Introduction

An existential catastrophe is an outcome where the potential of Earth-originating intelligent life is drastically and irreversibly curtailed. Such an outcome could be the result of human extinction, a permanent collapse in human civilization, or a permanent totalitarian world government (Baum et al., 2019; Bostrom & Ćirković, 2008). While many scholars across various disciplines agree that an existential catastrophe would be a negative outcome (Bostrom, 2003; Häggstrom, 2016; Ord, 2020; Scheffler, 2018), their supporting arguments are anthropocentric or focused on human values. For example, Scheffler (2018) argues that human extinction or a permanent collapse of civilization would deprive us of any future that makes our projects worthwhile. Bostrom (2003) argues that an existential catastrophe would drastically reduce the potential amount of human happiness and individual flourishing in the universe. According to perfectionist theories, art, science, and the human experience would also be lost in some existential catastrophes, the value of which cannot be reduced to happiness (Scanlon, 1998). While hedonism and perfectionism are not necessarily anthropocentric, they are often expressed in such terms, especially hedonic theories that distinguish between ‘higher and lower pleasures’ (Mill, 2002). This is also true for those views that welcome human extinction. For example, Benatar (2008) argues that non-existence is better than existence, and that we should not reproduce, even when this would in the long run imply human extinction. From this viewpoint, human non-existence is desirable, because to avoid human suffering matters more than anything else, including non-anthropocentric values.

Yet, not all normative theories are anthropocentric. Authors in the field of environmental ethics argue instead for the need to protect and preserve ecosystems such as coral reefs and mangrove forests, as well as the biosphere as a whole (Brennan & Lo, 2020; Callicott, 1999; Naess, 1973). This view is known as ecocentrism. Human civilization (both modern and pre-modern) poses a considerable threat to various ecosystems (Barnosky et al., 2011; IPBES, 2019), which motivates some proponents of ecocentrism to advocate anti-modernism and primitivism, or the view that human civilization should be materially simplified (Harrison & Tanner, 2011; MacCormack, 2020). By contrast, this article will argue that (1) the biosphere will be destroyed by the increased luminosity of the sun in about a billion years, and (2) that human civilization is currently necessary for extending the lifespan of the biosphere beyond this point. Consequently, there are strong instrumental reasons for ecocentrists to prevent the permanent collapse of modern industrial human civilization. Whereas there are many anthropocentric reasons to prevent such a collapse, we present here an argument that does not assume anthropocentrism.

The article will proceed as follows. First, we describe ecocentrism and its emphasis on holistic systems. We then describe some long-term astronomical threats to the biosphere before describing some potential technological responses. In the final section, we discuss objections, arguing that since human civilization needs to be preserved to prevent the premature (i.e., in one billion years) destruction of the biosphere, this provides ecocentrists with strong reasons to prevent a collapse of civilization.

Ecocentrism

Ecocentrism is an important normative theory in the field of environmental ethics. Fundamentally, ecocentrism argues that ecosystems matter, not only for the benefits they provide humans and non-human animals with, but for their own sake. In other words, ecocentrism rejects anthropocentrism, or the notion that all value is related to human interests. Ecocentrism is distinct from biocentrism (the view that all organisms, and not only humans, matter) by including ecosystems as holistic entities, including their non-living components (Washington et al., 2017). This is an important aspect of ecocentrism. While ecocentric philosophers believe that individual humans and animals ought to be respected as parts of their respective ecosystems, when the interests of ecosystems are in conflict with the interests of human and non-human animals, some ecocentrists believe that it could be more important to protect the ecosystems. Washington et al. agree with Stan Rowe that:

“Reflecting on the ecological status of all organisms, it [ecocentrism] comprehends the Ecosphere as a Being that transcends in importance any one single species, even the self-named sapient one” (Rowe 1994, quoted in Washington et al., 2017).

One such example has been discussed by Oscar Horta, concerning ongoing efforts at introducing predator species into ecosystems to create an “ecology of fear” (Horta, 2011). Here, it is clear that those advocating the introduction of predators see the interests of plant-eating animals as less important than the flourishing of the ecosystem as a whole.

