Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

New life for Sapir-Whorf? (and other language links)

0 views
Skip to first unread message

Prentiss Riddle

unread,
Apr 18, 2002, 7:05:34 PM4/18/02
to
Scientific American reports on recent research showing some support
for the Sapir-Whorf Hypothesis, or the notion that the language
we speak influences the way we think. I'm sure I'm just having
a layperson's typical reaction to the picayune sort of refutable,
reproducible experimentation upon which scientific progress is built,
but it seems to me that the experiments described are pretty far from
the heart of Sapir-Whorf.

I'm struck by the mountains of anecdotal evidence for some form of
a Whorfian effect, from folkloric attributions of language character
("Spanish is a loving tongue", etc.) to bilingual writers who argue
that their choice of language affects their work. Particularly strong,
it seems to me, are sociolinguistic arguments about nuances of meaning
within a language: as power words like "gay", "feminist", "negro /
black / Afro- / African-American" etc. gain or lose pejorative or
approbative connotations, surely thought is affected (as any good
pollster can demonstrate). I'm not sure why examples like these
don't crop up more often in discussions of Sapir-Whorf; perhaps it's
harder to design experiments around them, or perhaps linguists feel
compelled to search for an effect in the grammar of a language rather
than in its slippery social context.

In other intriguing linguistic links: A beautifully done exploration of
the Evolution of Writing, showing how media and implements changed the
Roman alphabet and created the font styles we know today. Speculation
about "Space English", or the linguistic drift likely to happen on a
hypothetical multi-generation interstellar flight. A dictionary of
prison slang. A review of David Crystal's Language and the Internet
(you'll be happy to know that we aren't necessarily going to hell
because of smilies and l33tsp33k). Two reports on the necessity and
the discontents of international English from a European perspective
("What passes as English in the European institutions bears about as
much resemblance to the language of the Anglo-Saxon world as a soufflé
to a steak-and-kidney pie"). The Fight the Fog campaign to encourage
people in the EU to write more clearly, complete with a song (anybody
have any idea what tune that should be sung to?), an obligatory
list of badly translated signs, and a self-defeating disclaimer.
I'm sure it's all considered very hip in Brussels these days.
(Many of these links via Enigmatic Mermaid.)

Links and discussion at:

http://www.io.com/~riddle/language/?item=20020411

-- Prentiss Riddle ("aprendiz de todo, maestro de nada")
-- rid...@io.com / http://www.io.com/~riddle/

Peter T. Daniels

unread,
Apr 18, 2002, 10:21:40 PM4/18/02
to
Prentiss Riddle wrote:

> In other intriguing linguistic links: A beautifully done exploration of
> the Evolution of Writing, showing how media and implements changed the
> Roman alphabet and created the font styles we know today.

What is "intriguing linguistic links"?

That's certainly not what a scholar of writing systems would consider
"evolution"; the roman alphabet has essentially not changed since they
got it from the Etruscans 2600+ years ago -- the shapes have been
regularized a bit, and sometimes letters have been added, but that's
just esthetics.
--
Peter T. Daniels gram...@att.net

Peter T. Daniels

unread,
Apr 18, 2002, 10:27:46 PM4/18/02
to

Turns out to be a link to something called "textism.com," which is
bizarre at best, which gives a capsule history of roman writing
illustrated not with genuine inscriptions/mss. but with modern
calligraphic imitations (including forms for the letters that didn't
exist when the styles were in use) that look far more like modern
writing than like the authentic stuff.

Prentiss Riddle

unread,
Apr 18, 2002, 11:50:18 PM4/18/02
to
Peter T. Daniels <gram...@att.net> wrote:
|
| Prentiss Riddle wrote:
|
| > In other intriguing linguistic links: A beautifully done exploration of
| > the Evolution of Writing, showing how media and implements changed the
| > Roman alphabet and created the font styles we know today.
|
| What is "intriguing linguistic links"?

Sorry, I guess it was bad form of me to post the miscellaneous items
from the end of my blog entry to Usenet. I'm still trying to figure
out a reasonable way to cross-pollinate between the (hypertext) weblog
world and (plaintext) Usenet. I did say: "Links and discussion at
http://www.io.com/~riddle/language/?item=20020411 ".

| That's certainly not what a scholar of writing systems would consider
| "evolution"; the roman alphabet has essentially not changed since they
| got it from the Etruscans 2600+ years ago -- the shapes have been
| regularized a bit, and sometimes letters have been added, but that's

| just esthetics. ...


|
| Turns out to be a link to something called "textism.com," which is
| bizarre at best, which gives a capsule history of roman writing
| illustrated not with genuine inscriptions/mss. but with modern
| calligraphic imitations (including forms for the letters that didn't
| exist when the styles were in use) that look far more like modern
| writing than like the authentic stuff.

True, the "Evolution" piece at Textism (URL http://textism.com/writing/ )
is from the point of view of designers, not scholars. Is it
factually off or just a bit too simplified for the sci.lang audience?
I'm neither a scholar nor a linguist myself, just an interested
layperson, so I'm easily fooled.

Mark Rosenfelder

unread,
Apr 19, 2002, 1:00:49 AM4/19/02
to
In article <ohIv8.101935$K5.84...@bin5.nnrp.aus1.giganews.com>,

Prentiss Riddle <rid...@eris.io.com> wrote:
>I'm struck by the mountains of anecdotal evidence for some form of
>a Whorfian effect, from folkloric attributions of language character
>("Spanish is a loving tongue", etc.) to bilingual writers who argue
>that their choice of language affects their work. Particularly strong,
>it seems to me, are sociolinguistic arguments about nuances of meaning
>within a language: as power words like "gay", "feminist", "negro /
>black / Afro- / African-American" etc. gain or lose pejorative or
>approbative connotations, surely thought is affected (as any good
>pollster can demonstrate). I'm not sure why examples like these
>don't crop up more often in discussions of Sapir-Whorf; perhaps it's
>harder to design experiments around them, or perhaps linguists feel
>compelled to search for an effect in the grammar of a language rather
>than in its slippery social context.

One linguist who studies this stuff (and agrees with Whorf) is Suzette
Haden Elgin; see her very readable _The Language Imperative_.

Unfortunately, I don't find all this very convincing, largely because
advocates don't bother to check whether it's that language influences thought,
or that thought influences language.

I'd say there's just as much anecdotal evidence *against* the Whorfian
view. Terms for mental retardation are an example: just as quickly as
people invent new supposedly non-perjorative terms, kids turn them into
pejoratives. (Indeed, 'retarded' was a favorite epithet when I was
growing up.) Isn't it possible that, rather than the words making people
prejudiced, people's prejudices color the words they use?

As for bilinguals finding that their choice of language affects their
work, see linguist Anna Wierzbicka's work on cultural differences.
But note that she finds it's a matter of culture, not language per se--
as is shown by the fact that you can find different stylistic or cultural
conventions within the same language.

Peter T. Daniels

unread,
Apr 19, 2002, 7:22:19 AM4/19/02
to

Basically, facts are _absent_ -- it's just a collection of pretty
pictures.

Do you know *The World's Writing Systems* (Oxford UP, 1996), coedited by
yours truly? Somewhere out on the web there's a site that has copied a
lot of the charts (it's not at all clear that they either requested or
received permission to do so) and some of the text, though I've no idea
where to find it.

Prentiss Riddle

unread,
Apr 19, 2002, 11:02:54 AM4/19/02
to
Mark Rosenfelder <mark...@enteract.com> wrote:

: Unfortunately, I don't find all this very convincing, largely because


: advocates don't bother to check whether it's that language influences
: thought, or that thought influences language.

