Language, Thought, and Space (V): Comparing Different Species

As I’ve talked about in my last posts, different cultures employ different coordinate systems or Frames of References (FoR) "which serve to specify the directional relationships between objects in space, in reference to a shared referential anchor” (Haun et al. 2006: 17568) when talking about space. As shown in my last post these linguistic differences boil down to certain cognitive differences.

Whether speakers mainly use a relative, ego-based FoR, a cardinal-direction/or landmark-based absolute FoR, or an object-based, intrinsic based FoR, also influences how they solve and conceptualise spatial tasks.

In my last post I posed the question whether there is a cognitive “default setting” that we and the other great apes inherited from our last common ancestor that is only later overridden by cultural factors – and, crucially, which Frame of Reference is might be the default one?


Haun et al. (2006: 16570) argue that

"there has been a great deal of speculation about the inherited structure of spatial relational thought. Immanuel Kant argued that the human body provides the source of our basic intuitions about the nature of space. In agreement, many cognitive scientists hold the assumption that spatial cognition is fundamentally egocentric [references omitted].”

In the same vein, Dirven et al. (2007: 1213), point out that

“The principle of corporeal deixis [i.e. body-based or body-centric spatial reference] is even so self-evident that it has been considered by most linguists, including cognitive linguists, as the default case, if not the universal one (Talmy 1983).”

I myself also thought this way and found Karl Bühler’s (1934) notion of the ego-based “Here-Now-I”-origo quite self-evident. In Bühler’s view, this deictic centre is the origin of a “coordinate system of subjective orientation,” which builds the basis and referential anchoring for all communicative acts and all social interaction and coordination,

However, we can still see the egocentric frame of reference as the universal default case and see absolute frames of reference as being cultural variations that transform our innate frame of reference. This would thus still be in line with the paradigm mentioned above as well as with developmental psychologist Jean Piaget’s central assumption that in cognitive development infants start with an egocentric perspective on the world and only later learns to coordinate multiple perspectives on the world.

To test whether this is really the case, Haun et al. used a similar but simplified setting like in their first experiment that I described in my last post, and tested gorillas, chimpanzees, bonobos, orangutans, and 4-year-old German children.

This time there were only 3 identical cups on the table so the object-centred and the geocentric conditions from the previous experiment were collapsed.
The subject was first shown where a reward is hidden
(X = Cups | = Screen E= Experimenter O=Hidden Object)
and was then rotated 180° and brought to an identical table behind the screen where they could then again locate the reward.

In the Relative condition the reward was in relation to the subject. If it was hidden to his left on table 1 it was also hidden to his left on table 2 :
In the Absolute condition “the hiding and finding cups maintained position Relative to the larger, surrounding environment” and also to “a salient landmark between the two tables, namely the screen or the experimenter” (Haun et al. 2006 : 17569).
It turned out that the great apes including the 4-year old children deployed an allocentric/environment-based layout more readily than an ego/self-based spatial strategy. At the very least, these data indicate that

“Despite common expectations, [...] Hominid spatial cognition is at least not always primarily egocentric“ (Haun et al. 2006: 15751).

To further corroborate this evidence, Haun et al. 2006 tried an even simpler version of this task on the non-human great apes as they are known to have problems with abstract rule-learning and generally scored quite low on the second experiment.

Thus, in a third experiment, the subjects were directed to the first table and in the baseline condition were rewarded for picking a cup no matter which it was for 10 times. The animals were then led to the table that was on the other side of the screen and the experimenter started to only reward one of the cups until the apes consistently picked a particular cup ten out of twelve times in a row. If they had succeeded in doing so, they were redirected to the original table and in the test trials were again rewarded for all of their ten choices.
In this way, the experimenters induced a “response” or “training bias” in the apes: In the test condition they tended to preferably pick the same cup as the one for which they were rewarded 10 times in a row in the training condition. That is, if they were rewarded for picking the middle cup in the training session, they also preferred to pick the middle cup when rotated 180° and brought back to the original table in the test condition.



The interesting question then is: If they were rewarded for picking say this cup:

in the training condition: which cup would they show the response bias for in the test condition? Would they interpret this as being ‘the cup to their left’? Or would they interpret it as being allocentrically/environmentally based?
As in the second experiment, great apes showed a preference for environmental cues as opposed to self-based cues. That is if they developed a preference for the cup in the upper illustration, they then preferred to pick the following cup in the rotated test condition.
Haun et al. (2006: 17572) conclude that gorillas, chimpanzees, bonobos, orangutans, and we all have inherited “preference for allocentric spatial strategies” from our shared ancestor.

