In this post I’m going to offer some follow up to two debates I recently completely won in victories so total, my interlocutors will probably just have to become baptized Lutheran pronatalist influencers.
The first debate is in the video above, moderated by Peter Singer vs. Jeff Sebo, in which I conclusively demonstrate the foolishness of donating to shrimp welfare. But in the post below I will provide additional arguments, because when your opponent is down, it is morally obligatory to kick him until he’s pulp.
The second debate was audio only, on Maiden Mother Matriarch hosted by Louise Perry vs. Diana Fleischman on the topic of eugenics. I so thoroughly destroyed Diana’s position that she actually agreed on air to donate money to help drug addicts access fertility treatment!!!
So with that, let’s go ahead and review these debates!
But before I go further, I want everyone to remember that the objective of a debate is to win. Learning is not the goal. Discerning truth together is not the goal. Pinning your opponent and making them surrender claims it was their job to defend is the goal, or else rendering their defense of those claims preposterous. Being fair and truthful are usually helpful to that end but are not always requirements. Sometimes you foul a player to slow them down; that’s how basketball is played. Several people have commented to me that they were disappointed I didn’t spend more time in the debates trying to draw out interesting lines of inquiry or seize the opportunity to follow an opponent’s reasoning to its more charitable conclusion. These commenters do not understand what debates are for. Debates can be won by being more correct or simply by being a better debater, both are valid, and in fact you should regard opinions which systematically appeal to people unable to make compelling public defenses of them under pressure as less likely to be true even if the claims seem factually plausible. The best test of truth is the public torture of the believer in the debating arena, and cracking for any reason, even reasons not logically related to truth claims, is an indictment of the truth of the position. I will not elaborate on this further at this time, but will just say that truth has no existence outside of minds and so to believe a claim you should in fact weigh the mind making the claim, not only the claim in independence.
Eugenics Is Bad And I Won
We debated eugenics. Diana Fleischman wrote a postmortem:
I will focus on rebutting her postmortem claims.
It Is Obvious That I Won
Diana offers a lot of arguments about what she was going to argue but it’s all sort of moot because the nice thing about a debate is you formally agree beforehand exactly what you’re going to argue about. So for the record, this is the exact text of the resolution:
Agree/disagree, people who carry genes that are likely to cause their children great suffering or render their children incapable of economic independence should be socially discouraged from reproducing
The resolution we agreed to debate in advance stipulated “social discouragement” (not coercion and not even necessarily government policy) of reproduction for people with genes rendering them incapable of economic independence or with genes causing great suffering. “Great suffering” and “incapable of economic independence” are not defined and I expected Diana to use that fuzzy leverage to her advantage; as it is she basically didn’t, she essentially conceded the economic independence point, spent almost no time on great suffering, and instead introduced an argument unrelated to the resolution, namely, third party harms. Diana talked a lot about drug addicts and murderers, which is interesting because drug addicts and murderers are not necessarily genetically incapable of economic independence nor are they necessarily genetically caused to experience great suffering. Many drug addicts and murderers have no genes predicting these outcomes, many people with genes predicting these outcomes don’t experience the phenotype, and many people experiencing the phenotype nonetheless have only moderate suffering or do have economic independence. Rich murderers exist.
As such, by any reasonable scorekeeping, I won the debate. Diana conceded half the resolution explicitly, didn’t defend half of it, and burned up her time on ancillary arguments unrelated to the resolution. Instead of actually debating eugenics in the sense of “Should we prevent people from being born for their own good” (great suffering) or “for economic reasons”, she asked “Should you murder baby Hitler?” to which the answer is yes. At one point to defend her position she literally had to posit a minority-report style perfect precrime report, a thing which, given the statistical rarity of the extreme events in question, is simply never going to exist.
But Was I Actually Right?
Having established my indisputable victory and the ignominious defeat of eugenics, it’s now fun to ask if my view is actually true, a separate and only somewhat related question.
As it turns out my view is true, which is very convenient.
Let’s go through Diana’s rebuttals.
Is Lyman A Total Utilitarian?
