Chapter 6 - What are we entitled to do other organisms?

In the following discussion, I shall be outlining an ethical position which is at once humanistic (insofar as it permits human beings to engage in activities that promote their own telos), Aristotelian (in its conception of what that telos consists of), and biophilic (in that it accords inherent value to each and every living thing).

In this chapter, I reject what I call the Instrumentality Thesis, or the principle that animals and other living things exist simply as resources to be used for our benefit. To justify our use of living creatures for food, clothing and medicine, I put forward an alternative Telos-Promoting Principle, which affirms that human beings are morally entitled to promote any of the basic human goods which constitute their telos. I then propose a Telos-Promoting Harm Principle, which limits the harm human beings may inflict on other living things, while pursuing their own good. I discuss further ethical constraints on our pursuit of basic goods. Finally, I propose a Justifiable Defence of Ecosystems Principle which allows human beings to defend the environment against the depredations of non-human organisms (e.g. introduced pests).

The clearest philosophical exposition of the idea that animals and other organisms possess a purely instrumental value is found in the writings of Kant, who regarded animals as things, or instruments to serve human ends:

If, as I have argued in this thesis, living things possess intrinsic value, then they cannot be regarded as mere resources that exist simply for the benefit of other beings. Intrinsic value is incompatible with IT.

Some philosophers would reject as absurd the claim that the lives of non-rational organisms can have any kind of purpose, meaning or value, apart from that assigned to them by their owners. But this objection misconstrues both the concept of finality (as extrinsic finality), and the concept of meaning (as an end that an agent strives for). Non-rational organisms do have to exist "for" anything else, in order to possess intrinsic value. And although they do not strive to find meaning in their lives, their internal goal-directed activities certainly have a meaning, which can be defined in terms of the organism's self-preservation and well-being - in other words, the realisation of its telos.

At first sight, the following statement from Aristotle's Politics (I.8, 1256a28ff.) might be taken as lending support to IT:

However, Aristotle's statements elsewhere about living creatures are clearly at odds with IT. Even from the above passage, it should be clear that Aristotle rejected the simplistic Kantian division of the world into persons and things. And since Aristotle certainly believed that all living things have a telos of their own, he can hardly have thought them to possess merely instrumental value, as Kant did. What Aristotle seems to have believed in was a natural hierarchy, whereby creatures lacking reason, though good in themselves, were also ordered towards (and could justly be used for) the benefit of rational beings.

From a contemporary perspective, there remains an unresolved tension in Aristotle's position: his biological views suggest that he regarded all organisms as possessing intrinsic value, but as the statement cited above reveals, he felt quite comfortable with the idea of humans using non-rational creatures for any purpose they please - not just for food but for "other accessories of life". Unless Aristotle's philosophy is able to exclude some ways of using creatures as immoral, his teleology is ethically superfluous, and does no extra philosophical work.

I shall re-visit Aristotle's views below. In the meantime, the question I propose to explore is: which (if any) usages of non-human life-forms can be sanctioned by someone who espouses that biocentrist position that I advocated in the previous chapter?

The principal objection to biocentrism is that it appears unable to justify making use of other living things, even for subsistence needs like food and clothing. For if all organisms possess intrinsic value, what could possibly justify one organism's making use of another?

I shall argue that this objection rests on the mistaken belief that the fundamental principles of ethics are negative rather than positive. The removal of this ethical misconception allows us to construct an ethic that is at once humanistic and biophilic.

This "negativist" approach to ethics is implicit in Comstock's (2000) argument that an ethic of intrinsic value (such as Paul Taylor's respect-for-life ethic) is incompatible with agriculture:

Comstock's argument is open to criticism on both emprical and moral grounds. First, his assertion that we could all live on a diet of "nuts, berries and vegetable products taken from dead or dying plants" is clearly erroneous if he is referring to the 6.4 billion people currently living on this planet. The reason why our Stone Age forebears (who, incidentally, ate lots of meat) switched to agriculture and took up a cereal-based diet (despite its nutritional inadequacies) was probably to satisfy the food demands of a burgeoning population (Diamond, 2003). (Whether we have a right to populate the planet at its current level is an issue that will be discussed below.)

In addition to its factual errors, Comstock's ethical argument is invalid: the premise that all living things are intrinsically valuable (and hence morally significant) does not imply the conclusion that we have an unconditional obligation to refrain from harming living things. Another premise is required: the premise that we have an unconditional obligation to refrain from harming morally significant beings. The Hippocratic injunction to "First do no harm" could be interpreted in this way. According to this account, the primary precept of morality is a negative one.

"First do no harm" may be an excellent cardinal rule for the practice of medicine, but it is clearly inadequate as a first principle of moral conduct. The notion of wrong is logically derivative upon the notion of right, whose content needs to be defined first. The notion that harming living things is a bad thing to do makes no sense unless one first recognises that promoting the well-being of an organism - including oneself - is a good thing to do.

An additional reason for rejecting the "First do no harm" principle is that it yields practical contradictions in a world where the interests of living things conflict. We can only eat by killing other living things. Should we then starve ourselves to death (thereby harming ourselves) or eat (thereby harming other living things)?

Finally, the notion that we have an over-riding obligation to refrain from harming other living things is ethically flawed, because it obliges us to put the lives of other living things before our own.

If we may sometimes harm other living things, then under what circunstances may we do so? According to environmental holists, we may inflict harm on other living things, if and only if our actions benefit or do no harm to the environment as a whole. Leopold's oft-quoted maxim encapsulates this ethic:

However, there are good grounds for believing that environmental holism cannot serve as a warrant for even our most basic needs, and that it is in any case ethically flawed.

Leopold's holism has been labelled as "environmental fascism" (Regan, 1988, p. 362), and not without reason. It is doubtful whether holism would allow most of us to satisfy even our vital needs. There is a widely held view among environmental holists that a human population of 6.4 billion puts a severe strain on the earth's biosphere, and that from a purely ecological point of view, it would be a good thing if most of us were not here. Some holists even claim that humanity's very existence is a threat to the biosphere. The Voluntary Human Extinction Movement (VHEMT)'s declaration that "phasing out the human race by voluntarily ceasing to breed will allow Earth's biosphere to return to good health", exemplifies the profoundly anti-human ideologies that are compatible with environmental holism. The cartoon below humorously illustrates the various shades of misanthropy endorsed by holists. (Misanthropic statements by some prominent spokespeople for animal rights and the environment are cited in the Appendix.)

