Behind any eyes

In the course of teaching, one good sign that arises in various course designs is the comfort and curiosity that leads people to ask certain questions. Meaning, the kinds of questions that they are pretty sure the course subject avoids or denies or is simply wrong about, or the ones that have always led them┬áprivately to wonder “how” with little faith that anyone could or would answer.

A nearly guaranteed topic for these is human eye color. How is it, they ask, that if eye color is “genetic,” that people can have eyes of different colors? Or that change through the course of life? Or aren’t “brown” exactly? Or … (the list goes on). Crucially: the student asks in such a way as to indicate challenge, by which I don’t necessarily mean hostility, but rather an underlying assumption that this feature is observationally different from what “biology says.” Eye color defies what they think genes are supposed to be doing.

This is an important issue in the interfaces among education, popular understanding or lack thereof, and actual biology, but to address it, let’s first take a look at mammalian, specifically human eye anatomy and composition.

  1. There’s the surface of the iris: smooth or rough. A smooth iris transmits light; a rough one reflects it, and as it happens, the particular sort of roughness we’re talking about reflects blue.
  2. There are the pigments: pheomelanin (red) and eumelanin (brown). Either or both can be present.

Eye color is the result of any degree of surface roughness with any degree of either pigment – wait, quick amendment to that: apparently, pheomelanin has an upper limit in density for eyes such that we don’t see red-pigmented irises, so think of it as various degrees of yellow or gold). But for example, a light density of pheomelanin plus a rough surface obviously reflects both yellow and blue, respectively, for the combined effect of green. Unsurprisingly when we are talking about combining yellow, blue, and varying degrees of brown, the possible nuances of “brown” are legion. Hence honey, amber, hazel, black, et cetera et cetera, with no Hollywood lenses necessary.

If one has smooth-surfaced irises with no pigment at all, they are transparent, so the blood vessels behind the pigment reflect red – this is why albinos have either blue eyes, due to the surface of the iris, or pink-to-red eyes, which are not otherwise observed in humans.

So … how about inheritance? How many genes? Yeesh. There’ve gotta be dozens, and probably several in sequence for any particular effect. Let’s dispose of one problem immediately: the whole blue/brown, dominant/recessive thing for eyes is simply arrant nonsense. I don’t know why or how it got started as standard pedagogy, but it ranks high in the annals of flat-out lying to students. Simplification is one thing; falsehood is another.

In reality, eye color is just like many other biological variables with multiple genetic influences: its inheritance is consistent enough to start thinking it’s fixed, until it isn’t. It’s a good example of how human experience (what you see, what you “know”) is statistically innumerate Yet again, high significance + low correlation coefficient – I should post more about that some time.

I’ll also hold off on discussing eye pigmentation vs. skin pigmentation, and for that matter, hair – the only thing to say for now is that all three show geography-specific variation, but also that there’s no physiologically necessary association among the developmental processes of each, and if there are ties among those processes, they are multi-variate and not absolute.

More nuances:

  • Different in different lighting: Obviously. This is such simple physics it bears no further comment.
  • Different via association/contrast: Also not surprising, although the mechanisms are more interesting – the perceiving brain may seek matches among fields it strongly associates (i.e. located on the same person) and adjust the input accordingly.
  • Different left/right: Yup. Slight asymmetry of all sorts of left-right bodily features is the norm, not the exception, and occasional past-norm instances are common. (No, David Bowie did not have one brown and one blue eye; he had a non-color-related condition in which his pupils were of differing sizes.)
  • Different colors for eyes as one matures: Again, so common as to be the norm, especially for pigmentation. My hair was white-blond as a little kid, quite gold through my teens and twenties, and straightforward brown ever since.
  • Different fields on the iris, sometimes “wedges” or flecks but very commonly central heterochromia, for a “ring” or “star” effect surrounding the pupil. Sometimes the result for the viewer is a gestalt, as their brain constructs a color which isn’t actually produced by the eye, and sometimes the two are perceived as such.

Almost like this but whiter around the pupil and without the dark-pigmented perimeter.

For instance, my eyes are darkish (“steel”) blue but with a lighter, essentially white field immediately surrounding the pupil. Both of my brothers and all three of my kids have the same eyes, as do a number of relatives on our father’s side. At the risk of being dramatic, people don’t like them very much; clearly the ordinary attention that people pay toward the direction of others’ gazes (which is easy to see in humans anyway, due to our small irises) is triggered by the extra visibility of the pupil. It takes very little actual looking before someone gets offended at “staring” or the “way” they feel they’re being looked at. I wore thick glasses as a kid and switched to contacts as soon as I could, but a few years later, I went back to glasses (fortunately tech had now made thinner lenses possible), because I was tired of people responding to casual eye contact with genuine primate threat-displays, exactly as if I’d just done the same to them.

