Impact statement

De-extinction is a conceptually and ethically challenging topic. Through careful philosophical reflection, we can recognize it is a logically coherent course of action, though it is not always ethically advisable. Sometimes it is morally impermissible.

What is de-extinction?

Many think that once a species goes extinct, the species ceases to exist. However, recent scientific work on de-extinction suggests the possibility of “resurrecting” lost species. From a philosophical perspective, this raises two questions. First, is de-extinction even possible? Second, should it be done even if we can do it? In this essay, we consider proposed answers to both questions.

To begin, we need to get clearer on what exactly de-extinction is. One problem that immediately appears is that de-extinction seems to be impossible by definition. Here is an example of this worry from Alastair S. Gunn.

If a species is extinct, then it cannot exist in the future. If it cannot exist in the future, then it cannot be resurrected. But de-extinction just is the resurrection of an extinct species. Therefore, de-extinction is impossible. Before I examine this argument, I want to consider a terminological worry. As Beth Shapiro notes, “de-extincted” is an awkward and unappealing verb. I will describe the process of de-extinction as resurrection when we can bring back the very species that disappeared. I will use the term recreation when we create organisms very similar to those of an extinct species but that are not necessarily the same species as it. Resurrection is sometimes associated with bringing back the very organism who has died, and this is not what is being discussed in debates regarding de-extinction. We are discussed higher levels of organization such as populations, species and the like.

To see the error in this argument, we must consider what “extinct” means. The Concise Oxford Dictionary of Ecology defines it this way, “Applied to a taxon no member of which is living at the present time” (Allaby, Reference Allaby1992, p. 150). However, this would imply when plants of a species are dormant, it has gone extinct. Yashina et al. (Reference Yashina, Gubin, Maksimovich, Yashina, Gakhova and Gilichinsky2012) found 30,000-year-old fruit tissue from narrow-leafed campion (Silene stenophylla) in the Siberian permafrost and were able to successfully create fertile plants. Likewise, gene banks such as the Svalbard Global Seed Vault serve as repositories for regenerating species.Footnote ¹ Thus, this definition does not capture what extinction is.

A different definition comes from paleobiologist Michael Hannah. He writes, “The extinction of a species occurs when the last individual belonging to that species dies” (Hannah, Reference Hannah2021, p. 36). His definition does not imply that the narrow-leafed campion went extinct when there were no living members. This is because the last member had not died. So, the simplest response to the argument that de-extinction is impossible by definition is that in cases where the resurrection of a species occurs, we were simply wrong that the species had gone extinct. We thought it was extinct, but it was not. This does occur occasionally in other contexts. For example, the black-browed babbler (Malacocincla perspicillata) had not been observed in 170 years and was presumed extinct. However, in 2020, a team of researchers found a specimen in Borneo (Akbar et al., Reference Akbar, Nugroho, Suranto and Fauzan2020).

One might object that this response trivializes the notion of extinction (and thus the impossibility of de-extinction). For any given species, they might be resurrected and so we are never able to declare a species extinct. For example, Ben Novak suggests that we consider “reproductively competent single cells” as members of an “evolutionary torpid species” (Novak, Reference Novak2018, p. 9). However, even in the best of circumstances, information from DNA cannot be extracted after approximately 1.5 million years (Allentoft et al., Reference Allentoft, Collins, Harker, Haile, Oskam, Hale and Bunce2012). Thus, for any species which has disappeared for that length of time or more, we simply cannot resurrect them. In some cases, we can confidently declare a species extinct. The concept of extinction is not trivialized by de-extinction.

Let us characterize what de-extinction is. De-extinction is the process of creating living organisms that are similar (often genetically similar) to members of extinct species (Sandler, Reference Sandler2017, p. 1).Footnote ² Some authors augment this definition with notions of ecological function. For example, Ben Novak writes,

These definitions of de-extinction are inclusive. The reason they are inclusive is that the techniques associated with de-extinction do not require that a species be resurrected but also may be recreated.Footnote ³ To see why, let us consider these three techniques (Shapiro, Reference Shapiro2015, Reference Shapiro2017). There is back-breeding, cloning through somatic cell nuclear transfer (SCNT), and genetic engineering. This also allows us to explore examples of current or future de-extinction.