“Based on this result, we argue that, at least for small-scale spatial relations, the inherited cognitive mode of operation is not, as argued by Kant and others, egocentric but preferably deploys environmental cues as common reference between object.”

I think this is a fascinating result and I find the argument very convincing. It is also strengthened by findings from cognitive and linguistic development: By 16 months, when they have become successful mobile navigators, children competently use non-egocentric cognitive strategies. 3- to 5-year old English-speaking children even appear to be “better at allocentric strategies than at egocentric ones.” Moreover, in cultures where an absolute frame of reference is the dominant one, children master “this system as early as 4 and certainly by 7 years of age.” Children who live in cultures that dominantly employ a relative frame of reference, on the other hand, do not seem to master the full use of a left and right system until about 11 years of age. (Haun et al. 17572).

At first sight, this might seem like quite a blow for embodiment theories in Cognitive Science and Cognitive Linguistics who see embodied, egocentric experience as our primary way of making sense of the world.
But as Dirven et al. (2007) point out, “given the universality of human bodily experience,” it would not be “astonishing” if corporeal deixis constituted “the default case” of spatial cognition,

“but as a more refined conception of embodiment, the notion of situated embodiment incorporates and integrates man’s physical and social environment in his or her holistic bodily experience” (Dirven et al. 2007: 1217).

References:

Bühler, Karl. 1982. Sprachtheorie: Die Darstellungsfunktion der Sprache. Stuttgart,
New York: Fischer (Uni-Taschenbücher 1159).

Dirven, René, Hans-Georg Wolf, and Frank Polzenhagen. 2007. “Cognitive Linguistics and Cultural Studies.” In: The Oxford Handbook of Cognitive Linguistics, ed. by Dirk Geeraerts and Hubert Cuyckens. Oxford: Oxford University Press: 1203-1221.

Haun, Daniel B. M., Christian J. Rapold, Josep Call, Gabriele Janzen, and Stephen C. Levinson. 2006. “Cognitive Cladistics and Cultural Override in Hominid Spatial Cognition.” In: PNAS 103: 17568–17573.

Levinson, Stephen C. 2003. Space in Language and Cognition: Explorations in Cognitive Diversity. Cambridge: Cambridge University Press.

Levinson, Stephen C., Sotaro Kita, Daniel B.M. Haun, Björn H. Rasch. 2002. “Returning the Tables: Language Affects Spatial Reasoning.” In: Cognition 84: 155–188.

Palmer, Gary B. 2007. “Cognitive Linguistics and Anthropological Linguistics.” In: The Oxford Handbook of Cognitive Linguistics, ed. by Dirk Geeraerts and Hubert Cuyckens. Oxford: Oxford University Press. 1045-1073.

Pederson, Eric. 2007. “Cognitive Linguistics and Linguistic Relativity.” In: The Oxford Handbook of Cognitive Linguistics, ed. by Dirk Geeraerts and Hubert Cuyckens. Oxford: Oxford University Press. 1012-1044.

Talmy, Leonard. 1983. “How Language Structures Space.” In: Spatial Orientation: Theory, Research and Application. ed. by Herbert L. Pick, Jr., and Linda P. Acredolo. New York: Plenum Books. 225–82.

Language, Thought, and Space: (IV) Comparing Different Cultures.

I wanted to continue posting on Stephen Levinson’s 2003 book, but unfortunately, I only have very limited access to the internet right now, and as my year abroad at the University of Nottingham is over, I don’t have access to the e-book version anymore (The Library at the University of Heidelberg has a copy of the book but I’m currently at home and can’t get it from there either).

Instead, I’ll post about a fascinating study done by people from the Max-Planck-Institute for Psycholinguistics (Nijmegen, Netherlands) and people from the Max-Planck-Institute for Evolutionary Anthropology (Leipzig, Germany) that pretty much blew me away (quite literally so: it pretty much destroyed a central assumption of the received theory of cognitive perspective that I wanted to work on).

As I already said, work done by Stephen Levinson and others on how different cultures talk about and conceptualise space has shown that not all of them employ a bodily, egocentric frame of reference or coordinate system as their dominant organizing principle for their experiences and thoughts. Speakers of “several indigenous languages of Australia, Papua New Guinea, Mexico, Nepal, and south West Africa,” in contrast, organize the axes of their dominant coordinate system by absolute principles such as fixed landmarks (e.g. uphill vs. downhill) or cardinal directions (e.g. move the chair to the north). In addition, there are also languages that primarily use “intrinsic,” object-centred Frames of References, such as in “The dog is at the front of the library.”
In a set of clever experiments Levinson and his colleagues have also shown that speakers of relative and absolute languages differ in how they solve non-linguistic spatial tasks. For example, in the “motion-maze task” (Pederson & Schmitt 1993) participants see a toy move on a table. They are then rotated 180° and asked to recognize, on a table with a maze-like diagram, the “the path traversed from within a maze-like diagram containing both absolute and relative possibilities“ (Levinson et al. 2002).