An ecosystem, in this context, is a community of organisms that interact with each other and the non-living environment as a system. For example, a forest is an ecosystem where many organisms are interconnected and interact with each other and the water, soil, and air in various complex ways. According to some ecocentrists, the value of the forest generates a prima facie moral duty on the part of humans to preserve it or promote its goods, be it resilience or biodiversity, as ends in themselves (Callicott, 1999; Curry, 2011; Naess, 1973; Rolston, 1985, 1992). In turn, the biosphere is the sum of all ecosystems on Earth. While no natural ecosystems are closed off, the biosphere represents a closed system, at least concerning matter. The biosphere can thus be understood as a distinct object composed of all the objects that, according to ecocentrism, are important.

According to mainstream deontology, duties are directed against right-holders, such as persons or individuals with interests. This view has been challenged in environmental ethics in general and in ecocentric ethics in particular, as ecosystems and other objects of value are often not sentient beings, as Sandler (2010) argues. Philosophers concerned about duties in environmental ethics, often conceptualize these duties as general principles of proper conduct. The duty to preserve ecosystems can in this view be seen as respect for those entities, rather than as a rights-based duty. This is one aspect where environmental ethics differs from other forms of normative theory (Taylor, 1981).

Other philosophers argue that non-sentient entities such as species and ecosystems can be seen as moral patients, even when they lack what we typically think of as interests. For example, Holmes Rolston III (1985) argues that holistic entities such as species and ecosystems, have interests that are distinct from but analogous to those of humans and animals, and that killing a species is the ‘super killing’ of a whole pattern of life. This line of reasoning could be extended to the entire tree of life. If pruning a branch on the tree (i.e., killing a species) of life is a “super killing”, then killing the entire tree would arguably constitute a “total killing”. This notion has certain similarities with the idea of ultimate harm, or the destruction of all value in a domain (Persson & Savulescu, 2008).

While some authors like Hiller (2013) endorse a form of ecocentric consequentialism, this view is unusual among ecocentrists (Hiller, 2017). However, this view should not be conflated with the view that we ought to maximize the objects that are value-bearers. This view is even more unusual among ecocentrists, although it has been defended by Tonn (1999), and more recently by Owe (2022) who have argued that we have an obligation to seed the universe with life. Maher and Baum (2013) also argue that ecocentric value could be vastly increased by interstellar expansion, which suggests an aggregative view of ecocentric value (i.e., the view that “more is more”). However, most ecocentric philosophers argue that we have a duty to preserve and protect (rather than promote) ecosystems, especially those that make up our planet’s biosphere. Similarly, Scanlon (1998) argues that some values ought to be protected and not maximized.

The implication of rejecting maximizing consequentialist ecocentrism is that while humanity may, in the distant future, choose to radically increase the number of ecosystems in the universe, this is not necessarily something that would count in its favor. From a mainstream (i.e., non-maximizing) ecocentric perspective, what matters is protecting the ecosystems that exist, not maximizing their numbers. In other words, the creation of life on other planets cannot compensate for the destruction of life on this one. This is especially true on the view defended by Holmes Rolston III (2012) according to which the value of ecosystems in parts depends on their particular evolutionary history, a history that cannot be continued unless this particular biosphere is preserved.Footnote 1

Human civilization and its relentless economic activity represent a considerable near-future threat to various ecosystems, according to the 2019 Global Assessment Report on Biodiversity and Ecosystem Services (Barnosky et al., 2011; IPBES, 2019). From an ecocentric perspective, then, it can be argued that modern civilization and industry are major sources of value destruction (Best, 2014). Many environmental ethicists and ecocentrists are also value pluralists and therefore recognize at least some anthropocentric values, while others simply deny that we can separate “nature” from the “human”, as the view that “nature must be protected from humans” implies (Morton, 2007). However, modern ecocentrists have been very critical of modern industrial civilization, and its immediate harms on the biosphere, while long term considerations have rarely been discussed. The implication of such views has sometimes been that modern civilization is considered harmful or undesirable (Harrison & Tanner, 2011; MacCormack, 2020).

