Odd linguistics argument

[Roboramma]

Hi all. I'm involved in a little debate about atheism on a yoga board. I'm putting this in the science forum because I'm asking mainly about the scientific claims, and whether they are acurate.

Anyway, someone brought up some points about linguistics to suggest that words and definitions change the world, or something like that. I'm wondering if anyone can tell me anything about the linguistics he's talking about, and how acurate what he's saying is.

I don't really know all that much about this, aside from what I've read from Steven Pinker, but as my knowledge is limited, I was hoping someone else could help me broaden it.

The full post his here: http://p196.ezboard.com/fyoga84291fr...art=61&stop=72

Quote:

In linguistics, one of the ideas advanced by the Sapir-Whoft hypothesis---which can list towards linguistic determinism---is that most Western languages in general tend to analyze reality as objects in space: the present and future are thought of as a "places," and time is a path linking them.

A phrase like "three days" is grammatically equivalent to "three apples" or "three miles."

This is in comparison with other languages---Whorf's famous example was of the Hopi Native American language, which is oriented towards process.

So how is a Hopi's reality shaped by language?

Whorf advanced the idea that a Hopi speaker would find relativistic physics easier to understand than an Western language speaker.

Um, empirically?

"Hunter-gatherers from the Pirahã tribe, whose language only contains words for the numbers one and two, were unable to reliably tell the difference between four objects placed in a row and five in the same configuration.

Peter Gordon, the psychologist at Columbia University in New York City who carried out the experiment, does not claim that his finding holds for all kinds of thought.

'There are certainly things that we can think about that we cannot talk about. But for numbers I have shown that a limitation in language affects cognition.'"

From New Scientist, 19 August 2004

I'm actually not entirely sure what he's saying. Obviously words affect thought, but I don't think they do so to a very high extent. If there's a concept we don't have a word for, but it would be useful to have a word for it, we're not incapable of thinking about it. We just come up with a new word.

[Shrinker]

The science as quoted says people who think differently, talk differently or vice versa. Big deal. What the science doesn't say is people who think/talk differently have a different physical experience of the universe, which seems to be what he's trying to claim.

If you don't have a word for donkey, it doesn't mean you can't ride one.

[Lord Muck oGentry]

Originally Posted by Shrinker

The science as quoted says people who think differently, talk differently or vice versa. Big deal. What the science doesn't say is people who think/talk differently have a different physical experience of the universe, which seems to be what he's trying to claim.

If you don't have a word for donkey, it doesn't mean you can't ride one.

There's some useful stuff about this at Wikipedia:

http://en.wikipedia.org/wiki/Sapir-Whorf_Hypothesis

I thought the language makes reality horse had expired years ago. But there are those who want to flog the poor beast back to life...

[Paul C. Anagnostopoulos]

Quote:

In linguistics, the Sapir-Whorf hypothesis (SWH) states that there is a systematic relationship between the grammatical categories of the language a person speaks and how that person both understands the world and behaves in it.

Sounds entirely reasonable to me. I would expect such a relationship. Cause and effect, now that's another matter.

~~ Paul

[Alkatran]

Somehow I find it hard to believe that people wouldn't be able to distinguish between 4 and 5 apples.

[Mercutio]

I will have to check around for it, but it seems to me I recently read where the "1,2, more than 2" cultures could indeed tell the difference between larger numbers. They used qualifiers--sort of a "much more than two", "a little more than two", "way way more than two".

Same hypothesis makes a big deal of the many words for "snow" that the eskimos have. We just have "snow"...which we modify, using "powdery", "fluffy", "sticky", "dense", "driving", etc.

[Ririon]

Originally Posted by Alkatran

Somehow I find it hard to believe that people wouldn't be able to distinguish between 4 and 5 apples.

Chimpanzees can do this with some training, so i share your doubt.

An excellent overview of words that don't exist in the English language:
http://www.douglasadams.com/creations/liff.html

Ririon

[Dr Adequate]

Hopi

[Paul C. Anagnostopoulos]

Originally Posted by Merc

Same hypothesis makes a big deal of the many words for "snow" that the eskimos have. We just have "snow"...which we modify, using "powdery", "fluffy", "sticky", "dense", "driving", etc.