No. Diana says I am, but this is because she’s so utilitarian she cannot understand how non-utilitarians think about consequences. Diana thinks that because I believe that aggregate “policies” for society should involve some kind of consideration of consequence, that I must be a utilitarian. But this is nonsense: it’s perfectly suitable for a virtue ethicist to believe individual action should not generally be consequence-depended and that judgments of the morality of action should not primarily be consequence-informed, and also to believe that in the particular role of lawgiver and sovereign, the sovereign has a duty to perform virtues which might require the consideration of consequences. A person should not generally lie because lying is bad for the soul and sets you on a path of deception— but the sovereign by accepting sovereign duty has ceased to be human, they now are State, and are morally obliged to perform the virtues of the state rather than the virtues of the man. Dr. Lyman Stone is a doctor of sociology so should not cut you open, but if Dr. Lyman Stone was a surgeon, cutting you open might be a very good thing to do. That which is morally licit depends on who you are; universals exist only at a meta-ethical level. Thus, we should all be virtue-ethicists, but the virtues of the sovereign will require making considerations of the greatest good which, yes, are often going to require some math.
So no, I’m not a total utilitarian. But as it happens, total utilitarianism is often a decent proxy for the things I believe the imaginary sovereign should do. If I said the real reasons the sovereign should do things, ah, well:
For the word of the cross is foolishness to those who are perishing, but to us who are being saved it is the power of God. For it is written: “I will destroy the wisdom of the wise, And the understanding of those who have understanding, I will confound.” Where is the wise person? Where is the scribe? Where is the debater of this age? Has God not made foolish the wisdom of the world? For since in the wisdom of God the world through its wisdom did not come to know God, God was pleased through the foolishness of the message preached to save those who believe. For indeed Jews ask for signs and Greeks search for wisdom; but we preach Christ crucified, to Jews a stumbling block, and to Gentiles foolishness, but to those who are the called, both Jews and Greeks, Christ the power of God and the wisdom of God. For the foolishness of God is wiser than mankind, and the weakness of God is stronger than mankind.
I would simply not be understand. It’s not useful to speak Japanese to an audience that speaks Italian; in the modern world, speaking the language of virtue to a society without any simply fails at communication. We must speak in utils because the utils are all the listener has left in our current cultural age.
Was that German Couple Offered Sterilization?
I repeatedly said that the German brother-sister couple in question was offered sterilization. Diana was unaware of this fact. In her follow-up, she insists I made this up. This is bizarre. The couple had four children and then Stuebing underwent a voluntary vasectomy. At one point I seem to recall Diana claiming they made him get a vasectomy? I didn’t rewatch the debate so that may be erroneous recall— but the key point is, Stuebing could have gotten that vasectomy at any time! Lack of access to it wasn’t the issue! And German health insurance will cover tubal ligations for medical reasons, and “likelihood of birth defect” is a valid medical reason! After their first child was born with such a defect and given that this was incest, she almost certainly would have been covered. None of the media reporting has ever suggested they were denied contraception, temporary or permanent, and Germany has a generous public healthcare system!
What’s going on here is Diana has an extremely low view of the autonomy of people she sees as cognitively less. It’s not enough that these procedures were available and likely very cheap or free. It’s not enough that LARCs were available and free to use. It’s not enough that obviously yes after the first baby and the hubbub around it everyone was telling them to do something about it. For Diana to count it as “offered,” somebody had to be knocking on their door with a check and a car to take them to a clinic. Here’s Grok’s take:
So Grok cites sources explicitly saying the reason they didn’t use contraception is because they wanted kids! That’s exactly what I said in the debate! I was right! This couple wasn’t having “whoopsie” babies they wanted children.
Also note that their sex was still a crime even if they were sterilized, which blows up the argument that incest laws are eugenic in intent.
So here, I’m clearly right. Yes, they were offered options for not having babies. They chose not to take them, because they wanted children.
Incest Laws Generally
Diana had us quickly move off incest laws which is sad because they’re the most important part of the debate and I was loaded for bear on them. And crucially, they’re extremely load-bearing in her view!
Most people do oppose incest, and do feel disgust at it. Diana wants you to believe that feeling of disgust you have is about eugenics. She wants to argue that your anti-incest impulses are a deep revulsion at dysgenic fertility.
But they aren’t! First of all, other primates don’t exhibit the breadth of Westermarckian-type-effects we see in humans. Humans are unique for the breadth of their care relationships and the breadth of their incest taboos, especially father-daughter. This tells us that whatever is going on with human incest taboos, it must be evolutionarily fairly “recent” in phylogenetic terms.
