Backs to the future?
June 12, 2006 4:04 PM   Subscribe

New analysis of the language and gesture of South America's indigenous Aymara people indicates they have a concept of time opposite to all the world's studied cultures -- the past is ahead of them and the future behind. The morphologically-rich language, of which you can hear samples here, may also prove useful to computer scientists due to its unique ternary logic system.
posted by youarenothere (42 comments total) 12 users marked this as a favorite
Marvelous. What a terrific link, and what a good week for anthropology on MeFi!
posted by fourcheesemac at 4:12 PM on June 12, 2006 [1 favorite]

> they have a concept of time opposite to all the world's studied cultures

Not quite all.
posted by jfuller at 4:16 PM on June 12, 2006 [1 favorite]

I haven't even read the links yet, but the idea of the past being ahead of us and the future behind makes a lot of sense: we can see/know the past, while we can't see/know the future.
posted by LeisureGuy at 4:24 PM on June 12, 2006

I'd recommend a serious look at Hanks' description of indexicality in Mayan grammar (see his Referential Practice) before being so sure the metaphorization of time::space boiled down to simple binaries like backward/forward and up/down in any actual grammar (accounting more fully for gestural and other deictic and metapragmatic dimensions of grammar, it is perfectly plausible that many languages model time as a spatial continuum, with uneven gradience phenomena (encoded in grammatical aspect, for example) standing for particular kinds of sequences being described. I need to read this paper before I comment further, though. Think about it -- although you may think of time as being "behind" and "in front of" you thanks to handy dandy metaphors and grammatical concepts (which, like word order defaulting to observed narrative sequences -- man bites dog, etc. - - would mostly be reversible in conceptual space) -- in expressive practice it is quite simple to engage in anachronism. We wouldn't have most of the Star Trek plots if we were truly entrapped by our basic metaphors.
posted by fourcheesemac at 4:26 PM on June 12, 2006 [1 favorite]

Or, one could simply walk backwards on a regular basis.
posted by fourcheesemac at 4:27 PM on June 12, 2006

The study's younger subjects, Aymara fluent in Spanish, tended to gesture in the common fashion. It appears they have reoriented their thinking. Now along with the rest of the globe, their backs are to the past, and they are facing the future.

Clearly, the Prime Directive has been violated here.
posted by interrobang at 4:29 PM on June 12, 2006 [1 favorite]

Aymara people indicates they have a concept of time opposite to all the world's studied cultures -- the past is ahead of them and the future behind.

If this is new, then why did Lakoff and Johnson say in the 1980 Metaphors we live by: quote
Though the polar oppositions up-down,in-out, etc., are physical in nature, the orientational metaphors based on them vary from culture to culture. For example, in some cultures the future is in front of us, whereas in others it is in back.
posted by MonkeySaltedNuts at 4:32 PM on June 12, 2006

fourcheesemac: Was trying to say the same thing, without the whole Hank's thing. Thank you for putting it so eloquently.

I am sorry to be dense, but what is the ternary logic / 'reverse concept of time' connection?
posted by batou_ at 4:38 PM on June 12, 2006

Interesting stuff, if inevitably overstated—as MonkeySaltedNuts points out, it's hardly a new concept, and the authors (Rafael E. Núñez and Eve Sweetser, to give credit where due) acknowledge that other languages (for example, Ancient Greek) have been said to exhibit the same model (past in front, future behind); I can't say I find their dismissal of such examples particularly convincing. But I'll grant them this is a striking and unusual case. For anyone who doesn't want to wade through the entire pdf, here are the first and last paragraphs of the conclusion:
How are our cultural models of time constituted, and how are they related to our models of space? In most cases, recent linguistic evidence suggests strongly that models of time are based on certain aspects of our spatial models, and that those aspects of spatial understanding are at least universally accessible cross-culturally. We argue here that gestural data provide a unique source of convergent empirical support for many of these claims, although also offering challenges to some of them.


