32 ms·

Language may shape human thought

44 points by b-man 15y ago

DanielRibeiro 15y ago
Sapir Whorf all over again?[1] [1] http://en.wikipedia.org/wiki/Sapir-Whorf_hypothesis#Computer_coding_language_conceptual_correlate http://en.wikipedia.org/wiki/Sapir-Whorf_hypothesis#Computer...
- jensnockert 15y ago
  Yes, and I think most programmers have profoundly experienced the effect, you are not thinking the same way before and after learning a new programming paradigm. You don't think the same way after learning LISP, Erlang, or...
  - timinman 15y ago
    Cross-reference this Google Tech Talk by Matz, the creator Ruby: http://goo.gl/rAxIL http://goo.gl/rAxIL.
  - DanielRibeiro 15y ago
    Precisely the arguments of many posts about programming: * http://brandonbyars.com/2008/05/13/beating-sapir-whorf/ http://brandonbyars.com/2008/05/13/beating-sapir-whorf/ * The creator of ruby on the "The power and philosophy of ruby: or how to create Babel-17" [1][2][3] However these usually imply the weak version of Sapir Whorf, not the strong one. [1] http://www.rubyist.net/~matz/slides/oscon2003/mgp00001.html http://www.rubyist.net/~matz/slides/oscon2003/mgp00001.html [2] http://en.wikipedia.org/wiki/Linguistic_relativity#cite_note-48 http://en.wikipedia.org/wiki/Linguistic_relativity#cite_note... [3]http://en.wikipedia.org/wiki/Linguistic_relativity#cite_ref-48 http://en.wikipedia.org/wiki/Linguistic_relativity#cite_ref-...
    - wisty 15y ago
      Didn't Einstein say something like "all mathematical advances are notational"?
  - DasIch 15y ago
    There is a difference between programming paradigms and programming languages.
- pnathan 15y ago
  I know. It's kinda tiresome hearing these old claims trotted out as "AMAZING NEW THEORY". As a note, the initial Sapir-Whorf hypothesis is considered to be based off of flawed research.
flashmob 15y ago
How many people did they test this on? one, two or many?
- rflrob 15y ago
  Many: "Gordon set seven Pirahã a variety of tasks."
rednum 15y ago
I remember reading about similar phenomenon concerning colours - people who have one word for red and purple in their language treat them as same colour; I can't find a link to those information provided in concise manner on this topic. However, I think both of these facts are smaller than the title could suggest - though it seems that usually we put 'things' we think about into buckets (so if you put both 4 stones and 5 stones into bucket 'many' you can't distinguish between them) labelled by words from language we use, it gives us little insight on how language shapes our cognitive algorithms, which is much more interesting question to me.
- Jun8 15y ago
  I think you are referring to the famous Berlin and Kay study (here's a chapter detailing it: http://www.icsi.berkeley.edu/~kay/KBMM.ps http://www.icsi.berkeley.edu/~kay/KBMM.ps).
heydenberk 15y ago
It tickles me the way this title blithely summarizes a great deal of 20th century postmodern philosophy.
- bchjam 15y ago
  well, maybe
tokenadult 15y ago
From the article: "Gordon says this is the first convincing evidence that a language lacking words for certain concepts could actually prevent speakers of the language from understanding those concepts." I've studied multiple different human languages over the years, and have also studied the methods of linguistics fieldwork used when scientists encounter previously unknown languages. I find it interesting that the article says Gordon's work is the "first convincing evidence" for strong linguistic relativity, as the strong linguistic relativity claim is a claim that has been made over and over again, so I guess the previous claims are mostly doubted. I doubt the claim here, as I doubt the previous claims. Peter Gordon has spent most of his career attempting to produce publications from his field work with this tiny tribe, but he hasn't even convinced other colleagues who do field work there. http://en.wikipedia.org/wiki/Pirah%C3%A3_people#Language http://en.wikipedia.org/wiki/Pirah%C3%A3_people#Language "Some researchers, such as Prof. Peter Gordon of Columbia University, claim that the Pirahã are incapable of learning numeracy. His colleague, Prof. Daniel L. Everett, on the other hand, argues that the Pirahã are cognitively capable of counting; they simply choose not to do so. They believe that their culture is complete and does not need anything from outside cultures." If the scientists and the subjects start out communicating using mostly gestures, and perhaps a language that one side or the other had acquired only poorly as a second language, how does either side know the full range of communication available in the language of the other? (This is one of the defects of much cross-cultural research in linguistics--something is declared to be "impossible" for speakers of one language by investigators who didn't grow up in a community of native speakers of that language.) Here's a link to a PowerPoint describing the study: http://ling.umd.edu/~zukowski/courses/Spring2005/hon218L/Chaz-Piraha.pdf http://ling.umd.edu/~zukowski/courses/Spring2005/hon218L/Cha... People do learn from engaging in new activities--including speaking new languages--and also from entering new environments while continuing to use their native languages. As always in such studies, the difficult task is disentangling what people learn from a new language and what people learn from the environment that surrounds them, whatever the language. Some forms of learning produce "imprinting," which then prevents highly malleable learning on the same issue, but it's not clear that any human language does much to constrain thinking about truly universal concepts of human experience. The submitted article was submitted here with the (tentative) original title. An even more cautious title might be "Researcher who has devoted his career to obscure tribe thinks that the tribe is proof of a controversial concept, but his colleague disagrees."