Certainly if they're not differentiating between cause and effect, that's
a problem. But why not consider the possibility that both are true?

: I'd say there's just as much anecdotal evidence *against* the Whorfian


: view. Terms for mental retardation are an example: just as quickly as
: people invent new supposedly non-perjorative terms, kids turn them into
: pejoratives. (Indeed, 'retarded' was a favorite epithet when I was
: growing up.) Isn't it possible that, rather than the words making people
: prejudiced, people's prejudices color the words they use?

Again, sounds like a feedback loop to me!

Thanks for the tips on Suzette Haden Elgin and Anna Wierzbicka.

By the way, the Scientific American article (for people who didn't
get to it via my webpage) is at:

http://www.scientificamerican.com/explorations/2002/032502language/

Wolf Kirchmeir

unread,
Apr 19, 2002, 10:38:58 AM4/19/02
to
On Fri, 19 Apr 2002 03:50:18 GMT, Prentiss Riddle wrote:

>Sorry, I guess it was bad form of me to post the miscellaneous items
>from the end of my blog entry to Usenet. I'm still trying to figure
>out a reasonable way to cross-pollinate between the (hypertext) weblog
>world and (plaintext) Usenet. I did say: "Links and discussion at
>http://www.io.com/~riddle/language/?item=20020411 ".

Why use a browser for Usenet in the first place??? Browsers are simply not as
useful as advertised for this purpose.

But if you must use a browser, set it to produce and post in text and in
HTML. The text will then include the URLs in the right places. Any current
newsreader will start the browser if requested to do so. EG, with mine I
highlight and LMB click, and get a small menu which lets me "Explore URL." A
few moments later, the browser is connecting to the referenced page.


Best Wishes,

Wolf Kirchmeir
Blind River, Ontario

..................................................................
You can observe a lot by watching
(Yogi Berra, Phil. Em.)
..................................................................


Greg Lee

unread,
Apr 19, 2002, 5:01:34 PM4/19/02
to
Prentiss Riddle <rid...@eris.io.com> wrote:
> Mark Rosenfelder <mark...@enteract.com> wrote:

> : Unfortunately, I don't find all this very convincing, largely because
> : advocates don't bother to check whether it's that language influences
> : thought, or that thought influences language.

> Certainly if they're not differentiating between cause and effect, that's
> a problem. But why not consider the possibility that both are true?

It's possible that both are true. Now, where does that get us? I've
never seen it argued that it is impossible that language influences
thought. The issue is whether it's true, not whether it's possible.

...
--
Greg Lee <gr...@ling.lll.hawaii.edu>

Wolf Kirchmeir

unread,
Apr 20, 2002, 11:06:09 AM4/20/02
to
On 19 Apr 2002 21:01:34 GMT, Greg Lee wrote:

> I've
>never seen it argued that it is impossible that language influences
>thought. The issue is whether it's true, not whether it's possible.


There's no question IMO that language influences attitudes and feelings.
That's why people on one side of a debate will deliberately coin terms that
make their position seem good and the opposing position seem bad. EG, pro-
and anti-choice, vs pro- and anti-life. --- Therefore, to the extent that
feelings are implicated in thought, language will influence thought.
(Assuming we know what we mean by thought -- sci.cognitive for a thread that
isn't getting anywhere on this issue.)

IMO it should be possible to devise experiments that illustrate how language
determines thought. An accidental real life one occurred the other day --
well, I _think_ it's relevant to this issue. I had just pointed out that
explicability is not the same as predictability. Just because you can explain
how something happened doesn't mean that you can predict when (or if) it will
happen again. "That doesn't make sense!" was the response. Clearly, that
person's _definition_ of the two terms determined what he thought. So to that
extent, language does determine thought.

Also, the history of philosophy may be relevant here. Most philosophy IMO
consists of attempts to redefine terms so that "correct" or at least "better"
thinking is possible. And understanding philosophers IMO consists chiefly of
being able to redefine terms and so becoming able to think the thoughts of
that philosopher. If you really want to.

Learning mathematics consists of learning terminology and how to think with
it.

Sapir-Whorf however holds that each language has built in biases in
definition and therefore has built in biases in thinking. This should be a
testable hypothesis, eg, can speakers of one language learn certain kinds of
mathematics more easily than speakers of another language? Or is that
question too vague?

Anyhow, I think Wittgenstein actually said it better:

"The limits of my language are the limits of my world."

Mark Rosenfelder

unread,
Apr 20, 2002, 3:01:00 PM4/20/02
to
In article <jbysxveflzcngvpbp...@news1.sympatico.ca>,

Wolf Kirchmeir <wwol...@sympatico.ca> wrote:
>There's no question IMO that language influences attitudes and feelings.
>That's why people on one side of a debate will deliberately coin terms that
>make their position seem good and the opposing position seem bad. EG, pro-
>and anti-choice, vs pro- and anti-life.

Er, how does that suggest that language influences attitudes, rather than
the reverse? It looks to me like people with anti-abortion convictions
choose language that flatters their position ("pro-life"), while people
who want to preserve access to abortion do the same ("pro-choice").
In other words, attitudes influence language.


Brian M. Scott

unread,
Apr 20, 2002, 3:46:54 PM4/20/02
to
On 20 Apr 2002 19:01:00 GMT, mark...@enteract.com (Mark Rosenfelder)
wrote:

I suspect that Wolf is assuming that the names are chosen at least
partly for their supposed ability to influence others' attitudes.

Brian

Wolf Kirchmeir

unread,
Apr 20, 2002, 6:44:52 PM4/20/02
to
On Sat, 20 Apr 2002 19:46:54 GMT, Brian M. Scott wrote:

>I suspect that Wolf is assuming that the names are chosen at least
>partly for their supposed ability to influence others' attitudes.
>
>Brian

Well, of course.

And any poet knows you can influence attitudes with words. So do advertisers.

I'm not talking about explicit persuasion here. Consider what would happen if
a "pro-choice" person starts thinking of abortion as "killing a baby" instead
of as "terminating a pregnancy." The phrasing is chosen not only to influence
others' feelings, but also to influence one's own. Or of a "pro-life" person
starts thinking about "taking control of a woman's body" instead of "saving a
baby." IMO people choose euphemistic language so as _not_ to be stimulated to
have feelings they don't want.

It also seems obvious to me that a large part, of the effect of language on
one's feelings is cultural. If "choice" didn't have such a high value for
Americans, the phrase "pro-choice" wouldn't work. It wouldn't have enough
resonance.

But these observations don't speak to the Sapi-Whorf hypothesis, which is
much tougher. Maybe Lakoff's theories about embodied thought are relevant to
Sapir-Whorf.

Greg Lee

unread,
Apr 20, 2002, 8:33:26 PM4/20/02
to

Yes, and that shows that people _think_ language influences thought.
But, of course, we knew that.

--
Greg Lee <gr...@ling.lll.hawaii.edu>

Mark Rosenfelder

unread,
Apr 21, 2002, 3:44:03 AM4/21/02
to
>On Sat, 20 Apr 2002 19:46:54 GMT, Brian M. Scott wrote:
>>I suspect that Wolf is assuming that the names are chosen at least
>>partly for their supposed ability to influence others' attitudes.
>>
>Well, of course.
>
>And any poet knows you can influence attitudes with words. So do advertisers.

I'm suspicious of things that "anyone knows". Lots of things make poems or
ads work... their content, for instance. Word choice may have an effect,
but I suspect much of it is negative: poorly chosen words distract from
the message.

