds) the events we all
experience are composed." "But the relatively modern art of
*directed thought*, which analyses events and gives separate verbal
(or other) symbols to each separate component, is *not* natural to
mankind; it is a comparatively modern mental trick which has to be
learnt in early youth by every hearing individual. Its importance is
enormous, for it has given to mankind the powers of logical thought,
and the gifts of imagination and invention. There seems good reason
to suppose that the mentality of Paleolithic man cannot have been
very different from that of the uneducated born-deaf of to-day."
from "The Origins of Language, with Special Reference to the
Paleolithic Age," Journal of World History 1 (1953), p. 404 [article
pp. 399-414].
Regards,
John Halloran
------------------------------
Date: Mon, 16 Feb 1998 21:12:23 -0700 (MST)
Subject: Re: EvolLang: Ancient & Tribal Languages/Thought
>Sent by: "Dr. John E. McLaughlin" <mclasutt@brigham.net>
>
>Actually, many "modern, technological" languages also have the
>passive, like English, Greek, French, Russian, etc. I don't recall
>ever seeing a language WITHOUT a passive construction.
I think the claim was more that some native languages LACK the
active or accusative construction. Have you seen that?
>Sent by: "Larry Trask" <larryt@cogs.susx.ac.uk>
>
>I happen to be a specialist in Basque, the world's most famous
>ergative language, and possibly the one with the most thoroughgoing
>ergative morphology of all. Basque-speakers absolutely do not
>perceive the world differently from their Romance-speaking
>neighbors, and the crazy statements about Basque made by
>19th-century linguists are laughed away by native speakers and
>modern linguists alike.
>
>Nor is there anything special about polysynthetic languages, nor
>any evidence that their speakers see the world differently from us.
Okay, you are saying that language does not determine thought. It
would be interesting to compare languages invented by modern minds,
such as sign languages, with languages invented a long time ago. Do
sign languages have a variety of verbal constructions, including
accusative, active, passive, ergative, etc.? If so, which forms did
the inventors implement first? Because saying that language does not
determine thought does not mean that thought has not determined the
forms of language.
>I haven't seen the article cited, but I'm not sure I want to. I
>simply do not believe in the reality of something called "the
>primitive mind", and I don't know of anybody else who does, either.
>This objectionable phrase is nothing more than an anachronistic
>remnant of 19th-century European racism.
I will make one point from my knowledge of ancient cultures. Over
the centuries there has been a trend to place increasing importance
on the individual. This has been at the expense of the family and
the tribe. There is much more value placed on the value and rights
of the individual now than was the case 4,000 years ago. Looking
at the simple names that Sumerians had, such as Servant of Baba (the
deity of a temple), I can imagine a time when people were like cows
in not having names. I believe that there was a time when only the
tribe had an identifying name and that name was the totemic symbol
that represented that tribe compared to other tribes.
I am a little fuzzy on how a deemphasis on the individual would
coordinate with use of the passive voice, but I believe they are
connected. The post started with a statement that the modern Western
mind emphasizes the actor as opposed to the action. Our Western
legal system certainly emphasizes the responsibility of the
individual as the actor. This would contrast with the village elders
gathering together to achieve a solution that restores harmony to the
village as a whole, a system that is still practised in rural India.
Can someone help me to make a more rigorous connection between the
active voice and emphasis on the individual?
>>>> Thanks for the reference - by passive voice do you mean the
>>>> same as the ergative construction?
>
>No. An ergative construction is utterly different from a passive --
>though it is possible that *some* ergatives have developed out of
>earlier passives.
True.
>> Proto-Indo-European was also an ergative language (see Robert
>> S.P. Beekes, Comparative Indo-European Linguistics: An
>> Introduction (1995), pp. 193 and 254).
>
[snip]
>
>Anyway, Beekes's account of ergativity is *seriously* in error. He
>asserts that an ergative language is one which distinguishes agents
>from non-agents. That is false. A language that distinguishes
>agents from non-agents is an *active* language, not an ergative one,
>and active languages exist; examples are Crow and Eastern Pomo (and
>possibly Sumerian -- Miguel?)
I don't know about that. In learning Sumerian we think of the
ergative marker on the subject as marking the agent, the person who
causes the action to happen to the patient [or non-agent].
>Sent by: mcv@wxs.nl (Miguel Carrasquer Vidal)
>
>The categories animate/inanimate do play a role in Sumerian
>(inanimate nouns do not take a plural ending, different verbal
>prefixes are used for agreement) but overall, Sumerian is not an
>"active/stative" language.
>
>Sumerian is split-ergative in that personal pronouns are
>"accusative", i.e. they take a nominative ending when they are the
>subject of either a transitive or intransitive verb [the
>"nominative" ending is identical to the ergative ending -e]. Nouns,
>including inanimates, are ergative. Marking of the verb is highly
>complex, but generally ergative [*]. Johanna Nichols, in
>"Linguistic Diversity in Time and Space", classifies Sumerian as E
>(for Ergative).
>
>[*] <imagine smaller print here:>
>The details are very interesting and will look vaguely familiar to
>a Vascologist like Larry:
>
>1) Intransitive verbs of any aspect mark the absolutive subject by
>a set of suffixes: -en, -en, -{zero}; -enden, -enzen, -esh.
>
>2) Transitive verbs of the maru^ aspect (imperfective, usually
>present tense) mark the ergative subject by a set of suffixes: -en,
>- -en, -e; -enden, -enzen, -ene [note that only the 3rd p. differs
>from the absolutive suffixes]. The 3rd. person object is marked by
>prefixes: -n- for animnate, -b- for inanimate.
>
>3) Transitive verbs of the hamt.u aspect (perfective, usually past
>tense) mark the ergative subject by a set of prefixes: -{zero}-,
>- -e-, -n-/-b-. The plural forms add a suffix: -enden, -enzen,
>-esh. The 3rd. person object is left unmarked, except in the 1st p.
>sg. and pl., where -n- and -b- fill the empty prefix slot of the
>1st. p. ergative.
The most masterful discussion of the Sumerian maru^ verb forms is
Joachim Krecher, "Die maru^-Formen des sumerischen Verbums" in Alter
Orient und Altes Testament 240 (1995; Fs. vSoden II) 141-200. He
says the maru^ or imperfect forms are later than the hamt.u or
perfect verb forms, saying that there are examples of other languages
that took primary verb roots (such as the aorist or preterite) and
through enlargement, elaboration, or lengthening created secondary
verb roots for the present tense or the unfinished aspect. Krecher
gives examples from ancient Greek, Akkadian, and Old Georgian showing
how the present tense forms are longer forms of the aorist or
preterite forms.
John L. Hayes says on p. 140 of his grammar, "The term split ergative
has been applied to Sumerian, because the personal-affixes behave in
an ergative way in the hamt.u, but not in the maru^; in the maru^,
they behave in a nominative-accusative way."
This split could indicate that the ergative pattern is primary
because the hamt.u forms are primary.
There are several relevant comments about ergativity and
Proto-Sumerian in a recent monograph by Jens Hoyrup [where the o has
a slash], Sumerian: The Descendant of a Proto-Historical Creole?,
1992. "Analyzing the system and in particular the exceptions to the
"ideal" system, Jacobsen [ZA 78 (1988):161-220] offers a tentative
explanation of how the ergative system may have developed. As an
example he analyses the sentence lu2.e e2.NULL mu.n.du3.NULL. He
characterizes it as "passive" and translates it as "by the man [lu2]
the house [e2] was built [du3]", taking the suffix /-e/ on lu2 as an
originally locative-terminative mark." "constructions originally
concentrating on the state into which the logical patient (the house)
has been brought will have been refocused, with the consequence that
an originally locative-terminative "by" has been reinterpreted as an
ergative mark on persons."
Hoyrup says that in Sumerian, "the perfective aspect is also the
unmarked aspect, which implies that the underlying unmarked voice
will have been passive." "This may appear as a revival of the
classical understanding of Sumerian as a "passive" language, a notion
which has otherwise been replaced by the concept of ergativity.
Revival of the outmoded idea may indeed be called for by the
differentiation of syntactical and morphological ergativity and by
the observation of the various kinds of splitting; but it shall be
observed that Proto-Sumerian and not Sumerian is the language where
we could speak of an unmarked passive voice: the very development of
morphological ergativity makes this description obsolete."
It sounds to me like the early form of the language lacked forms for
nominative-accusative statements. The emphasis was on what was
happening to the patient. It is a more descriptive way of looking
at reality, whereas in identifying agents language is introducing
causative explanation. In this case what one sees with one's eyes
and mind actually is influenced by the categories of language.
Regards,
John Halloran
------------------------------
Date: Tue, 17 Feb 1998 22:17:13 -0700 (MST)
Subject: Re: EvolLang: Ancient & Tribal Languages/Thought
>Sent by: mcv@wxs.nl (Miguel Carrasquer Vidal)
>
>>John A. Halloran wrote:
>>
>>It sounds to me like the early form of the language lacked forms
>>for nominative-accusative statements.
>
>To the contrary. If the Sumerian ergative derives from a passive
>construction in Pre-Sumerian, then Pre-Sumerian *must* have been a
>nominative/accusative language (or an active/stative language with
>"accusative bias" as Nichols puts it). Passive constructions
>require nominative/accusative.
That is a strong and interesting statement. Nominative/accusative
means subject/object.
I find the following statement regarding the passive construction in
Arabic in Lyovin's book on Languages of the World:
"In Arabic passive constructions, the agent NP cannot be expressed.
(This is not a rare phenomenon among the languages of the world.) If
agent NP must be expressed, the statement must be reworded in active
voice.
"qutila Zaydun 'Zayd was killed'
"'Zayd was killed by Mohammed' cannot be expressed using passive
voice."
So in the Arabic passive voice, subject and object relations are not
present.
You want to interpret Pre-Sumerian as a modern language and I want to
interpret Proto-Sumerian as the language of early Neolithic humans
learning how to speak. I would interpret the passive in
Proto-Sumerian as equivalent to the two word sentences that Crystal
describes for children learning how to speak.
Let me give one more quote to shake up people who think that
linguistic habits in the past were no different from in the present.
"Problem of Names in the Proto-Uralic Language. There is no
linguistic/material basis to suppose that the speakers of the
Proto-Uralic language used names. Names have no connotation, they
always refer to individuals and not to classes. Both the personal
names and the geographical names are relatively new in the member
languages and/or branches of Uralia. Some name giving habits can be
supposed for the Finno-Ugric epoch." Gyula Decsy, The Uralic
Protolanguage: A Comprehensive Reconstruction (1990), p. 94. He
dates Proto-Uralic to 4,000 B.C. and Proto-Finno-Ugric to 3,000 B.C.
Regards,
John Halloran
------------------------------
Date: Wed, 18 Feb 1998 23:15:29 -0700 (MST)
Subject: Re: EvolLang: Ancient & Tribal Languages/Thought
>Sent by: mcv@wxs.nl (Miguel Carrasquer Vidal)
>
>>Passive constructions require nominative/accusative.
>>
>>So in the Arabic passive voice, subject and object relations are
>>not present.
>
>Which is interesting but irrelevant. Arabic is a nominative/
>accusative language, and it can rephrase a transitive construction
>as intransitive by moving the accusative object to the nominative
>position by using the passive.
Do you explain the lack of subject and object in the Finnish passive
construction in the same way?
>>You want to interpret Pre-Sumerian as a modern language and I want
>>to interpret Proto-Sumerian as the language of early Neolithic
>>humans learning how to speak. I would interpret the passive in
>>Proto-Sumerian as equivalent to the two word sentences that Crystal
>>describes for children learning how to speak.
>
>Yes, but why?
If Jacobsen thinks that the Sumerian ergative construction evolved
out of an earlier passive construction, then besides that passive
construction, what evidence for other verbal constructions from that
early period would you say is preserved in historical Sumerian?
If the maru^ imperfect verb forms originated as elaborations of the
earlier hamt.u perfect verb forms, do you have evidence for any verb
forms other than the perfect verb forms from that earlier period?
If not, would you say that a language with only passive construction
and perfect verb forms qualifies as a modern language?
Regards,
John Halloran
------------------------------
Date: Wed, 18 Feb 1998 23:38:29 -0700 (MST)
Subject: Re: EvolLang: Ancient & Tribal Languages/Thought
>Sent by: "Larry Trask" <larryt@cogs.susx.ac.uk>
>
>Apart from one or two divine names, we cannot reconstruct any names
>for Proto-Indo-European, either, but I have never seen anyone
>suggest that the speakers of PIE therefore didn't use any names.
If one follows the continuum from modern arbitrariness in the meaning
of names back to earlier concrete, descriptive names, one eventually
ends up at no names. The name of the oldest Sumerian deity was An,
but in Sumerian the word 'an' just means 'sky, heaven'. So yes it is
a name if a word for an importance part of existence is a name. As
far as I know, all early names were either existing words for object-
or action-classes or were compound juxtapositions of existing words
for object- or action-classes. If one goes back far enough, there
are no arbitrary names.
Language-using zoologists like Dian Fossey may give names to the
mountain gorillas that they study, but mountain gorillas do not give
names to each other, nor is it necessary for their existence that
they do so.
Regards,
John Halloran
------------------------------
Date: Tue, 24 Feb 1998 23:06:00 -0700 (MST)
Subject: Re: EvolLang: On defining `language' and `communication'
>Sent by: John A. Halloran
>
>>Sent by: mcv@wxs.nl (Miguel Carrasquer Vidal)
>>
>>What does "symbolic language" mean?
>
>Communication which divides up the world into discrete objects and
>actions and uses symbols to refer to these discrete parts.
One cannot determine when language originated if one cannot define
language.
I would suggest to Larry Trask that he assigns the transition from
nonlanguage to language to our remote hominid ancestors because he
has not clearly defined what makes language different from
nonlanguage. In the above definition of language, I stress how the
lexicon of a language cuts up reality into discrete pieces. As a
language's lexicon expands, reality is cut up into more and more
pieces. Syntax can only come into play when a number of symbols have
been created with which to refer to discrete aspects of reality.
>Earlier, Sir Richard Paget made a similar statement, "As to the
>uneducated born-deaf, the absence of any training in the art of
>giving a *name* to each element of the events which they experience
>(such as all hearing children obtain in their early days) has this
>effect: viz., that the born-deaf naturally appreciate events as a
>whole. The consequence is that they cannot direct their thoughts
>to the separate items of which (to our minds) the events we all
>experience are composed."
This is the key to the definition of language. In the absence of
language, reality is experienced as an unbroken whole. When humans
began experimenting with language, they began breaking reality up
into little pieces. The ability to do this is clearly a left-brain
skill just like making a stone tool requires the left-brain to break
the task up into a separate series of steps.
To date when this occurred, one must look at transitions in the
mentality of humans. When did humans start to show analytical
capabilities missing from the uneducated born-deaf?
The fluid, naturalistic cave paintings of Paleolithic southern France
do not show analysis into discrete components, rather those paintings
are wholistic products of the right-brain, not of the analytical,
segmenting left-brain.
In the course of looking in Alta Vista for Sir Richard Paget, I ran
into a very interesting paper by Robin Allott on gesture and the
origins of language. The URL is:
http://www.percep.demon.co.uk/pfolpt1.htm
This paper would interest Sherman.
Robin summarizes many earlier writers on the mouth-gesture origin of
language theory, providing evidence from Polynesian, Indo-European,
etc. for the same theory of language origin that I arrived at from
analyzing the Sumerian vocabulary, at which time I knew of none of
these earlier findings.
Regards,
John Halloran
------------------------------
Date: Sat, 28 Feb 1998 14:55:23 -0700 (MST)
Subject: Re: EvolLang: On defining `language' and `communication'
>Sent by: "Jeremy Ottevanger" <ottevanger@msn.com>
>
>I have proposed several times on this and other lists the hypothesis
>that certain aspects of human thought may rely upon the presence of
>grammar - not specifically that thought process must use words, but
>that aspects of the mental grammar cannot be present without their
>analogue in language. This is a hypothesis as far as I'm concerned,
>not a belief. My hope was that, if this is so, one could isolate
>which aspects of language grammar are necessary for the performance
>of which mental operations, one could look for these operations in
>activities. I was inspired here by Thomas Wynn's work. I hoped that
>the chaine operatoire of stone tool manufacture, as revealed by the
>archaeological record and by experimental archaeology, could
>therefore reveal which aspects of linguistic grammar had to be
>available/developed by a given date (according to the technology
>observed at that date). So I liked your first paragraph, but would
>suggest that your own comments about stone tool manufacture might
>imply that certain aspects of language grammer may have been present
>far earlier than the time of palaeolithic cave paintings, if this
>hypothesis is correct.
>
>What do you think? I'm still looking for a linguist or psychologist
>that is willing to go out on a limb and try to develop the idea
>further, to see if there is merit in it, and possibly to open
>themselves to ridicule with a publication. Any takers?
Jeremy, in the book The Axemaker's Gift [by James Burke of TV
'Connections' fame and psychologist Robert Ornstein] it is said that
Gordon G. Gallup, Jr. [an evolutionary psychologist at New York
University at Albany] "has analyzed the limb-movement sequences of
tree-dwelling new-world monkeys and noticed "a kind of grammar" in
their motions, a succession of actions that must be made in the
correct sequence. After the original move out onto the savannah, the
underlying brain structure that originally evolved for all the
sophisticated, sequential swinging-around was then available for
other uses." p. 21.
This clearly says that syntactic ability, the ability to connect
elements to form a sequence, predates both tool-making and language.
I am sorry that I cannot give you a more precise reference for
Gordon's work - his academic web page only has his recent
publications in the last two years. But this could provide a
helpful direction for your research.
Regards,
John Halloran
------------------------------
Date: Sat, 28 Feb 1998 17:01:58 -0700 (MST)
Subject: EvolLang: Re: On defining `language' and `communication'
>Sent by: Jose-Luis Mendivil Giro <jlmendi@posta.unizar.es>
>
<snip>
>
>Language can, of course, have a cognitive function, but I think that
>even Saussure would deny that thought is not possible without
>language and, of course, he would deny that the uneducated born-deaf
>'naturally appreciate events as a whole' and that they 'cannot
>direct their thoughts to the separate items of which (to our minds)
>the events we all experience are composed', as assumed in Halloran's
>message. I, personally, can conceptualize things I cannot name.
You are right that I need to avoid stepping into the trap of making
language more important for thought than it is. The last quote is
from Paget from almost 50 years ago and more research has been done
into the uneducated born-deaf since then - I need to look in more
detail at the late 60's research by Hans Furth.
In The Oxford Companion to the Mind [ed. Richard L. Gregory 1987]
there is an article Language: Learning Meanings which discusses how
children learn language. It says, "What is clear is that, from a
very early stage, a child may use either perception of similar form
or knowledge of common function (or both) as the basis for a
generalization of word meaning -- a generalization which may go
beyond the limits of the accepted adult norm." p. 422. For example,
initially the child may speak the word 'ball' upon seeing a balloon,
an Easter egg, a small round stone, etc. Alternately 'ball' may be
applied to any object able to be bounced or thrown.
It would be too strong to say that a creature needs a symbolic word
in order to be able to group items according to form or function.
My dog does not need the English words 'bird' or 'squirrel' to react
very differently to squirrels than he does to birds. The mind of
any intelligent creature will naturally order the objects of its
experience according to form or function in the course of maturing.
This is basically Hans Furth's conclusion, "The evidence for
conceptual thinking in the linguistically deficient deaf has been
presented and leads to the direct conclusion that thinking develops
through living contact with the environment regardless of the
presence or absence of a ready-made linguistic symbol system." Hans
G. Furth, Thinking Without Language: Psychological Implications of
Deafness (1966). pp. 198-99. Saussure was correct in his conclusion
that linguistic symbols are not necessary for thought.
I think Paget's point is that language trains the young mind to
order the flow of experience into pieces as understood by the
culture in an artificially amplified manner which does not happen
without language. The culture carries the child's mind farther and
in a particular direction. It does not even have to be a child's
mind. What happened when you studied linguistics at the university?
Wasn't your mind taught how to cut up language into pieces and
weren't the words assigned to those pieces a necessary part of your
education? Could a person understand the inside of the human body
without the aid of a large medical vocabulary? Possibly one could
learn to group items on the basis of form or function, but by
learning words and categories invented by our predecessors we are
saved much time.
Berlin and Kay divided almost a hundred languages into eight types
according to how many color terms the languages have.
1. white, black
2. white, black, red
3. white, black, red, green
4. white, black, red, yellow
5. white, black, red, green, yellow
6. white, black, red, green, yellow, blue
7. white, black, red, green, yellow, blue, brown
8. white, black, red, green, yellow, blue, brown, purple and/or
pink and/or orange and/or grey.
[embarrassingly for this hypothesis, Russian and Hungarian contain
twelve basic color categories, having respectively two terms for
dark and light blue or for dark and light red]
"Berlin and Kay add a further "evolutionary" hypothesis, which
states that the types of vocabulary as ordered above represent a
fixed sequence of historical stages through which a language must
pass as its basic vocabulary increases." Geoffrey Leech, "Colour and
Kinship: Two Case Studies in "Universal Semantics"" in Language,
Culture, and Cognition: Anthropological Perspectives, ed. Ronald W.
Casson, 1981.
Color terms provide a good example of how a culture provides lexemes
to help its members make categorical distinctions. They also show
that languages pass through stages of historical complexity in the
number of lexemes that they use. Larry Trask wanted to know what is
a 'modern language', thinking that that indexical term has no
referent. What other index can one use to refer to a language that
has evolved through a number of stages away from ancient or primitive
languages such as those that belong to Berlin and Kay's types 1 and
2? Is Larry criticizing the term as an imprecise index, or is he
denying that languages pass through historical stages? I guarantee
you that Sumerian as an ancient language lacked the metalanguage
lexemes that we are using on this list. The inhabitants of this list
would have been unable to argue about nouns and verbs in Sumerian
because Sumerian just had the word 'inim' for 'word', but not
separate words for 'noun' or 'verb'. I think that the size of a
language's lexicon is actually a very good indication of how
advanced its speakers are. By learning the lexicon of an advanced
language, individuals can accelerate their own conceptual
sophistication.
I will also go so far as to suggest that if one could determine the
size of the most-advanced written lexicons that we know about at
different points in time going back to about 2000 BCE and plot those
sizes against the respective dates in a statistical curve fitting
program that one would end up with a function that would say pretty
accurately at what date anyone's most-advanced lexicon had a value of
zero.
I am asserting that vocabularies have been increasing in complexity
and in their number of entries since humans invented language. I
don't think there is any evidence for long periods of equilibrium or
quiescence in vocabulary formation.
Regards,
John Halloran
------------------------------
Date: Sun, 1 Mar 1998 16:16:38 -0700 (MST)
Subject: Re: EvolLang: On defining `language' and `communication'
>Sent by: "Peter T. Daniels" <grammatim@worldnet.att.net>
>
>> This clearly says that syntactic ability, the ability to connect
>> elements to form a sequence, predates both tool-making and
>> language. I am sorry that I cannot give you a more precise
>> reference for Gordon's work - his academic web page only has his
>> recent publications in the last two years. But this could provide
>> a helpful direction for your research.
>
>They simply repeat the W. H. Calvin error of believing that syntax
>is a matter of sequencing. Syntax is far more than sequencing; it is
>reference and coreference, anaphora and ellipsis, and much, much
>else.
"Syntax: The study of the rules governing the way words are combined
to form sentences; contrasts with morphology, the study of word
structure." David Crystal, An Encyclopedic Dictionary of Language and
Languages, p. 379.
If language were expressed through visual pictures of multiple
elements then sequencing would not be important. In human language
sequential actions and sequential memory are very important.
Contradicting your point, anaphora and ellipsis are good examples
where the meaning depends on the listener's sequential memory of what
subject was just being discussed.
The ability to form the precise sequences necessary to achieve the
desired meaning-end-result is evident even in roots and their
morphology, because we are talking about sequences of sounds. It is
nothing but the sequence that communicates the different meaning of
'dog' versus 'god'. [Remember that before writing words could not be
taken in visually at a single glance.] You can certainly remind us
that learning the syntax of a particular language involves learning
how it makes syntactic references, but do not imply that evolution
of mental ability for remembering and acting out specific sequences
was unimportant as a preadaptation for language.
Regards,
John Halloran
------------------------------
Date: Mon, 2 Mar 1998 17:18:31 -0700 (MST)
Subject: Re: EvolLang: On defining `language' and `communication'
>Sent by: William Edmondson <W.H.Edmondson@cs.bham.ac.uk>
>
>>If language were expressed through visual pictures of multiple
>>elements then sequencing would not be important. In human language
>>sequential actions and sequential memory are very important.
>
>Presuming he means that sequence is an inherent property of
>language, then it is simple to demonstrate that this is factually
>incorrect.
My thanks to William, Peter, Larry T., and Larry G. for their
responses. They forced me to do more research, in the course of
which I came up with the following quote from a linguist, Matthew
Saxton, discussing Theories of Grammar:
"With this [recent, i.e., 1993] shift towards lexical determination
of sentence structure [in theories of grammar], the role of purely
syntactic phenomena in determining grammaticality may prove
ultimately to be quite limited."
I have no problem with emphasizing the importance of concepts at the
expense of syntax. Once again this emphasizes the importance of
thought which is preverbal in nature, as the verbal part of a lexeme
is like the tip of the iceberg.
The human ability for fine control of sequential motor actions
appears be more important for producing linguistic communication
than it is for mental sentence formation. Judging from the ability
for both speech and signs to express language, we appear to believe
that the means of articulation are of only secondary importance to
the mental association processes that the symbols of language
express.
Regards,
John Halloran
------------------------------
Date: Mon, 2 Mar 1998 17:18:31 -0700 (MST)
Subject: EvolLang: Re: stadialism
>Sent by: "Larry Trask" <larryt@cogs.susx.ac.uk>
>
>Stadialist ideas were largely the creation of 19th-century European
>linguists who noticed that non-European languages were often
>different from European languages. Operating under the common
>19th-century perception that race, culture and language were part
>and parcel, and of course regarding themselves as the most advanced
>types on the planet, they reached the conclusion that the languages
>of those po' li'l colored folks were by definition less advanced and
>more primitive than their own. This stuff has thankfully
>disappeared from linguistics, as language after language has proved
>to be every bit as rich and complex as English or German.
I knew there was a reason to buy that unabridged dictionary. I can't
tell from the above what is the etymology or meaning of 'stadialism'.
[snip]
>> By learning the lexicon of an advanced language, individuals can
>> accelerate their own conceptual sophistication.
>
>There are no "advanced languages". There are only languages with
>more elaborated vocabularies. And any language can have its
>vocabulary elaborated if its speakers choose to do this: Finnish and
>Basque are just two examples of languages which, in this century,
>have changed from being local vernaculars with no elaboration to
>fully elaborated languages in which it is possible to speak, and
>write, on any subject at all.
You do not deny that speakers of a pidgin language get by with just
100 words at the earliest stage of a pidgin language, do you? All
I am saying is that languages evolve from limited lexical sets.
Study of creole languages that evolve from pidgin languages show
that they do so very quickly, within one or two generations.
Admittedly at the time that people began to speak, their brains
probably were not selected for speech to the extent that humans are
today, but the process of creating a fully functional language
probably did not take more than three or four generations.
All I have ever claimed is that I can see in the vocabulary of
Sumerian a core of about 200 words of V, VC, and CV structure that I
believe were the original words of that language. The lexicon
expanded over the generations from that base. The Sumerians loaned
many of their words into other languages. The pool of words and
therefore concepts available to speakers today has been accumulating
over the centuries.
It is very interesting that you say Finnish and Basque 'changed',
instead of saying that they 'evolved'. Bichakjian noticed that
linguists prefer to use the term 'linguistic change' instead of
'linguistic evolution'.
Regards,
John Halloran
------------------------------
Date: Tue, 3 Mar 1998 23:19:30 -0700 (MST)
Subject: Re: EvolLang: Re: stadialism
>Sent by: "Dr. John E. McLaughlin" <mclasutt@brigham.net>
>
>> Sent by: John A. Halloran
>>
>> It is very interesting that you say Finnish and Basque 'changed',
>> instead of saying that they 'evolved'. Bichakjian noticed that
>> linguists prefer to use the term 'linguistic change' instead of
>> 'linguistic evolution'.
>
>There is an important terminological distinction here. "Change" is
>the shift from one type of fully modern language (which includes all
>known recorded languages, both ancient and modern) to another.
>"Evolution" is the development of modern human language from its
>ancestors (none of which are either attested in the written record
>or extant today).
>Sent by: "Larry Trask" <larryt@cogs.susx.ac.uk>
>
>Yes, this usage is standard. We prefer to avoid `evolution' in this
>context because of its unfortunate associations. Languages are
>always changing, but they do not thereby become better, nor do they
>become significantly different in nature as a result. Whales
>evolved from four-legged land mammals, from which they differ
>greatly. Italian is *descended* from Latin, but we do not like to
>say that Italian "evolved" from Latin.
I am finally starting to appreciate what John L. and Larry T. are
saying. There is a reason why all the languages that we know of
appear functionally modern. There is a reason why we can see the
evolution of automobiles, but not the evolution of languages,
although both are human inventions. The reason is that language
expresses human thought, and the ability for thought is shared
equally by all human populations. Language is different from
other human inventions and tools because it must be able to match
itself to whatever thoughts humans are capable of thinking.
The need for language to mirror thought explains why, although
language could be nothing more than a tool for communication, humans
who start with a simple pidgin language feel and act upon the need
to elaborate the language into a full-fledged creole language so
quickly.
At the same time, you guys are setting up a Catch-22 situation as
far as the scientific study of language evolution is concerned. You
are saying that over the time period for which we have scientific
evidence there has only been language change, not language evolution.
By putting language evolution outside the time period for which we
have evidence, doesn't that reduce its study to unscientific armchair
theorizing?
The way out of this quandary is to focus on the lexicon. One can
maintain the 'modernity' of the manner of thinking expressed in a
particular language while recognizing that there has been a process
of escalating abstraction in the concepts and symbols available to
the speakers of languages. Starting from a base of referents to
concrete objects and actions, we have added increasingly abstract
vocabulary items. This has a parallel to what happens in the brain;
the individual evolves feature detectors which have as their input
the features detected by lower level feature detectors; abstract
thinking ends up depending on several layers of feature detectors.
At this time historically, the discipline of linguistics itself
appears to be shifting attention away from grammar and towards the
lexicon.
Until recently, there has been little interest on the part of
linguists in estimating the size of early language vocabularies as
of different dates. In The Indo-European Protolanguage: a
Computational Reconstruction (1991), Gyula Decsy says, "Concerning
the size of the Indo-European vocabulary we should keep in mind that
at least 20-25% of the comparative sets in a protolanguage dictionary
can be regarded by scholars representing different schools as
doubtful or uncertain. This should be considered in judging the
Indo-European - English Word Index printed in this book. It is
surprising how little attention was paid to Walde's and Pokorny's
works and the Proto-Indo-European vocabulary generally by
Indo-Europeanists of the last decades. In important handbooks, we do
not even find estimates about the size and semantic composition of
it. The reason is the extreme enthusiasm with which most
Indo-Europeanists focused on theoretical problems of phonetics,
especially in connection with the laryngeal theory." p. 15.
