Rafael Wlodarski, John Manning, R. I. M. Dunbar

性的関係において、ひとところに『留まる』型と、『さまよう』型の男女群を比較した。着目したのは性行動に関する二つの潜在指標だ。
一つのデータは、北米と英国で性的習癖について尋ねたインターネット・アンケートで、もう一つは、英国の男女1314人について行ったいわゆる「2D:4D比」に関する調査だ。
「2D:4D比」とは、薬指の長さで胎児期に影響を受けた男性ホルモン(テストステロン)の量が分かるとする暴露指標で、母体の胎内でさらされた男性ホルモンの濃度が高い多いほど、人差し指に比べて薬指が長くなるとされている。この人差し指に比べて薬指が長い人が、性的関係においては統計的に「不特定多数」傾向が高いとされてきた。
今回、二つの調査のデータを合わせたところ、男性では57%が「不特定多数」型で、43%が「一途」型だった。女性ではこれが逆になり、「不特定多数」型は47%、「一途」型が53%だった。
ただし、純粋に身体的特徴に基づく「2D:4D比」に関するデータだけでみると、男性67%、女性50%と、二つのデータを合わせたよりも男女共に「不特定多数」型の数値が高くなった。
ダーウィンの進化論的な視点でいえば、複数の相手との性行為は、子孫が生まれる確率を高め、従って遺伝子が引き継がれる確率が高まる。一方、決まった相手とのより長期にわたる関係は、もっと個人的な努力を要するが、性行為の結果として誕生した子孫の生存の可能性が高まる。チームは「今回の研究が示唆しているのは、男女それぞれに二つの異なるタイプの個人がいて、異なる交尾戦略をとっているということだ」と述べている。

2 thoughts on “Rafael Wlodarski, John Manning, R. I. M. Dunbar

  1. shinichi Post author

    人間の性行動は2タイプ、「不特定多数」型と「一途」型

    AFPBB News

    https://www.afpbb.com/articles/-/3038988

    人間は性行動について「不特定多数」か「一途」かの二つに大きく分かれるとする説を裏付ける初めての証拠を得たとする研究結果が5日、英専門誌バイオロジー・レターズ(Biology Letters)に発表された。

     性的関係において、人間以外のほ乳類は種によって完全に一夫多妻制か、一夫一妻制かのどちらかだが、現生人類ホモ・サピエンスは種全体でいずれかに納まることはない。人間は何故、性的関係になるとほ乳類の中で例外なのかは長年の謎だ。

     今回、その答えを見つけたとするのは、英オックスフォード大学(University of Oxford)の実験心理学者ラファエル・ブウォダルスキ(Rafael Wlodarski)氏らの研究チームだ。「性的関係において、ひとところに『留まる』型と、『さまよう』型の男女群を比較」した。着目したのは性行動に関する二つの潜在指標だ。

     一つのデータは、北米と英国の18~63歳の回答者585人を対象に性的習癖について尋ねたインターネット・アンケートで、もう一つは、英国の男女1314人について行ったいわゆる「2D:4D比」に関する調査だ。

    「2D:4D比」とは、薬指の長さで胎児期に影響を受けた男性ホルモン(テストステロン)の量が分かるとする暴露指標で、母体の胎内でさらされた男性ホルモンの濃度が高い多いほど、人差し指に比べて薬指が長くなるとされている。この人差し指に比べて薬指が長い人が、性的関係においては統計的に「不特定多数」傾向が高いとされてきた。

     今回、二つの調査のデータを合わせたところ、男性では57%が「不特定多数」型で、43%が「一途」型だった。女性ではこれが逆になり、「不特定多数」型は47%、「一途」型が53%だった。

     ただし、純粋に身体的特徴に基づく「2D:4D比」に関するデータだけでみると、男性67%、女性50%と、二つのデータを合わせたよりも男女共に「不特定多数」型の数値が高くなった。

