tapraid4/z2n-amepsy/z2n-amepsy/z2n00811/z2n3794d11z xppws Sϭ1 10/12/11 4:21 Art: 2010-2207

versal. Individual differences certainly neity rather than homogeneity in those cog- DOI: 10.1037/a0024626 exist in the efficiency, size, quality, sensi- nitive processes (Bartholomew, Deary, & Evolutionary Psychology and tivity, or performance of adaptations, but Lawn, 2009). Van der Maas et al. (2006) these differences are not the basis of their proposed a second plausible model that ex- Intelligence Research Cannot existence (Borsboom & Dolan, 2006; plains g by positing mutualistic develop- Be Integrated the Way Penke, 2010). The existence of g does not mental interactions among distinct cogni- Kanazawa (2010) Suggested indicate that general intelligence is present tive processes. These models are within every normal human, but that every irreconcilable with Kanazawa’s (2010) Lars Penke human occupies one of its levels, which is claim that g reflects a single domain-spe- The University of Edinburgh a completely different statement. In short, g cific adaptation. is not an adaptation or causal mechanism, Kanazawa’s (2010) second assump- Denny Borsboom but a variable. tion, that general intelligence is a specific University of Amsterdam A variable is not necessarily associ- adaptation to the domain of evolutionary ated with just one modular adaptation (or novelty, is also questionable. “Evolution- Wendy Johnson mechanism/process/cause). Running speed ary novelty,” which is defined by exclusion The University of Edinburgh is associated with bipedalism, but also with (i.e., as everything previously not encoun- the cardiovascular system, with the tered in our evolutionary past), is not a Rogier A. Kievit, Annemie Ploeger, and achievement motive to train harder, and so coherent characterization of an adaptive Jelte M. Wicherts forth—arguably adaptations in their own problem. Selection can only tailor domain- University of Amsterdam rights. A given variable can indicate pa- specific adaptations to common problem rameters of adaptations (Penke, 2010), but structures. So which structural feature do Evolutionary psychologists search for hu- the variable is never tantamount to the lightning, flash floods, television charac- man universals, differential psychologists adaption. In addition, a variable is not the ters, genetically unrelated groups, and elec- for variation around common human cause of the existence of an adaptation tric light have in common? According to themes. So far, evolutionary psychology (running speed is not the cause of bipedal- Kanazawa, their communality is that they and differential psychology seem some- ism), nor does the existence of an adapta- pose problems that are logically solvable. what disparate and unconnected, although tion explain the nature of individual differ- But what is logically correct about being Kanazawa (May–June 2010) is certainly ences (bipedalism is not the cause of politically liberal when living in unrelated not the first to attempt integrating them (see differences in running speed). groups or about being slightly more noc- The empirical observation of g in it- Penke, 2010, and references therein). Ka- turnal when having electric light? Further- self tells us nothing about the causal rea- nazawa uses intelligence to elaborate his more, there is no coherent computational sons why people show individual differ- view of integration. His evolutionary the- mechanism that embraces “methods of in- ences in g. Kanazawa (2010) assumed that ory of intelligence is based on two assump- duction...deductive reasoning..., anal- g is the result of a single domain-specific tions: (1) General intelligence (g) is both an ogy, abstraction, and so forth” (Kanazawa, adaptation. If this were true, then different individual-differences variable and a do- 2010, p. 283) and could thus be called a individuals with the same g score (or rank main-specific adaptation, and (2) the do- domain-specific general intelligence adap- on the dimension described by g) should main to which general intelligence is tation. Finally, novelty violates require- adapted is evolutionary novelty. Both have this score for the same reasons; that is, ments for rational decision theory (includ- claims are erroneous. g differences reflect differences in the per- ing logic), as by definition relevant Kanazawa (2010) defended his first formance of a single coherent adaptation. information is unknown or has to be esti- assumption by arguing that there are indi- The available biological evidence, how- mated from small samples when encoun- vidual differences associated with any ad- ever, points to causal heterogeneity under- tering novelty (Gigerenzer & Brighton, aptation. To use one of his illustrations: lying g and thus against the single-intelli- Bipedalism is a universal human adapta- gence-adaptation hypothesis: Different 2009). More plausible evolutionary re- tion, but there are individual differences in individuals seem to use their brains differ- sponses to novelty include simple heuris- running speed. In the same vein, Kanazawa ently to solve intelligence tests equally tics (Gigerenzer & Brighton, 2009), open claimed that general intelligence is a uni- well, and different rare (probably private or developmental programs (Mayr, 1974), and versal human adaptation but that there are family-specific) mutations likely contribute domain-specific adaptations supporting so- also individual differences in general intel- substantially to the genetics of g in differ- cial/cultural learning (instead of wide- ligence. He then inferred that g, an individ- ent individuals (Deary, Penke, & Johnson, spread individual reasoning) (Henrich & ual-differences variable, can be used as a 2010; Penke, 2010). Thus, although one McElreath, 2003). “measure” or “indicator” of a general intel- cannot rule out the single-intelligence-ad- To conclude, while evolved adapta- ligence adaptation. This line of reasoning is aptation hypothesis a priori, the biological tions can and often do vary in certain pa- troublesome. As Borsboom and Dolan evidence does not support it. rameters, an individual-differences vari- (2006) have spelled out, g is a psychomet- We need to keep in mind that the able need not correspond to a specific ric construct, reflecting positive correla- g-factor is just one among many models for underlying adaptation. Because g is an in- tions between scores on different cognitive describing the actual empirical observa- dividual-differences variable, it is uninfor- tests (i.e., the positive manifold). To inter- tion, namely, positive-manifold correla- mative about whether a domain-specific pret g as something other than a psycho- tions among cognitive tests. Equally plau- adaptation for evolutionary novelty exists. metric construct is to go far beyond the sible and explanatory models exist: For This undermines Kanazawa’s (2010) inte- data. Specifically, in contrast to adaptations instance, Godfrey Thomson (1881–1955) gration of evolutionary and differential such as language acquisition or color per- presented a “bonds” model based on ran- psychology as well as his empirical evi- ception, g refers exclusively to human in- dom sampling of cognitive processes for dence for his evolutionary theory of intel- dividual differences, not to a human uni- solving test items, thus positing heteroge- ligence—all based on g—completely. In