Back-breeding uses selective breeding to bring back ancestral traits in extant populations of organisms.Footnote ⁴ This technique has been used in bringing back traits in species similar to the auroch (Bos primigenius) which is the ancestor of modern cattle. The last known auroch died in Poland in 1627. Fascinated with the animal, German brothers Heinz and Lutz Heck selected for cattle that had large horns, large body size, and were more aggressive than most cattle. In 1983, a group of 32 “Heck cattle” were released into a nature preserve in the Netherlands, the Oostvaardersplassen.Footnote ⁵ Notice that back-breeding recreates the auroch by virtue of reestablishing auroch-like traits though the auroch is not resurrected. This is because in recreation we create organisms very similar to those of an extinct species even when they are not members of the same species.

Cloning uses SCNT to create a genetic copy of an organism. The nucleus from an adult somatic cell is placed into an enucleated egg cell. The host egg cell becomes an undifferentiated pluripotent stem cell and has an identical nuclear genome sequence to the donor of the somatic cell.Footnote ⁶ For example, in 2003, a cloned bucardo (Capra pyrenaica pyrenaica) calf was born. The bucardo is an extinct subspecies of the Pyrenean ibex. This individual died soon after they were born due to a lung deformity. One of the most discussed possibilities for de-extinction is the woolly mammoth (Mammuthus primigenius) which went extinct in Eurasia and North America about 8,000–10,000 years ago. The last population of dwarf woolly mammoths disappeared on Siberia’s Wrangel Island 4,000 years ago (Vartanyan et al., Reference Vartanyan, Arslanov, Tertychnaya and Chernov1995; Stuart et al., Reference Stuart, Kosintsev, Higham and Lister2004). Fortunately, there are many mammoth bones in the subarctic from which cells (or at least chromosomes) could be extracted. Woolly mammoths are more closely related to Asian elephants (Elephas maximus) than African ones (Loxodonta). In principle, they could be cloned with Asian elephant mothers. One serious ethical problem here is that the Asian elephant is endangered, and egg harvesting, implantation along with pregnancy would likely harm them. Of course, some surrogate species are not endangered. We will consider issues of animal welfare later.

For approximately 100,000 years, there was a large area of productive grasslands on which horses, bison, woolly rhinoceroses and woolly mammoths lived. This “mammoth steppe” disappeared 10,000 years ago (Zimov, Reference Zimov2005). Mammoths, and other megafauna, disappeared due to a combination of climate change and human predation. This in turn leads to the disappearance of these grasslands. These grasslands created carbon-rich materials such as dead plant roots frozen in the soil and permafrost. However, anthropogenic climate change is melting the permafrost and this will rot the plant material releasing this carbon (or methane) into the atmosphere. Sergey A. Zimov, founder of Pleistocene Park, proposes resurrecting woolly mammoths to slow down permafrost melting. Snow-covered forests have a lower albedo than do snow-covered grasslands. The recreation of snow-covered grasslands would prevent a greater amount of permafrost melt than currently existing snow-covered forests. Grazing animals would trample the snow which brings colder air to chill the permafrost (this air is often colder than the permafrost itself). Zimov and his colleagues have established herds of herbivores including horses, moose, reindeer, muskox and yak. Unlike woolly mammoths, however, these animals cannot knock down trees to create grassland. Thus, Zimov and his colleagues have been using a tank to bring down the trees. One remarkable proposal then is to resurrect or recreate woolly mammoths to combat anthropogenic climate change.Footnote ⁷

