What is the meaning of “should” here?

Exactly. You can’t get an ‘ought’ from an ‘is’, but you can get an ‘ought’ from an ‘is’ and a goal.

So the ‘correct’ number of teeth depends rather critically on what purposes the teeth are to be put to, and in what environment.

And what of design compromises? The human female pelvis, for example, which has to cope with both walking and large-headed babies. The human throat, which allows talking but is suboptimal with respect to keeping food out of the windpipe. Should humans have little toes, which contribute almost nothing measurable to balance?

That merely moves the question one stage back – “Should the human jaw be able to accommodate 32 teeth?”

What is the meaning of “should” here?

In the example given, could we say “humans ought to have 32 teeth”?

Ray,
There’s no need to hypothesize. My wife (like about 30% of the population) naturally has 28 permanent teeth. It’s not clear to me whether she is meant to have 32, according to Food/Skidelsky. We’re currently waiting to see if our kids will (should?) have 28 or 32 teeth.

“If everything goes well for them they should have 32 teeth” seems rather different. That “should” is doing a lot of work. Replacing it with “probably will” wouldn’t alter the descriptive accuracy but would make it look less like a normative statement.

Similarly, “Humans are meant to have 32 teeth” seems a bit fishy. That “are meant to” smuggles in the concept of intent (“are meant to” by whom?) and seems to assume you can get ought from is, rather than demonstrating it. Given the frequency with which wisdom teeth cause problems, one could, I think, argue coherently that humans ought to have 28 teeth, but don’t.

Humans have 32 teeth is a statement regarding frequency of a phenotype. It’s an ordinary descriptive statement, not a normative one, unless you insert an unnecessary “ought” as Skidelsky did. It seems quite different from “humans ought not induce vomiting to eat more,” and I doubt that the morality of the latter statement is derived primarily from a description of the biological function of eating.

You’re perverting the natural function of eating, which is to nourish yourself.

Wouldn’t smoking be a ‘sin’, or at least a ‘perversion’, in some sense, then? I mean, you’re perverting the natural function of breathing, which is to take in oxygen and expel carbon dioxide.

What about drinking alcohol? Where’s the nourishment there? More to the point, wouldn’t artificial sweeteners likewise have to be perverted?

Epidemiologists have the notion of the ‘wild type’ of a virus. Real viruses have lots of alleles, variations of the genes. Gene 1 has variants a, b, and c; gene 2 has variants x, y, and z, etc. The ‘wild type’ is the set of most common variants, say ’1a2y3p…’.

Here’s the thing – there may not be an actual virus that has that particular genome. The ‘wild type’ is kind of the ‘center of gravity’ of a multidimensional ‘cloud’ of viral variants.

If a mutation arose that caused some humans to develop 28 teeth, it seems unlikely that anyone would decide that such people weren’t human. If the gene spread and eventually all humans had 28 teeth, I likewise doubt we would say humanity had been replaced by a ‘post-human’ race.

Humanity has a ‘center of gravity’, too – a ‘wild type’. (Actually, it has two ‘lobes’ of the cloud with separate ‘centers of gravity’ – male and female.) The boundaries of that cloud are fuzzy. Travel far enough in one direction and you’ll find the common ancestor of humans and Neandertals… but you’ll be hard pressed to pick out an unambiguous line you crossed between them.

So saying ‘humans have 32 teeth’ doesn’t seem to have quite the normative force in practice that Foot supposes.