More than that, in a paper I haven’t published yet using over 10,000 ancient human genomes, the decline in incest really is historically recent! Incest rates were a lot higher in the past than they are now. This matters, because in the hunter-gatherer context, cousins were also reared together! In the ancestral human context individuals are not just related to most of their social ties but they are reared alongside many potential mates or even reared by their potential mates. Extensive non-kin or distant-kin alloparenting relationships mean that if it really were the case that humans had deeply-rooted incest-aversion, we’d have gone extinct long before agriculture.
Rather, humans very clearly preferred to mate with relatives and people they were raised with for most of our history. To the extent we developed strong preferences against mating with out cousins, it’s a phenomenon that must have begun in the last 10,000 years. As far as avoidance of mating with siblings or parents, humans still do this! Older siblings make up a large share of perpetrators of child sexual abuse. Here’s Grok’s take, again, if you don’t believe me:
It goes beyond that. A recent study in Britain shows that 1-in-3600 Britons in the latter 20th century were products of first-degree incest. Given nonconception, miscarriage, abortion, and selection out of sample via early mortality or stratification, it' seems extremely likely that more than 1-in-1000 20th century Britons were periodically having sex with a sibling, parent, or child, perhaps has much as 1-in-100 if you think that close-kin relationships have higher failure to conceive (debated). Diana brought up the example of Huntington’s Disease— but first-degree incest is literally multiple times more common than that!
A lot of scholars are really deeply wedded to the idea that close incest is a super rare thing, humans instinctually are repelled by it, etc. But this just isn’t the case. People bang their siblings, parents, and children with gross frequency. This behavior has declined over time, and we know this from large archaeogenetic samples. The best explanation for the decline is basically agriculture + Christianity. But the fact that a large segment of the world still has crazy high rates of consanguinity due to religious influence (Islam). Moreover, the fact that tons of countries legally allow first-degree incest should really give us pause about how deeply-rooted anti-incest norms actually are.
Why does this all matter? Well, Diana’s whole rhetorical project here is just to slide down the slippery slope slowly. She wants to convince you that you already accept eugenics, so why not a little more? That’s actually her argument! “People already think drug addicts shouldn’t reproduce.” “People already support incest laws.” etc
So it’s important to recognize that actually not everybody supports incest laws and those who do often don’t do so for genetic reasons! You can support incest laws for non-eugenic reasons, for example, “Because I think incest leads to really bad social norms in a lot of different ways and creates dangerous tribalism, and because to the extent I want Islam to spread in my country I want it to be Islam bereft of the Arab tribal cultural baggage which is so damaging so societal health.” This is my actual belief, I would like to extend laws against incest even further, and increase how punitive they are.
Does Legalized Abortion Reduce Crime?
We also disagreed about abortion and crime. Diana is a partisan of a widely-panned theory that abortion reduces crime. She already admitted that she massively overstated the effect size. But she’s also just wrong.
To understand why, we have to understand what the paper she cited did.
They basically looked at abortion rates over time vs. homicide rates lagged some nuumber of years and found a correlation. So abortions reduce homicides right? Wrong. Their first paper just had coding and data errors which eliminated the results. Moreover, the results they found turned out to be spurious: the falling crime rates were not actually among the specific birth cohorts exposed to higher abortion, but occurred across all cohorts. Eventually the authors responded to comments and doubled down, which was absurd because they 1) didn’t bother actually addressing the cohort-vs-period issue and 2) argued that Roe v wade was causing continuing convergence for decades after convergence in abortion rates had stopped, which defies their mechanism.
So then you want to ask, What’s the international evidence? In Romania, banning abortion and causing a MASSIVE increase in births had no effect on the criminality of the people born. This is important to grasp: abortion can reduce crime rates mechanistically by reducing birth rates, and young people do more crime. The eugenic mechanism operates on criminality of those born, not on total cohort size. In Romania, crime rose due to total cohort size, but the criminality of those born was unchanged.
What about in Canada? In Canada, within-cohort criminality also didn’t change when abortion rose. A 2025 study also says the crime-abortion link in Canada “cannot be substantiated.”
There are several European studies too, which find mixed results, the most Diana-favorable of which finds an effect size about 1/2 the size of what Donohue and Levitt claim, but the others find no effect.
In general, the evidence here is very weak. And what has not been done is to use something like international youth homicide rates vs. international abortion rates across countries to do this in very diverse contexts.
