Summary

• Utilitarianism is generally presumed to be the main “animal-friendly” philosophical framework.
• Its exclusive focus on hedonic goods and animal sentience provided strong reasons for more moral concern towards animals, in particular about their current treatment in factory farming.
• While initial utilitarian arguments in favor of better moral treatment for animals point to generally common sense reforms like welfare reforms or plant-based diets, current utilitarian arguments rightly point to another conclusion which is coherent with this moral framework after informed by recent scientific evidence: invertebrate welfare as the pressing problem due to its astronomical scale.
• The most detailed analysis up to date about this issue is the Moral Weight Project. Its implications point to the need of massive shifts in resource allocation toward invertebrate welfare.
• While it may be reasonable to accept that invertebrate welfare can be considered as one of many pressing problems, it seems to be against commonsense when the logical results of hedonistic and utilitarian framework suggests that invertebrate welfare has a dominant priority over other issues - which is very hard to challenge due to its huge scale.
• While a theory does not need to conform to commonsense in order to be true, it might be a sign to investigate further if there is something wrong with the general theory.
• My view is that a moral position that rejects hedonism and accepts incommensurability of values can provide a more reasonable and coherent framework that can on one hand give animals what they are due morally and still does not go wildly against common sense.
• This position is compatible with the idea that invertebrate welfare matters and is one of many areas that one can reasonably pursue to do good, but it rejects that invertebrate welfare deserves our almost exclusive moral concern, as well as moral reasons that lead this particular conclusion.
• This analysis also provides some general reasons against making brute calculations and cost-effectiveness estimates about different donation and career decisions -which is common in effective altruism.

Utilitarianism as an animal-friendly but an against common sense moral framework

Animal sentience is generally considered to be the main feature that many animal-friendly moral frameworks base its arguments on. It can be said that Peter Singer is the leading philosopher who started or at the very least popularized this position. And it can also be said that this was a needed wake up call since there was so much animal suffering in the world and so little moral concern towards animal suffering. 

Singer is an utilitarian and drives his conclusions from utilitarian reasons: 

• Pain is (morally) bad,
• All interests should be considered equally - regardless of race, sex, nationality, or species.
• Various forms of animal suffering are not currently considered morally equal to similar suffering beared by humans. This violates the principle of equal consideration.
• Various practices like factory farming and animal testing without any concrete outcome are immoral and need to be changed.

Singer’s theory is radical but his conclusions were not extremely unconventional: we should end factory farming, stop unnecessary animal testing that does not have clear tangible outcomes, choose plant based over extreme animal based diets, and so on. Although Singer's animal ethics were unconventional for some people who thought vegan diets are very unhealthy or who believed that the scientific community needs a blank check, his overall conclusions are not considered to be “crazy” - which also partly explains his elevation as a public intellectual. 

Yet after certain findings that support invertebrate sentience, if one takes Singer’s arguments seriously and “apply utilitarianism” more broadly, the conclusions become much less conventionally reasonable, even for many animal-friendly people. 

If invertebrates are sentient, and suffer (or the probability of invertebrate sentience is not close to zero), this group should dominate ethics under utilitarian assumptions. All other moral issues (including vertebrate animal suffering) become a sideshow or intellectual passtime. This is because of the astronomically high number of these animals (quadrillions) - relatively significantly higher than even vertebrate animals (“only” billions). 

Given that according to utilitarianism, there is no interesting difference between doing harm and allowing it to happen, this issue cannot be easily solved by “not killing” bugs only.  Effective altruists in particular who are well aware that their career (and relatedly their donations) is the main tool to make an impact, should conclude that they should either pursue a career to reduce invertebrate suffering, or donate exclusively and very heavily to this field - under these assumptions. One final but still perfectly coherent approach (under utilitarianism) might be to cause (or accelerate) the complete destruction of life on earth in order to end invertebrate suffering if one is pessimistic about the possibilities of increasing their welfare through other ways.

Of course, there might be some indirect ways to maximize total utility like mitigating existential risks or accelerate human progress to allow the increase of invertebrate welfare as well as human welfare in the future, or “build effective altruism” to get some form of “trickle-down effects”. So although one can still pursue somewhat conventional career paths, the motives cannot be but unconventional.

A moral framework does not need to be conventional to be true. One can bite the bullet and act according to these conclusions. One can also naturally think that given a lot of people do not pursue these careers or give all their money to this field in practice, there is not a major problem. But note that this (utilitarian) moral framework which is based on animal sentience is used for many purposes by many animal-friendly people and organisations, including effective altruists (even though their actions do not fit perfectly with the real and more pressing conclusions of this framework). 

