(1999) argues, emotions involve the cognitive appraisal by the cortex of emotional impulses generated by the limbic system, testosterone would potentiate the original impulses involved in attracting and competing for mates. DHEA/S would act to expand these impulses by promoting neural pathways that incorporate the positive judgment of social factors that lead to pride while cortisol would tend to lead to the negative awareness of social factors contained in shame. Such a process might also underlie the creativity that Miller (2000) argues is the basis of courtship in human males.
So far, I have argued, on largely theoretical grounds, that the existence of adrenarche in humans gives us special reason to expand our understanding of the processes involved in human reproductive maturation beyond simply the development of sexual behavior. The next section follows through on this argument by considering empirical evidence for gonadal and adrenal hormones in the sexual behavior of adolescent boys. We start with findings on the role of pubertal maturation and testosterone in the onset of sexual behavior and then move to evidence that both cortisol and DHEA/S may have a role in aspects of social behavior important for the full expression of sexual behavior.
Udry has developed a biosocial model of the determinants of sexual behavior among adolescent boys (Udry et al., 1985; Udry and Billy, 1987). In its original formulation the model argued that increasing levels of testosterone during puberty will result in increasing levels of endogenous sexual motivation in boys. To the extent that sexual motivation is not socially controlled, it will result in the onset of sexual behavior in boys. And compared to girls, it appears that there are relatively few social controls on boys.
The role of testosterone in pubertal maturation remains hard to pin down. Halpern (1997) was able to demonstrate a direct relationship between salivary testosterone and sexual behavior in adolescent boys, though an earlier analysis found that secondary sexual characteristics, not serum testosterone, was the more important predictor (Halpern et al., 1993). Differentiating behavior and physical effects of testosterone is difficult because testosterone levels rise quite quickly in individual boys (Nielsen et al., 1986). Thus, changes in motivation associated with testosterone stimulation of the brain may happen in the absence of social opportunities for sexual behavior.
Fortunately, the distinction between direct and indirect effects of testosterone may not be crucial to understanding the role of pubertal matura-