However, the ethical arguments advanced by environmental holists are unsound. In particular, the argument put forward by members of VHEMT that we, as a species, have somehow forfeited our right to exist by causing the extinctions of other species, begs the question: to what, or whom, are we beholden for this "right to exist"? "Mother Nature", after all, is not a moral agent.

The general problem with ethical arguments in support of environmental holism is that they rest on a false metaphor of the biosphere as a living organism. According to this metaphor, each individual is a part of a whole, so its interests should be completely subordinated to those of the whole. In the previous chapter, it was argued that this metaphor was inaccurate, and that ecosystems (and by extension, the biosphere) do not exhibit the kind of part-whole subordination that living things possess. Because the functioning of its parts is not directed or regulated by that of the whole, an ecosystem cannot be said to have interests, properly speaking. An ecosystem, then, is not a single organism but a loose network of organisms, whose interests, although inter-connected, are nevertheless separately identifiable.

I conclude that any ethically sound justification of the harms humans inflict on other living things has to be based on an individualistic ethic, rather than a holistic one.

By contrast, the modest claim that any harms we inflict on the environment should be sustainable in the long term, does not in any way presuppose holism: it merely assumes that we care about our descendants. But if we believe that each and every living thing matters, then this minimal claim can hardly serve to define which harms we inflict on organisms are morally wrong. The fact that a practice is sustainable does not make it right.

If we are going to construct a non-holistic ethic that allows us to make use of living things, we might start with a minimalistic formulation: human beings may harm other living things, but only for the sake of satisfying their vital needs. Naess (1989) seems to endorse this idea in his Deep Ecology Platform:

However, Naess's ethical proposal would be extremely limiting, if implemented. For instance, it would only allow people to build essential structures like houses, as the construction of buildings destroys millions of earthworms, plants and soil microbes.

Naess does not define what he means by "vital needs" in his platform. Does he mean biological needs? Such an equation is too simplistic for humans and many other animals: as we saw in chapter 2, educable animals have ends that are not purely biological (Beisecker, 1999). In the previous chapter, we drew up a list of basic animal goods. In addition to pursuing their own biological ends, animals may also seek practical knowledge. In chapter 3, we discussed evidence that social animals enjoy the company of other individuals of their kind, as well as engaging in play.

To avoid focusing exclusively on an ethically restrictive set of needs, some authors have suggested a broader account of human and animal goods. VanDeVeer (1979) divides interests into three categories: 1) basic interests necessary for survival, 2) non-basic but serious interests, and 3) non-basic, peripheral interests. In a similar vein, Taylor (1986) divides interests into central and peripheral interests. Unlike Naess, Taylor allows human "peripheral" interests to take priority over the vital interests of other living things, provided that certain principles (self-defense, proportionality, minimum wrong, distributive justice and restitutive justice) are adhered to. However, these principles fail to address the central problem of any broad-based account of human goods: how can one justify prioritising a non-basic interest - even a serious one - over a vital one? Wetlesen (1999) argues that it is "open to doubt" whether Taylor, as a biocentrist, can consistently allow this. The prioritising of peripheral interests over vital ones is even more problematic: as we argued in chapter 1, the satisfaction of an individual's desires may be morally good, bad or neutral.

Our dilemma, in a nutshell, amounts to this: if "human interests" are broadly defined, they is no obvious reason why their bearer should be morally entitled to satisfy them at the expense of living things, especially when the ends they encompass may be good, bad or indifferent. On the other hand, the satisfaction of "vital biological needs", while morally justifiable, is ethically restrictive, as it merely entitles us to subsist, and fails to encompass the gamut of animal (let alone human) thriving.

The Telos-Promoting Principle (TPP): which goods are we entitled to pursue?

The above dilemma could be resolved by defining "benefit" and "harm" in an objective way. Satisfying a human interest at the expense of other living creatures could then be justified by arguing that it is objectively beneficial to human beings, or by arguing that not satisfying the interest is objectively harmful to human beings. The concept of telos provides us with a way of objectively grounding the terms "benefit" and "harm": an organism's realisation of its telos is a public occurrence, which is amenable to scientific investigation.

Human beings, like every other species, have a unique telos of their own. It was argued in chapter 5 that this telos includes a range of irreducibly distinct "basic human goods" (variously enumerated by natural law theorists) whose defining feature is that they contribute to human thriving in an objectively knowable way: they are wanted because they are good, not the other way round. Human thriving is contingent on having the opportunity to realise these goods. When deprived of this opportunity, human beings fail to thrive and are thereby physically, emotionally, mentally or spiritually stunted - i.e. objectively harmed - even if they do not realise it.

What I am proposing, then, is that human moral agents are entitled to do more than merely subsist: they are entitled to pursue basic human goods, in order to thrive as humans do. Because some of these goods require them to use other creatures, we can say that the usage of other living things is part and parcel of human thriving. Human moral agents cannot deny their right to pursue these goods without putting the welfare of other living beings before their own flourishing.

The proposal that each of us is morally entitled to pursue his/her own thriving may offend those altruists who believe that only unselfish acts are morally virtuous. But self-denying altruism is morally incoherent: one cannot consistently hold that the promotion of another's welfare is good while arguing that the promotion of one's own is not.

The biocentrist's justification for harming other organisms is a broadly Aristotelean one: although they are valuable in their own right, so are we, and our flourishing - which is defined by our attainment of the basic human goods comprising our telos - requires us to make use of them.

The ethical model that I am endorsing here is thus individualistic (in opposition to environmental holism), self-loving (in opposition to self-denying altruism), objectivist (in opposition to ethical relativism) and naturalistic, where the "nature" of a living thing is defined in terms of its telos.

We can summarise the foregoing considerations in an ethical principle, which I shall refer to as the Telos-Promoting Principle (TPP):

TPP should be seen as a restricted self-preference principle: it affirms that human beings are entitled to put their own interests first, while restricting the grounds on which people may do so to the pursuit of basic human goods.

