Slashdot Log In
One, Two, Many - Language Shapes Thought
Posted by
CowboyNeal
on Fri Aug 20, 2004 07:01 AM
from the all-natural-language-processing dept.
from the all-natural-language-processing dept.
Chuck1318 writes "The Piraha tribe in the Amazon has only three words used in counting, that mean one, two, and many. A psychologist testing them has found that they are unable to accurately perform tasks involving quantities as few as four or five. He says that this shows that, at least for numbers, language shapes and limits how people can think." I can't help but be reminded of the gully dwarves from Dragonlance when reading this.
Related Stories
[+]
Amazonian Tribe Has No Word To Express Numbers 482 comments
In 2004 we discussed the Piraha, a tribe in the Amazon, when a study appeared characterizing their language as a "one, two, many" language. Now reader mu22le informs us of a new study of the Piraha pointing to the possibility that they use no number words at all. Instead they seem to use the word formerly thought to mean "two" to represent a quantity of 5 or 6, and the "one" word for anything from 1 to 4. The language has about 300 native speakers. "The study... offers evidence that number words are a concept invented by human cultures as they are needed, and not an inherent part of language, Gibson said."
This discussion has been archived.
No new comments can be posted.
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
Full
Abbreviated
Hidden
Loading... please wait.
Where have I heard this before? (Score:4, Informative)
[ George Gamow, "One, Two, Three...Infinity" 1953 ]
Re:Where have I heard this before? (Score:5, Informative)
Well yes, but if you read the article, it's not claiming to be a new theory, simply *proof* of an existing theory. From the article:
Parent
Babel-17 (Score:5, Interesting)
Parent
Re:Babel-17 (Score:5, Interesting)
It seems to me, as a layman and parent of two children, that the thoughts and ideas occur anyway regardless of the ability to express them in a language that can convey meaning to others.
Now, to throw another sci-fi reference into the mix, I seem to recall that Caesar, the ape that lead the revolution in the Planet of the Apes mythology, got his start with a single word: No.
Parent
Re:Babel-17 (Score:5, Insightful)
My 11 month old can understand a lot of words and commands, though she doesn't speak yet.
I have a two and a half year old who can tell you what she wants, and can understand nearly anything you might tell her. She can even express some abstract ideas "That's amazing", "This is fun".
It seems to me as a layman, that the only thoughts that occur naturally to children are "Feed me" and "make me comfortable again" (change diaper, make me warm, stop the thing that's hurting me). Everything else seems to be environmentally induced (most of the play I see in the 2.5 year old is mimicry of adult actions).
But evenso, I find it hard to grasp the concept of a language that goes to anything less than five-- because that's how many fingers you have and it seems to me that someone would want to count them sometime.
Also, personally, if I see a group of things, and it is five or less, I just know how many it is-- I don't have to consciously count them. Six though, I have to count. I do that by making two groups of three, so it's nearly instantaneous, but it is definitely not just "known".
On the other hand-- you only have two hands, so you'll rarely if ever manipulate more than two things at a time, so maybe that's it-- one, two, too many to do things with right now.
But for people who gather fruit and nuts... it seems like it would be a survival necessity to be able to tell the differnce between 4 cashews (I'm going to need to eat more) and 400 (I'm going to be so full).
And I think they can tell the difference. It seems, based on the article, that they just approximate volume. Because there is no need to tell the difference between 350 cashews and 400-- both of them will give a few people a snack. Similarly, who cares whether you have five avacodoes or six? That's a lot of avacadoes to eat.
Parent
Re:Chicken and Egg. (Score:5, Insightful)
Sorry, but that's just not the case. This happens to be my field - language is far more complicated than you might imagine. Linguistics, psycholinguistics, and visual cognition are not trivial just because you don't understand them on a serious level.
Parent
Re:Babel-17 (Score:5, Interesting)
It's an "it" to the French, too. It's not that the table 'has' gender - it's that the word belongs to a certain class of words that behave like it - that is, they're preceded by "la", "une", etc. In French, it's not "table", it's "la table" - the article is linked to the word itself (much like in English the infinitive is two words, but one idea).
In truth, it has nothing to do with the object itself. The French don't know why it's "la table" and not "le table", other than to tell you that it doesn't sound right as "le table".
