Van Lange et al. suggest that two primary ecological forces, temperature and variation in temperature, contributed to variation in the life history (LH) speed of humans, which ultimately contributed to cross-national variation in criminal propensity and behavior. We agree with some of what Van Lange et al. have to say. Yet, the CLASH model is flawed because it ignores relevant literature and because it does not grapple with genetics. Most evidence suggests that life history traits are at least partially genetically caused; we suspect that human population groups evolved different LH speeds, a plausible hypothesis Van Lange et al. ignore. We believe that it will prove difficult to explain variation in criminality without considering human population group differences and genetics. Below, we expand further on what we consider to be limitations of the target article.
J. P. Rushton was one of the first scholars, if not the first, to apply the concepts of LH evolution to the study of human population differences. His name is absent from the target article. Rushton (Reference Rushton1985a; Reference Rushton1985b; Reference Rushton2000b) argued that because of variation in ecological pressures, different human population groups evolved different LH speeds. Specifically, population groups located close to the equator evolved faster LH traits than those further from the equator. In support of his arguments, he catalogued a host of phenotypic differences in LH traits that emerged in the population using data drawn from various population groups (Rushton Reference Rushton2000b). Although debate remains over parts of his work, his theory was a powerful precursor of more sophisticated LH-based attempts to explain criminality. It is therefore remarkable that Van Lange et al. completely ignore him (see Penke et al. Reference Penke, Denissen and Miller2007). Shortly after Rushton, Ellis (Reference Ellis1988), and then more recently Boutwell et al. (Reference Boutwell, Barnes, Beaver, Haynes, Nedelec and Gibson2015), further linked LH speed to crime, yet these names too are absent from the target article. Failure to cite relevant work is not just an oversight, however. The larger consequence is a theory that is likely incorrect in key respects.
The CLASH model asserts that populations located nearer to the equator have faster LH speeds than those located farther from the equator. Yet, theirs is a very facultative explanation, one in which calibration of LH speeds appears to take place within an individual's lifetime as his or her body responds flexibly to ecological forces. A straightforward prediction of this model is that LH speeds should depend almost exclusively on where an individual is raised. Imagine an extreme version of this. According to the CLASH model, if a woman from an ancestral group that evolved near the equator moved to northern Michigan and had a child, her child's LH speed should, ceteris paribus, resemble the LH speeds of those who were raised in Michigan. The reverse should also be true. Evidence, however, strongly contradicts this. Human population groups that evolved closer to the equator for longer periods (in our ancestral past) have faster LH speeds than other groups even if they currently dwell far from the equator (Barnes et al. Reference Barnes, Boutwell, Miller, DeShay, Beaver and White2016; Rushton Reference Rushton2000b). It gets worse than this for the CLASH model. One can restrict one's focus to an individual state (even counties within states) and find variations in LH speeds and criminal propensities (Beaver & Wright Reference Beaver and Wright2011). These data are incongruous with the CLASH model. Furthermore, Figueredo et al. (Reference Figueredo, Vasquez, Brumbach and Schneider2004; see also Penke et al. Reference Penke, Denissen and Miller2007) have noted that LH speeds are at least moderately heritable. Therefore, although the environment may play some role in modifying LH traits, genetic endowment also seems to play a strong role, and LH traits are not nearly so malleable as the CLASH model appears to suggest.
A more biologically defensible hypothesis – one, in fact, already proposed by Rushton (Reference Rushton1985a) – would predict that variation in LH speeds across population groups exists in part because of biological adaption across those groups. After migrating from Africa, human groups lived in relative isolation from one another for thousands of years, each group facing different ecological pressures. One may object that evolution could not have fashioned differences between human population groups in such a short (in evolutionary terms) time. This does not appear true (Cochran & Harpending Reference Cochran and Harpending2009), as recent evidence continues to mount suggesting that natural selection can operate more quickly than once assumed, making changes even in the course of a few thousand years (Beauchamp Reference Beauchamp2016). There is no reason, and Van Lange et al. do not provide one, to dismiss Rushton's hypothesis that human population groups evolved different LH speeds in response to different selective pressures. One of the pressures might have been temperature, but Van Lange et al. are silent on the issue of whether they think that temperature differences acted to change the frequency of trait-relevant genes across different population groups. This is a proposition worth testing, of course. Either way, the CLASH model needs to take more seriously the likelihood that genetic variation between population groups also explains (at least partly) variation in LH speed and, thus, in criminality. Currently, the CLASH model is like a theory that explains variation in the color of bear fur by climatological variables, ignoring genetics and group variation. It isn't entirely wrong – bear fur is, after all, a response to features of the environment. It is just not very helpful because it ignores the most important variables (genetics, evolutionary history, etc.).
Van Lange et al. have propounded a theory that, from our perspective, appears to represent a minor step forward in our understanding of the variation of criminal propensities. Yet, the unique hypotheses they offered are almost certainly overstated or likely incorrect (e.g., the more facultative hypotheses of LH speed). The rest of their theory is a restatement of other analyses that are more fruitful and compelling than is this incarnation of an LH approach to crime. Our suspicion, however, is that this theory's shortcomings will have the positive effect of drawing attention to those more plausible, powerful, and parsimonious analyses.