I think they have about 20 words for ice and snow. It's a polysynthetic language, so you can just make them up as you go along.

Meanwhile: snow, ice, iceberg, snowflake, ice crystal, ice fog, hoarfrost, frazil, sleet, hail, graupel, powder, packed powder, wet powder, corn snow, frozen granular, wet granular, loose granular, hardpack, snow drift, snow flurry, snow shower, snow crust, sun crust, rain crust, wind slab, ice crust, film crust, dendrite, blizzard, firn, rime.

~~ Paul

[Ririon]

Snow and snow-related phenomena are important and generate a lot of words in areas with a lot of snow. It may create more words than seems intuitively necessary for people living closer to the equator, but it's really not that surprising when you think of it.

A preliminary list of about 400 snow-related Norwegian words:
http://folk.ntnu.no/ivarse/snjoord.html
(With English explanations.)

I'm sure you can make similar long lists for any language used in polar areas. Including English in Canada and the USA.

Ririon

[Ryan O'Dine]

As an aside, people can identify three objects without counting, but must “laboriously” count to identify four or more. This was detected by timing how fast people can determine #’s of objects in a lab. 1, 2 and 3 were almost instantaneous. 4 and more took several extra milliseconds.

This shows up in language. Ordinal #’s don’t regularize until 4: first, second, third, fourth, fifth, sixth...

This is in Stanislas Dehaene’s The Number Sense, which goes into more language/number relationships.

[UserGoogol]

Language might not mold your perception of reality, but math does. If you don't have a particularly well-defined concept of the number four, you're not going to be able to count things as well as other people. And as the above poster points out, around four or five is where people lose their ability to "instinctively" count the number of things.

Of course, I might be biased because I really like math. But still, math is different from language, and appears to be slightly less innate than language is. If you haven't learned the skill, you haven't learned the skill.

[Soapy Sam]

I thought T.H. Huxley invented the term "agnostic", not Bertrand Russell.

The Whorf claim I read was that pre- white man, the Hopi had no word for "time". I understand this has been shown to be a misunderstanding.
(What they did not have were mechanical clocks. Neither did white men until quite recently).

I've never understood how anyone can seriously think that language shapes reality. It undoubtedly affects our view of reality, but that's not the same thing at all.

[epepke]

Originally Posted by Paul C. Anagnostopoulos

I think they have about 20 words for ice and snow. It's a polysynthetic language, so you can just make them up as you go along.

Except that this is false.

[Paul C. Anagnostopoulos]

Originally Posted by Epepke

Except that this is false.

Indeed, the claim that they have 100 words for snow is false. I've managed to find about 20.

~~ Paul

Edited to add: Looks like a great book. I just ordered it. Thanks!

[Kotatsu]

Originally Posted by Ryan O'Dine

As an aside, people can identify three objects without counting, but must “laboriously” count to identify four or more. This was detected by timing how fast people can determine #’s of objects in a lab. 1, 2 and 3 were almost instantaneous. 4 and more took several extra milliseconds.

This shows up in at least the English language. Ordinal #’s don’t regularize until 4: first, second, third, fourth, fifth, sixth...

This is in Stanislas Dehaene’s The Number Sense, which goes into more language/number relationships.

I have corrected your post, as your rule is not universally applicable.

[CurtC]

I don't know the details of the Piraha tribe or the experiment that determined they have trouble differentiating between four apples and five, but what sounds more plausbile to me is that the researchers simply had trouble communicating the task to the Piraha that they were being asked to accomplish.

[drkitten]

Originally Posted by Paul C. Anagnostopoulos

Sounds entirely reasonable to me. I would expect such a relationship. Cause and effect, now that's another matter.

My understanding is that even showing a relationship -- ignoring causaility for the moment -- has been extremely difficult.

Basically, the Sapir-Whorf hypothesis is dead, but well-meaning idiots continue to indulge in necromancy. The best evidence we have against the Sapir-Whorf hypothesis are the experiments done by Brent Berlin and Paul Kay on basic color terms, which I will be happy to discuss at length if anyone really wants to see me in full-on lecture mode. But S-W will unfortunately go down in history as yet another beautiful theory brutally done in in a dark alley by an ugly fact.

