Extract of flowchart of female reproductive agency (for full schematic and explanatory notes, see below)


This post aims to reconcile conflicting aspects of economics and evolution. The intention is strong, testable predictions about human conduct, both individual and collective. 

Why bother? Because after millennia of effort, humans still struggle for comprehensive models of our behaviour, either individual or collective[1]. Only two fields claim to offer accounts which are broad enough to be interesting and which also lend themselves to Popperian falsifiability, that is testing with data. Unfortunately, neither is satisfactory. 

  • Economics and its subsidiary disciplines in finance, games theory etc, make robust predictions, in principle meeting the test of falsification by measurement. The trouble is that their concepts can be inconsistent, and their subject-matter is too unstable to be reliable and too narrow to be generally applicable. Nonetheless they offer useful tools relating to agency, information and risk. 
  • Darwinian biology is comprehensive, supported by evidence and testable, but is indifferent to - if not, at odds with - the thrust of economics, by denying individual agency and bearing only obliquely upon information and risk. In addition, evolution’s central assertion regarding individuals, that they are given to reproduce, offers only the disappointing irony of scant further fruit. 

Any such reconciliation must overcome the near two centuries of justifiable zealotry, which has seen off past attempts to modify the central element of Darwin’s model, whereby random mutations in genotype are selected only by the brute fact of survival. The zealots have been right to stand their ground: all such attempts have erred, either as whimsical in introducing advertence, generally divine, into the evolutionary process, eg, the various flavours of “intelligent design”; or as conflating genotype with phenotype, eg, Lamarck, Lysenko.  This post aims to avoid such errors[2]

Introduction to the issue

For countless generations, our species needed to understand other animals. First, homo sapiens had to survive predators and hunt game. Then, our forebears turned to domesticating animals into sources of clothing, energy, milk, meat, other raw materials and the variety of duties assumed by dogs. Eventually breeding, drafting, herding, milking, riding and shearing animals became a matter of course. At times, selected species have been venerated; others are routinely butchered. Nobles took the skills of mastering a warhorse as a guide to the leadership of men; priests took the responsibilities of a shepherd as a metaphor for the love of God. Even on earth, shepherds could prosper: at best they might end up a king like David of Judea. 

Things change. Dealings between homo sapiens and other animals, for most of human existence at the heart of everyday life, have been reduced in our era to the small number of those involved with agriculture and traditional ceremonials or competitions, together with the larger number of pet-owners. Talent with animals, once commonplace and essential, has been overtaken by skills with ideas, machinery or other people. How has this happened?

Any explanation must respect the evolutionary principle that characteristics acquired during life cannot be transmitted genetically. The argument of this post is that even so, the incidence of talents in the gene-pool may alter if their underlying genetic features enjoy reproductive preference. The twelve conjectures below explore the implications of such preferences, as stemming from the exercise of female agency.

Genetics and human reproduction

We start with the standard account of the relationship between genetics and human mating[3]. What men do doesn’t matter that much: by default, they select their reproductive partners for perceived attractiveness, a surrogate for expected reproductive success. Men may follow the conventions of the day, with the template of Beyoncé Knowles succeeding that of Kate Moss, in turn succeeding that of Marilyn Monroe. Alternatively, the criteria may be hard-wired, eg, clarity of complexion and facial symmetry as signs of health, or the ratio of hip to waist to indicate fertility. Conventional or hard-wired, the signals of health and fertility vary randomly with other genetic features, preventing men from playing a leading part in selection for the latter. In addition, men are unimplicated in the narrowly biological aspects of child-rearing. In the absence of cultural restraint, this combines with men’s relative indifference to phenotypes, to encourage the promiscuous mating which maximises reproductive certainty for an individual male.