On p. 14, Decsy says that proper Indo-European (without Hittite)
which dissolved around 3,000 B.C.E. "consists of the 2563 words
listed in the Proto-Indo-European - English Word Index in this book
below as well as of the ca. 200 pronominal forms (partly inflected)
and derivational and grammatical suffixes treated in the part
Morphology." On p. 13, he says that "4,000 B.C. corresponds
chronologically to the dissolution of the Uralic Protolanguage (which
had 472 reconstructed words), and 3,000 B.C. to the dissolution of
the Finno-Ugric Protolanguage (ca. 1,200 words, including the Uralic
ones)." Those vocabulary item count numbers are quite consistent
with the numbers for the Proto-Afroasiatic vocabulary and the
non-compound logogram Sumerian vocabulary.
Do you see that one can meaningfully focus on the *evolution* of
the lexicon into the 100,000 plus entries of modern dictionaries
without insulting the modernity of the thinking processes of our
ancestors, who while historically ancient were evolutionarily modern?
Regards,
John Halloran
------------------------------
Date: Fri, 6 Mar 1998 12:50:58 -0700 (MST)
Subject: EvolLang: Re: stadialism
>Sent by: Larry Gorbet <lgorbet@unm.edu>
>
>John A. Halloran wrote, concerning remarks by Dr. John E. McLaughlin
>on the absence of evolution of language in the recent historical
>past:
>
>>For you then, language evolution happened during a brief
>>moment in time, kind of like the Big Bang, and has not occurred
>>since.
>
>Unless I really missed something, this characterization seems a
>little distorted, calendar-wise. The "brief moment in time" is
>possibly well over a million years (from the first changes in
>hominid communication that made it more language-like to the
>existence in some hominid groups of "modern language"). The
>"since" is almost certainly quite a bit less, maybe as little as a
>few tens of thousands, probably no more than a little over 100,000
>years.
You missed the insight that evolution has made all humans modern in
the way that they think, and that once language becomes not just a
tool for communicating with other people, but becomes integrated with
thought, that its forms and capabilities rapidly elaborate to be able
to express all the types of object relations that modern people are
capable of thinking. Having language accompany thought is a nothing
or all proposition.
This is the reason why it is practically impossible to observe
evolution in the grammar of language.
Regards,
John Halloran
------------------------------
Date: Fri, 6 Mar 1998 12:50:58 -0700 (MST)
Subject: Re: EvolLang: Out on a Limb
>Sent by: "Dr. John E. McLaughlin" <mclasutt@brigham.net>
>
> and 5) the size of the lexicon of the average speaker shows no
>evidence of having increased over the last five millenia due to any
>external factors (in other words, my mother has some new computer
>terminology, but she doesn't know half the farming terminology
>that my grandmother knew).
That is completely wrong. Languages today have many more abstract
concept words than did languages of four or five thousand years ago.
Until 1963, what word did any language have to express the 'Catch-22'
concept? That word is not related to new technology. Four or five
thousand years ago languages mainly had words for concrete objects
and actions. Sure there was some abstraction, but abstract concepts
and words have been incrementing over time at a rapid rate. The
difference between the lexicons of Biblical Hebrew and Israeli
Newspaper Hebrew is not all new technology.
Earlier you made a semi-valid point about how for some reconstructed
languages we lack words for concepts that we are sure must have been
present in those languages. The only objective way that I can think
of to judge how well we have recovered most of a particular
language's lexemes is by trying to put together a Swadesh 100-word or
200-word list and seeing how close we come to 100 or 200 items (while
appreciating that some of Swadesh's words, such as 'swim', may not
apply to the culture).
Regards,
John Halloran
------------------------------
Date: Fri, 6 Mar 1998 12:50:58 -0700 (MST)
Subject: Re: EvolLang: Re: stadialism
>Sent by: "Dr. John E. McLaughlin" <mclasutt@brigham.net>
>
>> Sent by: John A. Halloran
>
>> What you define as language evolution can be studied in
>> modern times by observing how pidgin languages become creole
>> languages.
>
>I don't think so. Pidgins are not languages ex nihilo. Peter
>Daniels and I had a discussion a while back (just before Christmas)
>about how exposure to even a little writing changes people's
>perception of the sound stream. (If I'm not being fair to your
>view, Peter, I'm sorry and will accept correction.) While I'm
>pretty stubborn, he has a very good point. Human language evolved
>in a vacuum -- there was nothing in the world like it. These
>conditions NEVER existed again. Creoles don't develop in a vacuum
>either. By definition, they must arise from a pidgin situation when
>that pidgin becomes a native language. Both pidgins and creoles are
>generally assembled from pieces of the substrate and superstrate
>languages, not from original invention. This is NOT language
>evolution.
I accept the correction. That is a valid point. It is much harder
to invent a way of doing things than it is to borrow it.
However, that is exactly the reason why I think that at the time that
the Sumerians invented the vowel-only words A, U, O, and E to mean
'water', 'plant food', 'sleep, intercourse', and 'house, shelter'
that they had not come into contact with other language speakers. My
reasoning is that they would not have started with vowel-only words
if they had already been exposed to language speakers who were using
more structurally-complex words.
>We can't study any lexicon of a non-modern human language. Every
>single language we have presently extant or recorded during the 5000
>some-odd years of writing, and every reconstructed language to date
>is a fully modern human language. The lexicon has not "evolved"
>during the last 10,000 years. Lexicons have changed over time
>during the histories of individual languages to reflect changing
>environments, cultures and technologies, but every single lexicon
>shows exactly the same thing -- a complete capability to express
>everything that a particular culture needed to express.
I was willing to go along with you when you said that the way that
all peoples think and express their thoughts in language are equally
modern. But if you are going to say that all lexicons are equally
modern, I am not willing to go along with you. 'How' people think
has possibly not changed, but 'what' people think has evolved along
with culture. You cannot seriously say that human culture has not
evolved from the Paleolithic to 1998.
Regards,
John Halloran
------------------------------
Date: Mon, 9 Mar 1998 00:49:35 -0700 (MST)
Subject: Re: EvolLang: Lexicon evolution
>Sent by: John A. Halloran
>
>>Sent by: "Dr. John E. McLaughlin" <mclasutt@brigham.net>
>>
>> and 5) the size of the lexicon of the average speaker shows no
>>evidence of having increased over the last five millenia due to any
>>external factors (in other words, my mother has some new computer
>>terminology, but she doesn't know half the farming terminology
>>that my grandmother knew).
>
>That is completely wrong. Languages today have many more abstract
>concept words than did languages of four or five thousand years ago.
>Until 1963, what word did any language have to express the
>'Catch-22' concept? That word is not related to new technology.
>Four or five thousand years ago languages mainly had words for
>concrete objects and actions. Sure there was some abstraction, but
>abstract concepts and words have been incrementing over time at a
>rapid rate. The difference between the lexicons of Biblical Hebrew
>and Israeli Newspaper Hebrew is not all new technology.
Following is the evidence to back up this claim. It is taken from
Ernest Klein, A Comprehensive Etymological Dictionary of the Hebrew
Language For Readers of English (1987: Macmillan). This dictionary
of 30,000 entries uses the initials NH to note an entry that is only
found in New Hebrew. I will reference all the NH marked words from
the first six pages of the letter B. Many are abstract concept
words. Can Larry Trask comment on the type of words that Finnish and
Basque had to add to their lexicons in order to change from being
local vernaculars to modern, elaborated languages?
BAUT: representation.
B'AQRAY: accidentally.
BA'SHAN: cleome (plant).
BAESHET: halitosis.
BAV: gateway.
BUBAH: doll, puppet.
BABONAG: camomile.
BUBONET: dolly.
BUBAT.RON: marionette theater.
B'VAQASHAH: please.
BUBATIY: doll-like.
BUBATIYUT: dollishness.
BIGUD: clothing.
BAGIYR: adult.
B'GIYRAH: adolescence.
BAGIYRUT: adulthood.
BADAIY: lying, invented.
B'DOD: loneliness.
BIDUD: separation.
BIDUAH.: entertainment.
BEDVIY: bedouin.
BIDUL: distinction.
BADUN: canvas hut.
BADUNAG: linoleum.
BIDUQ: control, censorship, overhaul.
BIDUR: entertainment, recreation.
BADUR: entertained.
BIDURI: entertaining.
B'DUT: falsehood.
B'DUTAH: fiction.
BADEAH: playful, facetious.
BIDAHON: jesting.
BADH.IYT: farce.
BADH.ANUT: jesting.
BADH.ANIY: amusing.
BADIYD: a small branch of a tree.
BIDYON: fantasy, invented story.
BIDYONIY: fantastic.
B'DIYH.AH: joke, witticism.
BADYAN: one who invents stories.
B'DELUT: separation, detachment.
B'DALH.IY: crystal, crystalline.
BADLAN: separatist, isolationist.
BADLANUT: separatism, isolationism.
BADLANIY: separatist, separatistic.
BADAQ: censor.
BADQAN: a meticulous or overscrupulous person.
BADQANUT: meticulousness.
BADQANIY: meticulous, overscrupulous.
BIDOQET: censorship.
BADRAN: entertainer, artiste, comedian.
BADRANUT: entertaining, entertainment.
BIHUM: brutalization.
BAHUQ: shining, bright, flashing.
B'HIYAH: astonishment.
BEHAM: animal driver.
BAHAQAN: vitiliginous.
BAHAQANIY: bright, white.
BOESH: skunk.
BOGDANUT: treachery.
BOGDANIY: treacherous.
BODED: single, lonely; far away.
BOHEQ: shining, glittering.
BOZZAN: robber.
BOH.EN: examiner, tester.
BOH.ER: voter, elector.
BOT.EH: biting, sharp, strong.
BUL: (postal) stamp.
BUL: bull's-eye.
BULAUT: stamp collecting, philately.
BULAY: stamp collector.
BULAIY: philatelic.
BUM: boom.
BONEH: beaver.
BONNUT: insight (psychology).
BU`IT: a small bubble, a small blister.
BO`ER: burning; most urgent.
BUTSIYN: Verbuscum (botany).
BOTSIYTS: Butomus (botany).
BUQIYTSAH: elm.
BORIYT: Saponaria (botany).
BURSAN: stockbroker.
BORRUT: arbitration, arbitrage.
BAZBUZ: finch.
BIZBUZ: waste, extravagance.
Hebrew is a valuable language for tracing the progressive evolution
of lexemes and concepts down through historic times. Klein marks
many words, such as [A]ORLOGIYN, 'clock', as PBH, for Post Biblical
Hebrew. He marks other words, such as BE[A]UR, 'explanation,
commentary', as MH, for Medieval Hebrew. Many words are marked as FW
such as AZBEST.: 'asbestos', to indicate a recent Foreign Word. I
have another Hebrew dictionary, the Langenscheidt Pocket Hebrew
Dictionary to the Old Testament, which claims in just 8,000 entries
that it "contains the whole vocabulary of the Hebrew Old Testament".
Admittedly some of those entries could be expanded into additional
separate entries for participial forms. Do Peter or Izzy have other
figures for the number of vocabulary items in the Hebrew Bible? Lest
anyone get the wrong idea about the number of words in modern Hebrew,
Reuben Alcalay's 1974 The Complete Hebrew-English Dictionary, which
includes many technical and scientific terms, claims over 120,000
entries. So, in these lexicons of the same language separated by 2500
years, we have fairly complete guides to the evolution of mankind's
lexicon in general.
Recently, Peter remarked his belief that there would be 2000 Sumerian
words starting with the letter A alone when the Pennsylvania Sumerian
Dictionary finishes publishing the letter A. This number seemed high
to me, so I went to Steve Tinney's on-line Index to the Secondary
Literature, which is an up-to-date list of all the Sumerian words,
and references discussing them, on which the PSD project is drawing.
I constructed a search which obtained all words starting with the
letter A, which I saved to file. I then wrote a little program which
trimmed away all duplicate word entries, leaving about 1400 entries
starting with A, within which there are still a few duplicates. If
Peter or anyone wants to receive this file by e-mail, please contact
me.
In the compound logogram section of my own Sumerian lexicon, there
are 111 entries starting with the letter A, out of 1120 such entries,
for a total of about 10% of all Sumerian words. This indicates that
the total number of words that the Pennsylvania Sumerian Dictionary
is presently working with is not greater than 14,000. This includes
all words, even the most technical terms, from all periods, from the
language of the most advanced civilization of the third millenium
B.C. They are drawn from a staggeringly large corpus of texts. I'm
not sure how many tablets we have in total, but the largest corpus of
texts, taken from Girsu, was 400,000 texts. To compare apples to
apples, one can only compare the PSD's comprehensive collection of
14,000 Sumerian terms to a modern unabridged dictionary such as
Webster's Third International Dictionary with 450,000 headwords.
4,000 years of cultural and lexical evolution separate the two sets
of vocabularies.
Regards,
John Halloran
------------------------------
Date: Tue, 10 Mar 1998 11:33:45 -0700 (MST)
Subject: Re: EvolLang: Out on a Limb
>Sent by: "Peter T. Daniels" <grammatim@worldnet.att.net>
>
>> John A. Halloran wrote:
>>
>> Earlier you made a semi-valid point about how for some
>> reconstructed languages we lack words for concepts that we are
>> sure must have been present in those languages. The only
>> objective way that I can think of to judge how well we have
>> recovered most of a particular language's lexemes is by trying to
>> put together a Swadesh 100-word or 200-word list and seeing how
>> close we come to 100 or 200 items (while appreciating that some of
>> Swadesh's words, such as 'swim', may not apply to the culture).
>
>This is total nonsense. What is our database for the size of lexicon
>of a language of "4 or 5000 years ago"? 2 or 1 languages
>respectively (Akkadian and Sumerian). How well are the lexica of
>those languages at that period preserved? Pretty miserably.
Peter, sometimes you express yourself in an unfortunate qualitative
and emotional manner. It doesn't matter how many languages from 4
or 5000 years ago we have in our database if what I am interested in
is the most advanced lexicon at that time. How many Sumerian tablets
and how many Akkadian tablets would you estimate have been recovered?
Recently I quoted a figure of 400,000 tablets taken from Girsu. That
would be from the mid to early 3rd millenium BCE. 5000 years ago is
not doable from written records, but 4000 years ago is. Do you know
Sumerian and Akkadian well enough to say that there are concepts
which they must have had for which we lack words?
If so, what are they?
>And are you suggesting that they never saw fish locomoting in
>Mesopotamia? There are certainly enough of them depicted on seals.
Take it up with Robert Englund. He says that swimming was not and is
not part of Middle Eastern culture. He and I worked together on words
that Miguel Carrasquer Vidal was missing from his 200-word Swadesh
list of Sumerian. I will quote the relevant part of my message to
Miguel:
>swim: ? i7.de3 IL2 or such, 'float'.
>He says the Sumerians didn't swim, similar to people nowadays in the
>Near East, perhaps because of parasites. This despite the Tigris
>and Euphrates being slow-moving broad rivers. I should have asked
>him what fish are described as doing in the water. But since he
>wrote that book on fishing in the Ur III period, I think he would
>have known of any such word if it existed.
Of the missing words presented to Englund, this was the only one on
which we had to put a question mark. Later I saw Miguel's list and
he had 15 question marks out of 200 words. Looking at his list some
of those question marks do not need to be on the given Sumerian words
or blanks can be filled in from entries in the newer version of my
lexicon [such as 156 TO SUCK = sub; su-ub]. So I would question your
statement that Sumerologists have recovered the lexicon "pretty
miserably".
Regards,
John Halloran
------------------------------
Date: Tue, 10 Mar 1998 13:11:06 -0700 (MST)
Subject: Re: EvolLang: Out on a Limb
>Sent by: "Dr. John E. McLaughlin" <mclasutt@brigham.net>
>
>nauseum." You get a good list of personal and place names, a good
>list of economic and trade items, a lot of material culture, but
>very little thought or personal vocabulary. What is the Sumerian
>for the six basic needs of man? "Eat, drink, sleep, p.[urinate].s,
>s.[defecate].t, f.[copulate].k" The last one is especially hard to
>find in any early written records (and I don't mean euphemisms).
>What's the Sumerian word? I don't know the records, but I'd be
>willing to bet the last three weren't recorded (except
>euphemistically, which is NOT the way average people usually talk).
Those concepts are all in my Sumerian lexicon. My bibliography refers
to a source by Josef Bauer where he lists u5 and translates as "coire"
with a very clear animal example. This is one of those U words that
Stephen J. Lieberman argues was actually pronounced O. One also
finds other words and euphemisms for sex, although I admit that they
are not all in my lexicon at this point. I encourage you to visit
and look for those words or any other words in which you are
interested.
Regards,
John Halloran
------------------------------
Date: Tue, 10 Mar 1998 13:10:54 -0700 (MST)
Subject: Re: EvolLang: Re: stadialism
>Sent by: "Dr. John E. McLaughlin" <mclasutt@brigham.net>
>
>> Sent by: "Patrick C. Ryan" <proto-language@worldnet.att.net>
>>
>> We know there are ergative-type languages and nominative-type
>> languages. But when we see a language that combines both
>> ergative- and nominative-type traits, is it not legitimate to
>> think of an intermediate stage of an evolution from ergative-type
>> to nominative-type?
>
>This is change, not evolution. The language is just as capable of
>expression whether it is ergative, nominative, or somewhere in
>between. Nominative is not somehow more advanced than ergative.
Johanna Nichols has found that when given a choice, multilingual
speakers pick nominative-accusative over ergative or other
non-accusative alignment types. She does not find progressive
evolution of morphosyntactic structures within a particular language,
but she does find that in residual zones of contact between languages
of different types that "greater morphological complexity,
cliticization or other increase in head-marking patterns, and
propensity for *accusative alignment* [my emphasis] and SOV
word order evidently arise as speakers select from the inventory of
grammatical patterns made available to them by multilingualism." So
selection does favor these characteristics, but can only operate when
alternatives are available. She continues, "This kind of evolution
yields a standard statistical profile for certain features in each
individual residual zone, but there is not reason to think it affects
language generally." Linguistic Diversity in Space and Time, p. 276.
Would anyone care to argue that evolution through natural selection
is not evolution?
Regards,
John Halloran
------------------------------
Date: Fri, 13 Mar 1998 14:07:17 -0700 (MST)
Subject: Re: EvolLang: Out on a Limb
>Sent by: "Patrick C. Ryan" <proto-language@worldnet.att.net>
>
>>Sent by: "Dr. John E. McLaughlin" <mclasutt@brigham.net>
>
>Just, that is true. But if I see text after text that says "I place
>in your hand" where we would say "endow", it is legitimate to
>conclude that we are qdealing with a simpler stage of language.
Pat, that reminds me of how in Hebrew the same word, af, means both
'nose' and 'anger'. Dr. Hetzron told me they visualized someone's
nose getting red or hot and that meant anger.
>>But did Sumerian really lack abstract concepts? No, it didn't.
No one said that Sumerian lacked abstract concepts. As a language,
Sumerian had at least a five thousand year history. Of course the
Sumerians came up with abstract words. If you look at the vocabulary,
the noncompound words, which must be earlier in time than the
compound words of which they are elements, refer to objects and
actions that are more basic, more concrete, and more appropriate to
earlier time periods than do the compound words, which refer to more
sophisticated and abstract concepts that come from a diachronically
later period.
Sumerian for 'neck' is gu2.
Sumerian for 'door, entrance' is ig.
Sumerian for 'eyes, face, front' is igi.
igi, 'eyes', is probably reduplicated ig, 'door, entrance', with
elision of the final 'g', a common phenomenon in Sumerian, cf.,
lalla, 'shortage, deficiency', from reduplicated lal.
How far are we from the origins of language with a language that
derives the word for 'eyes' from the word for 'door'?
'Neck' and 'door' derived their meanings around the 8th millenium
B.C., in my opinion, from the way those word sounds are made down
in the throat.
>>I'm perusing Halloran's Sumerian web page while I write this and
>>I find roots form for "totality", "plant", "death", "joy",
>>"satisfaction", "offense", "work", "reliable", and "benefit". And
>>that's just on the first three pages of his site. I wasn't even
>>looking very hard, just skimming. That's a pretty high number of
>>abstract terms for a language that is preserved mainly in economic
>>records.
You're making statements about Sumerian texts off the top of your
head. Bendt Alster just published a two volume collection of Sumerian
parables. Samuel Noah Kramer spent his life translating Sumerian
literature. The late Thorkild Jacobsen produced soon before his
death The Harps That Once... Sumerian Poetry in Translation (1987)
which has 484 pages of nothing but translated Sumerian poetry.
If we do not know the Sumerian words for 'art', 'beauty', 'politics',
and 'religion', it is because they didn't have those words, not
because those words are not preserved.
>>We probably don't need to belabor this point anymore. There is
>>general agreement among the linguists that 1) Sumerian was a fully
>>modern language with a completely modern lexicon ("modern" is the
>>term the linguists are using for "fully capable"),
Is tautology the word to describe saying that the language of the
Sumerians was fully capable of talking about whatever they talked
about?
>>and 2) we don't have anywhere near a complete lexicon of Sumerian
>>as it was spoken.
Sumerian is not Panamint. I think you are projecting.
Before closing, I did want to say that I approved of your
description of the three stages of language. You have said yourself
that you have not thought too much about the dating of stage two.
The dating is something to which I have given a lot of thought. I
have not found any evidence to contradict the idea that stage two
was more recent than most people assume, namely at the start of the
Near Eastern Neolithic.
I believe the evidence supports the spread of language through
stimulus diffusion much better than it does a hundred-thousand year
genetic descent from a single Out of Africa ancestor.
Regards,
John Halloran
------------------------------
Date: Sun, 15 Mar 1998 13:23:17 -0700 (MST)
Subject: EvolLang: Re: Concepts Without Words
>Sent by: Robert Whiting <whiting@cc.helsinki.fi>
>
>> Sent by: Larry Gorbet <lgorbet@unm.edu>
>>
>> For a megadose of well-constructed and widespread concepts without
>> words
[with examples such as the hard tip on the end of a shoelace]
>
>Along the same lines, if you still don't believe that there can be
>concepts without words have a look at Douglas Adams and John Lloyd,
>*The Deeper Meaning of Liff* (expansion of an earlier version, *The
>Meaning of Liff*). You should find it in the humor section of your
>bookstore. Presented as a dictionary, it achieves its purpose by
>matching (in a somewhat iconic way) place names for which there is
>no definition with concepts for which there is no word.
>
>Some examples:
>
<snip of examples>
>
>I don't want to give any more for fear of running afoul of the
>copyright laws, but you get the idea. There are literally hundreds
>of documented concepts without words as well as the hundreds of
>concepts that Larry refers to that have words that most of us have
>either forgotten or never knew.
Both posts are very interesting. Would everyone agree that even
functioning individuals without language, such as the uneducated
born-deaf, could have such concepts? [Specifically those concepts
that do not involve speech and hearing]
This really makes the symbols of language the tip of the concept
iceberg.
An individual without language could recognize a person to be of an
annoying type encountered in the past or a shopping cart to be of a
defective type encountered in the past. A person without language
could use those conceptual recognitions as the basis of behavior.
It appears that recognition of similarity and grouping of experience
into discrete types and patterns is necessary for concepts. We
already noted that children group based on form or function. A
concept groups objects or actions of similar form or function.
Concepts can range along a continuum from concrete to abstract.
The concepts from Esquire that Larry G. quoted, such as the hard tip
on the end of a shoelace, were on the concrete end of the continuum.
The Douglas Adams concepts that Robert quoted are still concrete, but
getting more abstract.
Medium abstract would be the concept of children born out of wedlock.
Very abstract would be the concept of self-esteem.
Very abstract would be the ecological concept of the tragedy of the
commons.
Highly abstract would be the square root of a negative number.
Probably all of these concepts can be conceptualized without language,
although it does not appear that many individuals will conceive of
the more abstract concepts without learning them via learning the
words that stand for them.
On the function of words as tokens for concepts, I cannot do better
than to quote two analytical thinkers from the past.
"words are the tokens current and accepted for conceits [concepts],
as moneys are for values" Francis Bacon [1561-1626], Advancement of
Learning, Second Book, Section XVI.3.
"One of the uses of speech is "the Registring of the Consequences of
our Thoughts; which being apt to slip out of memory, and put us to a
new labor, may again be recalled, by such words as they were marked
by. So that the first use of names is to serve for Markes, or Notes
of remembrance." Thomas Hobbes [1588-1679], Leviathan, Bk. IV.
More recently, a historian wrote,
"As the invention of granaries made it possible to store corn
instead of living from hand to mouth, and as the invention of money
made it possible to accumulate the fruits of work and use them for
fresh enterprise, so the invention of conceptual speech made it
possible to save the fruits of experience and use them for the
formulation of new thoughts and theories. All alike are a form of
banking, a way of making the past serve the future." Jacquetta
Hawkes, History of Mankind, Vol. I, Prehistory and the Beginnings
of Civilization (1963), pp. 6-7.
It helps to be a student of history to appreciate the way in which
our bank account of conceptual words has grown cumulatively through
the years. When we are talking about the gradual growth, evolution,
or elaboration of the lexicon, linguists need to maintain an
historical perspective.
Getting back to the idea of people who have concepts without
language, could people have domesticated goats and sheep without
having the word "to domesticate"? Could people have built mud-brick
houses without having the word "to build"?
Regards,
John Halloran
------------------------------
Date: Sun, 15 Mar 1998 14:46:00 -0700 (MST)
Subject: EvolLang: spread of language
>Sent by: Kelly Jaakkola <jaakkola@biology.ucsc.edu>
>
>> On Fri, 13 Mar 1998, John A. Halloran wrote:
>
>> I believe the evidence supports the spread of language through
>> stimulus diffusion much better than it does a hundred-thousand
>> year genetic descent from a single Out of Africa ancestor.
>
>John (or anyone else), I would be really interested in knowing what
>evidence you think supports this idea. (I'm not being argumentative;
>I just don't know this literature very well.) Could you lay out what
>you think the relevant evidence is?
I was referring to the way in which the languages of populations
that have been in historical contact may share typological features
or subsets of their basic vocabulary but other large sections of
basic vocabulary show no genetic resemblance indicating that the
populations innovated most of their vocabulary on their own. This is
a problem with which all comparative linguists of primarily genetic
persuasion must grapple. The 'Altaic' languages are an example of
languages that are grouped together because they share morphology in
common, but unfortunately for the genetic descent model do not share
a common vocabulary.
I have seen no reference on this list to Robert R. Ratcliffe's blast
on the State of comparative linguistics which appeared on the
Linguist List on March 6, 1998, so I will quote the portions that
relate to your question. [Reposted with permission. Matt]
> I strongly share Manaster's concern for the current state of
>historical linguistic studies in the U.S. and elsewhere, and for the
>lack of attention paid to non-IE families. But having read several
>of his postings here and on other lists, I feel that his apparent
>obsession with GENETIC classification is both bizarre and
>anachronistic.
> The goal of historical linguistics is to answer two questions:
>What can language change tell us about language and what can
>language data tell us about historical movements and contacts among
>different populations. The classic genetic tree model of the
>nineteenth century reflects only ONE type of historical situation --
>that where a population speaking a more or less homogeneous language
>splits into two or more groups which then separate to the point that
>they completely and permanently lose contact with each other. But
>based on more modern research on sociolinguistic variation,
>dialectology, creoles and pidgins, and language contact, we have to
>conclude that the situation envisaged by the classic genetic model
>is by no means the most common or typical situation. That is why
>many historical linguists feel that rather than try to force the
>data into an inadequate model -- that is to force all langauges into
>bigger and bigger language families -- it might be more worthwile to
>try to develop relationship models which deal realistically with
>contact and variation.
[snip]
>> Likewise, many general linguists and others "know" from sources
>> such as Johanna Nichols' book from a few years ago that the Altaic
>> theory has been conclusively refuted and is now dead, but
>> almost no one outside the circle of scholars who actually
>> work on these languages knows that there are actually
>> many more proponents of Altaic now than there were in
>> the 1960's, when the great Altaic debate raged, or that
>> one of the two or three leading opponents of Altaic,
>> Janhunen, has just recently announced what appears to
>> be an endorsement of the relationship between Mongolic
>> and Tungusic, which is a part of the Altaic theory.
>
>I assume that Manaster is referring here to the book "linguistic
>diversity in space, and time," in which case he is seriously
>misrepresenting what Nichols says about Altaic. As I read her,
>Nichols nowhere says that the hypothesis of an Altaic family has
>been refuted. Rather she says that many scholars are now considering
>the possiblitiy that the similarities among so-called Altaic
>languages may be due to something other than a classic genetic
>relationship -- i.e. they may be due to shared typological
>properties or to areal contact. If this what she means, she is
>perfectly right. I can attest to this, although I am no specialist,
>because I have had the privilege over the last year to participate
>in a series of workshops on "Altaic" languages organized at our
>university under the chairmanship of the Turkologist Lars Johanson,
>involving specialists in Turkic, Tungusic, Mongolian, as well as
>languages or families in contact with "Altaic" languages such as
>Uralic, Slavic, and Semitic (the last represented by me). The stated
>position of the chairman was that the "Altaic" family remains an
>open question. And none of the members seemed particulary interested
>in arguing for or against the genetic grouping. Rather, what
>everyone was interested in, and the official theme of the panel was
>"Contact induced changes in peripheral 'Altaic' languages." In short
>language contact and typological universals are interesting areas of
>research right now, and until we know more about these things arguing
>for or against an Altaic family is utterly pointless.
I personally have compared the vocabularies of proto-Afroasiatic (in
Chris Ehret's book) to that of other languages such as Sumerian. I do
not find any point of contact that would justify saying that these
languages descended from a common ancestor. In The Origin of
Language, Merritt Ruhlen gave 13 words that he thought had global
cognates in languages around the world, indicating monogenetic descent,
but I couldn't find those words in Sumerian either, and linguists do
not take him or his methodology seriously.
Evidence does exist in the ancient world for independent elaboration of
other cultural inventions, such as animal domestication, whose concepts
were spread by what is called stimulus diffusion.
Regards,
John Halloran
------------------------------
Date: Sun, 15 Mar 1998 19:17:57 -0700 (MST)
Subject: Re: EvolLang: Abstraction in the Lexicon
>Sent by: "Ronald Kephart" <rkephart@unf.edu>
>
>> Sent by: John A. Halloran
>
>> ...Of course the Sumerians came up with abstract words...
>
>I repeat: all words in Sumerian are/were "abstract"; am I the only
>one who thinks the "abstract vs. concrete" thing is a false
>dichotomy?
We need to agree on terms. I am referring to how concrete or
abstract is the referent of whatever word I call an abstract word or
a concrete word. Is there a more precise term that I should be using?
In the sense in which I am using it, 'home' seems more abstract than
'house'. Both referents are more concrete than the referent of the
mathematical word 'pi'.