So in practice, the strategy of “just encourage people with unwanted pregnancies to abort more” is not in any sense consistently linked to selection against criminality. It appears to be basically close to criminality-neutral.
Is There A Contradiction Between My Favoring Banning Cousin Marriage And My Pronatalism?
No, obviously. Diana thinks there is because 1) she thinks I’m a total utilitarian and 2) she thinks cousin marriage will lead to additional people being born. But 1) I’m not a total utilitarian and I accept many policies which may reduce the number of babies born in some oblique fashion and 2) incestuous marriages likely have lower fertility than less-incestuous ones, ceteris paribus.
Was I Wrong On Conjugal Visits?
At this point it feels like Diana needs hearing aids. She found evidence showing that, yes, conjugal visits are a thing that some states provide to some inmates. She then claims this refutes my claim that… some states provide conjugal visits to some inmates. My exact verbatim claim is here:
I would say that there is not a presumptive right that the state should incur additional costs on the behalf of a criminal to enable them to reproduce. That is to say, when they committed the crime, they chose to do the time. But at the same time, the state, most states do allow conjugal visits for a range of certain types of crimes on the understanding that going to some length to allow people basic biological autonomy is not entirely unreasonable. So I would say, again, current law allows conjugal visits for some classes of criminals in some jurisdictions. And I don’t think that’s crazy.
Current law does indeed allow conjugal visits for some classes of criminals in some jurisdictions. That’s what Diana found. I don’t think that’s crazy. I did not argue that conjugal visits are allowed for all, most or even many prisoners, and I did not argue that all jurisdictions allow this for some prisoners.
Now, I was not aware, as the link Diana shares notes, that conjugal visits are being phased out for many states. That seems dumb to me. Being able to dangle family visitation for prisoners with good behavior seems incredibly useful to me, and I don’t think that the state should presumptively prohibit such a “carrot,” even as I don’t think prisoners have a presumptive right to them.
Does Diana Even Know What Eugenics Are?
She shares this graph from a survey:
I don’t support discouraging poor people from having children, but surely Diana recognizes that the correlation between the best available PGS score for income and actual income is nowhere close to 1! It’s not even vaguely in the neighborhood of ! PGS scores for e.g. income are strongly correlated with income or wealth at group means, but with enormous individual variance, such that the actual share of income explained by the PGS score is <5%. In other words, you could repeatedly select on phenotypic income in every generation and you’d basically never get much drift in the underlying PGS, because the “bad genes” remain so common among phenotypically “good” outcomes that they would continue to perpetuate; and “good genes” are so common among phenotypically “bad” outcomes you’d be weeding them out too! This is basically a Breeder’s Equation but with a continuous trait. Imagine a trait which is continuous from 0 to 1 (income percentile). For this trait, there is a PGS which explains 5% of the trait variance (this is about right for current income PGS scores). Suppose that there is moderately high assortative mating on the phenotype (maybe 0.4), which is not implausible. We want to compare two scenarios. In one scenario, across 10 generations, individuals at all phenotype levels have the same fertility. In the second scenario, across 10 generations, individuals in the bottom 10% of the trait have 50% lower fertility, consistent with some of them being eugenically sterilized. We want to see how the genotype distributions change across 10 generations for each scenario.
Doing way more math than I expected when I started that paragraph, I think this pencils out to approximately a gain of like 0.1 standard deviations in the underlying genotype across 10 generations. Given typical generational widths nowadays, that means in 275-400 years, you’d gain 0.1 standard deviations in whatever the PGS score for income is actually doing, biologically.
That’s absurdly slow for any kind of meaningful social policy.
And here’s the wild thing: in the survey results Diana cites, people are most eugenic about wealth and less eugenic about intelligence! They are more eugenic about wealth than about crime! This doesn’t support the idea that people have plausibly eugenic intuitions! They’re fine with dumb and reckless people breeding and don’t net-net tend to favor stuff aimed at those people, but they really hate paying for welfare. That’s all we learned here.
Diana repeatedly appeals to phenotypically targeted interventions when discussing eugenics, but, crucially, the word for phenotypically targeted interventions is “euphenics” not “eugenics.” Now, I also dislike most cases of euphenics, but the debate resolution was about people carrying genes not people presenting phenotypes.