It's different for women. In much of the past, their direct implication in childrearing has restrained their economic capacity. They are also physically slighter than men: if unsought reproductive overtures turn violent, women and their children are at risk. By default, women take it that they will do best with men who will recognise, stick with, defend, and for choice play a leading part in providing for their reproductive partners and the offspring arising; and who are competent to do so. The argument of this post is that this gives women a vital interest in male phenotypes, that is the individual expression of genetic features, for example qualities, eg, intelligence (subject to the limit of note 10), stature, or strength; aptitudes, eg, articulateness, dexterity or horsemanship; and traits, eg, impulse control, persistence, or reliability[4]

Thesis: conjectures (1) and (2) and "virtuous circle"

Now for a full statement of the argument of this post[5]

Where at liberty to do so, women practice selective mating for preferred phenotypes (shortened below to “selective mating”), guided by their judgement of the future, with the effect of inadvertently reinforcing the underlying genetic features. Over the span of many generations, selective mating may improve the reproductive outlook, sometimes also fostering cultural advance, all once again inadvertently. It should go without saying that selective mating largely occurs instinctively and unconsciously, rather than through a painstaking application of evidence to decision-making.

The argument originates with the premise that women face existential dilemmas in identifying satisfactory reproductive partners. For most of human existence, the future has been subject to turmoil and apparently insusceptible to individual agency, impeding the specification of men’s general and reproductive competence. Even today, it is tricky to know whether the challenges in store might best be met by persistence, piston, or prayer. On its face, reliability is always to be prized, but it might be trumped by aggression or charm. This leads to the conjecture that 

(1) in conditions of turmoil, women can generally do no better than to choose as randomly as men, preferring the certainty of successful reproduction to the uncertainty of seeking out a partner for preferred phenotype.

Particularly in recent centuries, reduced turmoil and improved agency enable women to make more accurate and frequent selections for preferred phenotype, giving rise to the conjecture that 

(2) the inadvertent consequences of these choices may enable women to make further such selections, giving rise to a virtuous circle.

These conjectures are set out schematically below. 

Notes on schematic

  1. the outlook for stability  means the (instinctive and unconscious) view of a female enjoying reproductive agency, about the certainty with which she may specify phenotypes competent for reproductive partnership.
  2. the outlook for lifetime agency  means the female’s view of the prospect for herself, her partner and her offspring to affect the circumstances of their lives.
  3. identifiably competent partners  means male phenotypes, viewed with sufficient clarity for specification, as favourable for foreseeable conditions, in sufficient number for a realistic prospect of selection.
  4. improvement in the reproductive outlook  means any or all of greater stability, lifetime agency or numbers of identifiably competent partners.
  5. cultural self-determination  means the command of resources and skills sufficient to maintain the prevailing culture. 

The right-hand section of the schematic captures the conditions whereby the reproductive choices of women improve recursively but inadvertently as levels of uncertainty diminish, the prospect for individual agency improves, the accuracy of specification of competent reproductive partners improves and their numbers increase. Thus, the virtuous circle of selective mating, set out in blue in the schematic. In this model, advances in cultural self-determination are an inadvertent but nontrivial subset of the virtuous circle, leading to conjectures (3) to (9) below, as amplified in note 7.

Conjectures (3) to (9): reproductive preference and cultural self-determination

The least ambitious approach for an individual woman is negative preference, the avoidance of genetic features expected to lead to unsuccessful outcomes. The maths is exponential. The chart below shows that a genetic feature universally present in the gene pool at the outset, but attracting twenty percent fewer reproductive opportunities in each generation (expressed in the chart as a level of reproduction of 80%), attaches to only one fifth of the gene pool after seven generations. If it attracts one third fewer opportunities (expressed in the chart as a  level of reproduction of 67%), it falls to that level after four generations. 