We argue, in line with ‘distant environmental ethics’ (Owe, 2022; Tonn, 2002), that from a long-term perspective, human civilization and its technological tools currently represent the best hope for the long-term survival of life on Earth, and perhaps in this galaxy. Specifically, the biosphere is likely to crash in about a billion years due to the increased luminosity of the Sun. From an ecocentric perspective, this is a significant concern since it implies ultimate harm (Persson & Savulescu, 2008), that is, the loss of all value structures and the irreversible loss of all potential to produce further value. Consequently, we argue that there are strong ecocentric reasons to preserve modern civilization.

Long-Term Threats to the Biosphere

Over the long term, the Earth and the Sun are irreversibly changing. The Earth’s core will cool down, leading to a weakened continental cycle, and the Sun’s luminosity will gradually increase, eventually transforming the Sun into a red giant. These trends will impact the biosphere and eventually cause its destruction. This is different from other past mass extinction events, as these were consistent with the possibility of life in the future. Here, the long-term changes under consideration will permanently end the Earth’s suitability for life.

Breakdowns of Biogeosphere Homeostasis

The carbonate–silicate cycle is an important aspect of the Earth’s thermal regulation. At present, carbon dioxide (CO2) provides a greenhouse effect that keeps Earth warm enough for water to be liquid at the surface, maintains temperature homeostasis, and makes photosynthesis possible. In the long term, this homeostasis will break down and likely produce a state in which normal photosynthesis is impossible, ending the biosphere as we know it.

As part of this cycle, carbon dioxide is removed from the atmosphere by the weathering of silicate rocks in warm and wetland environments. Dissolved carbonates and silicates eventually precipitate on the sea floor, forming sedimentary rocks. These sediments are subducted by continental drift; at the higher subterranean temperature, they recombine to produce the original silicates and CO2. The latter is released through volcanism, maintaining a higher concentration in the atmosphere than the weathering chemical equilibrium would entail (Pierrehumbert, 2010). However, this cycle is not necessarily a closed loop, since the amount of CO2 that flows back into the atmosphere depends on the rate of subduction and volcanism. Were these to slow down or cease, CO2 levels would fall as weathering would irreversibly remove the gas.This cycle also serves to regulate surface temperature, removing more CO2 at high temperatures and allowing it to build up during cold periods (Walker et al., 1981). While inorganic weathering is a slow process that currently accounts for far less carbon flow (about 0.2 gigatonnes per year) than the biosphere and anthropic processes (about 120 and 7 gigatonnes per year, respectively), it is linked to vast reservoirs that are at least 20,000 times larger than the combined atmospheric and biospheric carbon (Martin, 2017). This allows the carbonate–silicate cycle to regulate not only the temperature but, to some extent, the mass of the biosphere since it is in part constrained by the amount of CO2 available for photosynthesis.

The loss of radioisotopes and accretion heat in the Earth’s core will eventually end the recycling of CO2 through tectonic activity. In the absence of human intervention, this will lead to a significant loss of greenhouse effect and plant viability, presumably leading to a cold and barren environment in a few billion years.

Conversely, the Sun’s luminosity is also slowly increasing; the effects of this process will likely counteract the cooling effect of reduced tectonic activity, but also speed up the weathering effect. As hydrogen atoms are fused, the proportion of helium atoms in the Sun’s core gradually increases and the core itself becomes denser and hotter, moving hydrogen fusion outwards. At present, this process is increasing the Sun’s luminosity by around 1 per cent every 100 million years and will have a significant warming impact on the biosphere, even in the absence of CO2, which is not the only greenhouse gas. In particular, water vapor is a major greenhouse gas and hence a sufficiently hot and humid environment would tend to become hotter, since warm air accelerates evaporation. In certain scenarios, this produces drastically hotter surface temperatures, also known as a ‘runaway greenhouse’ (Goldblatt & Watson, 2012). In this runaway greenhouse scenario, a warmer and hence wetter atmosphere heats the planet up to temperatures that seem inconsistent with complex life, and ultimately up to a point at which the atmosphere becomes steam-dominated and the seas disappear. Moreover, as moisture and heat push water vapor into the upper atmosphere, UV radiation dissociates oxygen and hydrogen from the water molecules. The resulting free hydrogen molecules—which are very fast and thus can easily escape the Earth’s gravity—will be stripped from the atmosphere, eventually drying out the surface completely; this is somewhat confusingly known as the ‘moist greenhouse’ scenario (Kasting, 1988). If the moist greenhouse has not sterilized the planet by this point, the absence of water will make the biosphere even more difficult to sustain. Eventually, sufficient heat may bake CO2 out of carbonate rocks, further boosting the greenhouse effect and making Earth Venus-like. At this point, no lifeform, no matter how simple, seems remotely possible.

