When we look at a drug addict we don't think "poor guy - probably won't have many kids"

That's right. But I was addressing this point:

It is not at all clear that the people resistant to addictive drugs are reproducing at a higher rate than those who aren't.

Since someone had made this comment, and since I was addressing this comment, then it's neither here nor there whether we typically think about the drug addict's reproduction. That comment concerned that question, so regardless of whether it is something we normally think about, it's something that the commenter was thinking about.

, nor indeed would we accept a demonstration of his high reproductive capacity as evidence that being a drug addict isn't harmful.

Well, it's hard to really gather the relevant evidence. To really sort out cause and effect it's really very weak just to gather evidence from the population like that. So, sure, in the practical world we probably should doubt such evidence unless it was really overwhelming (such as a planet-of-the-apes scenario in which the meth heads take over the planet).

We can easily recognize most harms simply by looking - we can recognize when someone is hurt. E.g., they're bleeding, or they're bruised, or they have trouble walking, or they're disoriented, etc. We can tell. So we don't need to do a vast demographic study to see whether the particular harm in question would reduce probability reproduction. Nevertheless, I think we can be sure that it would in fact reproduce probability of reproduction, simply by considering it from an evolutionary standpoint. I don't think there can be any serious doubt that, on average, a recognizable harm does reduce the probability of reproduction (on average, over a large enough population).

I'm quite certain that harms reduce probability of reproduction. This is why I'm quite certain that if drug addicts are harmed by their addiction, this must reduce their probability of reproduction. Obviously, on average. If you come up with crazy scenarios such as one of the commenters did, then in individual cases addiction might lead to enhanced reproduction. For example, if someone puts a gun to your head and tells you they'll shoot you if you don't become an addict, then obviously in this specific situation, becoming an addict will increase your average lifespan and, if you're fertile, will give you more chance to reproduce. But strange hypotheticals aside, I am sure that harms adversely impact reproduction.

Most of the things we think of as harms also interfere one's heartbeat, with one's brainwaves, with one's breathing, with one's digestive systems, with one's sleep patterns, etc, since we're our biologies and every biological function is interconnected with the other.

Indeed, and for this reason, we might be able to measure harms indirectly by looking at heartbeat. A doctor actually does this sort of thing - he looks at your vital signs to see how you're doing.

Nothing I wrote should be interpreted as excluding these other ways of measuring harms. I was talking about one measure, but that doesn't mean there aren't others.

your whole argument [could be read as] trivially true...

Indeed, and I expected it to be read as trivially true, and therefore as the claim I was responding to as trivially false. I wasn't expecting any pushback on such an obvious point. If drug addicts really are harmed by drugs, we should expect this to show up in lower reproduction. It's really quite a trivial point.

I also don't expect they do well in the life expectancy or health or prosperity department, which is a more customary method of determining well-being and "harm" than "number of descendants" is.

Indeed, but these other respects aren't relevant to the claim I was addressing, which concerned the reproduction of drug addicts.

Recall that I referred to reproduction to several generations, the point being that the meth addicts need not only produce babies but also raise them well enough that the babies grow up to reproduce - and so on, for many generations.

If meth addicts really did better than the rest of us in that department, then yes, I would seriously question whether meth addicts weren't actually a superior breed of human. I mean, really, take this very seriously and extrapolate out many generations. After, I don't know, a hundred generations, the whole planet is populated with meth heads, with just small pockets here and there of non-meth-heads. It's a bit like the planet of the apes scenario, but with meth heads instead of apes. In this scenario, yes, I would have to seriously question whether meth heads' addiction is harming them.

But an argument can still be made. For example, it might be that meth-heads, however much better they do than the rest of us, would do even better if they kicked the habit. In that case, they would be better than the rest of us despite their addiction, not because of it. In this case, the addiction would be, possibly, some negative side-effect of an otherwise superior genetic makeup.

Is that enough of an answer or do you want more?

But it will make our intuitions about harm correspond to diminished capability to survive and reproduce well enough. Our intuitions about harm were formed as if for the purpose of ensuring we would avoid impairment to our capability to survive and reproduce, in the sense that our eyes were formed as if for the purpose of seeing.

Darwin then, after explicitly explaining the "as if", made the analogy with a human breeder consciously breeding to a purpose, and then proceeded with anthropomorphic language.

Your argument similarly condemns Darwin.

So, anthropomorphizing natural selection is scientific terminology with well understood meaning among the educated. When using this terminology to the less educated it is necessary to qualify, explain, and clarify, but such qualification and clarification should not be needed among the intelligent and educated.