Genetic engineering uses ancient DNA and genome editing to resurrect or recreate a species. To do this, scientists must first reconstruct the genome of the extinct species. As noted, DNA can often survive much longer than cells after an organism dies. This is especially so with organisms that die in cold environments.Footnote ⁸ Full genomes have been reconstructed for mammoths, aurochs and passenger pigeons (Ectopistes migratorius).Footnote ⁹ The next step in resurrecting a species using genetic engineering is to determine what parts of the genome are responsible for which phenotypes. Next one takes the cell with the edited genome and creates a living organism. This often combines genetic engineering with SCNT in groups like mammals. However, genetic engineering currently cannot be used with SCNT in egg-laying species like birds or reptiles. One target for de-extinction using genetic engineering is the passenger pigeon. It is an awe-inspiring species, which existed in the billions. Aldo Leopold called it a “biological storm” (Leopold, Reference Leopold1949, p. 104). One recorded flock was 300 miles long and a mile wide. Martha, the last passenger pigeon, died in 1914 at the Cincinnati Zoo. The cause of the passenger pigeon’s disappearance was predation by humans. For example, in 1878 hunters killed over 50,000 birds per day for 5 months straight.Footnote ¹⁰

Currently, Ben Novak along with George Church working with others like Stewart Brand is planning to bring back the passenger pigeon.Footnote ¹¹ The passenger pigeon’s genome has already been sequenced and its closest living relative is the band-tailed pigeon (Patagioenas fasciata). They have already identified which parts of the band-tailed pigeons’ genome that must be edited and replaced via CRISPR with the passenger pigeon’s genes. Thereafter they would grow band-tailed pigeon cells and edit the DNA within the nucleus with synthetic passenger pigeon DNA. As noted above, here complications arise since birds cannot be cloned. A fertilized ovum divides with its yolk moving down the oviduct, which is wrapped in albumen. When it reaches the isthmus, membranes are deposited around it and the eggshell is added. Currently, it is simply too difficult to enucleate and replace the nucleus of the egg without disrupting the process. Instead, primordial germ cell transplantation (PGCT) is used to create a chimera between the passenger and band-tailed pigeons. When PGCT is used where the donor and recipient are of distinct species, we have interspecies PGCT or iPGCT. From here, adult chimeras are mated together and scientists will work to establish a population of these birds.

A caveat of some importance is that these techniques for recreating or resurrecting species do not create an organism that is a copy of an extinct one. With back-breeding, at best an organism with the same phenotype of an extinct one is created. With genetic engineering, at best the same phenotype is generated from some of the same genes as the extinct organism, but many other parts of the genome are different (Sherkow and Greely, Reference Sherkow and Greely2013). Even organisms that are cloned from extinct ones using SCNT are not copies since the mitochondria of the two will be different. Finally, there are different gene–environment interactions that occur with respect to the resurrected organisms and the extinct ones that can change their respective developments. However, many proponents of de-extinction suggest that creating exact copies of extinct species is not the primary aim. Beth Shapiro writes,

Whether this is the aim of de-extinction research or not, can we nevertheless resurrect species?

Is de-extinction possible?

Philosophers have argued about whether de-extinction is even possible. This issue is not whether it is difficult or even unlikely that a species will be resurrected or recreated. The concern is that it is physically or conceptually impossible.Footnote ¹² On some views of what species are, they contend it is not possible (and possible on others) (Siipi and Finkelman, Reference Siipi and Finkelman2017; Finkelman, Reference Finkelman2018). Let us consider one of the strongest arguments for this claim.

One popular view amongst philosophers of biology is that species are “historical entities” or, as it is sometimes put, they are individuals (Hull, Reference Hull1978; Ereshefsky, Reference Ereshefsky2000).Footnote ¹³ Consider the element gold. Something is gold if, and only if, it has atomic number 79. According to many philosophers and biologists, species are not like periodic elements. First, they are temporally bounded in that they have a beginning and an end in time. Second, they are spatially localized having a geographical distribution. Third, the organisms of species interact causally at a time (cohesion) and through time as well (integration) via various intraspecies processes. Gold is not like this since something is gold simply in virtue of having atomic number 79. It can occur in principle anywhere and elements of gold need not have anything to do with one another.

