ozarque (![[info]](http://l-stat.livejournal.com/img/userinfo.gif){kind=small} ozarque) wrote,@ 2007-01-30 07:19:00 |
|
Linguistics; the Piraha mystery/controversy....
Next topic to come around on the guitar, requested bylyonnesse ...
"i'd like to hear your thoughts and feelings about... the 'simplicity' of piraha (sorry for the lack of appropriate char!)..." [Note: I don't have that character -- an "a" with a tilde above it, for the final vowel in "Piraha" -- either. We can manage without it.]
Background
In "Living without Numbers or Time," by Rafaela von Bredow, online at http://www.spiegel.de/international/spi
"The language is incredibly spare. The Piraha use only three pronouns. They hardly use any words associated with time, and past tense verb conjugations don't exist. Apparently colors aren't very important to the Pirahas, either -- they don't describe any of them in their language. But of all the curiosities, the one that bugs linguists the most is that Piraha is likely the only language in the world that doesn't use subordinate clauses. Instead of saying, 'When I have finished eating, I would like to speak with you,' the Pirahas say, 'I finish eating, I speak with you.'
Equally perplexing: In their everyday lives, the Pirahãs appear to have no need for numbers. ... The Pirahãs simply don't get the concept of numbers."
In my opinion, this article contains far too many linguistics-related distortions and misconceptions. I want to mention just one of them here: Piraha is a long way from being the "only language in the world" that uses structures like "I finish eating, I speak with you." There's English, for example, where all of the following are just fine:
"You go on like that, you'll be sorry."
"You break it, you own it."
"You go out looking like that, people are going to notice."
For more background, I recommend three interesting articles on Piraha at Language Log, all written by Mark Liberman: "Good Story, Bad Headline," at http://itre.cis.upenn.edu/~myl/language
Moving On...
Douglas Dee sent me "The Straight Ones: Dan Everett on the Piraha," by Geoffrey K. Pullum, online at http://itre.cis.upenn.edu/~myl/language
I'm baffled by the Piraha case; I think most linguists are baffled by it. (Everett says that he put off publishing anything about it for a long time because he knew how "weird" it sounded.) But suppose I take off my linguist
hat and put on my science fiction writer hat.... I am then well aware that any minute now we're going to see the novel -- maybe half a dozen novels -- with the Piraha scenario as plot. Either a human culture with a language so
unlike any other human language that it forces a paradigm shift in linguistics, or a human culture that is able to maintain an elaborate linguistics hoax for decades in spite of the close attention of linguists.
[Note: I know a little bit about linguistic hoaxes, on a small scale. I've been accused (falsely) of perpetrating a linguistic hoax myself from time to time; I've had distinguished linguists stand up in the audience after I gave a paper on Ozark English and say solemnly, "I feel that everyone present should know Dr. Elgin writes science fiction, and her remarks today should be considered in that context."]
My Opinion, Nothing More...
Douglas Dee also gifted me with the August-October 2005 issue of Current Anthopology that contains Daniel L. Everett's paper "Cultural Constraints on Grammar and Cognition in Piraha" on pp. 621-634, followed by comments on pp. 635-642, a reply from Everett, and a bibliography. I gave this paper my intense attention for the simple reason that I perceive Everett's publications on Piraha to be the most important work related to the linguistics/science fiction interface that has been published to date -- because if Everett's claims for the language are true they drastically change the set of specifications for what qualifies as a Terran language.
I've been uneasy about Dan Everett's claims for Piraha from the beginning, and now that I've read the paper I'm even more uneasy. I don't have access to Piraha speakers and have never worked with an Amazonian language; I have no personal knowledge base from which to judge the examples provided, and I certainly haven't read every word Everett has written on the language elsewhere. [It should be noted that on page 623 he says "This paper supersedes any other published or unpublished statement by me on those aspects of Piraha grammar here addressed."] I am prepared, however, to offer my opinion about this one paper. I have three major objections.
1. Everett's method in the paper is to say "Piraha does not have any word(s) for [X]" or "Piraha does not have [grammar feature X]" and "When [some other linguist] says it does, he/she is simply wrong." All right; that's a coherent claim. But a few example sentences, with his translations, do not constitute adequate evidence for the claim. His position is, I believe, that because he and his wife are fluent in the language we should simply trust his judgments and accept them. I would have to point out, respectfully, that we don't even do that for native speakers of English writing about English.
2. Where the claim that the Piraha have no numbers is concerned, there is one really glaring piece of absent evidence. The first question any linguist should have, in my opinion, is this one: "When you ask a Piraha woman 'How many children do you have?', what does she say?" That is, it's one thing to have no concept of how many nuts or pebbles or batteries some linguist is asking you to count; it's quite another not to know how many kids you have, or how many siblings you have, or how many domestic animals you have. If there'd been an example where Everett asked a Piraha mother how many children she had and her answer was "I don't know" or a "I have a bigness of children," I'd have found that convincing. Nothing of that kind appears in the article.
