New Omnivorism: a Novel Approach to Food and Animal Ethics

New omnivorism is a term coined by Andy Lamey to refer to arguments that – paradoxically – our duties towards animals require us to eat some animal products. Lamey’s claim to have identified a new, distinctive position in food ethics is problematic, however, for some of his interlocutors are not new (e.g., Leslie Stephen in the nineteenth century), not distinctive (e.g., animal welfarists), and not obviously concerned with eating animals (e.g., plant neurobiologists). It is the aim of this paper to bolster Lamey’s argument that he has identified a novel, unified, and intriguing position (or set of positions) in animal ethics and the philosophy of food. We distinguish new omnivorism from four other non-vegan positions and then differentiate three versions of new omnivorism based on the kinds of animal products they propose we consume. We conclude by exploring a range of argumentative strategies that could be deployed in response to the new omnivore.

are more harmful than some ways of sourcing animal protein. Thus, if reducing harm to animals is our goal, we should be eating some animal products, rather than relying upon (those forms of) arable agriculture.
New omnivores can be located outside of the standard argumentative space in food and animal ethics. 2 Crucially, new omnivorism represents an important challenge to veganism, but one different from others prevalent in the literature. Here, we follow Jan Dutkiewicz and Jonathan Dickstein (2021) in adopting a practice-based definition of veganism as simply an abstention from animal-based products, which does not necessarily carry with it any moral or political commitments. 3 Veganism can be justified in all kinds of ways, but prevalent in the western philosophical literature -with echoes of arguments in the tradition's Greek foundations -are "basic arguments" about harms to animals (see, e.g., Hooley & Nobis, 2016). Roughly, basic arguments for veganism go, it is wrong to cause significant harm to animals when we could pursue less harmful options, and producing meat and other animal products causes significant harm to animals. Given that we have a possible alternative in the form of veganism, the production of animal-based foods is wrong. Consequently -though we acknowledge that there are contentious questions to ask about the relationship between the consumption and production of animal products (Fischer, 2020) -we should be vegan. Basic arguments can draw upon lots of different ethical theories, and lots of different understandings of harm to animals.
Incidentally, basic arguments allow exceptions. For example, basic arguments do not say that veganism is required for those for whom a secure, safe, and healthy vegan diet is inaccessible. This might include, for example, those living in the far north (where arable agriculture is deeply impractical), those who are highly food insecure, and those with health problems creating particular barriers to plant-based eating. But basic arguments do conclude that relatively wealthy consumers in the industrialized west should be vegan. New omnivores aim to rebut even this limited conclusion.
Basic arguments for veganism have received a range of challenges in ethics, some of which will be explored later. It is as an alternative to these existing challenges that new omnivorism has emerged. 4 New omnivores reject veganism -indeed, not only do they say that we may be non-vegan, but that we should be -but not by simply arguing that we owe little to animals, or that what we owe them is consistent with farming or hunting them. We propose the following definitions: 3 Crucially, therefore, we reject that one is still vegan if one eats only (assuming such things exist) ethically acceptable animal products. The person who eats vegan but for ethically acceptable animal products might have a permissible or even laudatory diet -but it is not a vegan diet. 4 Again, we are here talking about philosophy in the western, and especially analytic, tradition. Might new omnivore arguments be present in other traditions? Margaret Robinson (2016) explores arguments adjacent to new omnivorism in Mi'kmaq philosophy. Though defending veganism as a practice consistent with indigenous beliefs about animals, she looks to cultivated meat. This could -she says -provide a way for indigenous people to eat the foods important to them and maintain valued relationships with animals without the need for harmful farming, fishing, or hunting practices. Specifically, she asks whether the moose is still a brother if moose meat is not eaten. In so doing, she opens the door to an argument according to which a kind of relational harm is done to (certain) animals if (certain) people do not eat (certain) animal products. As will be seen, this is a different sort of harm to the one that concerns new omnivores. Nonetheless, we could imagine new omnivorism playing a part in the conversation in which Robinson is participating.
New omnivorism: A family of arguments favoring the consumption of animal products to minimize or avoid (certain) physical harms to animals.

