"Space in Language and Cognition"
Sep. 1st, 2004 07:53 pm![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
stoutfellowThis book, by Stephen Levinson, discusses the various methods that languages use to describe location and motion, and the possible effects of these on the ways that people think. I'm not going to attempt a full-blown review, but I'd like to describe some of the results of his investigations.
To begin with, Levinson identifies several broad strategies. (Each can be implemented in a number of ways; a full typology would be quite large.) An object can be located by means of a placename; by "deixis" (literally "pointing"), using phrases like "here" and "there"; by what Levinson calls "topological" means, involving contact, containment, or proximity - in English, prepositions like "on", "in", or "near" play this role; or by methods involving "frames of reference" (FORs). It is the last which attract Levinson's attention, and there are three types. In each, the speaker wants to locate an object (the "figure") with reference to another object (the "ground").
The first type is the "intrinsic" FOR; the figure is located using some inherent feature of the ground. For example, in English one might say, "The cat is in front of the house"; the house (which is the ground) has a definite front, and the cat (the figure) is in that direction from the house. "The door is to the left" likewise uses an intrinsic FOR; this time, the ground is the addressee.
The second type is the "absolute" FOR; the figure is located according to some system of cardinal directions, as in "The house is north of the lake".
The third is the "relative" FOR, which involves three factors instead of two; along with the figure and ground, there is an additional viewpoint. For example, we say, "The ball is to the left of the tree"; "left" is not an aspect of the tree, but of the speaker or the addressee. If you say, "The dog is behind the truck", meaning that the dog is near the (intrinsic) back of the truck, that uses an intrinsic FOR, but if you mean that the dog is on the far side of the truck from the person you're talking to, that uses a relative FOR.
In English, we can use any of the three types of FOR, but the relative is probably the most common and the absolute the rarest. Other languages may use any or all of the different types, and with different frequencies. In particular, there are many languages which rely primarily on absolute FORs; some of them do not use relative FORs at all. Among them are: Guugu Yimithirr, an Australian Aboriginal tongue; Tzotzil, in the Mayan family; and a rural dialect of Tamil, which is one of the official languages of India. In Guugu Yimithirr, for instance, you might be warned that there is a scorpion north of your foot, but not that it is to your left. (For convenience, I'll refer to "absolute" and "relative" languages to mean languages in which absolute and relative FORs, respectively, are most common.)
The choice of types of FOR has interesting consequences. Levinson and his associates have shown that speakers of absolute languages typically have a remarkably good sense of direction; after being taken on a long and meandering journey, on foot or by automobile, they can identify the direction of their point of origin within about fifteen degrees of accuracy, and make a good estimate of the distance as well. By way of comparison, a similar experiment involving speakers of Dutch (like English, a relative language) yielded results that could not be distinguished from randomness. (Other, somewhat larger-scale experiments involving speakers of English and Japanese produced somewhat better results, but they were still far inferior to the results from speakers of absolute languages.)
Again, experiments were conducted in which the subject was shown a table on which small figurines of animals were lined up in a particular order. They were then turned around to face another table with similar figurines, and asked to arrange them in the same way. Speakers of relative languages would typically put the figurines in the same left-to-right order; speakers of absolute languages would put them in the same north-south order (i.e., in exactly the reverse of the order used by the relative speakers).
There are other consequences; for example, speakers of absolute languages tend to use broader, full-arm gestures, but tend not to move their torsos in such gestures. (A speaker of English, saying, "And then you turn right", is likely both to gesture rightward and turn his/her body slightly in that direction as well.)
The natural question, then, is the extent to which these linguistic choices affect thinking; Levinson gives reason to believe that this influence is quite large. (For those familiar with the Sapir-Whorf hypothesis, Levinson explicitly adopts an intermediate position between (neo-)Whorfianism and the lately popular nativism (a la Chomsky); he postulates a two-level system, with possibly innate semantic primitives undergirding a system of culturally-determined compound concepts, in which most speech draws directly on the upper level. He draws an analogy - possibly strained - to the relationship between assembler and high-level programming languages.)