We have good cross-cultural evidence for a range of experientially based ways of thinking for speaking about time. Often a given language manifests more than one of these patterns. The distinction between time-RP and ego-RP structures that we propose is itself a potential cognitive universal, because both kinds of structures are documented in many languages. One extremely dominant and salient pattern is the ego-RP metaphor Time Is Ego’s Motion Along a Path; its experiential basis is clearly universal, and the metaphor itself is so nearly universal that we cannot deny its cognitive accessibility to all humans. Instead of Time Is Ego’s Motion Along a Path, Aymara uses a static mapping of past and future onto the space in front of and behind ego, respectively. This mapping, although its underlying correlations are potentially accessible to any human, has distinctly less elaborate inferential mappings between source and target domains; this may account for its rarity as a primary metaphor of time. Unusual though it may be, gestural as well as linguistic data strongly and systematically attest to its cognitive reality in Aymara speakers. The study of the peculiar Aymara spatial construals of time provides an excellent opportunity to study how fundamental abstract everyday concepts such as time, although ultimately grounded in the same universal human bodily experience of the world, can get shaped in specific ways to generate cultural variability. Sadly, this rare pattern of linguistic and cognitive construal may be vanishing (at least from northern Chile), thus diminishing the rich cultural diversity of our world.
Nice post!
posted by languagehat at 5:06 PM on June 12, 2006 [1 favorite]

I'm curious to know how they've determined that "Andean logic" is a ternary logic. The final pdf seems to suggest that the third truth value means "perhaps true, perhaps false". But we are fully able to say that in English, and that doesn't mean that we implicitly employ a third truth value. Do the Aymara actually reason differently from us, or is their reasoning fully translatable into our own reasoning if their third truth value is understood as "perhaps true, perhaps false"? My guess is that something like "I don't know" or "perhaps true, perhaps false" in Andean is a short holophrastic utterance, which makes it look on the surface as if it's a whole new truth value.

Like MonkeySaltedNuts, I also thought I'd read somewhere that other cultures treated the past as before them. I have this pretty vivid memory of reading somewhere that some culture (at the time of ancient Greece?) thought of time as a river that they were standing in, facing downstream. But all these news services are shouting that the Aymara are the first. Maybe the group I'm remembering held the "the past is before us" metaphor as a philosophical position, and didn't have it deeply encoded in their culture or language.
posted by painquale at 5:10 PM on June 12, 2006

batou_: it's just that they're both weird. Most cultures have dualistic concepts of reality, at their core — day/night, left/right, man/woman, good/bad, hot/cold, true/false.

If one insists that there is a third state to truth, i.e., that something can be potentially either true and false, and unknowable ... pretend for a moment that there was a math question in seventh grade that the only right answer was 'maybe'. now pretend everything had that possibility.

The idea that propositions can only be either true or false, not both or something else, is called the Law of the Excluded Middle, and a hell of a lot is based on it. It may not be too hard to imagine things without it if you work in the humanities, but when you're deal with hard math or computer science, ternary/trinary logic makes things tricky and unintuitive, at least to my poor dualistic mind. It can do some neat tricks though.

This seems to be a good link on Aymara and trinary logic.
posted by blacklite at 5:47 PM on June 12, 2006 [1 favorite]

tibetan buddhists first inntroduced me to the concept that the past is in front of us--because we can see what has happened and learn from it, but the future is behind our backs and unknowable.

Dialecticism is as blacklite says the philosophy of oppositional binaries and the excluded middle. There are native Ameircan philosophical traditions as weell as those of tibetan and zen buddhism that have long expressed a "trenary logic" called by terms such as "the third way" or "both/and" and yes these are more applicable to humanistic endeavors than to plane geometry.

I am not commenting on the Aymara--I saw the article in the NYT I think--just noting that the "sweeping generalization" is just that, a sweeping generalization.
posted by klangklangston at 6:03 PM on June 12, 2006

Fantastic FPP. I love learning something new from this place instead of reading news I can get from CNN, recycled youtube video, and tech junk from digg.
posted by ninjew at 6:08 PM on June 12, 2006

The article seemed to say fairly clearly that they distinguish between things which were witnessed and things which were not. So I would gather their logic system is based on true, false, and unverified.
I don't quite grasp the whole "past is in front, future behind" significance. They seem to be implying it is merely a spatial relationship, but the wording they use could be misconstrued to mean "ahead/behind with respect to timeline" in which case that would mean the people believe the past is yet to come, and the future has already happened. Which I don't think they mean at all. So really, I felt the writing was unclear.
posted by nightchrome at 6:42 PM on June 12, 2006