majmun 15y ago
i always tought that it is other way around, that thoutgh shape language. maybe my language was not good enough.
bchjam 15y ago
http://www.edge.org/3rd_culture/boroditsky09/boroditsky09_index.html http://www.edge.org/3rd_culture/boroditsky09/boroditsky09_in... is a good article on how language can/does affect thought
latj 15y ago
I believe this man went to brazil and then came back and presented the same idea fifty years ago: http://en.wikipedia.org/wiki/Napoleon_Chagnon http://en.wikipedia.org/wiki/Napoleon_Chagnon
ern 15y ago
This seems to be the original paper from 2004: http://languagelog.ldc.upenn.edu/myl/surprise/llog/Gordon_Piraha.pdf http://languagelog.ldc.upenn.edu/myl/surprise/llog/Gordon_Pi...
neovive 15y ago
I wonder how this correlates to children growing up in bilingual households. Does access to multiple language alter the thought processes. Do they learn to think in one language vs. the other or do they think in both languages depending on the task.
Jun8 15y ago
Like Strong and Weak AI, there are stronger and weaker versions of the so-called Sapir-Whorf hypothesis (if you dig their writings, you'll see that neither Sapir nor Whorf didn't really say most things attributed to them, and definitely they did not propose the strong version, i.e. "language shapes thought". Up until recently, c. 90s, Whorf was often derided (and ridiculed) in most linguistic circles as being an amateur or worse, a nut job (he was a chemical engineer by profession), e.g. see Pinker's The Language Instinct. This was because of the prevalence of universal grammar idea of the Chomsky school, that held sway in Linguistics, especially in the US. Only recently researchers have started to revisit the idea thoroughly and gave it its due merit (for an early, ~mid 80s, and influential debate on the subject, read this paper about the debate between Bloom and Au about whether the lack of counterfactuals in Chinese hampers their analysis of complex counterfactual sentences http://groups.psych.northwestern.edu/gentner/papers/YehGentner05.pdf http://groups.psych.northwestern.edu/gentner/papers/YehGentn...). People resist the Whorf hypothesis from a purely political grounds, too, thinking that accepting it would lead to cultural relativism. That being said, although the idea seems intuitively correct, there are difficulties when you start to think more. The analogy with programming languages goes only so far, there you are trying to translate a task, described in your native language, e.g. language, into a programming language. If language limits thought, the expressions of thought should be done (partly) in a non-linguistic and more rich way. Not very many linguistics, cognitive scientists, or philosophers would take this view, I think, but there are some who do, e.g. Fodor, who proposes that thought are expressed in a "language of thought" distinct from native language. This is a fascinating subject, if you have a few hours (or days) to sink, check out the Language of Thought entry of the Stanford Encyclopedia of Philosophy: http://plato.stanford.edu/entries/language-thought/ http://plato.stanford.edu/entries/language-thought/
peregrine 15y ago
Fascinating topic and RadioLab did a really nice radio program on this here. http://www.radiolab.org/2010/aug/09/ http://www.radiolab.org/2010/aug/09/
totalc 15y ago
This story has been circulating the internets for years since the original article. Has there been additional research on the Pirahã?
dragoner 15y ago
I always that its the other way, human thought determines the language
ThaddeusQuay 15y ago
Language DOES shape human thought, as anyone who knows APL can tell you. "Language as an Intellectual Tool: From Hieroglyphics to APL" by Donald B. McIntyre in IBM Systems Journal Vol.30, No.4 1991, p.554-581. http://www.electricscotland.com/mcintyre/index_f/menu_f/j_f/sysj_r_29.pdf http://www.electricscotland.com/mcintyre/index_f/menu_f/j_f/... "Notation as a Tool of Thought" by Kenneth E. Iverson in his 1979 ACM Turing Award Lecture. http://elliscave.com/APL_J/tool.pdf http://elliscave.com/APL_J/tool.pdf
Typhon 15y ago
Maybe language shapes thought. Or maybe thought shapes language. Or maybe they both influence each other. Actually, that last one, I think, is the most likely, since it's less simplistic than the previous two. Now, much has been written about the Piraha, and whether or not studying their culture will revolutionize linguistics. See this very interesting link http://www.edge.org/documents/archive/edge213.html http://www.edge.org/documents/archive/edge213.html

