1. Low SOI scores may include not only those who follow a more monogamous mating strategy (sect. 7.5) but those who are not interested in having any sexual partners (up to 5% of the males in some of our samples). Furthermore, the SOI contains items using very different metrics, and a standardized composite is not formed: Instead, a weighing formula is used without a clear conceptual basis. In addition, many of the items are open-ended variables (e.g., number of partners desired in the next five years) that are heavily skewed (Pedersen et al. 2002), making them unsuitable for parametric analyses. The median test employed by Schmitt is known to be problematic for testing median differences (Miller & Wilcox, in preparation). The Mann-Whitney U test tells us that there are distributional differences between men and women, but not whether those differences are at the median or deep into the tails: Newer methods allow us to assess this (Miller & Wilcox , in preparation). In short, conceptually and psychometrically these measures could be improved.

2. Harlow's research (discussed by Bowlby [1969/1982]) provides a model of how diversity in mating outcomes can result from departures from the adapted-for environment (e.g., absent or impaired maternal caregiving). Clearly Harlow's monkeys (and apes) that were removed from their mothers by humans and given cloth alternatives did not evolve a sensitivity to environmental cues that produced the differential mating and sexual outcomes experienced by these primates.

3. Ultimately, relative support for alternative evolutionary theories will rest on providing models of the underlying biochemically based evolved mechanisms (and their control parameters) – and how these operate and are effected. We are learning enough about the biochemical underpinnings and genetic processes here to specify in more detail (than is provided) some plausible mechanisms. For example, regulatory genes seem to have evolved to directly impact mating strategies in voles (e.g., more monogamous versus more promiscuous) by ensuring (or not) that there are sufficient oxytocin receptor sites in the dopamine reward pathways (Insel 1997). This genetic mechanism would enable (or not) the specific partner preference phase and later attachment stages (mentioned in Figure 1) that are necessary in affording pair-bond formation. But, these effects occur between species and occur in embryonic brain development (Insel 1997; Young et al. 1998) – requirements that do not fit with either DA or SPT.

4. Solely promiscuous species may not have mechanisms for partner preference formation, whereas pair-bonding species are likely to have evolved chemical and biological mechanisms to support most, if not all, of these mechanisms. Some species, especially among primates, may evolve partner preference mechanisms and perhaps some, but not enough other mechanisms, to support enduring pair-bonds. It's an intriguing possibility that species may differ along a continuum of mechanisms that together afford pair-bonding.

5. Bowlby (1968/1982) said that, “although regarded as distinct behavioral systems, attachment behavior and sexual behavior are believed to have unusually close linkages” (p. 230). The sexual circuitry system, which is heavily impacted by positive and negative emotions, dovetails well with these systems (Miller et al., in preparation). Sustained negative emotions and/or insufficient positive emotions may serve as cues that the relationship is unlikely to last and offspring production should be avoided because, in the absence of biparental care, such offspring would be far less likely to survive.