The problem really comes because teachers like to teach it as if it really is confusing, and massively different from English, so you have to start seeing the gender in things. That's crazy. It's not different. It's just something you have to memorize, just like they have to memorize which adjectives you use "more", "most" with, and not "-er", "-est".
In truth, you can see the obvious bias in the study, as well. If your society has no language for counting above "two", then it likely has no need for counting above two, and so when presented with a situation where they need to count above two, they will be confused, because it's something they haven't done before. After all, when you're taught numbers, you're taught to count! So is it linguistic? I doubt it. I think the reverse is more true - thought (and society) shapes language.
I'd have to find out more about the study, but it seems really weak. You'd need a very careful control - that is, someone who lived in the same society as the Piraha, but spoke a different language that contained numbers higher than 2, and even that would be touchy because, as I said before, learning numbers above 2 means you were taught to count. But anyway, obviously no one like that exists, and I find it ludicrous that the psychologist made the leap "language directs thought" rather than "society directs language".
Parent
Re:Where have I heard this before? Whorf-Sapir ... (Score:5, Informative)
It has been discredited many times, as believable as it sounds. It is however a fascinating story; B.L. Whorf was an amateur linguist who was professionally a insurance claims inspector specialising in fire-related claims. He noticed that several fires where started when workers through cigarette butts into drums that in English we call "empty", even though they contained invisible and explosive fumes. Whorf realised that the workers knew this technically, but he wondered if being forced to think of the drums as "empty" changed their view of the drum. He did lots of research on languages of central america, and came up with interesting theories because many of these languages (eg Hopi) appear to have very different verb tenses; Whorf proposed that this gave their speakers almost-Einstein-like views of time and space.
A numbers of tests have been down over the years. Some languages have only a few words for color, for example. However, experiments show that this does not impair speakers of these languages from differentiating different shades of colors.
Parent
Re:Where have I heard this before? Whorf-Sapir ... (Score:5, Insightful)
I assume this is the same thing. Nomadic tribes don't deal with a lot of things, because everything they have they have to carry. So there's no need to count above two. If suddenly you ask a guy to keep track of four things, he's gonna have trouble: not because he doesn't have a word for it, but because he's going to have difficulty differentiating between the four things. It's no different than if I moved from driving a car to driving a semi trailer with no training. I'd get some of it, but important, non-intuitive concepts would be lost on me, and I'd probably crash. It's not because I don't have a word for them.
This is like the Inuit people and their umpteen words for snow. We outsiders can recognize the different types of snow with only a little practice, but since we don't get snow 8 months of the year, there's no need for it. English speakers understand foreign concepts like "esprit d'escalier" (the french term for all the cool things you wish you would have said when you leave somebody's house) or "bokeh" (the japanese term for the photographic effect that occurs with large aperatures in which the foreground is in sharp focus and the background is out of focus and fuzzy, thus drawing the eye towards the focus), even if we don't know what to call them.
It's experience that drives language, not vica versa -- althought the part of the brain that employees language is also responsible for the most critical human activity: symbolic logic.
Parent
Re:MOD PARENT UP (Score:5, Funny)
Parent
Re:MOD PARENT UP (Score:5, Insightful)
Parent
Re:Where have I heard this before? (Score:5, Interesting)
Which makes them smarter than Hottentot tribesmen....
Parent
Re:Where have I heard this before? (Score:5, Funny)
nope, it's easy...with crows, it'd be murder!
Parent
Re:Where have I heard this before? (Score:5, Interesting)
As a one-time owner of an African Grey, I can testify to this. I got my bird, Caesar, for my 11th birthday, and hand-fed him. Anyway, these days he lives with my grandparents (who have another African Grey) since it was too hard to keep him in a small New York apartment when my family moved here when I was 15 (I'm 25 now, so the bird is about 14 years old now, still a kid by Grey standards - they often live 50-60 years or even longer in captivity, sometimes as long as 70 or 80, barring illness. In fact, I'm pretty sure he'll outlive my grandparents and I'll end up with him again some day.
