Wikipedia article on The Wonder Weeks

Learning PCT Wikipedia
1

This is a brief account of the scientific basis of The Wonder Weeks, by Hettij Plooij, Frans Plooij, with their daughter Xaveira Plas-Plooij now a third co-author, and the Wikipedia article about the book. (References are provided in the current revision of that article.) The book is important to PCT because the underlying research provides independent confirmation of the theory. The Wikipedia article has been subject to persistent attempts to denigrate the book as unscientific and to discredit Frans Plooij, representing him as a failed academic. For nine years, misinformation in this Wikipedia article has spread to mothering blogs and articles in magazines and newspapers.

=The research and a controversy=

Frans and his late wife Hettij did ethological research with chimpanzees in the 1970s, working with Jane Goodall. Ethologists have observed in many mammals ‘regression periods’ in how infants and their mothers interact, each followed by the infant’s greater independence with newly emerged cognitive abilities. While they were doing an intense longitudinal study of communication by chimpanzees, the Plooijs observed these ‘regression periods’. They had become acquainted with PCT, and saw in it a scientific explanation. After returning from Africa, they applied the same ethological approach to human mothers and infants. Their scientific results are confirmatory of PCT. That is why this is of particular importance to IAPCT. (Their popular book and a volume of replication studies are not included in the ‘Books’ section of our website, but should be.)

They published a study with 15 Dutch mothers and their infants in 1992. In the same year, Hettij published her book for the general public, in Dutch, with Frans as co-author. It was not published in English until 2003, the year that Hettij died of a tropical disease contracted during that research in Africa. Dag posted to CSGnet his news of the publication.

A controversy blew up in the late 1990s. It had two bases. The first is the methodological difference that Phil Runkel describes in his book Casting Nets and Testing Specimens: Two Grand Methods of Psychology.

The second base of the controversy has persisted in developmental psychology from its beginnings, “Is the early childhood development process gradual or in leaps? This question has preoccupied researchers since the beginning of this century, when the systematic study of child development began.” Those are words of the editor of the journal Neuropsychiatrica, introducing a volume of that journal in which Frans and his academic opponent presented their views. Frans’s opponent in the controversy, Paul van Geert, claims that infant development is chaotic and can only be described by statistical methodology and chaos theory. Frans’s rebuttal was the last of the papers there. Over Frans’s objection, De Weerth’s dissertation was published jointly with van Geert as the first author.

The immediate social setting of the controversy was Frans’s employment of van Geert’s PhD student to replicate the study that Frans and Hettij had done with human mothers and their infants. The Wikipedia article as it stands in August 2024 describes how she failed to meet the requirements of a replication. I refer you to that article for more information about the research.

=The Wikipedia article=

Over the past nine years, two people have maintained a derogatory Wikipedia article titled The Wonder Weeks, and have repeatedly deleted every change which describes the book and its scientific basis accurately.

The article was created in 2015 by a Dutch Wikipedian with the user name Pizzaman, who claims expertise in neuroscience and occupational medicine. The Dutch Wikipedia is notorious for such corrosive behavior. As you can see in that 2015 version, he wrote a derogatory hatchet job aimed at discrediting the book and the Plooijs, relying on gossipy Dutch newspaper articles as sources. Here’s an unedited cut-and-paste quotation:

This theory has been disproved by Plooij's own PhD student, Carolina de Weerth. Plooij refused to acknowledge this, which led to his resignation as a professor. ... Despite the evidence that disproved it, Plooij kept defending his theory, to te point where no one in the scientific world takes him serious anymore.

At the end of that year, Marielle Voltz, a young woman from Boston working for Wikimedia as an intern in London, did some copy editing and added a link to de Weert’s dissertation as a “paper which debunked it”. She evidently accepted that the book was ‘debunked’ without looking farther, and appears to me to have seen herself as defending young parents from a misleading fraud. She did some more copy editing in 2017.