You see what’s revealed by the questions, right? Even putting the nonsense about Mendelian blue/brown aside, somehow, people have been getting the idea that because eye color is “genetic,” it must be fixed in inheritance pattern, immutable per person, invariant across people, and essentially tied to that person’s identity. This isn’t about eyes but about what people think the word “genes” means. It’s tied to the profound failure to teach (not to understand, to teach) that actual bodily features come from development and physiology, not genes. Genes are the inheritable component thereof, not the “core” or “essence” thereof.

I call special attention to what any bio teacher grapples with, that ordinary critical thinking is short-circuited – no one thinks it odd that hair color is an inheritable feature and also undergoes personal changes, but somehow for eye color to do so is jarring and “non-genetic.”

I categorized this post as “Talking it over” for a reason. The crude and counter-factual thinking is so deeply embedded that it’s impossible for people to state – why should they or how can they, state “the obvious?” It’s only revealed as assumptions behind questions. Ask me questions about genes, specifically for things which seem grossly physically obvious. Let’s get into this right here in the comments.

Links: The Tech

Next: So uplifting

9 thoughts on “Behind any eyes

  1. Is it the same with handedness? Multiple genetic influences, consistent enough to start thinking it’s fixed until it isn’t? My birth family has a mix of handedness that doesn’t line up very well with the right/left, dominant/submissive thing I was taught, and I always kind of wondered why.

    Liked by 1 person

    • Excellent example, and yes! There must have been some over-enthused push to shoehorn anything into Mendelian form, I’m thinking in the 1940s, when scientific texts underwent a big change and most of the modern currcula received their basic structure. One reason might have been the super-win with sickle-cell, which does follow that inheritance pattern and was one of the first genes-to-evoluition-to-ecology-to-medicine science moments in the context of Cold War U.S. grant culture.

      Um, I think you meant “recessive” not “submissive” … (oy)

      Frankly, almost nothing displays the AA/Aa/aa inheritance pattern. The few that do are well-known not because they’re biologically typical, but because they’re logistically easily identified and studied. To be such a trait, the following have to be the case: (i) high, high penetrance on the phenotype for that one gene, and positing that all the other relevant genes are constant across the individuals we’re looking at; (ii) one allele of that gene which “does” and one which “doesn’t,” which is why the misnamed “dominance” effect occurs with just one allele of the first type; and (ii) a visible effect of “does/doesn’t” but not enough mishap to the phenotype to kill the critter or obviate its development. All three of these are merely points on three independent spectra of actual features.

      On a slightly related note, I believe it was in Morgan’s lab (a century ago exactly) in which the term “mutant” became confounded between “rare variant, discovered by combing acres of individuals” and “novel phenotype due to altered genetic content specifically in this individual.”


        • I thought of one more point that applies well to right/left handedness: that some features develop to specific forms with no distinctive determinant factor as to which, only that they will emerge in some way in a particular range. Fingerprints are a good example. In such situations a given “way” may be the majority or even be considered the default in strictly statistical terms, but for every single individual it’s a matter of “dunno, see how it turns out.” Which is to say, although the trait is absolutely genetic in the sense that without genes, it wouldn’t happen, but the specific forms or outcomes cannot be traced to a specific genetic instruction – because there isn’t any; the genes’ input ends with “get it done” and stochastic tissue outcomes finish it off.

          Somewhat controversially, I think that this how gender preference for sexual contact works, insofar as that’s even a “trait” at all.


  2. Somewhere early on, I heard the joke/story about the student who suspects the professor is glossing over important details. “But professor, isn’t it actually more complicated than that?” The response: “Of course it’s more complicated – it’s ALWAYS more complicated. I’m not going to waste time reminding you all of the obvious!” But sometimes we do need to be reminded, and (as this blog does a great job of reminding me) sometimes there are details in HOW it’s complicated that are well worth noting.

    As far as genes and physically obvious stuff: growing up, I happened to know a number of folks who were missing a limb. From accident, disease, and one guy who explained it was “genetic.” I’m pretty sure he did NOT mean anything about heredity in saying that, so … variation? Is there a preferred/rigorous way to talk about what we mean when we say “genetic”?


    • Re: your more abstract point to start, it’s always a matter of accuracy. Scientists are understandably fixated on precision, and one subset thereof is completism. To teach, obviously, one must decompress a little and give up precision in favor of effective pedagogy. Getting fellow faculty to do that has been an interesting struggle over the years, and a lot of them have a fetish about “but but it’s not complete” as if that would be the collapse of western civ or something.