Extreme unconventionality might be a sign for a missing piece in a moral framework. So I think it is worth trying to figure out what these missing pieces might be if there are any. Below I will provide a shallow list of things that might be considered. I don’t claim that these are very novel points nor that I necessarily agree with them fully (although I do think they hold at least some weight). My main aim is to find out the main elements of a moral framework that can on one hand, provide reasonable concern for animal suffering, as well as that can fit reasonably well with conventional standards. If successful, I hope this framework can inform better moral reasoning and decision making.

Rejecting hedonism

Some sophistications aside, we can say that utilitarianism is hedonistic. This means that good and bad are exclusively defined by pleasure and pain. Certain things might not be conventionally or immediately be “named” as such, like friendship or love, but the value of such things are converted to and weighted as pleasure and pain under hedonism. 

As pointed out above, the conclusion of equal consideration of interests under a hedonistic framework leads to the dominating need for concern for invertebrate welfare. I think it would be understatement to say that the implications of utilitarianism is just “animal-friendly”

Rejecting hedonism does not necessarily need to reject the importance of suffering, especially extreme and long enduring pain. But the non-hedonistic position holds that mere pleasure and pain as themselves do not present as overwhelming or comprehensive forms of good. This position also holds that certain goods are not commensurable, meaning that these can be easily reduced or converted into hedonistic goods (more on that in the next sections). 

A few examples of non-hedonistic goods can be having friends, developing virtue, or acquiring knowledge. 

Moral Weight Project (which elaborately supported the hedonistic case and its conclusions) defends that non-hedonism is weak because the value of non-hedonistic goods do not seem to have a higher value if confronted with hedonistic goods. 

Its main argument and insight is shown through a thought experiment, “Tortured Tim”: “Now imagine Tortured Tim, an individual who is experiencing extraordinarily intense physical suffering. Nevertheless, he may have many other goods in his life: friendship, romantic love, knowledge, practical skills, and so on. However, it seems highly implausible that these goods put Tortured Tim into a net positive welfare state: no matter how valuable they are, they aren’t so valuable as to outweigh the welfare costs of intense physical suffering.” 

Bob Fischer (who is the main author of the Moral Weight Project) says in his podcast episode in 80000 Hours that  “… So that suggest that having all those non-hedonic goods isn’t actually more important, isn’t a larger portion of the welfare range than the hedonic portion — and that kind of caps how much good you can be, in principle, getting from all the non-hedonic stuff.”

I don’t think this thought experiment really proves its point. Given that this is a thought experiment of an extreme case, I think it is fair to test it in similar extreme cases. 

Let’s assume that Tortured Tim’s torturer is Tommy. Tommy is not tortured in any way. He simply tortures Tim. Can we say that Tommy’s life is better than Tim’s? It would be definitely true that Tommy’s life is better in terms of hedonic goods, but I don’t think we can say that Tommy’s life is better. In fact, a reasonable person (think of yourself to test the thought experiment) would grudgingly prefer the life of Tim over Tommy’s. 

Let’s modify the thought experiment even more. Let’s assume that an evil government presents Tim two options: torture an innocent person that will othervise live, or be tortured. I think under these extreme circumstances, accepting to be tortured rather than to torture is the reasonable option that can at the very least provide a better life than being a torturer. 

It is also suggested that it would be irrational to argue that it would be rational for Tim to choose to extend his life, meaning continuing to get tortured. Or it would be similarly irrational to argue that it would be rational for Tim to choose torture now for an equivalent number of normal days tacked onto the end of his life. 

I agree that the former claim is true. But I think this is related to the fact that without a certain level of “health”, it is equally impossible to actualise a significant number of non-hedonic goods, as well as hedonic goods. So while this observation is true (it makes sense for Tim to choose a way to end the torture even if it will also end his life), it doesn’t show hedonic goods have overwhelming weight. 

I don’t fully agree with the latter proposition. I think it might be reasonable to trade off high amounts of hedonic goods for a reasonably proportionate amount of non-hedonic goods. For example, it might be reasonable to accept some “torture days” in order to attain or preserve free and autonomous days of life, if these won’t be available at all in the alternative scenario. It also can be reasonable to accept some “torture days” for the wellbeing of a friend, or to achieve some level of artistic or professional perfection. I don’t think these choices can immediately be judged to be irrational. There are many life choices like these that can be argued as perfectly reasonable. 