44 points by b-man 15y ago

Sapir Whorf all over again?[1] [1] http://en.wikipedia.org/wiki/Sapir-Whorf_hypothesis#Computer_coding_language_conceptual_correlate http://en.wikipedia.org/wiki/Sapir-Whorf_hypothesis#Computer...

jensnockert 15y ago
Yes, and I think most programmers have profoundly experienced the effect, you are not thinking the same way before and after learning a new programming paradigm. You don't think the same way after learning LISP, Erlang, or...
- timinman 15y ago
  Cross-reference this Google Tech Talk by Matz, the creator Ruby: http://goo.gl/rAxIL http://goo.gl/rAxIL.
- DanielRibeiro 15y ago
  Precisely the arguments of many posts about programming: * http://brandonbyars.com/2008/05/13/beating-sapir-whorf/ http://brandonbyars.com/2008/05/13/beating-sapir-whorf/ * The creator of ruby on the "The power and philosophy of ruby: or how to create Babel-17" [1][2][3] However these usually imply the weak version of Sapir Whorf, not the strong one. [1] http://www.rubyist.net/~matz/slides/oscon2003/mgp00001.html http://www.rubyist.net/~matz/slides/oscon2003/mgp00001.html [2] http://en.wikipedia.org/wiki/Linguistic_relativity#cite_note-48 http://en.wikipedia.org/wiki/Linguistic_relativity#cite_note... [3]http://en.wikipedia.org/wiki/Linguistic_relativity#cite_ref-48 http://en.wikipedia.org/wiki/Linguistic_relativity#cite_ref-...
  - wisty 15y ago
    Didn't Einstein say something like "all mathematical advances are notational"?
- DasIch 15y ago
  There is a difference between programming paradigms and programming languages.
pnathan 15y ago
I know. It's kinda tiresome hearing these old claims trotted out as "AMAZING NEW THEORY". As a note, the initial Sapir-Whorf hypothesis is considered to be based off of flawed research.

flashmob 15y ago

How many people did they test this on? one, two or many?

rflrob 15y ago
Many: "Gordon set seven Pirahã a variety of tasks."

rednum 15y ago

I remember reading about similar phenomenon concerning colours - people who have one word for red and purple in their language treat them as same colour; I can't find a link to those information provided in concise manner on this topic. However, I think both of these facts are smaller than the title could suggest - though it seems that usually we put 'things' we think about into buckets (so if you put both 4 stones and 5 stones into bucket 'many' you can't distinguish between them) labelled by words from language we use, it gives us little insight on how language shapes our cognitive algorithms, which is much more interesting question to me.

Jun8 15y ago
I think you are referring to the famous Berlin and Kay study (here's a chapter detailing it: http://www.icsi.berkeley.edu/~kay/KBMM.ps http://www.icsi.berkeley.edu/~kay/KBMM.ps).

heydenberk 15y ago

It tickles me the way this title blithely summarizes a great deal of 20th century postmodern philosophy.

bchjam 15y ago
well, maybe

tokenadult 15y ago

From the article: "Gordon says this is the first convincing evidence that a language lacking words for certain concepts could actually prevent speakers of the language from understanding those concepts." I've studied multiple different human languages over the years, and have also studied the methods of linguistics fieldwork used when scientists encounter previously unknown languages. I find it interesting that the article says Gordon's work is the "first convincing evidence" for strong linguistic relativity, as the strong linguistic relativity claim is a claim that has been made over and over again, so I guess the previous claims are mostly doubted. I doubt the claim here, as I doubt the previous claims. Peter Gordon has spent most of his career attempting to produce publications from his field work with this tiny tribe, but he hasn't even convinced other colleagues who do field work there. http://en.wikipedia.org/wiki/Pirah%C3%A3_people#Language http://en.wikipedia.org/wiki/Pirah%C3%A3_people#Language "Some researchers, such as Prof. Peter Gordon of Columbia University, claim that the Pirahã are incapable of learning numeracy. His colleague, Prof. Daniel L. Everett, on the other hand, argues that the Pirahã are cognitively capable of counting; they simply choose not to do so. They believe that their culture is complete and does not need anything from outside cultures." If the scientists and the subjects start out communicating using mostly gestures, and perhaps a language that one side or the other had acquired only poorly as a second language, how does either side know the full range of communication available in the language of the other? (This is one of the defects of much cross-cultural research in linguistics--something is declared to be "impossible" for speakers of one language by investigators who didn't grow up in a community of native speakers of that language.) Here's a link to a PowerPoint describing the study: http://ling.umd.edu/~zukowski/courses/Spring2005/hon218L/Chaz-Piraha.pdf http://ling.umd.edu/~zukowski/courses/Spring2005/hon218L/Cha... People do learn from engaging in new activities--including speaking new languages--and also from entering new environments while continuing to use their native languages. As always in such studies, the difficult task is disentangling what people learn from a new language and what people learn from the environment that surrounds them, whatever the language. Some forms of learning produce "imprinting," which then prevents highly malleable learning on the same issue, but it's not clear that any human language does much to constrain thinking about truly universal concepts of human experience. The submitted article was submitted here with the (tentative) original title. An even more cautious title might be "Researcher who has devoted his career to obscure tribe thinks that the tribe is proof of a controversial concept, but his colleague disagrees."

