Lyman Stone's intellectually molested critique of shrimp welfare

*study shows that sky is blue*

Once upon a time, Philosophers were feeling sorry for themselves. “How can I trust my judgements on theoretical ethical questions — the Repugnant Conclusion, moral realism, the infinite loop-loop-loopedy-loop variant of the trolley problem? What if my judgements are out of touch with those of the common man? We must do like the social scientists, and survey the peoples of Earth. We’ll ask them questions — yes, questions! — and tally up their responses with chalk. (*cackles*) To think through these questions, they won’t need decades of reflective equilibrium in pursuit of conceptual clarity, in dialogue with a body of their peers. Good Heavens! They’ll rawdog that shit, hit on the answer right away, make us look like a band of fools. And at last? (*shrieks with sadistic mirth*) At last we shall know the Truth!”

Thus experimental philosophy was born.

X-phi — a discipline named after The X-files — slowly took over the world. Grants flew off the shelves; intuition factories closed down. Those mechanics who sold intuition pumps? On street corners, begging for nickels.

Decades later, in the year of our Lord 2025,

Lyman Stone

uploaded a Substack. When the world had stopped spinning on its axis, and I’d dizzily hoovered the vomit from my bed, I clicked on the blog post in question.

Good grief. I hadn’t been so bowled over since — in 2019, for the time in history — a study showed that the sky is blue. Stone’s graphs were revolutionary — in fact, they grafted into my eyes, bars of blue and yellow taunting me with results of world-breaking importance. A gap in the literature was filled. A 🚨 GAP 🚨 IN 🚨 THE 🚨 LITERATURE 🚨 WAS 🚨 FILLED.

Apparently, having surveyed approximately 450 people from all walks of the internet, none of them cared about the wellbeing of shrimp.

Here is a bit of background, for those of you who are uninitiated into the lore of shrimp welfare.

Meet Homer Shrimpson:

Currently, around 480 billion Homers are killed in shrimp farms. When you factor in wild caught shrimp, the number jumps to 25 trillion, making shrimp — by miles — the most common animal for humans to kill for food.

Unfortunately, it looks like shrimp might feel pain (and pain, famously, feels bad.) Two recent literature reviews — one from 2021, one from 2022 — reviewed the evidence for and against sentience in decapod crustaceans; both reviews concluded that there’s substantial — though limited — evidence that decapods like shrimp might feel pain.

If shrimp can feel pain, many of the farming and slaughter practices we employ are needlessly causing them pain. To take the most salient example, farmed shrimp are typically slaughtered by being dumped — in large batches — into boxes of ice, where they die by suffocation and freezing over the course of about 20 minutes.

I say this practice “needlessly” causes them pain because we could easily change this part of the process by electrically stunning the shrimp before they freeze and suffocate. That way, if shrimp do feel pain, they won’t suffer for that part of the process.

The Shrimp Welfare Project — an Effective Altruist animal charity — gives stunners to shrimp farms and gets retailers and shrimp suppliers (Tesco, M&S food, Sainsbury’s, etc.) to commit to stunning their shrimp before slaughter. Because the stunners are so cheap, and stun so many shrimp at once, donations are absurdly cost-effective. $1 to the Shrimp Welfare Project amounts to around 1,500 shrimp being stunned and killed unconsciously per year.

Last year, Substack went into shrimp mode.

Glenn

,

Bentham's Bulldog

,

Richard Hanania

, and myself all posted essays on the same day, pushing the case for caring about shrimp welfare, and for donating something — even a tiny amount — to the Shrimp Welfare Project.

Going in, we all knew no one cared about shrimp. That was why we made arguments for caring about shrimp. The argument was roughly as follows:

Pain is intrinsically bad, and its badness matters morally. Suppose you happen upon a goose by the side of the road, writhing and squawking in agony as rasping parasites consume the bird from within. Within a day, the goose will be gone. Fortunately, before you, there’s a convenient little button: if you push it, the goose will be anaesthetised — the parasites will still have their meal, but the goose won’t die in all-consuming agony. Intuitively — at least to me — you would have some moral reason to push the button, even if you had to heave your flabby arm up from your side to do so.

Your reason to anesthetise the goose, the thought goes, wouldn’t vanish if the parasites had damaged the goose’s brain so much that it’s IQ was that of a decapod. As long as the goose’s pain hurts intensely, the thought goes, that’s enough to generate some moral reason to press the button. Gratuitous pain is bad — relieving it is a good thing, ceteris paribus. And what’s sauce for the goose is sauce for the shrimp: analogously, when shrimp suffer, that fact has a non-zero moral weight.

Now consider the numbers: 480 billion farmed shrimp are killed annually. 480 billion. (For a sense of scale, that’s more subscribers than I have on Substack.) According to many moral philosophers — who have argued for this view after much trepidation and in painstaking detail — the badness of pain is aggregative. That is, if two sentient creatures are suffering, that’s worse than if one creature is suffering, ceteris paribus. According to this view, the badness of pain sums up, even when it occurs in separate minds. This view is not uncontroversial; nevertheless, the arguments in its favour are robust, and it’s hard to specify where they go wrong.