In any case, he had a vocabulary of at least 60-80 words when he was 3 or 4 years old. He exhibited exactly the kind of word combination that you reference - often semi-sensical, sometimes very amusing, sometimes scarily accurate and meaningful. They are fast to pick up on words or phrases, often times without a clear idea of what the words or phrases mean, but just as often they clearly DO associate meaning. "Caesar good boy", "Caesar good bird" or just "good bird" were often cooed out when he was feeling mellow after a meal. He seemed to take delight in yelling my name from across the apartment in my mother's voice to get my attention (they definitely learn names and associate them with people).
Interestingly, Greys have long memories - Caesar recognized my mother when she visited my grandparents in Florida recently even though he hadn't seen her in at least 3 or 4 years. The first thing he tried to do was regurgitate some food for her (yuck, but that's just their way of showing love).
I actually did a prize-winning middle school science project on Caesar, working on teaching him object differentiation skills, by color and shape, and associating them with words. He was pretty decent at simple object differentiation and fetching tasks administered verbally when you could get him to cooperate (he was less good at wanting to cooperate).
These birds can have AMAZINGLY strong personalities, be very willful and sometimes even nasty. Caesar was prone to losing his temper (okay, now I'm definitely ascribing human traits here, but he would have these fits of anger) and biting my fingers and ears. My fingers still bear the scars to this day. He was always timid or downright scared around strangers and could get nasty with even other less-favored family members who he saw every day, despite having been hand raised, lovingly treated, well fed and so on. He could also be very sweet and loving, desired affection, petting and human contact.
But Greys are the only animals I've ever seen capable of what I can only label "deceit". When a dog comes up to you and licks you, he wants to be petted, and if you pet him, he'll be happy. Caesar would sometimes play a nasty trick on people where he'd say "Rub my head" and cock his head like he wanted the attention. Somebody would slowly approach and gently extend their fingers to rub his head, then he's suddenly turn his head and take a big nip at their finger, usually accompanied by "OW! Stop That!" or "Bad bird!". I think it was mostly a way of getting more attention, which they do crave, but the effect was downright spooky coming from an animal.
Parent
yeah well, (Score:4, Funny)
funny but missing the point (Score:5, Interesting)
If this tribe calculated 0, 1, 2, many, many 1, many 2 or something like it there would be no trouble. Just confusing for base 10 users.
But it seems this tribe doesn't have/need the concept of higher numbers.
What I would like to know if they understand the concept of zero. The invention of 0 is a usually considered a pretty big step in western culture and one arabs like to claim as their contribution to the world. If this tribe wich can only count to 2 understands 0 then it would make an intresting find.
They may not have a need to count higher numbers but me thinks it is very important to know the difference between 1 fish and 0 fish.
What may also be intresting is that if you need language to count and animals can count does that mean that all animals that can count have a language. And not just a language of "food" "danger" "sex" but a language with "1" "2" "3" etc?
Parent
Re:funny but missing the point (Score:5, Informative)
Parent
Does this mean anything? (Score:5, Funny)
Perhaps they are not used to takss involving more than 3 items because usually it goes like this:
Hunt
Kill
Eat
Bang over head
Shag it
Sleep
Now I think some of thier ways of going about business is even more refined than ours.
Too Many replies (Score:5, Funny)
Re:Too Many replies (Score:4, Funny)
Parent
Sapir-Whorf (Score:5, Informative)
Re: Sapir-Whorf (Score:5, Interesting)
Oddly enough, that's pretty much how the word "Dutch" got into English: it's a corruption of "Deutsch" (one of the things Germans call themselves) -- the assumption at the time being all non-French continentals were the same people.
Back on topic, you can't take the "one -- two -- many" thing too far: almost every language shows at some stage of its development a "one two many" noun declension. Old English had specific dual endings; as did all the Germanic languages; dual was present in Proto Indo European and survived into most of the child languages.
Many semitic languages show vestiges of a 1 2 many number system (Arabic and Hebrew still retain a dual declension for some nouns). Swahili retains a separate noun class entirely for objects that come naturally in pairs (maono rather than *nyono, for instance).
I think all this points towards the fact that the distinction between one and two, and the distinction between two and many, is simply more important to people than the distinctions among various numbers greater than two, and that "one two many" is a natural linguistic response to the conditions of human life throughout most of human history -- people only develop more complex plural systems when agriculture and trade make it neccessary to develop them.