>I'm not talking about explicit persuasion here. Consider what would happen if
>a "pro-choice" person starts thinking of abortion as "killing a baby" instead
>of as "terminating a pregnancy."

What *would* happen? They'd magically become anti-abortion? Has this
happened to anyone you know?

>The phrasing is chosen not only to influence
>others' feelings, but also to influence one's own. Or of a "pro-life" person
>starts thinking about "taking control of a woman's body" instead of "saving a
>baby." IMO people choose euphemistic language so as _not_ to be stimulated to
>have feelings they don't want.

Well, I think that the categories people set up are important to them.
If someone puts a fetus into the "baby" category, that implies certain
behavior, such as not killing it. To me the categories belong to thought,
not language-- if nothing else, so far as I can see animals use categories
too (predator, family, edible food, etc.).

Wolf Kirchmeir

unread,
Apr 21, 2002, 1:56:27 PM4/21/02
to
On 21 Apr 2002 07:44:03 GMT, Mark Rosenfelder wrote:

>Well, I think that the categories people set up are important to them.
>If someone puts a fetus into the "baby" category, that implies certain
>behavior, such as not killing it.

Certainly -- but I note that pro-choice people are careful to talk about
"terminating a pregnancy", and not "abortion," even less "killing a baby."
IOW, terms are chosen to reflect categories and so to evoke
feelings/attitudes appropriate to such categories, and to suppress dissonant
feelings/attitudes. Language does influence thought and feeling -- if it
didn't, people wouldn't take such care in selecting terminology.

OTOH, it's not a one way relationship, which is the reason that many
euphemisms rapidly become as pejorative as the terms they replace. IMO it's
suggestive that one class of euphemisms that degrade refer to mental
disability. It seems we don't like to be reminded that our organic brains are
prone to organic and mechanical glitches which in turn change what we are
pleased to think of as our essential selves. Better to insult the living
evidence of this unwelcome thought. The fact that euphemisms for mental
defects always degrade suggest that the feelings/attitudes associated with
some categories are much less malleable than others.

And some euphemisms last - think of the ones dor "die/death", for example.
The fact that such euphemisms persist IMO demonstrates that we use language
to control our own and other people's feelings.

But this is still not relevant to Sapir-Whorf. :-)

Brian M. Scott

unread,
Apr 21, 2002, 6:41:21 PM4/21/02
to
On Sun, 21 Apr 2002 13:56:27 -0400 (EDT), "Wolf Kirchmeir"
<wwol...@sympatico.ca> wrote:

>On 21 Apr 2002 07:44:03 GMT, Mark Rosenfelder wrote:

>>Well, I think that the categories people set up are important to them.
>>If someone puts a fetus into the "baby" category, that implies certain
>>behavior, such as not killing it.

>Certainly -- but I note that pro-choice people are careful to talk about
>"terminating a pregnancy", and not "abortion,"

In my experience this simply isn't true: the usual expression is
'having an abortion'. But perhaps I move in circles less inclined
toward euphemism.

> even less "killing a baby."

But one can object to this use of the expression 'killing a baby' on
the grounds that it is factually inaccurate and still agree that if it
were accurate, the associated emotional response would be entirely
appropriate.

[...]

>And some euphemisms last - think of the ones dor "die/death", for example.
>The fact that such euphemisms persist IMO demonstrates that we use language
>to control our own and other people's feelings.

But -- as is in particular true of advertisers -- not always
predictably. For instance, I invariably experience a flicker of
annoyance when someone says something like 'He passed last Friday'; he
*died*, damn it!

[...]

Brian

Bobby D. Bryant

unread,
Apr 22, 2002, 11:13:03 PM4/22/02
to
On Thu, 18 Apr 2002 23:00:49 -0600, Mark Rosenfelder wrote:

> I'd say there's just as much anecdotal evidence *against* the Whorfian
> view. Terms for mental retardation are an example: just as quickly as
> people invent new supposedly non-perjorative terms, kids turn them
> into pejoratives. (Indeed, 'retarded' was a favorite epithet when I
> was growing up.) Isn't it possible that, rather than the words making
> people prejudiced, people's prejudices color the words they use?

Yeah, I've noticed the same phenomen in terms for physical impairment.
When I was a kid it was "crippled". It has been through an
ever-accelerating series of replacements since then, and the phenomen
is obvious enough that people make jokes about some of the more recent
stand-ins (e.g., various uses "challenged").


My experience in discussing S-W with various people, including
linguists, is that anyone will concede that the "strong" form of the
hypothesis is wrong but maybe some "weak" form is valid... but as the
discussion continues the "weak" form is progressively watered down
until it is meaningless.

Bobby Bryant
Austin, Texas

Bobby D. Bryant

unread,
Apr 22, 2002, 11:15:34 PM4/22/02
to
On Sat, 20 Apr 2002 09:06:09 -0600, Wolf Kirchmeir wrote:

> On 19 Apr 2002 21:01:34 GMT, Greg Lee wrote:
>
>> I've
>>never seen it argued that it is impossible that language influences
>>thought. The issue is whether it's true, not whether it's possible.
>
>
> There's no question IMO that language influences attitudes and
> feelings. That's why people on one side of a debate will deliberately
> coin terms that make their position seem good and the opposing
> position seem bad. EG, pro- and anti-choice, vs pro- and anti-life.

No, that's only evidence that people _think_ they can sanitize an idea
by sanitizing the language. I'm not aware of any evidence that it
actually works.


> Anyhow, I think Wittgenstein actually said it better:
>
> "The limits of my language are the limits of my world."

It seems to me that our species has never had any trouble thinking up
new ideas, and new words to go along with them.

Bobby Bryant
Austin, Texas

Mark Rosenfelder

unread,
Apr 23, 2002, 12:06:17 AM4/23/02
to
In article <aa2jhc$kg7$1...@geraldo.cc.utexas.edu>,

Bobby D. Bryant <bdbr...@mail.utexas.edu> wrote:
>My experience in discussing S-W with various people, including
>linguists, is that anyone will concede that the "strong" form of the
>hypothesis is wrong but maybe some "weak" form is valid... but as the
>discussion continues the "weak" form is progressively watered down
>until it is meaningless.

Suzette Haden Elgin notes, with some annoyance, that not even Whorfians
believe in the "strong" form.

Bobby D. Bryant

unread,
Apr 23, 2002, 12:33:03 AM4/23/02
to

The question is, is there a form that's "weak" enough for anyone to
believe and yet still "strong" enough to be worth mentioning?

I happen not to think so. Maybe someone will post something in this
thread that motivates me to reconsider.

Bobby Bryant
Austin, Texas

Peter T. Daniels

unread,
Apr 23, 2002, 7:36:45 AM4/23/02
to

That is, like "P.C.," the "strong form" is a straw person invented for
the detractors to make fun of.

However, neither the "strong form" nor the "weak form" is what Whorf
(Sapir had very little to do with it) was suggesting; he offered a
Linguistic Relativity Principle. See Dan Moonhawk Alford's postings some
years ago on LINGUIST List, and, in print, Penny Lee (is that the right
name?)'s *Whorf Theory Complex* (Benjamins, ca. 1998).

John M Sorvari

unread,
Apr 23, 2002, 3:08:52 AM4/23/02
to
How 'bout:

1--the educational/cultural canon so strongly concentrates on
verbal comprehension and expression than nonverbal modes of
thought become atrophied. (S-W Effect)

2--no useful S-W hypothesis for a (large enough) population, but a
range of individual capacity to engage in nonverbal(ized) thinking
or to switch between verbal and nonverbal modes. (S-Wism)
--

John M Sorvari
on my own time
nospam 2 johan

Greg Lee

unread,
Apr 23, 2002, 9:31:36 AM4/23/02
to
Bobby D. Bryant <bdbr...@mail.utexas.edu> wrote:
> On Sat, 20 Apr 2002 09:06:09 -0600, Wolf Kirchmeir wrote:
...