At the end of January 2018, Marco Plas rewrote it. Within days, Pizzaman restored his version, and blocked Plas from editing the page because of his conflict of interest. This was proper, since Plas is an employee and family member. (He is Married to Xaveira Plooij.) However, in accord with Wikipedia guidelines and advice he should have been invited and encouraged to contribute suggestions on the Talk page.

I got involved in September 2019. I posted about it here in IAPCT Discourse at that time. The ensuing five years are documented in talk pages for the article and in its version history.

This spring, after I had again rebuilt the article step by well-documented step, and MVoltz had again deleted it and pasted in its place her own prior version, I appealed for help in a community forum for Wikipedians (the ‘Village Pump’). The constructive advice was just to persist. Accordingly, I rebuilt the page, one step at a time, with long pauses between them. However, subsequent to my appeal for help those two people have not revisited the article to tear it down. I surmise that some communication happened that I could not see.

2

I find this controversy about the timing of regression periods to be rather puzzling because I think the evidence is that both sides are wrong. I think it’s highly unlikely that the regression periods are as precisely timed as they are shown to be in Plooij’s The Wonder Weeks. But the temporal irregularity of regression periods certainly doesn’t imply that the development of control skills is chaotic, as claimed by van Greet.

I think the Plooij’s most important findings relative to PCT was that control skills seem to develop in the order predicted by the notion that perceptual abilities develop from the bottom to the top of the hypothetical PCT hierarchy of control. Plooij describes very convincing observational evidence of infant chimps and humans going from being able to control only sensations to the ability to control configurations of sensations and then to the ability to control smooth transitions from one configuration to another, etc. There is apparently a regularity, across individuals, of the order in which the ability to control each type of perceptual variable develops – the ability to control of sensations always precedes the ability to control configurations which always precedes the ability to control transitions, etc. And this regularity of the order of the development of control abilities is nicely described in Wonder Weeks and it certainly contradicts the notion that development is chaotic.

But these “leaps” in control skill are typically preceded “regression periods” where existing control skill declines. These regression periods presumably correspond to the random reorganization process proposed in PCT as the way of developing new skill. But I have difficulty believing that these regression periods occur with the temporal regularity across individuals that is described in Wonder Weeks. I base this on the fact that there is wide variability in the timing of other regular developmental changes, such as the break through of baby teeth and then adult teeth. For most children, the first baby teeth appear at anywhere from 6 to 10 months. That’s a 4 month (16 week range). But for every child baby teeth appear before adult teeth. So the order of development is the same for all individuals even though the timing can vary considerably. I feel that this is likely to apply to the development of control skill as much as it does to the development of teeth.

I suppose it is possible that regression periods in the development of control skills occur with the temporal regularity described in Wonder Weeks. But this seems like an extraordinary claim, which would mean it requires extraordinary evidence. So I would like to see the data on which the claim of temporal regularity of regression periods across on individuals is based. Can you point me to that data Bruce?

Thanks

Best, Rick

1 Like
3

You are correct that opinions about the science cannot be based on a popular book. It does not provide data. It has a different purpose, a useful interpretation and application of results.

References are in the Wikipedia article.

Van de Rijt & Plooij (1992) is the first human study.

Reports of their earlier primate results are available too, as are the later human replications. The Heimann volume is available if you look for it. The TOC is in this PDF together with the Preface and Introduction. If you want a specific chapter on that basis I can provide a PDF. You are of course aware of the paper in the ABS issue that you edited, and Frans’s chapter in the Handbook LCS IV. Earlier studies of other species by ethnologists could be searched out.

4

Helpful post Rick, thanks.

5

There is a good reference list at

6

I have read the relevant papers by Rijt-Plooij & Plooij (1992) – RP&P – and by de Weerth and van Geert (1998) – W&G – and the winner is… RP&P. This is not because there was anything particularly wrong with the work of W&G (1998). RP&P and W&G were just measuring different things that they called by the same name: “regression periods”.