Parent
Obligatory discworld reference (Score:5, Funny)
"In fact, trolls traditionally count like this: one, two, three...many, and people assume this means they can have no grasp of higher numbers. They don't realize that many can be a number. As in: one, two, three, many, many-one, many-two, many-three, many many, many-many-one, many-many-two, many-many-three, many many many, many-many-many-one, many-many-many-two, many-many-many-three, LOTS.
Language is key (Score:5, Insightful)
I can only imagine that one in a completely different society would have a very different thought pattern. The common roots of Western languages indicates a similarity in thought, and people who learn foreign languages are far more adept at understanding and integrating with that society.
Similarily, in computer languages different grammatical structures lead different programmers to analyze and solve problems differently: i.e. functional vs imperative. Add the context-sensitive nature of human languages, and this becomes substantially more complex.
Ok, thats longer than my normal post, but this is a really interesting topic
Re:Language is key (Score:4, Interesting)
No, I can learn how to make a gun, plow a field, fetch water from a well from an Asian person with whom I have no common language - almost as easily as I could with an English speaking person.
I posted previously to this topic that it's all about the willingness of the people to learn,and the access to information that they are willing to subject themselves to is what forms thought and intelligence.
Parent
psych 101 (Score:5, Interesting)
No Orwell references? (Score:4, Insightful)
=Smidge=
superior language implies superiour thoughts? (Score:4, Interesting)
This makes one wonder if a another language would give us the ability to better reason about other things. Would we be smarter if we had a better language in which to think?
There is an artifical language called lojban (http://www.lojban.org/ [lojban.org]) based on predicate logic but which is meant to be used as other "real" languages (compare with eg. esperanto, interlingua and swahili). The question is, would native speakers of lojban be better a rational thought? As far as I know there are no native speakers of lojban but what would happend if I raised my (hypothethical) children to speak if from birth?
Mathias
Re:superior language implies superiour thoughts? (Score:5, Interesting)
That's the problem with this psychologist's study--it doesn't say whether or not they learned larger numbers and applied them effectively.
Parent
Personality depends on language, too (Score:5, Interesting)
Many fluently multilingual people will tell you that they are a slightly different person when they speak a different language.
I'm fluent in English and Japanese, and I can attest to this. In fact, there have been occasions when I was out of touch from Japanese speakers for a long time, and I began to miss my "Japanese self" because it hadn't had a chance to surface for so long.
So? (Score:5, Insightful)
But, even knowing that, is anything so dramatic going on? "Western" people with the proper training and experience could tell the difference at a glance between a screen full of C programming and a screen full of FORTRAN. My grandmother would struggle with that task. It would just all look like gibberish to her. Likewise, someone experienced in wine tasting can describe in detail the differences between two wines most of the rest us couldn't even tell apart.
A lot of what's necessary (or at least very helpful) in learning about programming or wines is the specialized language. When I'm told that the difference between two wines is that one is "fruitier" than the other, I've got something to look for. The nebulous and complex experience of tasting wine is brought into my understanding a little because I can now use a word to identify a part of what I'm sensing.
My point is, the idea that language affects how we think and what we perceive is not really all that novel.
There are also a lot of findings to the contrary (Score:5, Informative)
Example #1: Different languages divide up the color space differently. For example, Russian divides the color space covered by the English word "blue" into two separate color terms. However, language doesn't appear to affect the way people perceive color. For example, when researchers ask informants to judge color chips as "same" or "different," there appears to be no effect at all from the division of color space in the informant's native language.
Example #2: Chinese doesn't have a way of marking counterfactual or hypothetical statements as some languages do. One researcher (Bloom) had speakers of English and of Chinese read the same story in their respective native languages, and the speakers of Chinese did in fact have trouble answering whether such-and-such really happened. Bloom took this as evidence that language strongly affects thought. But another researcher said that the problem was just a bad translation into Chinese, and repeated the experiment with a better translation. Now the Chinese speakers had no difficulty saying "Of course such-and-such didn't happen."
On the other hand, the tense/aspect system of Russian does appear to have an effect on the way that speakers evaluate the temporal relationships in non-linguistic pictures of events. So it is occasionally possible to tease out a case where language does seem to have an effect on non-linguistic thought.
In sum, a blanket statement that "language determines thought" is much too strong. Even if the finding of the article mentioned above is accurate, the weight of the evidence seems to be that these cases are the exception, not the rule.