>> Anyhow, I think Wittgenstein actually said it better:
>>
>> "The limits of my language are the limits of my world."

> It seems to me that our species has never had any trouble thinking up
> new ideas, and new words to go along with them.

But suppose we can't actually think up a new idea unless we can find a
potential word to go with it, and suppose the potential words are
limited by our language. There could be an infinity of new words and
new ideas we can think up, yet that does not preclude there also being
an infinity of words and ideas we can't think up, because our language
limits us. The possible ideas might be easy to find new words for
simply because finding a potential word is already part of thinking
up a new idea.

I've been trying to think up an example to illustrate this ...

--
Greg Lee <gr...@ling.lll.hawaii.edu>

Bobby D. Bryant

unread,
Apr 23, 2002, 10:23:35 AM4/23/02
to
On Tue, 23 Apr 2002 07:31:36 -0600, Greg Lee wrote:

> Bobby D. Bryant <bdbr...@mail.utexas.edu> wrote:
>> On Sat, 20 Apr 2002 09:06:09 -0600, Wolf Kirchmeir wrote:
> ...
>>> Anyhow, I think Wittgenstein actually said it better:
>>>
>>> "The limits of my language are the limits of my world."
>
>> It seems to me that our species has never had any trouble thinking up
>> new ideas, and new words to go along with them.
>
> But suppose we can't actually think up a new idea unless we can find a
> potential word to go with it, and suppose the potential words are
> limited by our language.

Yes, "suppose". I'd really be interested in the idea, if someone
actually backed it up with evidence rather than supposition. However,
it looks to me that all the evidence points the other way.

Bobby Bryant
Austin, Texas

Wolf Kirchmeir

unread,
Apr 23, 2002, 10:20:21 AM4/23/02
to
On Tue, 23 Apr 2002 08:23:35 -0600, Bobby D. Bryant wrote:


[quoting Greg Lee:]


>> But suppose we can't actually think up a new idea unless we can find a
>> potential word to go with it, and suppose the potential words are
>> limited by our language.
>
>Yes, "suppose". I'd really be interested in the idea, if someone
>actually backed it up with evidence rather than supposition. However,
>it looks to me that all the evidence points the other way.

Anecdotal evidence:

A) Poetry
One of the common complaints about poetry is "If the poet mean that why
didn't (s)he just say so?" And the standard answer is: "(S)He didn't mean
exactly that, which is why (s)he didn't say exactly that."

One has to somehow stretch one's thought to fully comprehend poetry. (The
fact that one has to do the same for philosophy suggests that philosophy is a
type of poetry. Not very good poetry, though.:-)) That stretching of thought
is somehow made possible by the poetry. A feedback loop? And IMO one cannot
understand some poems until one first understood others.

B) Physics
One has to stretch one's thoughts to comprehend physics. Somehow, the
language of physics enable that stretching.

Here are two stories that provide possible supportive evidence:

A) One evening some years ago, alright, many years ago, a friend and I played
Dylan Thomas's "A Winter's Tale," a long narrative poem, on the record
player. At first it was just wonderful noise -- after all Thomas was reading
it. At some point, around the third or fourth playing, it became crystal
clear. And ever since I've found Thomas's poetry "easy to read", which was
not the case before.

B) I started out to be an engineer. First year physics was a pain. It
required calculus, which was not taught in Alberta high schools at the time.
We took calculus concurrently, but the physics of heat required more than we
had learned. Prof Harry Schiff, a gentle and ironic man, whose specialty was
nuclear physics, and who may have thought his having to teach uncouth louts
like us was a penance for unspecified sin, was lecturing on thermodynamics.
As he talked, he wrote the equations on the board, one by one, showing both
their mathematical derivations and their physical interpretations. When he
was finished, he asked, "Well, gentlemen, how many of you understood that?" A
few tentative hands went up. So Dr Schiff erased everything he'd written, and
started over, repeating his explication word for word as far as I could tell.
And repeated his question. And started over. At the third repetition, what he
said, and what the equations meant, made perfect sense to me. Dr Schiff
repeated the whole performance a fourth time before the 50 minutes were up,
and the fourth repetition seemed to me a model of clear exposition.

So these may be two cases where language, repeated several times, did in fact
enable thinking ideas that could not be thought before. In some ways the
click felt like the click you feel when you master a second language, and
start thinking in it.

Henry Polard

unread,
Apr 23, 2002, 12:15:59 PM4/23/02
to
I would like to know if the languages and clutures of northern
California have been studied with respect to the Sapir-Whorf hypothesis.
It seems to me that there are in that area languages from many language
families spoken by peoples with very similar cultures, so they would be
good test situations.

Or are the cultures not similar enough?

Also, wouldn't other situations where widely different languages adjoin
each other be useful test cases? I have in mind Swedish vs. Finnish,
Estonian vs. Latvian, Hungarian vs. its neighboring laguages, and Basque
vs. French and Spanish.

I am extremely skeptical about the Sapir-Whorf hypothesis, given the
widely different cultures or lifestyles of people who speak English
(_pace_ arguments about whether Vermonters and Texans, let alone Texans
and Yorkshiremen, speak the same language). I tend to ascribe most
cutlural differences to ecology rather than language, but am open to
change.

Henry Polard || Reality is an allusion.

Mark Rosenfelder

unread,
Apr 23, 2002, 12:58:47 PM4/23/02
to
In article <aa3nno$dv1$1...@news.hawaii.edu>,

Greg Lee <gr...@ling.lll.hawaii.edu> wrote:
>Bobby D. Bryant <bdbr...@mail.utexas.edu> wrote:
>> It seems to me that our species has never had any trouble thinking up
>> new ideas, and new words to go along with them.
>
>But suppose we can't actually think up a new idea unless we can find a
>potential word to go with it,

Do you really mean 'word' here? I.e., no one could think about robots
until someone invented the word 'robot'? How was the word invented, then?

This might be more plausible if you substituted 'phrase'; i.e., we can't
think of an idea we can't find words to express. Still, this just seems
wrong to me; the very fact that we struggle sometimes to express ourselves
suggests to me that ideas precede their articulation. And surely we can
(say) experience tastes or smells we're completely unable to describe.

However, I'd go along with the idea that having a word makes the idea
easier to recall and develop. Paraphrases are tedious after a while.

>and suppose the potential words are
>limited by our language.

I'm having trouble imagining how this could be.

Mark Rosenfelder

unread,
Apr 23, 2002, 1:06:10 PM4/23/02
to
>So these may be two cases where language, repeated several times, did in fact
>enable thinking ideas that could not be thought before. In some ways the
>click felt like the click you feel when you master a second language, and
>start thinking in it.

I really don't see how these stories show what you're trying to show. In
the case of the physics problem, there was a derivation the professor was
trying to convey; repetition simply made each step clearer, and allowed
them to connect up into a coherent whole. In the case of the poet, I'd say
you were getting accustomed to his way of speaking.

If all you mean is that ideas can be conveyed by language, I don't think
anyone could disagree; but there's nothing Whorfian about that.

Peter T. Daniels

unread,
Apr 23, 2002, 4:44:38 PM4/23/02
to

Surely any linguistic area would do? The best-studied examples are South
Asia and the Balkans. (Curiously, Hungarian doesn't participate in the
Balkan L.A.) What questions would you put to your hypothetical database?