The process by which increased solar luminosity will lead to a moist or runaway greenhouse (and ultimately the destruction of the biosphere) interacts with the carbonate–silicate cycle in two primary ways. First, warmer temperatures will interrupt the homeostasis of the cycle, increase weathering, and speed up the process of CO2 removal. Second, increased solar luminosity will compensate for the reduced greenhouse effect from the lost CO2 for a limited time; eventually, despite the lack of CO2, the planet will continue to heat up. While the warming induced by increased solar luminosity will for some time be offset by increased weathering of CO2, this will also reduce the potential for photosynthesis and eventually lead to a runaway greenhouse state with surface temperatures exceeding 100 °C. Even if some simple life forms may survive this, the Earth’s biosphere will ultimately be sterilized without any surface water.

Aside from these inevitable changes, other biosphere-ending disasters may also occur. Over the next billion years, we should expect a certain amount of large asteroid or comet impacts and supernovas that are severe enough to cause mass extinctions (Beech, 2011). Sufficiently large impacts to wipe out the entire biosphere are less likely, but not impossible (Sloan et al., 2017). Likewise, there is a small risk (less than 1 per cent) of orbit disruption through close stellar encounters (Malmberg et al., 2011) or orbital instability in the inner solar system (Laskar & Gastineau, 2009). We will not deal with these risks here, other than noting that interventions to prevent some biosphere-ending disasters are possible and raise similar ethical issues.Footnote 2

Estimates of Remaining Biosphere Lifespan

One of the first scientific estimatesFootnote 3 of future biosphere lifespan was made by Sagan and Mullen (1972), who predicted a runaway greenhouse state in 3–4.5 Gyr (billions of years) due to increasing water vapor as the planet heats up due to rising solar input. Since then, increasingly sophisticated models of interactions between the biosphere, geosphere, atmosphere, and solar input have been used to estimate the limits of the future biosphere. Lovelock and Whitfield (1982) estimated that CO2 levels will fall below the critical level for C3 photosynthesis (the most common type) in only 100 Myr (millions of years). Caldeira and Kasting (1992) re-examined the problem with a more elaborate model (and focusing on C4 photosynthesis), positing a lifespan of 0.9–1.5 Gyr. Franck et al. (1999) extended this model, finding that photosynthesis breaks down in 0.5–0.8 Gyr and surface temperatures reach 100 °C in 1–1.5 Gyr. Other models and assumptions have produced estimates of the remaining biosphere lifespan of 0.48–0.63 Gyr (Franck et al., 2000), 1.2 Gyr (Franck et al., 2002; Lenton & von Bloh, 2001; von Bloh et al., 2003), 0.84 Gyr (Rushby, 2015) and 1.08 Gyr (Ozaki & Reinhard, 2021) (Fig. 1 and Table 1).

Fig. 1
figure 1

Published time estimates for the end of Earth’s biosphere

Full size image
Table 1 Published time estimates for the end of Earth’s biosphere
Full size table