The Telos-Promoting Harm Principle (TPHP): which actions are we entitled to perform?

While TPP tells us which goods we may pursue, it does not tell us which acts we may perform to achieve these goods. If human beings are entitled to attain these basic human goods, then they must also be entitled to perform acts which inherently tend to promote one of these goods. Because these acts contribute to human flourishing in an objective way, they are unambiguously good, as regards their essential "whatness". However, these acts may be morally wrong as regards the manner in which they realise a basic good. For instance, building a house is good per se, but building one in an ecologically wasteful manner is morally wrong. In other words, while the pursuit of basic human goods is good in an unqualified sense, acts performed to attain these goods are subject to ethical constraints.

Nevertheless, a telos-promoting biocentric ethic must be capable of sanctioning certain human practices, if it is to do any ethical work. I would like to propose at the outset a litmus test for any minimally "self-loving" environmental ethic: at the very least, it has to be able to justify, under at least some circumstances, those actions that have historically promoted the survival of our own species - that is, practices without which most of us would not be alive today. We can legitimately question the pace or manner of implementation of some of these practices, as well as certain specific instances of them, but we cannot doubt that these practices, in their historical context, were morally justified on the whole, without doubting the goodness of the conditions that enabled our own species to survive and flourish.

On an individual level, one may, of course, regret some of the actions that made one's own existence possible, without regretting one's own existence. Each of us doubtless has ancestors who were conceived as a result of rape, but we can nevertheless regret that these violations occurred without wishing that we had never been born. One can make a distinction between the bad actions of one's ancestors and one's own flourishing, which is morally good. What I am claiming, however, is that this distinction cannot be made at the species level, because a species is an entity that includes all its individual members - past, present and future - and defines their telos, which makes them the kind of individuals they are. To wish that telos-promoting actions by members of our species that saved Homo sapiens from becoming extinct had not occurred, is to say that it would have been better if humans had died out. One cannot consistently believe that, and simultaneously be glad that one is alive as a human individual.

In addition to being glad about being alive, it makes sense for self-loving individuals to be glad about the reductions in human mortality rates that have given members of our species increased opportunities to realise the basic human goods. Practices that reduced mortality rates in the past should therefore be regarded as justifiable (for the most part, at least) in their historical context. Historically, it is these practices that have allowed our species to flourish.

Which human practices, then, are we committed to sanctioning? Roughly speaking, they would have to include the following:

A genuinely self-loving ethical account has to sanction, in broad outline, both the Agricultural Revolution and the Industrial Revolution. Put simply, agriculture allowed countless millions of people to survive who would otherwise have starved to death, while industrialisation made possible the mass production of all kinds of commodities promoting human welfare and survival. Obvious examples of the latter would include life-saving drugs, devices for measuring the quality of a town's water supply, surgical instruments, fertilisers, pesticides, disinfectants and stoves for boiling unclean water.

Broad approval of historical practices that promoted the human telos is nevertheless compatible with qualified criticism of them. For instance, sanctioning the agricultural and industrial revolutions should not be construed as approving all of their effects on our world, or endorsing the entire range of commodities that have been produced through these practices.

Additionally, certain telos-promoting practices have turned out to be unsustainable for the human species in the long run. Our ancestors thrived by becoming proficient hunters and fishers as well as gatherers, but it is no longer possible for most of the planet's human inhabitants to live by hunting and fishing. And while DDT has saved hundreds of millions of lives, its use in present circumstances is neither necessary nor justifiable.

Nevertheless, since the foregoing practices made it possible for our species to survive and flourish, then we should accept as an ethical "given" their morally permissibility in at least some circumstances. For these kinds of activities, then, the only appropriate moral question we can ask is not: "Are they permissible?" but "When are they permissible?"

What ethical constraints, then, apply to acts performed in pursuit of basic human goods? As a description of these constraints, I would like to propose the following principle, which I shall refer to as the Telos-Promoting Harm Principle (TPHP).

The rationale for TPHP's conditions requires careful, extensive argumentation, which I shall attempt to provide.

The first condition of TPHP stipulates that an inherently harmful act cannot be justified solely in terms of its consequences. Because the act itself is inherently harmful, there is something flawed about the act itself. The only way to vitiate the prima facie wrongfulness of such an act is to re-characterise the act itself as something inherently good. I have argued that only a telos-promoting act can be regarded as inherently good. In a human context, this means that the act must inherently tend to promote a basic human good.

The words "inherently tend" have been chosen with care. For a harm-inflicting act to be morally justifiable, it is not enough that it happen (by accident) to promote some human good. It must do so inherently - i.e. because it is an act of a certain kind. At the same time, it is unreasonable to expect the act to infallibly achieve this good. Human actions seldom bring about results deterministically. Their efficacy may be probabilistic, or it may be off-set by countervailing causes. A reasonable minimum requirement is that the act performed should tend to promote a basic human good.

The differences between TPHP and IT should be readily apparent. IT justifies actions that inflict harm on creatures in terms of their good consequences in relation to some human end, while TPHP justifies them in terms of their inherent goodness as telos-promoting acts. Whereas IT allows people to exploit and/or harm other creatures in any way that advances their (morally licit) ends, TPHP would only allow people to use creatures in a way which is essentially connected to their telos.

To illustrate the difference between the approaches, let us suppose that the government of a certain country sets aside a sum of money annually for "trivial" biomedical research on animals, performed without any particular objective in mind, on the off-chance that it might lead to a serendipitous medical discovery with a practical application. The government knows full well, however, that most of this research comes to naught. The animals in the experiments are thus harmed for no clear reason. Let us suppose that the research, against all expectations, results in the accidental discovery of a life-saving drug. An IT proponent would then say that the research was justified because it brought major benefits to human beings, whereas the first clause of TPHP implies that the research was immoral because it had no inherent connection to the promotion of any basic human good.

(The implications of the other clauses of TPHP for animal research will be discussed below.)

The second condition of TPHP is similar to one of the conditions required by the principle of double effect (PDE), except that I have added a disjunct: part (a). In PDE, as in (ii)(b), the harm done must be an unintended consequence. However, when one is dealing with other living things, this condition is too restrictive: conflicts of interest between organisms are in the very nature of things, and one must sometimes intend harm to some other living thing in order to promote one's own flourishing. The point of part (a) is that if the harm is intended, there must be an intrinsic connection between the harm done and the realisation of some basic human good, before the harmful act can be called morally justifiable.