majmun 15y ago

i always tought that it is other way around, that thoutgh shape language. maybe my language was not good enough.

bchjam 15y ago

http://www.edge.org/3rd_culture/boroditsky09/boroditsky09_index.html http://www.edge.org/3rd_culture/boroditsky09/boroditsky09_in... is a good article on how language can/does affect thought

latj 15y ago

I believe this man went to brazil and then came back and presented the same idea fifty years ago: http://en.wikipedia.org/wiki/Napoleon_Chagnon http://en.wikipedia.org/wiki/Napoleon_Chagnon

ern 15y ago

This seems to be the original paper from 2004: http://languagelog.ldc.upenn.edu/myl/surprise/llog/Gordon_Piraha.pdf http://languagelog.ldc.upenn.edu/myl/surprise/llog/Gordon_Pi...

neovive 15y ago

I wonder how this correlates to children growing up in bilingual households. Does access to multiple language alter the thought processes. Do they learn to think in one language vs. the other or do they think in both languages depending on the task.

Jun8 15y ago

Like Strong and Weak AI, there are stronger and weaker versions of the so-called Sapir-Whorf hypothesis (if you dig their writings, you'll see that neither Sapir nor Whorf didn't really say most things attributed to them, and definitely they did not propose the strong version, i.e. "language shapes thought". Up until recently, c. 90s, Whorf was often derided (and ridiculed) in most linguistic circles as being an amateur or worse, a nut job (he was a chemical engineer by profession), e.g. see Pinker's The Language Instinct. This was because of the prevalence of universal grammar idea of the Chomsky school, that held sway in Linguistics, especially in the US. Only recently researchers have started to revisit the idea thoroughly and gave it its due merit (for an early, ~mid 80s, and influential debate on the subject, read this paper about the debate between Bloom and Au about whether the lack of counterfactuals in Chinese hampers their analysis of complex counterfactual sentences http://groups.psych.northwestern.edu/gentner/papers/YehGentner05.pdf http://groups.psych.northwestern.edu/gentner/papers/YehGentn...). People resist the Whorf hypothesis from a purely political grounds, too, thinking that accepting it would lead to cultural relativism. That being said, although the idea seems intuitively correct, there are difficulties when you start to think more. The analogy with programming languages goes only so far, there you are trying to translate a task, described in your native language, e.g. language, into a programming language. If language limits thought, the expressions of thought should be done (partly) in a non-linguistic and more rich way. Not very many linguistics, cognitive scientists, or philosophers would take this view, I think, but there are some who do, e.g. Fodor, who proposes that thought are expressed in a "language of thought" distinct from native language. This is a fascinating subject, if you have a few hours (or days) to sink, check out the Language of Thought entry of the Stanford Encyclopedia of Philosophy: http://plato.stanford.edu/entries/language-thought/ http://plato.stanford.edu/entries/language-thought/

peregrine 15y ago

Fascinating topic and RadioLab did a really nice radio program on this here. http://www.radiolab.org/2010/aug/09/ http://www.radiolab.org/2010/aug/09/

totalc 15y ago

This story has been circulating the internets for years since the original article. Has there been additional research on the Pirahã?

dragoner 15y ago

I always that its the other way, human thought determines the language

ThaddeusQuay 15y ago

Language DOES shape human thought, as anyone who knows APL can tell you. "Language as an Intellectual Tool: From Hieroglyphics to APL" by Donald B. McIntyre in IBM Systems Journal Vol.30, No.4 1991, p.554-581. http://www.electricscotland.com/mcintyre/index_f/menu_f/j_f/sysj_r_29.pdf http://www.electricscotland.com/mcintyre/index_f/menu_f/j_f/... "Notation as a Tool of Thought" by Kenneth E. Iverson in his 1979 ACM Turing Award Lecture. http://elliscave.com/APL_J/tool.pdf http://elliscave.com/APL_J/tool.pdf

Typhon 15y ago

Maybe language shapes thought. Or maybe thought shapes language. Or maybe they both influence each other. Actually, that last one, I think, is the most likely, since it's less simplistic than the previous two. Now, much has been written about the Piraha, and whether or not studying their culture will revolutionize linguistics. See this very interesting link http://www.edge.org/documents/archive/edge213.html http://www.edge.org/documents/archive/edge213.html