Here then is an argument against the possibility of de-extinction (for a discussion, see Campbell and Whittle, Reference Campbell and Whittle2017; Slater and Clatterbuck, Reference Slater and Clatterbuck2018). A species can evolve only if traits of organisms in the species are heritable. But a trait is heritable in a species only if organisms between generations are causally connected. Any entity whose parts are causally connected is a historical entity. Therefore, species are historical entities. De-extinction requires that organisms of a species exist at two separate times but which are not causally connected between those two times. But this is impossible. Hence, de-extinction is impossible.

To make this more concrete, let us consider the last generation of woolly mammoths on Wrangel Island and some that are created through cloning or genetic engineering (assuming it is successful). One might think that the former does not give birth to the latter since the latter was produced in a laboratory. So, they are not causally connected by heredity and thus the two populations do not share a history. They are not parts of the same historical entity, Mammuthus primigenius. But this argument fails (Campbell and Whittle, Reference Campbell and Whittle2017; Slater and Clatterbuck, Reference Slater and Clatterbuck2018; Campbell, Reference Campbell2022). To see why, consider a human baby who is born through in vitro fertilization. There is a clear sense in which ordinary reproduction has not occurred; it is a form of human-facilitated reproduction. Nevertheless, we do not think that the baby is not a member or part of Homo sapiens. They are nevertheless human. The same should be said regarding de-extinction. There is a causal connection between populations through human-facilitated reproduction as well.Footnote ¹⁴

One might object that the woolly mammoth created through cloning or genetic engineering is a “mammophant” and not a genuine mammoth since it has Asian elephant DNA. It is a hybrid or a chimera. But hybridization and horizontal gene transfer occur throughout the natural world (Piotrowska, Reference Piotrowska2018). For example, all the brown bear (Ursus arctos) populations on the Admirality, Baranof and Chichagof islands possess polar bear mitochondria with less than 1% of their nuclear DNA from polar bear ancestry (Cahill et al., Reference Cahill, Green, Fulton, Stiller, Jay, Ovsyanikov and Shapiro2013). They are still brown bears. We know that humans for example interbred with Neanderthals and Denisovans, yet we are still members or parts of H. sapiens (Rogers et al., Reference Rogers, Harris and Achenbach2020). Additionally, mammophants would belong to a relevant woolly mammoth species according to some species concepts (Slater and Clatterbuck, Reference Slater and Clatterbuck2018, pp. 8–11).

Another objection comes from Katz (Reference Katz2022). He claims that with biological systems, there is no design. Humans, and only humans, design objects. However, designing something even in part changes it into an artifact. By recreating and resurrecting species, we are creating artifacts that lack the “integrity” that natural objects possess. Additionally, this sort of design contributes to the domination of the natural world by humans which is morally wrong. There are several problems with this argument (for responses to Katz, see Browning and Veit, Reference Browning and Veit2022; Lean, Reference Lean2022; Preston, Reference Preston2022; Reydon, Reference Reydon2022; Sandler et al., Reference Sandler, Stabell, Baylon, Lundgren, Weisbeek, Yelle and Zaba2022; Turner, Reference Turner2022). First, as many philosophers of biology and biologists argue, there can be design without designers (Kitcher, Reference Kitcher1993). Natural selection is one such designing process.Footnote ¹⁵ Second, the argument rests on a flawed dualism between humans and the rest of nature. Humans are an extraordinarily unique and complicated mammal, but we are an evolved species amongst many others. As a species, we have radically changed our planet. Given most of us are moral agents, we can be morally responsible for those changes.Footnote ¹⁶ This concerns our status as persons and not as humans. Finally, even if we are dominating the planet and it is morally wrong, it is not clear how much a few instances of de-extinction will contribute to this. Many promote de-extinction as a form of restoration ecology; we are restoring ecological functions to ecosystems.Footnote ¹⁷

