We welcome back philosophy contributor Eliot Michelson whose controversial first post stirred things up a bit. In his promised follow-up, Michaelson continues searching for solutions to the worries presented over the summer. I hope you’ll continue to share you thoughts below as we develop more sound arguments for ethical veganism.

– Joshua Katcher, Editor

By Eliot Michaelson, Ph.D. Candidate

After my last post, several of my friends in philosophy started pushing on me the thought that perhaps we have a duty not to eat animals — a duty stemming from the fact that animals exhibit certain qualities (sentience, an ability to feel pain, or whatever) that suffice to grant them moral significance.  (Interestingly, some of the most adamant proponents of this line of thought have been non-vegetarians, which has started to make me feel like I’m inhabiting a strange inverted world.)  These discussions got me thinking: might we have such a duty?  I should preface my remarks by saying that I’m very sympathetic to the idea that we have lots of duties to each other, and to animals.  I have a duty not to kill other people, for instance, nor should I kill animals.  In fact, I think I should also help people (and animals) out where I can, when I can.  And I don’t mean that I just think it’s nice for us to do these things if we feel like it.  No, I actually think it’s mandatory to help people out where you can.  And I think that ‘where you can’ is a lot more inclusive than just about any of us are living up to.  All that is to say, I’ve got nothing against duties; I think there are lots and lots of them.  And I think that recognizing these duties is an important part of coming to better understand our moral lives.  Still, I wondered, do we have a duty to be vegetarian?

Now, one might think that if we have a duty not to hurt or kill animals, which I just granted, that settles things: we should be vegetarian.  But, unfortunately, I don’t think that’s where the story actually ends.  The problem is that, in asking about vegetarianism, we’re asking about a different kind of duty than when we ask about whether each of us, individually, should refrain from killing animals ourselves.  With vegetarianism, we’re really interested in a duty not to take part in a system that results in many, many animal deaths.  That’s what we do when we purchase meat these days; most of us don’t kill animals ourselves (a few Brooklyn hipsters aside).  Whereas I think that I have a decent, if tentative, grasp of many of my individual duties, these systemic duties are trickier.  For instance, should I refuse to pay my federal taxes because I don’t support the war in Afghanistan or farm subsidies, even though I do support various of the social programs run by the federal government? (Or, should I insist on just paying the X% that goes to programs I approve of?)  I submit that the answer to this is not at all clear.  It may well depend on what value I generate with my life by not ending up in jail, or what value I might generate by going to jail.  (If I could write even half as well as some of my heroes did when in jail, then I almost undoubtedly should abstain from paying my taxes.)

In any case, though I hadn’t planned on heading down this route initially, the suggestion that we have a duty to be vegetarian has gotten under my skin.  So, it’s time for a brief foray into the possibility of a ‘Kantian’ justification for vegetarianism.  I call this sort of justification ‘Kantian’ (there’s not just one response here, but really a class of related ones) because they are collectively inspired by the work of the great 18th century German philosopher Immanual Kant.  Kant covered a lot of ground in his work, but here we’ll focus on a way of thinking about ethics he pioneered in the Groundwork of the Metaphysic of Morals and several subsequent works, a way of thinking about ethics that puts the notion of duty front and center. (It is worth noting that Kantian ethics is still an active research project today, and Kant’s influence can be found in a number of prominent animal rights theorists — including even contemporary consequentialists like Peter Singer.  Thus, in calling a sort of response ‘Kantian’, I am not suggesting that it amounts to a mere historical curiosity.)

I begin by recapping the futility worry.  That worry ran like this: first, if there is no causal connection between any particular choice to abstain from eating animals and reduced animal death, then one of the strongest reasons for being vegetarian or vegan would evaporate.  In other words, if we were motivated to be vegetarian by a desire to make the world a better place via directly reducing animal suffering and death through our economic decisions, then, if there turns out to be no such causal connection between these things, this would turn out to be a bad justification for vegetarianism.  The second part of the worry involved an argument to the effect that it is in fact extraordinarily likely that there is no causal connection between any particular decision to abstain from eating meat and decreased animal suffering.  The problem, in a nutshell, is that there is too much noise in the economic signaling system, combined with a high tolerance for waste — meaning that the expected effect of any of our particular choices to abstain from eating meat is zilch.  Abstaining from meat, according to this line of thought, turns out to be a futile gesture, one with no impact on actual animal welfare.  Thus, unless we can offer some different sort of justification for being vegetarian, we find ourselves without a good reason for thinking that eating meat is bad.  (It’s important not to confuse this with an argument for eating meat; that would be something entirely different from what I’ve provided.)