Robert Nozick’s experience machine thought experiment also provides good insights to this problem. In this thought experiment, the option to be plugged into a machine that provides an endless amount of hedonic goods as experiences is presented. Nozick argues that it would be irrational to agree to be plugged to this machine because the primary goods of life are not about positive hedonic experiences but rather are to be a certain person or do certain acs i.e. non-hedonic goods. 

Fischer argues against this point by stating that these discussions can be just a “game”: “...And then there’s a lot of me that just says, you know, this is all kind of a silly game, at the end of the day. Like, why does anything matter? It’s because it feels good or feels bad. And there’s a lot of intellectualising to try to make us feel like we’re more important and somehow different from other animals, when in fact we too are just chasing pleasures and trying to avoid pains, and we just dress it up in prettier language.”

But this response begs the question, why should we play this  “(morality) game” where we count and consider the interests of all parties equally rather than playing another game that maximizes our own pleasure and disregard “morality talk” altogether - since this is just a silly game? I think when one presses for hard skepticism against non-hedonic goods and morality around that, it becomes hard to hit the brakes at “moral” hedonism (just at the right time) and not fall into nihilism. 

Conceding a moderate hierarchy between moral status of different animals

Another problem in this topic is the question of how to apply equality and impartiality to different animals. Utilitarianism had an easy and straightforward answer to this question but as pointed out above, it is against common sense. 

An alternative solution to this is to propose a moral hierarchy. Hierarchies can be just if the criteria for the ordering are justified. It can be argued that different animals with different cognitive abilities that can actualize different goods (other than hedonic goods) should deserve different levels of moral concern. For example, a dog or a pig who can use their “animal reason” to acquire certain knowledge, to will certain goods (hedonic or non-hedonic) for themselves and others, and to act in a free manner (in their own limited capacities), can be considered to have a higher status than a worm or a fruit fly who don’t have such capacities (even if they are sentient and have the capacity to experience a high range of hedonic goods). 

Similarly, this would justify placing a much higher status for humans (given their much higher capabilities to actualise goods). This can also explain the “common sense” position that it can be justified to use animals for nutrition and health (vaccines or drugs) if these are necessary for maintaining good health (which is a precondition to actualise non-hedonic goods) to the point of need and sustenance (if there is really such need) (and with due care for their welfare), but not for mere taste or pleasure (hedonic goods).  This can also explain the "common sense" position that humans have "rights in the strong sense" and cannot be treated as a mere means for any amount of human or animal utility.

Due to length concerns, I will not go into more depth. For a more detailed and comprehensive analysis, Shellby Kagan’s “How to Count Animals, more or less” is a good source. 

Not giving too much weight to arguments from marginal cases

This proposition does not immediately resolve all common sense matters. Critics can rightly point out that there will be cases where certain individual members of certain species will have less capacity than the stereotypical members of its own group. According to this argument, this hierarchical ordering can only be coherent if one concedes to treat those members as individuals with “lower” moral status. For example, if hierarchism is to be coherent, one needs to treat a mentally handicapped human with equal moral consideration of an animal that has a similar intelligence (like an ape or a dog). 

Of course, this goes well against common sense. In order to remain coherent, one needs to make some adjustments. There are some philosophical moves available here. I am not going to go into detail about all of them. For brevity, I will mention one that striked me the most. This are two passages from Jeremy Waldron’s One Another’s Equal: The Basis of Human Equality:

“I remember that Peter Singer was once giving a presentation at Columbia University in which he was pursuing his line about disabled persons and well-functioning chimpanzees. A woman in the audience raised her hand to ask a question. She said something like this: 

“I have a profoundly disabled daughter, and you cannot understand the disability she suffers without understanding that this is a girl made for speech who can’t speak. This is a being made for thought that can’t think clearly. This is a person made for neural control of movement who can barely control motor functions. You can’t understand the situation she is in except for her being, in a sense, wired for speech, for reason, for neural control of movement and the wiring not working. You can’t understand her predicament unless you understand the potential that has been frustrated in her case. She is not like a well-functioning chimpanzee.””

… 

“When we deal with the profoundly disabled, we are determined to include them as humans and as our equals—grimly determined as a tribute to the nature they have so tragically failed to fulfill. I think we bend over backwards to give to those who are profoundly disabled as much of the benefit of basic equality, equal worth, and human dignity as we possibly can. No doubt there will be normative implications that cannot apply. There will be a need for continued guardianship and intense care; there will be less concern about independence. There will be choices that are normally privileged by rights that become problematic in the case of the profoundly disabled. A profoundly disabled person may not have or be able to conceive a view or a preference to express as a vote in an election, for example, though of course that does not mean that a vote to protect or advance his interests could not be cast in his behalf by a guardian. Other normative implications of basic equality stand firm: equal consideration of interests, for example—pursued directly rather than through voting—and other forms of equal concern. And issues of justice and fairness are important, too, because of the great vulnerability of the profoundly disabled and the danger of their being neglected, exploited, or taken advantage of.”