When these processes will ultimately destroy the last organism on Earth depends on assumptions about the heat tolerance of different kinds of life. Franck et al. (2006) found future extinction times of 0.8–1.2 Gyr for multicellular life, 1.3–1.5 Gyr for eukaryotes and 1.6 Gyr for prokaryotes using the same basic model. O’Malley-James et al. propose that while a moist greenhouse may set in 1 Gyr from now, small refugia at high latitudes and altitudes may sustain extremophiles up to 2.8 Gyr (O’Malley-James et al., 2012).Footnote 4 On the contrary, it may be objected that the biosphere has proved remarkably resilient over the most recent billions of years. It has persisted in the face of extreme and violent events such as abrupt climatic changes, massive volcanism and asteroid collisions (O’Malley-James et al., 2012). Moreover, the events described here will take place over a timespan that allows for the evolution of adaptive traits. Even if no existing species would survive a biosphere without CO2 or a moist greenhouse, such species could conceivably evolve. Consider the first (known) mass extinction event: the Great Oxygenation Event. Although very few species had evolved tolerance for oxygen (a highly reactive gas), life evolved and adapted nonetheless (Margulis & Sagan, 1997).While it is true that life is very resilient, especially in simple lifeforms, those ecosystems that can survive will be very simple and biodiversity will be very low. While some parts of the biosphere could conceivably survive for hundreds of millions of years after the runaway phenomena described above, no multicellular life is likely to ever emerge again. While some value-bearing systems would persist for a while, complex valuable ecosystems would not. Eventually, without intervention, even the most resilient organisms would disappear. Note that the outcome of these processes is not merely an ‘extinction event’ nor even a ‘mass extinction event’. Rather, we refer to the outcome as sterilization. On the one hand, extinction is where a species or a phylogenetic group dies out. On the other hand, sterilization is the permanent elimination of all lifeforms on a planet. Whereas extinction may be a natural part of the history of life, sterilization is the ending of that history. Whether sterilization is ‘natural’ or not depends on the exact definition of ‘natural’ in this context. Mass extinction events in a biosphere add to the richness of its history and enable new forms of life to emerge; consider, for example, the evolutionary radiations of mammals and birds after the Cretaceous-Triassic mass extinction. However, biospheres are causally isolated from each otherFootnote 5: one biosphere’s death does not benefit any other biosphere.

In summary, modern biosphere/geosphere models tend to suggest a minimum of 800 Myr of remaining lifespan and 1.2 Gyr as a plausible median. Crucially, all models imply that the biosphere’s future is shorter than its past. While future evolutionary adaptations will likely help life to persist where current life would perish, there are limits to what natural evolution can achieve.

Potential Responses

Technological interventions could postpone the biosphere’s demise in several ways. Here, we discuss their feasibility given the assumption that human civilization survives over the relevant periods. These are some suggestions found in the literature and should not be seen as an exhaustive list. Whether or not these suggestions are feasible in the future remains to be seen. The basic point is that something can be done at a moderate cost to a future civilization with adequate resources and technological tools.

Moving the Earth to a More Remote Orbit

While extreme amounts of momentum are required to change planetary orbits appreciably, the slowness of solar brightening also means the methods do not have to be particularly fast. Any orbit-change scheme will still by necessity involve managing energy on a scale comparable to the climate. McInnes (2002) suggests a gravitational tug approach to move the Earth, while Brin (2014) proposes using a solar-powered electrodynamic tether on the far side of the moon. One scenario consists in using streams of asteroids to transfer momentum between Jupiter and Earth using gravity assist maneuvers (Korycansky et al., 2001). While some of these solutions could be dangerous, there is ample time to develop technologies and study the problem over the coming millennia.

Tuning Atmospheric Chemistry

This would involve efforts either to reduce the greenhouse effect (adding up to 2.3 Gyr) (Li et al., 2009) or to counteract CO2 depletion by artificially releasing bound carbonates.Footnote 6 As the current climate change problem shows, human civilization is already technologically capable of climate-affecting atmospheric chemistry.

Solar Input Control

Sagan and Muller proposed aerosol geoengineering in 1972 but assumed it would be hard to achieve even with very advanced technology. From our current perspective, geoengineering may appear more feasible (Goldblatt & Watson, 2012). However, aerosols need to be replenished often and cannot realistically handle insolation beyond a certain critical level without becoming opaque, upsetting the light and temperature balance. Artificial rings around Earth have been suggested to reduce global warming (Bewick et al., 2013; Mautner, 1991; Pearson et al., 2006) with typical estimated insolation changes in the order of 1.7 per cent. While this may prove enough to deal with current anthropogenic climate change, it will be difficult to scale up to the degree needed to handle increasing solar luminosity.