BTW: I'm sure that somewhere in this discussion, someone is going to bring up the idea that the Inuit (Eskimos) have some huge number of words for snow. That claim almost always gets trotted out in this kind of context. This is a kind of academic urban legend that just won't die. The linguist Geoff Pullum thoroughly debunked this whole fable some time back, and traced the series of misunderstandings and exaggerations which had given rise to it. In fact, it appears that Inuktitut has just two words for snow.
Sapir-Whorf refuted (Score:5, Interesting)
The Sapir-Whorf hypothesis, however, even though long proven wrong, has shaped the thinking of a whole generation of people, including those in the feminist movement, proposing "politically correct" words (female forms e.g.) hoping that they would induce a new thinking.
Language may be a result of knowledge and cultural concepts, thus reflect it. But it does not shape it, because - and that's known as de Saussure's work - the word is not equal to the concept. Whether you call something a small feline animal or a cat, it's still the same entity that you are thinking of. Whether you call someone a nigger, an african american, a 'brother', a black person -- the name does force us to change our thinking. (It may prompt us to think about misconceptions, of course!)
Steven Pinker's book "The Language Instinct" is a good read.
Haven't read Feigenson's original article. But it seems painfully obvious to me that given all the other linguistic evidence, the Brazilian tribe might simply have established a culture of arithmetics that doesn't allow you to count more than two things.
Parent
Russian colors (Score:5, Interesting)
So what I know is that in Polish there are also two words for blue: niebieski and blekitni. (Okay so I had to strip-off the accents because slashcode did not like them.) They are light-blue and dark-blue respectively. (Really niebieski is related to the word for sky so you might think of this word as sky-blue, I do and that is what I meant earlier about learning Polish from a linguist probably was different from a native Polish speakers experience.)
Now you might think this is simple, well not really. Here is a translation of what happens in practice with some regularity. My wife says, "Bring me the blue one," where blue is the word for either light-blue or dark-blue depending on the color of the object. I oblige but then hear a response of, "No I said the blue one not the green one." Bizarre because notice I wrote blue and green. It is not like she said light-blue and I brought the dark-blue widget. Sometimes she claims I brought the purple thing instead. These exchanges are entirely in Polish because this what we speak predominantly at home.
Okay now I am not color-blind. For my work I need to pass a test every two years and in the report I always pass all of the tests, even those for which a certain percentage of people that are not typically considered color-blind would not pass. I can clearly distinguish between a wide spectrum of colors.
After a while of this my mother noticed it once so we did a little test with the family. My mother, father, uncle, aunt, and grandmother were all part of it. All of them had spent the majority of their lives in Poland and almost without fail they would agree with the colors that my wife gave to objects. Then we repeated the test with my brother and his girlfriend who except for a vacation had not spent any time in Poland. They agreed with me the majority of the time.
Now this test was not scientific in any way and it did involve alcohol because it happened during a family get-together, but I still think that native Polish speakers vs English speakers think of colors as different because of their languages. What I mean is that there are many shades of colors that are sort of between green and blue and others that are between green and purple and given a proper ambiguous color such as this Polish speakers will tend to identify it differently than English speakers.
So what I am trying to say after all of this is that the example of the Russian language having two words for blue is sort of a red herring. It is irrelevant to the real issues. In fact given two people that are not color blind, one a Russian and one an English speaker, they should not have any extra difficulty in being able to distinguish between color chips as being different or not. What I am saying is that they will think of the same color chip as a different color in their minds. Now this is subtle, and I tend to agree with the parent poster that it is a special case, but definitely an example of how language influences understanding and meaning. Here is a final true story to illustrate this idea.
My wife's favorite color is light-blue. Once I bought her a gift that was a light blue dress. When she got it she said that the dress was nice, but that, "Don't you know by now that I do not like the way I look in green?" Think about intend and effect in that example and you will see what I mean about language being important.
Parent
Gully Dwarves (Score:5, Informative)
I ran a similar experiment once (Score:5, Funny)
Evaluate the Study (Score:5, Insightful)
To contrast, let us imagine that the Pirahã are conducting a similar study on a member of another culture. As this site is of the
Additionally, the Pirahã have a phrase in their language which indicates a degree of certainty, usually applied at the end of a sentence:
My point here should be fairly obvious. We cannot assume that we know the critical details of the study based upon a web article which, between two columns of advertisements, still only takes two pages (on my monitor, at least).