One example would be killing animals for food. Here the harm done is intended. The good realised is nutrition (an aspect of human health). Although the death of the animal per se does not nourish the human agent who kills it, it does enable the agent to prepare a food made from the organism's body parts (e.g. by chopping them up to make a stew). Thus, the harm done inherently enables the promotion of a basic human good. The eating of meat satisfies condition (ii)(a) of TPHP, even if (as I shall argue) it often fails to meet other conditions of TPHP.

Another example would be the use of antibiotics. Here, the eradication of bacterial infections is intentional, but a basic human good is thereby achieved: the restoration of our health.

The use of pesticides also satisfies this condition. While pesticides do not make us healthy per se, they do enable us to protect our crops and ensure that we have enough to eat. As recently as 1960, almost a third of the Asian rice harvest was eaten up by insects (Lomborg, 2001, p. 63).

The harms produced by industrialisation may be either unintended side-effects (e.g. a factory's production of toxic pollutants) or an intended harms, in industries where organisms' bodies are utilised as resources. In the latter case, the harm done contravenes condition (ii)(a) if the good realised falls outside the scope of the basic human goods.

Two clearcut cases of industries that inflict unjustifiable harms are the cosmetics industry, which destroys large numbers of plants as well as animals, and the fashion industry, which continues to promote fur purely for the sake of looking glamorous. Smelling fragrant and looking glamorous are simply pleasures, and pleasure for pleasure's sake does not qualify as a basic human good.

Finally, the harm done to organisms in urbanisation is an unintended side-effect of urban construction(as per condition (ii)(b)), rather than a directly intended harm. Additionally, urbanisation is a net environmental blessing because it saves the lives of countless organisms, since high density living spares land that would otherwise have been destroyed (Ausubel, 2000; Ausubel, 2001).

The point of the third condition of TPHP is to exclude acts which incrementally improve human welfare, but at a high cost to other creatures. In essence, this is a proportionality requirement. One difficulty with a pure utilitarian framework, highlighted by Joshua Frank (2002), is that "a minor human interest can prevail over a survival animal interest if a large enough relative weight is given to human interests or if the number of humans affected is sufficiently large". My requirement that performing the act must make a significant difference to the agent's prospects of realising a basic human good, is intended to partly address this concern. Only basic human goods count against the vital interests of other creatures, and only an appreciable improvement (not a marginal one) in the prospects of attaining some basic good could justify the killing of other creatures.

I should acknowledge that my proportionality requirement places much stronger limitations on what isolated individuals can do than those it imposes on society. Consider the following example. Suppose there is a medicinal plant extract, obtained from a rare species of tree in a tropical rainforest, which (on average) adds an extra year to a person's life. From an individual perspective, harvesting the tree for a marginal increase in one's lifespan would be morally unjustifiable, but from society's perspective, giving everyone an extra year of life would result in a significant overall gain in productivity, as well as a reduction in health care costs, freeing up substantial sums of money for spending on basic human goods. It might then be morally permissible for the government to set up a tropical timber plantation and log the trees for their medicine.

However, it should be borne in mind that the foregoing case is atypical: the social benefits are both measurable (in monetary terms) and meaningfully aggregable over a group of individuals, yielding a substantial social benefit. More commonly, the benefits to society of a proposed course of action can only be described in qualitative rather than quantitative terms, and cannot be meaningfully aggregated over individuals. (I would therefore consider Frank's (2002) constrained-utility approach to animal welfare to have limited application, as his aggregation of individuals' welfares lacks objective significance in most contexts.) In these "typical" cases, a significant improvement in the quality of people's social life - defined in terms of their attainment of some basic human good - is required to justify a project that would inflict harm on other creatures.

As an example of how condition (iii) might work, we might look at Frank's (2002) example of building a highway, which he describes thus:

I agree with Frank that it often makes sense to build highways, but disagree with his diagnosis of the failure of what he calls "rights-based systems". Frank argues that these systems fail because "the peripheral human interests in terms of reduced travel time" will always take second place to "the basic interests of wildlife". To complicate matters further, in practice, humans will risk their own lives (a basic human good) by speeding on the old road, so as to get from A to B more quickly. Frank contends that this alters the moral equation: we now have a trade-off between the two basic goods of human life and animal life.

I believe the expansion of the highway can be justified even without suicidal drivers, as the goal of getting from A to B falls under the oft-neglected basic human good of practical reason, which covers goals that people need to achieve in order to realise the other basic human goods. People need to travel in order to realise basic goods such as work and play. For a large number of people living in a city, a major reduction in travel time would probably result in a significant improvement in their opportunities to realise these other basic goods - especially the good of work, as people can complete projects that they might not have otherwise attempted, creating new business opportunities and resulting in a substantial alteration in people's lifestyles. According to the third condition of TPHP, this significant gain would then justify the construction of the highway. If, on the other hand, no new business projects were being planned as a result of the highway, or if the additional jobs created were short-term ones, then the gains would be too insignificant to justify its construction.

TPHP's fourth condition is arguably the one with the most ecological "bite". It can be characterised as a harm minimisation requirement, with some similarities to Regan's (1988) mini-ride principle and Taylor's (1986) principle of minimum wrong. My proposed harm minimisation principle, unlike Taylor's, does not allow humans' non-basic interests to over-ride the basic interests of other living creatures, but this difference is more apparent than real, as the interests I include as "basic" are considerably broader in scope than Taylor's. For instance, building a park would fall under the basic human good of play.

My harm minimisation principle stipulates that whenever we are considering a course of action that inflicts harm on other creatures, we should first ask what basic human good we are attempting to realise. Because these basic human goods are categories of ends, they are general rather than specific. Having identified the basic good, we should then ask whether there are any other specific realisations of it which cause significantly less harm to creatures, and which we can achieve without significantly affecting our attainment of the other basic human goods. (This last proviso frees us from the obligation to seek a less harmful instance of the basic good, if it would be prohibitively expensive or would otherwise substantially reduce our quality of life.) Having selected a specific realisation of the basic good we are trying to realise, we should then seek to achieve it in a way that minimises the harm we inflict on creatures.