And:
New omnivore: An advocate of (one of the family of arguments making up) new omnivorism, including (but not limited to) someone who practices a non-vegan diet as endorsed by new omnivorism.
Understood in this way, we contend that new omnivorism is distinctive, thus warranting delimitation; novel, thus warranting the descriptor new; and interesting, thus warranting scholarly attention.
We note that this conceptualization of new omnivorism is a refinement of Lamey's. In Duty and the Beast, his book-length treatment of new omnivorism, Lamey engages with positions that are decidedly not captured by our definition. Some of what Lamey engages with is not particularly new. For example, his discussion of the logic of the larder (Lamey, 2019, chap. 7) engages with Leslie Stephen's nineteenth-century arguments about (future, hypothetical) pigs having an interest in the continued consumption of bacon, for without it, they would not come into being (Stephen, 1896, p. 236, cf. the discussion of "benign carnivorism" below). Some of what Lamey engages with is not particularly distinctive. For example, he responds to Temple Grandin's arguments about killing animals for meat in ways that do not involve excessive suffering (Lamey, 2019, chap. 5) -but these considerations are both mainstream and familiar (cf. the discussion of "compassionate carnivorism" below). And some of what Lamey engages with does not obviously relate to food. For example, Michael Marder's work on "plant neurobiology" (see Marder, 2013) could be deployed as a challenge to animal ethicists, but Marder is not obviously concerned with the ethics of eating. Nonetheless, we believe that our conceptualization of new omnivorism is in the spirit of Lamey's enquiry, which is subtitled Should We Eat Meat in the Name of Animal Rights?
It is the aim of this paper to bolster Lamey's argument that he has identified a novel, unified, and intriguing position (or set of positions) in animal ethics and the philosophy of food. But, more than this, it is our aim to clearly capture what that position is, sketch the key contributions of new omnivores, and outline possible routes of resistance. First, we identify new omnivorism, noting three versions of new omnivorism based on the kinds of animal products that advocates propose we consume. Second, we distinguish new omnivorism from other non-vegan positions that have played substantial parts in analytic food and animal ethics. 5 Third, we interrogate new omnivorism by exploring argumentative strategies that could be deployed in response to the new omnivore. 5 Arguments drawn from other traditions will not be explored in this paper. Religious ethics, for example, sometimes contains arguments purporting to show that veganism wrongfully denies God's provenance. Henrik Lagerlund (2018, p. 764) points to Augustine and Maimonides as making arguments like this. And similar arguments can be found in Islamic thought (e.g., Ahmad, 2010, p. 20). Importantly, however, the putative wrong, here, is to the divine (or perhaps humans), rather than to non-human animals (as in new omnivorism).

Identifying New Omnivorism
To appreciate the impetus for new omnivorism, it is instructive to observe that standard vegan arguments -including basic arguments -are inconclusive. 6 Or, put another way, there is an argumentative gap between animal rights (understood broadly) and veganism. Even if animals have rights (understood broadly), there are sources of animal protein that do not obviously violate those rights: the bodies of animals killed accidentally; eggs from well-treated chickens; animal foods that would otherwise go to waste; the bodies of non-sentient animals; the products of cellular agriculture (e.g., cultivated meat); and so on. Healthy debate exists about the ethics of each of these products (see below), but the point is the in-principle gap between animal rights and veganism.
New omnivores situate themselves in the argumentative gap, and they do so by arguing that not only are there (relatively) harm-free sources of animal protein, but that plant-based sources of protein are not harm-free. Typically, this will involve the observation that animals suffer in industrialized arable agriculture (or other ways of acquiring plant-based foods). "All aspects of consumption in late capitalism", Lori Gruen and Robert Jones observe, "involve harming others, human and nonhuman" (Gruen & Jones, 2016, p. 157). Pesticides, field traps, land clearing, and mechanical harvesting contribute to harming various animals. Some of this harm is intentional (e.g., pest control), while other harms are unintentional (e.g., accidentally running over a mole). Examples are easy to come by: the burning of orangutans through deforestation for palm oil plantations (Nellemann et al., 2007); the effects of pesticides on Brazilian tapirs (Gonzalez, 2019); field animals killed by mechanical harvesting (Archer, 2011;Davis, 2003); fertilizer and pesticides killing birds and fish (Fischer & Lamey, 2018); land clearing for agricultural use killing mammals, birds, and reptiles (Finn & Stephens, 2017;Fraser & MacRae, 2011); and more (Hampton et al., 2021).
The insistence on animal harms in arable agriculture allows the new omnivore to argue for the consumption of some animal products. New omnivorism builds upon 7 the following empirical observation and moral claim: 1. Industrialized farming, including industrialized arable agriculture, involves physical harms to sentient animals. 2. We should act to minimize harm to sentient animals in food production.
These propositions support the incorporation of some animal protein when doing so promises to minimize harm to animals in food production. New omnivores posit that 1 and 2 support: 7 Here, we make an observation about existing literature, rather than a conceptual claim. 6 James Rachels observes there isn't a "sweeping" argument for vegetarianism; rather, "we will have to consider the things we eat one at a time" (2004, p. 79).

