Maori Memories

In last February's issue of Child Development, I found a paper from a team that investigates the problem of childhood memories among the Maoris. It turns out that when you ask them, Maoris produce the earliest childhood memories on record: 2.5 years on average (the average American has 3.5, Asian memories being even older on average). Their Pakeha neighbours also have first memories 3.5 years old, with Maori first memories at least 10 months earlier.

 Pictures:a Maori man (top), and writer Georges Perec, whose book W ou le souvenir d'enfance (W : a childhood memory) is a twisted autobiography dwelling on his absolute lack of childhood autobiographical memories.

Pictures: a Maori man (right), and writer Georges Perec (left), whose book W ou le souvenir d'enfance (W : a childhood memory) is a twisted autobiography dwelling on his absolute lack of childhood autobiographical memories.

One possible explanation might be that Maori children hear more stories concerning their early childhood than do other children, as we know that these narrations can help stabilize, or even recreate, first memories. Reese, Hayne and MacDonald investigated that option by recording Maori and Pakeha mothers while they were discussing episodes of their children's lives ith their children. They try to push for this explanation in their paper, but as they acknowledge, the data fall short of confirming their view.

Compared to Pakeha mothers, Maori mothers' narrations of their children's early past are not more specific, detailed or abundant. Weirdly enough, though, their account of childbirth are much more vivid than Pakeha mothers' – and this correlated nicely with the age of the earliest autobiographical memories in children. The problem, of course, is that being born is hardly a typical episode in autobiographical memory (unless one is Tristram Shandy). I very much doubt that the results could be due to the fact that Maori informants would present their mother's account of their birth as a personal memory, thus creating the one year on average difference with Pakehas and Americans. The authors would tell us if such an obvious explanation was available.

So the jury is still out on what makes Maoris remember earlier episodes of their life. I have no theory and will just make a remark: as often happens, cultural and social influence in childhood is reduced to parental influence (in this case, the mother). But there is much more to achild's social environment than her parents. Slightly older peers may be somewhat better placed to observe, record, and endlessly repeat trivial episodes of the early life of a child (being lost in a mall, peeing oneself, etc.). I speculate that children's peer culture might be a possible key to the problem.


  • Helen De Cruz
    Helen De Cruz 26 October 2008 (01:00)

    As you pointed out, Olivier, the continuing presence of peers could have a substantial influence. This turns your hunch into a testable hypothesis, namely that second-born children should have earlier childhood memories than first-borns. This seems to be the case for me and my sister (6 years my junior) but I am unaware of any systematic studies on this issue. I seem to remember that theory of mind, which is closely associated with autobiographical memory, emerges earlier in second-born children, but I forgot the source of this.

  • Dan Sperber
    Dan Sperber 26 October 2008 (01:00)

    Ok, I cannot resist: I love it when a regular anthropologist takes us, populationist thinkers (in particular cultural epidemiologists), to task for essentialising groups (a classical anthropological practice totally incompatible with a populationist approach). There is hope, there is hope! Now, to compare groups statistically implies no essentialising at all, all that is involved is the trivial assumption that some features(e.g. being claissified in some contexts as a Such-and Such and scoring high on a given psychological trait) tend to correlate (positively or negatively). So if there is a statistical difference in childhood memories among people locally classified as Maori and as Pakeha, this is a potentially interesting datum that does not at all depend on the two ethnic categories being essential ones, or even well-defined ones, or uncontested ones. It might be that even more interesting correlations would be found with more fine-grained categories, but this would have no essentialist implications either. (If I may: This reminds me of the violent rejection by many anthropologists of the Nisbett et al.’s work on East-West. There are many good reasons anthropologists should have reservations about this work, but accusing it of assuming an essentialist characterisation of ’Asians’ and ’Westerners’ is not one of them. Nisbett is using these categories in a loose manner (that makes us anthropologists cringe) precisely because all he is looking for at the start are statistical correlations between identities, however loosely defined, and psychological dispositions, in order then to interpret these statistical findings. The problem should be with the interpretation and not (most of the time anyhow) with the findings, but part of the reason anthropologists focused on the methodology is that the conclusions are embarssingly close to anthropologists’ common relativistic claims.)

  • Hugo Mercier
    Hugo Mercier 26 October 2008 (01:00)

    I would very much agree with your last suggestion. Maybe a possible twist: if abilites such as those that allow children to pass standard false belief tasks mature partly in response to a richer social context (other kids playing tricks on you), then it might be the same for the abilities underlying autobiographical memory.I would even be tempted to argue that autobiographical memory could be most useful for argumentation, and that maybe Maori kids start to argue earlier? (Ok, this is probably just me being biased, but then you mentionned children’s peer culture, so let’s call it a draw).