If we are going to minimise harm done to creatures, then we need a way of evaluating the relative harms wrought by alternative courses of action. Some animal rights philosophers have proposed measures which appear relatively straightforward: counting the number of sentient animals or subjects-of-a-life affected. Such proposals can be faulted on both practical and moral grounds. If one confines one's scope to mentally normal mammals aged one year or more (Regan, 1988), then counting the number of victims is easy but morally arbitrary, as it excludes morally significant creatures with minds, such as fish, cephalopods and insects. If, however, one includes these creatures, then counting becomes logistically infeasible.

Additionally, the foregoing proposal simply ignores the biological interests of the other inhabitants of the biosphere: bacteria (many of which are beneficial to human beings and most of which are harmless), protoctista, plants, fungi and simple animals.

Finally, the animal-counting proposal only handles relatively simple cases. While I would agree that the wrongful killing of a sentient animal is worse than the wrongful killing of another organism, in real life, one seldom faces such straightforward choices. Usually one has to choose between killing different populations of organisms (some sentient and some not) which may live in different habitats.

I propose that we evaluate the relative harms caused by alternative courses of action by comparing their overall ecological impacts. There are different measures in use, such as the Ecological Footprint (Wackernagel and Rees, 1995; Redefining Progress, 2004) and the Sustainable Process Index (Narodoslawsky and Krotscheck, 1999). However imperfect they may be, they can serve as a general guide to our actions.

A good example of how TPHP's fourth condition works would be one's choice of transport. Aside from the good of practical activity (in this case, getting from A to B), the other human goods realised by travelling are the ones we realise at our destination (work, play, the pursuit of knowledge, etc). Trains and buses are 4 to 10 times more energy efficient than automobiles, when the occupancy rate is 1.5 persons per car (Baird and Hayhoe, 1993). Trams and light rail transport have a lower environmental impact than trains and buses (Encyclopedia of Sustainable Development, 2000) but this difference is relatively slight, compared with the environmental cost of car transport, so according to TPHP they could regarded as morally equivalent.

For most Australian city dwellers, reliance on public transport would marginally increase travelling time to and from work, but would otherwise make little difference to their attainment of the basic goods they pursue in their everyday lives, so according to TPHP's fourth condition, they should travel by bus or train whenever possible. By contrast, the lack of regular buses and trains in the countryside renders them impractical for most people living there. Also, people such as doctors, who have to respond to emergency calls, cannot rely on public transport.

For their part, the owners of public transport have a moral obligation to take whatever affordable measures they can to minimise pollution from their vehicles.

Finally, governments have an obligation to remove incentives for people to pollute. For instance, car and truck drivers should pay directly for road maintenance, as railway freight users do, to remove the bias toward road freight, which is 4 times less energy efficient than rail freight. Governments also have a responsibility to plan towns so as to reduce people's need to travel. Studies show that cities with a high urban density use far less petrol per capita than North American and Australian cities, and the separation of industrial areas into different zones also increases people's travel requirements (Baird and Hayhoe, 1993).

Urbanisation is often regarded as an evil, but from an ecological perspective it is a blessing: "By enduring crowding, urbanites spare land for nature" (Ausubel, 2000).

One might object that although the construction of built-up cities would save billions of organisms' lives, it would be at the cost of another basic human good: the enjoyment of beauty. But this objection is beside the point, as people living in built-up cities have lots of other opportunities to enjoy beauty in their everyday lives. In other words, the basic good has alternative instantiations.

In any case, many urban facilities that blight our landscape could be built underground over the next 100 years - not only underground shopping malls, but also underground inter-city magnetically levitated trains that would be ten times as energy-efficient as present transport systems. Already, the Swiss maglev plan links all Swiss cities within 10 minutes (Ausubel, 2001).

The fifth condition relates to long-term projects. A succession of telos-promoting acts, each of which meets the conditions of TPHP, may be performed as part of a long-term project, and yet because of poor planning, the project as a whole may fail to satisfy the conditions of TPHP - in particular, the fourth condition, which relates to harm-minimisation.

Family planning is a case in point. Even if each individual act by a couple of planning for another child satisfies the conditions of TPHP, their long-term project of planning for a family may fail to do so.

Because it is aimed at bringing a new human life into the world, the individual act of planning for another child promotes the basic human good of procreation, so the first condition of TPHP is met. The environmental harm resulting from the act - the inevitable burden that another individual places on the biosphere - is surely not intended by the parents-to-be, so condition (ii)(b) is satisfied. Condition (iii) is unproblematic: peformance of a procreative act, by definition, makes a significant difference to the agent's prospects of realising a basic good (procreation). Condition (iv)(a) of TPHP is not applicable to the decision to have another baby, because the good of procreation, unlike many other basic goods, is a singular good, which can only be realised in one form (having a baby). Thus there are no alternative instantiations. Condition (iv)(b) implies that if parents choose to have another child, they should endeavour to minimise the ecological consequences of their decision, as far as is practicable. This condition may also be satisfied. (For instance, the need for extra housing, which takes a toll on plants and animals in the environment, may be minimised by living in an extended family and putting the new child in the same bedroom as its siblings.)

The ethical paradox stems from the fact that it would be theoretically possible for a couple to have an unlimited number of children while observing all these conditions. Each child could be created and brought up in a way that minimises environmental harm, yet the aggregate environmental harm caused by creating such a large family would be enormous.

The proper way to avoid such a paradox is to include a requirement that TPHP's conditions should apply not only to the individual act of planning for another child, but also to the long-term project of planning for a family. In particular, the family as a whole should be planned in a way that minimises environmental harm.

It needs to be remembered, however, that an individual's moral status does not depend on how she originated, but rather on her nature. (The concept of nature was defended in chapter 5.) A child, once created, has the same moral status as any other child, even if her parents' decision to create her was morally irresponsible. While TPHP can set some ethical limits on the procreation of children, it can never be invoked to justify their destruction.

This condition complicates TPHP, but its inclusion is unavoidable. Briefly, TPHP as it stands is unacceptably selfish.