Peter T. Daniels

unread,
Apr 23, 2002, 4:47:38 PM4/23/02
to
Mark Rosenfelder wrote:
>
> In article <jbysxveflzcngvpbp...@news1.sympatico.ca>,
> Wolf Kirchmeir <wwol...@sympatico.ca> wrote:
> >So these may be two cases where language, repeated several times, did in fact
> >enable thinking ideas that could not be thought before. In some ways the
> >click felt like the click you feel when you master a second language, and
> >start thinking in it.
>
> I really don't see how these stories show what you're trying to show. In
> the case of the physics problem, there was a derivation the professor was
> trying to convey; repetition simply made each step clearer, and allowed
> them to connect up into a coherent whole. In the case of the poet, I'd say
> you were getting accustomed to his way of speaking.

Nope, with Dylan Thomas, the way of speaking definitely doesn't help!

> If all you mean is that ideas can be conveyed by language, I don't think
> anyone could disagree; but there's nothing Whorfian about that.

Whorf did so believe that.

Bobby D. Bryant

unread,
Apr 23, 2002, 5:18:56 PM4/23/02
to
On Tue, 23 Apr 2002 10:15:59 -0600, Henry Polard wrote:

> I would like to know if the languages and clutures of northern
> California have been studied with respect to the Sapir-Whorf
> hypothesis. It seems to me that there are in that area languages from
> many language families spoken by peoples with very similar cultures,
> so they would be good test situations.

If language has any substantial impact on thought, why would the
speakers of different languages have similar cultures in the first
place?


> Or are the cultures not similar enough?

I think anyone who tries to demonstrate that any nontrivial version of
SW is true is going to have a _very_ difficult time avoiding a circular
argument.

Bobby Bryant
Austin, Texas

Bobby D. Bryant

unread,
Apr 23, 2002, 5:23:59 PM4/23/02
to
On Tue, 23 Apr 2002 08:20:21 -0600, Wolf Kirchmeir wrote:

> On Tue, 23 Apr 2002 08:23:35 -0600, Bobby D. Bryant wrote:
>
>
> [quoting Greg Lee:]
>>> But suppose we can't actually think up a new idea unless we can find
>>> a potential word to go with it, and suppose the potential words are
>>> limited by our language.
>>
>>Yes, "suppose". I'd really be interested in the idea, if someone
>>actually backed it up with evidence rather than supposition. However,
>>it looks to me that all the evidence points the other way.
>
> Anecdotal evidence:

...


> So these may be two cases where language, repeated several times, did
> in fact enable thinking ideas that could not be thought before. In
> some ways the click felt like the click you feel when you master a
> second language, and start thinking in it.

It sounds to me like your anecdotes argue the other side of the case.
Did you have to learn French to understand the poetry and Korean to
understand the math? No, you simply stretched your mind into
unfamiliar areas.

Bobby Bryant
Austin, Texas

Mark Rosenfelder

unread,
Apr 23, 2002, 5:35:19 PM4/23/02
to

Of course he did. Who doesn't? That doesn't make it a Whorfian idea.
It's not useful to label as 'Whorfian' things that everyone else believes
as well.

Greg Lee

unread,
Apr 23, 2002, 7:30:00 PM4/23/02
to
Mark Rosenfelder <mark...@enteract.com> wrote:
> In article <aa3nno$dv1$1...@news.hawaii.edu>,
> Greg Lee <gr...@ling.lll.hawaii.edu> wrote:
>>Bobby D. Bryant <bdbr...@mail.utexas.edu> wrote:
>>> It seems to me that our species has never had any trouble thinking up
>>> new ideas, and new words to go along with them.
>>
>>But suppose we can't actually think up a new idea unless we can find a
>>potential word to go with it,

> Do you really mean 'word' here? I.e., no one could think about robots
> until someone invented the word 'robot'? How was the word invented, then?

Yes, I really meant word, and I really meant to include the qualifier
"potential". What I had in mind were arguments given at one time by
some generative semanticist or other (I can't recall -- but probably
mentioned in Lakoff's "Linguistics and Natural Logic") that possible
words are governed by the Ross movement constraints.

> This might be more plausible if you substituted 'phrase'; i.e., we can't
> think of an idea we can't find words to express. Still, this just seems
> wrong to me; the very fact that we struggle sometimes to express ourselves
> suggests to me that ideas precede their articulation.

I have no doubt that sometimes ideas precede their articulation. But
I did not make reference to the _articulation_ of words.

> And surely we can
> (say) experience tastes or smells we're completely unable to describe.

Sure. So can dogs.

> However, I'd go along with the idea that having a word makes the idea
> easier to recall and develop. Paraphrases are tedious after a while.

>>and suppose the potential words are
>>limited by our language.

> I'm having trouble imagining how this could be.

Think of a noun whose only apppropriate single N-bar definition would require
violation of the Coordinate Structure Constraint. If you can't, then
you've imagined it.

--
Greg Lee <gr...@ling.lll.hawaii.edu>

Henry Polard

unread,
Apr 23, 2002, 8:02:22 PM4/23/02
to
In article <3CC5C7...@att.net>,

"Peter T. Daniels" <gram...@att.net> wrote:

> Henry Polard wrote:
> >
> > I would like to know if the languages and clutures of northern
> > California have been studied with respect to the Sapir-Whorf hypothesis.
> > It seems to me that there are in that area languages from many language
> > families spoken by peoples with very similar cultures, so they would be
> > good test situations.

<snip>


> Surely any linguistic area would do? The best-studied examples are South
> Asia and the Balkans. (Curiously, Hungarian doesn't participate in the
> Balkan L.A.) What questions would you put to your hypothetical database?

I don't know - I'm just asking for pointers to studies. I think the
Sapir-Whorf hypothesis is fundamentally untestable, but in light of the
purported "new life" I thought that I could find material to test my
prejudice.

Henry Polard

unread,
Apr 23, 2002, 8:06:03 PM4/23/02
to
In article <aa4j5g$oh7$1...@geraldo.cc.utexas.edu>,

"Bobby D. Bryant" <bdbr...@mail.utexas.edu> wrote:

> On Tue, 23 Apr 2002 10:15:59 -0600, Henry Polard wrote:
>
> > I would like to know if the languages and clutures of northern
> > California have been studied with respect to the Sapir-Whorf
> > hypothesis. It seems to me that there are in that area languages from
> > many language families spoken by peoples with very similar cultures,
> > so they would be good test situations.
>
> If language has any substantial impact on thought, why would the
> speakers of different languages have similar cultures in the first
> place?
>

Indeed. So the situatio in northern California in my mind disconfirms
SW, but I would have have something mre solid than my own ideas.

> > Or are the cultures not similar enough?
>
> I think anyone who tries to demonstrate that any nontrivial version of
> SW is true is going to have a _very_ difficult time avoiding a circular
> argument.

Yup - hence my skepticism.

> Bobby Bryant
> Austin, Texas

-- Henry Polard

Greg Lee

unread,
Apr 23, 2002, 8:51:51 PM4/23/02
to

And what would all that evidence pointing the other way be?
--
Greg Lee <gr...@ling.lll.hawaii.edu>

Brian M. Scott

unread,
Apr 23, 2002, 9:25:36 PM4/23/02
to
On 23 Apr 2002 23:30:00 GMT, Greg Lee <gr...@ling.lll.hawaii.edu>
wrote:

>Mark Rosenfelder <mark...@enteract.com> wrote:
>> In article <aa3nno$dv1$1...@news.hawaii.edu>,
>> Greg Lee <gr...@ling.lll.hawaii.edu> wrote:
>>>Bobby D. Bryant <bdbr...@mail.utexas.edu> wrote:
>>>> It seems to me that our species has never had any trouble thinking up
>>>> new ideas, and new words to go along with them.