This matters, because a eugenic policy would be something like “You carry criminal genes, you should be socially discouraged from reproducing.” Or to put it another way, “If your parent or sibling committed a crime, you should consider sterilizing yourself.” That’s the actual eugenic argument for polygenic traits! That’s what we were supposed to be debating! The fact that Diana instead wanted to debate which people are bad enough of people that we should bully/bribe them into sterilizing themselves is, I think, a massive own goal, and a big reason why people are often uncomfortable with eugenics. It turns out, the eugenicists really are motivated by simple animus.
Shrimp Must Die And I Won
On shrimp, my victory was sweeping and total. He conceded the key point that there’s no actual science on the species the welfare projects are aimed at. He introduced only the goofy studies about acid on antennae and anaesthesia etc, which 1) have failed to replicate in at least one case and 2) are only demonstrations of nociception not pain. He failed to introduce any evidence relevant to the shrimp nervous system which might compel us to think shrink have pain. He asserts that the evidence the shrimp “might” have something like pain we should care about is strong, but introduces no actual such evidence and, furthermore, as I clearly proved, “might” isn’t good enough. The ReThink Priorities report does not give an explicit % rating to shrimp because insufficient data was available, so even if you think that report is credible (it isn’t), it couldn’t give you a plausible baseline. It is true people sometimes take “insurance type” actions based on tail-risk probabilities, but they usually do so only in cases where tail risk is extremely well demonstrated. There is a robust market for life insurance because the fact that mortality occurs is well known and indisputable. Disability and unemployment insurance, likewise, are tail risks, but the fact that the risk exists is basically an infinity z-score probability event. The probability that I am wrong that “disability is a thing which theoretically could happen because it exists in the world” is so low as to be equivalent to the probability that I’m wrong about my own consciousness and existence, in which case, there’s no point debating anyways.
But utilitarians and pansychists (the latter being the more honest extension of the former), of which I am not one, want you to accept a different and absurd kind of risk. The risk that maybe totally undemonstrated phenomena are real. Maybe the shrimp have complicated emotional lives. Maybe God is a shrimp and shrimp pain matters a lot to God for empathetic reasons. Maybe the trees are talking to each other. Maybe the earth cries out for the harms we do her. The list of things that we can imagine being catastrophic harms is very great. Maybe Alpha Centauri is a population of quadrillions of sentient aliens wireheaded to a torture computer which could be turned off by a human with a single push of a button and we should invest everything we have in space travel. Maybe Christian Hell is real and we should all be missionaries. What these risks all have in common is that there is really no way you can go and actuarially account for the risk of these things being true, efforts to do so are just open and frank dishonesty because there is no estimable hazard function, and thus it is impossible to price the insurance product. The reason there is no estimable hazard function is because we do not possess a way to measure and demonstrate the phenomena in question. And thus, “maybe shrimp are conscious” just isn’t an argument. At all. Do not pass go. Do not donate $200 to shrimp welfare. Until a shrimp the shrimp swims up and pushes the button labeled, “I am in pain and I mourn the ablation of the eyestalks of my sister Shrimpette,” they’ve got nothing. Literally nothing.
At times I felt like I was arguing with a crazy person, because again, there just isn’t any evidence there!
So anyways, Jeff conceded the key points, his only line of defense was to gesticulate at the camera at the end as if to argue that he really thinks we don’t know what it’s like to be a fruit bat, when we absolutely do know what it’s like to be a fruit bat.
Beyond that, some of my core arguments went unrebutted. Jeff basically didn’t engage with my “physical pain isn’t so bad” argument, which means it was left unrebutted, so it’s points in my column.
Some Additional Arguments
I triggered a jobless philosopher right there so let’s elaborate.
Fruit Bats Are Solved
The invention of NeuraLink lowkey settles the debate on whether or not we can collapse consciousness to matter. Now, it does not mean we can create consciousness from matter. Reversal may not work. But, provided we have an entity we think is conscious, we are almost certainly on the brink of being able to completely map every element of their consciousness and save it in a .raw file or whatever. It’s not clear if you could upload that consciousness somewhere else, if editing it outside of the biological substrate would work, what kind of machine could read the file well, whether reading it would constitute consciousness, etc. But we really are at a place where we can poke certain parts of the brain and make thoughts happen, or, vice versa, make thoughts happen and those manipulate, say, a mouse on a screen. Nonbiological brain-world interfaces are going to get better.