Claude Levi-Strauss, on page 1 of The Savage Mind, denies that
abstraction is more characteristic of the languages of civilized
peoples. "It has long been the fashion to invoke languages which lack
the terms for expressing such a concept as 'tree' or 'animal', even
though they contain all the words necessary for a detailed inventory
of species and varieties. But, to begin with, while these cases are
cited as evidence of the supposed ineptitude of 'primitive people'
for abstract thought, other cases are at the same time ignored which
make it plain that richness of abstract words is not a monopoly of
civilized languages. In Chinook, a language widely spoken in the
north-west of North America, to take one example, many properties and
qualities are referred to by means of abstract words: 'This method',
Boas says, 'is applied to a greater extent than in any other language
I know.' The proposition 'The bad man killed the poor child' is
rendered in Chinook: 'The man's badness killed the child's poverty'"
Levi-Strauss discusses uncivilized cultures' abstract systems of
totemic classification which utilize homologies and dichotomies to
set up systems of correspondences.
Certainly, a high level of abstraction is displayed by the early
creators of the different language vocabularies. This is
particularly apparent in the system of 'phememic' correspondences
that Mary LeCron Foster claims to have discovered for the phonemes of
primordial Indo-European [M.L. Foster, "The Symbolic Structure of
Primordial Language," in Human Evolution: Biosocial Perspectives, ed.
S.L. Washburn and E.R. McCown (Perspectives on Human Evolution, vol.
IV, Menlo Park, 1978), pp. 77-121.] Different consonants are
related to spatially different types of actions in the world. Jess
Tauber has given us examples of abstract iconicity for phonemes in
several different languages.
The iconism of the proto-Sumerians was not so systematic, but still
shows a level of abstraction. The word for 'female breast', UBUR,
is a compound, but only a compound of abstractions, not of words for
concrete objects.
AB1: window; niche.
AB2: domestic cow.
UB1: corner, angle, nook.
UB3: a drum.
UB4: cavity, hole; pitfall.
IB2, EB2: n., middle; waist; loins; thighs.
UR: to flood.
IR: sweat.
ER2: tears; to weep.
The words are referents to concrete things, but they manifest the
abstract ideas of 'hollow container' and 'liquid secretion', ideas
which come together in UBUR to refer to a 'female breast', although
it is not a compound of any of the concrete words above.
>> ...How far are we from the origins of language with a language
>> that derives the word for 'eyes' from the word for 'door'?
>
>Isn't this possibly an example of metaphor?
Humans constantly use metaphor to create new vocabulary, although
in this case I would suggest that the abstract idea of 'entrance'
symbolized by the throat lay behind IG as 'door' and reduplicated
IG to mean 'eyes'.
In discussing human ability for abstraction, we are not discussing
rocket science. A young child is similarly abstract when she
mistakenly extends the word 'ball' to apply to anything with a round
shape or to anything that bounces.
Instead of using the terms 'concrete' and 'abstract', what I really
want to say is that human languages first developed words for the
most obvious or important phenomena of human existence and that ever
since then we have been coining words that make finer distinctions or
are conceptually more difficult, less obvious ideas [often with
abstract referents]. We have been filling in pieces and building
upon or combining previous word ideas. Language has served as an
assistant to cultural invention. Elsewhere I quoted Bacon, Hobbes,
and Hawkes regarding words as tokens for ideas. Civilized humanity
has gradually built up an inherited bank account of concepts, denoted
by the 450,000 words in Webster's Third.
>> ...If we do not know the Sumerian words for 'art', 'beauty',
>> 'politics', and 'religion', it is because they didn't have those
>> words, not because those words are not preserved.
>
>Two questions:
>
>(1) Is it possible that we don't have these words because for
>Sumerians these things did not exist as seperate *categories* of
>meaning apart from other aspects of their lives? Art and religion
>are especially suspicious in this regard. "Art" as a thing
>separate, standing on its own, is a rather western (and perhaps
>capitalist) notion isn't it? In other cultural contexts, "art"
>doesn't exist apart from its functions in ritual, etc.
Yes.
>(2) How can anyone be so sure that these words didn't exist in
>Sumerian, given the nature of what is preserved (I'm asking here,
>since I'm not that familiar with the archeological record)?
All I can tell you is that in the 13 years during which I have been
assembling my lexicon of Sumerian, I have never seen any words so
translated.
I wouldn't worry too much about what was preserved. Contra Patrick
Ryan's suggestion of perishable writing tablets, I was recently
reading Klaas R. Veenhof's introduction to Cuneiform Archives and
Libraries (1986) in which he emphasizes how fortunate we are that
clay was the only writing material of Mesopotamia until late in its
history. About the 4th century B.C., a royal edict apparently cut
off much of our historical information by a shift to papyrus for
administrative records. He does say that starting around the 10th
century B.C. wax tablets were apparently used for scientific writing,
so we may have lost a lot there. But clay just gets harder and better
preserved when exposed to flames and that is all they wrote on during
the time of Sumerian civilization. We have extensive tablet archives
of everything that the Sumerians wrote for almost one millenium
down to about 1800 B.C.
>> Is tautology the word to describe saying that the language of the
>> Sumerians was fully capable of talking about whatever they talked
>> about?
>
>There are two separate issues here: grammar and lexicon.
Agreed.
>The *grammar* of Sumerian, as far as I know, was fully modern, i.e.
>it exhibited all the features of language. The *lexicon* of Sumerian,
>of course, was not "modern" in the sense that it lacked words for
>'hard drive', 'carburetor', and many others which we in the
>"modern" world need to get along. The lexicon is a relatively
>trivial part of a language in the sense that it can very easily add
>needed words for new concepts thru coinage, metaphor, borrowing, and
>so on.
I tend to agree with you, my only point having been that the lexicon
grows in a cumulative manner that can help date the origin of
language.
You should know, however, that a review on the back of The Symbolic
Species says, "The bottleneck is *not grammar*, which can be
simplified indefinitely. The real hurdle is *reference*: the yawning
chasm on the rocky evolutionary-neurobiological road from *icons* and
*indices* to *symbols*." Admittedly, Terrence Deacon is not a
linguist, but he appears to have a view opposite to the one you
express above.
Regards,
John Halloran
------------------------------
Date: Wed, 18 Mar 1998 17:54:37 -0700 (MST)
Subject: Re: EvolLang: Re: stadialism
>Sent by: "Larry Trask" <larryt@cogs.susx.ac.uk>
>
>Johanna is only talking about the distinctive polyglot regions she
>formerly called `residual zones' and now calls `accretion zones'.
>In these zones, and only in these zones, she finds a tendency for
>speakers to shift toward what she calls the "standard profile",
>characterized by "greater morphological complexity, cliticization or
>other increase in head-marking patterns, and propensity for
>accusative alignment and SOV word order".
>
[snip]
>
>What Johanna is saying is *not* that languages tend universally to
>change in certain directions -- indeed, she expressly denies that.
>Instead, she is merely saying that linguistically complex zones tend
>to shift toward certain statistically prevalent types. In other
>words, departures from statistical norms tend to be ironed out under
>intense contact. That's all.
So there are statistical norms to which languages approach or from
which they diverge. Humans must mentally prefer these patterns or
they would not be the most commonly found patterns. And under
intense contact, the fluidity of the situation allows humans to
regroup around the preferred patterns.
Did the languages with patterns that diverge from the norm have
those divergent characteristics from the start?
If so, what does that say about language evolution, about the initial
creation of language(s)? Doesn't that mean polygenesis?
Is there or is there not diachronic change in these characteristics
outside of the linguistically complex accretion zones? Nichols finds
that diversity equally characterizes the linguistically simple spread
zones of both past and present. This suggests that in these areas
these characteristics do not change with time. If they are changing,
they should change in the direction of the norm, but she doesn't find
unidirectional change in the spread zones. This contradicts the
conventional wisdom that languages are constantly changing from type
to type without favoring any particular direction. It says that left
unperturbed languages do not change. This means that languages whose
characteristics diverge from the norm have been that way since they
began - they did not gradually diverge into what they are from one
fully elaborated ancestor language.
This says to me that modern humans did not already have a spoken
language at the time that modern humans migrated around the earth,
except where those migrating humans can be shown to share the same
language, such as Austronesian, spoken by relatively recent waves of
migrating humans. When the languages are of different types and
cannot be shown to be genetically related, it is because humans had
already dispersed and there were separate pockets of language
innovation among the far-flung groups, stimulated by diffusion of the
concept of language.
Regards,
John Halloran
------------------------------
Date: Wed, 18 Mar 1998 17:54:37 -0700 (MST)
Subject: Re: EvolLang: Short Chomskyan Credo
>Sent by: "Larry Trask" <larryt@cogs.susx.ac.uk>
>
>> Jose-Luis Mendivil writes:
>
[snip]
>> you know, this innate Universal Grammar (which is not exactly the
>> same as human language) is a hypothesis intended to explain why
>> human beings naturally acquire or learn only human languages and
>> not, for example, formal languages or dog's dialects,
[snip]
Has anyone else read the recent book reporting on the two young girls
who in India in the 1940's were discovered to have been living with
wolves? Contrary to the statement above, the book basically said that
the girls considered themselves to be wolves, and had learned wolf
culture and wolf vocalizations.
>Yes, but Chomsky's particular version of innateness is anathema to a
>sizeable number of people, including many who would regard themselves
>as cognitively oriented.
[snip]
Regards,
John Halloran
------------------------------
Date: Sat, 21 Mar 1998 21:05:17 -0700 (MST)
Subject: Re: EvolLang: Re: Concepts Without Words
It surprised me how many people contributed the same impression,
namely that one does not need the words for an activity in order to
do it, such as the important Neolithic accomplishments of
domesticating animals and building mud-brick houses. Besides Chris
and Randy below, Sherman, Larry G., Mark, and John McL. also
contributed.
>Sent by: Chris Cleirigh <cleirig@speech.usyd.edu.au>
>
>I would agree that conceptual abilities precede and are expanded by
>language during development, but note that two different issues are
>being conflated in this discussion:
>
>(1) concepts without words
>(2) concepts without language.
>
>Concepts that an individual can't express as a single word
>but can express as a clause or noun phrase
>might be taken as evidence for (1),
>but I need a little help to see that
>they count as evidence for (2).
That is a valid point. Your sentence itself exemplifies the
possibilities of periphrasis and circumlocution.
>Sent by: GRFoote <GRFoote@aol.com>
>
>> John Halloran writes:
>
>> Getting back to the idea of people who have concepts without
>> language, could people have domesticated goats and sheep without
>> having the word "to domesticate"? Could people have built
>> mud-brick houses without having the word "to build"?
>
>I may be missing the meaning here, but cannot a bird build a nest
>without having the word "to build". One can act upon a concept,
>internal image or instinct, without being capable of expressing any
>such thing.
That is an especially relevant analogy. Unspeaking animals (and
uneducated born-deaf humans) are able to engage in many constructive
and cooperative activities without using symbolic language.
>However, being able to express the concept, to verbalize the image,
>to expand upon the instinct, opens up the way to rapid change and
>improvement. As in "Honey, I want you to add a bedroom so we don't
>have to sleep with the sheep." "And maybe a window ... and if the
>bedroom was on the South side it would be warmer, but the window
>should face west so that the east wind doesn't bring in the dust ...
>the neighbors did that and it worked real well"
>
>I have done a little bit of construction in my days, and I can well
>imagine that such a conversation cannot long postdate Language.
If the elements of language quickly elaborated to be able to express
the different object relations about which biologically modern humans
normally think, then you must be correct.
>And then the expressed concept of "to build" leads (rapidly) to
>plans drawn on paper or earth or clay, and everything "builds" upon
>what was just learned, all because the "critical mass" of true
>Language has been reached, feedback becomes immediate, and
>exponential change occurs. I suspect that this change occured long
>before Sumeria.
For everyone's information, it is Sumer, not Sumeria. Not to worry,
though, as it is quite natural to make that assumption.
I also agree that true language began long before the historical
Sumer of 3,000 B.C., if you consider four or five thousand years to
be a long time.
>Language is that which makes possible the transcendence of Instinct.
>Lascaux and Altamira demand the pre-existence of Language.
I don't agree with these two sentences. If you want to maintain
them, you will have to defend your reasoning. A year or two ago
there was news about finding a 600,000 year-old spear associated
with Homo Erectus in Germany. That wasn't fashioned from instinct.
As soon as humans broke out of their ecological niche in Africa by
using fire and tools to live in just about any niche on the planet,
humans were no longer living purely on instinct. I believe that the
important advantage conferred by an extended childhood is that it
allows time for the greatly expanded human cerebral cortex to
assimilate noninstinctual information about the idiosyncratic
characteristics of the birth environment. I think we are agreed
that the long evolutionary process that lengthened the human
childhood and expanded the human brain did not require the overlay
which is language.
Based on the responses, most contributors do not find it
inconceivable that humans could start to live together in permanent
settlements, cultivate small meat animals, and plant grain without
yet having symbolic language.
Regards,
John Halloran
------------------------------
Date: Sat, 21 Mar 1998 22:49:49 -0700 (MST)
Subject: Re: EvolLang: Abstraction in the Lexicon
>Sent by: mcv@wxs.nl (Miguel Carrasquer Vidal)
>
>>John A. Halloran wrote:
>
>>The iconism of the proto-Sumerians was not so systematic, but still
>>shows a level of abstraction. The word for 'female breast', UBUR,
>>is a compound, but only a compound of abstractions, not of words
>>for concrete objects.
>>
>>AB1: window; niche.
>>AB2: domestic cow.
>>UB1: corner, angle, nook.
>>UB3: a drum.
>>UB4: cavity, hole; pitfall.
>>IB2, EB2: n., middle; waist; loins; thighs.
>>
>>UR: to flood.
>>IR: sweat.
>>ER2: tears; to weep.
>
The above words are the historical Sumerian concrete expressions of
the proto-Sumerian abstract concept morphemes V+b and V+r, 'hollow
container' and 'liquid secretion', which meanings derive from
articulatory iconicity during an early stage in language evolution.
<snip>
>
>You cannot prove in any way that the word <ubur> means "sweating
>drum" or any iconic permutation of the above. Why not u2 "food,
>nourish" + bur12 "to draw"? Why not u5 "to be on top of" + mur
>"lungs"? And indeed why not simply the unanalyzable word <ubur>
>"breasts"?
Because through analysis we defeat ignorance and penetrate farther
into the past.
ablal(3): bird niche, nest (lal, 'to be high; to hang').
absin(3): (seeded) furrow (sin2, 'to sift').
Would you prefer to leave these good Sumerian words unanalyzed as
well, or is the idea of a 'hollow container' also appropriate for
them?
Regards,
John Halloran
------------------------------
Date: Sat, 21 Mar 1998 23:34:33 -0700 (MST)
Subject: Re: EvolLang: spread of language
>Sent by: "Patrick C. Ryan" <proto-language@worldnet.att.net>
>
>But, the bottom line is this: if the Sumerians "invented" language
>(a bit late, for my taste)
rather the people who later became the Sumerians
>, and the language they invented spread by contact, this requires
>that the a-linguistic groups with whom they came in contact, were 1)
>language-capable but 2) not using language for if they were not
>language-capable, how could they adopt the invention of the
>Sumerians?
>
>I find these two premises extremely difficult to accept together.
Yes, that is the problem in a nutshell. But if there was
polygenesis of language rather than monogenesis, it had to happen at
some date. If all modern humans alive today are not descendants of
one already speaking population, then there must have been language
capable humans who separately innovated the details of a new
language (it would have been easier for a population to make this
conceptual leap through exposure to the concept of language than it
would have been to independently originate the concept of language).
I think everyone has trouble with the idea of a language capable
human not using language.
Regarding Sumerian being connected to Afroasiatic via Indo-European
and a common 'Nostratic' ancestor, I have seen many reasons why
Indo-European and Afroasiatic are not descended from a common
ancestor. I have seen nothing but loanwords, not basic vocabulary,
connecting them, words which could have been a shared property of
the Fertile Crescent trading area which both populations occupied.
Comment?
Regards,
John Halloran
------------------------------
Date: Thu, 26 Mar 1998 13:10:32 -0700 (MST)
Subject: Re: EvolLang: Abstraction in the Lexicon
>Sent by: "Peter T. Daniels" <grammatim@worldnet.att.net>
>
>> Sent by: John A. Halloran
>>
>> We have been talking about the preservation of words, not texts.
>> You are implying that many words occurred only in one particular
>> text, pointing out correctly that we have just fragmentary copies
>> of some important texts. I think, to the contrary, that as more
>> and more tablets are published and studied, we are recovering the
>> entire common Sumerian vocabulary. I am not much interested in a
>> word that only occurs in one text.
>
>No; we have been talking about the supposed preservation of
>virtually the entire corpus of Sumerian literature.
>
>As for the lexicon, what percentage of entries in the Chicago
>Assyrian Dictionary comprise words attested *only* in the lexical
>lists?
My recollection is zero percent in the CAD. They provide usage
examples for all headwords. Examples are taken from actual texts,
not lexical lists. Do you have any examples to prove your point?
I looked at the 20-some volumes of the CAD for 12 hours a day for
about a month in the early 1980's, when I was starting to compile my
lexicon of Sumerian. My procedure was to go in the morning to UCLA's
Graduate Reserve Library where the staff put the volumes of Deimel,
the CAD, the AHw, and Labat on a cart which I just had to wheel to a
table in order to methodically go through all the Akkadian words that
Deimel listed for each Sumerian logogram. I gave a lot of credence
to the CAD in particular because its material was much newer than
Deimel and more complete than AHw (although at that time especially
the AHw was necessary for the missing letters in the CAD). The CAD
in the references section for each headword helpfully lists known
Sumerian logogram correspondences to the Akkadian words. It was a
long process for which probably no one else had the time or
inclination, but I felt I was doing something real and important and
was sorry when I came to the last word.
>What percentage of the lexical lists have not yet been recovered?
>Thus what percentage of the vocabulary of Akkadian _and_
>Sumerian words is missing?
I would not trust a word that appeared only in a lexical list.
Regards,
John Halloran
------------------------------
Date: Fri, 27 Mar 1998 02:07:49 -0700 (MST)
Subject: Re: EvolLang: Re: Australia
>Sent by: "H.M.Hubey" <hubeyh@montclair.edu>
>
>What I really wanted to discuss was the languages spoken by the
>indigenous peoples of Australia (the aborigines). The reasons are
>quite plain. Here we have an almost perfect laboratory. Humans
>apparently reached it about 40,000 to 50,000 years ago and probably
>no other people until the arrival of the white men (English etc.).
>It is one place where we can be reasonably sure that no "external
>effects" (i.e. contact induced change) could have occurred on a
>large scale (unless there were more arrivals later) and even if so,
>those arrivals would have to be pretty much restricted to one or two
>families in that region.
>
>I think we should all discuss this in detail. We might even go over
>Dixon's book chapter by chapter.
>
>What say you all?
The linguists on this list who know the Australian language
situation are not those claiming that Australia supports the
existence of language as of 50,000 BP.
"The modern distribution of Australian languages points to an
almost explosive expansion of the Pama-Nyungan speech-area 4,000
to 5,000 years ago. This expansion led eventually to the
establishment of Pama-Nyungan speech communities over seven-
eighths of the area of Australia." G.N. O'Grady, "Preliminaries
to a Proto Nuclear Pama-Nyungan Stem List," in Australian
Linguistic Studies, ed. S.A. Wurm (Pacific Linguistics Ser. C,
No. 54, 1979): 114.
Modern Australian languages display "a common phonemic basis.
In passing it is only necessary to remark that this feature can
hardly be explained from any idea except that of a common origin."
A. Capell, The History of Australian Languages, p. 423. This
means that no current Australian language is likely to be a
survival of an unrelated substrate language.
"In the original classification in terms of lexicostatistics by
O'Grady and F.M. and C.M. Voegelin in 1966, the Australian
languages were divided into 29 phylic families, of which 28
were found in Arnhem Land [the northern tip of Australia]."
"According to Dyen's principles of relationships (Dyen 1956)
the original 'homeland' should be sought here." Capell, p. 585.
Based strictly on the evidence, language in Australia does not have
to be older than 5,000 BP.
Regards,
John Halloran
------------------------------
Date: Sat, 28 Mar 1998 01:07:27 -0700 (MST)
Subject: Re: EvolLang: Abstraction in the Lexicon
>Sent by: Robert Whiting <whiting@cc.helsinki.fi>
>
>> Sent by: John A. Halloran
>>
>> >Sent by: "Peter T. Daniels" <grammatim@worldnet.att.net>
>> >
>> >As for the lexicon, what percentage of entries in the Chicago
>> >Assyrian Dictionary comprise words attested *only* in the
>> >lexical lists?
>>
>> My recollection is zero percent in the CAD. They provide usage
>> examples for all headwords. Examples are taken from actual
>> texts, not lexical lists. Do you have any examples to prove your
>> point?
>
[snip]
>
>So I got to the seventh word (not counting cross references) in the
>lexicon before I hit a "lex. only" word, but the 5th and 6th words
>were also hapax legomena known only from lists. The 9th word,
>aba:ja (a water-fowl) is also a "lex.*" word; the next is not. So
>of the first 10 words in CAD, 5 are known from lists only. That's
>50% for those who don't know much math. While the percentage will
>go down as the number of words looked at increases, I would guess
>that it will still end up in the 20-30% range.
I thank you and Peter for the education. You made it
unnecessary for me to go to UCLA to check up on this. Peter was
right about there being a good percentage of words that are attested
only in lexical lists, so therefore it is possible that we are
missing a percentage of vocabulary due to incomplete preservation
of the lexical lists. I do not yet know it all and am glad
that there is still more to learn.
[snip]
>Certainly a daunting task, and I can appreciate why you feel that
>there can't be many words missing. But anyone who thinks that
>there are very few Sumerian words whose meanings are not known has
>simply never tried to read a text written in Sumerian. I don't
>mean following someone else's reading of a text where all the rough
>edges have been smoothed out and readings and interpretations have
>been proposed, but picking up a tablet and making sense of it from
>your knowledge of Sumerian grammar, syntax, and lexicon. Even
>though much of it may be easy, there are inevitably places where
>you end up guessing and looking for parallels to substantiate your
>guesses. Even the formulaic and repetitive economic/administrative
>texts almost always have some tiny point that is not entirely clear
>from the point of view of meaning or usage.
I have had some experience with this. Part of the problem is that
with the little time available to us we are forced to be generalists,
but a general knowledge of Sumerian does not qualify to penetrate
down into the writing habits of a particular place, time, and
occupation within the thousand year span of Sumerian culture. I
studied the different categories of Ur III administrative texts with
Englund last quarter, but we just had enough time to translate a
couple of representative texts for each category. One must learn the
peculiar vocabulary and conventions of each category of text. Part of
the difficulty was not having an adequate lexicon, so at least a
portion of the Ur III administrative vocabulary is now part of the
lexicon that I have made easily accessible on the net.
>Have a look at a translation with a philological commentary and
>see how much effort is invested in justifying readings, meanings and
>interpretations. This is the material that usually gets filtered
>out in the glossary or the popular translation.
I have Bendt Alster's masterful new Proverbs of Ancient Sumer
(1997) in two volumes which follows the earlier (1959) and equally
masterful edition by Edmund I. Gordon and Thorkild Jacobsen. Is the
commentary in these volumes the kind to which you are referring?
>> I would not trust a word that appeared only in a lexical list.
>
>Why not? Do you suspect that the ancients just made up words to
>put in lexical lists?
That is a good question. I think the cautious Michigan Sumerologist
Piotr Michalowski put the fear of God in me by warning me of
metaphorical lexical lists like Idu and late lists like antagal.
Let's just say that I would feel a lot better about a word if two
separate texts which do not derive from the same source have the
same word with approximately the same meaning. If a word is
attested in actual use, you have to admit that its credibility goes
up significantly.
Thanks for your comments.
Regards,
John Halloran
------------------------------
Date: Fri, 3 Apr 1998 10:04:08 -0700 (MST)
Subject: Re: EvolLang: Monogenesis vs. polygenesis
>Sent by: "Dr. John E. McLaughlin" <mclasutt@brigham.net>
>
[snip]
>sure, I looked at some other language families I have material
>on--Proto-Uto-Aztecan *cam "good", Eastern Keres (Keresan family)
>ra'wa': (a' is acute accents over vowel) "good" and ?a'n'e': (n' is
>glottalized n; ? is glottal stop) "tasty", Lushootseed (Salishan
>family) ha'?hl (hl is voiceless lateral fricative) "good", Nez Perce
>
<snip>
>
>family) washte' "good, beautiful", Eastern Pomo (Pomoan family,
>Hokan stock?) q'o:di'y (q' is glottalized q) or c'i:qo'm (c' is
>glottalized c) "good", Seneca (Iroquoian family) wi:yo:h "good",
>Wichita (Caddoan family) acs "good", Coahuilteco (isolate) -shap'a:n
>(p' is glottalized p) or mana:m "good", and Hebrew t.ob (t. is
>teth) "be good, beautiful, pleasant".
[snip]
Am I the only one who thinks that these dissimilar roots are
evidence for polygenesis?
Who else thinks that many separate unspeaking populations must have
independently invented the words of their language?
The evidence for it stares us in the face in the forms given above.
Some would assert that these dissimilar roots and common meanings all
ultimately descend after a long and tortured history from one
original proto-language. But isn't that an article of faith, without
support from evidence?
What do linguists specializing in the Americas or New Guinea really
believe is the explanation for the existence of language isolates?
It appears to me that the presence of morphological or phonological
areal features in otherwise unrelated languages can best be
explained, not just by contact among language speakers, but by those
features being present in the local concept of how a language should
work, a concept that spread like a cultural wave across unspeaking
populations.
Regards,
John Halloran
------------------------------
Date: Sun, 5 Apr 1998 14:25:22 -0700 (MST)
Subject: Re: EvolLang: Monogenesis vs. polygenesis
>Sent by: "Dr. John E. McLaughlin" <mclasutt@brigham.net>
>
>> Sent by: John A. Halloran
>>
[snip]
>
>> The evidence for it stares us in the face in the forms given
>> above. Some would assert that these dissimilar roots and common
>> meanings all ultimately descend after a long and tortured history
>> from one original proto-language. But isn't that an article of
>> faith, without support from evidence?
>
>You are exactly right. It is an article of faith because we can't
>prove it.
[snip]
>depth. No linguist would ever claim that. Let's take the language
>of North America as an example. We can firmly establish about
>thirty language families and a couple dozen isolates. We know that
>the peopling of North America was a short term event and the
>circumstantial evidence strongly points to the peoples of the
>Americas being related to one another. Therefore the majority of
>languages of the continent are probably ultimately related. But we
>can't prove it. We don't have the evidence (and may never).
Good response. At a minimum, your linguistic article of faith
includes the ability of two or three languages to transform
themselves into thirty language families and a couple dozen isolates
within the time span that elapsed between the human occupation of the
Americas and the recording of the American languages. As a specialist
in the Americas, what is the current thinking on the depth of that
time span?
Are you familiar with Jacques Guy's article on variations in word loss
rates among different language communities? He concluded that there
are "unequal rates of lexical retention across languages". Where
glottochronology discovered a rate of about 65% retention after 1500
years for Indo-European, he quotes studies of Icelandic, Norwegian,
Georgian, and Armenian where the vocabulary retention rate after 1500
years ranged from a low of .816 for the 200-word list in an urban
Norwegian community to a high of .990 retention for the 100-word list
in a rural Icelandic community. The address where you will find
Jacque's article is
http://garbo.uwasa.fito
in the MS-DOS section and look in the pc/linguistics area.
Wouldn't the American populations whose languages are classified as
isolates be small populations comparable to the rural Icelandic
community, a community that changed one word out of one hundred over
a 1500 year period? Correct my math if necessary, but doesn't that
mean that at least 150,000 years would be required to change all one
hundred words?
I think we have been blinded by the example of Indo-European, the
language family of a highly mobile population whose degree of change
is not typical for languages in general.
Regards,
John Halloran
------------------------------
Date: Sun, 5 Apr 1998 14:25:22 -0700 (MST)
Subject: Re: EvolLang: Monogenesis vs. polygenesis
>Sent by: "Larry Trask" <larryt@cogs.susx.ac.uk>
>
>> John Halloran writes:
>
[snip]
>
>> It appears to me that the presence of morphological or
>> phonological areal features in otherwise unrelated languages can
>> best be explained, not just by contact among language speakers,
>> but by those features being present in the local concept of how a
>> language should work, a concept that spread like a cultural wave
>> across unspeaking populations.
>
>Sorry, but this strikes me as incoherent. It appears to require
>speechless populations to possess agreed notions of what languages
>should be like before they have any languages.
What I mean by a cultural wave is an unspeaking population learning
basic things such as phoneme repertoire and morphology from a
neighboring population that had recently learned and started to
implement the concept of language, inspired by the example of
another neighboring population that was closer to the center of the
conceptual innovation.
You mentioned that the existence of geographically large language
families was a problem given the small, isolated bands of ancient
humans. Personally, I suspect that the speakers of the large and
diverse Afroasiatic language family did not all descend from a single
population speaking this language, but rather that this language
spread like a changing wave into different neighboring populations in
Africa.
I think cultural innovations spread easier where a cultural vacuum
exists than where the new area is already filled with tools
performing comparable functions.
Regards,
John Halloran
------------------------------
Date: Fri, 17 Apr 1998 23:01:01 -0700 (MST)
Subject: Re: EvolLang: Monogenesis vs. polygenesis
>Sent by: "Dr. John E. McLaughlin" <mclasutt@brigham.net>
>
>> Sent by: John A. Halloran
>>
>> Good response. At a minimum, your linguistic article of faith
>> includes the ability of two or three languages to transform
>> themselves into thirty language families and a couple dozen
>> isolates within the time span that elapsed between the human
>> occupation of the Americas and the recording of the American
>> languages. As a specialist in the Americas, what is the current
>> thinking on the depth of that time span?
>
>The archeological evidence points to a subglacial habitation of both
>Americas about 14,000-12,000 BP. There is arguable evidence for an
>Alaskan occupation about 30-40,000 BP. Some point to subglacial
>evidence of 40,000 BP, but all the "evidence" is VERY problematic
>and uncertain.
John,
You are probably familiar with the following quotes regarding
American proto-languages. You can say if they still reflect the
views and concerns of American linguists. They indicate that
American proto-languages have been reconstructed going back as far as
5,000 years ago. You appear to be saying that many of these
proto-languages do not resemble each other and cannot be shown to be
genetically connected.
According to the theory according to which humans were already
speaking at the time that they entered the Americas, we should be
looking at two or three languages whose speakers transformed them
into 30 + 24 = 54 unconnectable languages during the period between
14,000-12,000 and 5,000-2,000 years BP, approximately a 9,000 year
period. If we can connect languages together to form proto-languages
that go back 5,000 years, I don't understand why the resulting
proto-languages cannot be connected together to go back to a farther
period.