Solar shades located between the Earth and the Sun have also been suggested as a remedy for current climate change (Early, 1989; Mautner, 1991; McInnes, 2010; Teller et al., 1996). Angel (2006) suggests swarms of thin film refractive discs, while other proposals involve an occulting disc or swarm (Hudson, 1991; McInnes, 2002, 2006). In all these cases, active control is needed for station-keeping, or the system will slowly drift out of alignment (a problem already solved for current satellites). These proposals typically suggest that the cost is within near-future reach for current human civilization (a few per cent of global GDP at most). Using Gough’s (1981) solar luminosity formula posits that, over the next Gyr, insolation will increase by 9.6 per cent. This is within the range of what these solar shades could address without too radical an engineering challenge. The 21 per cent and 35 per cent increases in luminosity over the following two billion years, respectively, may be more challenging. A workable solar shade would need to both slowly increase in radius (since the solar radius is increasing) and opacity (to maintain the same effective insolation). In principle, it would appear that this system could function up until the Sun starts to affect the Earth more directly in 7.4 Gyr.Footnote 7

The main drawback of this strategy is that the solar shade would need to be continually maintained. Self-replicating technology could autonomously replenish the swarm (for example, by mining asteroids and sending new components to Earth using solar sails when the shade is too thin). The total mass involved does not have to be enormous. An Earth-sized total shade would mass about 2 billion tonnes, or about 31 years of current annual aluminum production. The required growth rate is modest. For the next billion years, about one 100×100-m section would need to be added per year, weighing 15 kg (at the time of writing, the Planetary Society LightSail-2 craft that orbits the Earth has an area of 32 m2 and weighs 5 kg).

Discussion

Based on the above, it would seem that it is physically possible for an advanced civilization to reduce insolation or other factors in such a way that the biosphere could persist for at least an additional billion years. Although it is not at present technically feasible to do so, the aforementioned techniques for solar shades are straightforward extrapolations of current satellite station-keeping methods and solar sails. The most speculative part is automatic mining and manufacturing, but it does not appear implausible that such systems could be built over the coming centuries or millennia given current work on in-situ resource utilization in space. Even based on conservative assumptions, it seems feasible that a species with advanced technology could, through various engineering projects, extend the viability of Earth’s biosphere for at least one billion years. We have transitioned from pre-agricultural technology to the Anthropocene in a mere 10,000 years, and it is hard to imagine what capabilities would be available were human civilization to survive for another 10,000 years, or even 100,000 years. The same is true for our understanding of the biosphere and the factors that affect its survival. Granted, the methods outlined in this paper are likely to be superseded by something more workable, but this does not mean that protecting the biosphere should not be done in the here and now.

It should be noted that these interventions are not a restoration but a preservation of the biosphere. Building a solar shield does not create a biosphere; it is preventing the biosphere’s premature destruction. In other words, installing a solar shield is not equivalent to replanting a forest or recreating a marshland. Rather, it is the equivalent of building a levee to prevent saline water from the sea from destroying the ecosystem in a freshwater lake. It is important to distinguish between restoration and preservation here because some prominent authors have argued against the idea that natural ecosystems can and should be restored and that this restoration would reverse the environmental harm caused. For example, Elliot (1982) argues that the value of an object depends partly on its origins or how it came about. Similarly, Katz (1996) states that the attitude that constitutes the restorative ideal is such that restoration confers no natural value at all, only instrumental value. Katz and Elliot could therefore argue that preservation would also destroy the value of the protected ecosystem, but it does not follow from the claim that restored ecosystems lack value. Thus, even if we accepted the argument against restoration, it would not apply to the moral value of protecting the biosphere.

A similar but more recent discussion concerns whether humans should intervene to protect wildlife against disasters such as forest fires. Here, interventionists argue that the natural world and human civilization are now so intertwined that what happens in the natural world becomes our concern (McCumber & King, 2020). Other interventionists have argued that large-scale interventions are justified because it is our primary duty to prevent harm (Johannsen, 2020). These pro-intervention arguments apply not only to the current era (the Anthropocene) but also to the far future. It is a fact that humanity has irreversibly changed the history of life on this planet, and that we will continue to do so for as long as we exist. On the other hand, it is equally important to note that the survival of human civilization is currently necessary for the life extension of the biosphere to happen. While we may develop (at some point in the future) technology advanced enough to autonomously preserve the biosphere in our absence, such technology is far beyond our current grasp. Thus, to be of use to the biosphere, human civilization must survive at least until the necessary technology has been developed.

