[rirze]

I personally dislike the definitions here-- traits and genetic changes are, by default, unattributed to external pressures. They don't "arise" from the environment, rather the environment selects for traits through survival/reproduction, if applicable.

That means, if a trait is unaffected by the environment, then there's no attribution of why it developed; this is the default state.

I feel often in biology, there's a mindset of figuring out why features developed and that's great for pushing the field but it runs into a limit. Not every trait developed from a environmental pressure.

[andrewflnr]

Technically yes, kind of, but if a feature is truly unaffected by the environment, then there's no particular reason it would get fixed, and you have to make the rather shaky assumption that it became widespread purely by chance. Most features have some cost, and the simplest explanation for that feature being widespread is that it has some counteracting benefit that leads to it being selected for. That's a reasonable first hypothesis, to say the least, which means the "spandrel" hypothesis makes sense as the less common special case.

[rirze]

> Most features have some cost

I don't agree and that's totally fine-- we have millions of gene edits from parent to child and not every change is environment-selected. Certain features can be dominant by the mechanics of genes, not due to selective pressure.

Blood types are an example of this-- they aren't major environmental pressures for why one blood type is more common than others, it mostly comes down to population mechanics.

My belief is that "spandrel" features are less selected for studies because they have a harder burden to prove; there exists no external reason they exist and this must be verified through proof by contradiction. Its a high bar to prove.

[andrewflnr]

Blood types aren't fixed across the population. It's kind of a tricky question which of our points they make a better example for. :D

But also, I guess by "feature" I tend to mean "noticeable feature". You might be more right about that, at least as far as terminology goes. At the bottom end you have a point mutation to a codon for the same amino acid, and then eventually you get something like a slightly different color pattern. Those are probably free.

But as soon as you have to, say, add or increase the size of a physical feature, there's a metabolic cost. And when those features aren't used, that (small!) cost drives them to disappear in relatively short order (for instance, all the convergently legless lizards). That strongly implies that for those, there needs to be a persistent reason for them to keep existing.

Entirely aside, I wonder how long we need to have blood transfusions as part of society before blood type compatibility starts exerting selective pressure.