3
3. A certain mixed diet of plants and animal protein will minimize harm to animals in food production.
New omnivores disagree over what the "animal protein" is, and here is where the different kinds of new omnivorism are found. There are at least three kinds championed, though there is considerable space for novel proposals.

Large Mammals
The first kind of new omnivorism posits that (1) and (2) support the intentional killing of certain sentient animals for food. Steven Davis, whose challenge to Tom Regan's veganism represents an important starting point for new omnivorism, argues for a mixed diet of meat from large herbivores (i.e., beef) and plants. The basis for this claim is the observation that animals suffer in a variety of ways in arable agriculture. He estimates the number of animals killed in the United States to produce a plant-based diet to be around 15 field animals per hectare per year. He calculates that approximately "1.8 billion animals would be killed annually" (Davis, 2003, p. 390) to support plant-based diets for Americans. By contrast, adopting a diet that includes the meat of large herbivores -specifically, beef from grassfed cattle -would result in 1.35 billion animals killed. Mike Archer (2011) also supports the consumption of beef from pasture-raised cattle, given harms to mice in arable agriculture. The hunting of large mammals, too, has been supported using new omnivore arguments: such hunting, it is argued, is a less harmful source of protein than arable agriculture (Archer, 2011;Cahoone, 2009;Demetriou & Fischer, 2018).

Waste
A somewhat less controversial strand of new omnivorism 8 looks to the consumption of animal products that would otherwise go to waste. One proposal comes from Donald Bruckner (2015), who looks to roadkill. He argues that strict vegetarianism is immoral while roadkill is available, given the harms attributable to arable agriculture. A similar, more general, argument could be made about the eating of any animal products that would otherwise go to waste (Fischer, 2018). If food -even meat -would otherwise go to landfill, better to eat it than to financially support harm-causing arable agriculture. And this captures practices from the relatively mundane (eating a roommate's leftovers, pocketing sandwiches after catered events) to the more unusual (collecting roadkill, dumpster diving).

Invertebrates
Some new omnivores argue that animals who (or that) are not sentient (roughly, who lack the capacity to experience pain) may not be owed any particular moral consideration. Now, there are difficult questions to be asked about which animals are not sentient. Insects are a contender. C.D. Meyers argues that, because (first) insects are not capable of experiencing a conscious, felt experience of pain and (second) insect farming is less harmful than at least some forms of arable agriculture, "we ought to engage in and encourage entomophagy: the practice of eating insects" (Meyers, 2013, p. 119). Fischer (2016), meanwhile, bugs the strict vegan (and offers a twist on Davis's arguments) by observing that the quantity of insects killed in arable agriculture mean that opponents of entomophagy have some explaining to do. Even if there is a chance that insects are sentient, the numbers may come out in favor of eating insects, rather than killing them by the million (billion, trillion) in arable agriculture.
But insects are neither the least likely animals to be sentient, nor, perhaps, the most appetizing. Oysters are considered (almost) fair game by plenty of animal ethicists. 9 Their consumption is defended on new-omnivore grounds by the journalist Christopher Cox: No forests are cleared for oysters, no fertilizer is needed, and no grain goes to waste to feed them-they have a diet of plankton, which is about as close to the bottom of the food chain as you can get. Oyster cultivation also avoids many of the negative side effects of arable agriculture: There are no bees needed to pollinate oysters, no pesticides required to kill off other insects, and for the most part, oyster farms operate without the collateral damage of accidentally killing other animals during harvesting. (Relatedly, although it's possible to collect wild oysters sustainably, the same cannot be said for other bivalves like clams and mussels. These are often dredged from the seabed, disrupting an entire ecosystem. For that reason, it's best to avoid them.) (Cox, 2010) Melissa Kravitz echoes Cox's concerns about scallops/clams -though adds that there are sustainable gathering practices -but cleverly observes that This, however, brings up an entirely new facet to the scallop debate-if you're not eating wild bivalves because of bycatch (which PETA [People for the Ethical Treatment of Animals] stands by as an essential problem with eating seafood), what's the justification for eating kelp and other sea plants, which are harvested from the same depleting oceans? (Kravitz, 2018) On the face of it, then, the "ostrovegan" not only has a compelling case that ostroveganism is permissible, but that vegans should be joining them. Or, at least, they have a case that oysters and other bivalves are morally preferable to some plant-based foods. 10