     この違いは調査結果の解釈に慎重となる必要性を示していると研究者たちはいう。オックスフォード大のチームのロビン・ダンバー(Robin Dunbar)教授は「人間の行動は環境や人生経験など多くの要因に影響される。母親の胎内にいる間に起きることは、性的関係のような複雑な事柄に対して非常にわずかな影響しかもたらさないのかもしれない」と述べた。

     では、性行動の違いはどのように説明できるのだろうか。ダーウィンの進化論的な視点でいえば、複数の相手との性行為は、子孫が生まれる確率を高め、従って遺伝子が引き継がれる確率が高まる。一方、決まった相手とのより長期にわたる関係は、もっと個人的な努力を要するが、性行為の結果として誕生した子孫の生存の可能性が高まる。チームは「今回の研究が示唆しているのは、男女それぞれに二つの異なるタイプの個人がいて、異なる交尾戦略をとっているということだ」と述べている。

    Reply
  2. shinichi Post author

    Stay or stray? Evidence for alternative mating strategy phenotypes in both men and women

    by Rafael Wlodarski, John Manning and R. I. M. Dunbar

    Biology Letters

    https://royalsocietypublishing.org/doi/10.1098/rsbl.2014.0977

    Abstract

    In all comparative analyses, humans always fall on the borderline between obligate monogamy and polygamy. Here, we use behavioural indices (sociosexuality) and anatomical indices (prenatal testosterone exposure indexed by 2D : 4D digit ratio) from three human populations to show that this may be because there are two distinct phenotypes in both sexes. While males are more promiscuous and display higher prenatal testosterone exposure than females overall, our analyses also suggest that the within-sex variation of these variables is best described by two underlying mixture models, suggesting the presence of two phenotypes with a monogamous/promiscuous ratio that slightly favours monogamy in females and promiscuity in males. The presence of two phenotypes implies that mating strategy might be under complex frequency-dependent selection.

    1. Introduction

    Whenever comparative analyses of mammalian mating systems are undertaken, humans invariably fall midway between monogamous and polygamous species [1,2]. Although no explanation has ever been offered for this, one plausible explanation is that humans actually consist of a mix of short-term (promiscuous) and long-term (monogamous) mating phenotypes. The extent to which any one individual pursues a short-term mating strategy (‘unrestricted’ strategy involving promiscuous mating with multiple partners) or a long-term mating strategy (‘restricted’ strategy favouring the formation of exclusive and extended pair-bonds) has been referred to as their ‘sociosexual orientation’ [3].

    Cross-cultural research widely confirms that males, the lesser-investing sex, are typically more sociosexually ‘unrestricted’ (promiscuous) than females [4]. While long-term mating strategies that involve increased parental investment may enhance a male’s chances of offspring survival [5], they do so at the expense of lost mating opportunities [6,7]. Although these two strategies could well just be opposite ends of the same continuum, it has sometimes been assumed (albeit without any real evidence) that these represent two distinct male phenotypes: those that pursue a more promiscuous, unrestricted mating strategy (‘stray’) and those that focus on investing more heavily in their offspring in long-term relationships (‘stay’) [811]. Although individual differences in female mating strategies have sometimes been noted in the literature [10,12], the possibility that women might also exhibit contrasting mating strategies has received considerably less attention.

    We tested the hypothesis that there are distinct mating strategy phenotypes in both men and women using two large datasets: a North American and British sample of 595 individuals who completed the sociosexual orientation inventory (SOI-R) [13] and a British sample of 1314 individuals whose 2D : 4D digit ratios were measured. The SOI-R indexes an individual’s psychological degree of sexual promiscuity on a continuum running from restricted (monogamous) to unrestricted (promiscuous). The 2D : 4D ratio is an anatomical marker for fetal testosterone exposure and testosterone receptor-site density [14,15], and reflects the level of prenatal testosterone effects in the adult phenotype [16]. Across primates, 2D : 4D ratio correlates with mating system [2,17] and provides a biological marker for mating strategy. To test the hypothesis, we first determined whether or not the distribution of each index for each sex was most likely to be unimodal or bimodal, with the latter being indicative of a mixture of two underlying distributions. Where the latter was the case, we determined the mean and variance for each distribution.