Earlier we rejected IT because it fails to take account of the intrinsic value of organisms, ascribing an absolute value to moral subjects only, and because it allows human beings to use living things for any purpose, however trivial. A telos-promoting biocentric ethic seems to face the opposite problem: it restricts the reasons for which we may harm living things to the promotion of basic human goods, but would allow us to inflict any harm whatsoever on other human beings (or sentient animals), if doing so promoted these same goods. Two cases bring out this point.

(1) Lifeboat dilemmas are a commonplace of ethics textbooks. In the famous case of Regina v. Dudley and Stephens, three sailors (including Dudley and Stephens), who were stranded on a lifeboat, ate a fourth sailor in order to stay alive. Most moralists would agree that the cannibal sailors should be found guilty of murder. Only if one of the sailors starved to death would the others be justified in eating his flesh. However, the first five conditions of TPHP would have allowed the three shipwrecked sailors to kill and eat their companion, because the act promoted the basic human good of health.

(2) In October 1984, the living heart of a seven-month-old female baboon was transplanted into an infant who was nicknamed "Baby Fae" by the press. (The baby died 21 days later.) Even if this procedure had succeeded, many ethicists would condemn it as unacceptably cruel. However, the first five conditions of TPHP set no limit on the kind or degree of cruelty we may inflict on animals, while pursuing basic human goods.

The question we now have to answer is: are these problem cases fatal for TPHP, or do they merely demonstrate its incompleteness as an ethical principle? To answer this question, we need to clearly understand why the first five conditions of TPHP yield wrong answers in these problem cases. The root of the problem, I would suggest, lies in three simplifying assumptions that have been made in this chapter about how human beings attain their telos.

First, the realisation of an individual's telos has been characterised from an external, third-person perspective, in order to guarantee its knowability to outsiders and its amenability to scientific investigation.

Second, the basic goods comprising an individual's telos have been treated as if they were independent of one another - as if realising one's telos were a bit like shopping in a supermarket, where the over-riding aim is to grab a bundle of goods (i.e. basic human goods) required for one's own satisfaction.

Third, each individual's telos has been treated as if it were completely independent of that of other individuals. The possibility that harming other individuals in order to procure some good might have a negative effect on one's own well-being has been overlooked.

In reality, a sentient creature's realisation of its telos is something which is not only achieved outwardly but also felt inwardly, on an emotional, mental and/or spiritual level. We thus have to consider the possibility that a harmful act on our part which realises one of the basic goods comprising our telos may have an adverse internal effect on our enjoyment of another good. In particular, the act of interfering with other individuals' realisation of their telos so as to procure a basic good (e.g. bodily health) may weaken or destroy our capacity to enjoy another basic good. This is especially true if the individuals harmed are sentient beings. Because we can empathise with them, we cannot separate their emotional, mental and/or spiritual well-being from our own.

In short: the first five conditions of TPHP supply deficient answers to the lifeboat and Baby Fae cases, because they neglect the morally desensitising and spiritually self-destructive effects of killing someone for food, and of inflicting extreme cruelty on animals. In the lifeboat case, the sailors saved their lives at the cost of hardening their hearts and becoming indifferent to the killing of a fellow human being. In the process of saving their bodies, they lost their souls. Likewise, the infliction of extreme cruelty on animals may destroy our human capacity to empathise with them.

My account of TPHP's shortcomings is one that seems to accord well with virtue ethics. If one assumes, as part of a natural law theory of ethics, that the moral virtues are among the basic goods that human beings must strive to realise, one might then want to append the following condition to the first four conditions of TPHP: the performance of the telos-promoting act must not make the agent less virtuous.

However, this formulation invites some objections. First, what is so wrong with a trade-off between the basic human goods? Why shouldn't you forego some of your moral virtue if it makes you healthier - especially if it saves your life, as in the case of Regina v. Dudley and Stephens? One possible answer is that the moral virtues are distinctively human goods: to forego your virtue is thus to throw away some of your humanity. It could then be argued that if your performance of an act threatens to irrevocably destroy your capacity for a certain moral virtue, then you should resist such dehumanisation even at the cost of your life. The sixth condition to TPHP could then be tightened to read: the performance of the telos-promoting act must not threaten to irrevocably destroy the agent's capacity for any moral virtue.

A second objection to a special "virtue clause" is that it blurs moral boundaries between people, animals and other organisms. For instance, we can empathise with animals, so killing them could weaken our capacity for empathy too. Butchers were once excluded as jurors from the trial of capital crimes for this very reason. Why, then, are we permitted to kill animals for food if no other sources are available, but not people?

In the previous chapter, we argued that human beings belong in a distinct moral category from other animals, because their telos embraces a range of goods - including the moral virtues - that the telos of other animals does not. Animals, in turn, have more "dimensions of value" than plants. We are back, then, to Aristotle's idea of a moral hierarchy, but with a difference. What we have here is a hierarchy of wrongful harm. The idea is not that plants and animals exist for the sake of human beings. A biocentric virtue theorist will strive to respect the flourishing of all living things - especially that of animals with whom we can empathise. However, because humans have a special kind of flourishing that includes virtuous conduct, the virtue theorist will avoid equating the intentional killing of a sentient being with that of a human being.

A third objection to virtue ethics is that there is considerable latitude of opinion among virtue theorists about which kinds of acts destroy one's virtue. Louden (1984) argues that certain acts are so intolerable that we need to draw up a special list of proscribed offenses. Since virtue theory fails to provide such a list, Louden argues that an act-centred morality has a distinct advantage over virtue ethics, because of the latter's inability to offer clear guidelines regarding our obligations.

A virtue theorist, for instance, might be hard-pressed to give a clear explanation of why killing in self-defence is justifiable, but killing someone for food in a case of dire necessity is not, whereas an act-centred account of morality could make a distinction based on the Principle of Double Effect.

It is not my intention in this thesis to adjudicate between virtue ethics and act-centred accounts of morality. In any case, TPHP can also be harmonised with act-centred accounts: TPHP's first condition, for instance, focuses on the nature of the act, and not its consequences. Both virtue theory and the various (non-consequentialist) act-centred accounts of morality provide us with complementary ways of viewing the same moral act. For instance, the wrongfulness of murder can be explained from:

The scope of parts (a), (b) and (c) in the sixth condition includes animals as well as people. Animals fall within the scope of the virtues insofar as we can harm or mistreat them. As I argued above, the performance of telos-promoting acts that cause extreme distress to animals has the potential to deaden our moral capacities.