Potential Arguments
These three strands of new omnivorism -large mammals, food waste, and (plausibly) non-sentient animals -are merely examples of foods that new omnivores might defend. All the new omnivore needs, however, is an animal-based foodstuff less harmful than a 1 3 comparable plant-based foodstuff. So, for example, Lamey (2019, chap. 9) supports the consumption of cultivated meat. A new omnivore case in favor of consuming cultivated meat (or other products of cellular agriculture) over plant-based protein sources is not hard to envision -but will depend upon empirical information currently lacking.
Similarly, too, we could imagine new omnivore arguments in favor of the kind of unusual eating (or, to coin a phrase, unusual farming) practices that are explored in the animal ethics literature. Could genuinely humane dairy or egg farming be less harmful than arable agriculture? What about farming animals and then eating their bodies when they die naturally? Or even farming genetically modified animals who die natural deaths while young? We are not aware of anyone who makes new omnivore arguments for these practices, 11 but it is not hard to imagine that someone could.

Distinguishing New Omnivorism
New omnivorism is distinct from veganism (at least, the practice-based understanding of veganism noted in the introduction) because it endorses the consumption of (some) animal products. But it is worth motivating our claim that the position is distinct from other nonvegan positions. In particular, we want to contrast it with three (more or less) familiar sets of views in food and animal ethics according to which it is permissible to eat animal products, and another set of views according to which we are (prima facie, at least) required to eat animal products.

Non-veganism Is Permissible, 1: Committed Carnivorism
Some reject veganism by claiming we have few or no obligations to animals. This committed carnivorism has a long pedigree, but a recent defender is Timothy Hsiao (2015). He argues that, since morality is about action in pursuit of the good, it follows that membership in the moral community requires that a being be capable of (a) knowing what is good and evil, and (b) acting for the sake of that knowledge. Of whatever the moral life consists, morality is something that can only be articulated, understood, and practiced by rational beings. Accordingly, a being has moral status only if it possesses a rational nature. Because animals lack a rational nature, animals lack moral status. This has far-reaching implications. Though Hsiao accepts that killing animals or making them suffer harms animals, he denies that this involves a moral wrong.
New omnivores deny Hsiao's claim that animals' pain is morally irrelevant. For new omnivores, suffering matters, no matter who suffers. Instances of inflicted pain require moral justification. The principle that inflicting harm requires justification nicely captures our intuitions about causing harm. Shoving a being lacking moral agency for fun is morally wrong, but shoving the same being out of the way of an incoming train to prevent her from being hit is morally laudable -in both cases, there is inflicted harm and an evaluable reason for causing that harm. Intuitively, the principle is as true in the case of an animal as 11 Lamey includes Jeff McMahan's proposals about genetically modified animals in his book on new omnivorism, and Alasdair Cochrane (who defends corpse farming) shows awareness of the harms to animals killed in arable agriculture -but neither McMahan nor Cochrane link their proposals with the harms of arable agriculture and neither argues that we should eat such animal flesh. See also Zamir, 2007;but cf. Milburn, 2019. 1 3 a toddler. New omnivores endorse these intuitions. Inflicting harm to another creature is permissible only with good justification, and greater harm requires greater justification. To deny that causing harm to another creature requires good justification is to court moral callousness that new omnivores are unwilling to accept.

Non-veganism Is Permissible, 2: Compassionate Carnivorism
Compassionate carnivorism, which goes by many names, is the position that it is permissible to raise and slaughter animals if it is done humanely. Michael Pollan summarizes the position by saying that " [p]eople who care about animals should be working to ensure that the ones they eat don't suffer, and that their deaths are swift and painless" (Pollan, 2006, p. 328).
Thus, like new omnivores and unlike committed carnivores, compassionate carnivores allow that we have duties towards animals. But unlike new omnivores, compassionate carnivores see animals' interests or status as insufficiently weighty to rule out routinely killing animals for human consumption. Instead, our duties to the animals we intend to eat are limited to ensuring that they live relatively good lives. For example, compassionate carnivores may hold that it is wrong to inflict (much) suffering on farmed animals, or that it is wrong to deny farmed animals the chance to engage in (certain) species-specific behaviors. But neither of these, says the compassionate carnivore, rule out farming animals. 12 Compassionate carnivorism thus starts to sound very much like the "ethic of animal welfare", which "accepts that humans are morally superior to animals, but that because the latter are sentient and conscious beings, they have some moral worth, albeit less than humans" (Garner, 2013, p. 78). Painless killing is (typically) deemed unproblematic, and even infliction of suffering is permissible if it is "necessary" for some human goal. What this means in practice is that humans are "entitled to sacrifice the fundamental interests of animals if by so doing benefits to humans, and perhaps other animals, accrue" (Garner, 2013, p. 78).
What separates compassionate carnivores from new omnivores is that the former's arguments are distinctly anthropocentric. This is true in two senses. First, compassionate carnivores do not assign the same moral weight to animals and their interests as they do to humans and their interests. New omnivores will assign this weight -or, at least, they will assign sufficient weight to offer strong prima facie reasons against animal agriculture. Second, compassionate carnivores favor the consumption of animal products because of the benefits said consumption brings to humans. 13 New omnivores, however, will say that we should eat animal products for the sake of animals. Eating certain animal products will allow consumers to minimize or avoid (physical) harms to particular animals. 12 Conceivably, a compassionate carnivore could rule out farming animals, and favor hunting and/or fishing (cf. Cerulli, 2012). This branch of compassionate carnivorism, incidentally, may start to sound closer to new omnivorism or unusual eating (the latter is introduced below). We allow that distinctions may be difficult to draw at the margins. 13 Indeed, a variation on compassionate carnivorism might say that we should eat animal products, because it would be wrong to turn down these benefits.