One possibility is that if humanity goes extinct, a future species with similar ability to use technology will evolve and extend the biosphere lifespan. While certainly possible, the time remaining is finite, and the appearance of species able to use technology appears to be rare and random.Footnote 8 Reliance on future species thus represents a major gamble with the biosphere. And we have no reason to believe that any hypothetical future species would be more inclined to avoid environmental harm or to extend the lifespan of the biosphere in the future.

To clarify, we are not suggesting that it is possible to prevent the eventual sterilization of the biosphere. Even if it would be possible to reduce the amount of radiation to offset the increase in luminosity over the next billion years, the Sun will eventually become a red giant and swallow the Earth. And even if the Earth could be moved to another orbit, the fusion in the Sun’s core will eventually cease, and the Sun will become a white dwarf. Sooner or later, then, the biosphere will be sterilized, and humanity (or, rather, post-humanity) will no longer be able to prolong the lifespan of the biosphere. What we are suggesting is that the natural sterilization of the biosphere could be delayed. If the finitude of the biosphere contributes to its value, we can rest assured that this source of value is not threatened by human activity.

In summary:

  1. (1)

    According to ecocentric ethics, the biosphere is what matters.

  2. (2)

    The continued existence of human industrial civilization is currently necessary to prevent the premature destruction of the biosphere.

So,

  1. (3)

    Assuming ecocentric ethics, we have strong reasons to ensure the continued existence of human industrial civilization.

Objections

Anthropogenic destruction of the biosphere Human industrial civilization may cause the premature destruction of the biosphere. Human industrial civilization has caused much harm to the biosphere, even with today’s limited technology. There is a risk that future technologies could destroy the biosphere long before the geophysical processes outlined in this article. While it is possible that human civilization could save the biosphere, it is also possible that it could destroy the biosphere. When we can’t assign probabilities to two possible outcomes that are symmetrical and opposite in value, we should suspend judgement about which outcome is preferable.

Response: The outcomes are not symmetrical. In the absence of technological intervention, the destruction of the biosphere is almost certain. By contrast, while the destruction of the biosphere with anthropogenic means is certainly possible in the long-term future, it would require much more effort and resources than extending the lifespan of the biosphere. The biosphere is not fragile and has withstood several massive chocks. Not even the most apocalyptic nuclear winter scenarios (see Toon et al., 2014) in the literature or the most pessimistic scenario suggested by the IPCC (RCP 8.5, see Newbold, 2018) would suffice to destroy the biosphere. Taking this into consideration, we believe that it is far more likely that future humans (or posthumans) decide to extend the lifespan of the biosphere than they are to destroy the biosphere prematurely. However, this conjecture cannot be proven. Under the assumption that human civilization is more likely to destroy the biosphere than to extend its lifespan, there may indeed be strong instrumental reasons to prefer the irreversible collapse of human civilization.

It is possible that a future non-human civilization may intervene to extend the lifespan of the biosphere in the event of human extinction. But the probability of such a civilization emerging and surviving for long enough is unknown. While it is possible that human civilization destroys or fails to prevent the premature destruction of the biosphere, it is presently the only object that can prevent premature destruction. Here is an analogy.

A man suffers from a slow growing but lethal cancer. Currently, chemotherapy is necessary for his survival. The chemotherapy may not cure the disease, and it may kill the man. Moreover, there is an unknown probability that some treatment in the future may be just as effective. Yet, chemotherapy is presently necessary for the man’s survival. While it may be rash to begin treatment immediately, before it is strictly necessary, it would also be rash to eliminate the option to use chemotherapy entirely. Making sure that human industrial civilization persists would be like keeping chemotherapy as an option.

Hubris Humanity has destroyed much value in its hubristic endeavors. Our interactions with the biosphere should avoid hubris, ambitions of ‘playing God’ and efforts to subject the biosphere to our control. To suggest that we ought to preserve the biosphere is an extreme form of the same hubristic attitude and as such is a dangerous endeavor.