As morally significant beings, animals also fall within the scope of over-riding obligations. While TPHP certainly justifies the killing of animals for food in the absence of practicable alternatives, there may (as we argued in chapter 5) be an over-riding duty not to kill one's pet, even in extremis, because it seems to qualify as a companion, and the killing of companions for food is wrong.

Over-riding obligations may sometimes force us to choose a course of action other than the least ecologically harmful one, and settle for a second-best option. Controlled burning may be a very efficient way of removing excess undergrowth in a forest, but one may have to consider another alternative if the forest is the habitat of an endangered animal species (e.g. the koala).

The question of whether animals qualify as rights-bearers requires an analysis of the concept of "right", which I shall attempt below.

The above clause prohibiting intrinsically immoral acts poses a potential danger to any genuinely self-loving environmental ethic. Because it does not spell out what these acts are, its absolute prohibition threatens to undo the philosophical work accomplished by the first five conditions of TPHP, which were designed to guarantee our right to engage in certain kinds of self-promoting acts. To counter this threat, I have decided to add two riders to the clause:

The first rider is a modest one: the only practices it exempts from moral controversy are those promoting the survival of the species.

The second rider is meant to guarantee that people's interests in realising basic human goods shall never be over-ridden by the basic interests of other organisms. Without this guarantee, TPHP could hardly be described as a truly "self-loving" ethic. The only possible exception to this "people-first" rule is where the good in question is one which by its very nature the organism has a prior claim to. I shall discuss this below, in connection with rights relating to one's body.

Despite the avowedly humanistic bias of the second rider, there are still some situations where it might be rational for me, as a human being, to set aside my basic interests in favour of other organisms:

Hart (2000) lists several definitions of "sustainability". It might be asked why TPHP does not include a "sustainability clause" stipulating that an activity harming creatures should be ecologically sustainable before it can be deemed morally justifiable. What I have included instead is a requirement that in cases where an action forms part of a larger project, that project should also conform to the conditions of TPHP - including the harm-minimising condition (iv). However, even a harm-minimising long-term project may not turn out to be sustainable, as even the best available means of reducing harm for the project may still go beyond what is sustainable over an indefinite period. Should the project then be discontinued because it is unsustainable?

I think not. Human patterns of activity change over the course of time, in such a way that patterns of activity that would be unsustainable over the long term are usually discontinued before they become a threat to the biosphere. Most of the historical practices that contributed to the survival of the human species would not qualify as "sustainable" if extrapolated into the indefinite future. It would be ethically burdensome to proscribe an ecologically unsustainable action that is only meant to serve as a "stop-gap" until a sustainable alternative is found.

For instance, it would be misanthropic to condemn industrialists living in the 18th century for building ecologically unsustainable smokestacks for their factories at a time when less polluting methods of waste disposal were unavailable. As well as providing people with work, making them more affluent and enabling them to purchase the necessities of life for their families, many of these factories produced goods that, directly or indirectly, saved millions of human lives: disinfectant, cheap soap, toilet paper, stoves for boiling water, and medical instruments, to name just a few.

I have also omitted Regan's worse-off principle from TPHP, because its implicit ranking of basic as well as peripheral goods is potentially extremely prejudicial to human interests. The principle stipulates:

If we follow this line of reasoning, then any proposed development (such as a highway) that: (a) enhances a lot of people's opportunities to realise basic goods, (b) does not save any human lives, but (c) results in the deaths of a few animals, will forever be stymied.

Two other principles that I have not included in TPHP are Taylor's (1986) principles of distributive and restitutive justice. Distributive justice is meant to apply to cases where self-defence is irrelevant, and where the basic interests of both human and non-human organisms are at stake. In chapter 6 of his book, Taylor argues that the basic interests of both parties should be given equal moral weight. The inclusion of such a proviso in TPHP would be incompatible with its avowedly humanistic tenor, and even Taylor concedes that this principle is often too rigorous to apply in practice. In that case, he suggests, the principle of restitutive justice then applies: if we decide to promote basic human interests at the cost of destroying lots of non-human organisms, we can make restitution to them. But as Sam von Mizener points out (Nolt, 2000), we cannot "make it up" to the organisms we killed - they are already dead! Alternatively, we could try to replace the organisms we destroyed with an equal number of organisms - a notion von Mizener rejects as impracticable, given the vast numbers affected. Lastly, we could try to rescue some other ecosystem, but this goes against Taylor's argument that it is individuals, not ecosystems, that are of primary importance.

While not obligatory, restitution to a species or an ecosystem can nevertheless be seen as a praiseworthy attempt to restore balance to nature for harm done to individuals.

The first condition of TPHP only allows us to inflict suffering on creatures when performing an act that inherently tends to promotes a basic human good, and the second condition further stipulates that the suffering itself must either inherently tend to promote a basic human good, or be an unwanted side-effect of the activity. Let us consider the case of a sadist who delights not only in inflicting pain but in placing his victims under such duress that they violate their own moral norms, by doing things they never thought they would do. The sadist finds a human victim (who happens to be an animal lover) and an animal victim (a dog). Instead of torturing the dog himself, the sadist orders the victim to torture the dog, who has been tied up for the purpose. The sadist says that if the victim does not comply, the sadist will torture and then execute the victim's family members. Torturing the dog cannot be excused here as an unintended side-effect of an action: the victim has to do it, to humor the sadist into releasing his family. Nor is it an act that promotes a basic human good per se: making the dog suffer does not save any human lives as such, but merely humors the sadist into releasing the victim.

Some readers may object to my suggestion that the victim should let his/her entire family be tortured to death rather than torture the dog. It may seem a terrible waste of life to let people die in place of one dog. "Isn't a person's life worth more?", they may ask.

The moral concept that is being introduced here is that of intrinsic value, or more precisely, inherent worth, as Taylor (1986) calls it. It has been argued in the previous chapter that comparisons of the intrinsic values of living things on a single sliding scale are inappropriate: instead of saying that human beings have "more value" than other animals, we should say that they have more "dimensions" of value, as they can realise a greater number of different categories of good.