"One of the great American expansions is that of the
Athapaskans in northern and western North America. About
thirty languages can be distinguished, and their time depth
is about sixteen to eighteen centuries [Athapaskan includes
the Navajo language]. Another group is Algonkian in
central and eastern North America, including about a dozen
languages with divergences in the neighborhood of two
thousand years. The Siouan group includes some two dozen
languages, and the divergence for the greater part of this
group is about three thousand years. Uto-Aztecan involves
a similar number of languages, with a time depth of over
four thousand years.
"In Central and S. America the geographically great groups
are Chibchan, Arawakan, Tupian, and Carib. The numbers of
languages are about 30, 60, 40, and 30 respectively. The
depths of divergence run from 3,000 to 5,000 years." Morris
Swadesh, The Origin and Diversification of Language
(Chicago: Aldine Atherton, 1971), p. 222.
"suggesting relatively small homelands for language
families raises a number of problems. Even including the
known isolates, and assigning them homelands in their
present locations, much of the North American continent
is left unaccounted for. However, archaeological sites
show that most of the continent was indeed inhabited at
various early prehistoric times." M. Dale Kinkade and
J.V. Powell, "Language and the prehistory of North America,"
World Archaeology 8 (1976): 92-93. The Algonquian homeland,
for example, existed in a small area around modern Toronto
at about 1200 B.C. Algonquian languages subsequently
spread out to encompass middle and eastern Canada and the
northeastern seaboard of the United States.
In Meso-America the Archaic period of plant domestication
and early agriculture lasted from 8000 to 1500 B.C. At
the beginning of the Formative or Pre-Classic period
(1500 - 0 B.C.), ceramics, highly productive agriculture,
public buildings and clearly marked regional
differentiation of culture patterns all make their
appearance. Terrence Kaufman, "Archaeological and
linguistic correlations in Mayaland and associated areas
of Meso-America," World Archaeology 8 (1976): 102. He
dates proto-Mayan to 2200 B.C. Then he says, "When
Mayan is reduced to a small proto-area and the rest of
the Meso-American families as well, we may see that there
is a lot of unoccupied space, linguistically speaking.
Was it really unoccupied?" ibid, p. 111.
>> Correct my math if necessary, but doesn't that mean that at least
>> 150,000 years would be required to change all one hundred words?
>
>Only if language change occurred at a constant rate. While language
>is constantly changing, the rate of change can speed up or slow down
>for various reasons. Also, you are just looking at lexical change.
>While the lexicon may be very stable in this particular community,
>other parts of the language could be changing very rapidly. Lexicon
>is only ONE part of the equation. You also have to look at
>phonological, morphological and syntactic change as well.
Are you saying that there actually is a good percentage of shared
lexicon between those 54 American language families? That it is
their different phonology, morphology, and syntax that keep them from
being connected?
Regards,
John Halloran
------------------------------
Date: Sun, 26 Apr 1998 15:24:41 -0700 (MST)
Subject: Re: EvolLang: Monogenesis vs. polygenesis
>Sent by: "Dr. John E. McLaughlin" <mclasutt@brigham.net>
>
>> Sent by: John A. Halloran
>>
>> They indicate that American proto-languages have been
>> reconstructed going back as far as 5,000 years ago. You appear
>> to be saying that many of these proto-languages do not resemble
>> each other and cannot be shown to be genetically connected.
>
>That is exactly what I am saying. I am not saying that they are NOT
>ultimately related, but that beyond about 5-6k we can't definitive
>prove that anything is related to anything else anywhere in the
>world.
>
[snip]
>> If we can connect languages together to form proto-languages
>> that go back 5,000 years, I don't understand why the resulting
>> proto-languages cannot be connected together to go back to a
>> farther period.
>
>Because there comes a point where one cannot separate genetic
>cognates from random chance. Check out the website at
>http://www.tezcat.com/~markrose/chance.htm for some mathematical
>argumentation along these lines.
Mark Rosenfelder provides good linguistic education materials via the
Internet - I have had links to his sci.lang faq and ambitious
collection of numerals in the world's languages for a couple of years.
It is precisely my awareness of the problems that he discusses which
caused me to tell you earlier that I don't trust myself to do original
language comparison work.
However, his materials do not support the logic of what you say. I
think the fact of not being able to connect languages beyond a
certain time depth is combining with your assumption that language
must be older than that depth to affect your logic in a deleterious
way.
Obviously, there could be too much noise to connect modern language A
with modern language B if more than 5,000 or 6,000 years separate
them [although, as the studies of Icelandic, Norwegian, Georgian, and
Armenian show, fairly rapid language change would be required to lose
over half the lexicon in this time], but this argument does not
apply if one has a reconstructed protolanguage A and a reconstructed
protolanguage B. If one has reconstructed protolanguages for the
same ancient time period, say 5,000 years BP, it should be possible
to connect the reconstructed protolanguages in the same manner as has
been done for more recent languages that have been separated for less
than 5,000 or 6,000 years.
Dendrochronologists have reconstructed a sequence of tree-rings for
central Europe that goes back 10,000 years by combining sequences
from modern and ancient trees. If fully-developed spoken language is
everywhere at least that old, it should be possible to extend the
genetic sequence back farther than 6,000 years by combining
reconstructed ancient protolanguages. But when that can no longer be
done, it is a strong indication that the genetic sequence does not go
back any further.
If experts in comparative reconstruction like yourself, using a
rigorous methodological manner, cannot connect the reconstructed
protolanguages, then either the reconstructions are flawed or we are
looking at early Neolithic language polygenesis.
Regards,
John Halloran
------------------------------
Date: Wed, 29 Apr 1998 23:25:48 -0700 (MST)
Subject: Re: EvolLang: Monogenesis vs. polygenesis
>Sent by: Larry Gorbet <lgorbet@unm.edu>
>
>Nope. A "reconstructed protolanguage" is not really possible, if by
>that we mean a complete record of what the common ancestor of a
>group of related language was like. What we mean by "reconstructed
>protolanguages" is really "reconstructed fragments of
>protolanguages". First, by hypothesis, only that portion of its
>vocabulary which is shared by the daughters can be reconstructed.
>If the time depth is great --- as in Halloran's hypothetical case
>--- then that fraction is small. Moreover, some aspects of the
>reconstruction will be uncertain (i.e. of the form "X or Y"). Then,
>of course, we will lack much information about morphology and
>syntax, not to mention the details of its phonology.
>
>Reconstruction using very deep protolanguages as "starting points"
>is somewhat analogous to doing reasonable reconstructions from
>living or recent languages about which we have very limited data of
>dubious quality.
I could try to assemble a Swadesh 100-word list from the three books
that I have on Proto-Uralic, Proto-Indo-European, and
Proto-Afroasiatic, so that we could see just how large a fraction we
are talking about, but before I do that, I would like to ask, have
any other list members already assembled one or more of these lists?
>The analogy to dendrochronology is invalid, since a dendrochronology
>spanning a limited range of time need have no loss of information
>--- the oldest sub-spans within it may be just as "complete" and
>certain as the latest.
I see you are not asleep, but appreciate that the position to which
I was responding appeared to assert that one could go back no farther
than protolanguages reconstructed from modern languages, without
acknowledging that one should logically be able to go back even
further by comparing protolanguages.
>Sent by: "Dr. John E. McLaughlin" <mclasutt@brigham.net>
>
[snip]
>This is a very good logical question. If we look at extant language
>A with a recorded lexicon of, say, 5000 words (for a
>"well-documented" Native American language) and at extant language B
>with a similar-sized recorded lexicon, and assuming that their
>common ancestor was 6k BP, we would probably wind up with a
>reconstructed lexicon of about 1000 forms (given good, but not
>necessarily "tight" semantic matches). (This seems to be about
>right for Native America based on my experience with reconstructions
>on this continent. Wick Miller's collection of Uto-Aztecan cognate
>sets was about 1,000 at the time of his death.) Now take language C
>(same info as A & B) and language D (ditto) and reconstruct their
>ancestor, also at 6k coincidentally, and you have about 1000 for
>Proto-CD. At that ratio, with good quality lexicons of 1000
>protoforms from Proto-AB and Proto-CD, and assuming that they are
>related at about the same 6k depth, you only have 200 good cognate
>sets.
C. Ehret has 1011 items in his reconstruction of Proto-Afroasiatic.
G. Decsy has 2563 words and 200 items of morphology for
Proto-Indo-European. Decsy mentions 472 words for Proto-Uralic at
4,000 B.C. and ca. 1200 words for Proto-Finno-Ugric at 3,000 B.C.
So yes 1000 words is a realistic estimate for 5000 BP.
Let me ask you as an experienced comparative linguist, when you find
one out of five modern words in the protolanguage, is that one out of
five words in Swadesh's lists of basic vocabulary, or are the missing
words more likely to be less basic, more elaborate concept words,
indicating that they could be words for new concepts innovated since
the time of the protolanguage?
Your hypothetical example requires that the two protolanguages be
separated by 6,000 years, but surely there must have been some
protolanguages that were separated for less time than that, according
to the branching genetic tree ad infinitum model which you are
defending.
[thanks for the wonderful survey of North American protolanguage
reconstruction, indicating that few reconstructions go back as far as
5000 B.P.]
>If we look at the Old World, most linguists have been hypnotized by
>the quality of the reconstruction of Proto-Indo-European. Remember,
>however, that the reconstruction is the product of 200 years of
>hundreds of linguists' efforts. No other language family on the
>planet is so well-attested. James Matisoff is just now starting to
>produce Proto-Sino-Tibetan materials. Proto-Afro-Asiatic and
>Proto-Uralic materials are also starting to appear, but not in the
>quantity that Proto-Indo-European already has. In time,
>Proto-Nostratic may be provable and reconstructable, as well as
>other deeper groupings. I emphasize the word "may". It all depends
>on the size of the reconstructed vocabulary of Proto-Afro-Asiatic
>and Proto-Uralic. If all that we can find are 200 firm cognates,
>then the relationship will always only be "suggestive" and
>"probable".
If no one else has done so, I will have to assemble those lists. I
have looked at the vocabularies of Proto-Afroasiatic and Proto-Uralic.
Those vocabularies are just not related to Proto-Indo-European or to
Sumerian. Read the interview given by Don Ringe who has looked at the
evidence for Nostratic at
http://www.lexicon.net.au/~opoudjis/Work/ringe.html
He says that the hard core needed to prove the relationship is so
small in Nostratic that it could be the result of chance.
The only reason that the failure of these comparisons interests me is
because I deduced on completely different grounds that the early
vocabulary of Sumerian was invented from the ground up. Early
Sumerian was playing by a different set of rules. They arrived at
their word for 'egg', nuz, by combining the idea of 'discrete
individuality' which adhered to the phoneme /n/ with two other
phonemes /u/ and /z/ which meant 'food' and 'to cook' [as reflected
in uzu, 'meat'], phonemes which derived their meanings from iconic
symbolism. I am not just nay-saying based on the same comparative
evidence that everyone else is looking at. I can predict these
results based on the in-depth analysis that I did of Sumerian.
You sound more optimistic about the long-range potential of other
linguists' fields than you are about your own. Maybe they have felt
the same way about yours. At least you have been blunt in asserting
the reality of what you call the 6K boundary. I am just saying that
if the genetic tree model is correct, it should logically be possible
to push back farther than that by comparing reconstructed
protolanguages.
Regards,
John Halloran
------------------------------
Date: Tue, 5 May 1998 23:55:09 -0700 (MST)
Subject: Re: EvolLang: Monogenesis vs. polygenesis
>Sent by: "Larry Trask" <larryt@cogs.susx.ac.uk>
>
>> John Halloran writes:
>
>> I could try to assemble a Swadesh 100-word list from the three
>> books that I have on Proto-Uralic, Proto-Indo-European, and
>> Proto-Afroasiatic, so that we could see just how large a fraction
>> we are talking about, but before I do that, I would like to ask,
>> have any other list members already assembled one or more of these
>> lists?
>
>Unfortunately, no substantial reconstruction of Proto-Afro-Asiatic
>exists. There have been two serious attempts at this task in the
>last few years (Orel and Stolbova 1995, Ehret 1995), but these are
>very different from each other, and neither commands anything like
>general assent. The widespread feeling is that the time-depth of
>PAA is just too great to permit substantial reconstruction: I have
>seen this time-depth estimated at anything from 8000 to 13,000
>years.
Please provide any references that you have for reviews of the above
two books.
Alice Faber's chapter on the grouping of the Semitic languages in The
Semitic Languages volume edited by Robert Hetzron [my teacher for
four years] questions whether the comparative methods developed by
Indo-European linguists are even applicable to the Semitic languages,
stressing in particular the unreliability of lexemes as a guide to
genetic grouping. She attributes this fluidity to cultural and
geographic connections rather than to great time depth. In either
case, it does make it difficult to pin down the details of the
proto-language.
Is it correct that Orel and Stolbova's work is more influenced by
Semitic than is Ehret's work?
>And I'm not sure that we even have a reconstruction of Proto-Uralic
>that is as substantial as the one we have for PIE.
>
>> Let me ask you as an experienced comparative linguist, when you
>> find one out of five modern words in the protolanguage, is that
>> one out of five words in Swadesh's lists of basic vocabulary, or
>> are the missing words more likely to be less basic, more elaborate
>> concept words, indicating that they could be words for new
>> concepts innovated since the time of the protolanguage?
>
>Not sure what you mean by "missing words". We can only reconstruct
>what we have evidence for.
Read the words to which I am responding, not just my words. Those
said that 1,000 out of 5,000 extant words were found in the
proto-language. I am asking if the 1,000 words represent concepts
that are more likely to be found in Swadesh's lists of basic
vocabulary than do the 4,000 extant words whose ancestors are not
found in the proto-language. This is an important question because
if it were the case, then we probably haven't seen attrition down to
20% of the proto-vocabulary, rather there has been a growth in the
size of the vocabulary. I am not saying that there has been no
attrition, just that it has not been as extreme as dividing the size
of the proto-vocabulary by the size of the extant vocabulary would
indicate.
>> Your hypothetical example requires that the two protolanguages be
>> separated by 6,000 years, but surely there must have been some
>> protolanguages that were separated for less time than that,
>> according to the branching genetic tree ad infinitum model which
>> you are defending.
>
>Sure. The majority of secure families are less than 6000 years old,
>and there is no case of a proto-language which has been
>substantially reconstructed to general satisfaction which is
>believed to be more than 6000 years old.
Once again, read the words to which I am responding. The
hypothetical example was of 6,000 year old proto-languages that
themselves had been separated for 6,000 years, for a total time
depth of 12,000 years. I am pointing out that if we can ever
reconstruct a sufficient number of 6,000 year-old proto-languages,
some of them should show obvious connections through having
separated just 2,000, 3,000, 4,000 years previously, pushing
us back to 8,000, 9,000, or 10,000 B.P. [if the branching genetic
tree truly does go back ad infinitum as many linguists assume].
[snip re: Sumerian meaningful phonemes]
>I'm sorry, but you will find no linguist on earth who takes this
>seriously. The idea that languages have words built up from
>meaningful single sounds was very popular in the 17th and 18th
>centuries, before historical linguistics was invented, but it is
>long dead.
I am going to quote a paragraph that I previously posted while you
were a member of this mailing list. If you pay attention this time,
you will avoid making such wildly incorrect statements.
Foster, Pulleyblank, Hewes, von Raffler-Engel, Fonagy, and de
Grolier all published papers devoted to exploring early isomorphisms
between sound and meaning in the 1983 Glossogenetics volume edited
by Eric de Grolier.
Pulleyblank writes, "Chinese also has many words meaning "turn",
"bend", "round", etc., that seem to have had a uniconsonantal root
w. Many words meaning "dark" either begin or end in m, and so on."
ibid, p. 382. On page 404 he writes, "A side result of our present
investigation, however, is that the earliest recoverable forms of
Sino-Tibetan and Indo-European do show a transparently consonantal
structure that is found in only a few remote and isolated languages
at the present day but is closer to what spoken language was probably
like when it first developed."
Do you remember reading this before? Look back at your words above.
Not very accurate, are they?
Regards,
John Halloran
------------------------------
Date: Tue, 5 May 1998 23:55:09 -0700 (MST)
Subject: RE: EvolLang: Monogenesis vs. polygenesis
>Sent by: "Patrick C. Ryan" <proto-language@email.msn.com>
>
>>The only reason that the failure of these comparisons interests me
>>is because I deduced on completely different grounds that the early
>>vocabulary of Sumerian was invented from the ground up. Early
>>Sumerian was playing by a different set of rules. They arrived at
>>their word for 'egg', nuz, by combining the idea of 'discrete
>>individuality' which adhered to the phoneme /n/ with two other
>>phonemes /u/ and /z/ which meant 'food' and 'to cook' [as reflected
>>in uzu, 'meat'], phonemes which derived their meanings from iconic
>>symbolism. I am not just nay-saying based on the same comparative
>>evidence that everyone else is looking at. I can predict these
>>results based on the in-depth analysis that I did of Sumerian.
>
>I certainly cannot argue with the significance you attach to
>phonemes generally, but in the example you give, it seems to me that
>several really important questions are unanswered.
>
>Firstly, I have determined from my own research that PL na meant
>'nose' (really, even more basically: 'nostril'), and then, because
>of the body-counting method in force at a very early time: 'one'.
>This is certainly very close to your 'discrete individuality'.
I have no information on this semantic symbolism. I see the early
meaning of the phoneme to derive from the gesture of firmly touching
with just the tip of the tongue. This appears to be the most deictic
gesture that the tongue can make during speech.
>We see finally -n as an individualizing element in many languages
>around the world. Frequently, initially, it is the base for
>demonstratives or articles.
>
>However, in the many languages I have analyzed with /n/ as an affix,
>I have never come across one that uses it initially.
>
>So, my first question is in how many other Sumerian words does
>initial n- function as an individualizer, and what are they?
You yourself quote one down below, na4, 'counting pebble or token'.
>The second question is since we both know that there is some
>evidence for AdjN in Sumerian (though the usual pattern is NAdj),
>are you asserting that n- is a preposed modifier of what follows
>(-uz)?
>
>And, if instead, you analyze /n/ as the element of a compound, on
>what basis are you asserting compounds were constructed?
Are you asking if the order of elements was significant when they
were starting to create their vocabulary? I think word order would
have become important gradually over time, but not at the beginning.
>Your example above suggests an analysis 'individual-food-cook'. It
>seems to me that the most likely main element of such a compound
>would be 'food', which would put modifiers of 'food' in both pre-
>and post-placement. Typologically, I would expect both to be on one
>side or the other.
>
>You might want to consider whether nuz, 'egg', is a compound built
>on na, 'stone' + uz, 'goose' since goose eggs were quite popular
>before the introduction of the chicken.
I have always been cautious about using uz, 'goose', in etymologies
because I had seen the suggestion that it could be a Semitic loanword
[Akkadian usu^, ussu]. However, tonight I looked up aleph-vau-zayin,
'goose', in Ernest Klein's Hebrew etymological dictionary, and he
traces all the Semitic forms of this word back to Akkadian and
Sumerian. Actually, I like 'pebble' + 'goose' as a straightforward
etymology for nuz, 'egg'. Thank you for the suggestion.
>We have a number of examples of compounds of this form in Sumerian
>without any overt sign of the genitive (e-2=dub, 'house-tablet' =
>'library, archive').
I don't understand this. It is usually e2-dub-ba, where the ba
syllable represents the a(k) genitival suffix.
>In a general way, I question whether /z/, '*cook', is likely to
>have been associated with 'egg' in early days. It is my impression
>that the egg was valued highly because it was easily digestible raw.
'cook' for /z/ was my abstraction from:
uzu: flesh; cut of meat.
uzu5: evening, sunset.
izi: fire.
uz3, ud5, ut5: she-goat.
where the goat was the meat animal, par excellence, of the ancient
world.
It is hard not to see a connection between 'fire' and 'meat', but
since there are /z/ words also meaning 'teeth', 'sharp stone' and
'to cut', the word for 'meat' could also derive from those
associations.
I do think that even the vowels possessed semantic meaning in early
Sumerian. The fact is that u2 means 'food' and /i/ generally has the
meaning that it has in:
ed2, e3; i: to go out; to send forth; to rise; to sprout; to be or
become visible (the final d appears in mar conjugation).
This meaning of /i/ can be seen in both izi, 'fire', and in the
following:
zi: breathing; breath (of life); throat; soul (cf., zid, z
g).
Regards,
John Halloran
------------------------------
Date: Tue, 5 May 1998 23:55:09 -0700 (MST)
Subject: EvolLang: Neandert(h)al Speech
>Sent by: Matt Fraser <mattf+@pitt.edu>
>
>So? No *intense* discussion of the importance/implications of the
>hypoglossal canal evidence/conclusions about Neandertal Speech?
>
>Yes, no, maybe? Everyone buy it? Does it make any of the other
>religious ceremonies and the musicality of the flute more palatable
>to skeptics?
>
>Speech mediated language outside of our species? Hmmmm?
The representation of important disciplines such as physical
anthropology should be greater than it is on this list. Your
question is addressed by an important book that no one on this list
has mentioned, probably because it does not have the word 'language'
in its title, namely, Merlin Donald, Origins of the Modern Mind:
Three Stages in the Evolution of Culture and Cognition (Harvard
University Press, 1991), 413 pp. He devotes pages 38 - 41 to the
evidence supporting Charles Darwin's proposal that rudimentary
song provided the impetus for primates to evolve the greater vocal
control that the findings you reference indicate.
I will quote a small part of these pages. "Darwin assumed that the
voluntary vocalizations of early hominids were initially achieved
with the standard mammalian vocal apparatus possessed by modern apes.
He spoke of the refinement of the modern human vocal apparatus as a
product of, rather than a precursor to, the earliest human
vocalizations. This being the case, Darwin expected that the first
use of the evolving new skill would have been in producing cadences,
or modulations, that would have resembled singing more than speech.
He pointed out that gibbons produce such modulations during
courtship." p. 38.
Various lines of evidence are adduced which support the long
existence of song and music prior to language and symbolic skills.
Regards,
John Halloran
------------------------------
Date: Wed, 6 May 1998 21:12:56 -0700 (MST)
Subject: EvolLang: Re: group identity markers
>Sent by: Marshall Pease <mpease@eng.Sun.COM>
>
[snip]
>
>I am proposing that observable characteristics of language make it a
>useful tool for identifying your co- group members.
That's interesting. sibboleth versus shibboleth, Shevy versus Chevy.
Ancient Israelite and U.S.-Mexican-border tests for group identity.
What would be the consequences of emphasizing group identity for
language evolution?
>Your language is directly linked to how you were brought up (your
>culture) and native-speaker status is not at all easy to
>counterfeit (compared with dietary preferences or religious
>observances or whatever.) Not the whole picture by any means, but
>evolution is nothing but opportunistic.
>
>It remains to be shown *how* language *got* that way. I personally
>doubt that much progress will be made thataway until we understand
>a lot better how it works.
How language got *what* way?
Regards,
John Halloran
------------------------------
Date: Sat, 9 May 1998 17:53:26 -0700 (MST)
Subject: Re: EvolLang: Monogenesis vs. polygenesis
>Sent by: "Larry Trask" <larryt@cogs.susx.ac.uk>
>
>[on the two proferred reconstructions of Proto-Afro-Asiatic]
>
>> Please provide any references that you have for reviews of the
>> above two books.
>
>Unfortunately, I have seen no reviews, though a specialist journal
>will surely provide some. There might have been one in Language.
>My information comes from an article by Johanna Nichols, but I've
>read about eight of her articles in the last year, and it will
>take a while to dig out the right one. I'll see if I can locate
>it.
Examination of the Journal of Semitic Languages did not turn up any
reviews.
I do have Orel and Stolbova now and will put together the first few
words of a Swadesh list to see how different their book is from
Ehret.
>First of all, where does the mysterious figure of "5000 extant
>words" come from?
This is John M.'s estimate for vocabulary available from the
average American Indian language family.
I thank John for his informative response from two days ago, and
will now ask him a question that has been on my mind. Since a
proto-language is usually constructed from several modern
descendants, what are the chances that a particular modern or extant
language will have a particular word from the proto-vocabulary?
Would a particular modern language have just 80% of the
proto-vocabulary? If so, I can certainly appreciate that that would
be a source of loss and difficulty in trying to compare
reconstructed proto-languages, as we would probably be limited to
just the two proto-languages being compared, and so the situation
would be comparable to having just two modern languages which one is
trying to connect through shared vocabulary.
If we have in the reconstructed proto-vocabulary almost all of the
items on Swadesh's 100-word or 200-word lists, then I don't see that
there has been much attrition in our ability to connect languages
on the basis of those vocabulary items. It may be a linguistic and
scientific tragedy that we have lost the specialized terminology to
which you referred, but it doesn't matter for tracing the origin of
languages.
<snip>
>
>On the general point, yes, there is some reason to believe that
>basic vocabulary is more than averagely resistant to replacement,
>which is the point of the Swadesh list. However, no version of the
>Swadesh list is larger than 200 words - a far cry from 1000 -- and
>anyway there is no shortage of linguists who are deeply skeptical
>of the supposed stability of basic vocabulary, and even of the
>cross-linguistic validity of the Swadesh list.
>
>There is no doubt that languages expand their vocabulary as the
>world of their speakers becomes more complex, but nevertheless, as
>another respondent has pointed out, vast amounts of earlier
>vocabulary can be lost from a language, even in quite a short time.
>And, if the point of the question is supposed to be "Can we
>reconstruct most of the vocabulary of a proto-language?", the
>answer is a resounding "no".
Not sure what you mean by 'most'. I would think that we have enough
to compare, to say yea or nay if we have 95% of Swadesh's 100 or 200
words.
[snip of Larry Trask's belief that we can reconstruct nowhere near
to the origins of language]
>
>> >I'm sorry, but you will find no linguist on earth who takes this
>> >seriously. The idea that languages have words built up from
>> >meaningful single sounds was very popular in the 17th and 18th
>> >centuries, before historical linguistics was invented, but it is
>> >long dead.
>
>> Foster, Pulleyblank, Hewes, von Raffler-Engel, Fonagy, and de
>> Grolier all published papers devoted to exploring early
>> isomorphisms between sound and meaning in the 1983 Glossogenetics
>> volume edited by Eric de Grolier.
>
>I take it that I am supposed to understand from this paragraph that,
>contrary to my statement, there are just bags of linguists out there
>actively working on the idea that a "single speech sound = single
>meaning" stage is still detectably represented in living or
>reconstructed languages.
>
>Wrong.
>
>There is no Foster in the book. I assume the reference must be to
>Fischer -- who is an anthropologist, not a linguist.
Mary LeCron Foster is a linguist at Berkeley. Her article appears
on pages 455-480. It is quite possible that she has had trouble
getting published in the journals that you normally read. However,
she has published a number of papers and monographs since the 1970s
on this subject. Let me quote a little bit of what she says in her
article in this 1983 Glossogenetics volume:
"I rejected the received linguistic wisdom that early linguistic
horizons could not be reconstructed by means of the comparative
method because I saw little or no attrition of vocabulary in
reconstructed Indo-European, virtually all of the roots of which have
been recovered and are available for comparison with those of other
languages, reconstructed or not. I reasoned that roots were unlikely
to disappear, since derivational processes virtually assured their
preservation in at least some words of any language.
This assumption has proved correct.
"An unexpected, but especially significant byproduct of this
research, from the standpoint of uncovering the sources of symbols,
has been the discovery that sound and meaning were originally
isomorphic, in the sense that spatial actions and relationships were
replicated gesturally through manipulation of the oral cavity as the
gestural instrument. Since these earliest linguistic forms were
neither phonemes nor morphemes in the conventional sense, I applied
the term phememe to this smallest unit."
>That leaves Pulleyblank. He never seems to admit to any particular
>academic discipline or university department. As far as I know, he
>is a specialist in Chinese language and culture. However, he has
>unquestionably published in linguistics, and even in historical
>linguistics. The works of his that I have seen are decidedly
>speculative in nature, and this one is no exception: it is
>*extremely* speculative, indeed breathtakingly so. Still, I am
>willing to count him as a linguist.
>
>So, to contradict my statement, you've got Pulleyblank and a brief
>and non-specific article by one other person. Not much, is it?
How did you miss Foster?
Regards,
John Halloran
------------------------------
Date: Sun, 10 May 1998 16:38:16 -0700 (MST)
Subject: EvolLang: The Swadesh List is NOT Synonymous with Glottochronology
>Sent by: "Dr. John E. McLaughlin" <mclasutt@brigham.net>
>
>I think that I have read two or three posts lately talking about
>(and criticizing) the Swadesh List and Glottochronology in the same
>breath. (I love the speech metaphor here.) The Swadesh List
>(depending on which version one uses) is a list of 100 or 200 words
>that tend to be more resistant to replacement in the language than
>other words. It provides a very useful place to begin looking for a
>relationship between two languages or language families. It is also
>useful in the sense that it provides a framework on which an
>evaluation of a proposed relationship can be build. If a
>reconstruction of a proto-language includes much of the Swadesh
>List, then it is generally more widely accepted at first glance than
>a reconstruction that includes only very few forms from the list.
>While it is not universally accepted as a valid list, it seems to
>be more useful than not and many historical linguists use it without
>reservation.
As a standardized list, its use helps to ensure that the comparative
linguist has not chosen to compare just those word meanings that
support the linguist's hypothesis.
>Glottochronology, on the other hand, is a use of the list that has
>been widely discredited. One compares the items in the list from
>two languages and then calculates how many of the items between the
>two lists are replacements. A formula then gives us a number of
>centuries that the two languages have been separated. No one relies
>on this method of dating language separation anymore.
Isidore Dyen et al. have a very informative collection of data for
Indo-European up at a web site. Particularly interesting is their
rating of how many word changes particular meanings are likely to
experience over a 2400 year period unit. This is based on their
data for 95 Indo-European language families, data which can be
examined at the web site. The data and their rating of it advances
considerably on Morris Swadesh's 1952 selection of a basic
change-resistant vocabulary, in that they identify more precisely
those words on which a linguist who wants to make long-range
comparisons should concentrate.
http://www.ldc.upenn.edu/ldc/service/comp-ie/IE-RATE1
050 0.01 FIVE
057 0.01 FOUR
078 0.01 I
169 0.01 THREE
176 0.01 TWO
182 0.01 WE
188 0.01 WHO
075 0.03 HOW
100 0.05 NAME
184 0.06 WHAT
186 0.07 WHERE
109 0.08 ONE
168 0.08 THOU
172 0.09 TONGUE
- -
[intermediate rate word meanings]
- -
153 3.28 STICK (OF WOOD)
120 3.61 ROAD
141 3.63 TO SMELL (PERCEIVE ODOR)
193 3.71 WIPE
009 4.93 BECAUSE
150 5.05 TO STAB (OR STICK)
149 5.28 TO SQUEEZE
029 6.23 DIRTY
The first number is the word number in the Swadesh 200 word list.
The second number is the rate at which this word was replaced during
a 2400-year time unit period. The data is not presented at their web
site sorted in this order, but anyone who wants to can save the page
and use the DOS SORT command to sort on column 5, as I did.