Second, and more breifly, the assumption that counting capacity defines intelligence is inherently flawed. The Pirahã have no need for counting; this is not to say they are not capable of it. Most Americans don't need to know what a coral snake looks like or that touching the little yellow-and-black frog is a bad thing. This doesn't mean they couldn't learn.
In summary, while the study definitely presents an interesting idea, one must evaluate it critically before accepting it as fact. Mistakes can be made.
That was a lot more than I meant to type. Thanks for the time.
Re:Discworld... (Score:4, Informative)
In fact, trolls traditionally count like this: one, two, three... many, and people assume this means they can have no grasp of higher numbers. They don't realize that many can be a number. As in: one, two, three, many, many-one, many-two, many-three, many many, many-many-one, many-many-two, many-many-three, many many many, many-many-many-one, many-many-many-two, many-many-many-three, LOTS.
- Men at Arms
Parent
Re:Inca's and Zero (Score:5, Funny)
Parent
INDIA (was Re:Inca's and Zero) (Score:4, Informative)
You see, in American English, you have only one word for Indians, unlike in other languages where they can actually tell the difference between Native Americans and the people who invented the decimal system, grammar, and many other useful things, like "Karma".
Parent
Re:INDIA (was Re:Inca's and Zero) (Score:5, Funny)
Parent
Re:INDIA (was Re:Inca's and Zero) (Score:5, Funny)
Parent
Re:I wanna be a "researcher" too. (Score:5, Informative)
were tested for practical skills involving
quantities of items or events larger than 3.
Parent
Actually, it is surprising (Score:5, Informative)
Parent
Re:Actually, it is surprising (Score:5, Interesting)
Without counting, or saying any number to yourself, or using a word to describe the concept of quantity, or referring to your fingers, tell me how many X's are on the next line:
X X X X X X
Think about it. You want to count them up to six, then remember "Six" and not remember " X X X X X X". This is the problem with the study. Language is, by definition, symbolic. That' the whole point of it, to not have to remember each experience in its totality, but to be able to share it symbolically with someone, so it is a) easier to remember and b) easier to transfer. Otherwise you'd be telling stories with models and pantomime. Now then, back to our experiment. How many were there? Draw it. See? Much harder than just saying "Six. I see Six X's" (you might have said "two groups of 3 X's", which you HAVE words for, but still, harder than "Six". The problem is that, as describe above, not that language affects how we think, but our vocabulary affects how we are able to recall and describe the world. You can still tell the difference between five X's and six X's... and you may even be able to build up "groupings" by using your own words for things. Does it mean you're not as smart as people who can describe "six"? No, you just are less able to recall and describe parts of the world you don't have words for.
I'm sure some day, aliens will come down and say to us "Electrons do not orbit nuclei, fools! Slithy toves gyre and gimbol in the wabe!" and then laugh into their tentacle-sleeves at us. (apologies to David Gribbin)
Parent
Re: Troll (Score:5, Funny)
> Or the trolls in Terry Pratchett's Discworld books
Finally, a
Parent
Re:Get those research grants! (Score:5, Funny)
2. Leave aformentioned slashdot poster in amazonian jungle with same level of technology as amazonian tribe s/he ridiculed.
3. Wait for slashdot poster to die in hostile environment which ridiculed tribe thrives in
4. Collect his/her life insurance
5. Profit!!
Parent
Re:Sapir-Whorf hypothesis (Score:4, Insightful)
Seems like no-one takes the Sapir-Whorf hypothesis serious these days, but I always thought it makes sense.
Why? It seems to me that all manner of spontaneous word creation (and outright theft from other languages) is hobbled if it were true. I mean, if thoughts of 0 or 3+ things were important to these people, they would have that thought long before they came up with a clean word to express it. As another poster joked, a computer isn't hobbled by only having 0 and 1 at its disposal. I think it is more correct to say that these people are not Turing-complete (for whatever reason) rather than blaming the language.
Parent
Re:Could it be? (Score:5, Funny)
Parent
Re:Could it be? (Score:5, Funny)
I would have thought it was more along the lines of
"In the land of the only-to-2-counting, the 3-counter is burnt at the stake for being a witch."
Parent