Sunday, July 10, 2005

The Flynn Effect: Flynn, Lynn… or Vernon?   posted by DavidB @ 7/10/2005 03:06:00 AM

The ‘Flynn Effect’ is the name generally given to the long-term trend for average scores on IQ tests to increase. Recently some writers have begun to refer instead to the ‘Lynn-Flynn Effect’. So is this new usage desirable?

To recap a little history, in 1984 James R. Flynn, a New Zealand political scientist, published a long paper [1] showing that there had been large increases in mean IQ scores in the USA between 1932 and 1978. In 1987 he published a further major paper [2] showing that the same trend could be observed in more than a dozen other developed countries. These two papers have stimulated a great deal of research and discussion, including at least one book [3].

In 1994 Herrnstein and Murray’s The Bell Curve [4] discussed the rising trend in IQ and said as follows: ‘We call it “the Flynn Effect” because of psychologist [sic] James Flynn’s pivotal role in focussing attention on it, but the phenomenon itself was noticed in the 1930s when testers began to notice that IQ scores often rose with every successive year after a test was first standardised’. Herrnstein and Murray’s term ‘the Flynn Effect’ has been generally adopted, for example by Arthur Jensen [5], who says ‘This upward trend in the population’s mean test scores has been aptly dubbed the “Flynn Effect”.’

So where does Lynn come in?

Richard Lynn of the University of Ulster is one of many psychologists who have worked on the Flynn Effect since Flynn’s original paper, but those who refer to the ‘Lynn-Flynn’ Effect evidently mean more than this. The first reference I have seen to ‘Lynn-Flynn’ is in a paper of 1999 by J. P. Rushton [6], where he says ‘The rise in IQ phenomenon might be better named the “Lynn-Flynn effect” because it was actually the Lynn (1982) paper in Nature that first identified the trend in recent times (among the Japanese)’. Rushton says that he had made the same suggestion in a book review in 1997. Rushton’s suggestion does not seem to have been widely followed by other psychologists ( I got no hits for the search phrase ‘Lynn-Flynn’ on Google Scholar), but it has been picked up by some internet commentators. For example, Steve Sailer said in August 2004 that ‘This phenomenon was first noticed in the 1940s, but Lynn was one of the first researchers to call lasting attention to it. Later, New Zealand political scientist James Flynn did important work on the subject. It is now usually called the Lynn-Flynn Effect or simply (and somewhat unfairly to Lynn) the Flynn Effect'. The clear implication here, as in Rushton’s paper, is that Lynn identified the ‘effect’ before Flynn.

Lynn’s own website list of his publications includes only two papers before 1984. The one Rushton refers to is a paper of 1982 in Nature. Curiously, the publication details for this paper on Lynn’s website are actually those of a later discussion [7]: I give the correct details for the paper itself below [8]. I have read Lynn’s 1982 paper and in my view it is far from justifying the claim that Lynn anticipates Flynn. As its title suggests, the purpose of Lynn’s paper is to show that IQ has increased in Japan relative to the USA. It compares Japanese IQ scores with American scores on the same tests, shows that the gap between them has increased by about 7 points, and concludes that ‘over the course of a generation the mean IQ in Japan has risen by ~7 IQ points’. This implies that American IQ was static over the same period. Lynn goes on to suggest that improved health and nutrition in Japan may be responsible. But this is only one of many ad hoc observations about rising IQ which have been made sporadically since the 1930s. There is nothing in Lynn’s paper to suggest awareness of a large, widespread, long-term trend, which is the key point of the Flynn Effect. Indeed, it suggests the contrary, because if Lynn had been aware of such a general trend his paper would surely have taken account of it. Notably, he would have had to consider the likelihood that IQ had risen in the USA, and not just in Japan, between the dates of the various tests. Flynn himself commented on Lynn’s paper and gave evidence of rising IQ in the USA in a letter to Nature in 1983 [9]. In his reply [7] Lynn described Flynn’s claim of rising American IQ as ‘more contentious’, and only grudgingly accepted it. Any suggestion that Lynn anticipated the Flynn Effect as we now understand it - a trend found throughout the developed world - can therefore be dismissed.

So unless anyone can produce better evidence for Lynn’s priority, I see no good reason to change the established terminology of the ‘Flynn Effect’. But even if we were to accept Lynn’s work on Japan as a full recognition of the Flynn Effect, before conceding priority to Lynn it would be necessary to show that Lynn’s 1982 paper preceded Flynn. This might seem obvious: 1982 precedes 1984. But this assumes that Flynn’s famous paper of 1984 was his first work on the subject. This is not the case. As already mentioned, Flynn commented on the subject in 1983. This is still after Lynn’s paper. But Flynn’s letter also says that ‘the evidence for American IQ gains has been published in detail elsewhere’, and gives a reference to a paper of his own in the Bulletin of the British Psychological Society for 1982 [10]. I have not seen this paper (the journal is not as widely available as the British Journal of Psychology), but I see no reason to doubt Flynn’s description. Barring the minutiae of publication dates within the same year, the papers of Lynn and Flynn were effectively simultaneous, and the claim that Lynn anticipated Flynn collapses.

But if we must hunt for ‘anticipators’ of Flynn, there is a better candidate than Lynn. In 1979 Philip E. Vernon, a distinguished British psychologist, who had been studying IQ since the 1940s, wrote that ‘A similar increase [to the widespread long-term increase in body height] has undoubtedly taken place in intelligence… There is good reason to believe that the average intelligence of the human race will continue to rise as education improves in underdeveloped countries; and that, even in western countries, further gains may occur…’ [11, p. 207] This statement is both earlier and more general than Lynn’s 1982 remarks, so if there is to be any change in usage I would suggest the ‘Flynn-Vernon’ effect as a more appropriate term. But it remains true, as Herrnstein and Murray put it, that Flynn was ‘pivotal’ in drawing attention to the phenomenon, and it was Flynn who did most of the work necessary to document it. I therefore think that the ‘Flynn Effect’ is still the best designation.

There is one proviso. Lynn is well known for the hypothesis that improved nutrition, especially in early childhood, is the main reason for the rising trend of IQ scores, not least because the phenomenon is found in quite young children, before schooling can have affected it. If Lynn is ever proved right (which in my view is quite possible), then it would be appropriate for Lynn and Flynn to share the credit for the ‘effect’: Flynn for identifying the trend, and Lynn for identifying its cause.

David B

[1] James R. Flynn (1984): The mean IQ of Americans: massive gains 1932 to 1978. Psychological Bulletin, 95, 29-51.
[2] James R. Flynn (1987): Massive gains in 14 nations: what IQ tests really measure. Psychological Bulletin, 101, 171-91.
[3} Ulric Neisser (ed.) (1998): The Rising Curve.
[4] R. Herrnstein and C. Murray (1994): The Bell Curve.
[5] Arthur Jensen (1998): The g Factor.
[6] J. P. Rushton (1999): ‘Secular gains in IQ (etc)’, Personality and Individual Differences, 26, 381-9.
[7] IQ in Japan and the United States: ‘Matters arising’, Nature (1983), 306, 291-2.
[8] Richard Lynn (1982): ‘IQ in Japan and the United States shows a growing disparity’, Nature, 297, 222-3.
[9] James R. Flynn (1983): ‘Now the great augmentation of the American IQ’, Nature, 301, 655.
[10] James R. Flynn (1982): Bulletin of the British Psychological Society, 35, 411.
[11] Philip E. Vernon (1979): Intelligence: Heredity and Environment.