Awesome post. Thanks!
posted by rollbiz at 6:50 PM on June 12, 2006

There was an FPP a while back about a group of people for whom a fact was only true if you also stated how you came to know it. All else was unverifiable. Perhaps similar in a sense.
posted by Freen at 6:55 PM on June 12, 2006

Is Terniary logic similar to answering "Mu" to a question such as "Have you stopped beating your wife?"
posted by Freen at 6:57 PM on June 12, 2006

Thanks nightchrome. But isn't the differentiation between things witnessed or not in African languages too? I cannot recall where I read that.
posted by batou_ at 7:08 PM on June 12, 2006

Truly an excellent post, but (I hate to say it) it really helps to warn of PDFs as they love to crash certain browser configurations.
posted by moonbird at 7:45 PM on June 12, 2006

I want to know what the third truth value means to the Aymara and how we've determined that they actually use a third truth value. The essay blacklite links to suggests that the Aymara use something like Lukasiewicz's 3-valued logic. But there are a whole bunch of deviant multivalent systems of logic. Why not Bochvar's, or Reichenbach's, or Kleene's systems of 3-valued logic? What do they use the third truth value to mean? I'm unconvinced that we couldn't just formalize the Aymara language as we do our own, but that's because I haven't seen any of the data.

You see this a lot: some field linguist wanders into the jungle, meets some tribe with odd customs, and proclaims that their armamentarium of concepts is totally foreign to our own. In this case, they speak using a completely mysterious truth value! It's spooky! But if their "third truth value" is best translated as "unverifiable" -- a concept we have ourselves -- then why not formalize their language as we do ours? The chapter that blacklite links to says that "Spanish-thinking people feel that "uncertainty is unbearable' and has nothing to do with logic, whereas for an Aymara-thinking person, "ina" is a part of reality, and is as logical as "jisa" or "jani"." I really don't know what sort of data drives the author to write this. "Spanish-thinking people" don't repudiate the very existence of uncertainty or modality. They're entirely comfortable saying things like, "it's unknown whether there is life on other planets," or, "there might be a natural vaccine to this disease." Why not translate Aymara sentences using "ina" into sentences like these? What drives the field linguist to translate "ina" as a truth value and claim that the Aymara think in a completely different way than we do?
posted by painquale at 8:20 PM on June 12, 2006

The trolls in Terry Pratchett's Discworld series have the same sort of conception of time. The reasoning provided is that they believe one moves through time backwards because you can only see the past, not the future.

posted by Pseudoephedrine at 8:21 PM on June 12, 2006

Why not translate Aymara sentences using "ina" into sentences like these? What drives the field linguist to translate "ina" as a truth value and claim that the Aymara think in a completely different way than we do?

I don't know the answer to that, but I suspect it does not have much to do with whether a translation is possible or even accurate/appropriate. After all, 3-valued logic systems are articulable (& comprehensible) in non-Aymara languages but no one claims all those languages are thinking "in a completely different way than we do."

I'm guessing it has more to do with the contexts in which "ina" is commonly used.
posted by juv3nal at 8:42 PM on June 12, 2006

There was an FPP a while back about a group of people for whom a fact was only true if you also stated how you came to know it. All else was unverifiable. Perhaps similar in a sense.

Many languages have some degree of mandatory evidentiality, which doesn't alter the facuality of information, but conditions the truth-value of the statement of fact. It's problematic when the evidence of modern metapragmatics and semantics is used to make claims about "groups of people for whom" experience is somehow conditioned by grammar. That sort of direct pseudo-Whorfianism (since Whorf never claimed anything so simple) vastly simplified the much more complex subject of linguistic relativism. (The work of John Lucy is crucial if you really want to pursue this topic.)

The concept of differential truth value is common to all human cultures, and expressible in all languages. Western cultures formalize it in legal language. "Hearsay" vs. "eyewitness" testimony, for example, is accorded different weight. The same statement of fact -- "he killed her," for example -- can be uttered in the same grammatical form to represent any possible way the utterer might have known this (s/he could also have inferred it from evidence, for example, as well as seeing it, hearing about it, or receiving a vision), but we can certainly insist on clarification, and any good defense lawyer does. We *conceptualize* the difference just fine without a mandatory grammaticalization of it.

