What is the nature of the relationship between thought and language?

April 10, 2006 5:38 PM

What is the nature of the relationship between thought and language?

Any answers - psychological, philosophical, or linguistic - are par for the course. Conjecture is welcome, though experimental evidence would be great too. I understand that this is not exactly answerable per se, and that many bright people have had much to say on the subject (Chomsky, Humboldt, Lacan... any other references would be appreciated). My question could be split up into several child questions:

Is (linguistic) representation a necessary condition for thought?
What is the evidence for and against the Sapir-Whorf hypothesis (strong and weak)?
Do monolinguistic speakers of analytic languages think differently than monolinguistic speakers of synthetic languages? More generally, do any morphological or syntactic differences among languages correlate with any differences in the form of thought (expressed in the philosophy native to that language [Heidegger to German for instance], in cultural norms, etc.)?
Does the structure of language parallel the structure of our thoughts (a la the Structuralists, and even some post-structuralists to a certain extent)?
Logographic vs. Phonetic?

Okay, I'm done.

posted by Frankieist to writing & language (30 comments total)

This has been brought up many times in AskMe. The wikipedia entry on Sapir-Whorf might be a good place to start. It's a common belief but I have never seen much evidence for it (but some evidence against).

I personally don't believe language influences thought other then as another source of input. For example, someone thinking about the color blue might start thinking about being sad if they were an English speaker, due to the linguistic coincidence that the words are the same, but they also might not. They don't really associate the color with the emotion.
posted by delmoi at 5:46 PM on April 10, 2006

Sapir-Whorf is a great albeit controversial theory.

Just today I overheard people complaining about a cassette player because the < button fast-forwarded the tape and the>> button rewound it. It works that way because when you fast-forward a tape, the tape's insides move to the left. But because english speakers read left to right, when something points left, they think "back," therefore rewind. This doesn't apply exactly because writing is something entirely seperate from language, but it reflects the kind of cultural consciousness that results from a particular sort of language.
posted by BuddhaInABucket at 5:58 PM on April 10, 2006 [1 favorite]

There's evidence against Sapir-Whorf for color perception, at least. The S-W Hypothesis would say that people might have trouble perceiving colors they don't have a name for. But what we find is that people of all cultures have comparable ability to distinguish colors. Also, if a culture has only two color words, they will correspond to black and white. If a culture has three color words, the third is red. And so on, down a universal ordered list.

The SWH is a pretty awesome insight, but it wasn't really based on science. Whorf was a fireman when he came up with it. But I like to think that it's still true in certain situations. It may not be true for areas where there are definite concrete perceptual analogues for words (like colors), but what about words that are more abstract? What about words which stand for something that may or may not actually exist, like "the soul" or "the state"?
posted by Laugh_track at 6:11 PM on April 10, 2006

There's this recent article from the straight dope. Fascinating subject. I must go do a lot more reading
posted by moonshine at 6:14 PM on April 10, 2006

You might check out Lera Boroditsky, who studies this sort of thing.


One of her papers (pdf)
posted by logicpunk at 6:24 PM on April 10, 2006

linguistic relativity hypothesis (a supplement to relativism)
language of thought hypothesis
private language and wittgenstein
mental imagery
quotation

the stanford encylopedia of philosophy (1) kicks ass and (2) needs your support
posted by andrew cooke at 6:29 PM on April 10, 2006

Here's the anecdote that regained my interest in the subject:

It was after having read George Lakoff, and his notion that abstract thought is (always?) mediated through metaphor (of more tangible things) struck me. I don't know why I noticed it, but there are many different modifiers of the word "stand" that represent relatively abstract things: a stand-up citizen, standing one's ground, one's standing in society, one's stance on an issue. Incidentally, these are all meanings of the (ancient) Greek word stasis, according to LSJ lexicon. After looking more into it, it seems that all of these were derived from the PIE root "sta-", meaning "stand" or "to stand." Instance, substance, resistance (standing-up for sthng), distance, etc. are all, according to OED, derived from "sta-". Now, to me, these prefixes to the word "stance" don't seem to have any logical connection to the change in the meaning of the word. Even estar and many Indo-European "to be" verbs are supposed to be derived from sta-.

So I wondered, are these vestigial structures in language, or does this reflect a certain characteristic of cognition?
posted by Frankieist at 6:31 PM on April 10, 2006

actually, not certain about "resistance" but everything else above stands, I believe (no pun intended).
posted by Frankieist at 6:49 PM on April 10, 2006

It works that way because when you fast-forward a tape, the tape's insides move to the left.