Accepting the incommensurability of values and the possibility of different reasonable ways of doing good better

Another issue that needs to be addressed for “effective altruism purposes” is to give a reasonable framework for prioritising between non-hedonic goods, as well as between different animals. 

Utilitarian and hedonic approaches can give a simpler and mechanistic answer to this question: add the utilities and check for results. Given that the currency of all goods are the same (pleasure and pain), this is only a matter of calculus. A good example of how this can work out is the cross cause cost-effectiveness model by Rethink Priorities. 

But again, the probable outcome of the calculus goes way against common sense as pointed at the beginning. But the begging question here is that what is the alternative solution for figuring out how one should act to do the most good? 

I think the common sense (and more correct) answer lies in the incommensurability of values. Incommensurability of values does not deny that hedonic goods or bads experienced by invertebrates do not constitute any value, but it argues that other goods, such as the use of reason of a human or the friendship between two dogs cannot be reduced to the same common measure as the suffering (or the lack of this suffering) of a worm. To rephrase, incommensurability of values does not need to exclude hedonic goods or hedonic goods experienced by invertebrates, but it also includes other goods as well and rejects the existence of a common measure to compare their “weight”. 

Incommensurability of values also begs many questions. Firstly, one can question the validity of this approach by examining its own possible outcomes. One may question that if incommensurability of values is true, it may be coherent to disregard extreme levels of suffering or need while pursuing other goods. For example, it may be coherent under this framework to read the newspaper while watching a child drown in a pond since “knowledge” and “health” are incommensurable. 

To reply, incommensurability of values is not incompatible with other moral principles, including absolute ones. One can perfectly justify the moral obligation to save the child instead of reading the newspaper in that context by applying the “Golden Rule”, rather than a utility calculus. If a person would have wanted to be saved by another person nearby reading a newspaper if he was drowning in a pond, one should treat that person who is drowning similarly and save that person.  

A related second begging question might be the problem of prioritisation. An effective altruist can question the validity of this approach by checking the reasonableness of its probable outcomes in career or donation recommendations. One may question that if incommensurability of values is true, it may approve cases where one chooses to donate to ineffective charities or not impactful careers since there is no measure to compare them by a common measure. While a person might pursue a good with good intentions, their actions would inevitably create the side-effect of not realising other goods. For example, when one pursues a career in a field, it leaves other fields less talent-dense than otherwise could have been. Would it be a mistake to dismiss these outcomes all together by relying on the idea that we can’t compare them according to incommensurability of values?

First response here is that incommensurability of values, does not argue that one should not pursue for efficiency (or effectiveness) when one pursues to achieve a good. There can be common measures that can allow comparison for the same (or very similar) values and it would be perfectly reasonable and coherent to make comparisons in those cases. For example, if one pursues to relieve poverty of the needy, it makes perfect sense to prioritize interventions or charities that relieve more poverty rather than less. To make it fit more into the EA context, it makes perfect sense to prefer donating to Give Well charities rather than Play Pumps in Africa if one aims to help the world's poorest. Or it also makes sense to pursue a career in an organisation that is more effective at actualising certain goods than pursuing a career in another organisation that is less effective at actualising the same or similar goods. 

But it is true that incommensurability of values does not provide immediate tools to weigh in different values. For example, how should we prioritise funding or reform efforts between health, education or housing?

Similarly, and more related to our topic, incommensurability of values does not provide immediate tools to weigh in different values that are actualised by different species. For example, should we help chickens by staging corporate campaigns or help humans by distributing bednets? Which one has more value?

Let’s extend the list: 

Helping millions of shrimp by ending their suffering at the end of their lives or helping thousands of chickens by stopping the industry from putting them in cages? (Helping animals that have “lower” moral status with higher numbers which suffer from shorter yet higher intensity pain vs. helping animals that have “relatively higher” moral status with lower numbers which suffer from longer yet relatively less intense pain.)

Working on corporate campaigns or working on wild animal welfare research? (Helping some animals now with relatively higher probability of success or working to find ways to help more animals in the future with relatively lower probability of success.)

I don’t think there are any clear “better” or “worse” decisions between these options, and I don’t think this is even intelligible to find such correct answers through abstract reasoning. The reason for this outcome is that the elements in these dilemmas don’t have common metrics that allow comparing them mechanically through abstract reasoning. 