  • Olivier Morin
    Olivier Morin 26 October 2008 (01:00)

    I think Michael Scott has a point when he says that comparative psychologists tend to exagerate the coherence and homogeneity of the cultures and societies they study, and I agree that it is a significant problem for comparative psychologists. I am really glad the issue was raised. Actually, I have made the same point a number of times against relativist philosophers. I disagree with what Dan wrote about the work of comparative psychologists (for example, the Child Dev. study or Nisbett’s work). These authors are not merely showing correlations between some psychological measure and some ethnic classification. They also tend to assume that their results can be explained by ”culture”. Here, ”culture” is a set of ideas and practices that tends to be possessed by individuals ethnically classified as X, and not by individuals ethnically classified as Y. That variable is never measured, it is assumed to be positive in all subjects classified as Xs, by virtue of their being Xs.

    For example, Nisbett will argue that the behavior of his Chinese subjects can be explained by habits of mind that they share with, say, Lao Zi. The only reason why a random Chinese individual would share the views of Lao Zi is because both he and Lao Zi are classified by Nisbett as Chinese (although I doubt Lao Zi would have defined himself as Chinese).

    Now, as a philosopher, I tend not to use labels like ”essentialist”, because I don’t know for sure what they mean. But certainly, people like Nisbett are assuming that many cultural features massively correlate with ethnic boundaries. That assumption might in many cases be wrong, and it is never questioned. It is entirely possible that the differences such studies measure could be ascribed to non-cultural causes: for example, they might be explained by geographical environment, by social conditions shared by Xs compared to Ys, etc.

    Assuming that differences between individuals with different social labels are due to the culture they are supposed to share by virtue of having that label is indeed a strong hypothesis. And when it comes with a general disregard for the complexity of ethnic labelling (the authors of the Maori study don’t even explain how they recruited their Maori subjects, as opposed to the Pakehas, and how they divided the two groups), it is problematic.

  • Charles Stafford 26 October 2008 (01:00)

    I thought it might be interesting to show the post about ”Maori memories” – which I found very good and very thought provoking – to my colleague Michael Scott, who works in the Pacific and is familiar with many of the societies/cultures of the region. Here’s his reaction, which I think does contain serious points for culture & cognition researchers to reflect on: ”Olivier’s discussion of Maori memories makes what may be an important point that ’children’s peer culture’ is likely to be central to the development of children’s first memories. But, a lot will, of course, depend on how we define ’culture’ for these purposes. If we assume that distinct social groups have their own cultures we will be open to a range of critiques of the culture concept that are now widely accepted by anthropologists. That is to say, use of the culture concept may cause us to exoticize, essentialize, highlight differences, and so on. Such problems are apparent in the article in Child Development (not having read the article, I rely on Olivier’s characterization). For example, to treat Maori and Pakeha as two homogeneous groups is an act of problematic essentialism that not only ignores the complex dynamics of contemporary life in New Zealand, but appears naive to the political contestations over who is or is not deemed to be authentically Maori.”

  • Dan Sperber
    Dan Sperber 26 October 2008 (01:00)

    I am afraid Olivier’s apparent disagreement with my aside on anthropological reactions to Nisbett’s East-West work misses the point. I agree with everything critical he says about Nisbett’s work, but there (assuming ’culture’ as a whole to have causal effets on cognition), of course, Nisbett is not that different from most anthropologists, and they dont take him to task for this. They criticise his data as if it were anathema, much less his interpretation of his data.

  • Paulo Sousa 27 October 2008 (21:35)

    Perhaps it is important to distinguish in this discussion two different moments of ”mainstream” anthropology and its relation to the concept of culture. Once anthropologists were confortable in appealing to culture as a kind of vague independent variable that could explain differences. This is somewhat similar to some current work in cross-cultural psychology. The reaction of Michael Scott seems typical of a more contemporary post-modern rejection of the concept of culture (a good debate on the issue appeared in Current Anthropology, 1999, 40), where questions about the essentializing implications of a certain interpretation of the concept culture are mixed with political issues–they are not discussed as a pure methodological issue.

  • guest guest 28 October 2008 (18:12)

    I was wondering if one explanation of the differences observed in the different populations couldn’t be due to differences in diet. At the stage of my proposition it is obviously completely speculative but we already know that diet can have strong effects on development (e.g. myopia ). I think that for this kind of problem, direct-social-environment hypotheses are very important but do not have to be privilegied. We cannot push aside a less direct-social-environment hypotheses like diet influence on brain development. I don’t know whether or not these possible links have been studied.