What, was the tape upside-down or something?
posted by kindall at 6:59 PM on April 10, 2006

I'm a total amateur, and there's a great deal of formal scientific inquiry being linked to here, so treat this as a layman's opinion at best.

Back when I was taking martial arts, I noticed a couple of things. First, the mind is extraordinarily intelligent without using words at all. It is perfectly possible to have a problem, go and get the tool or device you need to fix the problem, and repair it, without ever thinking about it consciously. And that kind of thinking is incredibly fast. It's more than just reflex, it's actual intelligence. I've lived in my rational mind most of my life, and it was quite frightening at first to realize just how smart and how fast that nonverbal mind is. Conscious thought is at least an order of magnitude slower, and I'm not sure it's that much smarter.

Second: it is the nature of the rational mind to ascribe all things to itself, particularly the decisions made by the nonverbal mind. The rational mind tells stories, and often it claims responsibility for decisions that it didn't make at all. It doesn't seem to be able to understand that it's not turned on all the time.

The rational mind is simultaneously our single greatest tool and a huge set of blinders. Rational thought is the process of ignoring nearly everything that's true about a given thing or set of things, in order to manipulate a subset of symbols about them. My particular breakthrough in understanding that was looking at a flashlight after a sleepless night.... I thought, "_why_ is this a flashlight?". (the kind of dumb question you have to be extremely tired to even think of). And in that moment, I realized that it was a flashlight because I _called_ it one. If I took the top off and emptied it, I could call it a drinking cup; if I filled it with dirt, it was a planter; if I stuck it in someone's tailpipe, it was a bomb. And when I was seeing it as a flashlight, I couldn't see it as a planter; and when it was being a planter, it couldn't be a flashlight. The thing itself changed not a bit, just my perception of it.

Ever since then, I've been a little distrustful of hyper-intellectual pursuits, because I realized that they are, in many cases, castles built on foundations of sand. Rational thought is _incredibly_ useful, but useful and _true_ are not the same thing. Thinking about things means ignoring almost everything about them.

Most books that I have read about thought have looked at it from the standpoint of the rational mind. One book, rather ludicrous in most respects, but good in this one analogy, summed it up this way: Using the rational mind to examine the complete thought process is like being a flashlight and trying to study the dark.
posted by Malor at 7:07 PM on April 10, 2006

Languagelog had a series of great posts about this, such as this one. Search the archives for "sapir-whorf" for more.
posted by Brian James at 7:11 PM on April 10, 2006

If SW is true, how is it that we can and do create new words, to describe new concepts for which we previously had no words?

For me, the big problem with SW is bootstrapping: if our thinking is limited by our language, then that requires that language must have existed as long as we have. SW seems to force us to become creationists, with humans coming into being with language already in place.
posted by Steven C. Den Beste at 8:43 PM on April 10, 2006

My interest in this very subject has been spurred recently by reading about the Piraha tribe of the Brazilian Amazon.This small group of people is quite unique in having no words to express many abstract concepts we find fundamental, such as numbers, colour, past and future tenses. The arguement rages over the implications of this for Sapir-Whorf, but it certainly gives you a lot to think about.

Links: Mefi, Mefi, Language Log, Whyfiles.

William Burroughs has a lot to say about this matter, most (in)famously his theory of language as a virus.

"My basic theory is that the written word was literally a virus that made spoken word possible. The word has not been recognized as a virus because it has achieved a state of stable symbiosis with the host...(This symbiotic relationship is now breaking down for reasons I will suggest later.)" From The Electronic Revolution

I cannot find a decent summary of Burroughs' ideas, but the above link is a starting point. I also found this essay, which looks interesting from skim-reading, but also rather heavy. The ideas may seem extreame, but there is method in his madness. I suspect the 'meme' concept may have been influenced by Burroughs' theories, for example.

The fusion anomaly language node is thought-provoking.

I have some of ideas of my own, but will restrain myself from ranting sharing here until I can get the thoughts into some sort useful arrangement.
posted by MetaMonkey at 8:45 PM on April 10, 2006

The interesting findings in the SW tradition have little to do with lexicalized concepts. Any determining influence from language on perception or thought (two different things, to some extent) is very likely mediated by syntax and other non-lexical modules of grammar. This is not a subject for common sense debate, it is scientificially specialized. Opinions don't really matter.
posted by fourcheesemac at 9:17 PM on April 10, 2006

Steven C. Den Beste : "if our thinking is limited by our language, then that requires that language must have existed as long as we have."

