I’ve got to stop reading opinion articles in the (online) paper.
But I haven’t. So here we are.
Painting with a broad brush, for any gross characteristic common to the vast majority of the human species, people can be divided according to the characteristic. Your average human* will possess either testicles or a vagina. Skin darker or lighter than the Chosen Reference Shade. An IQ test result above or below 100. And so on. Each of these are reasonably unremarkable and inoffensive characteristics in and of themselves, which can be observed, noted, and in various cases altered. We will assume that all societally-caused differences have already been negated for this discussion, and deal only with “true” differences.
[* I am well aware that this does not in fact cover all the humans, and that there are plenty of edge cases that can’t be neatly divided into Category A or Category B. That doesn’t detract from the overall direction of this discussion, which is whether or not one should care in the first place.]
At this point, there are two and only two relevant considerations. Either the characteristic is irrelevant – that is to say, the bell curves of other characteristics for members of Category A have significant overlap with the bell curves for the same characteristics for Category B – or it is relevant (in which case the bell curves are a significant distance apart). Of course, all bell curves won’t overlap 100%, so you have to choose a standard measure, say 1σ, as your boundary value. And if I go into that any further we’ll be off on a tangent, so I won’t. In any case, the only thing then worth considering is the consequences of each option. (In an ideal world we would know these. In this messy real world we do not fully know each of these distributions, and so it comes down to “what is the best evidence the person is aware of” and to some extent “which would the person most like to be true”. Also I’ve just assumed that everything follows a normal distribution for simplicity. You’re welcome to replace the graphs with top-weighted or bottom-weighted variants as you like, or to give the two groups different standard deviations, according to what you think is realistic. It doesn’t change the base argument.)
The Characteristic is Irrelevant
In this case, the spread of differences within the group is far larger than the differences between the groups. It is highly unlikely that the A/B characteristic can be used to predict any given person’s attributes.
In this case, it logically follows that the characteristic should be ignored except when it is directly relevant (for example, to one’s doctor, who may well need to take it into account to prescribe the correct medication). All social inferences and expectations based on the characteristic should immediately be dropped.
Since the curves do not perfectly overlap, in a small number of extreme cases it may still appear that people with Version X of the characteristic are significantly favoured. For example, in a sporting population extremely selected for height, it may prove that Category B individuals are notably more common than Category A individuals. This is a natural result of the high/low tail differences of the two distributions (if the high tail of B is even fractionally higher than the high tail of A), and will also periodically appear as a false positive by chance. In populations not so stringently selected, the effect will promptly vanish. This extreme-selection effect is true of all distributions that are not completely identical, and can safely be ignored.
The society we would construct here has a wide variety of people with all distinct A/B characteristics filling all roles in that society according to the unrelated personal characteristics suiting them to the role. Membership of Category A is not assumed to mean anything other than the literal statement in question. Since the categories are irrelevant (other than to the individual and perhaps their doctor), membership of the two categories, or of neither, may be as fluid as each individual likes. All of these conclusions are reasonably logical follow-on effects of this base premise.
The Characteristic is Relevant
In this case, it is clear that the A/B characteristic can be used as a predictor of other attributes. Perhaps of a propensity to violence, or of high compassion, or of physical strength, or whatever. Since this is so, it logically follows that any task for which such an attribute is preferred (for example, physical strength in non-mechanically-assisted construction work) should select preferentially from the category that predicts superiority in that attribute.
It additionally follows that other connections should be formed along the same lines. For example, an individual seeking only high-compassion friends should look for friends primarily amongst Population A. In fact, it might even prove wise in some circumstances to segregate the populations from one another: if population B is of significantly higher violence propensity than Population A, they should probably be kept apart in situations where violence is plausibly expected, as members of B will more probably than not attack members of A, who will be commensurately unable to defend themselves.
Similarly, since we are assuming that we have removed all social influences and therefore these remaining differences are set in biological stone, it would be sensible to prevent individuals attempting to alter which category they are placed in. Taking on the outward appearance of a Category A would not remove the Category B traits that the member of Category B would unavoidably be biologically determined to express. Indeed, it would be outright dangerous in some cases, for example wherein people expecting to find a low-violence-propensity individual instead found themselves interacting with a “disguised” high-violence-propensity individual, and got punched.
The society we would construct here must have two parallel streams for dealing with these two significantly different populations. Different private areas, different allowable jobs (having a Category A working in a role requiring Category B traits would, after all, only ensure underperformance, since even the best Category A would be barely as good at the job as a very mediocre Category B), and so on. Contact between the categories may need to be monitored or chaperoned in cases where the trait of one category is likely to result in negative effects upon the other. The occasional difficult-to-assign individual must either be treated as a third category with its own specific rules and permissions, or assigned to one of A or B. All of these conclusions are reasonably logical follow-on effects of the base premise here.
In the broad-brush sense, these are the only two options. Either a categorisation is relevant, in which case all consequences that logically follow must be taken into account, or it is not relevant, in which case there should be no difference drawn between the categories other than the mere fact of their existence. You cannot, as they say, have your cake and eat it. You can’t argue that A and B are so utterly different that they can’t possibly be allowed to intermingle freely and then in the same breath demand that they be given equal treatment in jobs that select for linked characteristics. And you can’t argue that A and B are essentially the same and then pronounce that A and B should still be, say, physically segregated. If you do either of those things, you are a hypocrite, and should re-evaluate your stance immediately.
Observing others have these discussions (often in circles around one another), it seems to me that most people don’t think these things through. [Disclaimer: it’s not like I can ask. Stepping into a guns-blazing forum fight and demanding everyone lay out their stances in logical format starting from a base premise is something only the forum owner can hope to do and get away with.] Instead they pick whatever collection of ideas “feels right” to them, for whatever long and complicated half-subconscious reasons they may have, without ever, it seems, analysing them for contradictions. I say again: you can’t have your cake and eat it – and you can’t hold two logically contradictory positions at once and not expect to be called a hypocrite. The only logical reason to believe that Category A individuals should be separated from Category B individuals is if you believe that there is a significant difference between members of Category A and Category B – and if you believe that there is a difference significant enough to react to, why then, that difference must be significant enough to react to, and it’s worth, for example, using category membership as a means of selecting job candidates. You can’t have both. Choose. And if you don’t like it, explain why, logically, clearly, consistently, and coherently. I’ll listen, if you can. (It’s not like I’m infallible.)
If you are wondering about my current stance on such things, a combination of my preferences and the available evidence I have seen to date (bearing in mind that I have not done a literature search regarding the correlations of any human characteristics with other human characteristics, and that therefore I am primarily relying on only the sources cited by the people I read, who are primarily those whose opinions I agree with and thus may have provided a selection of sources to favour their stance) leads me to take the position that the majority of these characteristics do not have relevant levels of difference between their bell curves, and that therefore save under extreme selection pressures there is no point or purpose in distinguishing between most human characteristics for anything other than the characteristic itself, which is most commonly the province of one’s doctor. That said, as always, if the evidence comes to show that I am wrong about that for a given categorisation, then I will, reluctantly, update my worldview to suit, and follow the dictates of a situation where the characteristic is relevant. Reality cares not for my emotional desires, after all.