We turned to this paper next for two main reasons: Firstly, its importance as a case study in Roepstorff and colleagues’ paper where they claimed that culture should be understood as “patterned practices” in order to understand how cognition is “enculturated”. And secondly, because of its wider importance in debates about the nature of psychology as a discipline. Namely, that the overwhelming majority of research conducted in psychology is limited to only studying a fraction of the entire human species and yet it makes claims about the entire species. A meta-analysis by Arnett showed that 68% of all samples come from only 5% of the world’s total population (2008, pp. 602, 605). Henrich and colleagues label this the problem of WEIRD participants (Western, Educated, Industrialised, Rich, and Democratic). And Arnett mockingly asks whether we should rename psychology literature: Studies of American Undergraduate Students (2008, p. 604).
Much of our discussion was taken up with this latter issue as opposed to the former. We were joined by Greg Downey who gave his views on how anthropology could be incorporated more into cognitive science and how there is a certain clash of ideals between the differing fields into regarding human psychology and culture. You can see his commentary on the paper in his blog post at Neuroanthropology. Here he offers the acronym MYOPIC (Materialist, Young, self-Obsessed, Pleasure-seeking, Isolated, Consumerist, and Sedentary) which is intended to pick out a more negative set of features.
The paper itself consists in a nice telescoping progression of meta-analyses of a range of studies in different cognitive domains comparing the differences between: industrial and small-scale societies; Western and non-Western societies; and lastly, contemporary Americans and the rest of the West. This strategy really rams home the point that undergraduate American psychology students – who make up a disproportionate amount of the samples of psychology – really are a minuscule percentage of the human population. And even more damagingly, that this sample population are outliers (insofar as they are often at one end of a tail of distribution across differing populations) and are therefore a particularly egregious starting point from which to attempt to draw inferences across humans as a species. This is made especially clear in the diagram below (see p. 64) which shows the results of 14 small-scale societies and some South African and US undergraduate students to the Muller-Lyer illusion. This chart shows that the WEIRDs have disproportionately higher PSE (“point of subjective equality”) scores and are far from the norm in understanding how this illusion effects visual perception/judgements.
Here I would like to just pick out one of the studies that I found particularly of interest given my own research re-analysing Ed Hutchins (in)famous navigation team case study (1995). Henrich and colleagues discuss a range of studies that looked at the ways that agents from different cognitive-cultural niches tackle spatial cognitive tasks (p. 68). Examinations of linguistic terms for spatial reference have revealed that languages tend to be refer to space in one of two ways:
- egocentric – space relative to the speaker (e.g. left, right, ahead, behind, etc.)
- allocentric – space either as object centric (e.g. to the left or right of a cup on the table) or geocentric (e.g. north, south, west, and east)
WEIRD languages (Indo-European) are predominately egocentric, whereas many other languages are allocentric. For instance, it has been found that several Indigenous language groups in Northern Queensland use geocentric spatial terms fluently (Haviland 1998). And that these agents are able to reliably point in the direction of North even when inside with no reference point outside, when blindfolded, and, in one extreme case, even when spun around so much that they were on the verge of throwing up! (although this last one could be an urban myth)
What I find particularly interesting is how these linguistic differences appear to significantly influence how agents tackle and comprehend spatial reasoning tasks. In the above and below diagrams (taken from Majid and colleagues 2004 – also see Haun and colleagues 2006), speakers of either predominately allocentric or egocentric languages are presented with spatial reasoning tasks that do not have a clear cut answer but instead have one of two solutions – each favoured by either allocentric reasoning or egocentric reasoning (respectively, these are the relative and absolute solutions). The researchers found that most speakers of a language favoured their predominate way of tackling the problem. The inference being then that one’s language group structures how one comprehends and thinks about space.
I could discuss this much more because I think it is fascinating. But I will limit myself to just one more point which references back to the navigation team. As Tim Ingold has convincingly argued, map-reading is an incredibly complex cultural skill (2000, pp. 219-242). And the sophisticated activities on-board a US naval vessel prior to the widespread use of GPS-devices is a rarefied example of this. Of particular note is the fact that map-reading (or chart reading) on board a ship involves two forms of allocentric spatial frames of reference: the cardinal directions (NSWE) and also object-centric frames of reference dependent on the ship (bow, port, starboard, stern, etc.). As such, an Indo-European speaker, in learning to use a map alters their predominate relationship to space through a process of enculturation that reverses a previous default patterned way of thinking. And Hutchins relates that experienced navigators speak in terms of being able “to think like a compass” (1995, p. 141). This becomes even more interesting if the research by Haun and colleagues (2006) is valid, since this compared several primates and young human participants and found that they all preferred allocentric spatial reasoning. Older children in languages with predominately egocentric spatial terms then altered from this. This would indicate that allocentric spatial reasoning is the primate default and so egocentric language is a strong enculturating factor that overrides this (only to be complicated further by map-reading!). However, I think this is a speculative claim at this stage.
Henrich et al 2010 The weirdest people in the world? Behavioral and Brain Sciences, 33, 61-83.
Hutchins 1995 Cognition in the wild. Cambridge, MA: MIT Press.
Here is a recent interview with Joseph Henrich discussing and promoting his recent book The Secrets of Our Success. This book is a really interesting discussion of his wider work which emphasises the notion of the cultural niche.