Change 'limited' to 'shaped' and the objection disappears.
posted by Gyan at 10:08 PM on April 10, 2006

I see my answer was deleted because a mod may have thought that my Star Trek reference was a joke.
(the aliens with the big heads that don't need vocal communication)

So I'll try again.

Human thought can be larger than the universe but our capacity to communicate thought's is limited by about 5,000 + words in the English language.

Therefore, the deepest intellectual thoughts must be thought out twice to communicate them concisely within the boundaries of human speech.

I thought I had communicated that with my glib remark but apparantly I had a failure to communicate with the MeFi mod.
posted by SwingingJohnson1968 at 10:53 PM on April 10, 2006

try 1,000,000 words.
posted by Frankieist at 12:47 AM on April 11, 2006

Granted, the average english speaker contains a vocabulary between 10,000 and 20,000 words.
posted by Frankieist at 12:49 AM on April 11, 2006

Maybe you will be interested in this recent question about symbols and understanding, if you didn't see it.
posted by teleskiving at 1:32 AM on April 11, 2006

try 1,000,000 words.
posted by Frankieist at 12:47 AM PST on April 11 [!]


Granted, the average english speaker contains a vocabulary between 10,000 and 20,000 words.
posted by Frankieist at 12:49 AM PST on April 11 [!]

There are really only about 5,000 words that are ever used in English in every day conversations.

Even the deep conversations.
posted by SwingingJohnson1968 at 2:01 AM on April 11, 2006

My attempt from a few years back to work through these ideas, in context of features of the Korean language, is here, for what it's worth.
posted by stavrosthewonderchicken at 2:39 AM on April 11, 2006

You should definitely read Daniel Everett's Cultural Constraints on Grammar and Cognition in Pirahã: Another Look at the Design Features of Human Language (PDF link). Features of the Pirahã language include:

* One of the smallest phoneme inventories of any known language (perhaps surpassed only by Rotokas), and a correspondingly high degree of allophonic variation, including two very rare sounds, [ɺ͡ɺ̼] and [t͡ʙ̥].
* The pronunciation of several phonemes depends on the speaker's sex.
* An extremely limited clause structure.
* No grammatical numerals, not even "one" or "two"; the closest the language comes to numerals are general quantity words like 'many'.
* No abstract color words other than terms for light and dark.
* Few specific kin terms; one word covers both "father" and "mother".
* The personal pronouns (and seemingly no other words) may have been borrowed from an unrelated Tupian language. Many linguists, however, find this claim questionable, noting that there is no historical-comparative evidence indicating the non-existence of pronouns in a previous period of the history of Pirahã; also, the overall lack of Tupi-Guarani loanwords in areas of the lexicon more susceptible to borrowing (such as nouns referring to cultural items, for instance) makes this hypothesis even less plausible.
* Pirahã can be whistled, hummed, or encoded in music.
(from Wikipedia)

In support of S-W, the tribe's lack of numerals seems to impede their ability to count, unable to even grasp putting objects in a one-to-one correspondence. See this article from the Economist and this one from ScienceDaily, both explaining some of the tests the Pirahã underwent, and scroll down on this page for an enlightening video of one lesson.

More on the phonetics of Pirahã
Also, some drawings by the Pirahã.
posted by youarenothere at 5:41 AM on April 11, 2006

Interesting clip of Piraha sung speech (MOV link)
posted by youarenothere at 5:52 AM on April 11, 2006

youarenothere: Watching this movie [.avi] (supposedly documenting the inability of a Pirahã gentleman to accurately reproduce a sequence of knocks), I wonder how much of this is due to a difference of "experience." Is it really the lack of number-words that impedes the man's ability to knock on the wall a correct number of times? Or is it that there is a misunderstanding of intention? What has he been asked to do by the "knocker"? Repeat the "same" number of knocks? The word "same" has different meanings in different cultures, especially in oral cultures where exact repetition may not be valued or meaningful. Could it be that he has never had to perform such a task?

Many languages share the features you have listed above and the speakers seem to be able to exist in their environments quite well. And as has already been mentioned (in regards to color differentiation), the absence of vocabulary does not inhibit perception or even categorization.
posted by imposster at 8:06 AM on April 11, 2006

According to Lawrence Barsalou, words cannot be separated from our perceived representations of the world which are modal, schematic, and productive (through simulation.)