Kantian-leaning vegetarians seem to have an appealing way out of this conundrum.  Namely, they would seem to be able to offer us an explanation of why purchasing and eating meat is bad even if any particular decision to eat meat is highly, highly unlikely to result in increased animal suffering and death.  We shouldn’t eat meat, claims the Kantian, because we have a duty not to.  In other words, Kantians look well situated to simply grant all the assumptions that drove the futility worry, but to argue that we should be vegetarian anyway.  Plausibly, not all of our duties are justified in terms of expected outcomes.  Sometimes, you should just do X because it’s the right thing to do, even if it is futile.   (One important side-note: Kant himself was no vegetarian, and many contemporary Kantians also eat meat, and think themselves to be justified in doing so.  Both Kant and these non-vegetarian Kantians will reject certain of the premises below, but I won’t dwell on that here.)

Here’s how this sort of justification for vegetarianism might go: according to Kant, whenever we act, we should ask if the maxim under which we are acting is universalizable — and we should act only according to those maxims that are, in fact, universalizable.  This https://i2.wp.com/upload.wikimedia.org/wikipedia/en/thumb/3/3f/Immanuel_Kant_3.jpg/220px-Immanuel_Kant_3.jpg?w=840probably just sounded like gibberish to many readers.  Fear not.  I will do my best to explain what each of the key terms here (‘maxim’ and ‘universalize’) means in plain English.  Then I’ll consider how this rubric for good action can potentially be used to motivate vegetarianism, even in the face of the futility worry.  One thing I do not propose to do, however, is to explain why exactly we might be justified in endorsing this ‘universalizability principle’.  Kant’s reasons for endorsing the principle were complicated, and subsequent attempts to justify it have been complicated as well.  Once I’ve explained the technicalia, it will hopefully hold at least some intuitive appeal.  But if you don’t end up feeling this intuitive pull, all I can really do is point you towards the (vast) literature on the subject.

So what is a ‘maxim’?  In the sense in which we are interested here, a maxim is essentially a way of describing an action, paired with both (i) a motivation for that action and (ii) an endorsement of acting in that way for the given reasons.  Crucial to understanding this notion of a maxim is the observation that there are typically multiple ways of describing any particular action.  For moral evaluation, what often matters is the sort of background practice in which that action is embedded — something that may be highlighted by some descriptions and downplayed by others.  For instance, to borrow an example from J. L. Austin, when I say ”I do”, I might be described as merely saying ”I do”, or as undertaking a vow or oath.  The latter description will tend to be significantly more important, more useful, for purposes of moral evaluation; saying ”I do” is, in itself, typically neither good nor bad.  But some vows are good things to undertake, while others are not.  It would therefore be odd to endorse or criticize saying ”I do” in general, but it is perfectly reasonable to either endorse or criticize undertaking certain kinds of oaths.

Next we move on to ‘universalizability’.  We’ll say that a maxim is universalizable just in case it is the sort of action that one can consistently approve of everyone doing whenever it arises as a possible action (note: what this notion of ‘universalizability’ amounts to is extremely controversial in Kant scholarship; I’ve done my best here to capture the gist of the idea, but this certainly isn’t all there is to the story).  One standard example of a non-universalizable action is supposed to be lying.  Consider acting under the maxim: it is okay to lie whenever it proves convenient to do so.  Now suppose we universalize on this maxim: we thus conclude that it is okay for anyone to lie whenever it is convenient for him or her to do so.  But in a world where everyone lives by this maxim, there will be no reason for anyone to trust anyone else; after all, they may be lying if it happens to be convenient for them.  For reasons along these lines, Kant thinks that it is self-undermining to endorse a maxim like ”it’s okay to lie whenever it proves convenient to do so.”  This, in turn, means that this maxim fails to be universalizable.

Now we’re ready to try out a Kantian defense of vegetarianism.  Consider any particular decision to purchase meat.  If one decides to purchase that meat, for instance, then one will be acting under, and thus implicitly endorsing, a maxim such as: it is okay to purchase meat when one desires it.  Does everyone acting on this maxim lead to an intolerable situation?  Well, if we accept that we have a duty not to kill animals, then yes.  If we universalize on this maxim, then everyone will turn out to be permitted to purchase meat when they desire it.  And this is very likely to lead to quite a few animals being killed.  That, in turn, means that some specific persons are going to have to violate their duty not to kill animals — which is bad.  So it would seem that the Kantian has a reason to claim that we should not purchase meat: we cannot universalize on the maxim that it is okay to purchase meat supposing that one desires it without the result that at least some members of our society will have to violate their duty not to kill animals.  It would seem, therefore, that we have reason to be vegetarian.