    2. Methods

    SOI-R data were collected from British and North American Caucasian participants using an online questionnaire [18]. These comprised 134 male and 186 female British participants and 68 male and 187 female North American participants (ages 18–63, mean = 24.7, s.d. = 7.9). To assess participants’ preferred mating strategy (sociosexual orientation), questions forming the ‘attitude’ and ‘desire’ subscales of the SOI-R were used [3,13].

    Data on 2D : 4D ratios were previously collected from right-hand photocopies in a large-scale study of a British Caucasian population (n = 1314, 572 males, 742 females).

    Statistical analyses were carried out with R and the mixtools package for finite mixture model analysis using model-based clustering (version 0.4.6) [19]. This analysis assumes underlying Gaussian distributions of any modelled modes—an assumption that seems reasonable as most genetic and behavioural human variation is typically normally distributed. To determine whether each dataset was a mixture of two distributions, we first used a likelihood ratio test (asymptomatic χ2) to compare multicomponent distributions with a single-mode model. Maximum-likelihood estimation (MLE) was then used to iteratively estimate multimodal Gaussian distributions that maximized model fit to the observed data, varying component mixing proportions (λ), means (μ) and standard deviations (σ). Mean best-fit model parameter estimates and standard errors were further estimated using parametric bootstrapping (1000 realizations).

    3. Results

    The likelihood ratio χ2 tests confirm that the British and North American male and female sociosexuality datasets each have an underlying bimodal distribution (table 1). Modelling confirmed the existence of two phenotypes within each sex, one of low (restricted) sociosexuality and the other of high (unrestricted) sociosexuality. High-sociosexuality males make up a slightly larger proportion of the male distribution in each case, and low-sociosexuality females make up a slightly larger proportion of the female distributions (table 1).

    While the χ2 tests confirm that the male 2D : 4D data also have an underlying bimodal distribution, the female data just fail to reach statistical significance (p = 0.079). Nonetheless, modelling still supports the existence of two underlying phenotypes for both sexes (table 2), with low 2D : 4D males making up a larger proportion of the male distribution, and the female 2D : 4D phenotypes being more evenly distributed.

    4. Discussion

    This study is the first, to the best of knowledge, to show statistically that both men and women exhibit two reproductive phenotypes of varying proportions. This would seem to provide a principled explanation for the fact that humans always appear midway between monogamous and polygamous species on all anatomical indices of mating system. Overall, our results suggest that the proportional split in males slightly favours an unrestricted (short-term) mating strategy, with a 57 : 43 split on average for the three datasets, whereas females have a reversed split (47 : 53). However, the mixing proportions in the 2D : 4D digit ratio dataset suggest that a slightly higher proportion of the unrestricted phenotype is present in both sexes (males approx. 62%, females approx. 50%). Note that although males are, overall, more unrestricted than females in all three datasets (as has widely been reported to be the case [4,20]), there is in fact considerable overlap: unrestricted females have more extreme (i.e. more promiscuous) indices than restricted males in each sample. This would not have been predicted on conventional views of human mating style. Of course, while the limitations of the available statistical tools have obliged us to approach the data in the way we have, our analysis does not formally allow us to determine whether the phenotypes we identify represent two separate subpopulations, each with their own normal distribution, or a single population with two modes. Deciding between these, and other, options will require further biological study.

    While it has been widely suggested that males divide into two mating types (‘cads’ versus ‘dads’ [8]) and there is some evidence for a genetic basis for this distinction [11], this study is the first, to the best of our knowledge, to provide quantitative evidence on their proportional distributions in natural populations. More importantly, it is the first to suggest that a similar partition may also exist in females. Although the genetic variation underlying sociosexual behaviour in a female twin population had been previously found on visual inspection to be clearly bimodally distributed [10], quantitative evidence for distinct phenotypes underlying such bimodality has been lacking.