The takeaway from this is that once China decides TortureLink is a great way to deal with Uyghurs, we will have solved all our questions about “To what extent is the cerebrum involved in pain.” The answer is going to be “A whole lot.” Decerebated children have extremely weak evidence of pain in the way we feel pain, and, of course, deactivating even a small part of the cerebrum is doing the lion’s share of the “work” of general anaesthesia. Species with cerebral activities and vertebrae simply do experience pain rather like we do, and to the extent their brains contain analog structures to ours, they likely experience other affective states similar to ours. Now, because they generally have WAY lower compute than us, we can plausibly guess they experience most things “less” than we do, many species also have different flicker rates meaning they likely experience time differently than we do; in shrimp’s case, in a way that means “1 second of pain” in our time is actually less than “1 second of pain” equivalent in their time. Species vary widely in memory capability and ability to self-recognize, and these likely influence the scope of affective states they can experiences.
But the key point is this: it just is the case, empirically, that the sentence of a species is a function of its neurological structure. Where there’s not hardware, there’s not software. It’s not quite as simple as “cerebral neuron count X flicker rate” but honestly that’s like 75% of what you need to know. And by that standard the humane treatment of bumblebees rises surprisingly high on the list of things a utilitarian may want to care about.
It’s also worth noting that sense most studies suggest that the cerebrum intensifies pain signals, that shrimp likely experience less pain than we would from a given neuronal stimuli. We can actually observe this in the shrimp-acid experiment! Rubbing acid on the shrimp’s most sensitive primary receptors didn’t even cause all shrimp to react to the injury, and their reactions were nowhere near the maximum reactive intensity of which shrimp are capable. If you poured the same acid into human eyes, virtually 100% of people would react, and quite intensely. Likewise, cutting a human arm off would provoke much more response than shrimp eyestalk ablation causes. Bereft of eyes, shrimp simply breed like crazy. Humans don’t respond to mutilation that way, which should tell us the shrimp experience of pain is nothing like ours, and if it provokes an extremely high-evolutionary-reward-behavior (intensified reproduction), we have to consider the possibility that shrimp may be sadomasochists who enjoy certain types of pain (we’ll come back to that).
We Didn’t Get To Talk About Memory!
Compute matters. So does memory.
Even if I granted that shrimp feel pain in a way morally similar to how I feel pain, conventional calculations of the pain in a shrimp’s life are literally just wrong. They do not account for the temporality of pain. Most of us feel a lot of pain when we come out of the womb—but in hindsight few of us would see the pain of coming out of the womb as worse than being alive. But the key reason for this isn’t just that being alive is good, but that we have no memory of being born. Memory is a key component for weighting pain: the moral weight of pain is not just the instantaneous pain, but the pain weighted by some memory-related factor. Nobody thinks the only harm of a pain is the instantaneous pain! Because shrimp are demonstrated to have very short memories, if any “memory” at all (they have no cerebrum of the sort that would store memories as we think of them for other creatures, but they do seem to have some stimuli-recollection ability), we must place a very low weight on this. If you think remembered pain is even 1% of initial harms, the brevity of shrimp memories vs. the great length of e.g. cow memories (i.e. cows plausibly remember thousands of times as much stuff as shrimp do), then the identical neuronal stimulation for a cow is vastly worse than for a shrimp, because the cow remembers it for orders of magnitude longer, perhaps even giving rise to emotions like shame. Because the cow remembers it, the cow also anticipates it, likely having emotions like fear. Talk to people who have survived any kind of traumatic experience and ask them “Of the share of the stuff that happened to you, how much of the badness you experienced was because of the physical pain vs. the fear/shame/horror/guilt/whatever?” All of those are basically memory-contingent. Shrimp, being bereft of most memory, blessedly do not “dwell” in their pain. Rip off an eyestalk, hurts for a little while, then you just go and churn out shrimplings and move on.
Neuron Scaling
If every single neuron a shrimp possesses had no function other than pain-reception (this is between 10,000 and 100,000), triggering every single one all at once would trigger about the same number of pain receptors as are triggered when a human licks their fingers and puts out a match: that just isn’t very “much” pain, and the volume of pain is so low it should be considered non-additive. Utilitarians insist on adding pain at the individual creature level, but they should obviously actually be adding pain at the nociceptor level! 5,000 nociceptors triggered across 5,000 individuals or 1 individual should be equivalent to utilitarians; same pain sum.