But, unfortunately, things aren’t quite this simple.  The problem is one that the reader may have already anticipated: it hardly looks plausible that there is just one monolithic way of purchasing meat.  And the moral valence of buying meat may well vary depending on the sort of practice in which any particular meat-buying is embedded — much like the moral valence of saying ”I do” seems to depend on when and why one is doing it.  Even most vegetarians, I take it, are going to find it intuitively plausible that my purchasing scavenged road-kill in order to feed the ocelot that I rescued from a wildlife smuggling ring (and which lives in a special enclosure in my ample back yard — certainly not in my house) is very different from buying some pink plastic-wrapped stuff in a grocery store.  But these are both acts of buying meat — so why does the one seem so different from the other?

Again, the Kantian has a plausible answer here: the maxim you are actually acting under is slightly more specific than just ‘it’s okay to purchase meat when you desire it’.  Rather, it’s something more like: it’s okay to purchase meat when you desire it and as part of a wider socio-economic practice of raising and killing animals for food, often causing them intense pain and suffering in the process.  Now we’re in business!  If we universalize on this maxim, we clearly end up endorsing a world filled with animal suffering and death.  Not only is this plausibly bad in itself, it also requires that there be people violating their duties not to kill animals (notice how this stands in contrast to road-kill purchasing, which requires no such thing even if we all start purchasing road-kill).  So, it seems that we plausibly cannot universalize on the maxim under which we seem to be acting when we buy meat at a store.  In other words, we have a duty not to purchase meat.  Once more, vegetarianism would appear to be vindicated.

But, once again, things are not quite so easily settled.  The problem is that, once we notice that even decisions to purchase non-road-kill meat don’t necessarily result in any more animal suffering or death — due to the noisy, waste-tolerant nature of the  contemporary meat production and sales system — it’s no longer so clear that a meat purchaser will necessarily be endorsing a practice of killing animals for food, even when he or she purchases perfectly ordinary meat in a perfectly ordinary manner.  One way of putting the problem is this: it’s easy to classify which practice an action is a part of when there’s only one option.  But now, there seem to be two competitor practices when considering the act of purchasing meat: one might involve endorsing a wider practice of killing animals for food while the other might only involve endorsing the practice of reducing meat-waste so long as it is available.  Now, plausibly, most people when they purchase meat really are endorsing the practice of raising and killing animals for food.  That I grant.  But suppose that you want to eat meat and you find yourself aware of the economics of the meat-production system.  Now, it seems that you might reasonably fail to endorse the practice of raising and killing animals for food while still purchasing meat.  You realize that if you don’t purchase that meat, it really is going to be thrown away (or, at least some identical-looking piece of meat is).  You might even — consistently it seems — engage in political advocacy to end the practice of selling meat in your society.  You’d prefer that the meat-wast not be there at all.  But, given that it is, you decide to engage in a project of meat-waste reduction, knowing full well that your project is not going to lead to any increase in animal suffering or death.  If meat-waste were not available, you would in fact make no attempt to purchase and consume meat.

What’s going on here?  Basically, we’re facing a deep problem for both Kantian ethicists as well as ethicists of many other stripes: the need to discern which practice some particular action is best conceived of as being a part of, from amongst several competing candidates, for the purpose of moral evaluation.  One appealing option is to allow the agent’s intentions to determine which practice he/she was taking part in.  This can seem quite natural: if I ride my bicycle to work because I want to be a part of a wider movement to rely less on cars, then it seems like I am a part of that movement.  But if I ride my bike to work simply because I like biking, it seems like I am not a part of this wider movement.  If it’s laudable to be a part of this movement, then it seems like I can be praised in the first instance, but not in the second.  In the second case, I just happen to like riding bikes.

But notice that, if we allow the agent’s intentions to determine which practice she is taking part in, and thereby endorsing — killing animals for food, or merely reducing meat-waste so long as it is available — then we don’t have a Kantian defense of vegetarianism after all.  What we have is Kantian grounds to not eat meat while thinking of one’s meat purchases as part of a wider practice of killing animals for food.  Once one is appraised of the facts, however, one can simply continue to eat meat while thinking about things slightly differently.  That is, one can continue to eat meat, but conceive of this meat eating as part of a wider practice of meat-waste reduction, given that such meat-waste is widely available.  Or, so it seems at least.

At this point, it is open to Kantians to claim that, for some reason, the practice that one is really taking part in is one of killing animals for food, even when one knows that one’s individual choices to eat or abstain from meat have absolutely no effects on how much meat is produced in the world.  The Kantian might claim, perhaps, that this is the default practice in which one is engaged — since it is the way that most people in our society think about purchasing meat.  But, again, this doesn’t seem to preclude one from being slightly more radical: from both eating meat and loudly advocating for its no longer being available for sale.  In that case, it seems like one would have a good claim not to be participating in the practice of killing animals for food, but only in the practice of reducing meat-waste.  If that’s right though, then the Kantian doesn’t have an argument for vegetarianism; rather, she has an argument for either vegetarianism or for meat-eating coupled with political advocacy of a certain sort.