[Roboramma]

Originally Posted by drkitten

My understanding is that even showing a relationship -- ignoring causaility for the moment -- has been extremely difficult.

Basically, the Sapir-Whorf hypothesis is dead, but well-meaning idiots continue to indulge in necromancy. The best evidence we have against the Sapir-Whorf hypothesis are the experiments done by Brent Berlin and Paul Kay on basic color terms, which I will be happy to discuss at length if anyone really wants to see me in full-on lecture mode. But S-W will unfortunately go down in history as yet another beautiful theory brutally done in in a dark alley by an ugly fact.

I'd love to hear about it. I hope I can keep up, of course, but I'll absorb what I can.

To all who've posted so far, thanks for the responses, it's more than I expected. I think I'll have to really go over all this a little more before I can make a meaningful contribution.

[Paul C. Anagnostopoulos]

Ooh, lecture, lecture! I'm interested to hear about the experiments, Dr. K.

~~ Paul

[Mercutio]

Originally Posted by drkitten

Basically, the Sapir-Whorf hypothesis is dead, but well-meaning idiots continue to indulge in necromancy. The best evidence we have against the Sapir-Whorf hypothesis are the experiments done by Brent Berlin and Paul Kay on basic color terms, which I will be happy to discuss at length if anyone really wants to see me in full-on lecture mode. But S-W will unfortunately go down in history as yet another beautiful theory brutally done in in a dark alley by an ugly fact.

Um...given that I teach about this at times, and would love to steal your notes for my class, I would love to discuss this. I would even be happy to play devil's advocate (patsy) and try to poke holes in your lecture.

[Piscivore]

I want to hear it too.

[drkitten]

Originally Posted by Roboramma

I'd love to hear about it. I hope I can keep up, of course, but I'll absorb what I can.

All right. To some extent, I'm doing this from memory as my copy of the book is in my other office (or possibly in a box somewhere). But basically, what Berlin and Kay studied were the cross-linguistic properties of "basic color terms."

These are defined as single words that describe colors, with the following three properties.

1. the words are monomorphemic, meaning you can't break them apart into meaningful sub-units (which rules out words like "blueish-gray").
2. the words are not the names of any objects they purport to describe (which rules out "teal," "olive," and "cornflower").
3. the words are universally applicable across domains (which rules out words like "roan"," which can only be used to describe a horse).

In English, there are (IIRC) eleven such basic color terms:

black, white, red, yellow, green, blue, brown, purple, orange, pink, gray

However, this number is not "universal" -- many languages do not have words for all of these distinctions. What does seem to be universal is the hierarchical ordering of these words, which is to say, if a language has one of these terms, it will have all all of the terms to the left as well. No language, for example, has a word for blue, without having words for black, white, red, yellow, and green -- but a language might have words for green and yellow without having words for blue. (Japanese is such a language.)

So we've got clear evidence that color naming is, in fact, language-bound. However, Berlin and Kay did a number of other experiments, and found that there is no relationship between color naming and color perception. People had no problem distinguishing (or even describing) different "color," even when the same basic term was used in their language. The classic and off-cited example is from a language that had only three color terms -- black, white, and red -- but yellow things were distinguished from red things by being "red like banana." English-speakers do the same thing; we have "sky blue" and "navy blue" and "royal blue" and "midnight blue."

[LW]

Originally Posted by drkitten

black, white, red, yellow, green, blue, brown, purple, orange, pink, gray

However, this number is not "universal" -- many languages do not have words for all of these distinctions. What does seem to be universal is the hierarchical ordering of these words, which is to say, if a language has one of these terms, it will have all all of the terms to the left as well.

The ordering is not completely universal. For example, Finnish has had a word for grey for quite a long time (hundreds of years) but the word for pink is a very recent loan word: even I can remember time when we used 'light red' instead.

[LordoftheLeftHand]

This reminds me of Orwell's writings where the Party was trying to destroy certain ideas (like thoughtcrime) by destroying words. The theory was that if you had no terms to express dissent; dissent would be impossible. While I don't really believe this, it sure is interesting.