Incidence of feature in the gene pool over eight generations 
at selected levels of reproduction

The most ambitious approach for an individual woman is positive preference, that is selecting for genetic features expected to lead to successful outcomes. Naturally, some women are more insightful than others, some men are accomplished deceivers, and mistakes are bound to be made – about the future, about individual men. Even so, where women are at liberty to select the fathers of their children, a genetically significant fraction of them will succeed in looking forward in just this way. To take the example of the earlier paragraphs, two hundred years ago, a young woman might have found herself wondering if a suitor with a bent for those new-fangled steam engines might not be a better proposition than his rival, the cowman. And after a few generations, that’s it for talent with animals, though mind you a bent for steam engines shortly turned into a dead-end. This argument opens the door to further conjectures about culture, economics and the history of both, to the effect that

(3) selective mating by women has been promoted for centuries in northwestern Europe and its cultural offshoots by the style of courtship prevailing, whereby men vie for the favour of women[6];

(4) the accumulation of the consequent changes in the gene-pool offers an explanatory mechanism for the industrial revolutions originating in the region[7];

(5) where the culture impedes female reproductive agency, adaptation to modernity has been slower, as in Catholic Europe until the last couple of generations and much of Hispanic America to this day; 

(6) the prolonged maladaptation to modernity of Moslem cultures occurs because their closeted women are prevented from exercising reproductive agency; 

(7) where arranged marriages prevail, as in religious enclaves, the subcultures arising are given to ossification;  

(8) conditions of acute change are perceived by women as turmoil, deterring them from specifying competent partners or adding to their perceived rarity, leading women to revert to their conduct as conjecture (1) above, selecting as randomly as men; and 

(9) the descendants of enslaved populations bear the cultural residue of the loss of lifetime agency during the generations of bondage. Thus, transported Africans and Russian serfs. Recent generations of the former also bear subcultural turmoil following migration to cities. This has led a genetically significant fraction of women to choose reproductive partners as randomly as men, reducing the competitiveness of subculturally endogenous phenotypes against the general population; and failing to alleviate - at worst intensifying - subcultural maladaptation[8]

All the conjectures set out above lend themselves to testing with data[9].

Brief comments re evidence

Persuasive evidence of the mutual relationship between the reproductive interests of women and culture comes from the accelerated pace of development since the agricultural revolution. This follows the congruence since then between the standard account of human mating and a division of responsibilities between the sexes, whereby women take principal responsibility for childrearing as do men for economic support, within a cultural context promoting sexual fidelity. Anthropologists tell us of other approaches to the reproductive project: extended families, mutually supportive communities of women and much else, but these attach almost invariably to ossified hunter-gatherer cultures. 

On the other hand, we see that the division of responsibilities between the sexes prevailing since the agricultural revolution is now altering in our own, post-industrial, culture. Even so, selective mating explains modern complaints about lack of reproductive opportunities: “where have all the good men gone?”; “why can’t I get laid?” The women are after men whose phenotypes signal genetic features expected to make for good outcomes, always uncertain and all the more so in times of change (conjecture (8) suggests a limit). The men are stuck with phenotypes whose dysfunction or obsolescence is evident to the women who care about such things. 

The future: conjectures (10) to (12)

So, what next? Over the last century or so, women have taken up economic agency. This makes it less vital for the women so engaged to subscribe to the default of choosing partners with a view to future outcomes. This leads to further conjectures, at this point possibly less readily testable with data, but still worth ventilating, that 

(10) the continued, if not accelerating, pace of change shows that any defection by women from selective mating has had negligible consequences; 

(11) alternatively, it is offset by an increased incidence of men selecting women who demonstrate economic agency or display appealing phenotypic characteristics other than perceived attractiveness, noting that this impairs the reproductive choices for women less well qualified; and 

(12) as women with economic agency feel at liberty to defect from the default of selecting reproductive partners for such traits as reliability, the incidence of such traits in the gene pool will decline.