Month 2011 ● American Psychologist 1 © 2011 American Psychological Association 0003-066X/11/$12.00 Vol. ●●, No. ●, 000–000 DOI: 10.1037/a0024626 tapraid4/z2n-amepsy/z2n-amepsy/z2n00811/z2n3794d11z xppws Sϭ1 10/12/11 4:21 Art: 2010-2207

addition, neither “evolutionary novelty” Deary, I. J., Penke, L., & Johnson, W. (2010). individual differences. In D. M. Buss & P. H. nor “general intelligence” as a computa- The neuroscience of human intelligence dif- Hawley (Eds.), The evolution of personality and tional mechanism is a coherent concept. ferences. Nature Reviews Neuroscience, 11, individual differences (pp. 243–280). New Any integration of evolutionary and differ- 201–211. doi:10.1038/nrn2793 York, NY: Oxford University Press. doi: Gigerenzer, G., & Brighton, H. (2009). Homo 10.1093/acprof:oso/9780195372090.003.0009 ential psychology must acknowledge that heuristicus: Why biased minds make better van der Maas, H. L. J., Dolan, C. V., Grasman, there can be more than one causal adapta- inferences. Topics in Cognitive Science, 1, R. P. P. P., Wicherts, J. M., Huizenga, H. M., tion or mechanism underlying any dimen- 107–143. doi:10.1111/j.1756-8765.2008 & Raijmakers, M. E. J. (2006). A dynamical sion of individual differences. .01006.x model of general intelligence: The positive Henrich, J., & McElreath, R. (2003). The evo- manifold of intelligence by mutualism. Psy- lution of cultural evolution. Evolutionary An- chological Review, 113, 842–861. doi: REFERENCES thropology, 12, 123–135. doi:10.1002/ 10.1037/0033-295X.113.4.842 evan.10110 Bartholomew, D. J., Deary, I. J., & Lawn, M. Kanazawa, S. (2010). Evolutionary psychology (2009). A new lease of life for Thomson’s and intelligence research. American Psychol- Correspondence concerning this comment bonds model of intelligence. Psychological ogist, 65, 279–289. doi:10.1037/a0019378 should be addressed to Lars Penke, Centre for Review, 116, 567–579. doi:10.1037/a0016262 Mayr, E. (1974). Behavior programs and evolu- Cognitive Ageing and Cognitive Epidemiology, Borsboom, D., & Dolan, C. V. (2006). Why g is tionary strategies. American Scientist, 62, Department of Psychology, The University of not an adaptation: A comment on Kanazawa 650–659. Edinburgh, 7 George Square, Edinburgh EH8 (2004). Psychological Review, 113, 433–437. Penke, L. (2010). Bridging the gap between mod- 9JZ, United Kingdom. E-mail: lars.penke@ doi:10.1037/0033-295X.113.2.433 ern evolutionary psychology and the study of ed.ac.uk

2 Month 2011 ● American Psychologist JOBNAME: AUTHOR QUERIES PAGE: 1 SESS: 1 OUTPUT: Wed Oct 12 04:21:21 2011 /tapraid4/z2n-amepsy/z2n-amepsy/z2n00811/z2n3794d11z

AUTHOR QUERIES

AUTHOR PLEASE ANSWER ALL QUERIES 1