>>>But suppose we can't actually think up a new idea unless we can find a
>>>potential word to go with it,

>> Do you really mean 'word' here? I.e., no one could think about robots
>> until someone invented the word 'robot'? How was the word invented, then?

>Yes, I really meant word, and I really meant to include the qualifier
>"potential". What I had in mind were arguments given at one time by
>some generative semanticist or other (I can't recall -- but probably
>mentioned in Lakoff's "Linguistics and Natural Logic") that possible
>words are governed by the Ross movement constraints.

Even if true, why couldn't this be a cognitive constraint on language
rather than the reverse?

[...]

Brian

Mark Rosenfelder

unread,
Apr 24, 2002, 12:36:31 AM4/24/02
to
In article <aa4qpo$mq5$1...@news.hawaii.edu>,
Greg Lee <gr...@ling.lll.hawaii.edu> wrote:
>Mark Rosenfelder <mark...@enteract.com> wrote:

>> Greg Lee <gr...@ling.lll.hawaii.edu> wrote:
>>>and suppose the potential words are
>>>limited by our language.
>
>> I'm having trouble imagining how this could be.
>
>Think of a noun whose only apppropriate single N-bar definition would require
>violation of the Coordinate Structure Constraint. If you can't, then
>you've imagined it.

Someone (I forget the name) suggested that the best definition of "tincture"
is "solution of alcohol and".

One can of course construct other definitions, which I suppose would run
afoul of your "only" above. But that seems weaselly anyway; when is there
only one way to do something in syntax? (I'm sure you remember McCawley's
book, _Thirty Million Theories of Grammar_...)

Greg Lee

unread,
Apr 24, 2002, 1:04:27 AM4/24/02
to
Brian M. Scott <b.s...@csuohio.edu> wrote:
>>words are governed by the Ross movement constraints.

> Even if true, why couldn't this be a cognitive constraint on language
> rather than the reverse?

I don't think I can distinguish those alternatives. The reason for
the Ross constraints, I guess, is a limitation on how we can parse
phrases. Since we don't have real syntactic transformations, to
simulate moving things around we have to distinguish the grammatical
types of phrases related by movement rules. But then those distinct
types can't be coordinated -- hence the Coordinate Structure
Constraint. Is that a language constraint or a cognitive constraint?
Since it has specific reference to the form of language expressions,
in my formulation anyway, I would ordinarily think of it as a language
constraint, but I don't see anything wrong with branding it as
cognitive.
--
Greg Lee <gr...@ling.lll.hawaii.edu>

Greg Lee

unread,
Apr 24, 2002, 1:14:56 AM4/24/02
to
Mark Rosenfelder <mark...@enteract.com> wrote:
> In article <aa4qpo$mq5$1...@news.hawaii.edu>,
> Greg Lee <gr...@ling.lll.hawaii.edu> wrote:
>>Mark Rosenfelder <mark...@enteract.com> wrote:
>>> Greg Lee <gr...@ling.lll.hawaii.edu> wrote:
>>>>and suppose the potential words are
>>>>limited by our language.
>>
>>> I'm having trouble imagining how this could be.
>>
>>Think of a noun whose only apppropriate single N-bar definition would require
>>violation of the Coordinate Structure Constraint. If you can't, then
>>you've imagined it.

> Someone (I forget the name) suggested that the best definition of "tincture"
> is "solution of alcohol and".

Cute example.

> One can of course construct other definitions, which I suppose would run
> afoul of your "only" above. But that seems weaselly anyway; when is there
> only one way to do something in syntax? (I'm sure you remember McCawley's
> book, _Thirty Million Theories of Grammar_...)

Not weaselly. Careful. Cautious. Fearful of counterexamples. Besides,
the underlying idea is that we actually understand words relative to
their derivations from phrases, and one predicts a phrase to be acceptable
if it has at least one correct derivation (regardless of how many incorrect
ones that could be imagined for it).
--
Greg Lee <gr...@ling.lll.hawaii.edu>

Brian M. Scott

unread,
Apr 24, 2002, 1:40:07 AM4/24/02
to
On 24 Apr 2002 05:04:27 GMT, Greg Lee <gr...@ling.lll.hawaii.edu>
wrote:

>Brian M. Scott <b.s...@csuohio.edu> wrote:

>>>words are governed by the Ross movement constraints.

>> Even if true, why couldn't this be a cognitive constraint on language
>> rather than the reverse?

>I don't think I can distinguish those alternatives.

Then I don't see how you can talk about the S-W hypothesis at all, at
least in connection with this example.

> The reason for
>the Ross constraints, I guess, is a limitation on how we can parse
>phrases. Since we don't have real syntactic transformations, to
>simulate moving things around we have to distinguish the grammatical
>types of phrases related by movement rules.

Sorry; this is completely opaque to me. Can you expand it a bit?

>But then those distinct
>types can't be coordinated -- hence the Coordinate Structure
>Constraint. Is that a language constraint or a cognitive constraint?
>Since it has specific reference to the form of language expressions,
>in my formulation anyway, I would ordinarily think of it as a language
>constraint, but I don't see anything wrong with branding it as
>cognitive.

Brian

Peter T. Daniels

unread,
Apr 24, 2002, 7:22:22 AM4/24/02
to
Mark Rosenfelder wrote:
>
> In article <aa4qpo$mq5$1...@news.hawaii.edu>,
> Greg Lee <gr...@ling.lll.hawaii.edu> wrote:
> >Mark Rosenfelder <mark...@enteract.com> wrote:
> >> Greg Lee <gr...@ling.lll.hawaii.edu> wrote:
> >>>and suppose the potential words are
> >>>limited by our language.
> >
> >> I'm having trouble imagining how this could be.
> >
> >Think of a noun whose only apppropriate single N-bar definition would require
> >violation of the Coordinate Structure Constraint. If you can't, then
> >you've imagined it.
>
> Someone (I forget the name) suggested that the best definition of "tincture"
> is "solution of alcohol and".

Which would be fine if we said "tincture iodine," but we don't; it's
"tincture of iodine."

Greg Lee

unread,
Apr 24, 2002, 8:38:33 AM4/24/02
to
Brian M. Scott <b.s...@csuohio.edu> wrote:
> On 24 Apr 2002 05:04:27 GMT, Greg Lee <gr...@ling.lll.hawaii.edu>
> wrote:

>>Brian M. Scott <b.s...@csuohio.edu> wrote:

>>>>words are governed by the Ross movement constraints.

>>> Even if true, why couldn't this be a cognitive constraint on language
>>> rather than the reverse?

>>I don't think I can distinguish those alternatives.

> Then I don't see how you can talk about the S-W hypothesis at all, at
> least in connection with this example.

Well, I don't see what cognition has to do with it. The hypothesis is
not that language influences cognition. The previous discussion had
mentioned "thought" and "ideas". The coordinate structure constraint,
it seems obvious, is not per se a constraint on ideas. You brought
in the term "cognitive", and probably I've guessed wrong at what
you might mean by it.

>> The reason for
>>the Ross constraints, I guess, is a limitation on how we can parse
>>phrases. Since we don't have real syntactic transformations, to
>>simulate moving things around we have to distinguish the grammatical
>>types of phrases related by movement rules.