Examination of the data shows that for long-range comparisons one
should stick with the most change resistant word meanings at the top
of the list, as even for replacement rates of 1.0, many I-E languages
had innovated unrelated words for the meanings in question. I like
the quantitative foundation and definition that this work gives to
long distance language comparisons. It is quite satisfying to look
at the 95 language exemplars for the most change resistant words, to
see how little innovation there has been for these words over
approximately 5 or 6 K years, and to compare their forms to what G.
Decsy shows in his dictionary of Proto-Indo-European.
Regards,
John Halloran
------------------------------
Date: Tue, 28 Jul 1998 14:16:51 -0700 (MST)
Subject: Re: EvolLang: Re: Iconicity
>Sent by: "Larry Trask" <larryt@cogs.susx.ac.uk>
[snip]
>We call a sign `iconic' whenever we can detect a direct correlation
>between its semantic value and its form, but `arbitrary' when we
>cannot. As a general rule, no sign is ever totally iconic, for the
>reason I gave earlier: the only totally iconic representation of a
>non-linguistic object is the object itself. So, what we call an
>`iconic' sign is necessarily only partly iconic, and still partly
>arbitrary.
>
[snip]
>
>OK. The English word `tree' appears to be totally arbitrary, like
>most words in spoken languages: I can detect no iconic element at
>all. The same goes for Spanish <arbol>, Basque <zuhaitz>, German
><Baum>, Russian <derevo>, Hungarian <fa>, Modern Greek <dhendro>,
>Turkish <agac>, and every other word I know of in a spoken language.
[snip]
I hadn't thought about the iconicity of Sumerian gish for 'tree'
before, but this word fits in with the rest of Sumerian's basic
vocabulary in this regard. The iconism is to go from a
representation of long and narrow to a representation of a multitude.
In Sumerian gu meant 'needle', gu2 meant 'neck', and gi meant
'reed'. esh meant 'much, many'. One goes from making a sound where
the throat is long and narrow to making a splashy, mushy continuous
sound. Here we have an iconic representation of an object with a
long, narrow trunk and a scattered multitude of branches and leaves.
The basic vocabulary of Sumerian dates back to a time before the
relation between sound and meaning could be completely arbitrary.
Regards,
John Halloran
------------------------------
Date: Wed, 29 Jul 1998 15:23:32 -0700 (MST)
Subject: Re: EvolLang: Re: Iconicity
>Sent by: mcv@wxs.nl (Miguel Carrasquer Vidal)
>
>>John A. Halloran wrote:
>
>>I hadn't thought about the iconicity of Sumerian gish for 'tree'
>>before, but this word fits in with the rest of Sumerian's basic
>>vocabulary in this regard. The iconism is to go from a
>>representation of long and narrow to a representation of a
>>multitude. In Sumerian gu meant 'needle', gu2 meant 'neck', and
>>gi meant 'reed'.
>
>But isn't "tree" <g~is^>, with g~- (something like [N], [Nw] or
>[gw]) instead of g-?
You must know that g~is^2, g~es^2 means 'penis'. I think that this
word influenced the pronunciation of gis^1, 'tree; tool; organ'.
g~is^2 has an etymology that exactly parallels the etymology of
nitah, 'male'.
The g~ phememe meant 'self' as in g~e26; g~a2 meaning 'I; myself' and
in g~u10 meaning 'my; mine'. The appropriateness of the iconic
representation comes from g~ being the most internal of the resonants.
The s^ phememe meant 'many' as in es^, 'many, much; to anoint'. The
combination of the phememes describes the penis, by referring to an
object that turns the self into many.
The correctness of this etymology is shown by the parallel to nitah,
'male; man' from ni2,'self', + tah,'to multiply'.
So I think that the g~ sound is original to g~is^2, 'penis', and that
the g sound as in gu, gu2, and gi was original to gis^1, with the
specific meaning of 'tree' as something long and narrow which
branches into a multitude. But you can see how the analogies between
what a penis does and what a tree does are similar in the Sumerian
language and way of thinking.
Regards,
John Halloran
------------------------------
Date: Thu, 30 Jul 1998 13:33:23 -0700 (MST)
Subject: RE: EvolLang: Re: Iconicity
>Sent by: "Patrick C. Ryan" <proto-language@email.msn.com>
>> The appropriateness of the iconic representation comes from g~
>> being the most internal of the resonants.
>
>So far as I know, /g/ is not a resonant nor yet even a sonorant.
/g/ is an obstruent. /g~/ represents the Sumerological convention of
depicting IPA ENG, the voiced velar nasal [e.g., sing, thing, think],
as a g with a tilde over it. It is not an obstruent, but a resonant or sonorant.
>> The s^ phememe meant 'many' as in es^, 'many, much; to anoint'.
>> The combination of the phememes describes the penis, by referring
>> to an object that turns the self into many.
>
>This type of philosophical interpretation of ancient etymologies is
>interesting but surely wrong. Our ancient ancestors were always much
>more concrete and direct in naming the objects of their environment.
Modern children have been shown to use new words incorrectly for both
objects of similar shape as well as similar function. You are
denying that the inventors of language attempted to fashion icons that
represented objects by their function.
You asked about the distinction in meaning between g~ as self and n
as self. g~ is the first person self while n is the third person self.
/n/ formed with the tip of the tongue refers to discrete others. /g~/
is formed deep in the throat in a way that makes it inappropriate for
speakers using sounds as icons to represent 'an outside object or
self'.
The Sumerian word for 'word', inim, combines this idea of 'discrete
individual' with eme, the word for 'tongue; speech'. Some
comparative linguists have seen a borrowing between Sumerian inim,
'word', and Indo-European enomn, 'name'. If true, because the word
is analyzable in Sumerian, it is a borrowing from proto-Sumerian into
proto-Indo-European, not the other way around.
Regards,
John Halloran
------------------------------
Date: Fri, 31 Jul 1998 00:36:17 -0700 (MST)
Subject: EvolLang: Using Indo-European to Date Language Origin
Linguist Isidore Dyen has a web site that details the 200 words of
the Swadesh list in 95 Indo-European languages. It shows that for
certain basic words there has been extremely high conservation among
all 95 languages.
The address for this web site is:
http://www.ldc.upenn.edu/ldc/service/comp-ie/IE-RATE1
(For more details, see my post to this list of May 10, 1998.)
The most change resistant words are shown below, where the first
number is the ID in the Swadesh list and the second number is the
rate of change in a 2400 year period. At a change rate of .01, the
words at the top of the list have an even chance of being replaced
only every 240,000 years.
050 0.01 FIVE
057 0.01 FOUR
078 0.01 I
169 0.01 THREE
176 0.01 TWO
182 0.01 WE
188 0.01 WHO
075 0.03 HOW
100 0.05 NAME
184 0.06 WHAT
186 0.07 WHERE
109 0.08 ONE
168 0.08 THOU
172 0.09 TONGUE
The degree of conservation over at least a 5,000 year period is so
high among the 95 different languages for these words that it is
inconceivable that there would not have been similar conservation for
these words if one went back in time even farther to the hypothetical
ancestor of Proto-Indo-European and looked at these words in the
language families that might descend from that ancestor. But one
does not find these same words even in the languages supposedly
descended together with PIE from 'Nostratic'. Because of this, one
cannot look at the conservative similarity of these words in all 95
IE languages and honestly believe that Indo-European is descended
in an organic way from a single world proto-language spoken 40,000 or
100,000 years ago. If Indo-European extended back in time like that,
it would share these conservative words with other language families,
which it does not. Therefore, the Indo-European vocabulary and
language began relatively recently, just like the other
proto-languages, probably dating no farther back than the Neolithic.
Regards,
John Halloran
------------------------------
Date: Thu, 30 Jul 1998 13:33:24 -0700 (MST)
Subject: Re: EvolLang: Re: Iconicity
>Sent by: "Dr. John E. McLaughlin" <mclasutt@brigham.net>
>
>> Sent by: "Patrick C. Ryan" <proto-language@email.msn.com>
>>
>> In my view, the Proto-Language underlies all languages, including
>> Sumerian.
>>
>> I believe that John's view is that the Sumerians invented
>> language, which spread by contact to non-Sumerians.
>
>Just to add context to the naming conventions here without
>increasing syntax :-), John H(alloran) thinks that the Sumerian
>language was some kind of independent invention by the
>Proto-Sumerians. John McL(aughlin), me, thinks that John H is wrong
>and that Sumerian is just one of all the other human languages with
>a common ancestor somewhere far, far back in the mists of time.
John M.,
I lost interest in pursuing these arguments on this list when I
discovered that the information contradicting the view of gradual
evolution has been known and available to linguists since the 1970s.
On May 10, 1998 I posted to the EvolLang list my discovery of Isidore
Dyen's web site containing details on the 200-word list in 95
Indo-European languages. It shows that there has been extremely high
conservation among all 95 languages for certain basic words.
The degree of conservation over at least a 5,000 year period is so
high among 95 different languages for these words that it is
inconceivable that there would not have been similar conservation
for these words among all the language families that descended from
the hypothetical ancestor of Proto-Indo-European. The most change
resistant words are shown below, where the first number is the ID in
the Swadesh list and the second number is the rate of change in a
2400 year period. At a change rate of .01, the words at the top of
the list have an even chance of being replaced only every 240,000
years.
It shows intellectual dishonesty for linguists to refuse to analyze
or account for these facts. One cannot look at the conservative
similarity of these words in all 95 languages and honestly believe
that Indo-European is descended in an organic way from a single world
proto-language spoken 40,000 or 100,000 years ago. If Indo-European
extended back in time like that, it would share these conservative
words with other language families, which it does not.
I guess I feel that if linguists have known these facts for 30 years,
and have not drawn any conclusions from them, then it is hopeless
talking to linguists.
050 0.01 FIVE
057 0.01 FOUR
078 0.01 I
169 0.01 THREE
176 0.01 TWO
182 0.01 WE
188 0.01 WHO
075 0.03 HOW
100 0.05 NAME
184 0.06 WHAT
186 0.07 WHERE
109 0.08 ONE
168 0.08 THOU
172 0.09 TONGUE
The address for Isidore Dyen's web site is:
http://www.ldc.upenn.edu/ldc/service/comp-ie/IE-RATE1
Regards,
John Halloran
------------------------------
Date: Fri, 31 Jul 1998 11:44:31 -0700 (MST)
Subject: RE: EvolLang: Using Indo-European to Date Language Origin
>Sent by: "Patrick C. Ryan" <proto-language@email.msn.com>
>
>> Sent by: John A. Halloran
>>
>> Linguist Isidore Dyen has a web site that details the 200 words of
>> the Swadesh list in 95 Indo-European languages. It shows that for
>> certain basic words there has been extremely high conservation
>> among all 95 languages.
>>
[snip]
>>
>> But one does not find these same words even in the languages
>> supposedly descended together with PIE from 'Nostratic'.
>
>Absolutely incorrect. I am just finishing off an essay at my website
>where I show *easily* a hundred cognates between IE and two AA
>languages: Arabic and Egyptian. And I have just included the ones
>that require no real explanations.
>
><http://www.geocities.com/Athens/Forum/2803/comparison.AFRASIAN.3.htm>
There is a very important lesson to be learned from the data that
Dyen has collected and organized at his web site. That is that the
words in the Swadesh list are extremely different in their retention
rates. If you are making comparisons involving words with a high
rate of change, it is impossible that those words have retained their
form and meaning over the time periods involved, so that any cognates
involving such words are subjective and coincidental. When the list
below says that the retention rate for 'to smell (perceive odor)' is
3.63, that means that the word used for this concept is likely to
change 3.63 times in a 2400 year period. If one looks at the
different words for these high-change concepts in the 95 IE languages,
one sees very few cognates. This is to be contrasted with the .01
rate words shown in yesterday's message, where almost all of the words
have similar form and meaning across the 95 languages, even though in
some cases five or six thousand years separate the languages involved.
WORDS WITH THE HIGHEST RATES OF CHANGE IN IE
153 3.28 STICK (OF WOOD)
120 3.61 ROAD
141 3.63 TO SMELL (PERCEIVE ODOR)
193 3.71 WIPE
009 4.93 BECAUSE
150 5.05 TO STAB (OR STICK)
149 5.28 TO SQUEEZE
029 6.23 DIRTY
To do real long-distance comparison, a comparative linguist must
avoid word concepts that have a high rate of change. Does your work
make an effort to do this? What are the change rates assigned by
Dyen et al. to the 100 IE-AA cognates in your list?
I looked at your list. Your entries are too diffuse. Can you
simplify them into one or two word definitions, similar to the
Swadesh list words?
Regards,
John Halloran
------------------------------
Date: Sun, 2 Aug 1998 10:25:25 -0700 (MST)
Subject: Re: EvolLang: Using Indo-European to Date Language Origin
>Sent by: mcv@wxs.nl (Miguel Carrasquer Vidal)
>
>Limiting myself to Proto-Uralic, I find:
>
>*mi(nV) "me, I" (PIE *me(ne))
>*si(nV) "you" (PIE *tu:)
>*me "we" (PIE *me:s/*we:s)
>*ken- "who" (PIE *kwi-s, and the other "[w]h" words)
>*nima" "name" (PIE *Hnomn-)
>
>So, apart from "tongue" (which is problematical in PIE itself) and
>the numerals, we have reasonable potential cognates in Uralic for
>all the "stable items" above.
It is historically possible for PIE and Proto-Uralic to have
descended from an earlier language. Both proto-languages date to
about 4,000 B.C. There are no natural barriers between the
Proto-Uralic homeland in eastern Russia and the Proto-Indo-European
homeland somewhere north of Anatolia. Furthermore, if I am correct in
dating the simplest vocabulary of the proto-Sumerians to about 7,000
B.C., that leaves 3,000 years for language to develop to the north.
>The numerals as we know them in PIE indeed probably go back to the
>Neolithic.
Why do you think that is? Decsy does not give proto-Uralic forms
for the numerals, but looking at Finnish and Hungarian, the words
for four, five, and six could be cognate. Are number words just not
necessary to a hunting-gathering existence? Are number words part
of being civilized, part of the Neolithic revolution?
The clay counting tokens that appear throughout the Near East
starting in 8,000 B.C. never appear as far north as Anatolia.
The probable nonexistence of numeral words among the speakers of
proto-PIE-Uralic is certainly something to think about for those who
believe that humans already had fully modern speech in the
Paleolithic.
Regards,
John Halloran
Date: Mon, 03 Aug 1998 12:18:06 GMT
Subject: Re: EvolLang: Using Indo-European to Date Language Origin
Sent by: mcv@wxs.nl (Miguel Carrasquer Vidal)
>John A. Halloran wrote:
[mcv:]
>>The numerals as we know them in PIE indeed probably go back to the
>>Neolithic.
>
>Why do you think that is?
Because the stability of the numerals seems to be a post-Neolithic
phenomenon, at least in the language families I'm more or less
familiar with from Europe, Northern Africa and Western Asia.
It's true that the words for 2-10 are remarkably stable in
Indo-European, but the most archaic member of the group, Hittite,
shows a quite unrelated form for "4" (meiu-). Finno-Ugric has common
words for 1-6, but the Samoyed numerals appear to be quite unrelated.
In Afro-Asiatic, Semitic, Egyptian and Berber show several parallel
forms, but they have few if any in common with Chadic, Beja-Cushitic
or Omotic.
There is also evidence for quite extensive borrowing of numerals:
Kartvelian seems to have borrowed 4-10 from Indo-European and/or
Semitic, the Indo-European (also Etruscan and Basque) words for 6 and
7 are likely to be directly or indirectly Semitic borrowings, while
Uralic has borrowed words for 7 and 10 from Indo-European.
>Are number words just not necessary to a hunting-gathering
>existence? Are number words part of being civilized, part of the
>Neolithic revolution?
I don't know. In North-Western Eurafrasia, at least, it seems that
earlier systems were quite generally replaced by decimal systems of
counting [assembled from pre-existing elements, neologisms and
borrowings] sometime in the Neolithic (after PAA or PU, before PSem
or PFU).
=======================
Miguel Carrasquer Vidal
mcv@wxs.nl
Amsterdam
Date: Mon, 3 Aug 1998 11:05:52 +0100 (BST)
Subject: Re: EvolLang: numerals
Sent by: "Larry Trask" <larryt@cogs.susx.ac.uk>
> John Halloran writes:
> It is historically possible for PIE and Proto-Uralic to have
> descended from an earlier language.
This is possible, but it is not known to be so. There is a tiny
amount of evidence for an IE-Uralic link, mainly the personal
pronouns, but the available evidence is too sparse to provide solid
support for such a link.
> Both proto-languages date to about 4,000 B.C. There are no natural
> barriers between the Proto-Uralic homeland in eastern Russia and the
> Proto-Indo-European homeland somewhere north of Anatolia.
> Furthermore, if I am correct in dating the simplest vocabulary of
> the proto-Sumerians to about 7,000 B.C., that leaves 3,000 years
> for language to develop to the north.
No comment.
[somebody else]
> >The numerals as we know them in PIE indeed probably go back to the
> >Neolithic.
> Why do you think that is? Decsy does not give proto-Uralic forms
> for the numerals, but looking at Finnish and Hungarian, the words
> for four, five, and six could be cognate.
Specialists have reconstructed Proto-Finno-Ugric numerals from one to
six, and reportedly also for 20 and 100. I haven't seen any
reconstructions for 7-10; it appears that borrowing has interfered
here. But these reconstructions apply only to PFU. I have seen no
reconstructed numerals for Proto-Uralic, and I have the impression
that numerals cannot in general be reconstructed for PU. Certainly
the Samoyed numerals don't look much like the Finno-Ugric ones, for
what that's worth.
[An aside: Samoyed and Finno-Ugric are the two main branches of
Uralic.]
> Are number words just not necessary to a hunting-gathering
> existence?
Apparently not.
> Are number words part of being civilized, part of the Neolithic
> revolution?
Large sets of counting numbers appear to be a rather recent
development in human affairs. For PIE, we can reconstruct numerals
from one to ten, though there appear to be at least two different
words for `one'. But we can't, in general, reconstruct numerals
above ten. The several branches of IE have very different
formations for these numbers, variously involving addition,
subtraction, multiplication and even fractions. It rather looks as
though the higher number-names were created independently in the
daughter languages after the break-up of PIE.
There *is* a reconstructible PIE word which is often assigned the
meaning `100', but not all specialists are persuaded that's what it
meant in PIE. Some think it only meant something like `a large
number', which is exactly what it does mean in Homeric Greek.
Moreover, there is reason to believe that the PIE numerals from one
to ten may have been a late creation in that language.
Plausible-looking etymologies have been offered for the lower
numerals, while W. P. Lehmann, in his well-known book on PIE, is
deeply suspicious of the numerals above `five', which he regards as
being of very late formation or, in some cases, even of being
borrowed.
> The probably nonexistence of numeral words among the speakers of
> proto-PIE-Uralic is certainly something to think about for those
> who believe that humans already had fully modern speech in the
> Paleolithic.
The existence of proto-IE-Uralic is very far from being established,
and the label should not be used as though it represented a piece of
reality.
However, as I noted above, it seems doubtful that numerals can even
be reconstructed for Proto-Uralic, let alone for anything more
remote.
But this is no problem. It is simply not true that a fully modern
human language requires a large set of numerals. Very many languages
which are spoken today have no counting system beyond two or three,
or sometimes as high as six or ten by reduplication, such as
`two-two-two' for `six'. This is true of all Australian languages,
of some Amazonian languages, of some Khoisan languages, of some New
Guinean languages, and, with complications, of quite a number of
languages on every continent except Europe. Probably all of these
are languages spoken by foragers who lack pastoralism.
The obvious guess is that these languages lack large counting systems
because their speakers have nothing they need to count. It has been
suggested that large counting systems first arose among pastoralists
for whom counting animals was vital. I wouldn't want to defend this
specific view, but I do endorse the more general view that you
acquire numerals only when you need them.
Pastoralists typically have large vocabularies for naming varieties
of domestic animals, distinguished by sex, age and reproductive
status (at least). Foragers have no such words because they don't
need them. Complex technological societies have large sets of color
terms in their languages; the languages of foragers usually have
quite small sets of color terms.
Counting numbers, then, are not an intrinsic feature of human
languages, on a par with deixis or anaphora. They are strictly a
cultural development which may, and usually will, arise in a society
which finds it has need of them.
In our own society, the demands of scientists. mathematicians and
engineers have led to the introduction of fractions, decimals,
negative numbers, zero, irrational numbers, imaginary numbers,
complex numbers, and all sorts of further exotica, not to mention the
standard scientific way of naming and writing very large and very
small numbers. But this a cultural development, not a strictly
linguistic one. Most languages lack the terminology required to talk
about these things, but they can readily acquire it when they need
it. Thirty years ago, Basque had no way of talking about these
things; today it does, because Basque-speakers have decided they want
to talk about these things in their own language.
Hence the absence of counting numbers is no argument for the
non-modern status of a language. Very likely, around 10,000 years
ago no language had a large set of counting numbers, but that's not
because the languages were less than fully modern: it's because
speakers didn't need them.
Larry Trask
------------------------------
Date: Sun, 2 Aug 1998 10:25:34 -0700 (MST)
Subject: Re: EvolLang: Using Indo-European to Date Language Origin
>Sent by: "Larry Trask" <larryt@cogs.susx.ac.uk>
>
>> John Halloran writes:
>
>> There is a very important lesson to be learned from the data that
>> Dyen has collected and organized at his web site. That is that
>> the words in the Swadesh list are extremely different in their
>> retention rates.
>
>> To do real long-distance comparison, a comparative linguist must
>> avoid word concepts that have a high rate of change.
>
>Yes, this has long been realized, and linguists have tried to
>address it by compiling shorter lists of seemingly more stable
>words. But there's a big problem with this. The shorter your list,
>the smaller your database, and the less secure are any results you
>might obtain.
The reality of change creates what you and other linguists have
called the 10K barrier. For almost all of the rapidly changing
vocabulary, this is a possible argument.
I was pointing out that Isidore Dyen et al. found words in
Indo-European that had a change rate per 2400 years of .01, which
would smash through that barrier, if it were real.
>About twelve years ago Aharon Dolgopolsky made a study of 140
>Eurasian languages and came up with a list of what he regarded as
>the words absolutely most resistant to replacement. This list
>contained just 15 words, which is hardly enough to do a lot of work
>with. Moreover, of those 15 words, he found that the most stable
>(`I/me') had been replaced in none of the languages, while the least
>stable (`dead') had been replaced in 25% of them. Sorry I can't
>recall what his time-unit was, but it must have been several
>thousand years.
And how did you and other linguists react at the time to his work?
What conclusions did you draw regarding the prehistory of those
Eurasian languages?
>Another point. Counts like these can only be made on families whose
>prehistory is so well known that cognate words can be identified
>with certainty. We cannot look at word-lists from arbitrary
>languages and decide which words from different languages are
>historically the same word. That is impossible.
Are you saying that one should only look to compare these words
between reconstructed proto-languages? Or are you saying more than
that? It is true that if there has been significant sound shift, the
words may have the same ultimate origin, but with few exemplars it
would be difficult to set up the sound correspondences and history.
In a related message, Pat Ryan asked to see samples from Dyen's
database of words that illustrate low change versus high change. I
will paste in the 95 language exemplars for 'two' and 'three' with
change rates of .01 per 2400 years and 'dirty' with a change rate of
6.23 per 2400 years. The web site source is:
http://www.ldc.upenn.edu/ldc/service/comp-ie/IE-DATA1
If you want to see all of the available documents, just eliminate
IE-DATA1 from the address.
a 176 TWO
b 002
176 59 Gujarati BE
176 56 Singhalese DEKA
176 58 Marathi DON
176 30 Swedish Up TVA
176 31 Swedish VL TZVA
176 24 German ST ZWEI
176 35 Icelandic ST TVEIR
176 34 Riksmal TO
176 32 Swedish List TVA
176 33 Danish TO
176 36 Faroese TVEIR
176 29 Frisian TWA
176 28 Flemish TWEE
176 25 Penn. Dutch ZWAY
176 26 Dutch List TWEE
176 27 Afrikaans TWEE
176 38 Takitaki TOE
176 37 English ST TWO
176 09 Vlach DWAWE
176 17 Sardinian N DUOS
176 18 Sardinian L DUOS
176 15 French Creole C DE
176 68 Greek Mod DHIO
176 66 Greek ML DUO
176 70 Greek K DUO
176 67 Greek MD DUO
176 69 Greek D DUO
176 08 Rumanian List DOI
176 11 Ladin DUOS
176 19 Sardinian C DUSU
176 10 Italian DUE
176 23 Catalan DOS
176 20 Spanish DOS
176 12 Provencal DOUS, DOS
176 14 Walloon DEUS
176 16 French Creole D DE
176 13 French DEUX
176 21 Portuguese ST DOUS
176 22 Brazilian DOIS
176 87 BYELORUSSIAN P DVA
176 45 Czech DVA, DVE
176 90 CZECH P DVA
176 43 Lusatian L DWA
176 44 Lusatian U DWAJ
176 93 MACEDONIAN P DVA
176 50 Polish DWA
176 88 POLISH P DWA
176 51 Russian DVA
176 85 RUSSIAN P DVA
176 54 Serbocroatian DVA
176 92 SERBOCROATIAN P DVA
176 46 Slovak DVA
176 89 SLOVAK P DVA
176 42 Slovenian DUA
176 91 SLOVENIAN P DVA
176 86 UKRAINIAN P DVA
176 94 BULGARIAN P DVA
176 41 Latvian DIVI
176 40 Lithuanian ST DU, DVI
176 39 Lithuanian O DU, DVI
176 52 Macedonian DVA, DVE
176 47 Czech E DVA
176 49 Byelorussian DVA
176 48 Ukrainian DVA, DVOJE
176 53 Bulgarian DVE
176 55 Gypsy Gk DUI
176 73 Ossetic DUUAE
176 57 Kashmiri ZAH (DOYE = OBL.)
176 64 Nepali List DUI
176 61 Lahnda DU
176 78 Baluchi DO
176 02 Irish B DA
176 01 Irish A DO (DHA)
176 03 Welsh N DAU
176 04 Welsh C DAU
176 05 Breton List DAOU (M), DIOU (F)
176 06 Breton SE DEU
176 07 Breton ST DAOU
176 77 Tadzik DU
176 76 Persian List DO
176 63 Bengali DUI
176 62 Hindi DO
176 60 Panjabi ST DO
176 65 Khaskura DUITA, DUI
176 74 Afghan DVA
176 75 Waziri DWA
176 79 Wakhi BOJ
176 81 Albanian Top DY
176 80 Albanian T DY
176 83 Albanian K DI (M.), DII (F.)
176 84 Albanian C DI
176 82 Albanian G DY
176 95 ALBANIAN DY
176 71 Armenian Mod ERKU
176 72 Armenian List YERGU
a 169 THREE
b 002
169 71 Armenian Mod EREK`
169 72 Armenian List YEREK
169 30 Swedish Up TRE
169 31 Swedish VL TRI
169 02 Irish B TRI
169 01 Irish A TRI
169 03 Welsh N TRI
169 04 Welsh C TRI
169 05 Breton List TRI (M), TEIR (F)
169 06 Breton SE TRI
169 07 Breton ST TRI
169 24 German ST DREI
169 35 Icelandic ST THRIR
169 34 Riksmal TRE
169 32 Swedish List TRE
169 33 Danish TRE
169 36 Faroese TRIGGIR
169 29 Frisian TRIJE
169 28 Flemish DRIE
169 25 Penn. Dutch DREI
169 26 Dutch List DRIE
169 27 Afrikaans DRIE
169 38 Takitaki DRI
169 37 English ST THREE
169 17 Sardinian N TRES
169 18 Sardinian L TRES
169 15 French Creole C THWA
169 69 Greek D TREIS, TRIA
169 67 Greek MD TRIA
169 70 Greek K TREIS, TRIA
169 66 Greek ML TRIA
169 68 Greek Mod TRIA
169 40 Lithuanian ST TRYS
169 39 Lithuanian O TRYS
169 41 Latvian TRIS
169 08 Rumanian List TREI
169 11 Ladin TRAIS
169 19 Sardinian C TRESI
169 10 Italian TRE
169 23 Catalan TRES
169 20 Spanish TRES
169 12 Provencal TRES
169 14 Walloon TREUS
169 21 Portuguese ST TRES
169 22 Brazilian TRES
169 13 French TROIS
169 16 French Creole D TWA
169 86 UKRAINIAN P TRY
169 91 SLOVENIAN P TRI
169 42 Slovenian TRI
169 89 SLOVAK P TRI
169 46 Slovak TRI
169 92 SERBOCROATIAN P TRI
169 54 Serbocroatian TRI
169 85 RUSSIAN P TRI
169 51 Russian TRI
169 88 POLISH P TRZY
169 50 Polish TRZY
169 93 MACEDONIAN P TRI
169 44 Lusatian U TRI
169 43 Lusatian L TSI
169 90 CZECH P TRI
169 45 Czech TRI
169 87 BYELORUSSIAN P TRY
169 94 BULGARIAN P TRI
169 52 Macedonian TRI
169 47 Czech E TRI
169 49 Byelorussian TRY
169 48 Ukrainian TRY
169 53 Bulgarian TRI
169 55 Gypsy Gk TRIN
169 81 Albanian Top TRI (GRA)
169 09 Vlach TRE
169 57 Kashmiri TREH
169 61 Lahnda TRAE
169 79 Wakhi TROI
169 74 Afghan DRE
169 80 Albanian T TRE, TRI
169 83 Albanian K TRE (M.), TRII (F.)