Van Lange et al. suggest that two primary ecological forces, temperature and variation in temperature, contributed to variation in the life history (LH) speed of humans, which ultimately contributed to cross-national variation in criminal propensity and behavior. We agree with some of what Van Lange et al. have to say. Yet, the CLASH model is flawed because it ignores relevant literature and because it does not grapple with genetics. Most evidence suggests that life history traits are at least partially genetically caused; we suspect that human population groups evolved different LH speeds, a plausible hypothesis Van Lange et al. ignore. We believe that it will prove difficult to explain variation in criminality without considering human population group differences and genetics. Below, we expand further on what we consider to be limitations of the target article.
J. P. Rushton was one of the first scholars, if not the first, to apply the concepts of LH evolution to the study of human population differences. His name is absent from the target article. Rushton (Reference Rushton1985a; Reference Rushton1985b; Reference Rushton2000b) argued that because of variation in ecological pressures, different human population groups evolved different LH speeds. Specifically, population groups located close to the equator evolved faster LH traits than those further from the equator. In support of his arguments, he catalogued a host of phenotypic differences in LH traits that emerged in the population using data drawn from various population groups (Rushton Reference Rushton2000b). Although debate remains over parts of his work, his theory was a powerful precursor of more sophisticated LH-based attempts to explain criminality. It is therefore remarkable that Van Lange et al. completely ignore him (see Penke et al. Reference Penke, Denissen and Miller2007). Shortly after Rushton, Ellis (Reference Ellis1988), and then more recently Boutwell et al. (Reference Boutwell, Barnes, Beaver, Haynes, Nedelec and Gibson2015), further linked LH speed to crime, yet these names too are absent from the target article. Failure to cite relevant work is not just an oversight, however. The larger consequence is a theory that is likely incorrect in key respects.
The CLASH model asserts that populations located nearer to the equator have faster LH speeds than those located farther from the equator. Yet, theirs is a very facultative explanation, one in which calibration of LH speeds appears to take place within an individual's lifetime as his or her body responds flexibly to ecological forces. A straightforward prediction of this model is that LH speeds should depend almost exclusively on where an individual is raised. Imagine an extreme version of this. According to the CLASH model, if a woman from an ancestral group that evolved near the equator moved to northern Michigan and had a child, her child's LH speed should, ceteris paribus, resemble the LH speeds of those who were raised in Michigan. The reverse should also be true. Evidence, however, strongly contradicts this. Human population groups that evolved closer to the equator for longer periods (in our ancestral past) have faster LH speeds than other groups even if they currently dwell far from the equator (Barnes et al. Reference Barnes, Boutwell, Miller, DeShay, Beaver and White2016; Rushton Reference Rushton2000b). It gets worse than this for the CLASH model. One can restrict one's focus to an individual state (even counties within states) and find variations in LH speeds and criminal propensities (Beaver & Wright Reference Beaver and Wright2011). These data are incongruous with the CLASH model. Furthermore, Figueredo et al. (Reference Figueredo, Vasquez, Brumbach and Schneider2004; see also Penke et al. Reference Penke, Denissen and Miller2007) have noted that LH speeds are at least moderately heritable. Therefore, although the environment may play some role in modifying LH traits, genetic endowment also seems to play a strong role, and LH traits are not nearly so malleable as the CLASH model appears to suggest.
A more biologically defensible hypothesis – one, in fact, already proposed by Rushton (Reference Rushton1985a) – would predict that variation in LH speeds across population groups exists in part because of biological adaption across those groups. After migrating from Africa, human groups lived in relative isolation from one another for thousands of years, each group facing different ecological pressures. One may object that evolution could not have fashioned differences between human population groups in such a short (in evolutionary terms) time. This does not appear true (Cochran & Harpending Reference Cochran and Harpending2009), as recent evidence continues to mount suggesting that natural selection can operate more quickly than once assumed, making changes even in the course of a few thousand years (Beauchamp Reference Beauchamp2016). There is no reason, and Van Lange et al. do not provide one, to dismiss Rushton's hypothesis that human population groups evolved different LH speeds in response to different selective pressures. One of the pressures might have been temperature, but Van Lange et al. are silent on the issue of whether they think that temperature differences acted to change the frequency of trait-relevant genes across different population groups. This is a proposition worth testing, of course. Either way, the CLASH model needs to take more seriously the likelihood that genetic variation between population groups also explains (at least partly) variation in LH speed and, thus, in criminality. Currently, the CLASH model is like a theory that explains variation in the color of bear fur by climatological variables, ignoring genetics and group variation. It isn't entirely wrong – bear fur is, after all, a response to features of the environment. It is just not very helpful because it ignores the most important variables (genetics, evolutionary history, etc.).
Van Lange et al. have propounded a theory that, from our perspective, appears to represent a minor step forward in our understanding of the variation of criminal propensities. Yet, the unique hypotheses they offered are almost certainly overstated or likely incorrect (e.g., the more facultative hypotheses of LH speed). The rest of their theory is a restatement of other analyses that are more fruitful and compelling than is this incarnation of an LH approach to crime. Our suspicion, however, is that this theory's shortcomings will have the positive effect of drawing attention to those more plausible, powerful, and parsimonious analyses.