For the last fifty years, women in an increasing number of rich countries have been defecting from the reproductive project altogether, bearing insufficient children to maintain the level of population. We cannot know how long this will continue; some applaud it as reducing the pressure on resources. Regardless, policy makers are currently most alive to the transitional problems. Their responses are public intervention, so far either unavailing or disastrous; immigration which, if extensive, is said to challenge the integrity of local cultures; or increased capital intensity, which, if including AI (and why wouldn’t it?), is said to threaten us all. Much of this sounds overwrought but as ever, humanity is on a journey with no maps.

Notwithstanding the preceding paragraphs, we may surmise that where they can, most women will continue to subscribe to the default, at least making decisions about their partners which give rise to negative preference. Genetic features already seen as obsolescent (and likely to become more so) include the impaired intellectual capacity signalled by a limited education, manual dexterity and physical strength. More ambitious women will continue to attempt positive preference. Genetic features already prized (and once again, likely to become more so) include communication skills and mastery of symbols. In the early-adult life which is the arena of courtship, these are most readily taken as signalled by articulateness and credentials. Articulateness can be deceptive; credentials vary in quality but may be taken as more reliable in reducing the uncertainty of specification, accelerating the pace of positive preference, which otherwise suffers from punishing maths, with the phenotypes arising difficult to specify and at the outset rare[10]

Finally on this score, technology. Widespread contraception has enabled women to assert reproductive agency embracing national populations, apparently disqualifying a fraction of men from reproductive opportunities. Fertility treatments involving donor sperm deny selective mating but occur too infrequently to affect populations. Gene-splicing and suchlike promise further to accelerate the effects of selective mating, also removing it from the province of individual women, the level of populations and the timespan of many generations. That cannot but lead to fireworks.

Personal conclusion

This post is the fruit of ideas emerging during the Covid epidemic[11]. I am not the right person to take them further. Although I still write the odd financial paper for think-tanks, I retired from the City (Britspeak for financial services) over twenty years ago and I lack professional standing as a biologist or an economist. Nor do I command the resources to organise experiments or crunch the numbers. In consequence, I present this post as conjectures, in the hope that they may inspire others, better qualified and resourced than me, to take them up. I wish them well. 

[1] For the purposes of this post and with apologies for such rough and ready synopses, my summary of the state of play is as follows. 

  • The Greek philosophers introduced reason and provided taxonomies, some helpful, others stultifying for millennia. Their heirs largely offer amplifications, most interesting to each other.
  • Religions and their latter-day camp-followers, Freud, Marx et seq, provide metaphors under the seductive guise of revealed truth, once again some illuminating, others not so much. 
  • Anthropologists and sociologists assemble data and occasionally present noteworthy concepts, but all fall short of comprehensive or testable science; they are also given to aim off our own culture, respectively preoccupied with the extraneous or the transgressive. 
  • Of late, western historians have fled the pseudo-science of Marx for outright subjectivism, though many continue to echo his underlying critique; literary practitioners and critics follow a similar course. 
  • Archaeologists are now assembling enough data to test their patchy narratives, but in the nature of things they address the past. 
  • Neurology and psychiatry stick to treating dysfunction, mainly with drugs just about meeting double-blind standards, while their underlying subject, the brain, remains a black box to medicine.

[2]  Briefly to expand, this post aims to dispel four complications attaching to the paradoxical union of inadvertency and ex post teleology, or purposiveness, in its account of evolutionary processes.

  • First, the history of failed attempts to modify Darwinian randomness brings any such exercise into disrepute. By contrast, the conjectures of this post deny advertence in the evolution of genotypes, thus working with the standard model, rather than challenging it.
  • Second, any argument engaging with human purposes trespasses upon the sphere of religious faith. This risks upsetting both the devout and those with a principled antagonism to devotion. Partisans might do best to restrain themselves until these conjectures have been tested with data.
  • Third, the heart of the argument of this post is the conduct of women. If the account is well-founded, there are bound to be further wrinkles. Examples immediately coming to mind include the identification of inherent preferences on the part of women, or developments causing men to act similarly. The current intellectual climate for relations between the sexes threatens heat rather than light, so here too partisans might better wait for the results of testing.
  • Fourth, the recursive effect, over generations and populations, proposed in this post, embraces an uncalculated degree of redundancy, by way of behavioural and genetic false starts. Once again, if the conjectures of this post turn out to be well-founded, they argue for deliberation, if not outright reluctance, as to intervention to affect the process, whether cultural, political, or technological.