Fascinating topic.
posted by fourcheesemac at 10:13 PM on June 12, 2006 [1 favorite]

Ferdinand De Haan's home page has some excellent stuff on cross-linguistic studies of evidentialitym, including an awesome map of evidential systems.
posted by fourcheesemac at 10:16 PM on June 12, 2006 [1 favorite]

I have a vague memory that Mandarin uses an up/down metaphor for future & past (though not necessarily in that order) - eg one might ask "what are you doing below work?" for "after work". Then again, my head might be making this up.
posted by UbuRoivas at 10:22 PM on June 12, 2006

No, you're right. One of my first Mandarin textbooks described the language's view of time as "a man walking up a stairs backwards."

Afternoon is "below noon". Morning is "above noon". Next week is "down week". Last week is "up week".

But not consistently so. Last year is "go year" for example.
posted by mono blanco at 10:57 PM on June 12, 2006

The paper linked to here was unimpressive, in my opinion. There's no logical connection to anything related to the Aymara culture, it's just a sophomoric discussion of a 3-value logical calculus from the point of view of set theory.

The author makes one very fundamental error in particular, he assigns values of 0, 1, and 2 to "unknown", "true" and "false". This mis-assignment unavoidably indicates at least two mathematical relationships between "true" and "false" that are simply not valid: ie, (1) that "false" = 2x"true"; and (2) that "unknown" < true false. the values 0 and 1 are, for a very good reason (namely, absence and presence) associated already with false and true respectively, and to move them about simply confuses the reader. better to use 0, 1 and ?. if there is an ordinal relationship at all--there may be, for some purposes--it is clearly 0 ? 1. following this path of logical thought, we are led to the conclusion that this ternary logic as the author presents it is in fact nothing more than a subset of a href="">probabilistic logic.
posted by aeschenkarnos at 12:47 AM on June 13, 2006

What the hell is the posting form doing to my text layouts?
posted by aeschenkarnos at 12:48 AM on June 13, 2006

Reminds of something I heard about musical melodies: that some cultures have considered what we call a descending meoldy as a rising melody, and use that term, because the listener moves up relative to the pitch that moves down.
posted by StickyCarpet at 6:51 AM on June 13, 2006

The visual metaphorization of musical space is a very interesting subject. "High" and "low" seem like such naturalistic terms to westerners. But that is in part because of a long history of visually notating music on a vertical scale. Music provides a fine example of what I was talking about above, since actual musical sound occupies three dimensions in space, with none of its most salient features significantly significantly encoded in vertical displacement, or "height," any more than in horizontal displacement. (Overall displacement of volume is, of course, the relevant feature we perceive as "loudness.") Indeed, we translate (as "height") the periodic frequency of sound, but never speak of high pitches as "faster" than lower ones. On the other hand, musicians (and anyone who sings) can relate to the naturalistic fast/slow parameter by considering the physical effort required to excite pitched vibrations in various media at various frequencies. The lay mistake is usually, in matters semantic, to confuse maps with territories.
posted by fourcheesemac at 7:46 AM on June 13, 2006 [1 favorite]

Afternoon is "below noon". Morning is "above noon". Next week is "down week". Last week is "up week".

Ever seen a daily or monthly calendar? These spacial relationships map exactly to one, and both types of calendar are considered highly intuitive.

Last year is "go year" for example.

Sorry, I'm all out of clever for this one.
posted by majick at 7:47 AM on June 13, 2006

Of course, you can always play music upside down.
posted by fourcheesemac at 7:48 AM on June 13, 2006

Very nice post.
posted by OmieWise at 8:42 AM on June 13, 2006

4ChMac, have you got any good references about visual metaphorization of musical space? I'd like to read more...
posted by sneebler at 12:33 PM on June 13, 2006

I recall from many years ago that one native African concept of musical pitch (might have been Zulu or Xhosa) uses 'fat' and 'skinny' for what we call 'low' and 'high'. Wish I had a better citation for you.
posted by eritain at 2:54 PM on June 13, 2006

tibetan buddhists first inntroduced me to the concept that the past is in front of us--because we can see what has happened and learn from it, but the future is behind our backs and unknowable.

Can you quote the text or source? As I've never heard any Tibetan Buddhist saying anything like this.
posted by nickyskye at 4:52 PM on June 13, 2006

4ChMac, have you got any good references about visual metaphorization of musical space? I'd like to read more...