Response: If we understand ecocentrism as a theory of how to value certain entities, such as ecosystems, rather than as a more general theory of human virtues, then it is not contrary to ecocentrism to have ambitions that could be described as hubristic. Whether hubris is good or bad is irrelevant from the perspective of the biosphere. Hubris, arrogance, and delusions of godhood are human attitudes and mental states. These are good or bad, depending on which anthropocentric notion of which human vices and virtues we should avoid. However, unless ecocentrism is understood as a theory of human virtues, it is silent on the intrinsic value of different human attitudes. Of course, there is a prudential reason to be wary of playing God: grand interventions can go wrong and cause commensurate damage. Clearly, any attempt at extending the lifespan of the biosphere requires extreme care and caution (especially if it is irreversible, for example, by ceding control over to pre-programmed robots). The high stakes imply that having humans deliberate over or continually monitor the relevant processes for extensive periods is desirable. Starting too early may be a serious mistake, but then humanity needs to survive long enough to start at an appropriate time.

Animal suffering Many animals suffer greatly in the biosphere: disease, predation, injury, and starvation are common (Horta, 2015). Extending the lifespan of the biosphere would increase this suffering and is hence undesirable.

Response: If one accepts the objection at full strength, then the conclusion would seem to be that interventions to shorten the lifespan of at least a complex animal-hosting biosphere would be favorable (Schopenhauer, 1958). Others have argued that nature selects for ‘beneficial pain’, and without it there could not be as much value, so one should accept the status quo (Rolston, 1992), in which case the objection loses power. However, a moderate position (still rejecting anthropocentrism) would be that humans have duties of assistance towards animals suffering in the wild (Moen, 2016)—a further reason for human civilization to persist. Note also that while some ecocentrists acknowledge the importance of animal suffering, most argue that it is less important than the continued existence of ecosystems. Indeed, the interests of animals and what benefits an ecosystem’s resilience may sometimes come into conflict in other contexts. For example, it has been suggested that ‘purging’ harmful genes from an inbred population of wolves could make the overall population, and by extension the ecosystem of which the population is part, better off (Liberg, 2005).

DiscountingFootnote 9 Some ecocentrists might argue that we should not value ecosystems in the distant future of hundreds of millions to billions of years from now in the same way that we should value the near-term future of tens to hundreds of years from now.

Response: The idea of discounting is plausible in some contexts, especially with regards to economic investments. It might for example be better to delay making renovations to your house for a year, if the costs of delay increase less than the costs of the interest on the same sum if it was saved or invested elsewhere. Some economists, most famously William D. Nordhaus argues that we should not only consider the interest rate, but also that we should also include “time discounting” (Nordhaus, 1997). Time discounting is what a person may indulge in when postponing a visit to the dentist, as it involves giving less weight to future well-being. Many philosophers have been very critical to this idea, as it seems to amount to a disregard for objects and individuals depending on when they happen to exist. Just as an object’s location in space is normatively irrelevant, so is an object’s location in time (Caney, 2014, pp. 323–327; Parfit, 1984, pp. 480–486; Rawls, 1999, pp. 259–262).

Conclusions

A permanent global collapse of human industrial civilization would be a moral disaster according to several normative theories. Thus far, this conclusion has mostly been supported by anthropocentric arguments. However, if human industrial civilization is currently necessary to avoid premature sterilization, then this would imply a strong instrumental reason to avoid this outcome, according to ecocentrism. Depending on how much importance is assigned to ecocentric values relative to anthropocentric values, these considerations may override anthropocentric or biocentric reasons for extinction (Benatar, 2008). Even if we believe that the persistence of humanity is bad for humans, the fact that only humanity could protect the biosphere may prove more important. In this case, the instrumental value of continued existence would be greater than the negative intrinsic value of human suffering.Footnote 10

Some authors have argued that a permanent collapse of human civilization would be very bad (Baum et al., 2019; Häggström, 2016; Ord, 2020). These arguments have often assumed an anthropocentric outlook. Ecocentric views have been more ambivalent towards the value of human civilization, sometimes even welcoming its destruction. We have argued here that proponents of ecocentrism have strong normative reasons to prevent this outcome as long as humanity remains the only species able to develop the capabilities of preventing premature destruction of the biosphere.

The timescale of human civilization is vanishingly short when compared to the timescale of the biosphere. But just as it is possible to do harm that has consequences far beyond the present, it may be possible to do good that stretches eons into the future.