      The trouble is the reverse: losing accuracy. My argument has always been that more precision adds nothing to accuracy unless you’re focusing on accuracy for its own sake anyway. And if you’re doing that, then whatever you establish is a valid platform for later precision, so relax about it.

      Meaning, too, that way too much scientific teaching traditions are the worst: diminishing accuracy while claiming it’s “simplification” (no, falsehood is not simplification), and piling on precision while claiming it’s necessary for accuracy (when it’s not).

      Re: your point about “genetic:” yes, that is the question. Does the term actually mean anything considering that to be a creature at all, genes are obviously involved? Is there any such thing as “not genetic,” considering that no gene ever did a thing, at all, without an environmental and material context in which it was activated and upon which it acts? (And too, that “acts” is a brute chemical mechanical thing, as far as a given gene is concerned; genes do not behave.)

      I like the way I put it: creatures and their features are a matter of development and physiology; genes are the inheritable components of those precise things. The degree of inheritability varies. The degree of penetrance, per gene, varies. The degree of variance of the outome, per gene, varies. Constricting the meaning of “genetic” to the most invariant for all of those things is patently absurd.

      I think most of the discourse is trapped in very old and counter-factual constructs concerning choice and social responsibility, and embedded in unspoken argument concerning what actions or identities are or are not going to be socially attacked (which includes physical harm). With material like this loading into the discussion, there’s not much hope that the actual thing being discussed will see any light at all.


      • Based on the blog entry, I’m not sure it really IS useful to talk about eye color in dominant/recessive terms, but if it is … sure, some precision about how that is only ONE PART of the equation seems necessary. I’m not sure I ever got that caveat from a teacher, so I point to a “it’s ALWAYS more complicated” baseline to explain why the explanation in this post didn’t utterly shock me.

        Checking on ‘genetic’: My childhood buddy had, as a fact of his physiology (and assuming I understood what he meant by “genetic”, probably as a consequence of his in utero development), one leg. Knowing the degree of inheritability – does the fact that it happened to him change the chance of it happening for his offspring? – is understandably of interest. But depending on specifics, his kids might have more, the same, or less chance of having one leg, for the same reason it happened to him or (some small chance) entirely different reason(s). Knowing that his one-leggedness was genetic doesn’t tell us much about the degree of inheritability. Knowing that it happened in an accident at age 12 – that WOULD (probably? certainly?) tell us that the degree of inheritability is zero/really, really low. Hmm … if I’m on track, I really like the idea of focusing on the degree issue, across inheritance/variability/etc.


        • There’s a useful term: “genetic anomaly.” It can refer to an allelic variant which happens to be rare, to an effect due to multiple and probably not-well-understood genes, to incidents likely enough to have a specific outome and name (e.g. Down’s syndrome), and to any developmental outcome which is unexpected (or rather, rare) for a given (or any) genetic profile. In other words, not a causal term at all. It’s a shame that that’s why it’s useful, because it’s a good “settler” in a conversation when over-causation is the driving assumption.

          I don’t know your friend’s medical condition, obviously. But the effect you describe is not a variation in the sense that attached/unattached earlobes are variations, highly penetrated by the allelic profile of a single gene. There are literally dozens of known causes that one might investigate, some of which do involve specific genetic (i.e. coding) and some of which don’t.

          However, I’m going to suggest the term “genetic” in this case is a cultural phrase. It means, “didn’t have an accident,” and “couldn’t help it,” and is a techno-ed stand-in for the older phrase, “born with it.” It communicates important social information to the listener, situates the speaker relative to them in terms of personhood, and provides a hip-pocket narrative for these points to stand on.

          That narrative needs one thing: to distinguish between fetus and infant, at the moment of birth. “Genetic” in this sense doesn’t have anything to do with genes in the sense of DNA, RNA, and proteins. It means “happened as a fetus” and therefore “don’t kick in your exclusion mechanisms.” It is literally a certificate for inclusion in a social world in which anatomical incompleteness is often grounds for exclusion.


        • (In reply to Ron)
          Yes, I think you nailed it re: genetic as primarily a social phrase here. I guess I was wondering if it was ALSO accurate in the sense of variation or the like, but “genetic anomaly” does seem to best capture what I think he meant, scientifically. It seems like a good term – maybe I can say a good explanation, without implying it provides a cause? Regarding eye color … is there an outcome best labelled “genetic anomaly” for that?

          Ideologically, I’m all for tools that mitigate exclusionary mechanisms. OK, that’s over-simplification – extreme example, exclusion of serial killers would be just fine by me. But Bill was a great guy, I’m glad this post reminded me of him. Thinking about him and exclusion, while no doubt accurate/appropriate … I hope the certificate worked!


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s