Assuming there’s a decent chance that shrimp feel pain, that pain is intrinsically bad, that the badness of pain generates moral reasons for those who can easily lessen it in others to do so, and that that badness of pain can aggregate, the moral risk-based argument for donating to shrimp welfare starts to look pretty darn strong. If you’re not sure which step of the argument for shrimp welfare you reject, you should consider donating just in case (or so we defenders of shrimp welfare would claim.)

The case intensifies when you realise that — assuming shrimp are sentient — pro-shrimp conclusions can be derived from a variety of moral theories. The argument I just sketched has consequentialist overtones — albeit of the sort that you can accept even if you’re a committed non-consequentialist, as I am. (Recently, for example,

Bentham's Bulldog

and I interviewed the Kantian philosopher Florence Bacus, who made the case for shrimp welfare from a staunchly Kantian perspective.)

You don’t have to agree with the moral risk argument for shrimp welfare to agree with what I’m going to say next: polling 450 internet people about a complex practical ethics question that hinges on complex empirical and normative ethical questions and gathering up their off-the-cuff responses is not how one ought to do moral philosophy.

Why is polling 450 internet people about a complex practical ethics question that hinges on complex empirical and normative ethical questions and gathering up their off-the-cuff responses a bad way to do moral philosophy? Here are just a few of the reasons, zooming in on the question of shrimp welfare:

1. Intuitions — here I agree with Stone — are prima facie evidence in ethics; when people report intuitions that shrimp don’t matter, for example, that’s prima facie evidence that shrimp don’t matter. However, there’s a reason that ethical intuitionist philosophers sometimes accept results that are prima facie unintuitive, rather than steadfastly sticking to the first reaction they had to every ethical question they’ve ever been faced with. The reason is that moral philosophy, done right, relies on reflective equilibrium. Instead of just going with your first reaction to an ethical question, you’re supposed to gather up a wide range of ethical intuitions that you have — consider whether there’s any reason to think that some are likely to be less reliable than others, so that they don’t all have equal weight when you sort through them — and then weigh them against each other. As you do this, you’re supposed to weigh intuitions about specific cases against intuitions about abstract moral principles. An example of an intuition about a case is the intuition that you ought to save a nearby drowning child at minimal expense; an example of an intuition about a principle is that how morally important a child is shouldn’t hinge on whether her belly button is an innie or an outie. Over time, you’ll notice that you can’t possibly accept every intuition about cases and principles as true, since some of them are contradictory — sometimes, you will have to surrender some intuitions about cases or principles, generating verdicts that are counterintuitive.

2. In the case of shrimp welfare, defenders of shrimpkind have given positive reasons to think that our case-intuitions about shrimp are unlikely to be reliable, since they are probably delivered by heuristics that we have independent reasons to think are unreliable guides to moral truth. For example: (1) shrimp are small — many pro-lifers complain, quite rightly, that pro-choicers who write things like “zygotes don’t matter, since they’re smaller than the full stop at the end of this sentence.” are being silly, since they’re relying on a size-matters heuristic that we have no reason to think would be truth-tracking; (2) shrimp are the most eaten animal alive by the numbers, so we’d expect status quo bias to creep in when we’re considering how important their interests are; (3) many people believe — falsely — that shrimp aren’t cute: the fact that humans commonly assign more moral weight to cute animals than crusty ones gives us independent reason to think that shrimp cuteness-deniers would have unreliable intuitions about cases involving shrimp; (4) many people believe — for reasons that have nothing to do with the relevant scientific evidence — that shrimp either don’t feel pain or barely feel any if they do.

3. Stone’s cases were ones where people were asked whether they’d save a drowning child over “[1,000 ] / [10,000] / [100,000] / [1,000,000] freshwater shrimp”. Note two things about this: we have independent reason to think that people’s moral intuitions would track any moral difference between different large numbers of shrimp, because of the well-documented phenomenon of scope insensitivity. When numbers get big enough, humans stop being sensitive to the difference between the numbers — including to differences in moral value. At some point, our brains just start saying Big Number, and fail to track salient differences in size. Second, many people — rightly or wrongly — seem to endorse the view that human lives/ interests are lexically prior to either some or all animal lives/interests. That is, many people are disposed to place infinitely greater weight on human lives/interests compered to the lives/interests of some or all animals, depending on who you talk to. Importantly, many utilitarians endorse lexical priority views, so even utilitarians may be suckered in to this way of thinking. Now: such views could be right — they have their defenders; but equally, lexical priority views face formidable theoretical challenges. If you’re serious about figuring questions of moral status out, you can’t just report your intuition that human lives are lexically prior to some animal lives and be done with it: there are plenty of weeds to hack through.

4. Which is why what Stone says of shrimp I am hyperbolically inclined to say about the views of his 450-something survey respondents: “They do. Not. Matter. [Epistemic] weight of zero.” Some questions in ethics are hard: that’s why legions of trained philosophers dedicate their entire lives to reflective-equilibriuming the bejesus out of them. If you asked 450 people if they found Parfit’s Repugnant Conclusion intuitive, you would get 450 noes. That’s why they call the conclusion “repugnant”. But that would not be philosophically probative. The reason philosophers still puzzle over the Repugnant Conclusion is that there are multiple, extremely theoretically tempting lines of argument in its favour that employ extremely hard-to-deny premises.

5. That is what shrimp welfarists claim when it comes to shrimp welfare — as a result, Stone’s survey is dialectically inert, regardless of what you happen to believe about the fantastic and awesome Shrimp Welfare Project.

   ---