[3]   The next two paragraphs draw on the analysis originating with Robert Trivers, Natural selection and social behavior, Harvard University, Cambridge, Mass. 1972. https://hollis.harvard.edu/primo-explore/search?tab=everything&search_ scope=everything&vid=HVD2&lang=en_US&mode=basic &offset=0&query=lsr01,contains,003992263. For a later synthesis, see Buss, D. M., & Schmitt, D. P. Sexual Strategies Theory: An evolutionary perspective on human mating. Psychological Review, 100(2), 204–232.9 (1993) https://doi.org/10.1037/0033-295X.100.2.204.

[4]  The evolution of genotypes through the reproductive preference of phenotypic features is observed in humans and other species, as assortative mating. See Robinson, M., Kleinman, A., Graff, M. et al. Genetic evidence of assortative mating in humans. Nat Hum Behav 1, 0016 (2017). https://doi.org/10.1038/s41562-016-001616. For female agency, see the literature on false paternity. The compendium of surveys most cited on the Internet was published in 2005 and plays down the practice, with the abstract stating that, “Rates vary between studies from 0.8% to 30% (median 3.7%, n = 17).” The authors also concluded, “This paper identified a distinct lack of well designed…surveys”. Source: Mark A Bellis, Karen Hughes, Sara Hughes, John R Ashton. Measuring paternal discrepancy and its public health consequences. Centre for Public Health, Faculty of Health and Applied Social Sciences, Liverpool John Moores University (2005).  https://jech.bmj.com/content/59/9/749. By contrast, a more recent double-blind survey of a community of Himba pastoralists in Namibia showed a rate of 48%, with the authors arguing that earlier studies have been skewed for European descent groups. Source. B. A. Scelza et al. High rate of extrapair paternity in a human population demonstrates diversity in human reproductive strategies. Science Advances, 19 Feb 2020. Vol 6, Issue 8.  https://www.science.org/doi/10.1126/sciadv. aay6195. This revives the topic but leaves open the relationship between false paternity, culture and turmoil.

[5]   The balance of this post follows my own line of reasoning.

[6]   If this conjecture holds, it outweighs the anecdotal or sporadic counterexamples of bullying families, coercive courtship, or dynastic alliances, however familiar from campaigning historicism, cautionary tales, or romantic literature.

[7]   My interest in these matters was rekindled by Oded Galor’s The Journey of Humanity, Dutton, New York, and Bodley Head, London, 2022 (see also note 11). His central explanatory mechanisms for economic take-off are the accumulating effects of a parental preference for quality over quantity; and “goldilocks” cultural diversity balancing conformity and chaos. His findings are detailed in The Demographic Transition: Causes and Consequences, Oded Galor, NBER Working Paper 17057, May 2011. Other economists have challenged him, for example Joshua Angrist, Victor Lavy, and Analia Schlosser, Journal of Labor Economics, Vol 28, No 4, October 2010, as to the evidence for quality over quantity; and Steven N. Durlauf, Population and Development Review, Vol 49, Issue 2, 27 June 2023, as to the evidence for diversity, as well as methodology and theory. Many others have weighed in.

It is not clear how to reconcile these debates to this post. Nothing definitively pins down the timing whereby the accumulation of genetic effects reached a tipping point between 1650 and 1750. We may speculate from the description thirty generations earlier, of the reproductive agency enjoyed by the noblewomen among the dramatis personae of medieval romances. Unfortunately, the Dark Age record is too sketchy for us to gauge the novelty of such agency as emerged during the medieval period.