A different sort of worry is that resurrected species are in some sense “inauthentic” (Siipi, Reference Siipi, Oksanen and Siipi2014). When we say something is inauthentic, we are saying either that it is not identical to the object of interest or it does not share many of the properties that the target of interest possesses (Siipi, Reference Siipi, Oksanen and Siipi2014, pp. 77–78).Footnote ¹⁸ For example, there is no analog community for the woolly mammoths since many of the species that it would have existed with no longer exist. But the same can be said of translocation or assisted migration (Novak, Reference Novak2018, p. 5). For example, wolves (Canis lupus) were reintroduced (or relocated) to Yellowstone National Park for the purposes of restoration and the recreation of trophic cascades (Fritts et al., Reference Fritts, Bangs, Fontaine, Johnson, Phillips, Koch and Gunson1997). Though the pack reintroduced did not descend from the last wolves eradicated in Yellowstone around 1926, they are of the same species. Likewise, as climate change alters the environments of many species, we may relocate them to places where their peer species are not present (Schwartz et al., Reference Schwartz, Hellmann, McLachlan, Sax, Borevitz, Brennan and Zellmer2012). This does not mean that they are a different or “inauthentic” species taxa any more than species aided with assistant migration.

In the end, there are a variety of technical issues associated with resurrecting or recreating species using de-extinction technologies. Nevertheless, it does appear possible to recreate or resurrect some lost species.

Is de-extinction morally permissible?

We now consider the most important arguments for and against de-extinction (for surveys, see Sandler, Reference Sandler2014; Rohwer and Marris, Reference Rohwer and Marris2018).

One common argument for de-extinction comes from considerations of restorative justice (Cohen, Reference Cohen2014; Jebari, Reference Jebari2016). When a moral agent harms a moral subject, the former owes the latter restitution. Humans harmed woolly mammoths for example by driving them extinct in combination with others factors like climate change (Martin, Reference Martin2005). Thus, humans owe them restitution or compensation for that harm. The most apt form of restitution would be to resurrect the species. Therefore, we should resurrect woolly mammoths if it possible. As we saw in the last section, though there may be technical impediments to resurrecting the woolly mammoth, it is possible.

One immediate objection to the argument is the individual organisms harmed are long dead and they cannot be resurrected. This is true but beside the point since the harm of extinction befalls a species and not individual organisms.Footnote ¹⁹ A different objection to the argument from restoration is that we do not owe extinct species restitution since we did not drive them extinct. We were not alive then. We thus do not owe them anything (Campbell and Whittle, Reference Campbell and Whittle2017).Footnote ²⁰ This argument has force when we consider woolly mammoths, but it might have less force when we consider the passenger pigeon. After all, Americans killed passenger pigeons for their own material benefit. Thus, Americans who benefited from their extinction (even if they did not drive them extinct) might owe them restitution through de-extinction.Footnote ²¹

Here is another argument for de-extinction from considerations of the conservation of biodiversity (Campbell and Whittle, Reference Campbell and Whittle2017; Iacona et al., Reference Iacona, Maloney, Chadès, Bennett, Seddon and Possingham2017). We should conserve and restore biodiversity. De-extinction is a means to conserve and restore species (and thus biodiversity). Thus, we should resurrect and restore species. Conservationists recognize the importance of preventing species from becoming threatened, endangered and eventually extinct. De-extinction raises the possibility of taking an extinct species and changing it conservation status (Campbell and Whittle, Reference Campbell and Whittle2017, p. 91). Of course, our obligation to conserve and restore biodiversity is a prima facie obligation; it can be outweighed by other morally factors.Footnote ²² For example, if resurrecting a species creates a harmful invasive, then we may have good reason not to do it. As a special case of restoration (or rewilding), consider the restoration of the mammoth steppe (Josh Donlan et al., Reference Josh Donlan, Berger, Bock, Bock, Burney, Estes, Foreman, Martin, Roemer, Smith, Soulé and Greene2006; McCauley et al., Reference McCauley, Hardesty‐Moore, Halpern and Young2017). By resurrecting or recreating the woolly mammoth, we restore ecological functions performed by megaherbivores in the recreation of grasslands that will help prevent climate change. Of course, this assumes that those functions can be performed in a changed environment, and there is not another appropriate species that could do it short of de-extinction.Footnote ²³

Let us turn to some of the most important objections to de-extinction.