3. About the alleged absence of embedding in Piraha. Most linguists have worked with at least one language in which the structure you see is strings of consecutive clauses with no complementizer or relativizer -- strings like "I know it /John forgot to bring the book" and "I saw the woman/she was carrying the child." The standard practice is to analyze those sequences as having either a complementizer/relativizer that has been deleted, a complementizer/relativizer that is not lexicalized, or a zero complementizer/relativizer. I have never known any linguist to just say "Oh, this language has no embedding." Evidence has to be provided; I don't see that evidence in the article. [For nonlinguists: In "I know that John left" the word "that" is a complementizer; in "I know the woman who left" the word "who" is a relativizer. And the sentence "I know John left" -- without any overt complementizer -- is fully grammatical.]
I wish that Everett had taken each of his claims -- one at a time -- and published a detailed paper on that single claim. In each paper I would have liked to see the following: a sizable array of example sentences, thoroughly analyzed; some actual dialogues; a transcript of an account of some event from one Piraha speaker -- something that would constitute an example of extended discourse, with complete morpheme-by-morpheme analysis. His case would then have been firmly established, and it would have been time for him to take up the possible implications of his work for linguistic theory as a whole. As matters stand, I am not convinced.
Update...
The October 2006 issue of Scientific American Mind, on pp. 74-77, had an interview with Dan Everett conducted by Annette Lessmoellmann titled "Don't Count on It." Lessmoellman's first question is, "How does a Piraha mother count her chidren?" And Everett answers: "She would never say, 'I have five children.' But she does not need to do so, either. After all, she knows her offspring by name and face. If she wants to take them somewhere, she always looks them over first. She does not have to count to do so."
This doesn't convince me, either. I would have expected Everett to say something like this: "I checked that very carefully. I asked 25 Piraha mothers the question 'How many children do you have?". Not a single one said "I have [some number of] children. Instead, the women said [whatever they said]."
Finally, I want to make one thing absolutely clear here. I am not suggesting that Daniel Everett, for whom I have great respect, is distorting the data, or deliberately making false claims. Nor am I suggesting that the Piraha are playing some sort of trick. I am saying only that I am totally baffled. I cannot, even with my science fiction hat on, come up with anything that strikes me as a plausible explanation.
Next topic to come around on the guitar, requested by
lyonnesse ... "i'd like to hear your thoughts and feelings about... the 'simplicity' of piraha (sorry for the lack of appropriate char!)..." [Note: I don't have that character -- an "a" with a tilde above it, for the final vowel in "Piraha" -- either. We can manage without it.]
Background
In "Living without Numbers or Time," by Rafaela von Bredow, online at http://www.spiegel.de/international/spi
"The language is incredibly spare. The Piraha use only three pronouns. They hardly use any words associated with time, and past tense verb conjugations don't exist. Apparently colors aren't very important to the Pirahas, either -- they don't describe any of them in their language. But of all the curiosities, the one that bugs linguists the most is that Piraha is likely the only language in the world that doesn't use subordinate clauses. Instead of saying, 'When I have finished eating, I would like to speak with you,' the Pirahas say, 'I finish eating, I speak with you.'
Equally perplexing: In their everyday lives, the Pirahãs appear to have no need for numbers. ... The Pirahãs simply don't get the concept of numbers."
In my opinion, this article contains far too many linguistics-related distortions and misconceptions. I want to mention just one of them here: Piraha is a long way from being the "only language in the world" that uses structures like "I finish eating, I speak with you." There's English, for example, where all of the following are just fine:
"You go on like that, you'll be sorry."
"You break it, you own it."
"You go out looking like that, people are going to notice."
For more background, I recommend three interesting articles on Piraha at Language Log, all written by Mark Liberman: "Good Story, Bad Headline," at http://itre.cis.upenn.edu/~myl/language
Moving On...
Douglas Dee sent me "The Straight Ones: Dan Everett on the Piraha," by Geoffrey K. Pullum, online at http://itre.cis.upenn.edu/~myl/language
I'm baffled by the Piraha case; I think most linguists are baffled by it. (Everett says that he put off publishing anything about it for a long time because he knew how "weird" it sounded.) But suppose I take off my linguist
hat and put on my science fiction writer hat.... I am then well aware that any minute now we're going to see the novel -- maybe half a dozen novels -- with the Piraha scenario as plot. Either a human culture with a language so
unlike any other human language that it forces a paradigm shift in linguistics, or a human culture that is able to maintain an elaborate linguistics hoax for decades in spite of the close attention of linguists.