Non-veganism Is Permissible, 3: Unusual Eating
Another set of positions that are conceptually close to new omnivorism are what we can call, following Bob Fischer (2020), unusual eating. Unusual eating is simply about identifying those ways of acquiring animal products that do not, as it happens, violate our duties (or similar) towards animals, and arguing that they are permissible. It thus exists on a continuum with compassionate carnivorism, but, unlike compassionate carnivorism, starts from a non-speciesist (or at least strongly animal protectionist) position, like animal rights. Among animal protectionist philosophers, support for unusual eating may not, itself, be that unusual. Indeed, it has been noted (Milburn & Fischer, 2021) that there is a surprising (near) consensus among veg(etari)an philosophers that there is room for exceptions to vegetarianism, specifically in cases of meat that would otherwise go to waste. And "waste" meat is not the only animal product permitted by many ostensibly vegan philosophers. Others, for example, will leave room for the consumption of eggs from hens afforded good lives (Fischer & Milburn, 2019;Hooley & Nobis, 2016). Alasdair Cochrane (2012, pp. 86-9), meanwhile, defends the genuinely humane farming of animals for their eggs, milk, and even corpses.
New omnivores, on the other hand, are not merely arguing that it would be permissible to consume certain unusual animal products; instead, they are arguing that, for the sake of avoiding physical harms to animals, it would be good (or right, preferable, obligatory…) to consume these products. This is the distinctiveness, and novelty, of new omnivorism relative to unusual eating -and part of what earns it the label new.
Unusual eating, though newly named, is not a particularly new position. Indeed, it is present in Peter Singer's early work, making it familiar to animal and food ethicists. For example, Cora Diamond (in a very widely read article) sharply delineates her own vegetarianism from Singer's by observing that the latter would "be perfectly happy to eat the unfortunate lamb that has just been hit by a car" (Diamond, 1978, pp. 471-2) -Singer concedes the point (Singer, 1980, pp. 327-8).
However, we allow that new omnivorism is clearly related to unusual eating. Indeed, new omnivore arguments are deployed to bolster arguments that are more concerned with unusual eating (e.g., Milburn, 2018, p. 276). That is, the argument is made at length that the consumption of a given animal product is permissible, and then new omnivore arguments are introduced to indicate that the product may even be morally preferable to competing plant-based products.

Non-veganism Is Required: Benign Carnivorism
New omnivorism can be distinguished from the arguments so-far surveyed in that it concludes that we should be eating animal products, rather than simply that we may. There is, however, at least one other position that has played a significant part in food and animal ethics concluding that we are required to eat animal products. This position, going back at least to Leslie Stephen in the nineteenth century, has been recently defended by Nick Zangwill, who puts it like this: The basic claim, to put it crudely at first, is that eating meat is morally good primarily because it benefits animals. Of course, the practice does not benefit a particular animal that we eat at the time that we eat it. Nevertheless, the existence of that animal and animals of that kind depends on human beings eating animals of its kind and, hence, that meat-eating practice benefits them. (Zangwill, 2021, p. 295) Borrowing from McMahan (2008), we call this position benign carnivorism. The benign carnivore argues (with the new omnivore) for the consumption of animal products based on animals' interests. But there is no mention of the harms present in arable agriculture and the interest they appeal to is not the interest against physical harm. Instead, it is the putative interest that animals have in being born. This interest is much more controversial than that appealed to by the new omnivore. While it is clear that animals have interests in being free from physical harm (if they have interests at all), 14 it is much less clear (to us) that they have an interest in being born.
Why separate benign carnivorism from new omnivorism? Despite the different kinds of interests appealed to, they could be lumped together; arguably, this is what Lamey does in exploring Stephens's arguments in Duty and the Beast. However, we believe that it is a mistake to label benign carnivorism a kind of new omnivorism, given that it goes back (at least!) to Stephen. What's more, Stephen's arguments (via Henry Salt and, more immediately, Peter Singer) have become familiar to animal and food ethicists over the last few decades. Thus, including benign carnivorism under the banner of new omnivorism risks undermining the latter's claim to novelty -both qua recentness, and qua unfamiliarity.