169 84 Albanian C TRI
169 82 Albanian G TRE
169 95 ALBANIAN TRE
169 59 Gujarati TRAN
169 75 Waziri DRE
169 56 Singhalese TUNA
169 64 Nepali List TIN
169 58 Marathi TIN
169 63 Bengali TIN
169 62 Hindi TIN
169 60 Panjabi ST TINN
169 65 Khaskura TIN
169 78 Baluchi SAI
169 77 Tadzik SE
169 76 Persian List SE
169 73 Ossetic AERTAE
a 029 DIRTY
b 000
029 52 Macedonian
b 001
029 57 Kashmiri ACHOLU, ASORSHU, MALABORUTU
029 72 Armenian List AGHDOD
029 70 Greek K AKATHARTOS
029 11 Ladin ASCHER
029 77 Tadzik CIRKIN, IFLOS
029 25 Penn. Dutch DRECKICH
029 31 Swedish VL DYNGI
029 81 Albanian Top FELIKUR
029 84 Albanian C GHORDU
029 71 Armenian Mod KELTOT
029 56 Singhalese KILUTU
029 14 Walloon MASSI
029 08 Rumanian List MURDAR
029 09 Vlach (NE)LATE ("UNWASHED")
029 41 Latvian NETIRS
029 83 Albanian K I PERGUAM
029 55 Gypsy Gk PISI
029 79 Wakhi RIM
029 86 UKRAINIAN P SKVERNYJ
b 002
029 49 Byelorussian BRUDNY
029 48 Ukrainian BRUDNYJ, POGANYJ, PASKUDNYJ
029 46 Slovak BRUDNY
029 50 Polish BRUDNY
029 88 POLISH P BRUDNY
b 003
029 45 Czech SPINAVY
029 90 CZECH P SPINAVY
029 89 SLOVAK P SPINAVY
029 47 Czech E SPINAVE
b 004
029 35 Icelandic ST SKITUGUR
029 34 Riksmal SKIDDEN
029 36 Faroese SKITIN
029 33 Danish BESKIDT
b 005
029 04 Welsh C BRWNT
029 03 Welsh N BUDR, BRWNT
b 006
029 43 Lusatian L MAZANY
029 44 Lusatian U MAZANY
029 42 Slovenian UMAZANU
029 91 SLOVENIAN P UMAZAN
b 007
029 94 BULGARIAN P MRUSEN
029 53 Bulgarian MRESNO
b 008
029 40 Lithuanian ST NESVARUS, PURVINAS
029 39 Lithuanian O PURVINAS
b 009
029 54 Serbocroatian PRLJAV
029 92 SERBOCROATIAN P PRLJAV
029 93 MACEDONIAN P PRLAV
b 010
029 24 German ST SCHMUTZIG
029 30 Swedish Up LORTIG, SMUTSIG
029 32 Swedish List SMUTSIG, OREN
b 011
029 38 Takitaki DOTI, MORSOE, MOTOMOTO
029 37 English ST DIRTY
b 012
029 07 Breton ST LOUS
029 06 Breton SE LOUS
029 05 Breton List LOUS, LOUAN, LOUDOUR, FANK, LASTEZ-, LOUI-DIK.
b 013
029 74 Afghan CATAL, XIREN
029 75 Waziri KHACHEN, KHIRAN
b 014
029 51 Russian GRJAZNYJ
029 85 RUSSIAN P GR AZNYJ
029 87 BYELORUSSIAN P HRAZKI
b 015
029 01 Irish A SALACH
029 02 Irish B SALACH
b 016
029 15 French Creole C SAL
029 13 French SALE
029 16 French Creole D SAL
029 12 Provencal SALE, ALO
b 100
029 76 Persian List KASIF
029 73 Ossetic C"IZI, C"YF
b 200
c 200 2 201
029 80 Albanian T I, E NDOHTUR
b 201
c 200 2 201
c 201 2 202
029 95 ALBANIAN ZHGRYM, NDYT
b 202
c 201 2 202
029 82 Albanian G TROK, ZHGRYM
b 203
c 203 2 204
029 19 Sardinian C BRUTTU
029 17 Sardinian N BRUTTU
029 18 Sardinian L BRUTTU
b 204
c 203 2 204
c 204 2 205
c 204 2 206
029 23 Catalan BRUT, PORCH
b 205
c 204 2 205
c 205 2 206
029 10 Italian SPORCO
b 206
c 204 2 206
c 205 2 206
c 206 2 207
029 21 Portuguese ST SUJO, PORCO
b 207
c 206 2 207
029 20 Spanish SUCIO
029 22 Brazilian SUJO
b 208
c 208 2 209
029 29 Frisian FIIS
b 209
c 208 2 209
c 209 2 210
029 28 Flemish VIES, VUIL
b 210
c 209 2 210
029 26 Dutch List VUIL
029 27 Afrikaans VUIL, SMERIG
b 211
c 211 2 212
029 59 Gujarati GENDU
029 58 Marathi GHAN, GHANERDA
b 212
c 211 2 212
c 212 2 213
029 62 Hindi GENDA, MELA
b 213
c 212 2 213
029 60 Panjabi ST MELLA
029 61 Lahnda MAELA
029 64 Nepali List GUHE, MAILO
029 78 Baluchi MELAR
029 65 Khaskura PHORI, MAILO
029 63 Bengali MOELA
b 214
c 214 2 215
029 66 Greek ML LEROS
b 215
c 214 2 215
c 215 2 216
029 67 Greek MD BROMIKOS, LEROMENOS
b 216
c 215 2 216
029 69 Greek D BROMIKOS
029 68 Greek Mod VROMIKOS
It is instructive to look at the raw data like this. Even words with
a change rate of 1 per 2400 years change too much to be used in long
distance comparison. Of course, the languages of sedentary peoples
do not change as rapidly as that of the far flung Indo-Europeans. My
point was that if a volatile language family such as IE has words
with change rates of .01 per 2400 years, then at a time depth of
40,000 years, that is still only a .17 likelihood of change for these
words. Language families that are claimed to have a relationship
within this time frame should be tested against these low change rate
words.
Regards,
John Halloran
------------------------------
Date: Mon, 3 Aug 1998 18:29:46 -0700 (MST)
Subject: Re: EvolLang: Using Indo-European to Date Language Origin
>Sent by: "Larry Trask" <larryt@cogs.susx.ac.uk>
>
>> I was pointing out that Isidore Dyen et al. found words in
>> Indo-European that had a change rate per 2400 years of .01, which
>> would smash through that barrier, if it were real.
>
>Unfortunately, there isn't enough information on Dyen's site for me
>to work out just what his numbers are supposed to mean, or how they
>were arrived it.
You probably just looked at his IE-RATE1 page. You can look at all
the available pages by clicking on the different pages shown at:
http://www.ldc.upenn.edu/ldc/service/comp-ie/
>But note that one or two people have complained that using only
>Eurasian languages gives a false picture. Outside Eurasia, lexical
>replacement is known to have been very much faster in some areas
>than in any known part of Eurasia; this is true of Australia and of
>parts of North America (at least). Hence it is doubtful that D's
>results can be extrapolated to languages generally.
I was concentrating on Indo-European as an argument against those who
claim monogenesis within say the last 100,000 years. Among the
speakers of Indo-European certain words have exhibited great
stability. If they are descended along with other language speakers
from a population speaking a parent language, I would expect the
stability in these words to extend back to the parent language.
>Sent by: mcv@wxs.nl (Miguel Carrasquer Vidal)
There is nothing to clarify or disagree with in this well-informed
post regarding the history of numeral words.
>Sent by: "Larry Trask" <larryt@cogs.susx.ac.uk>
>
>> John Halloran writes:
>
>> It is historically possible for PIE and Proto-Uralic to have
>> descended from an earlier language.
>
>This is possible, but it is not known to be so. There is a tiny
>amount of evidence for an IE-Uralic link, mainly the personal
>pronouns, but the available evidence is too sparse to provide solid
>support for such a link.
My recollection is that the grammatical features are different,
meaning that the populations separated before much in the way of
grammar had been elaborated.
Miguel has met my requirement that the .01 change rate words be
similar between the language families. I would not expect the faster
changing words to show similarity.
[snip]
>However, as I noted above, it seems doubtful that numerals can even
>be reconstructed for Proto-Uralic, let alone for anything more
>remote.
>
>But this is no problem. It is simply not true that a fully modern
>human language requires a large set of numerals. Very many
>languages which are spoken today have no counting system beyond two
>or three,
[snip]
>
>Hence the absence of counting numbers is no argument for the
>non-modern status of a language. Very likely, around 10,000 years
>ago no language had a large set of counting numbers, but that's not
>because the languages were less than fully modern: it's because
>speakers didn't need them.
The most stable number words in Indo-European are for two and three.
Yet cognates to these number words are not found in other language
families. Above you appear to say that all languages spoken today
have at least the numbers two or three. So the problem in saying
that languages were fully modern 10,000 years ago is not the absence
of a large set of numerals, but the absence of two and three in the
Eurasian languages before about 6,000 years ago.
Regards,
John Halloran
------------------------------
Date: Wed, 5 Aug 1998 10:30:45 -0700 (MST)
Subject: Re: EvolLang: Using Indo-European to Date Language Origin
>Sent by: "Larry Trask" <larryt@cogs.susx.ac.uk>
>
[snip]
[re: Proto-Uralic and PIE]
>> My recollection is that the grammatical features are different,
>> meaning that the populations separated before much in the way of
>> grammar had been elaborated.
>
>Also doesn't follow. It is established beyond dispute that the
>grammatical system of a language can change radically over time.
>Look at IE: we *know* the IE languages are related, and we know a
>great deal about the elaborate grammar of PIE, but the grammatical
>systems of (say) English, Welsh, Russian and Hindi don't have much
>in common today.
That is what I would like to see a study of, similar to the lexical
study of Dyen et al. I would like to see what if any features have
been stable across IE languages over time.
[snip]
>Anyway, the putative stability of `two' and `three' in IE *does not
>prove* that these number names are universally stable. Take
>Oto-Manguean.
>
>Oto-Manguean is one of the most secure genetic families in the
>Americas, and one of the largest. The ancestral POM has been
>substantially reconstructed to general satisfaction, and the date of
>POM has been estimated at around 6000 years or so BP -- comparable to
>PIE.
>
>Now there is a list of number words in OM languages available on the
>Web. I'm looking at these, and in particular at the words for `two'
>and `three'. Here is a sample (minus diacritics). First, for `two':
>
>we / tii / tan'en / yoho / hinowi / mno / uvi / zhu / chope / tucua /
>jami
>
>Now, for `three':
>
>nde / hniuu? / tinhun / hishu / phyu / wa?ni? / hya / nihe / chona /
>sna / enhij
>
>Now, in all likelihood, some of these words are genuinely cognate,
>but I can't spot the cognation because I don't know the phonological
>histories of the languages. But what is the probability that *all*
>of them are cognate, or even most of them? Bear in mind that some
>of the most diverse-looking forms occur in OM languages which are
>regarded as particularly closely related, such as `tii' and `mno'
>for `two', and `enhij' and `phyu' for `three'.
>
>So, barring some pretty impressive phonology, the OM languages,
>which are provably related at a moderate time-depth, do not have
>cognate words for `two' and `three', and sometimes even close
>relatives within the family do not.
>
>Conclusion: words for `two' and `three' are not universally stable,
>any more than any other words are.
This is a solid contribution of certain facts. Your conclusions,
however, derive from your assumption that the speakers of POM were
using numeral words 6,000 years ago. In reality, the situation is no
different from that of Proto-Uralic, for which it has not been
possible to reconstruct numeral words. Once numeral words were
invented, at the time of Proto-Finno-Ugric, they exhibited stability.
If the descendants of POM all independently developed numeral words
at a date significantly more recent than 6,000 years ago, it is
probably because numeral words correlate with civilization, and
civilization was slower to develop on the American continent than in
the Old World.
Thank you for your contribution regarding the history of numeral
words in the Americas. It does not, however, prove that number words
can be unstable.
Regards,
John Halloran
------------------------------
Date: Mon, 10 Aug 1998 10:04:33 -0700 (MST)
Subject: Re: EvolLang: Using Indo-European to Date Language Origin
>Sent by: mcv@wxs.nl (Miguel Carrasquer Vidal)
>
>>John A. Halloran wrote:
>
>>The Uralic and Indo-European languages could have different
>>grammatical and morphological features because such features
>>evolved differently after the time that they separated. However,
>>if certain basic lexical items are shared due to descent from a
>>common ancestor language, but no grammatical and morphological
>>features are shared, then for your position to be true, the
>>grammar and morphology must have changed faster than did the
>>lexicon.
>
>That wouldn't be surprising. Languages typically have a lot more
>lexemes than they have morphemes. I'd say Modern English shares
>only a handful of morphemes with Old English, but plenty of
>vocabulary.
"With regard to the syntax, the members of the Uralic language family
are more closely knit than any other comparable language family.
There are many striking similarities even between such geographically
distant units as Hungarian and Ostyak or Selkup. These similarities
can only constitute retention of ancestral features from the time
of the Uralic Protolanguage. The number of the secondary (late)
syntactic innovations in the branch units and single languages is
relatively small." Gyula Decsy, The Uralic Protolanguage (1990),
p. 79.
In his 1991 book, The Indo-European Protolanguage, Decsy says
"Although the archaic parts of the sentence (predicate, subject) are
probably as old as language itself, numerous further (and, apparently
late) syntactic extensions (attribute, object, adverbials) were
built up according to identical principles in the languages of the
world." p. 55. Decsy has a lot of experience looking at ancient
languages so it is interesting that he regards most syntax as a late
development.
There are a number of syntactic differences that one quickly notes
between PU and PIE. PU had a simple sentence word order of S + P +
O, whereas Decsy says the ancient word order of PIE must have been P
+ S. Also, PU displays rigid word order while PIE had flexible word
order. Possessives in Uralic generally attach to the object, while
PIE has the genitive construction which Decsy derives from the
ablative. Given the conservative nature of syntactic change in the
Uralic languages, if PU and PIE share basic lexical items because of
a common origin, either the syntax developed after that common time
or the syntax of PIE evolved away from that of PU. Which do you
think is the cleaner solution?
Regards,
John Halloran
------------------------------
Date: Mon, 17 Aug 1998 09:59:56 -0700 (MST)
Subject: Re: EvolLang: Re: Symbol and Syntax: Overview
>Sent by: "Dr. John E. McLaughlin" <mclasutt@brigham.net>
>
[snip]
>
>Proto-Language. That period of time after vocalizations and gestures
>became more formal and less iconic or emotive. A very long period
>of time in which tone, context, gesture and vocalization all worked
>together to disambiguate utterances. Probably altogether lacking
>irrealis states (future, counterfactural, possibility, probability,
>etc.). No grammar (syntax or morphology).
>
>Human Language. That period of time that we are currently living in
>after the evolution of grammar and irrealis. Gesture, intonation,
>and context continue to play roles in disambiguation.
Before vocalizations [and gestures ?] can become less iconic, they
must first be iconic. There is a conceptual barrier in deciding that
mouth activities that do not at all resemble external objects shall
be mapped to and represent external objects and activities.
Quite a few writers going as far back as Sir Roger Paget have
proposed that this conceptual leap was only made possible by trying
to make the mouth activities be iconic gestures that were in some
way analogous to the object or activity that was being represented.
Originally such gestures must have been like logograms in only being
one-to-one mappings, and only later combined several mouth activities
to express more complicated activities or objects.
Regards,
John Halloran
------------------------------
Date: Wed, 19 Aug 1998 15:44:33 -0700 (MST)
Subject: Re: EvolLang: Writing and the evolution of language (long) (but worth every line)
>Sent by: Robert Whiting <whiting@cc.helsinki.fi>
>
[snip]
>
>Most opinion seems to be that the use of names for individuals
>was a very late development, but I know of no cogent or
>compelling reason for this. Indeed if the interpretation that
>some cetacean whistles are "signature tunes" is correct, it would
>indicate that personal names or identification is one of the
>earliest applications of language.
I fully agree that there were many parallels between the spread of
the concept of writing and the rise of different writing systems
around the globe and the spread of language around the world, and so
support your thesis.
In the above case, however, I would call your attention to our
genetically evolved ability to recognize individuals on the basis of
the unique timbre and tonal characteristics of their voices. Humans
have pretty much the same skills for visual, auditory, and olfactory
recognition of individuals that our animal cousins do. If you are
referring to ambiguities in reference, surely one must already have
sophisticated speech to even be referring to different individuals
in conversation.
There is a conceptual gap between what linguists call common nouns
and proper nouns (or names). Common nouns refer to classes of
objects, such as rock, house, tree, goat, or person. I would need to
see some actual data supporting the early use of proper nouns,
especially as proper nouns in the early written languages with which
I am familiar all seem to derive from common nouns.
Regards,
John Halloran
------------------------------
Date: Wed, 19 Aug 1998 16:09:53 -0700 (MST)
Subject: RE: EvolLang: Re: Symbol and Syntax: Overview
>Sent by: "Patrick C. Ryan" <proto-language@email.msn.com>
>
>> Sent by: John A. Halloran
>>
>> My only disagreement with Pat is about CVC words or roots always
>> deriving from the combination of two earlier CV roots. If the
>> logographic meaning adhered to the mouth gesture that formed what
>> we call the consonant, then two such consonants could be combined
>> to express a more complicated idea without having to combine
>> 'words' that were C + V pairs.
>
>I do not think that early speakers actually would have been able to
>define the significance of a consonant except in combination with a
>vowel -- a little to abstract for our ancestors, I think.
Levi-Strauss found much abstract thinking in the totemic systems
established by humans living in primitive societies. I would not
underestimate our ancestors. If you say that Proto-Language speakers
used labial consonants to refer to things involving a surface, is
that not abstract thinking?
>But, let me give you an example from Sumerian.
>
>sh and h are both reflexes of earlier /x/, sh is /x/ before a front
>vowel; h is /x/ before /a/ or a back vowel.
Technically, Sumerian had /sh/ and /x/. It did not have /h/.
Would you please define or give examples of front vowels and back
vowels? And what is your point in bringing this up?
>In your analysis, you should have been able to find a commonality of
>semantics between sh and h. Did you?
I believe you know that I found a dichotomy in the same semantic
range between /sh/ and /x/, a dichotomy between 'much' and 'many'.
Regards,
John Halloran
------------------------------
Date: Thu, 20 Aug 1998 00:04:18 -0700 (MST)
Subject: EvolLang: Re: NeuroSci: images and memory
>Sent by: Matt Fraser <mof@prophet.pharm.pitt.edu>
>
>> Sent by: apapp@email.gc.cuny.edu
>>
>> It is possible that recalling something requires the recreation
>> of an image on some level but the fact that it had to be
>> RE-created suggest that it had been created before and it gets
>> Re-reated by using some form of trace of it somewhere in the
>> brain.
>
>Yes, yes, yes! This is what I am after! Any jargon out there to
>describe this? Is this what another list-member was asking about
>when he asked about "engrams", or is this a totally different
>thing. Are "engrams" still in vogue?
The Scientologists appear to have co-opted the term 'engram'.
Seriously, the mechanics of the mind and memory was a hobby of mine
at one point, so let's discuss this. Let's try to avoid jargon,
which after all substitutes cliches for clear, plain thinking and
speaking.
The original writer is correct about memory, pattern recognition,
and identification working through the recreation of a stored past
experience.
The mind consists of a multitude of feature detectors, cells that
fire for vertical lines, horizontal lines, red, green, rough, shiny,
high-pitched, low-pitched, moving, not moving, known, unknown,
important to survival, not important to survival, etc. Cells for
different feature detectors are arranged in levels of abstraction,
with the output from concrete feature detectors providing input to
more abstract feature detectors.
An experience will cause a crescendo of feature detector cells to
fire, depending on the characteristics of that experience. That
event reduces the stimulus or firing threshold between all of the
cells that were involved in the experience. Then, if there is a
new experience which triggers a significant percentage of the same
cells, some of the energy released by that new experience will flow
into the reduced threshold pathways created by the original
experience. The result will be a new firing crescendo along
the old pathways, creating a memory experience of the first
experience.
Recognition of an object or activity as similar to a previously
experienced object or activity depends entirely on whether the
feature detector pathways are the same or different. There will
always be commonality between the classes of objects and activities
that humans in different cultures build up in their minds simply
because their brains work in similar fashion, especially at the
level of the concrete feature detectors. At higher abstract levels
one can learn different associations depending on one's culture
(which includes language).
It is interesting that in speaking humans, associations have been
established between abstract class feature detectors and feature
detector cells for the sounds of speech and the shapes of writing.
The crescendo of cell firing that occurs when one smells a rose
includes the word sounds that one learned as a child are associated
with that flower and if one is literate the shapes of that word in
writing. Then if one hears just those word sounds or sees just
those written word shapes by themselves, they constitute enough of
a percentage of the entire set of cells in the pattern to trigger
firing of the entire pattern, so that the sight and smell of the
flower recur in memory as a result of the language word 'rose'.
My description of this process feels very crude compared to the
finely balanced, complex reality of what the brain/mind does.
Any questions?
Regards,
John Halloran
------------------------------
Date: Thu, 20 Aug 1998 23:13:07 -0700 (MST)
Subject: Re: EvolLang: Proper nouns
>Sent by: Robert Whiting <whiting@cc.helsinki.fi>
>
[snip]
>
>Nicknames are what are used within a group when you want to
>attract someone's attention. If there are identical nicknames,
>some way is found to disambiguate them (as we now have Larry,
>the other Larry, and the other other Larry; but at least we don't
>have Larry, Curly, and Moe :>). But to return to your earlier
>point about needing sophisticated speech to refer to individuals
>in conversation, I don't think that this is necessarily true. In
>fact, you don't even need speech if each individual in the group
>has a gesture associated with them. If you want to communicate
>with #^\ (= Suzie), but she is not in sight, you gesture to
>whoever is in sight to find/get #^\ and it happens. Names then
>may very well have been an important part of early intragroup
>communication.
Social animals such as wolves and our nonspeaking primate cousins
appear to manage okay without names. Perhaps an ethologist has
already done a study of how such animals accomplish the social tasks
that speaking humans accomplish with speech and proper names.
Decsy wrote that "Both the personal names and the geographical names
are relatively new in the member languages and/or branches of
Uralia." The oldest personal name that he finds attested in both
Hungarian and Finnish is 'sleepy', which would have originated as a
nickname.
In Mesopotamia, the geographical names are the most likely enduring
witnesses to proper names before writing. It is my impression that
when subjected to analysis these proper names have turned out to
derive from Sumerian common nouns and speech. For example, Englund
derives the name of the Tigris river, idigna, from id2, 'river', +
i3, 'impersonal verbal conjugation prefix', + g~in, 'to go', +
nominative a, "the river that goes".
When I mentioned about a conceptual gap between common nouns and
proper nouns, I was suggesting that the idea of names for specific
individuals, as opposed to the idea of names for classes of things,
did not necessarily occur together.
The idea of tribal individuality, however, appears to be very old and
was symbolised by the tribal totem and by clan or tribe marks placed
on the person, such as tattoos or scars.
If you have any more thoughts on the stages that language and writing
might have had in common, I am sure that everyone will be interested
to read them.
Regards,
John Halloran
------------------------------
Date: Fri, 21 Aug 1998 11:45:11 -0700 (MST)
Subject: Re: EvolLang: Re: NeuroSci: images and memory
>Sent by: Matt Fraser <mattf+@pitt.edu>
>
>> Sent by: John A. Halloran
>
>> It is interesting that in speaking humans, associations have been
>> established between abstract class feature detectors and feature
>> detector cells for the sounds of speech and the shapes of writing.
>> The crescendo of cell firing that occurs when one smells a rose
>> includes the word sounds that one learned as a child are
>> associated with that flower and if one is literate the shapes of
>> that word in writing. Then if one hears just those word sounds or
>> sees just those written word shapes by themselves, they constitute
>> enough of a percentage of the entire set of cells in the pattern
>> to trigger firing of the entire pattern, so that the sight and
>> smell of the flower recur in memory as a result of the language
>> word 'rose'.
>
>Right, this is the association of patterns that is important. In
>fact, it is the formation of "hyper-patterns", which once a physical
>object, event or idea are associated with a lexical item or lexical
>field, you have language at some stage or another.
The substitution of the 'part for the whole' enables both thinking
and language. I encountered this idea most strongly in books such as
Hans G. Furth, Thinking Without Language: Psychological Implications
of Deafness (1966) and Aleksandr Romanovich Luria, Higher Cortical
Functions in Man (1980). Linguists call the part for whole
substitution synecdoche, such as 'aisle' for 'marriage'. Thinking
can substitute the 'part' of the culture's ready-made linguistic
symbols for the 'whole' of a set of experience patterns, or it can
substitute idiosyncratically developed images which represent a
larger set of experiences within the mind.
The following quote refers to how the meaning that a part has for a
person depends on their particular knowledge and experience of the
whole that the part is supposed to represent.
"Relativists like Kuhn and Hanson emphasize the importance of meaning
in terms of conceptual networks. Terms and concepts derive their
meaning from their multiple relationships with other terms: they are
nodes in a complex network. Scientific revolutions, changes of
paradigms, involve radical reorganizations of the system of
interrelated concepts in a given discipline. To understand theories
or statements produced within such a framework, one has to acquire
the semantic network. It is not sufficient to have clear empirical
references in order to have a meaningful statement for everybody.
Meaning is restricted to those who are conversant with the conceptual
framework." "Words and sentences not only apply to objects and states
of affairs; they form part of a system and they derive their meaning
as much from their relationships with other words as from their
reference to extra linguistic phenomena." Marc De May,
"Incommensurability of Theories and Untranslatability of Languages,"
in Universalism versus Relativism in Language and Thought, Rik
Pinxten (ed.), p. 272 (1976).
Lexical items can only function as indexes when there is a larger
set of experience patterns for them to index. Small children may be
able to speak words for which they have no referents. I imagine that
this fact is part of the problem in getting computers to handle
natural language.
Regards,
John Halloran
------------------------------
Date: Mon, 24 Aug 1998 22:00:25 -0700 (MST)
Subject: Re: EvolLang: Re: NeuroSci: images and memory
>Sent by: Matt Fraser <mattf+@pitt.edu>
>
>> Sent by: John A. Halloran
>
><snip of good stuff>
>
>> Lexical items can only function as indexes when there is a larger
>> set of experience patterns for them to index. Small children may
>> be able to speak words for which they have no referents. <snip>
>
>Or as infants they may have referrents without lexical indices.
>
>Would we consider a dog to have a lexicon because it can respond
>correctly to various commands, even though it cannot communicate
>back in the same fashion? Does this imply that even without
>training, such associations exist in non-human species, or do we
>actively rewire their brains?
I didn't train my dog to become excited when I put on my shoes. But
he associates it with going out in the same way that he associates
the words "go for a walk" with going out. (He is also smart enough
that he puts a damper on his excitement when I ask him if he wants to
go for a walk and I have not yet put on my shoes.)
There is no question that dogs and by extension other animals make
associations between remotely related phenomena. Wasn't that the
point of Pavlov's bell ringing experiments?
My dog knows the words "dog", "cat", and "squirrel", but I don't
recall making any great effort to teach him these words. He does
appear to look for a different type of creature depending on which
word he hears.
An expert on dogs says that they can tell from scent the age, sex,
and health of visitors to the communal post. For them it is like
reading the morning newspaper. I don't know exactly what association
circuits are being stimulated by these smells, but it would be
interesting to learn what association cortexes, such as visual,
auditory, etc., if any, fire as a result of what they learn. I think
one of the reasons that dogs go into such a marking frenzy when they
meet another dog is because it is important for them to be able to
identify a particular dog with that dog's scent. In this case the
scent functions as an index to the collection of animals in
the individual dog's experience.
Dogs and humans also have the ability to identify what caused
particular sounds. If I hear a high pitched squeal in the afternoon,
I no longer have to look out to know that the mail truck is
delivering the mail to each box along the street. If my dog hears a
knocking sound, he associates that with someone being at the front
door, and barks to alert the rest of the house. If I hear thunder, I
know that it will soon be raining. With spoken language, we have
taken this evolved ability to have a sound be an index to a much
larger set of images and events and developed an organized set of
referents and associations for purposes of group-wide communication.
Regards,
John Halloran
------------------------------
Date: Tue, 25 Aug 1998 13:51:04 -0700 (MST)
Subject: Re: EvolLang: Proper nouns
>Sent by: Robert Whiting <whiting@cc.helsinki.fi>
>
>>Sent by: John A. Halloran
>
>>>Sent by: Robert Whiting <whiting@cc.helsinki.fi>
>>>Names then may very well have been an important part of early
>>>intragroup communication.
>
>>Social animals such as wolves and our nonspeaking primate cousins
>>appear to manage okay without names. Perhaps an ethologist has
>>already done a study of how such animals accomplish the social
>>tasks that speaking humans accomplish with speech and proper
>>names.
>
>If we actually had some idea of what various wolf howls, barks,
>yips, and growls, or primate chatterings, calls, and gestures
>actually meant, I would feel a lot better about accepting this
>statement at face value. We just don't know what these animals
>get along with or without because we don't have a clue what any
>of these things mean (vervet warning calls excepted).
Don't forget body language - such as how a wolf holds his body and
tail. I was not just thinking of vocalizations. Because I see no
evidence that animals call out the names of individuals, and yet
socially they seem to function fine, I am not convinced of the
social necessity of names.
>But even Moonhawk would agree that animals don't use Human Language;
>and names seem to be an intrinsic part of human language (at least
>certainly nowadays). But in a part of the message that you
>snipped, I pointed out that researchers believe that porpoises,
>who have brains closer to humans than any other species (unless I
>have my facts wrong), use an equivalent of names, and that this
>may be a good indication that names were in fact an early part of
>Human Language.
Aren't the 'signature whistles' to which you alluded earlier merely
devices for signaling one's identity and location, similar to the
calls of creatures in the dense jungle? If you are claiming that a
single animal produces different whistles depending on who it wants,
you need to reference this much better.
Date: Wed, 14 Oct 1998 09:39:07 -0700 (PDT)
From: Kelly Jaakkola <jaakkola@biology.ucsc.edu>
Subject: Re: EvolLang: Proper nouns
Sent by: Kelly Jaakkola <jaakkola@biology.ucsc.edu>
Hi all. I've been going back through the list messages from before
our recent rather long hiatus. So this may be a little off current
topic.
Actually, no, signature whistles aren't quite the same as other
animals' contact calls. They are different in one crucial respect.
That is, most of the time, a given dolpin produces its own signature
whistle, which signals identity & location, much like other animals'
calls. However, a given dolphin will *also* produce the signature
whistles of *other* dolphins. The current thinking on this is that
these whistles are functioning as vocal labels (i.e., "name-like"),
and that dolphins use them both to broadcast individual identity &
to establish/maintain contact with other specific individuals.
If you want refs for this, you might try either of the following:
Tyack, P. L. (1986). Whistle repertoires of two bottlenosed
dolphins, Tursiops truncatus: Mimicry of signature whistles?
Behavioral Ecology and Sociobiology, 18, 251-257.
Tyack, P. L. (1993). Animal language research needs a broader
comparative and evolutionary framework. In H. L. Roitblat, L. M.
Herman, & P. E. Nachtigall (Eds.), Language and communication:
Comparative perspectives (pp. 115-152). Hillsdale, NJ: Erlbaum.
Cheers,
Kelly
- --------------------------------
Kelly Jaakkola
jaakkola@biology.ucsc.edu
http://www-bcs.mit.edu/~kelly/
- --------------------------------
Date: Sat, 17 Oct 1998 15:31:00 -0700 (PDT)
> Does anybody know what a Dolphin does when he hears his own
> signature whistle coming back to him?
>
I checked with a colleague who has more direct experience with this
than I do. Here's what he said:
"From my experience they always seem to respond with their own
signature whistle. In other words, if dolphin A hears its signature
whistle W produced by animal B, dolphin A will also produce its own
signature whistle W in response."
I hope that helps.
Cheers,
Kelly
[regarding pre-writing place names in Mesopotamia]
>>It is my impression that when subjected to analysis these proper
>>names have turned out to derive from Sumerian common nouns and
>>speech.
>
>And it was Landsberger's impression that many of these names were
>non-Sumerian in origin that led him to postulate a pre-Sumerian
>proto-population in the area.
And Landsberger was so influential that he affected the thinking of
a whole generation of Assyriologists. I saw for myself that he was
wrong a long time ago. Now others are seeing it.