Interestingly, speakers of relative languages such as Dutch or Japanese recognize the path based on the relative frame of reference they employ in their language whereas speakers of absolute languages such as the Australian Aboriginal language Arrernte or Tzeltal recognize the path based on their absolute frame of reference (Levinson et al. 2002, Levinson 2003).
Results like this were achieved on a wide array of space-based tasks, so the overall findings that relative speakers prefer relative FoR in spatial tasks and that absolute speakers prefer absolute FoRs seems to be quite robust. (Haun et al. 2006).
In a recent set of experiments Haun et al. (2006) tested how soon a cognitive bias in spatial cognition manifests itself. They tested 7-11 age year old children from a Dutch village, who mainly use an egocentric FoR) and from a Khoisan hunter–gatherer community in Namibia called =/= Akhoe Hai||om (to be honest, I don’t have any idea how to pronounce this), who almost always use absolute spatial descriptions.
They also tested adults of both cultures “to see whether differences were not only initial variations of an emerging cognitive skill but were actually stable across the life span.”
What they did was the following: The subject was placed in front of a table on which there were five identical cups in a “five dice”-constellation:
(X = Cups | = Screen E= Experimenter O=Hidden Object)

They were then shown the location of an object that was hidden under one of the cups.

In the next step, the Subject was turned 180° and brought to an identical table behind the screen:

Then, and this is the crucial bit, they were asked to indicate the spot where they thought the object was hidden this time. In the experiment, there were three conditions. First, an egocentric one: if the object was hidden to the left of the subject, it was also hidden to the subject’s left after she was rotated to her new position.



Secondly, there was an object-centred condition in which the “hiding and finding cups maintained position in relation to a salient landmark between the two tables, namely the screen or the experimenter” (Haun et al. 2006 : 17659).



And finally, there was a geocentric condition where, if the hiding cup was to the north-west, it would also be the to the north-west in the rotated position.


After repeated trials on all the conditions the following picture emerged: Hai||om children and adults were faster to learn and made the fewest errors in the absolute condition and Dutch children and adults were best in the egocentric condition.
“This correlation is fully robust by age 8 and persists into adulthood. In sum, Dutch and Hai||om subjects varied in their preferred cognitive strategy to solve a spatial relational learning task, and their preference matched the preferred mode of description in their respective language” (Haun et al. 2006 : 17570).

Although it is quite difficult to interpret these results in terms of the relationship between language and thought (see e.g. Pederson 2007, Palmer 2007: 1059ff.), the results are certainly thrilling and give rise to a further question: which of these frames of reference is the primary and basic one that infants have? Is there a cognitive default setting that we and the other great apes inherited from our last common ancestor which is only later overridden by cultural factors?
I’ll return to this question in my next post tomorrow.

References:

Haun, Daniel B. M., Christian J. Rapold, Josep Call, Gabriele Janzen, and Stephen C. Levinson. 2006. “Cognitive Cladistics and Cultural Override in Hominid Spatial Cognition.” In: PNAS 103: 17568–17573.

Levinson, Stephen C. 2003. Space in Language and Cognition: Explorations in Cognitive Diversity. Cambridge: Cambridge University Press.

Levinson, Stephen C., Sotaro Kita, Daniel B.M. Haun, Björn H. Rasch. 2002. “Returning the Tables: Language Affects Spatial Reasoning.” In: Cognition 84: 155–188.

Palmer, Gary B. 2007. “Cognitive Linguistics and Anthropological Linguistics.” In: The Oxford Handbook of Cognitive Linguistics, ed. by Dirk Geeraerts and Hubert Cuyckens. Oxford: Oxford University Press. 1045-1073.

Pederson, Eric. 2007. “Cognitive Linguistics and Linguistic Relativity.” In: The Oxford Handbook of Cognitive Linguistics, ed. by Dirk Geeraerts and Hubert Cuyckens. Oxford: Oxford University Press. 1012-1044.

Pederson, Eric & B. Schmitt 1993. Eric’s maze task. In Cognition and Space Kit Version 1.0 (pp. 73–76).Nijmegen: Cognitive Anthropology Research Group at the Max Planck Institute for Psycholinguistics.