[8]  Similar conditions may apply to maladaptive regional subcultures, eg, the Appalachians or the Mezzogiorno, contributing to an explanation of their multi-generational dysfunction.

[9]  The avenues of enquiry arising out of these conjectures are not exhaustive; others include:

  1. does mating randomly with respect to phenotype lead more to indiscriminate or to promiscuous mating; how otherwise does the conduct of random mating vary as to age, sex, under turmoil, in other conditions?
  2. how does selective mating spread within cultures; specifically, do elites take up selective mating ahead of the population as a whole?
  3. how do the effects of selective mating upon reproductive success bear upon intergenerational competitiveness; is it possible to identify phenotypic features whose selection has made for one, the other, both or neither? 
  4. how do the effects of selective mating upon cultural advance bear upon population levels; once again, is it possible to identify phenotypic features making for one, the other, both or neither? 
  5. how does the history of East Asian mating practices bear upon the region’s unequalled adaptation to modernity?

[10]The maths also challenges the applicability of “general intelligence”, or G, for these purposes. This is because a feature augmenting or underpinning almost every competence is bound to have already attracted positive selection. Let us consider the thought-experiment, whereby the incidence of an elevated level of G in the gene pool, at a hypothesised outset after which homo sapiens bred true, was one-tenth of one percent; and that it subsequently attracted a reproductive preference of one one-hundredth of one percent from generation to generation. Of course, these assumptions of incidence and preference are ludicrous, but they are too small to dismiss. On this basis, the elevated level of G would be universally present in the gene-pool after slightly over 6,900 generations, or some 173,000 years. The conclusions are that either G hasn’t attracted reproductive preference - on its face, far-fetched; or that the normal distribution of G in human populations, breeding true for at least that long, has reached its reproductive limit - behavioural, genetic, or both. 

This means that the dilemma for women persists: effective selection for intellectual qualities must seek to specify the applications or subsets of G which promise competence in variable conditions. This runs into the familiar problems of difficulty of specification and rarity, as well as denying prolonged cycles of positive preference.

[11] It was on 30 September 2020, that I found myself rooting through the nether reaches of my memory, for watchwords which once meant something to me. One of them was “every family is a hypothesis about the future”, which in undergraduate years seemed to explain much. My late-life annotation had it that “this might be one of the smartest things I ever said”. Smart or not, the perspective evidently abides. In the first half of October 2023, I made an initial draft of these remarks, drawing together lifetime preoccupations about the mystery of underperforming cultures or regions; relations between individuals, families and culture; the role of agency; the roots of economic take-off; and a view that evolution has something to offer the humanities, despite the disreputable cul-de-sac of Social Darwinism. I even managed to sneak in a side dish of teleology. 

I now see that these ideas reflect my undergraduate notion, giving rise to this mixed salad of text, graphics, and footnotes, all in all a testament to an idée fixe going back over fifty years. Thus have the bees in my bonnet become the raw material for that most unreliable of intellectual constructs, a single overarching theory. So let me concede pre-emptively that however much there might turn out to be to these notions, there are bound to be many other moving parts.

New Comment
2 comments, sorted by Click to highlight new comments since:

in conditions of turmoil, women can generally do no better than to choose as randomly as men, preferring the certainty of successful reproduction to the uncertainty of seeking out a partner for preferred phenotype.

Not sure if I understand your point, but this just seems plainly wrong. Not sure what qualifies as turmoil, but generally, a strong partner is better than a weak one, a smart partner is better than a dumb one, etc.

Do you have any evidence of women actually choosing their partners randomly?

  1. You are right: other things being equal, strong is better than weak; smart than dumb. 
  2. I haven't defined turmoil and if I'm honest my definition would be a tad circular: ie, (to take this example) conditions where it is impossible for a woman to choose between strong and smart.
  3.  My concept is "as randomly as men", ie for subjective attraction rather than other phenotypic features, where I hope you will agree we hardly want for evidence.