Interrogating New Omnivorism
New omnivores, despite their variety, share an argument: the empirical claim that animals are physically harmed in arable agriculture, coupled with the moral principle to avoid harm to animals, and the claim that adopting certain non-vegan foods would allow individuals to avoid harm. Much work around new omnivorism has focused on particular iterations of the position, and some new omnivores are critical of one iteration but not others. Cheryl Abbate (2019) is critical of (human) consumption of roadkill, but more open to entomophagy. Lamey (2019) is critical of Davis, but open to cultivated meat. Milburn and Fischer (2021) draw distinctions between different scavenged foods.
When faced with new omnivorism, the vegan could bite the bullet. After all, the new omnivore is likely advocating a near-vegan diet; they just identify a few exceptions. The freegan who is vegan but for the occasional venison sourced from roadkill, or the ostrovegan who is vegan but for oysters, is far closer to veganism than to someone following the (appropriately named) SAD: the Standard American Diet. 15 Even Davis, who defends beef farming, offers a proposal "radically unsupportive of the status quo in animal agriculture", calling, effectively, "for the complete abolition of intensive confinement and an end to poultry and pork production" (Matheny, 2003, p. 506). Or, provocatively, Davis's new omnivorism is a greater challenge to meat-eaters, including compassionate carnivores and benign carnivores, than it is to vegans.
Nonetheless, lots of vegans and vegetarians will want to resist new omnivorism. There are several argumentative routes open to them. Some of these dispute empirical premises in new omnivore arguments, such as by challenging claims about the numbers of animals impacted by arable agriculture. Some of them dispute the philosophical assumptions that new omnivores make, such as by offering arguments beyond basic cases for veganism against the non-vegan practices that new omnivores champion.

Disputing Numbers
It may seem that new omnivore claims about the number of animals killed in arable agriculture could be disputed. Gaverick Matheny argues that Davis assumes that "the two systems -crops only and crops with ruminant-pasture -using the same total amount of land, would feed identical numbers of people … In fact, crop and ruminant systems produce different amounts of food per hectare" (Matheny, 2003, p. 506). Correcting for this reveals, Matheny argues, that Davis has underestimated the number of animals killed in ruminantbased agriculture. Comparable mistakes are made by Archer, who, in his calculations of animals killed in the field, conflates the chances of a population explosion of mice over a large area with the likelihood of an explosion on a single farm (Fischer & Lamey, 2018). Both authors, meanwhile, could be accused of oversimplifying the calculations by considering only the number of deaths. One would also need to account for the extent to which the animals are harmed by death, and whether the animals are harmed in other ways (Milburn, 2022, p. 116).
The thought is this: once the numbers are properly accounted for, the new omnivore case falls down. But there are three problems. First, we simply do not know how many animals are killed in arable agriculture (Fischer & Lamey, 2018). Second, even if we did, the numbers may not favor the vegan case. Third, even if the numbers do side with the vegan here, they may not there. If the vegan is right about plant-based food A versus animal product X, the new omnivore could switch to plant-based food B, animal product Y, or both. There will (almost) always be a less harmful animal product or a more harmful plant product. The thought that the least harmful animal product is more harmful than the most harmful plant product is, we fear, implausible.

