When does heritable low fitness need to be explained?

Epistemic sta­tus: spec­u­lat­ing about things I’m not fa­mil­iar with; hop­ing to be ed­u­cated in the com­ments. This post is a ques­tion, not an an­swer.

ETA: this com­ment thread seems to be lead­ing to­wards the best an­swer so far.

There’s a ques­tion I’ve seen many times, most re­cently in Scott Alexan­der’s re­cent links thread. This lat­est var­i­ant goes like this:

Old ques­tion “why does evolu­tion al­low ho­mo­sex­u­al­ity to ex­ist when it de­creases re­pro­duc­tion?” seems to have been solved, at least in fruit flies: the fe­male rel­a­tives of gayer fruit flies have more chil­dren. Same thing ap­pears to be true in hu­mans. Un­clear if les­bi­anism has a similar ae­tiol­ogy.

Obli­gate male ho­mo­sex­u­al­ity greatly harms re­pro­duc­tive fit­ness. And so, the ar­gu­ment goes, there must be some other se­lec­tion pres­sure, one great enough to over­come the dras­tic effect of not hav­ing any chil­dren. The com­ments on that post list sev­eral other pro­posed an­swers, all of them sug­gest­ing a trade­off vs. a benefit el­se­where: for in­stance, that it pays to have some pro­por­tion of gay men who in­vest their re­sources in their nieces and nephews in­stead of their own chil­dren.

But how do we know if this is a valid ques­tion—if the situ­a­tion re­ally needs to be ex­plained at all?

For ob­vi­ous poli­ti­cal and so­cial rea­sons, it’s hard to be sure how many peo­ple are ho­mo­sex­ual. Note that we are in­ter­ested only in obli­gate ho­mo­sex­u­al­ity—bi­sex­u­als pre­sum­ably don’t have strongly re­duced fit­ness. The Wikipe­dia ar­ti­cle doesn’t re­ally dis­t­in­guish obli­gate ho­mo­sex­u­al­ity from bi-, pan- and even trans-sex­u­als. The dis­cus­sion in the SSC com­ments used an (un­sourced?) range of 1%-3%, which seems at least con­sis­tent with other sources, so let’s run with that.

The rate of ma­jor birth defects in the US, as re­ported by the CDC, is also about 3%. This counts both de­vel­op­men­tal and ge­netic prob­lems, and in­cludes ev­ery­thing from anen­cephaly (in­vari­ably fatal) through Down syn­drome (se­vere but sur­viv­able) to cleft palates (minor). But most of these, at least 1.5% of births, were always fatal be­fore mod­ern medicine, and many of the oth­ers re­duced fit­ness (via mate se­lec­tion, if noth­ing else). Var­i­ous other defects and dis­eases, which only man­i­fest later in life, are also thought to be in­fluenced or de­ter­mined dur­ing early de­vel­op­ment. And so is sex­ual prefer­ence.

(Whether ho­mo­sex­u­al­ity is a de­vel­op­men­tal di­s­or­der is not the point; I’m com­par­ing the effect of se­lec­tion pres­sure on fatal ter­a­tol­ogy with its effect on re­duced-fit­ness ho­mo­sex­u­al­ity.)

Em­bry­olog­i­cal de­vel­op­ment is a com­plex and frag­ile pro­cess, and there are many ways for it to go wrong. We don’t won­der how it is pos­si­ble that se­lec­tion pres­sure al­lows anen­cephaly to oc­cur in 1 in 4859 births. There are cer­tainly di­rect causes of anen­cephaly, ex­pla­na­tions of why it hap­pens when it does, but (I think) we don’t a pri­ori ex­pect them to be due to trade­offs yield­ing benefits el­se­where. It’s just as plau­si­ble that the trade­offs in­volved are against even worse (coun­ter­fac­tual) prob­lems el­se­where—or that there are just no available mu­ta­tions that don’t have these or equally se­vere prob­lems.

Could it be that link­ing sex­ual prefer­ence to the biolog­i­cal gen­der is, for some com­plex de­vel­op­men­tal rea­son, frag­ile enough that it goes wrong de­spite all se­lec­tion pres­sure to the con­trary, that it has no re­deem­ing qual­ities from the view­point of evolu­tion, and that is all there is to it?

When faced with any phe­no­type with re­duced fit­ness, how can we judge if there is some­thing to be ex­plained—a benefi­cial trade­off el­se­where to search for—or merely a hard prob­lem evolu­tion couldn’t solve com­pletely? And is there a way to quan­tify this ques­tion, re­lat­ing it to the known math­e­mat­i­cal mod­els of ge­net­ics?


1. I’m post­ing this in the spirit of re­cent sug­ges­tions to post more and ac­cept lower qual­ity of (our own) posts to Dis­cus­sion.

2. I’m go­ing to sleep now and will start re­ply­ing to com­ments about 10 hours from now; sorry for the in­con­ve­nience.