But this does not mean that one should be blind to certain matters. It is evident that there are multiple goods, and one can act to actualise multiple goods. Incommensurability does not mean that there cannot be certain reasonable “proportions” between goods, even if these proportions cannot be mechanically represented, and it is in fact practically reasonable to seek actions and outcomes that are proportionate. While values (or virtues) like friendship, knowledge, health, justice, welfare, aesthetics, or charity are not commensurable, it does not mean one cannot act for outcomes that are proportionate. For example, since we are living in this world and at this time, it is reasonable to be considerate about extremely neglected and large scale problems like factory farming and invertebrate suffering.

The solution of these dilemmas are a matter of practical wisdom, and require deeper considerations of individual aspects of each context for each person. If we have to force to square all these philosophical considerations that are built through centuries of inquiry into an EA jargon it can also be called “personal fit”, in a very broad sense. 

This last move might be viewed as an unfair “explaining away”. But note that the same moves are played by the proponents of utilitarianism, hedonism and commensurability, in order to avoid the obvious extreme outcomes that few people would reasonably accept. For example, it is widely argued that donating %10 of one’s income to effective charities is a very good act. I agree with this, since in a lot of cases where people have achieved a decent level of welfare and financial security, it seems “proportionate” to donate %10. 

But as pointed out in different ways above, this is not what an utilitarian, hedonist and commensurability moral framework proposes. The only correct moral action in this framework is to maximise total hedonic goods. This can mean donating much more than %10 (anything above essential needs) in order to purchase electrical stunners to increase shrimp welfare or donating similar amounts for more research into wild invertebrate welfare on the expectation that this might deliver more utility. Same logic can apply to career decisions. The difference between doing what would maximize expected (hedonic) value and %10 less, is not that different from the difference between doing %10 better and doing nothing - which explains the “impact anxiety” of some people who believe they don’t do enough even when they do a lot.  

Of course, in real life discussions these propositions are modified to make them appear somewhat more convincing. Many people might find the idea of donating %10 or choosing a career from a range of options that suits them reasonable. But almost no one would agree that one should exclusively dedicate their entire income or career for invertebrate welfare as the only legitimate option. Hence, it is good to “take the %10 pledge” or “choose an impactful career that fits you”. 

But as you can see these are also “explaining away” moves. And I think they should be considered as illegitimate moves (like moving bishops sideways instead of diagonally in chess) under a utilitarian framework. The accurate description of these moderate positions can only be “these are still very wrong and people who spend their money and time with their families, for humans or even for vertebrate animals are in fact immoral but this is still better than doing nothing (for invertebrate welfare)” or “since defending the correct and coherent utilitarian version will probably create a backclash or demotivation which will reduce overall utility, it is better to publicly argue for an incorrect and incoherent version which can hopefully deliver better outcomes”. It is important here to note that it is incoherent and contradictory to accommodate “moderation with reasonable background motives” in utilitarianism but it is coherent and non-contradictory to accommodate “efficiency” and “proportionality” considerations in a moral framework that is not utilitarian. 

Of course, one can also try to use “moral uncertainty” or “worldview diversification” considerations to water down the radical implications of utilitarianism. But I don’t think these “procedural” reasoning tools are enough. Even if one places a moderate credence on utilitarianism, it would be hard to reject its “normal” conclusions like donations or career choices (unless one faces very extreme cases like the option to end life on earth in order to end seemingly impossible to fix animal suffering). Given the overwhelming results of utilitarian calculus, one needs to discount utilitarianism extremely highly in order to reach a common sense conclusion - which would be artificial. 

I even think that these “moral uncertainty” or “worldview diversification” considerations “backfire” if one’s aim is to arrive at a common sense outcome. Even when one has a low (but of course not close to zero) credence in utilitarianism, one should still be putting a lot of effort into invertebrate welfare - again, because of their overwhelming numbers  [this might also be used as an argument against these considerations too - but I will not go deeper into this]. Without a substantial rebuttal of utilitarianism that really discounts its credence level by a lot, it is not possible to reach a commonsense result, even accounting for moral uncertainty and worldview diversification.  

Finally, I would like to make it clear that I don’t think that donating large amounts of money to shrimp welfare or wild invertebrate research, or pursuing a career in these fields are irrational or wrong. These can be one of many reasonably good donation and career options - with due care for proportionality with other goods of course (for example, it would be an injustice if a dependent of a person is really sick and needs financial help but that person donates a large amount for shrimp (or chicken or even human) welfare and dismiss the plea of help from their dependent). My point is that it is unreasonable to claim that these are the only reasonable options. And there needs to be a different moral framework  to account for this.