All in all, if you're interested in the relationship between culture and language, this is a fascinating book. I am not much more than a half-trained amateur of linguistics, and did not find it an easy read, but if you have the interest, a little background, and enough determination, it's definitely worth the effort.
To begin with, Levinson identifies several broad strategies. (Each can be implemented in a number of ways; a full typology would be quite large.) An object can be located by means of a placename; by "deixis" (literally "pointing"), using phrases like "here" and "there"; by what Levinson calls "topological" means, involving contact, containment, or proximity - in English, prepositions like "on", "in", or "near" play this role; or by methods involving "frames of reference" (FORs). It is the last which attract Levinson's attention, and there are three types. In each, the speaker wants to locate an object (the "figure") with reference to another object (the "ground").
The first type is the "intrinsic" FOR; the figure is located using some inherent feature of the ground. For example, in English one might say, "The cat is in front of the house"; the house (which is the ground) has a definite front, and the cat (the figure) is in that direction from the house. "The door is to the left" likewise uses an intrinsic FOR; this time, the ground is the addressee.
The second type is the "absolute" FOR; the figure is located according to some system of cardinal directions, as in "The house is north of the lake".
The third is the "relative" FOR, which involves three factors instead of two; along with the figure and ground, there is an additional viewpoint. For example, we say, "The ball is to the left of the tree"; "left" is not an aspect of the tree, but of the speaker or the addressee. If you say, "The dog is behind the truck", meaning that the dog is near the (intrinsic) back of the truck, that uses an intrinsic FOR, but if you mean that the dog is on the far side of the truck from the person you're talking to, that uses a relative FOR.
In English, we can use any of the three types of FOR, but the relative is probably the most common and the absolute the rarest. Other languages may use any or all of the different types, and with different frequencies. In particular, there are many languages which rely primarily on absolute FORs; some of them do not use relative FORs at all. Among them are: Guugu Yimithirr, an Australian Aboriginal tongue; Tzotzil, in the Mayan family; and a rural dialect of Tamil, which is one of the official languages of India. In Guugu Yimithirr, for instance, you might be warned that there is a scorpion north of your foot, but not that it is to your left. (For convenience, I'll refer to "absolute" and "relative" languages to mean languages in which absolute and relative FORs, respectively, are most common.)
The choice of types of FOR has interesting consequences. Levinson and his associates have shown that speakers of absolute languages typically have a remarkably good sense of direction; after being taken on a long and meandering journey, on foot or by automobile, they can identify the direction of their point of origin within about fifteen degrees of accuracy, and make a good estimate of the distance as well. By way of comparison, a similar experiment involving speakers of Dutch (like English, a relative language) yielded results that could not be distinguished from randomness. (Other, somewhat larger-scale experiments involving speakers of English and Japanese produced somewhat better results, but they were still far inferior to the results from speakers of absolute languages.)
Again, experiments were conducted in which the subject was shown a table on which small figurines of animals were lined up in a particular order. They were then turned around to face another table with similar figurines, and asked to arrange them in the same way. Speakers of relative languages would typically put the figurines in the same left-to-right order; speakers of absolute languages would put them in the same north-south order (i.e., in exactly the reverse of the order used by the relative speakers).
There are other consequences; for example, speakers of absolute languages tend to use broader, full-arm gestures, but tend not to move their torsos in such gestures. (A speaker of English, saying, "And then you turn right", is likely both to gesture rightward and turn his/her body slightly in that direction as well.)
The natural question, then, is the extent to which these linguistic choices affect thinking; Levinson gives reason to believe that this influence is quite large. (For those familiar with the Sapir-Whorf hypothesis, Levinson explicitly adopts an intermediate position between (neo-)Whorfianism and the lately popular nativism (a la Chomsky); he postulates a two-level system, with possibly innate semantic primitives undergirding a system of culturally-determined compound concepts, in which most speech draws directly on the upper level. He draws an analogy - possibly strained - to the relationship between assembler and high-level programming languages.)
All in all, if you're interested in the relationship between culture and language, this is a fascinating book. I am not much more than a half-trained amateur of linguistics, and did not find it an easy read, but if you have the interest, a little background, and enough determination, it's definitely worth the effort.