> Sorry; this is completely opaque to me. Can you expand it a bit?

Extraction constructions, e.g., relative clauses, can be described in
a context free phrase structure grammar, without appeal to grammatical
rules that change constituent structure, by assigning special
types. Instead of being an instance of S, a relative clause is,
let's say, an instance of RELATIVECLAUSE. A basic constraint on
coordination is that only things of the same grammatical type can
be coordinated. S and RELATIVECLAUSE differ, so things of those
respective types can't coordinate.

>>But then those distinct
>>types can't be coordinated -- hence the Coordinate Structure
>>Constraint. Is that a language constraint or a cognitive constraint?
>>Since it has specific reference to the form of language expressions,
>>in my formulation anyway, I would ordinarily think of it as a language
>>constraint, but I don't see anything wrong with branding it as
>>cognitive.

> Brian

--
Greg Lee <gr...@ling.lll.hawaii.edu>

Mark Rosenfelder

unread,
Apr 24, 2002, 11:40:37 AM4/24/02
to
In article <3CC695...@att.net>, Peter T. Daniels <gram...@att.net> wrote:

>Mark Rosenfelder wrote:
>> Someone (I forget the name) suggested that the best definition of "tincture"
>> is "solution of alcohol and".
>
>Which would be fine if we said "tincture iodine," but we don't; it's
>"tincture of iodine."

Which they'd presumably derive from "solution of alcohol and of iodine".

Brian M. Scott

unread,
Apr 24, 2002, 12:41:35 PM4/24/02
to
On 24 Apr 2002 Greg Lee <gr...@ling.lll.hawaii.edu> wrote in
news:aa6909$avh$1...@news.hawaii.edu in sci.lang:

> Brian M. Scott <b.s...@csuohio.edu> wrote:
>> On 24 Apr 2002 05:04:27 GMT, Greg Lee
>> <gr...@ling.lll.hawaii.edu> wrote:

>>>Brian M. Scott <b.s...@csuohio.edu> wrote:

>>>>>words are governed by the Ross movement constraints.

>>>> Even if true, why couldn't this be a cognitive
>>>> constraint on language rather than the reverse?

>>>I don't think I can distinguish those alternatives.

>> Then I don't see how you can talk about the S-W hypothesis
>> at all, at least in connection with this example.

> Well, I don't see what cognition has to do with it. The
> hypothesis is not that language influences cognition. The
> previous discussion had mentioned "thought" and "ideas".
> The coordinate structure constraint, it seems obvious, is
> not per se a constraint on ideas. You brought in the term
> "cognitive", and probably I've guessed wrong at what you
> might mean by it.

I was using it loosely as an adjective meaning something like
'related to thought'. Why can't the CSC, a linguistic
constraint, be a result rather than (as I took you to be
suggesting) a cause of a constraint on the ideas that we can
have?

[...]

Brian

Bobby D. Bryant

unread,
Apr 24, 2002, 8:22:41 PM4/24/02
to
On Tue, 23 Apr 2002 18:51:51 -0600, Greg Lee wrote:

> Bobby D. Bryant <bdbr...@mail.utexas.edu> wrote:
>> On Tue, 23 Apr 2002 07:31:36 -0600, Greg Lee wrote:
>
>>> Bobby D. Bryant <bdbr...@mail.utexas.edu> wrote:
>>>> On Sat, 20 Apr 2002 09:06:09 -0600, Wolf Kirchmeir wrote:
>>> ...
>>>>> Anyhow, I think Wittgenstein actually said it better:
>>>>>
>>>>> "The limits of my language are the limits of my world."
>>>
>>>> It seems to me that our species has never had any trouble thinking
>>>> up new ideas, and new words to go along with them.
>>>
>>> But suppose we can't actually think up a new idea unless we can find
>>> a potential word to go with it, and suppose the potential words are
>>> limited by our language.
>
>> Yes, "suppose". I'd really be interested in the idea, if someone
>> actually backed it up with evidence rather than supposition. However,
>> it looks to me that all the evidence points the other way.
>
> And what would all that evidence pointing the other way be?

Lessee.

Mentioned elsewhere in this thread is the fact that we have groups which
have very similar cultures and not only speak different languages but
languages from different families. If S-W has any nontrivial effect I
would find this very surprising; indeed, I would expect that the less
closely related two languages are, the more different the cultures of
their respective speakers would be.

Then there's the fact that our species very often comes up with new
ideas and has to invent names for those ideas. The linguistic gap does
not seem to offer any noticable barrier to conceptual innovations.
(And new ideas get around quite well without respect to linguistic
borders, it seems. For instance, automatic rifles have virtually
saturated the world over the past half century, but no one seems to
have any difficulty recognizing what they are or how to use them.)

Is there any motivation for thinking that any nontrivial version of S-W
is valid?

Bobby Bryant
Austin, Texas

Wolf Kirchmeir

unread,
Apr 24, 2002, 9:26:25 PM4/24/02
to
On Wed, 24 Apr 2002 18:22:41 -0600, Bobby D. Bryant wrote:

>Is there any motivation for thinking that any nontrivial version of S-W
>is valid?
>
>Bobby Bryant
>Austin, Texas

Well, I did come across the claim that symbolic logic was invented by English
speakers because in English syntax the structure of proportional logic is
easier to state and perceive. But IMO that claim is akin to a liquid useful
for the cleaning of hogs.

Peter T. Daniels

unread,
Apr 24, 2002, 11:22:43 PM4/24/02
to

Which is fine in French or Italian, but very iffy in English. We don't
generally repeat prepositions without a very good reason.

Greg Lee

unread,
Apr 25, 2002, 1:08:20 AM4/25/02
to

My supposition concerned a connection between ideas and potential words.

> Then there's the fact that our species very often comes up with new
> ideas and has to invent names for those ideas. The linguistic gap does
> not seem to offer any noticable barrier to conceptual innovations.
> (And new ideas get around quite well without respect to linguistic
> borders, it seems. For instance, automatic rifles have virtually
> saturated the world over the past half century, but no one seems to
> have any difficulty recognizing what they are or how to use them.)

> Is there any motivation for thinking that any nontrivial version of S-W
> is valid?

I don't know. I do have the opinion that your observation about our
species not having trouble thinking up new ideas and new words is
irrelevant, and I don't see here any of this evidence "pointing the
other way" from the supposition I offered.

--
Greg Lee <gr...@ling.lll.hawaii.edu>

Bobby D. Bryant

unread,
Apr 25, 2002, 1:17:37 AM4/25/02
to
On Wed, 24 Apr 2002 23:08:20 -0600, Greg Lee wrote:

> I don't know. I do have the opinion that your observation about our
> species not having trouble thinking up new ideas and new words is
> irrelevant, and I don't see here any of this evidence "pointing the
> other way" from the supposition I offered.

Yes, my mistake. I see now that I took your supposition to be an
indication that you subscribed to S-W, so I erroneously responded with
S-W in mind rather than with your more narrow supposition.

And even at that, I acknowledge that my evidence is weak. Still, if
anyone here subscribes to a non-trivial version of S-W I'd like to hear
what that version is, and what compells them to believe it.


But back to your supposition. That's going to be _very_ hard to examine
empirically, no?

Bobby Bryant
Austin, Texas

Greg Lee

unread,
Apr 25, 2002, 7:34:54 AM4/25/02
to
Bobby D. Bryant <bdbr...@mail.utexas.edu> wrote:
...

> But back to your supposition. That's going to be _very_ hard to examine
> empirically, no?

Yes, so long as we have only human languages to examine.