Experimental studies have repeatedly shown that pain is spatial, such that sheer number of nociceptors activated predicts more intense pain. Now, mammalian nociceptors are not binary, they allow gradations of pain due to stimuli intensity, but nobody has ever studied if shrimp nociception works this way, not least because nobody has ever actually materially found a shrimp nociceptor! We infer them from behavioral data without direct observation. I think it’s very plausible shrimp nociception is also not strictly binary, so my “lick your fingers and put out a match” comparison is not strictly correct. Something like accidentally knicking your finger peeling an apple is probably more equivalent. But these are still trivialities.
Regardless, this kind of low-grade pain occurs on an infinitely vast scale on our planet across all species. Every animal bigger than shrimp experiences that much pain a hundred times a day, making these kinds of harms our moral guide is absurd on its face and suggests advocates taking that view just don’t understand what moral intuitions around pain are about. If I pinch myself somewhat hard, I am actually sending more pain signals to my nervous system than would be the case if every single shrimp neuron fired all at once. Activists for shrimp insist on counting individuals because they know if we talked about neurons people would realize that there’s just no “there” there for shrimp. Maximum Shrimp Pain just isn’t much pain, and that’s assuming it’s something morally similar to our pain at all.
Shrimp Stunners May Not Work, Maybe?
This was one my big arguments I wanted to whip out to blindside Jeff with but it didn’t come up.
Shrimp stunners cause pain! Shrimp welfare projects pay for stunners, but when shrimp stunners are used, they cause a brief second of massive seizing by shrimp. Humans who have experienced major electric shock report that it is very painful even if they rapidly lost consciousness. For shrimpists who don’t want to downweight for memory factors, this instance of extreme pain has to be considered: it’s not actually clear if the split-second of literally every neuron a shrimp has being literally electrocuted is necessarily less painful than the alternatives. We humans prefer short-sharp pain because we have memory and a continuous sense of self, but shrimp have little memory and no cerebrum with which to have a continuous sense of self. Shrimp might plausibly prefer the pain peak to be lower but across more moments since they may not remember much across moments.
Shrimp Orgasms Matter
Humans think our lives are net positive because evolution has an incentive to convince animals to keep living. This is likely to be true across all species to some extent: to the extent population is stable or growing, we can assume that individuals are receiving evolutionary rewards of some kind. For humans, the biggies here are orgasms, but broadly we could think about hormones like serotonin and dopamine which, hilariously, shrimp apparently also have. Now, shrimp don’t have the cerebrum which could enable them to have the cerebral effects of these hormones— but if you’re like me, your cerebrum isn’t the only place you feel it when evolution gives you the “reward for reproduction.”
There’s no reason to think shrimp are different. To the extent they feel pain and we infer pain from behavior, we should also presume them capable of pleasure and, thus infer pleasure from behavior. Reproduction is extremely likely to provide organisms with positive rewards which may be like our pleasure. To that end, we should take the reproduction rate of shrimp as a proxy for shrimp welfare.
If they are reproducing without direct mechanical intervention (sorry, vertebrate farm animals), i.e. by their own actions, we should assume they’re getting the normal evolutionary rewards. Thus, I said it in jest, but tbh I meant it: eyestalk ablation may be welfare positive for the ablated shrimp. If eyestalk ablation really causes for reproduction, maybe shrimp eyesight is shrimp-happiness-negative! In an ecosystem free of predators, it’s actually easy to imagine that shrimp spending energy on eye sight really would be fitness-negative!
But this gets to a wider critique of the shrimp welfarist position. They want to abolish shrimp aquaculture in the long run!
But if a tiny intervention like shrimp stunners can alleviate so much harm, it stands to reason similarly tiny interventions could probably create massive amounts of shrimp pleasure, and creating that pleasure would be no less obligatory. The ultimate problem here is by using individual-counts instead of flicker-and-memory-adjusted-cerebrum-neuron-counts, the shrimpists have uncovered a classical utility monster! You can adjust shrimp utils way cheaper than human utils, and therefore total utility really is maximized by pouring all resources into shrimp wellbeing.
The shrimpists often elide this point, but I think it’s inescapable. Once you believe in shrimp pain, you almost are required to believe in shrimp pleasure. And once you do the math on shrimp pleasure, you’re not just hypothetically wondering about insect wellbeing, you realize that making shrimp farms marginally net-positive for shrimp would be literally the greatest thing ever. And making them even more net positive would be even better. And so on. This is absurd to me. Maybe it isn’t to you. But for me, I choose mankind.