[Note: I know a little bit about linguistic hoaxes, on a small scale. I've been accused (falsely) of perpetrating a linguistic hoax myself from time to time; I've had distinguished linguists stand up in the audience after I gave a paper on Ozark English and say solemnly, "I feel that everyone present should know Dr. Elgin writes science fiction, and her remarks today should be considered in that context."]
My Opinion, Nothing More...
Douglas Dee also gifted me with the August-October 2005 issue of Current Anthopology that contains Daniel L. Everett's paper "Cultural Constraints on Grammar and Cognition in Piraha" on pp. 621-634, followed by comments on pp. 635-642, a reply from Everett, and a bibliography. I gave this paper my intense attention for the simple reason that I perceive Everett's publications on Piraha to be the most important work related to the linguistics/science fiction interface that has been published to date -- because if Everett's claims for the language are true they drastically change the set of specifications for what qualifies as a Terran language.
I've been uneasy about Dan Everett's claims for Piraha from the beginning, and now that I've read the paper I'm even more uneasy. I don't have access to Piraha speakers and have never worked with an Amazonian language; I have no personal knowledge base from which to judge the examples provided, and I certainly haven't read every word Everett has written on the language elsewhere. [It should be noted that on page 623 he says "This paper supersedes any other published or unpublished statement by me on those aspects of Piraha grammar here addressed."] I am prepared, however, to offer my opinion about this one paper. I have three major objections.
1. Everett's method in the paper is to say "Piraha does not have any word(s) for [X]" or "Piraha does not have [grammar feature X]" and "When [some other linguist] says it does, he/she is simply wrong." All right; that's a coherent claim. But a few example sentences, with his translations, do not constitute adequate evidence for the claim. His position is, I believe, that because he and his wife are fluent in the language we should simply trust his judgments and accept them. I would have to point out, respectfully, that we don't even do that for native speakers of English writing about English.
2. Where the claim that the Piraha have no numbers is concerned, there is one really glaring piece of absent evidence. The first question any linguist should have, in my opinion, is this one: "When you ask a Piraha woman 'How many children do you have?', what does she say?" That is, it's one thing to have no concept of how many nuts or pebbles or batteries some linguist is asking you to count; it's quite another not to know how many kids you have, or how many siblings you have, or how many domestic animals you have. If there'd been an example where Everett asked a Piraha mother how many children she had and her answer was "I don't know" or a "I have a bigness of children," I'd have found that convincing. Nothing of that kind appears in the article.
3. About the alleged absence of embedding in Piraha. Most linguists have worked with at least one language in which the structure you see is strings of consecutive clauses with no complementizer or relativizer -- strings like "I know it /John forgot to bring the book" and "I saw the woman/she was carrying the child." The standard practice is to analyze those sequences as having either a complementizer/relativizer that has been deleted, a complementizer/relativizer that is not lexicalized, or a zero complementizer/relativizer. I have never known any linguist to just say "Oh, this language has no embedding." Evidence has to be provided; I don't see that evidence in the article. [For nonlinguists: In "I know that John left" the word "that" is a complementizer; in "I know the woman who left" the word "who" is a relativizer. And the sentence "I know John left" -- without any overt complementizer -- is fully grammatical.]
I wish that Everett had taken each of his claims -- one at a time -- and published a detailed paper on that single claim. In each paper I would have liked to see the following: a sizable array of example sentences, thoroughly analyzed; some actual dialogues; a transcript of an account of some event from one Piraha speaker -- something that would constitute an example of extended discourse, with complete morpheme-by-morpheme analysis. His case would then have been firmly established, and it would have been time for him to take up the possible implications of his work for linguistic theory as a whole. As matters stand, I am not convinced.
Update...
The October 2006 issue of Scientific American Mind, on pp. 74-77, had an interview with Dan Everett conducted by Annette Lessmoellmann titled "Don't Count on It." Lessmoellman's first question is, "How does a Piraha mother count her chidren?" And Everett answers: "She would never say, 'I have five children.' But she does not need to do so, either. After all, she knows her offspring by name and face. If she wants to take them somewhere, she always looks them over first. She does not have to count to do so."
This doesn't convince me, either. I would have expected Everett to say something like this: "I checked that very carefully. I asked 25 Piraha mothers the question 'How many children do you have?". Not a single one said "I have [some number of] children. Instead, the women said [whatever they said]."
Finally, I want to make one thing absolutely clear here. I am not suggesting that Daniel Everett, for whom I have great respect, is distorting the data, or deliberately making false claims. Nor am I suggesting that the Piraha are playing some sort of trick. I am saying only that I am totally baffled. I cannot, even with my science fiction hat on, come up with anything that strikes me as a plausible explanation.