RP&P measured these periods by having mothers say, after each week, whether the week was 1) more “tiring”, “difficult” or “demanding” than usual (if it was then the week got scored 1, otherwise it got scored 0) and 2) if a regressive behavior had happened during that period (again, if it did then the week got scored 1 otherwise it got scored 0). The results for 14 mother-infant pairs are shown in this table:

image
image726×588 174 KB

What this table shows is a remarkable consistency across mothers in terms of the weeks labeled regressive. The graph at the bottom shows the proportion of mothers rating each week as regressive (1) rather than non-regressive (0). The peaks in the graph, which represent weeks where there was maximum agreement across mothers on the week being regressive, occurred at weeks 5, 8, 12, 17, 26, 36, 44, 52, 61/62 and 72. These are the center points of what all mothers consistently identified as regressive periods (regressive sequences of weeks) and they are the basis for the regressive periods discussed in The Wonder Weeks.

W&G measured regressive periods by having the first author of their report (de Weerth) visit each of the 4 infant-mother pairs weekly and note down aspects of the infants behavior during the visit. So, W&G measured regression periods using an outside observer who saw the infant for only a few hours once a week while RP&P measured these periods using the mothers who were constantly present with the infants all week. Their measures of regression periods were the proportion of time during the once-a-week observer visit that the “regressive” behavior occurred. This is quite different than the RP&P method, which was the proportion of mothers who reported that a particular week was regressive. Given these differences in measures of regression it’s not surprising that the W&G results don’t look anything like those of RP&P (above). Here are W&Gs results showing one measure of regression --crying – as a function of weeks for two of 4 subjects:

image
image720×889 31 KB

The peaks in these graphs are not as well defined as are those in the RP&G graph above but this is a consequence of W&G’s use of a more continuous measure of regression. More important is that there seems to be little consistency in terms of the occurrence of regression periods across infants. This inconsistency is more apparent after about 20 weeks. And it’s the consistency across infants in the temporal occurrence of regression periods that is most striking in the RP&P data and is the basis for the main argument of The Wonder Weeks, which is that regression periods occur in a temporal sequence that is remarkably consisted across healthy infants.

As I said in a previous post, I find the claim that regression periods occur in a precise temporal sequence to be quite extraordinary, but I find the RP&P results to be the kind of extraordinary evidence that justifies this claim. And I find the W&G study to provide a very weak challenge to that claim. There are so many fundamental differences between the RP&P and the W&G studies that the latter can barely be considered a replication.

So my conclusion is that, based on mothers’ evaluation of the situation, babies go through difficult stages of development that last for about 2 to 4 weeks and occur near weeks 5, 8, 12, 17, 26, 36, 44, 52, 61/62 and 72. Assuming this finding holds up in future replications, it does not, in and of itself, support PCT. What is needed is research showing that this orderly progression of regression periods results from perceptual functions for higher- and higher-level control systems coming “on-line”, the regression periods being the time when the infant develops the ability for control these higher-level perceptual variable.

Best, Rick

1 Like
7

Those are excellent observations.

Another and I think more fundamental reason that de Weerth did not replicate van de Rijt and Plooij (1992) is that she admitted confounding variables that obscure the data, extrinsic sources of stress. This is detailed in Plooij & van de Rijt-Plooij (2003) and Woolmore and Richer (2003), among other places. It is particularly striking to me that in the conditions imposed by three of the four mothers an infant, unable to control by crying and clinging, substitutes smiling. I think it likely that the child is controlling for the mother to provide stable support and reassurance while reorganization is in process, but to validate that or any other interpretation is a challenge I haven’t taken on nor know quite how one would. Some aspects are taken up in van de Rijt and Plooij (1993).

There is considerable documentation of the sets of skills that emerge at each stage, and from this can be inferred what has become available cognitively at that stage. Some steps in that direction are in Plooij (2003) , Trevarten & Aiken (2003).