I am looking at a well-referenced article by Gonzalo Rubio which he
presented at this year's AOS meeting in New Orleans called "Was
Sumerian a Creole?" It is divided into three sections: 1. Language
Contact and Creolization; 2. Hoyrup's creolization theory; and 3. Is
There an Identifiable Pre-Sumerian Substratum? After presenting and
discussing all of Landsberger's alleged substratum words, Rubio
writes, "Actually, if one looks carefully at Landsberger's list of
substratum words, many of them happen to be Semitic loanwords,
Hurrian, Arealwo"rter or Wanderwo"rter, or have the /nu-/ prefix."
"On the other hand, one of the features of the structure of most
allegedly substratum words, a medial consonant cluster, seems to
disappear in several cases (nig~ir, nag~ar)." "It is interesting
to observe how strongly Landsberger has tried to show that dam-gar3
'commercial agent' is a typical Substratwort (/-ar/ ending and medial
consonantal cluster) which would belong to the Proto-Euphratic stock.
However, despite Landsberger's arguments, it is a clear Semitic
loanword, from Akkadian tamkaru [five references given]." "The
critiques of the substratum theory are sound and aim at the lack of
coherence of the alleged substratum [refs. to Kraus and Lieberman]."
"Since toponymy is very conservative, one should expect to find
traces of any substrate languages among place nouns." "Nevertheless,
these non-Sumerian toponyms - if they are such - do not present any
coherent pattern and little can be deduced from their existence,
besides that probably Sumerians were not the only inhabitants of the
area and perhaps not even the first ones - although it is impossible
to draw any sound conclusions on migrations, chronology, etc." "Only
a tapestry of various substrata or perhaps just adstrata comes out
of the study of these toponyms."
>>For example, Englund derives the name of the Tigris river,
>>idigna, from id2, 'river', + i3, 'impersonal verbal conjugation
>>prefix', + g~in, 'to go', + nominative a, "the river that goes".
>
>And the Assyrians analyzed the name of Erbil (A/E/Urbi[l]l[um] in
>late third millennium sources) as Arba+ilu: "four gods" as
>indicated in their writing of the city name with the signs for
>"four" and "god". But the name is not of Semitic origin. The
>city is still there (modern name Arbil or Erbil) but the modern
>Arabic spelling is <?arbi:l> with alif rather than ayin.
Do you know what language the city name *is* from?
>I am reminded of a tale that the late Klaus Baer used to tell
>about a four hour drive from Ann Arbor back to Chicago after
>an AOS meeting during which he had to endure Edmund Gordon's
>(_Sumerian Proverbs_) Sumerian etymologies of all the native
>American geographic names that they passed. "He just never
>ran out of them" was Klaus' lament.
I would like to have heard that.
>The point is that etymologizing place names is more of a game
>that anyone can play than an exact science. Like so many
>other things in language reconstruction, there is just no way
>to tell when you are wrong.
When I read the Hebrew Bible with Robert Hetzron we often ran
across folk etymologies of place names, often in the form of
etiological myths to explain why the place had that name. Some of
these etymologies were clearly fanciful. This is similar to the
Assyrians with their 'four gods' for Erbilum. Although the
Sumerians called it ka2-dingir-ra, some scholars have expressed the
belief that bab-ilani, 'gate of the gods' was a Semitic
reinterpretation of an earlier non-Semitic name for Babylon,
similar to your Assyrian example.
>>The idea of tribal individuality, however, appears to be very old
>>and was symbolised by the tribal totem and by clan or tribe marks
>>placed on the person, such as tattoos or scars.
>
>Yes, such marks served much the same purpose as uniforms for
>sports teams or armies. Helps to tell one of them from one of
>us. But such marks often told much more, not only about
>tribe/clan affiliation but also about family history and lineage.
>Medieval heraldic devices served much the same function. But
>totems are mnemonic-identifying devices and presuppose a certain
>level of language. Makes one wonder if tribes/clans would have
>names (totems) and individuals not.
Did you ever read Jean Auel's fiction book Clan of the Cave Bear?
In that book, not only does the clan have the cave bear totem, but
individuals pick their own totem, such as the lion for the heroine.
Is this what the Sumerian personal name Ur-Nisaba [Ur = dog, servant;
Nisaba = the grain goddess] originally meant, that the deified grain
was that individual's personal totem?
Totem symbols do not require spoken language. They function as
indexes, but they are not tied to the vocal-auditory channel. We
have to be open to the idea that humans used physical items and
images such as tokens and totems as pre-linguistic symbolic indexes.
These items are not nearly as abstract or as systematized as is a
system of indexical speech sounds.
Regards,
John Halloran
Date: Thu, 15 Oct 1998 17:14:02 +0300 (EET DST)
Subject: Re: EvolLang: Proper nouns
Sent by: Robert Whiting <whiting@cc.helsinki.fi>
[I wrote this last August and then did not post it because of the
hiatus. Kelly Jaakkola's recent post reminded me of it and I have
excavated it from its file and let it continue its journey. RMW]
>Sent by: John A. Halloran
>>Sent by: Robert Whiting <whiting@cc.helsinki.fi>
>>
>>>Sent by: John A. Halloran
>>
>>>>Sent by: Robert Whiting <whiting@cc.helsinki.fi>
>>>>Names then may very well have been an important part of early
>>>>intragroup communication.
<snip>
>>If we actually had some idea of what various wolf howls, barks,
>>yips, and growls, or primate chatterings, calls, and gestures
>>actually meant, I would feel a lot better about accepting this
>>statement at face value. We just don't know what these animals
>>get along with or without because we don't have a clue what any
>>of these things mean (vervet warning calls excepted).
>
>Don't forget body language - such as how a wolf holds his body
>and tail. I was not just thinking of vocalizations. Because I
>see no evidence that animals call out the names of individuals,
>and yet socially they seem to function fine, I am not convinced
>of the social necessity of names.
Body language is also a communication device, one that in fact
can be used on an interspecies level and as such is heavily
exploited by animal trainers. Body language even finds a place
in human language as a metaphor ('to take a ... stance'). As
an epistemological position, it is good not to be convinced in
the absence of evidence, but my point is that we have no evidence
for or against so there is no basis for making a decision. As
far as animals are concerned, yes there is no evidence of names
being used to call individuals, but since we don't know what the
calls being used are used for, we can't say that they don't.
As far as humans are concerned, names occur in the earliest
writing (which is in fact essentially lists of names and
commodities for accounting purposes) and therefore predate it
(one presumes that names were not invented just so that they
could be written down); the existence of pre-writing place names
also suggests the use of names well before this time; the word
for name itself would seem to be very old, arguably having a
similar form in several of the large and very old language
families. A name is an anaphor that allows an individual to be
called or discussed when not present. I would expect that names
came into use as soon as the capacity for using anaphoric
expressions existed. It may very well be that animals do not use
anaphoric expressions and therefore don't use names; presumably
their communication systems are so limited that they don't have
the bandwidth to waste on names. But we just don't have evidence
to prove or disprove this.
I think that the idea that (personal) names are late stems from
the idea that in earlier societies there was no concept of the
individual or individual awareness. I have even encountered
classicists who maintain that this is true until the time of the
Greeks. I find this utterly incomprehensible. Show me a
language that doesn't have a marker for the first person
singular, and then we can talk about the lack of the concept of
the individual.
>Aren't the 'signature whistles' to which you alluded earlier
>merely devices for signaling one's identity and location, similar
>to the calls of creatures in the dense jungle? If you are
>claiming that a single animal produces different whistles
>depending on who it wants, you need to reference this much
>better.
I had best just drop this idea and let those who are more familar
with the field and the data do the speculating.
[And I thank Kelly Jaakkola for commenting knowledgeably.]
>[regarding pre-writing place names in Mesopotamia]
>
>>>It is my impression that when subjected to analysis these proper
>>>names have turned out to derive from Sumerian common nouns and
>>>speech.
>>
>>And it was Landsberger's impression that many of these names were
>>non-Sumerian in origin that led him to postulate a pre-Sumerian
>>proto-population in the area.
>
>And Landsberger was so influential that he affected the thinking
>of a whole generation of Assyriologists. I saw for myself that
>he was wrong a long time ago. Now others are seeing it.
Gosh, John, don't hurt yourself trying to pat yourself on the
back :). I think if you research it, you will find that there
was immediate and continuous opposition to his position. At
least Jacobsen, Gelb, and Edzard have written against it,
probably before you were even born. Look especially at
Albright's position in _Cambridge Ancient History_ I/1 (1970,
but originally published in fascicle form earlier) where he
strongly denies Landsberger's proposed substratum. I don't think
you can claim that he "affected the thinking of a whole
generation of Assyriologists" except in the sense that he gave
them something to argue against.
I didn't say that Landsberger's postulate was correct. What I
said was that the starting point was non-Sumerian place names.
And even if one dismisses Landsberger's reconstruction as a
complete fabrication, one is still left with them.
>I am looking at a well-referenced article by Gonzalo Rubio which
>he presented at this year's AOS meeting in New Orleans called
>"Was Sumerian a Creole?" It is divided into three sections: 1.
>Language Contact and Creolization; 2. Hoyrup's creolization
>theory; and 3. Is There an Identifiable Pre-Sumerian Substratum?
<snip>
I have snipped the first part of your quotation because, although
it is sound and agrees with what I was taught (by Gelb), it is
not relevant to the present discussion. But the next part is, so
I have retained it:
>"Since toponymy is very conservative, one should expect to find
>traces of any substrate languages among place nouns."
>"Nevertheless, these non-Sumerian toponyms - if they are such -
>do not present any coherent pattern and little can be deduced
>from their existence, besides that probably Sumerians were not
>the only inhabitants of the area and perhaps not even the first
>ones - although it is impossible to draw any sound conclusions on
>migrations, chronology, etc." "Only a tapestry of various
>substrata or perhaps just adstrata comes out of the study of
>these toponyms."
Which brings us back to my original point: attempts to
etymologize toponyms do not, in the main, yield useful scientific
results, especially when we are dealing with unknown or poorly
known languages. Even if the language is well known, there is
no way of knowing, in most instances, that the name has not been
changed from an earlier, not understood, form to something that
was similar in form in the native language of the current
inhabitants and that therefore made more sense to them.
<snip>
>>And the Assyrians analyzed the name of Erbil (A/E/Urbi[l]l[um] in
>>late third millennium sources) as Arba+ilu: "four gods" as
>>indicated in their writing of the city name with the signs for
>>"four" and "god". But the name is not of Semitic origin.
<snip>
>Do you know what language the city name *is* from?
Not really. The location suggests Hurrian, but I have no
linguistic evidence that would back this up. It may be from
a pre-Hurrian stratum. If one looks at another very early
toponym from this general area, Ninua, and compares it with names
like Padua and Genua, one could say that there is Illyrian
influence in the area.
>>The point is that etymologizing place names is more of a game
>>that anyone can play than an exact science. Like so many
>>other things in language reconstruction, there is just no way
>>to tell when you are wrong.
>
>When I read the Hebrew Bible with Robert Hetzron we often ran
>across folk etymologies of place names, often in the form of
>etiological myths to explain why the place had that name. Some
>of these etymologies were clearly fanciful. This is similar to
>the Assyrians with their 'four gods' for Erbilum. Although the
>Sumerians called it ka2-dingir-ra, some scholars have expressed
>the belief that bab-ilani, 'gate of the gods' was a Semitic
>reinterpretation of an earlier non-Semitic name for Babylon,
>similar to your Assyrian example.
Again, this is exactly the point. With place names there is just
no way to tell. In general, place names lack context, the great
disambiguator of meaning. Names of cities founded in historical
times may have a clearcut and unquestionable etymology (e.g.
Constantinople), but ones that are earlier (e.g., Paris, London)
and other types of topographical features may have names that go
back to the earliest human inhabitants (with unknown shifts as
the languages of the inhabitants changed).
Etymolgizing names like this is, as I say, not science, but a
game. The problem is that there is no way to check the answer
in the back of the book. Of course, if you are one of those
who believe that if it can't be disproved it might as well be
true, this is an advantage rather than the other way around.
If you can't tell if it is wrong, neither can anybody else.
To see how far you can get with this procedure, you ought to
have a look at a book called _The Key_ by Philip Cohane. Cohane
uses Ryan's methodology (actually I believe Ryan uses Cohane's,
since I think the latter has temporal priority), in that he
posits the earliest syllables (6 in this case), assigns meanings
to them, then shows how they show up in the languages of the
world in words that demonstrate the correctness of his assigned
meanings. But this is just preliminary. He then shows how
one or more of these syllables turn up in place names from all
over the world as a final proof. Unlike Ryan, he makes no
pretense to rigor of method, ignores (apparently never heard of)
the concept of regular sound change, and since he is dealing
mainly with toponyms, does not have to worry about meanings. As
a result he has thousands of toponyms that he can explain through
his original 6 syllables, and no one can prove that he is wrong.
This is one reason why I consider etymologies of place names as
"just-so" stories; they are plausible stories that account for
the known facts. Given any set of circumstances, there are an
infinite number of scenarios to explain those facts; some will
just be more plausible than others. Now not all "just-so"
stories are false; the point is that there is no way to tell when
they are, so it really doesn't make any difference. The thing
that people tend to lose sight of is that plausible stories are
not evidence or facts -- they are substitutes for evidence.
<snip>
>>But totems are mnemonic-identifying devices and presuppose a
>>certain level of language. Makes one wonder if tribes/clans would
>>have names (totems) and individuals not.
>
>Did you ever read Jean Auel's fiction book Clan of the Cave Bear?
>In that book, not only does the clan have the cave bear totem, but
>individuals pick their own totem, such as the lion for the heroine.
Yes I did, but in that book all the individuals have names as
well as clan and individual totems so I don't think it is fair
to draw any conclusions from it. It's a good story, though.
>Is this what the Sumerian personal name Ur-Nisaba [Ur = dog,
>servant; Nisaba = the grain goddess] originally meant, that the
>deified grain was that individual's personal totem?
Who knows? But I would be more inclined to see totems reflected
in the names of the rulers of the first dynasty of Kish in the
Sumerian King list; although all Semitic, they are also all
animal names.
>Totem symbols do not require spoken language. They function as
>indexes, but they are not tied to the vocal-auditory channel. We
>have to be open to the idea that humans used physical items and
>images such as tokens and totems as pre-linguistic symbolic indexes.
>These items are not nearly as abstract or as systematized as is a
>system of indexical speech sounds.
I should have said a certain level of cognition rather than
language. Specifically, they require symbolic reasoning, which I
think we have agreed on this list is a necessary condition for
language as well as any other kind of symbolic index. I'm afraid
that I tend to link the development of cognition and the
development of language in my own mind, but I ought not to use
the two interchangeably, and I agree that (spoken) language does
not have to precede the use of visual symbolic indexes. In fact
if we want to put gestures into this latter category, it almost
certainly doesn't.
Bob Whiting
whiting@cc.helsinki.fi
------------------------------
Date: Mon, 5 Oct 1998 10:23:16 -0700 (MST)
Subject: EvolLang: Re: Before grammar, babble on, fie, etc.
>Sent by: Zylogy@aol.com
>
[snip]
>
>Bonobo's lack vocal chords only after particularly nasty radical
>surgery. What they lack is the same facility of manipulation of the
>supralaryngeal airspace, much smaller of course because they've
>never had the lowering of the larynx that our species developed as
>part of our ontogeny.- But then they've got this really flexible
>lip-muscle complex, maybe they've got some ability to signal
>with that?
Human primates diverged from bonobos by expanding the territory over
which they ranged. I think that this adaptation is responsible for
much of our brain increase as well as changes to our voice box to
enable long-distance vocalizations.
Regards,
John Halloran
------------------------------
Date: Thu, 15 Oct 1998 13:35:46 -0700 (MST)
Subject: Re: EvolLang: Proper nouns
>Sent by: Robert Whiting <whiting@cc.helsinki.fi>
>>>>>Sent by: Robert Whiting <whiting@cc.helsinki.fi>
>
>>>>>Names then may very well have been an important part of early
>>>>>intragroup communication.
>
>the word for name itself would seem to be very old, arguably
>having a similar form in several of the large and very old language
>families.
Yes, in Sumerian, PIE and Proto-Uralic. But in Afro-Asiatic, the
dictionary of Orel & Stolbova list nab- and sum- [with umlaut]. The
Sumerian word for name, inim, can be analyzed into component elements
[discrete unit + tongue], so it must precede the PIE and Proto-Uralic
usages. Speech in the west [either Palestine or Egypt/Africa]
developed independently of speech in the east [neolithic Iran]. No
one should expect otherwise given the great geographical distances
involved.
>A name is an anaphor that allows an individual to be
>called or discussed when not present. I would expect that names
>came into use as soon as the capacity for using anaphoric
>expressions existed. It may very well be that animals do not use
>anaphoric expressions and therefore don't use names; presumably
>their communication systems are so limited that they don't have
>the bandwidth to waste on names. But we just don't have evidence
>to prove or disprove this.
>
>I think that the idea that (personal) names are late stems from
>the idea that in earlier societies there was no concept of the
>individual or individual awareness.
Yes, that has been an intriguing idea that has been floating around.
Julian Jaynes explicitly wrote about it. But I am really arguing
about the order in which language-inventing humans invented common
nouns and proper nouns. Everything that I have seen indicates that
common nouns came first, and that early proper nouns were
constructed from existing common nouns.
There is nothing to make me think that individual humans had
linguistically meaningless signature sounds.
For one thing, it is very difficult to remember a person's name if
it is not chosen from an existing repertoire. I worked on a job once
with a young black woman who invented a name for her newborn son that
I had never heard before. That name was very hard to remember. I
cannot tell you today what it was. When she told other people her
new son's name, the common reaction was to be appalled, shocked, and
concerned about the welfare of a child with such an unusual name.
Isn't it the case that native, non-technological peoples all had
linguistically meaningful names?
Thank you for your snipped observation that the names of the Semitic
kings [from Kish ?] on the early Sumerian King List all had animal
and therefore possibly totem names.
Regards,
John Halloran
------------------------------
Date: Mon, 19 Oct 1998 14:37:21 -0700 (MST)
Subject: Re: EvolLang: Re: Culture and Language
>Sent by: Martin Byers <ambyers@sprint.ca>
>
>Of course, our salutations are symbols. But does this necessarily
>mean that the signature calls and whistles of other species are
>symbols? I take symbols to be arbitrary. Are personal dolphin
>signatures arbitrary, even though each dolphin may have a unique
>signature? I wonder.
My reply is taken from Tyack's article in the 1993 book that Kelly
recently recommended, in H.L. Roitblat, Language and Communication,
pp. 115-152..
He says that the signature whistles of young dolphins are ill-defined
until about age six months. Then the female dolphins develop
signature whistles that are very different from that of their
mothers, while the male dolphins develop signature whistles that are
very similar to that of their mothers. The female dolphin whistles
need to be different because the females live together in a group
while the males roam independently and need to be able to recognize
brothers for banding purposes. So the males just learn their
signature whistle from their mothers, but the females are programmed
to creatively invent a signature whistle. The female whistle is not
completely arbitrary in that it needs to be different from her
mother's.
Tyack gave one illustration of a female dolphin vocalizing the
signature whistle of another, older female dolphin, when both were
with other dolphins in a new and upsetting environment. The observed
result was that the older dolphin still emitted her own signature
whistle, but began to do so in a way that was timed to the younger
dolphin's whistles. Nothing is said about changing the whistles,
which are described as pure tones, or about any social behavior other
than the change in timing of whistle generation. Tyack did
hypothesize that the younger dolphin was looking for reassurance
from the oldest dolphin, who the researchers nicknamed Granny.
The signature whistles do allow mothers and daughters to find each
other, but it is by broadcasting their own names, not by calling out
each other's name. In order for male dolphins to learn their
mother's whistles, dolphins must be born with the ability to mimic
the sounds that they hear, just as can songbirds, so this is what
gives the ability to reproduce other dolphins' signature whistles.
I don't see that the signature whistles are functioning here as
part of any abstract system of representation.
Regards,
John Halloran
------------------------------
Date: Tue, 20 Oct 1998 15:27:04 -0700 (MST)
Subject: EvolLang: yet more whistles (was: Re: Culture and Language)
>Sent by: Kelly Jaakkola <jaakkola@biology.ucsc.edu>
>
>> I don't see that the signature whistles are functioning here as
>> part of any abstract system of representation.
>
>Maybe so, maybe not. But let me ask you this: What kind of evidence
>would it take to convince you that signature whistles are
>representational/symbolic? This is not a rhetorical question. I
>really would like to know people's thoughts on this.
Kelly,
I know that you have also posted on other examples of animal
language, such as the Sue Rumbaugh results. What criteria have
researchers used to declare that chimpanzees and bonobos have learned
language as a result of human intervention? Do we see examples of
those criteria being met among dolphins not instructed in language by
humans?
Do you have a published example of a dolphin vocalizing the signature
whistle of another dolphin and having that second animal come up to
and interact with the first animal? Does it appear that the first
animal used the whistle as a tool to get what it wanted? Humans are
tool users. We are accustomed to manipulating tools to make things
that we want happen. Our long possession of hands with an opposable
thumb and fingers has caused our minds to evolve to think in terms of
manipulating tools. Names and words are tokens that we manipulate
to get what we want. Does the evidence show that dolphins manipulate
different whistles to get what they want, or are their minds
unfortunately pure, without guile?
Regards,
John Halloran
Date: Tue, 20 Oct 1998 21:03:54 -0400
From: Martin Byers <ambyers@sprint.ca>
Subject: EvolLang: yet more whistles (was: Re: Culture and Language)
Sent by: Martin Byers <ambyers@sprint.ca>
>Sent by: John A. Halloran
>
>>Sent by: Kelly Jaakkola <jaakkola@biology.ucsc.edu>
>>
>>> I don't see that the signature whistles are functioning here as
>>> part of any abstract system of representation.
>>
>>Maybe so, maybe not. But let me ask you this: What kind of
>>evidence would it take to convince you that signature whistles are
>>representational/symbolic? This is not a rhetorical question. I
>>really would like to know people's thoughts on this.
>I know that you have also posted on other examples of animal
>language, such as the Sue Rumbaugh results. What criteria have
>researchers used to declare that chimpanzees and bonobos have
>learned language as a result of human intervention? Do we see
>examples of those criteria being met among dolphins not instructed
>in language by humans?
>Do you have a published example of a dolphin vocalizing the
>signature whistle of another dolphin and having that second animal
>come up to and interact with the first animal? Does it appear that
>the first animal used the whistle as a tool to get what it wanted?
>Humans are tool users. We are accustomed to manipulating tools to
>make things that we want happen. Our long possession of hands with
>an opposable thumb and fingers has caused our minds to evolve to
>think in terms of manipulating tools. Names and words are tokens
>that we manipulate to get what we want. Does the evidence show that
>dolphins manipulate different whistles to get what they want, or are
>their minds unfortunately pure, without guile?
What John Halloran has said about the relation between tools and
cognition is particularly germaine. As I read what he is saying is
that the way our communication system works flows from the way
hominids habitually acted by using tools. If so, then we have
inherited our action orientation towards the world (change it to
satisfy our needs) from our hominid ancestors' way of using tools to
materially address the world by way of bringing about changes that
will satisfy their needs. Satisfy subsistence needs by way of making
a tool to scavenge and butcher meat. And this tendency appropriated
communicative practices to similar ends - use signs to change the
social world so as to satisfy wants and needs.
This is a pragmatic view of communciation. I have been examining
how we can understand communication in pragmatic terms because I
think it will help us to relate tools and cognition along the lines
suggested by John Halloran. As he points out, we use words to fulfill
our wants by making promises, giving orders, making delcarations,
and so on. Referencing is secondary here since it is often done by
way of promising, or demanding. That is, referncing serves the
pragmatic need - in order to make a promise, we have to reference the
conditions that will satisfy that promise.
This suggests that dolphins and other non-tool using species may
have communication systems that are only partially commensurate with
the communication systems of habitual tool-making/tool-using species,
such as hominids. Instead of asking the question do dolphins make
reference? maybe we should be asking whether they make declarations
or promises or ask questions. Then if we can say yes to this, we have
already claimed that they make reference since none of these
communicative acts can be performed without reference. If they cannot
do any of these communicative acts, then how or why would they go
about referencing anything at all? And if this is the case, then
maybe the signature whistles of dolphins are not forms of referencing
but simply forms of broadcasting for the purpose of socializing??
Regards,
Martin Byers
------------------------------
Date: Thu, 5 Nov 1998 23:22:12 -0700 (MST)
Subject: Re: EvolLang: List of Potential Preadaptations
>Sent by: "C. Isa Kocher" <786isa@gto.net.om>
>I think it is very clear, and in fact crucial that LANGUAGE not be
>considered to be strictly speaking, one thing.
>
>The neurological distribution of language all over the brain so to
>speak would it seems to me to demonstrate the usefulness of
>considering how the parts of language developed and evolved from
>various unconnected and interconnected pre language abilities. Just
>as the eye is made of independently evolved parts which came
>together over time, so it is with behavior. Ethologists are able to
>compare behaviors in related species and unrelated species to
>untie the connections that bring together a current behavioral
>complex and make reasonable accounts of their evolution. Language
>is certainly a more difficult case, but the basic principles of our
>various disciplines need not be sacrificed but rather applied in
>detail and carefully and the gradual emergence of evidence will
>bring insight and understanding. IMHO.
This suggests a new list for Matt to start assembling.
A list of the preadaptations that made human language possible.
Possible items for inclusion would be:
1) the exaggerated split between the sequential parsing functions of
the left hemisphere and the spatial recognition functions of the
right hemisphere that occurs in humans versus the slight asymmetry
observed in other primates (a split possibly increased by the need
of wide roaming human hunters to recognize terrain in order to find
their way home, a la Doreen Kimura's theory that it was the right
hemisphere, not the left hemisphere, that drove the split);
2) development of extensive neural circuitry for using our human
hands to manipulate tools (this appears to shift the emphasis back
upon the left hemisphere and its specialized functions);
3) descent of the larynx (perhaps due to an increase in the size of
the voicebox to make louder sounds which relates back to the
increased roaming territory of humans and the need for individuals
to stay connected);
4) increased control over vocalizations for the sake of singing;
5) increased plastic intelligence for mastering new ecological
niches, connected to a longer childhood;
6) and social, anthropological factors such as increased population
density or other changes in the human mode of living.
What other potential preadaptations for language is this list
missing?
Put out a call for contributions, Matt. Let's approach this in the
way that Isa has suggested, recognizing that it is probably not
just one thing that separates us from our non-speaking animal
cousins.
Regards,
John Halloran
[Looks like you just did, John! :) Matt]
------------------------------
Date: Sat, 7 Nov 1998 16:12:17 -0700 (MST)
Subject: Re: EvolLang: yet more whistles (Part 2)
>Sent by: Kelly Jaakkola <jaakkola@biology.ucsc.edu>
>
>> Sent by: Martin Byers <ambyers@sprint.ca>
>> Instead of asking the question do dolphins make reference? maybe
>> we should be asking whether they make declarations or promises or
>> ask questions. Then if we can say yes to this, we have already
>> claimed that they make reference since none of these communicative
>> acts can be performed without reference. If they cannot do any of
>> these communicative acts, then how or why would they go about
>> referencing anything at all? And if this is the case, then maybe
>> the signature whistles of dolphins are not forms of referencing
>> but simply forms of broadcasting for the purpose of socializing??
>
>As a philosophical discussion, this is all well and good. But I've
>got to say that from the perspective of a scientist who's actually
>trying to analyze dolphin communication, you're not giving me much
>to work with. If you (or anyone) can come up with any specific
>suggestions for how in the world you could hope to tell if a dolphin
>was making a promise or asking a question, then I'll be happy to
>look for it. Until then, I'll be **thrilled** if I can just
>demonstrate representational whistles.
Kelly,
Instead of trying to match what dolphins do to the whole of what
humans do with language, can you match what dolphins do to what
prelinguistic humans might have done with singing?
By focusing on what animals do with calls, songs, duets, and
conversations, you would be researching an important part of what
preadapted humans for language, while recognizing that this animal
behavior does not itself involve using or manipulating indexical
symbols.
Human symbolic speech may not have been possible without the
refinement given to our vocal apparatus by the social importance of
song. Merlin Donald summarizes Charles Darwin's suggestions regarding
the importance of song in human evolution in his book, Origins of the
Modern Mind: Three Stages in the Evolution of Culture and Cognition
(Harvard University Press, 1991), at pages 38 - 41.
The ancient Sumerians had several occupations or classes of priests
devoted to singing, lamenting, wailing, and mourning. It is a
cultural tradition that has fallen out of favor, except as seen in
the ululating mourning wail or keening of Arab women. Does anyone
know of comparable behavior by other peoples?
If it has not been done, someone on this list should assign a Ph.D.
student to collect the available evidence and records regarding early
non-linguistic vocalizations in humans as it may have evolved from
our animal beginnings.
Regards,
John Halloran
------------------------------
Date: Fri, 13 Nov 1998 21:50:56 -0700 (MST)
Subject: Re: EvolLang: yet more whistles (Part 2)
>Sent by: "Ronald Kephart" <rkephart@unf.edu>
>
>> Sent by: John A. Halloran
>
>> ...The ancient Sumerians had several occupations or classes of
>> priests devoted to singing, lamenting, wailing, and mourning...
>> [...]
>> If it has not been done, someone on this list should assign a
>> Ph.D. student to collect the available evidence and records
>> regarding early non-linguistic vocalizations in humans as it may
>> have evolved from our animal beginnings.
[snip]
>
>In the case above, we have the implied claim that we can learn
>something from the Sumerians about language evolution. I would like
>to once again point out that the Sumerians were totally modern human
>beings who spoke a totally modern human language. Any non-
>linguistic vocalizations they made, such as laughing, crying,
>screaming, wailing, etc. do not represent our "animal beginnings"
>any more faithfully than the cries, laughs, screams, etc. that we
>ourselves produce.
>
>So, it would be just as fruitful I think to set those PhD candidates
>to studying these things in contemporary peoples. That way, they
>could get video and audio tape (we can't get that from those
>ancient Sumerians). Any such study should include cross-cultural
>and also cross-species components to help us tease out what is
>inherited (i.e., primate), what is derived (i.e., human), and to
>what extent culture shapes these vocalizations.
Ron,
Can you or your wife produce the sound that Arab women make?
Middle Eastern women specialized in wailing because the men were supposed to be stoic fighters. Not sure when Middle Eastern male priests stopped wailing, but the Sumerians lived a long time ago.
Darwin pointed out that gibbons produce modulations that resemble
singing during courtship. It is logical to think that before the
invention of speech, our human ancestors elaborated various
nonlinguistic vocalizations that can be roughly described as singing.
By suggesting that we study the vocalizations of present-day peoples,
you are missing the point that words and language have infected the
singing behavior of present-day peoples due to the omnipresence of
language. The vocalizations of Arab women are not genetically
mediated. They are culturally learned, but not language. These
are the type of vocalizations that may have a long history going back
to a time before language. I don't care about genetically mediated
vocalizations of present day peoples. Humans have been capable of
learned culture for much longer than they have possessed language,
in my opinion. The vocal tract of humans was capable of and probably
was used for culturally learned vocalizations that were not language
prior to the invention of language. If historical records show that
ancient peoples had a stronger tradition of non-language
vocalizations than we are accustomed to in the present, this could
reflect continuity of traditions from a time before language was as
omnipresent as it is today.
