Burns discusses several physiological measures in relation to psychometric g, but questions the importance of the observed correlations between these measures and the g factor. I argue that these correlations are worthy of attention, replication, and further investigation to discover their meaning. The positive findings are too significant to ignore. We are not in an entirely know-nothing position, but have some promising leads of where to look further for the physical basis of g.
2. The findings that tests' g loadings predict to a high degree those tests' correlations with the wave-form complexity (i.e. the 'string measure') of the average evoked potential and with the habituation index of the evoked potential are so far beyond chance and are so consistent with the hypothesis that g, more than any other factor in IQ test batteries, best reflects the biological component of individual differences in intelligence, that they should not be dismissed out of hand. Every attempt should be made to replicate them; if this is successful, these paradigms will serve as a wedge for further investigation into the biological basis of g. Resistance to this line of exploration seems nearly endemic among psychologists, as if physiology threatened to put psychology out of business. It won't, of course, because the measurement and analysis of behavior per se is just as essential for investigating its physical basis as is brain physiology itself.
3. Of course, a statistical correlation between two variables doesn't imply direct causality of the one variable by the other, nor does it imply the absence of causation. A lack of correlation, however, is more apt to imply the absence of causation. Investigation can't afford to eschew correlations as clues to causation, which of course must be established by other kinds of analysis. At an exploratory stage of investigation, reciting the "correlation is not causation" mantra is a premature criticism.
4. I agree with Burns that what Reed and Jensen (1992) measured and labeled as nerve conduction velocity (NCV) could just as well be labeled the visual evoked potential corrected for head length, but this makes no important substantive difference, as the procedure was clearly described in the original paper. What it amounts to is the subject's head length, measured with calipers, divided by the time interval between stimulation of the retina and the resulting change in electrical potential in the visual cortex. We called this 'nerve conduction velocity' (NCV). Most of the conduction time by far reflects axonal speed, as there are only four synapses in this nerve pathway, which have exceedingly short transmission times, as does the retinal reaction to the stimulus. Yet in a sample of male college undergraduates, this measure was significantly correlated with Raven IQ, an untimed and highly g-loaded test of nonverbal reasoning.
5. This experiment has since been replicated, with one refinement in procedure (Andres-Pueyo et al., 1999). In addition to measuring head length, the precise length of the pathway from retina to visual cortex was directly measured in each subject by magnetic resonance imaging. The correlation between their measure of NCV and Raven IQ was +.38. One hopes for further replications of this paradigm in representative samples of the general population. Reed & Jensen (1992) were fully aware of all the anatomical details and complexities of the visual evoked potential as mentioned by Burns and as found in every textbook of brain anatomy and physiology. This information does not in the least detract from the most interesting finding that the obtained measurements of involuntary reactions of very short duration (70 to 100 msec.) are correlated with scores on an untimed psychometric test of complex reasoning.
6. Since 'The g Factor' was published, it has been found that inspection time (IT), which is the shortest exposure of a discriminative stimulus that will permit the subject to make the discrimination correctly at some specified above-chance frequency, is probably less correlated with g per se than with a perceptual speed factor that figures in some tasks that are typically loaded on a spatial ability factor or the nonverbal 'performance' factor of the Wechsler IQ battery. Although IT is correlated about -.50 with IQ, this correlation is more attributable to the visual-spatial tests than to the verbal tests or to g, independently of the verbal and spatial factors (Deary & Crawford, 1998). It remains to be determined whether the NCV measure used by Reed & Jensen (1992) has the same property in this respect as IT, since NCV was correlated with the Raven test, which is nonverbal and involves reasoning based on visual-spatial figures. Both NCV and IT may reflect only the early input stage of information processing. This conjecture is consistent with the hypothesis that speed-of-processing is a component of both g and spatial ability.
7. There is no empirical evidence to support the conclusion that g (and IQ) is not heritable. The substantial heritability of g (between .4 and .8) is no longer even controversial, given the now massive amount of evidence on this subject. Those who are familiar with it no longer argue the issue.
8. Burns (par.#8) claims that I to hold a "unidimensional description of human abilities." This would mean, in other words, that I believe that Spearman's original "two-factor theory" (i.e., g + test specificity) is correct. I explicitly contradict this unidimensional view of abilities in "The g Factor" (Jensen, 1998, pp.31- 32; and Jensen, 1999) and subscribe fully to J. B. Carroll's (1993) 3-stratum theory of abilities, which embraces, besides g, 6 to 8 second-order group factors, and some 40 or so first-order factors, although the number of first-order factors is theoretically almost unlimited. The existence of many lower-order factors at the first and second strata of Carroll's model is not at all incompatible with g at the third-stratum. Also, pages 572-578 of 'The g Factor' discuss the limitations of g and the importance of other factors of ability and personality that are influential in many spheres of human performance.
Andres-Pueyo, A., Boastre, R.M., & Rodriguez-Fornells, A. (1999). Brain nerve conduction velocity, magnetic nuclear resonance and intelligence: New data. Paper presented at the 9th Biennial Convention of the International Society for the Study of Individual Differences, Vancouver, B.C., July, 1999.
Burns, N.R. (1999). Biological correlates of IQ scores do not necessarily mean that g exists. PSYCOLOQUY 10(73) ftp://ftp.princeton.edu/pub/harnad/Psycoloquy/1999.volume.10/ psyc.99.10.073.intelligence-g-factor.15.burns http://www.cogsci.soton.ac.uk/cgi/psyc/newpsy?10.073
Carroll, J.B. (1993). Human cognitive abilities: A survey of factor-analytic studies. Cambridge, UK: Cambridge University Press.
Deary, I.J., & Crawford, J. R. (1998). A triarchic theory of Jensenism: Persistent, conservative, reductionism. Intelligence, 26, 273-282.
Jensen, A.R. (1998). The g factor: The science of mental ability. Westport, CT: Praeger.
Jensen, A.R. (1999). Precis of: "The g Factor: The Science of Mental Ability" PSYCOLOQUY 10(23). ftp://ftp.princeton.edu/pub/harnad/Psycoloquy/1999.volume.10/ psyc.99.10.023.intelligence-g-factor.1.jensen http://www.cogsci.soton.ac.uk/cgi/psyc/newpsy?10.023
Reed, T.E., & Jensen, A.R. (1992). Conduction velocity in a brain nerve pathway of normal adults correlates with intelligence level. Intelligence, 16, 259-272.