This last observation points, I think, towards a larger and more interesting question: given that we, as enlightened vegetarians, know that our choices fail to directly reduce animal suffering and death, in what sense are we not also tacitly endorsing the practice of killing animals for food?  That is, in what sense are we actually obstructing the production of meat in our society?  One might be tempted to say that we do so by denying the meat industry our resources.  But, again, none of us, as end-product consumers, ever hand resources over to the meat production industry — even when we buy meat!  What we do is deny grocers and restaurants some resources, maybe.  But we also buy more vegetables and legumes, etc. in order to make up for these calories.  So we don’t really deny grocers much of anything.  And, given the waste-tolerance that both grocers and restaurants generally exhibit, it’s not at all clear that we are actually denying the meat-industry any resources with our individual choices not to eat meat.  So, if the dominant attitude in our society is that killing animals for food is acceptable, and if we are not doing anything to actively disrupt that attitude, it is unclear why we should count as not taking part in that practice as much as our animal-eating peers.  In other words, we risk being complicit in this immoral practice.

Now, it might be that there is a similar way out available to us as there was for the meat-eater: we must advocate for change in order to really count as not condoning the practice of killing animals for food.  But if we were looking to justify vegetarianism, this looks like an odd justification.  The way we seem to be able to best comply with our obligations to animals, it turns out, looks to have little or nothing to do with whether or not we, individually, eat meat.  Rather, it looks to have to do with how we conduct ourselves politically.  That’s strange — at least to the extent to which it is not much of an argument for vegetarianism.  And recall that this was on the assumption that we were going to need to appeal to something beyond the agent’s intentions in order to determine which practice her action was a part of.  If the agent can simply intend that her meat purchases be conceived of as part of the practice of reducing meat-waste, given its widespread availability, then she would seem to be exempt from criticism whether or not she engages in further political advocacy.  So, again, it seems that the Kantian assumptions and methods with which we began have left us without a solid argument for vegetarianism.

Here then, in summary, is the crux of the problem: Kantians typically start by acknowledging that our activities can’t be morally evaluated without knowing something about the wider practices in which they are embedded.  One standard way of determining which practice some action is a part of is to defer to the agent’s honest intentions on the matter.  But that in turn opens the door to eating meat so long as one is aware of the fact that one’s decision to do so is will have no discernible effect on animal welfare.  So, unless the Kantian can offer some other way of determining which set of wider practices one’s actions should be embedded in, Kantians seem to lack an adequate way of motivating vegetarianism.  They have an argument that so long as one conceives of oneself as endorsing killing animals for meat in purchasing meat, then one should be vegetarian.  But, so long as one has put some thought into things, one can consistently and honestly disclaim any endorsement of this practice while still purchasing meat.  It’s true that many people might only endorse such  a claim disingenuously.  The problem though is that we wanted a plain and simple argument for being vegetarian — and we haven’t found one here.  (A further problem is that, if we can figure out an intention-free way of describing which practices one tacitly endorses in one’s actions, it’s unclear why non-disruptive vegetarians will count as failing to endorse a practice of killing animals for food.)

While none of this is to say that Kantians will not be able to rise to this challenge, to me at least, this challenge looks fairly daunting.  I take it that one of the good aspects of Kantian ethics is its ability to account for why our intentions do — within certain bounds at least — matter for determining whether our actions are right or wrong.  They matter because we care about the sorts of wider social practices we are endorsing in acting the way that we do.  But what we endorse is something that we, as agents, would seem to have some say in.  Thus, given what we know about how our individual meat purchases affect animal welfare (in all likelihood, they don’t), room inevitably seems available here to eat meat without endorsing the social practices we find most troubling about meat consumption.  That makes the Kantian program look rather unappealing, at least on the face of it, as a way of motivating vegetarianism.

That said, there are many flavors of Kantianism — and no doubt some Kantians will have proposals for how to amend their views so as to face down the futility worry.  I look forward to seeing such attempts.  In the meantime, however, I’m going to push ahead in further posts and explore other sorts of arguments we might give in favor of vegetarianism.  As we’ve seen, consequentialist accounts aren’t going to do it without further modification, and Kantianism, at least of the sort we explored here, doesn’t give us the sort of straightforward argument for which we might have hoped.  Still, thankfully, these don’t exhaust the sorts of arguments that can be mustered in favor of vegetarianism.  And, hopefully, their failures will serve to help us develop better alternatives.