Regards,
John Halloran
------------------------------
Date: Mon, 16 Nov 1998 00:45:23 -0700 (MST)
Subject: Re: EvolLang: yet more whistles (Part 2)
>Sent by: mcv@wxs.nl (Miguel Carrasquer Vidal)
>
>>John A. Halloran wrote:
>
>>If historical records show that ancient peoples had a stronger
>>tradition of non-language vocalizations than we are accustomed to
>>in the present
>
>Do they?
>
>My mother tells me that when she was young you could still order the
>services of wailing women (planyidores, I think) at funerals.
>I really have no reason to think there's anything very special about
>Sumerian wailing priests. And, given that the historical records
>are in language (by definition), what might we learn in practice
>about those ancient non-language vocalizations from the texts?
Miguel,
Look in your copy of The Sumerian Language by Marie-Louise Thomsen
in the Catalogue of Verbs under za. What Thomsen says there
indicates a tradition of making vocal sounds that were not
specifically linguistic in nature.
For those who do not have access, I will quote the entry:
"This verb [za] occurs always in compounds with onomatopoetic words
like for instance dum dam...za 'to howl' (see ex. 822), others are:
bu.ud-ba.ad, bul3-bal, dub-dab, du.bu.ul-da.ba.al, gum2-ga.am3,
gun(KUN)-ga.an, hu.um-ha.am, mul-ma.al, pu.ud-pa.ad, pu.ug-pa.ag,
suh3-sah4, zur-za.ar, wu.wa. See [Miguel] Civil, 1966, p. 119: 'All
these forms mean 'to make noise', usually a repeated, monotonous
kind of noise'."
I submit that this is a window on a behavior by our ancestors about
which we in the 20th century are unaware because we have lost the
tradition.
I am not familiar with the name of the wailing women that you know
from your mother. Were they like the Arab women in making sounds
that were non-linguistic in nature?
In Anne Draffkorn Kilmer's article on Musik in the Reallexikon der
Assyriologie, vol. 8, 5/6 (1995) appears on page 469 the following:
"By the time of the canonical Lu2 lists [classes of human
occupations], there are more than 90 terms for musicians among other
types of entertainers. The transvestite entertainers, like kurgarru^
and assinnu, emit "twitters and chirps" (hababu, tsabaru) while
performing; whether these sounds were considered as "music" we do not
know."
The language and civilization of Sumer was mature, modern, and
fully-realized. A.D. Kilmer's article on the sophistication of their
music makes one appreciate this. The reason that it is worthwhile to
examine their civilization for traditions of non-linguistic
vocalizations is because having flourished over four thousand years
ago, they are that much closer to the origins that we are
investigating.
Regards,
John Halloran
------------------------------
Date: Tue, 24 Nov 1998 11:46:02 -0700 (MST)
Subject: Re: EvolLang: yet more whistles (Part nth)
>Sent by: "C. Isa Kocher" <786isa@gto.net.om>
>
>> Sent by: John A. Halloran
>
>> What cultural richness prior to the Neolithic are you thinking of?
>
>We have had over the last 150 years on the ground participant
>observation research among non-farming, small scale societies on
>every continent including such societies as the Ojibwa, the Bushmen,
>the Australians, the Amazon, California mountains, and so forth. The
>complexity and richness of these cultures is a matter of record.
<snip>
>The richness and complexity of these cultures is so great that its
>study has taken up the full energies of a great many scientists and
>others in its understanding and explanation. Many scientists have
>spent their whole scientific lives studying these cultures as
>cultures. These cultures could not have achieved this richness and
>complexity without language.
I can see that communal philosophizing would require language. I am
not so sure about the others.
>It is also significant that virtually none of this culture is
>reflected in the material culture, that which would be found in a
>physical archaeological site. There is no reason whatsoever to
>assume without very good contrary evidence that all humans
>everywhere have always shown this degree of cultural complexity
I agree.
>and in fact it is this cultural complexity and richness is what
>makes humans humans. On the contrary, whatever non-lithic material
>culture as has survived shows every indication that pre-agricultural
>societies and cultures have always been thus. Paintings, carvings,
>art exists many millennia, many thousands of millennia prior to
>Sumer.
Have you ever read a book or an article about the born deaf who do
not acquire language? They are quite capable of paintings,
carvings, and art. They don't need language to make them fully human.
They just need to be raised by loving parents in a cultural
environment.
>You also seem to be saying that this same process occurred all over
>the world even among human groups long in isolation (namely
>Australia) as well as all the other human groups. In any case, the
>human species did not have language, but just language capability
>for the large bulk of its history. If the Sumerians invented it, and
>it diffused to others, how did it diffuse to all humans regardless
>of the degree of social interaction, and why did it not diffuse more
>or less the same. Where have all the utterly unrelated and isolated
>languages come from in such a short time. If languages are all for
>the most part independently invented, then why is language so
>similar in the way it is utterly integrated into the culture totally
>independently of the degree and nature of socio-economic complexity
>and why does language seem to vary independently of socio-economic
>complexity. It just doesn't make sense. There is no rhyme nor reason
>to it.
On the one hand you are pointing out the diversity of the world's
languages and then you are pointing out how similar languages are.
All I can say is that humans are all pretty much alike and when
exposed to the concept of language the separate human groups
innovated systems that explore the different possible ways of
combining sounds and words to represent the world, but they were
under the constraint that their systems had to parallel common human
mental processes.
>Everything we know shows us that all humans are language users
>and all languages are human. The overlap is perfect
Is this supposed to be a proof of something, that only humans have
invented languages?
>and there is no evidence anywhere of any humans who have anything
>other than fully modern language. You can not provide one hint of
>evidence of any human group which did not have normal human language
>anywhere ever.
How would you classify the Australian Arunta/Aranda women who can
conduct a conversation in almost perfect silence with fingers, hands,
arms, and elbows. "One observer lists 454 words rendered by Arunta
gestures." A. Sommerfelt, "Speech and Language," in A History of
Technology, ed. Charles Singer (Oxford: Clarendon Press, 1954), p.
94. Is that language 'fully modern' and 'normal'?
Don't forget that Moonhawk has found evidence that native American
sign language appears to precede vocabulary in unrelated native
American languages.
>In addition, for me, there is something profoundly repulsive about
>the idea that for hundreds of millennia, modern humans went around
>in dumb pantomime and grunts.
The connection that you are asserting seems very fuzzy. What exactly
is the relationship between humans who are anatomically modern and
language? Is there a causal relationship? Did language make humans
anatomically modern two hundred thousand years ago? Did being
anatomically modern cause humans to invent language? Does being
language-capable cause one to invent and use language?
If that is true, were humans anatomically incapable of writing until
the last five thousand years?
Regards,
John Halloran
------------------------------
Date: Tue, 24 Nov 1998 11:46:02 -0700 (MST)
Subject: Re: EvolLang: yet more whistles (Part nth)
>Sent by: "Ronald Kephart" <rkephart@unf.edu>
>I thought the physical evidence was pretty solid: the "modern human"
>vocal tract with lowered larynx, and the "modern human" brain.
>Aren't both of these attested in the fossil record, and don't they
>in fact go back between 100 and 200 thousand years (long before any
>identifiable "proto-Sumerians")? The modern human vocal tract, in
>particular, I think is striking evidence for language, precisely
>because it makes possible a range of speech sounds unavailable to
>the "standard" primate vocal tract but at the cost of increased
>susceptibility to choking, difficulty in breathing, crowded wisdom
>teeth, and so on.
We have little evidence as to the age of the modern human vocal
tract, but I would not be surprised if this adaptation started as
soon as humans started wide-ranging foraging and hunting because I
think that selection pressures were greater on humans who became
separated from each other because their voiceboxes were too small
to make loud vocalizations. The descent of the larynx makes a
larger reverberating chamber. That has a lot more to do with
survival than does this business about different vowel sounds.
Regards,
John Halloran
------------------------------
Date: Tue, 24 Nov 1998 11:46:02 -0700 (MST)
Subject: Re: EvolLang: yet more whistles (Part nth)
>Sent by: "C. Isa Kocher" <786isa@gto.net.om>
>
>> >>"John A. Halloran" wrote:
>
>> >> "One of the most baffling and precipitous phyletic trends in
>> >> Homo Sapiens has been the appreciable shortening and moderate
>> >> widening of the head." Edward E. Hunt, Jr. "The Continuing
>> >> Evolution of Modern Man," Cold Spring Harbor Symposia on
>> >> Quantitative Biology, 24 (1959): 250.
>>
>> >> Couldn't this recent, rapid change qualify as an instance of
>> >> "adaptations in the human body to language use"?
>
>Emphatically, no it cannot. Humans all have language, and all humans
>have language, and all human languages are naturally spoken by
>humans. Homo sapiens speaks languages everywhere, so it is not a
>relevant variable. There are more than enough changes that occur in
>human societies when they begin to practice agriculture that any
>number of more reasonable and more importantly more testable
>hypotheses. It doesn't even merit my response. Language in humans
>is a species specific behavior and there is no evidence whatsoever
>of the contrary.
Are you saying that language defines the species? Are you saying
that the non-linguistic born deaf are not human?
>> >Show me how and why- what is the connection and what is the
>> >evidence.
>>
>> It has been explained to me that it is racist to research
>> cognitive or linguistic differences between human populations that
>> might be connected with varying amounts of cortex in the different
>> lobes of the brain. So this investigation will have to wait for a
>> more secure, confident age of science.
>
>So, in short, you have no evidence of any kind, and do not want to
>consider that that line of research has been shown repeatedly by
>scientist in any number of disciplines to be valueless and
>unscientific because it goes in the face of all known scientifically
>accepted and verified fact.
Do you have any references for these repeated studies in 'any number'
of disciplines?
The main question about the shift in head form from dolichocephaly to
brachycephaly has been whether the brain lobes just move around to
adapt within a skull that changes form for its own reasons or if the
skull adapts its form in response to the need to cover changing brain
lobes. Despite your assertion, scientists have avoided this subject
for the reasons that I mentioned. It was very difficult to find any
study of underlying brain differences between populations. I did
manage to locate one study which answers the main question, however.
C.U. Ariens Kappers, "On Some Correlations Between Skull and Brain,"
Philosophical Transactions of the Royal Society of London, ser. B,
vol. 221, pp. 391-429, reports on studies of the brains of
dolichocephalic and brachycephalic individuals from Europe, Africa,
and Australia. What he finds is that the temporal lobe is much
shallower in dolichocephalic skulled individuals and the parietal
lobe has less depth in brachycephalic skulled individuals. The
temporal lobe is important for acoustic tasks and verbal
comprehension. The parietal lobe is important for spatial mapping.
Why did it take so long for humans to map the world to a communally
shared set of vocal sounds? Conservatism and inertia. Why would a
mighty elephant hunter want to learn a set of sounds when it would
just undercut the power that he has from elephant hunting? The
already franchised and powerful would not have encouraged a new
communal system such as language. Also, when one studies ancient
civilizations, one sees that innovations are never virgin born, but
are always built upon and are adaptations of some prior system.
Change is a gradual process. I have already suggested that there
was a conceptual leap to mapping the world into categories of objects
which was bridged in the Near East by the token system. Do you
think that the concept of language was self-evident to our
non-speaking ancestors?
Regards,
John Halloran
------------------------------
Date: Sat, 28 Nov 1998 01:06:11 -0700 (MST)
Subject: Re: EvolLang: yet more whistles (Part nth)
>Sent by: "C. Isa Kocher" <786isa@gto.net.om>
>
>>"John A. Halloran" wrote:
>
>>Are you saying that language defines the species? Are you saying
>>that the non-linguistic born deaf are not human?
>
>First of all, born-deaf are not non-linguistic.
Okay, so you have studied nonscientific peoples living in small
groups, but you are not familiar with studies of the linguistically
deficient born-deaf. The following paragraph may have appeared on the
newsgroup before you joined it:
Quoting from an excellent book on the subject, "The evidence for
conceptual thinking in the linguistically deficient deaf has been
presented and leads to the direct conclusion that thinking develops
through living contact with the environment regardless of the
presence or absence of a ready-made linguistic symbol system."
Hans G. Furth, Thinking Without Language: Psychological Implications
of Deafness (1966). pp. 198-99. He says that any theory that would
explain thinking in terms of verbal or other symbolic units is
flawed and rests on false assumptions.
Here is a new quote from a later article:
"The remarkable fact is that children without linguistic skills do
seem to make use of whatever symbolic instruments are required in
the service of thinking. This illustrates strikingly that symbols
are basically produced by the thinking person and used according to
his available thinking skills. No radical distinction between verbal
and other symbols is indicated. That is, just as the existence and
the use of private symbols by deaf children argues for an internal
source, similarly the use of verbal symbols by hearing children must
be subordinate to the thinking of the children, even though the
existence of the symbols derives from an external source. In this
perspective, no symbols in themselves, verbal or otherwise, can
explain thinking." Hans Furth and James Youniss, "Congenital
Deafness and the Development of Thinking," in Foundations of Language
Development, vol. 2, Eric H. Lenneberg and Elizabeth Lenneberg (eds.)
(New York: Academic Press, 1975), p. 171.
"The spontaneous production of symbolic instruments is indeed the
most striking aspect of the intellectual development of deaf
children; it powerfully illustrates the subordinate role of all
symbols in the developing structures of thinking." ibid, p. 175.
If you are really interested in whether Homo sapiens is capable of
being Homo sapiens without language, you owe it to yourself to
study the research on linguistically deficient deaf children.
>> Do you think that the concept of language was self-evident to our
>> non-speaking ancestors?
>
>Not self-evident: species specific.
I am starting to suspect that you have a religious explanation for
the origin of Homo sapiens as a language using species.
Does that explain why you feel as strongly about this as you do?
To me, the barrier to achieving symbolic language is achieving the
concept of symbolic language. You blow right past that as if it is
not important. The concept of language is like the concept of fire,
the concept of the wheel, the concept of writing, or the concept of
relativity. It is not self-evident to those who have never seen it
in action. But once a group does see language in action, they
understand the concept and they want it. For a time in human history,
there were groups who had the concept and there were groups that
didn't have the concept. If you don't see that, then you must be
grinding some private ax that you have not yet shared with the rest
of us.
Regards,
John Halloran
------------------------------
Date: Sat, 28 Nov 1998 17:29:53 -0700 (MST)
Subject: Re: EvolLang: yet more whistles (Part nth)
>Sent by: "C. Isa Kocher" <786isa@gto.net.om>
>
>> Sent by: John A. Halloran
>>
>> >Sent by: "Ronald Kephart" <rkephart@unf.edu>
>>
>> >I thought the physical evidence was pretty solid: the "modern
>> >human" vocal tract with lowered larynx, and the "modern human"
>> >brain. Aren't both of these attested in the fossil record, and
>> >don't they in fact go back between 100 and 200 thousand years
>> >(long before any identifiable "proto-Sumerians")? The modern
>> >human vocal tract, in particular, I think is striking evidence
>> >for language, precisely because it makes possible a range of
>> >speech sounds unavailable to the "standard" primate vocal tract
>> >but at the cost of increased susceptibility to choking,
>> >difficulty in breathing, crowded wisdom teeth, and so on.
>>
>> We have little evidence as to the age of the modern human vocal
>> tract, but I would not be surprised if this adaptation started as
>> soon as humans started wide-ranging foraging and hunting because
>> I think that selection pressures were greater on humans who became
>> separated from each other because their voiceboxes were too small
>> to make loud vocalizations. The descent of the larynx makes a
>> larger reverberating chamber. That has a lot more to do with
>> survival than does this business about different vowel sounds.
>
>Hunting requires silence and stealth. Gathering in groups is
>involved with small talk and gossip. I beg you to find out SOMETHING
>about pre-agricultural societies and how they live. Or just imagine
>what it was like when you were a kid and went berry picking. I mean
>you KNOW where everyone is- and the idea of shouting back and forth
>to each other to keep track of everyone is ridiculous.
Most monkeys spend their lives in an area of two or three square
miles. The gorilla and chimpanzee have a range of roughly 15 square
miles. S.L. Washburn has pointed out that one of the remarkable
characteristics of human beings is that even the most primitive of
men operate over hundreds of square miles rather than these small
areas [ref. to The evolution of behavior, in The Uniqueness of Man,
J. Rolansky, ed., 1969].
When primates expanded the territory over which they range, they
grew into a life style similar to that of wolves (who in gradually
modified form became their companions). The primate ability for
long-range vocalizations increased because it had the same utility
for them that this ability does for wolves. Wolves range over great
distances as they follow game. They also vocalize to communicate
during group hunting. When primates evolved to do the same, it
helped their survival to evolve a similar ability to vocalize over
greater distances than could their ancestors.
The gossipy, berry picking humans that you are describing do not
sound like the nomadic hunting groups who travelled all the way
from Africa to America. The humans that you are describing sound
very sedentary. Humans did not start living in permanent, modified
dwellings until the end of the Paleolithic. It is a mistake to
project the behavior of present-day humans so far back into the
past. My claim is that humans have had language for at least three
hundred generations now. It is misleading to call present-day
nonagricultural peoples 'pre-agricultural', as if they are the same
as people who lived during the Upper Paleolithic and Mesolithic
time periods. Almost all present-day humans, including
nonagricultural peoples, are the product of at least three hundred
generations of ancestors who used language. Just as Ilse
Schwidetsky has shown that brachycephalization is the evolutionary
product of selection over a short time period, so other
evolutionary adaptations to favor language could have occurred as
a result of selection during this time period.
Regards,
John Halloran
------------------------------
Date: Sat, 28 Nov 1998 23:26:43 -0700 (MST)
Subject: Re: EvolLang: yet more whistles (Part nth)
>Sent by: Larry Trask <larryt@cogs.susx.ac.uk>
>
>>On Tue, 24 Nov 1998, John A. Halloran wrote:
>
>> Don't forget that Moonhawk has found evidence that native American
>> sign language appears to precede vocabulary in unrelated native
>> American languages.
>
>Sorry; I'm afraid I have no idea what this is supposed to mean.
That is from January 17, 1998, from Moonhawk's post entitled
Metaphor.
The first part was John Halloran. The second part was Moonhawk's
response.
>> It must be frustrating to work with American languages in which
>> most words cannot be analyzed, and then have some guy say that in
>> his language the words can be analyzed down to a much lower level.
>This is the position I feel in when I talk about how words in
>entirely different languages in Native America are made up from the
>same exact morphemes said differently in different language, when I
>find a common factor to be Native American Sign Language, which
>contains the same exact morphemes in sign. (E.g., "God" from "big"
>+ "medicine" + "upward".)
Moonhawk asked other linguists to help him understand how this could
be, but no one volunteered any help.
The logical explanation is that a preexisting common sign language
influenced the inventors of the separate spoken languages.
Isn't it the case in the Americas that the time depths of the
reconstructed protolanguages gets shallower as one goes north from
South America? Are there social, demographic explanations for that,
or are we looking at the spread of the concept of language?
Regards,
John Halloran
------------------------------
Date: Sun, 29 Nov 1998 15:07:43 -0700 (MST)
Subject: Re: EvolLang: Jaw Evolution
>Sent by: "Ronald Kephart" <rkephart@unf.edu>
>
>> John H. writes:
>
>> I don't see any other list members proposing physical evidence to
>> support the existence of language since the time of physically
>> modern humans. Maybe the linguists assume that the physical
>> anthropologists have the evidence while the physical
>> anthropologists assume that the linguists have it.
>
>I thought the physical evidence was pretty solid: the "modern human"
>vocal tract with lowered larynx, and the "modern human" brain.
>Aren't both of these attested in the fossil record, and don't they
>in fact go back between 100 and 200 thousand years (long before any
>identifiable "proto-Sumerians")? The modern human vocal tract, in
>particular, I think is striking evidence for language, precisely
>because it makes possible a range of speech sounds unavailable to
>the "standard" primate vocal tract but at the cost of increased
>susceptibility to choking, difficulty in breathing, crowded wisdom
>teeth, and so on.
Ron,
Your statement about crowded wisdom teeth caught my attention. This
change is caused not by language but by humans using tools to prepare
food and cook meats instead of needing to expend the energy to
maintain massive teeth for chewing plant material.
The following quotes are from J.J. Wymer, The Palaeolithic Age
(1982), pp. 138-139.
"Three species of Homo can be distinguished: Homo erectus, Homo
sapiens neanderthalensis and Homo sapiens sapiens. There are no
objective definitions for any of them and, even if there were, the
fragmentary nature of the remains as found would almost certainly
prevent any general application of such definitions. All that can
really be said is that certain characteristics are more likely to
belong to one species than another; for instance, a mandible with a
chin eminence is sapiens sapiens and a heavy supra-orbital ridge at
the base of the frontal bone coupled with a large brain capacity is
sapiens neanderthalensis. In many cases it is impossible to classify
the remains as they do not fit neatly into any of the sub-specific
categories, but seem intermediate, or slightly more like one
than the other."
"The gradual trend, at least in Europe, Africa and the Near East,
from about 400,000 years ago, was for the human skull to become
larger, with the occipital becoming more rounded and brow ridges less
prominent. This may have been through a reduction in the size of
the jaw; large canines and massive molars were no longer an advantage
for tearing up foodstuffs as stone tools did the job more
efficiently. As jaws decreased in size, so the necessity for thick
muscle attachments on the skull may have diminished and the skull
changed shape accordingly. Whatever the reason, this happened,
and is a non-erectus characteristic. It may be seen as a Neanderthal
characteristic or, if accentuated, a sapiens characteristic.
Similarly, the reduction of jaw size leading to the relatively
gracile mandible of modern man, another sapiens characteristic or
trait, can be seen by the end of the Last Interglacial at least [Last
Interglacial extended from 130,000 to 75,000 years BP]. By the Late
Pleistocene it seems that individuals were likely to possess either
sapiens, neanderthal or even erectus traits in varying proportions."
Despite the prominent position that linguists might like to give it,
all evolution does not revolve around language. Jaw size relates
to what the population is eating and how efficient is its tool kit.
Regards,
John Halloran
------------------------------
Date: Sun, 29 Nov 1998 22:45:57 -0700 (MST)
Subject: Re: EvolLang: yet more whistles (Part nth)
>Sent by: "Peter T. Daniels" <grammatim@worldnet.att.net>
>
>> John A. Halloran wrote:
>>>First of all, born-deaf are not non-linguistic.
>> Quoting from an excellent book on the subject, "The evidence for
>> conceptual thinking in the linguistically deficient deaf has been
>> presented and leads to the direct conclusion that thinking
>> develops through living contact with the environment regardless
>> of the presence or absence of a ready-made linguistic symbol
>> system." Hans G. Furth, Thinking Without Language: Psychological
>> Implications of Deafness (1966). pp. 198-99. He says that any
>> theory that would explain thinking in terms of verbal or other
>> symbolic units is flawed and rests on false assumptions.
>
>Look at the date. Furth (apparently a psychologist, not a linguist
>anyway) perhaps was unaware that signing deaf people are not
>"linguistically deficient" -- general awareness even among linguists
>that ASL (in particular) is a language was only to blossom in the
>next few years.
No, the researcher tested deaf children whose circumstances had not
exposed them to sign language. Find and read his book or his
articles before you leap to conclusions about his work.
>> external source. In this perspective, no symbols in themselves,
>> verbal or otherwise, can explain thinking." Hans Furth and James
>> Youniss, "Congenital Deafness and the Development of Thinking," in
>> Foundations of Language Development, vol. 2, Eric H. Lenneberg and
>> Elizabeth Lenneberg (eds.) (New York: Academic Press, 1975),
>> p. 171.
>
>Same author. Old dogs/new tricks? I believe Lenneberg had already
>done away with himself at that date, or perhaps he wouldn't have
>let the passage stand in a book his name was on.
You know the linguists of the world, but are not familiar with these
studies because they were not carried out by linguists. In the
modern world it is probably becoming the case that deaf children
who do not learn sign language are rare, so these scientific studies
from the 1960's become more valuable.
I do recall that back in January, 1998 Sherman Wilcox posted a story
about a Korean girl that he knew who because of being shuttled among
foster homes until the age of 4 was late to learn language, yet
once she did learn language had complete recollection of how she had
reacted as a thinking person before the time when she learned
language.
Regards,
John Halloran
------------------------------
Date: Fri, 4 Dec 1998 14:07:14 -0700 (MST)
Subject: Re: EvolLang: Putting it in perspective
Has anything become more clear after the recent flurry of posts and
responses?
One of the key questions is if there was any delay in time between
when humans became language-capable and when humans invented and
began using language.
In a closely reasoned article, Merlin Donald says that humans appear
to have been language capable by 45,000 years ago.
http://baserv.uci.kun.nl/~los/Articles/donald.html
This is the same Merlin Donald whose 1991 book, Origins of the Modern
Mind: Three Stages in the Evolution of Culture and Cognition, was so
good.
This 1998 article emphasizes evolution of the human capacity for
mimesis, which the dictionary defines as 'imitation, mimicry'. So
that needs to be added to the list of developments that preadapted
humans for language.
It is clear that some people have difficulty believing that if
humans were capable of language 45,000 years ago, they would have
waited until 10,000 years ago to actually invent and use language.
The only real reason why I believe that humans may have waited that
long is because of my analysis of how the Sumerian vocabulary evolved
from scratch. That is what motivated me to look at the whole field of
language from a slightly different viewpoint than other researchers.
I was trying to falsify my theory, trying to find evidence that would
disprove an early Neolithic origin for language. But no matter on
which continent I looked or in what fields of brain evolution, the
available evidence did not falsify such a recent origin.
If in fact language was delayed until 10,000 years ago, the barrier
to achieving language was achieving the concept of language, a
concept which is not self-evident to a nonspeaking world. And the
evidence found by Hans Furth on the mental capability of deaf humans
who have not been exposed to sign language makes it quite conceivable
that such humans were capable of all of the art and accomplishments
of the Paleolithic.
I am continuing to expand my lexicon of Sumerian and will think
about how I can better communicate what I see as being its early
features. That is better than engaging in rhetorical wars.
I thank John McL. for responding with his knowledge of the history
of Plains Sign Language and the dates of the reconstructed American
protolanguages. I had known about Athapaskan at 3500 yrs BP, but I
guess that linguists now accept Na-Dene with a much greater time
depth. I have two questions about that:
Do the archaeologists have an IE-like explanation for the expansion
of Athapaskan over such a wide area? Do they have evidence for a
particular cultural innovation that would cause that particular
language to replace other existing languages?
And, a question that I have never seen any linguist anywhere
address, how can a linguist distinguish between languages that share
features because they are related by common descent and languages
that share features because a nonspeaking population borrowed
elements from a population that already had language? In the latter
case, even the elements at the core of the new language could be
borrowed. Linguists such as Malcolm Ross and Semiticists such as
Robert R. Ratcliffe who stress contact models of linguistic change
as opposed to descent models are getting closer to addressing this
question.
As a postscript, list members should know that there are about 40
high quality articles about language origin at the Language Origins
Society web site that is hosting Merlin Donald's article referred
to above.
As a second postscript, Science News, Nov. 28, is advertising a
book, Children of Prometheus: The Accelerating Pace of Human
Evolution, by Christopher Wills, which provides evidence for how
quickly our genetic composition reacts to environmental stimuli,
especially as it relates to mental processes. Of course, one just
has to look at the differences in brain mass in different parts of
the brain between domesticated animals and their wild congeners to
see how quickly the brain can change over a short period of time.
To answer Pat Ryan's query, I found the reference to
brachycephalization being due to selection in an article by N.
Kondova and S. Cholakov, "Brachycephalization in Bulgaria", Homo
45/1 (1994), pp. 63-73. They reference a 1989 study by Fw Roesing
and Ilse Schwidetzky, the great master of statistical European
physical anthropology, on "Causative factors of brachycephalization
process". Their summary says, "It is selection that is considered
to be the main factor in the process of brachycephalization in
Central Europe during the period between the 8th and the 18th
centuries."
Regards,
John Halloran
------------------------------
Date: Sun, 6 Dec 1998 13:11:39 -0700 (MST)
Subject: Re: EvolLang: material culture, language and symbolism
>Sent by: Martin Byers <ambyers@sprint.ca>
Martin,
What about looking at when groups emerged that were culturally
distinct from other groups? Today, language is the most important
factor in determining culture and group membership. In the
archaeological record of Mesopotamia one clearly sees regions of
different cultures marked by different styles of pottery, such as
the Hassuna pottery culture and the Samarra pottery culture. At
what point back in time can archaeologists identify culture
that distinguished neighboring groups from one another?
Regards,
John Halloran
Date: Tue, 08 Dec 1998 18:30:29 -0500
Subject: Re: EvolLang: material culture, language and symbolism
I have made the argument - Current Anthropology 94 - that we can
discern symboling in the lithics with the emergence of "isochrestic
styles," a term arising from a notion first developed in the early
80's by James Sackett. He spoke of functionally equivalent but
recognizably different material assemblages as displaying different
isochrestic (functionally equivalent) variation. He uses this notion
to speak of the different Mousterian facies as marking ethnic
distinctions. But he (and Philip Chase later) also argued that the
emergence of different congeries of isochrestic variatios in
different or even the same regions, does not mark symboling styles
since, for him, the continuity of a given tradition that we mark as
a style could be simly the consequence of habit. Philip Chase
makes the same point more emphatically.
I have argued that, to the contrary, the emergence of locally
stabilized isochrestic variation, so that it is recognizably a style
distinct from neighbouring assemblages, marks convention.
Stabilization cannot be sustained by habit unless this habit is
addressing something more than utilitarian function. I argue that
these form Style-2 assemblages, utiitarian assemblages that are
subsumed under cultural rules. But I also argue that these styles
do not necessarily mark ethnic groups. They can do this, but
usually as a by-product of a more basic meaning, namely, the
symbolic constitution of mateerial practices into social practices.
What this means is that we cannot identify an assemblage as
Style-2 in a vacuum because, as utilitarian assemblages, their
physical properties are exhaustively accounted for in practical
terms. Only if we can do a regional comparison and note that despite
the utilitarian properties and the same basic environmental
variables, localy stabilized isochrestic variants emerge can it be
claimed that, in fact, these are Style-2 assemblages. I am
inclined to seeing early aurignacian, chatelperronian, uluzzian,
and other pre- or early Upper Palaeolithic assemblages as the
range of isochrestic styles that demonstrate Style-2. As things now
stand in terms of dating this would put the appearance of symbolic
style in this region at the circa 40 kya range.
I also argue that this type of conventional style is grounded on
the same neurological capacities as symbolic language and,
therefore, it also marks the emergence in this region of fully
symbolic language - language as we know and use it.
Regards,
Martin Byers
------------------------------
Copyright © 2026 John Halloran, Los Angeles, California
Last modified on May 19, 2026.
https://www.sumerian.org/evolution_of_language_as_of_1998.htm