The first argument against de-extinction concerns animal welfare (Kasperbauer, Reference Kasperbauer2017; Browning, Reference Browning2018; Browning and Veit, Reference Browning and Veit2022). It is morally wrong to cause a sentient being unnecessary suffering. De-extinction will cause unnecessary suffering. Therefore, it is morally wrong to recreate and resurrect species. As a case in point, consider the bucardo cloned and born in 2003. Due to its deformed lungs, it lived for a brief time in tremendous pain (Cohen, Reference Cohen2014). De-extinction could lead to “miscarriage, stillbirth, early death, genetic abnormality and chronic disease” as the result of cloning (Browning, Reference Browning2018, p. 789). And these are just the beginnings of the ethical issues since there are others regarding rearing and reintroducing these animals. However, conservation always involves a balance between protecting and restoring species and the welfare of individual organisms. For example, in places such as New Zealand endemic biodiversity is protected from predatory mammals through multi-kill traps and aerial poisons (Butler et al., Reference Butler, Lindsay and Hunt2014). Our obligations to conserve and restore biodiversity can be outweighed by conservations of animal welfare, but the extent to which there can be “compassionate conservation” is an active debate in conservation practice.

Another argument against de-extinction is the argument from hubris (or “playing God”) (Minteer, Reference Minteer2014, Reference Minteer2019; Diehm, Reference Diehm2017). It is morally wrong to be hubristic (i.e., overly self-confident). De-extinction is hubristic insofar we overestimate our ability to predict and control resurrect or recreated species or ecosystems more generally. As Ben Minteer writes,

There are several ways to respond to this argument. First, recreating or resurrecting species can be hubristic, but it is not necessarily so. For example, proponents of de-extinction are very much aware of the limitations of all the various techniques involved. As we discussed, we could only resurrect only a tiny fraction of all the species that have gone extinct. Second, this objection serves as an important moral call to continually avoid overestimating our abilities or underestimating the uncertainties or risks. Third, if sound, this argument would suggest we should avoid all manner of technologies not the least of which is de-extinction. After all, cloning and gene editing alone are only the most dazzling of our current technologies but conservationists also use camera traps, tracking tags, remote sensing, acoustic sensors, drones, eDNA and artificial intelligence. Should we avoid those as well because they may involve hubris?

A final ethical argument against de-extinction is that it is a poor allocation of conservation resources (Bennett et al., Reference Bennett, Maloney, Steeves, Brazill-Boast, Possingham and Seddon2017). Conservation is an underfunded practice. Resources should be allocated to that which will do the most good in conserving and restoring biodiversity. However, de-extinction will do little for that aim. Thus, funds should not be allocated to it. There are several objections that might be raised to this argument. First, there is a crucial assumption here: If money (or other resources) had not gone to a de-extinction project, then they would have gone to conservation work. But this may not be true since some de-extinction projects will be funded by those would not otherwise give to conservation. Efforts to bring back the heath hen in Martha’s Vineyard appears to be such a case (Sandler, Reference Sandler2017, p. 2). Second, there is an assumption that ordinary conservation and de-extinction are mutually exclusive and that need not be the case. For example, there are good reasons to take tissue samples with DNA for “frozen zoos” or seed banks independent of whether they are used for de-extinction (Ryder et al., Reference Ryder, McLaren, Brenner, Zhang and Benirschke2000). After all, if a species disappears “in the wild,” conserving tissue samples may allow us to prevent them from going extinct.

We can now take stock of the moral arguments for and against de-extinction. The general conclusion we have found is that it is morally permissible to resurrect or recreate species except when there are outweighing morally relevant considerations. These moral considerations concern animal welfare, hubris and poor allocation of resources will sometimes outweigh obligations to recreate or resurrect species but not always. Thus, moral decisions regarding de-extinction will often need to be decided on a case-by-case basis.