LLH

[ClusterBoy]

Originally Posted by drkitten

but a language might have words for green and yellow without having words for blue. (Japanese is such a language."

Fascinating. But is it true? I just asked the japanese native speaker in my office if that was true and she said no. Do you mean it is a recent import? Historically they didn't have one?

[drkitten]

Originally Posted by ClusterBoy

Fascinating. But is it true? I just asked the japanese native speaker in my office if that was true and she said no. Do you mean it is a recent import? Historically they didn't have one?

The word that the Japanese commonly use (today) for green is "midori," which literallly means "honeydew melon" -- and hence is not a "basic color term" as defined by Berlin and Kay. It's not "new" in the sense of being an imported loan word, but its use as a common descriptor instead of an unusual and somewhat metaphorical word is relatively new (past fifty years or so, I believe). The traditional Japanese word for "blue" (which I can't remember offhand) historically covers both blues and greens, and it is only within living memory that people have started not using it to refer to greens.

If you want an English analogy, think of "avocado," a word that really only became popular as a color descriptor in the 1960s and 1970s when it became the new hot designer shade for about fifteen minutes. The difference is that "avocado" didn't really have a new semantic space to move into, since "green" already existed in English, and it's very hard to displace existing basic color terms.

[drkitten]

Originally Posted by LW

The ordering is not completely universal. For example, Finnish has had a word for grey for quite a long time (hundreds of years) but the word for pink is a very recent loan word: even I can remember time when we used 'light red' instead.

The actual ordering as found by Berlin and Kay is not entirely linear, but I lack the appropriate notation facilities to describe it accurately in this forum (and wanted to simplify to avoid making the post fifty meters long). A better description would be the following (top to bottom)

black
white
red
green | yellow
blue
brown
purple | orange | pink | grey

(assuming I remembered that right)

If a language has a word, then it has all the words above it, but not necessarily all the words on the same level. So a language may have a word for green or a word for yellow, but not necessarily both -- but if it has a word for blue, it will have separate words for green and for yellow.

[Avita]

Russian has no word for "privacy," but I have a clear memory of discussing with a friend of mine once how unhappy she was to be living in the same room as her mother. Since we couldn't call her uncomfortable feelings "lack of privacy," we instead talked about her mother being there all the time, etc., and understood perfectly what was so bad about that. So not having the word did not prevent either of us from experiencing and understanding lack of privacy, and later, when my family immigrated and I did acquire the word, I had no difficulty applying it to describe that situation.

[Piscivore]

Originally Posted by drkitten

The actual ordering as found by Berlin and Kay is not entirely linear, but I lack the appropriate notation facilities to describe it accurately in this forum (and wanted to simplify to avoid making the post fifty meters long). A better description would be the following (top to bottom)

black
white
red
green | yellow
blue
brown
purple | orange | pink | grey

(assuming I remembered that right)

If a language has a word, then it has all the words above it, but not necessarily all the words on the same level. So a language may have a word for green or a word for yellow, but not necessarily both -- but if it has a word for blue, it will have separate words for green and for yellow.

How much of this is due to relatedness in the evolution of the languages, though? No language evolved in perfect isolation, right?

ETC: "languange"?

[Paul C. Anagnostopoulos]

Originally Posted by Drkitten

If you want an English analogy, think of "avocado," a word that really only became popular as a color descriptor in the 1960s and 1970s when it became the new hot designer shade for about fifteen minutes.

Actually, it was an entire hour, namely one installment of Rowan and Martin's Laugh In.

~~ Paul

[aggle-rithm]

Originally Posted by Paul C. Anagnostopoulos

I think they have about 20 words for ice and snow. It's a polysynthetic language, so you can just make them up as you go along.

Meanwhile: snow, ice, iceberg, snowflake, ice crystal, ice fog, hoarfrost, frazil, sleet, hail, graupel, powder, packed powder, wet powder, corn snow, frozen granular, wet granular, loose granular, hardpack, snow drift, snow flurry, snow shower, snow crust, sun crust, rain crust, wind slab, ice crust, film crust, dendrite, blizzard, firn, rime.