--
Greg Lee <gr...@ling.lll.hawaii.edu>

Greg Lee

unread,
Apr 25, 2002, 7:45:19 AM4/25/02
to
Brian M. Scott <b.s...@csuohio.edu> wrote:
...

> I was using it loosely as an adjective meaning something like
> 'related to thought'. Why can't the CSC, a linguistic
> constraint, be a result rather than (as I took you to be
> suggesting) a cause of a constraint on the ideas that we can
> have?

Not knowing the nature of ideas, there's no way I could answer
this. I have no idea (so to speak) of what a non-linguistic
constraint on ideas might be like.
--
Greg Lee <gr...@ling.lll.hawaii.edu>

michael farris

unread,
Apr 25, 2002, 9:29:19 AM4/25/02
to

Henry Polard wrote:

> Also, wouldn't other situations where widely different languages adjoin
> each other be useful test cases? I have in mind Swedish vs. Finnish,

I remember an article (well I didn't read it, a student did a report on it)
that claimed that Finnish speakers have a higher per capita rate of
industrial accidents than Swedish speakers (IIRC the article examined
monolingual areas of Sweden and Finland as well as an area of Finland where
both Swedish and Finnish speakers are found). Again, IIRC the author claimed
this was related to something about the way location is expressed in the two
languages. I can't remember if the article (don't expect me to be able to
find it) mentions SW at all.

-michael farris

Bob LeChevalier

unread,
May 1, 2002, 3:44:19 PM5/1/02
to
"Bobby D. Bryant" <bdbr...@mail.utexas.edu> wrote:
>On Wed, 24 Apr 2002 23:08:20 -0600, Greg Lee wrote:
>
>> I don't know. I do have the opinion that your observation about our
>> species not having trouble thinking up new ideas and new words is
>> irrelevant, and I don't see here any of this evidence "pointing the
>> other way" from the supposition I offered.
>
>Yes, my mistake. I see now that I took your supposition to be an
>indication that you subscribed to S-W, so I erroneously responded with
>S-W in mind rather than with your more narrow supposition.
>
>And even at that, I acknowledge that my evidence is weak. Still, if
>anyone here subscribes to a non-trivial version of S-W I'd like to hear
>what that version is, and what compells them to believe it.

I am not sure that I subscribe to it, but Lojban is intended to test a
version of the Sapir-Whorf hypothesis: that the structure of language tends
to restrict or limit thinking. The language is designed to significantly
remove certain categorical restrictions found in natural languages, while
otherwise having the capabilities of natural languages. This would, if this
version of S/W is true, result in some measurable increase in thinking of the
type that is no longer restricted.

Lojban is noted for its focus on predicate logic, and thus one would conclude
that fluent speakers of Lojban might more easily grasp the thinking involved
in predicate logic without resort to the (non-linguistic) analytic methods
used by logicians (or perhaps they might learn those analytic methods more
easily). Unfortunately, few people study Lojban who don't already know a lot
about predicate logic, and the ways that we teach the language tend to negate
the possibility of a test, so a S/W would have to await development of a
speaking population that learns the language without resort to explicit logic
training.

Two other areas where Lojban has removed restriction seem more amenable to
testing in the short term.

There is a sublanguage of attitudinal indicators primarily used for emotional
expression, related to the supersegmentals and other methods used in various
natural languages to convey emotion, but carried to an extreme in both
capability and isolation from the rest of the language. I suspect that even
managing to internalize this system linguistically, no matter how it is done,
would show some effects on thinking and expression, perhaps even when
speaking other languages. Psychological testing of emotional responses
exists and is largely independent of language, so there is a means to test
for a S/W effect in this realm.

The other area is in the realm of creativity. Lojban minimizes distinctions
in word categories, while making some distinctions (object raising) more
explicit than in (most?) natural languages. The immediate result many
students of Lojban notice, is that they tend to start combining words in
unusual ways "to see what happens", and that this thinking feeds back on both
their Lojban and on their native language usage. Both in direct tests of
creativity, and in more subtle tests of the effects of removing some
distinctions could provide good S/W tests.

But I am hardly "compelled to believe" that these effects exist. I merely
have observed that there are signs that they may exist, and am interested to
see if I can get the project to the stage where linguists and psychologists
are willing to set up the sorts of tests that would lead to more definite
results.

If such results were found, there might be pragmatic implications for the
teaching of second languages, as well as applications for Lojban, or
Lojban-based methods in dealing with problems in either logic teaching or
emotional management/expression, as well as providing further basis for
research in S/W in other languages by showing the sorts of things to look for
in identifying S/W effects.

lojbab

Bobby D. Bryant

unread,
May 2, 2002, 1:26:17 AM5/2/02
to
On Wed, 01 May 2002 13:44:19 -0600, Bob LeChevalier wrote:

> "Bobby D. Bryant" <bdbr...@mail.utexas.edu> wrote:
>>On Wed, 24 Apr 2002 23:08:20 -0600, Greg Lee wrote:
>>
>>> I don't know. I do have the opinion that your observation about our
>>> species not having trouble thinking up new ideas and new words is
>>> irrelevant, and I don't see here any of this evidence "pointing the
>>> other way" from the supposition I offered.
>>
>>Yes, my mistake. I see now that I took your supposition to be an
>>indication that you subscribed to S-W, so I erroneously responded with
>>S-W in mind rather than with your more narrow supposition.
>>
>>And even at that, I acknowledge that my evidence is weak. Still, if
>>anyone here subscribes to a non-trivial version of S-W I'd like to
>>hear what that version is, and what compells them to believe it.
>
> I am not sure that I subscribe to it, but Lojban is intended to test a
> version of the Sapir-Whorf hypothesis: that the structure of language
> tends to restrict or limit thinking. The language is designed to
> significantly remove certain categorical restrictions found in natural
> languages, while otherwise having the capabilities of natural
> languages. This would, if this version of S/W is true, result in some
> measurable increase in thinking of the type that is no longer
> restricted.

[...]

Thanks for that post. Maybe we'll know more before too terribly long.

Bobby Bryant
Austin, Texas

F Cameron Sipston

unread,
May 22, 2002, 8:57:44 AM5/22/02
to
"Wolf Kirchmeir" <wwol...@sympatico.ca> wrote in message news:<jbysxveflzcngvpbp...@news1.sympatico.ca>...
> On 21 Apr 2002 07:44:03 GMT, Mark Rosenfelder wrote:
>
> >Well, I think that the categories people set up are important to them.
> >If someone puts a fetus into the "baby" category, that implies certain
> >behavior, such as not killing it.
>
> Certainly -- but I note that pro-choice people are careful to talk about
> "terminating a pregnancy", and not "abortion," even less "killing a baby."
> IOW, terms are chosen to reflect categories and so to evoke
> feelings/attitudes appropriate to such categories, and to suppress dissonant
> feelings/attitudes. Language does influence thought and feeling -- if it
> didn't, people wouldn't take such care in selecting terminology.

In reality this is more to with
medical English than any kind
of political correctness.

By this I mean that what a non-medic
would refer to as a "miscarriage"
is, in medical terms, strictly speaking,
an "abortion" when it comes to writing
up patients' notes, whereas what vernacular
has as an "abortion" is, medically speaking,
strictly referred to as a "termination".

Maybe this has something to do with the
old practices of downing litres of Gin
in the backstreets of Conan-Doyle's
London and the linguistic hegemony of
English in clinical practice but, whichever
way, the contributing factors informing
lexes here are a bit more complex than
simply a quirk of pro-choice vs pro-life
sociolinguistic identities.

<snip> (no pun intended)

FCS
--

0 new messages