Disputing the Comparison
Alternatively, the vegan could argue that the new omnivore is counting the deaths of wild animals caused by arable agriculture in the wrong way. This is a slightly more conceptual variant on the above strategy -though the literal numbers are not disputed, the relevance of the numbers are. Lamey (2019, chap. 4) deploys the doctrine of double effect (DDE) to challenge Davis, drawing a moral line between the intended harms to cattle in pastoral agriculture and the foreseen but not intended harms to field animals in arable agriculture. The latter, DDE's advocates will hold, are a less weighty moral problem than the former. If one accepts the distinction -DDE is contentious -then Davis's case is weakened. DDE does not entail that non-intended harm is unimportant, however, meaning that it is not a knockdown argument against new omnivorism. But it does muddy the comparison between harms to field animals and harms to farmed animals. Perhaps the harm to farmed animals is more morally weighty even if the harm to field animals is greater.
A bolder version of the argument comes from Abbate, who argues that field animals are not all-things-considered harmed by arable agriculture. This is because these animals live decent lives made possible by arable agriculture and experience quick deaths -deaths sufficiently non-harmful that their lives are still, overall, worth living (Abbate, 2019, pp. 176-8). 16 Abbate is treading a similar path to the benign carnivore, as she is assuming that animals (here, field animals) can benefit from coming into existence. But she rejects benign carnivorism, which creates animals with the intention of later killing them. The creation (and killing) of field animals, however, is a foreseen but not intended side effect. Thus, Abbate's case also depends upon the DDE. Abbate's argument -granting the DDE and the claim that animals can be benefitted by coming into existence -means we may have good reason to discount some of the harm to animals caused by arable agriculture.
Some, as Abbate's argument refers only to harm that is 1) Non-intentional (thus, it cannot capture the deliberate poisoning of rodents); 2) To individuals whose life is worth living (thus, it cannot capture painful deaths of young animals); 3) To individuals who exist because of arable agriculture (thus, it presumably does not capture the harms to fish caused by agricultural run-off). May, because -even granting the contentious premises of Abbate's argument -we could hold that we should help animals once they exist. Compare a human case. Imagine nuclear waste was buried on Schrunit Island, 17 and it was foreseen, but not intended, that this waste could one day cause disaster. Schrunit Island is sold off, and people come to live on it. Generations pass. Schrunitians live good lives only possible because of buried nuclear waste. (Present Schrunitians would not have been born if Schrunit Island had not been sold, and it would never have been sold had it not been used for nuclear waste.) We might think nonetheless think that Schrunitians have a right to be rescued when nuclear waste starts leaking, heralding imminent catastrophe. Perhaps similar is true of field animals. Once they exist, even if they benefit from existence and would not exist without a particular harmful practice, they may have claims on us relative to that practice.

Toward a More Ideal World
Another move available to the vegan is advocacy for less harmful arable agriculture. On this view, animal deaths in arable agriculture are "fixable problems: we just need better regulation and further efforts on the part of animal advocacy groups" (Fischer & Lamey, 2018, p. 416). New practices and new technologies help minimize the number of field animal deaths. For example, no-till methods cut down on the use of heavy machinery; contraception as a form of pest control does not kill animals; and sophisticated fencing and deterrence can keep animals out (Cochrane, 2012, pp. 101-2;Fischer & Lamey, 2018, pp. 424-5). One of us (Milburn, 2022, chap. 5) argues that harms to field animals provide an argument for shifts to indoor, vertical agriculture.
Even if the problems are fixable -fixing them would be a monumental task -this response is limited. Most do not presently have access to plants grown without animal harms. Advocates of this strategy are clear that humane agriculture is a future possibility, not a present reality for most people. If future arable agriculture is completely harm-free, then future new omnivores have no case against future vegans. But that's not much help in deciding how to behave now.
Nonetheless, this imagined future may be action-guiding in two ways. First, it may entail present political obligations to push for changing agricultural practices and to speak 1 3 up for field animals. Second, it may entail obligations to favor more humane forms of arable agriculture when available. For example, perhaps the deaths of field animals suggest that, where possible, we should be growing our own food (Abbate, 2019, pp. 173-6;Schedler, 2005). While only those with time and space could grow enough food to remove themselves from reliance on industrial arable agriculture, many may be able to limit said reliance -some mix growing their own food, minimizing food waste, or favoring nonindustrial sources of plant-based food will be accessible to many consumers.