"Perceptual symbol systems," BEHAVIORAL AND BRAIN SCIENCES (1999) 22

"Like a perceptual symbol, a linguistic symbol is a schematic memory of a perceived event, where the perceived event is a spoken or a written word. ... As selective attention focuses on spoken and written words, schematic memories extracted from perceptual states become integrated into simulators that later produce simulations of these words in recognition, imagination, and production."
posted by imposster at 8:27 AM on April 11, 2006

Numbers and numeracy are much more abstract ideas than color, which is rooted in senory experience. We are born with the ability to percieve 1, 2, and 3 (subitizing), and the experiments with the Piraha fell in line with this, despite the fact that they don't have precise words for those numbers. The ability of the Piraha to accomplish many of tasks for numbers up to three proves, I think, the fact that they understood the intention behind the tasks. However, without words or some other symbolic representation of exact quantities, numeracy - along with the Piraha's ability to complete most of the tasks put before them - stops. As Peter Gordon put it, "the practice of counting is inextricably entwined with the words (or signs) we use for number."
posted by youarenothere at 8:50 AM on April 11, 2006

A few anecdotes:

(1) I lived with some Susu and Fulani people in Guinea. Fula is a pretty complicated language with a very rich vocabulary. Susu not so much. However, it does have an amazing depth of vocabulary in three areas. (Probably more, but those were the three I noticed.) Sex, snakes, and rice. They have words to express certain snakes, and certain stages of rice production, that I would never think and would never need a word for. I believe they have more complex thoughts about snakes and rice anyway, then I am capable of having, based on my experiences and the paucity of English words on the subject.

(2) Both susu and fula have a pronoun that refers to the exclusive we (meaning "my friend and I, not you") and an inclusive we (meaning "me friend, you and I"). "Muxu" in susu and "men" in fula if I remember right. It's lovely, bc there aren't any awkward misunderstandings when someone says something like "we're going to lunch now." It's obvious whether you're invited or not. Obviously that is a concept we can think in English, but we don't really have a word for it.

Anyway, my conclusion drawn from purely anecdotal evidence is that there are some ways in which thought influences language - e.g. susu ppl think about sex, snakes and rice a lot, and consequently have a lot of words for those subjects - and other ways in which thought doesn't influence language as much as might be expected.
posted by Amizu at 9:49 AM on April 11, 2006

This is one of those subjects (like the Meaning of Life) that's fun to talk about, especially over beers, but for which there is insufficient evidence to draw firm conclusions, and probably always will be. I've talked about it enough to be bored with the whole thing, so I'll just say that Amizu summed it up pretty well:

there are some ways in which thought influences language... and other ways in which thought doesn't influence language as much as might be expected.
posted by languagehat at 11:12 AM on April 11, 2006

For what its worth, I've observed in my own life that vocabulary and depth of language always accompanied deeper understanding of things. I'm not sure if one produces the other, but the ability to encapsulate ideas in words allowed further thoughts and understanding in many subjects.
posted by walljm at 12:04 PM on April 11, 2006

If I could add my two cents (albeit an utterly novice and mostly philosophical two cents)...

Thought On Language:
The downright inability of language to fully express all of the thoughts & emotions we, as humans, experience is just one of the many insight I’ve gained from my experience (discussed below).

Some (Weak) Evidence:
Language is only one way in which we communicate what’s going on “inside” of us. One only need look at the millions of musicians, chefs, dancers and artists to see their brilliance; especially that of Helen Keller.

The Roots of These Thoughts:
While in college I was tested by a neurologist who claimed that I showed signs of a language based learning disability (I think it's a bogus diagnosis based on my history of high standardized test scores & the early age I began reading at...toot toot - tooting my own horn :o). Although, throughout my schooling, I have consistently had problems with non-standardized tests (I feel this is based on poorly worded questions). One particular question I recall from the neurological test is as follows: Paraphrase this sentence; "The boy saw a mouse running in his pajamas" (stated verbally). There must be at least 3 or 4 ways to interpret this - see if you can find them - but apparently only one is 'correct' and won't label you as 'disabled.'

Conclusion:
Language is a beautiful, but limiting tool. This gift not only enhances the precious moments we get to spend with other human beings, but provides us with a more effective method to share our knowledge. Unfortunately, no part of our physical existence can completely do our thoughts & experiences the justice they deserve.

One thing is for sure, without language, we certainly wouldn’t be having this fascinating discussion. However, I bet it would make a pretty picture!

Thanks for a very interesting thread & good luck. Sorry if this was a little too off topic for you
posted by Jhaus at 8:38 PM on April 11, 2006