    The statistical method used here assumes that underlying mixture distributions are normally distributed and does not rule out the possibility that the two phenotypes are skewed or represent two separate peaks on a single underlying distribution. Unfortunately, it is not possible to investigate alternative statistical distributions: the near-infinite number of possible permutations and combinations involved makes statistical analysis impossible. These methods, however, are still robust enough to point to the existence of alternative mating phenotypes in both sexes, and it is likely that these have distributions across the phenotypic continuum (in all likelihood reflecting the fact that they are predispositions rather than categorical types). More importantly, which of these alternatives is biologically the case does not affect our claim that statistical analysis of three separate datasets reveals that each sex seems to exhibit two different phenotypes in roughly equal proportions.

    There has been some debate concerning the distinction between trait and type views of personality dimensions [21], with some evidence to suggest that what have previously been seen as types (e.g. extrovert versus introvert) are in fact part of a trait continuum where differential binning of data can create the illusion of two underlying mixture distributions. Our analyses are inevitably subjected to the same risk, of course. However, we rest our claim not on the way we cast the data as a histogram or, as in the case of personality types, on particular theoretical preconceptions, but rather on a purely statistical method based on a quantitative approach to the data that we use simply to demonstrate that the data are not best described by a unimodal normal distribution. Exactly what this means in terms of the underlying biology we leave to future research to discover.

    Accepting our analyses as offering at least prima facie evidence for the existence of distinct mating phenotypes in the two sexes prompts a number of predictions for future investigation. If the two phenotypes essentially represent stable and unstable pair-bonding predispositions (see Walum et al. [11]), we might expect there to be some tendency for assortative mating between the phenotypes. We might also predict that stable–stable pairings are less likely to divorce than other pairings, with unstable–unstable pairings having the shortest durations. The existence of two phenotypes raises a number of further evolutionary questions. One is whether there are within-sex fitness differences between the two strategies. There is some evidence to suggest that reproductive success is linearly related to 2D : 4D ratios, but in opposite directions in the two sexes [14], but whether this is enough to drive the evolution of such a pattern has yet to be determined. While it is possible that the four-way division is an evolutionary stable strategy (ESS) and in evolutionary balance, it is also possible that the distribution is inherently unstable, because the two sexes are in conflict over the optimal balance between mating and parental investment.

    Finally, we noted above that 2D : 4D ratios (a biological marker) are slightly more biased towards a promiscuous (unrestricted) strategy than the SOI index (a psychological behavioural index). While the magnitude of the difference is small in each case, the discrepancy suggests that mating strategy inclinations might also be subjected to a modest degree of cultural modification. Previous research has found that female sociosexuality is more responsive to environmental shifts than male sociosexuality [4,22], and our data confirm this: while both sexes exhibit a shift (towards a restricted strategy in males, but towards unrestricted in females), the magnitude of the shift is larger in women than in men. While there is strong evidence that additive genetic factors best predict adult sociosexuality [23], differences in behaviour are in part likely to reflect cultural or environmental fine tuning of underlying genetic strategies in response to local circumstances as each sex tries to maximize overall fitness.

    Ethics statement

    Research on human participants was carried out with ethical approval from the University of Oxford’s Ethics board (CUREC) and Liverpool University local ethics committee, and participant consent was obtained.

    Acknowledgements

    We thank Professor Alison Etheridge for statistical advice. The data are available as the electronic supplementary material.

    Authors’ contributions

    R.W. was responsible for idea conception, study design, analyses and manuscript preparation, R.D. contributed to idea conception, analysis interpretation and manuscript preparation, J.M. supplied the 2D : 4D dataset and contributed to manuscript preparation.

    Funding statement

    R.W. and R.D. are supported by a European Research Council Advanced Grant to R.D.

    Conflict of interest

    The authors declare that they have no competing interests.

    Footnotes

    © 2015 The Authors. Published by the Royal Society under the terms of the Creative Commons Attribution License http://creativecommons.org/licenses/by/4.0/, which permits unrestricted use, provided the original author and source are credited.

    Reply

Leave a Reply

Your email address will not be published. Required fields are marked *