~~ Paul

In Texas, we only have one word for snow: "Itllneverstick".

[dogjones]

Originally Posted by drkitten

The actual ordering as found by Berlin and Kay is not entirely linear, but I lack the appropriate notation facilities to describe it accurately in this forum (and wanted to simplify to avoid making the post fifty meters long). A better description would be the following (top to bottom)

black
white
red
green | yellow
blue
brown
purple | orange | pink | grey

(assuming I remembered that right)

If a language has a word, then it has all the words above it, but not necessarily all the words on the same level. So a language may have a word for green or a word for yellow, but not necessarily both -- but if it has a word for blue, it will have separate words for green and for yellow.

1. Does the fact that "orange" is included imply that the colour existed in English before the fruit?

2. Where does indigo fit in, assume same level as purple?

[drkitten]

Originally Posted by Piscivore

How much of this is due to relatedness in the evolution of the languages, though?

As little as possible. Standard practice in doing this kind of study is to select languages (and language groups) that are as areally and geographically separate as possible. So if I were to use English as a data point in this study, I wouldn't use German or Dutch as well -- and if I were to be really really careful, I wouldn't use any other Indo-European language at all, so no Russian, no Armenian, and no Greek, either. Similarly, I will only pick one language out of the dozens in use in China, and if possible I will pick a Chinese language used well away from the Korean border if I plan to use Korean as another language in my study.

Selecting samples for linguistic typology studies is something of an art form in and of itself -- but it's a well-known and well-understood art form. I don't remember the exact samples that Berlin and Kay used, but they have been held up as models to follow, so I assume that they did their work well.

[drkitten]

Originally Posted by dogjones

1. Does the fact that "orange" is included imply that the colour existed in English before the fruit?

Or that I misremembered "orange" as being treated a basic color word in English; if Berlin and Kay didn't treat it as "basic," then English has only ten basic terms, and other languages have a non-descriptive BCT term for that color that we don't have in English.

Quote:

2. Where does indigo fit in, assume same level as purple?

I'm not sure "indigo" is commonly enough understood to be used in this experiment. Part of the experiment involved asking people to point to an "X" paint chip. How many people would be able to identify or use "indigo"?

Of course, it's also the name of a plant as well (from which the color derives its name), so they may have excluded it on those grounds. I would need to check the journals for a question this specific.

[pgwenthold]

Originally Posted by Ririon

Chimpanzees can do this with some training, so i share your doubt.

Crows, otoh, can only count to about 3. They don't seem to be able to distinguish between 4 and 5.

ETA: I should add that they do this without any training, and can do it naturally.

[Piscivore]

Did they suggest a "why" those colours are named first?

[drkitten]

Originally Posted by Piscivore

Did they suggest a "why" those colours are named first?

Not really. There have been a few attempts to explain these results in terms of cognitive psychology since then, but none that I really find convincing or credible.

[Piscivore]

Here's a linguistics question then- how can we be certain, give our common genetic ancestry, that all existing languages did not have a common basis at some point?

[drkitten]

Originally Posted by Piscivore

Here's a linguistics question then- how can we be certain, give our common genetic ancestry, that all existing languages did not have a common basis at some point?

We cannot. Nor can we be certain that faeries are not the true reason why bread rises.

On the other hand, we have extreme difficulty in reconstructing anything about language -- even the existence of it -- about 10,000 years ago. A few linguistics, most notably Ruhlen and Greenberg, have seriously proposed some reconstructions of a so-called proto-World language, but very few mainstream linguists take these reconstructions seriously, mostly due to a variety of methodological problems in their reconstruction.

So aside from the fact that the proto-World hypothesis is itself rather questionable, it then raises the issue of why there's no evident pattern in the distribution of color word schema. In particular, given the isolation of the various continents, we would expect to see characteristically "American" color word patterns, "Australian," and "Eurasian-African." We don't.

(As it happens, we also know that not all languages have a common genetic basis, because we've seen instances of language being created ex nihilo. NIcaraguan Sign Langauge is probably the best known and best documented. But these are isolated enough cases that the question as phrased is still meaningful.)