Not exactly down the middle of the plate, but Fred Lerdahl's Tonal Pitch Space gets into some of the relevant cognitive science. Back in he 50s, Victor Zuckerkandl had some interesting things to say about this in Sound and Sense. Steven Feld's work on Kaluli (Papua New Guinea) acoustic metaphors (influenced by Zuckerkandl, and published in famous papers like "Flow like a waterfall: the metaphors of Kaluli music theory" from 1981 and "Aesthetics as Iconicity of Style" from 1988, both originally in Yearbook for Traditional Music and subsequently widely republished) are necessary reading. Ben Brinner's work on Javanese gamelan and Hugo Zemp's on 'Are'Are (Solomon Islands) panpipe music are also worth a look. There's a huge body of stuff in cog sci and psych I don't know so well, too.
posted by fourcheesemac at 5:13 PM on June 13, 2006 [1 favorite]

The point about the three value logic might be that, while we can translate some things to English and really can express the same ideas in all languages using whatever means necessary (be they aymara, english, or what have you) is that perhaps the Aymara system represents something that is less restrictive than English. On the other hand, I can't really say English is restrictive in any way when we use intonation more to express doubt or certainty, and hearsay is certainly easy "I hear soandso is coming with tomorrow".

I don't know how the Aymara system works, but its likely that there's something that justifies it being different. Speaking from my perspective on Finnish at least, there are some grammatical categories that we have in both languages that in Finnish may be more productive; for instance, in Finnish it's possible to say more verbs in "passive" than in English. In Finnish one can say, "There was swam around in the lake for a while, and then was gone home", whereas in English you can't really do that in the same way. It's not that both of these languages can't express it, as we can say, "'They'/People swam around for a while and then went home,", but perhaps these various languages just represent different strategies.

Estonian for instance, has something called an oblique mood, in which you can express hearsay with a simple suffix on the verb, and while this isn't the way English does it, it's not like it's untranslatable. Perhaps what tri-value logic and evidentiality (stating how you know some information) is that it's something that is more clearly stated in Aymara, and that it's something that is required for grammaticality of the sentences, or something that happens FAR more often. Thus, we generally apply tense to our sentences to say when something happened, and in these languages it would be just as automatic to say how you know this. In English for instance, you can just state something and you don't have to say how you know it, you can just state it as a fact, as I am doing. Aymara would require me to insert particles to describe this as hearsay, as opposed to something I've experienced myself.

Still don't know about the 3rd value logic, and I can't claim hearsay for that either
posted by taursir at 10:17 PM on June 13, 2006

Geoff Pullum responds at Language Log:
People who say that some language is unique for this or that attribute are mostly wrong. (Not always, of course. But mostly, if the claim is at all general.) And already Language Log's Asian and Pacific desk has been informed by Anthony Jukes of London's distinguished School of Oriental and African Studies:
I suppose everyone will be writing in to say that 'their' language works like this too. And so will I. Makassarese and the other South Sulawesi languages also consistently refer to the past as in front of ego — minggu ri olo week PREP front = "last week", minggu ri boko = week PREP back = "next week". And while I haven't really looked into this, I get the impression that this is not that unusual in Austronesian languages. I've been told that Sasak does it the same way, for instance.
So the notion that Aymara is entirely unique among languages and cultures is almost certainly false.
posted by languagehat at 8:10 AM on June 14, 2006 [1 favorite]

People who say that some language is unique for this or that attribute are mostly wrong

Too true. Pullum is the author of The Great Eskimo Vocabulary Hoax, which does not demolish Whorf, just all the pseudo-Whorfianism out there.

I've been saying something like that throughout this thread too. Whether and how a concept is grammaticalized does not end the question of how a concept is operationalized in discourse or thought. The Sapirian tradition of seeing grammaticalization as an important lens on the way language interacts with thought and experience is no weaker for not reying on the strong correlation.
posted by fourcheesemac at 8:53 AM on June 14, 2006 [1 favorite]

This is not unique. See Bauschatz, The Well and the Tree. Out of print, but well worth reading.
posted by QIbHom at 12:19 PM on June 14, 2006

A defense of the uniqueness of Aymara (by Russell Lee-Goldman at Language Log).
posted by languagehat at 8:25 AM on June 16, 2006 [1 favorite]

« Older Damn, Sun!   |   Net neutrality: Meet the winner Newer »

This thread has been archived and is closed to new comments