Feed, don't Eat
A further general strategy open to the vegan that does not involve disputing harms to field animals is to argue that there is something more harm-reducing that the vegan could do with animal products than eating them. Abbate, for example, argues that caring about animals should not motivate one to consume roadkill (and the same argument might apply beyond roadkill) because there are other options that could lower overall animal harm (Abbate, 2019, pp. 172-3). Her reasoning is compelling. A vegan diet results in less harm than a diet that includes meat from a factory farm. Thus, when a vegan comes across edible roadkill, she should donate the flesh to her meat-eating neighbors, who in turn will consume less flesh from factory-farmed animals. Or, if the neighbors do not want it, it could be donated to a local animal shelter, which would lower the amount of meat purchased to feed animals. And there are other options.
This point regarding roadkill (and more) is well taken. However, it does not show that a person should adopt a strict plant-based diet. As one of us (Bobier 2021) argues, there might be situations in which the best course of action would be for the person to consume roadkill; if, for instance, neighbors are away and the local animal shelter does not accept roadkill. If there is no option available that minimizes animal harm more, the new omnivore has a compelling case that it is obligatory for the vegan to consume roadkill. Abbate concedes this, but wonders "how important it is to defend this very uninteresting point" (Abbate, 2019, p. 173).

Challenge the Animal Products
Many responses to new omnivores focus on challenging the specific animal products championed. Either these products are more harmful than envisaged, or their consumption/production would be wrongful regardless of harm. Let us consider some examples.
It might be objected that we are unsure whether insects can feel pain (Klein & Barron, 2016;Birch, 2020). The science is unsettled, so we should err on the side of caution (Birch, 2017;Knutsson & Munthe, 2017). After all, it matters if an animal can suffer, and so we should be confident that insects lack sentience before doing something that would harm them. Alternatively, one does not need to take a strict precautionary principle -i.e., give insects the benefit of the doubt -to worry about eating them. An expected utility principle, assigning only some chance that a given invertebrate is sentient, may well come out in favor of not eating them (Sebo & Schukraft, 2021). Recall, though, that this could be disputed by observing that insects, too, are killed in arable agriculture (Fischer, 2016). We cannot hope to settle this argument. Suffice it to say that there a range of resources that the vegan could draw upon to argue that we should not be eating insects (or bivalves, or what have you) if, as is surely plausible, there is a chance that these beings are sentient (Sebo, 2018).
These challenges concern new omnivore arguments for the consumption of (plausibly) non-sentient animals. Other new omnivore arguments need a different response. For example -though they are somewhat orthogonal to this enquiry -there are many published cases against cultivated meat. Carlo Alvaro (2019) contends that the desire to create cultivated meat is unvirtuous; John Miller (2012) worries that cultivated meat reinforces both capitalist and speciesist social structures; Rebekah Sinclair (2016) argues that "meatless meats" reinforce animals' edibility and killability. There is much less scholarly work on (or public awareness of) other forms of cellular agriculture -even though they, unlike cultivated meat, are already here in the form of (for example) animal-free rennet (Newman, 2020).
More firmly in the conversation about new omnivorism, Milburn and Fischer (2021) offer an argument against the consumption of (some) scavenged meat regardless of the extent to which it is harm-free. They argue that, if animals are a part of our political community, their bodies are owed respectful treatment. It is not (as it happens) considered respectful to consume the corpses of our human co-community members in (most?) existing communities. Thus, we should consider it disrespectful to consume the corpses of our animal co-community members, ruling out the consumption of leftover or scavenged meat when it comes from animals who are co-members. (Milburn and Fischer follow the lead of Donaldson and Kymlicka (2011), arguing that domesticated animals, but not wild animals, are members of our communities.) While this does not prove that the new omnivore case for scavenged meat fails, it does indicate that the argument is more complex than first thought.
Of course, there are other considerations that may tip the scale in the vegan's favor. The point here is that a proper account of the harms associated with different possible diets will not count only the harms directly caused by the practices, but other, less obvious harms. Maybe there are opportunity costs. No doubt an activist could do more good for animals by working overtime and donating to effective charities than spending an hour out dumpsterdiving. If the activist would have been working overtime for charity, better, likely, if they do that and stay vegan. Or maybe there are harms to third parties that need to be considered. For example, roadkill, animal products in bins, and invertebrates harvested from the wild need not be foods that will go to waste (Milburn & Fischer, 2021, fn. 2). If they are not eaten by humans, they will be eaten by other animals.

Conclusion
New omnivorism is a coherent, novel position that occupies the following logical space: because of concern about physical harms inflicted on animals -field animals included -we should eat certain animal products. The farming, fishing, and hunting of animals is (generally) morally wrong, but there are other, relatively morally innocuous ways to procure animal products. And, more, these innocuous routes to animal protein are seemingly less harmful than (some forms of) arable agriculture.
New omnivorism is not, it should be clear, a diet. Instead, it is a family of justifications for a family of (more or less) near-vegan diets. Collectively, these arguments represent some of the most compelling challenges to veganism, as they are able to turn the vegan's arguments against her. We call for greater engagement with and attention to new omnivore arguments, and predict that it will be forthcoming.