In any case, neither the statement "A is inherently more valuable than B" nor the statement "A has more dimensions of intrinsic value than B" entails the conclusion, "C may sacrifice B for the sake of A", which is what the sadist wants his victim to do. To justify this conclusion, an additional premise is required: "B exists for the sake of A" - in other words, the Instrumentality Thesis which we rejected at the beginning of this chapter. What the foregoing premises do imply is that it would be worse for C (the victim) to kill (or torture) A (a member of the his family) than to do the same to B (the dog). However, that in no way establishes that it would be right to torture the dog.

On the other hand, TPHP implies that the obligation not to intentionally inflict pain and suffering on animals is only a qualified one. One can envisage an extreme situation - say, a fight to the death with a dangerous animal - where one possessed the means to make the animal desist and run away, but only by inflicting such pain on the animal as to make it too concerned for its own welfare to continue fighting (e.g. fighting off a shark by gouging its eyes). Here, one has to not only inflict suffering on an animal, but intend its suffering as a means to an end: its abandonment of a life-threatening course of action. However, it was argued in chapter 5 that we may not, under any circumstances, intend an animal's suffering as an end in itself.

Apart from defending one's life, what other situations (if any) justify the infliction of cruelty on animals?

Regrettably, TPHP does not give a clearcut answer to this question, but we can set some boundaries in the moral debate on this issue. Since TPHP has been designed to articulate a self-loving ethic, it has to sanction, in broad outline, practices that are cruel to animals but historically contributed in a significant way to human survival at times when no alternative was available (e.g. hunting for food and animal husbandry). At the same time, TPHP prohibits cruel practices that do not promote a basic human good, or practices for which there is an alternative, practicable, cruelty-free way of realising the basic good. (For instance, even if one argued that bull-fighting was a sport and promoted the human good of play, the practice would still fall foul of TPHP's clause (iv), as many other sports inflict no suffering on animals.)

On the other hand, there may be legitimate disagreement regarding practices which necessarily inflict cruelty on animals while promoting basic human goods, but which have never been of critical importance for human survival. Some proponents of TPHP might regard these practices as morally debasing and destructive of virtue (hence disallowed under clause (vi)(a) of TPHP), but alternative interpretations of TPHP are possible.

In chapter 5 it was argued that human beings have duties towards other living things, but the duties described were prima facie rather than absolute. In this chapter, I have attempted to develop two ethical principles (TPP and TPHP) affirming that people are entitled to perform certain actions in pursuit of basic human goods, while limiting the harms they may inflict on other creatures in pursuit of these goods. However, because TPHP is an avowedly "self-loving" ethical principle, it is open to the criticism that it is too susceptible to abuse to provide animals and other living things with reliable protection against the depredations of human beings. Only a robust affirmation of the rights of other living things, it might be argued, can offer this kind of protection.

At first blush, TPHP and the "rights" approach might appear mutually antagonistic. On the contrary, I hope to show that the naturalistic ethic I have been defending actually entails that all organisms have moral rights. Angus Taylor (1999, p. 47) defines a right broadly as "an entitlement to have, use or do something". My position is that all living things have rights in this sense, and that these rights are natural but not absolute. The thrust of the preceding chapter was that we have a prima facie duty towards each and every living thing: the duty to respect its interests. This is equivalent to the assertion that a living thing is entitled to have its interests respected by us. This entitlement is grounded in its possession of a telos. The range of goods which comprise an individual's telos is defined by its nature. Thus if we call the prima facie entitlements of a living thing rights, then these rights must be natural rights.

Additionally, TPHP is compatible with legal recognition of rights for animals and other living things, whether these rights are regarded as qualified claims that they have on us, or (for animals capable of intentional agency) liberties to pursue basic animal goods. I also argue that our legal system should allow lawyers to act as advocates for organisms harmed by proposed human developments (e.g. the construction of a highway), in order to offset any biases inherent in TPHP.

Certain distinctions relating to rights need to be drawn at the outset of any discussion of rights. As it is not my intention to develop a "theory of rights" in this thesis, I shall address them briefly.

The rights that philosophers generally concern themselves with are moral rights rather than legal rights: although the laws of countries such as Germany may accord certain legal rights to animals, advocates for animal rights argue for their position precisely because they believe on moral grounds that animals at all times and places should be accorded the same respect.

Nevertheless, legal rights are of practical importance: bestowing them upon a group may bring about due recognition of its members' interests and remove biases against the group within the legal system. For instance, proposed developments are often evaluated by a committee that decides whether they bring about a measurable improvement in social welfare despite the harm they will inflict on other living things. It might be fairer to give these organisms a legal advocate who can argue their case against a legal promoter of human interests, and let a jury decide whether the harm done to these organisms can be justified under the terms of TPHP.

Human rights scholars distinguish between what they call claim rights and liberty rights (Fagan, 2003). (Hohfeld (1919) identified two further categories of rights - power rights and immunity rights - but many scholars, including Jones (1994), have argued that these kinds of rights can be reduced to claim and liberty rights.)

Jones (1994) defines a claim right as being owed a duty. A claim right is a right one holds against another person or persons who owe a corresponding duty to the right holder. The definition put forward by Joel Feinberg (1974) in his now-famous defence of rights for animals, is essentially identical:

A claim right imposes on others the duty to respect that claim. Claim rights may be positive rights to some specific good or service, which some other has a duty to provide, or negative rights against others' interfering in or trespassing upon one's life or property in some way. Claim rights may be held in personam against some specifically identified person who has the duty duty holder, or in rem against no one in particular, in which case they apply to everyone (Jones, 1994).

In the previous chapter, I argued that human beings have duties towards animals and other organisms. This is equivalent to saying that they have claim rights against us. For instance, a pet's right to be fed and looked after by its owner is a positive claim right, held in personam. An animal's right not to be treated cruelly is a negative right held in rem. Jon Lowry (1975) argues that natural rights are claims to conditions necessary for living the good life, and that the concept of natural rights is applicable to animals as well as human beings. If the argument I advanced in the previous chapter is correct, all living things possess these natural claim rights, which should be construed as negative rights held in rem.

These natural claim rights of other organisms may sometimes be legitimately over-ridden by those of human beings. TPP affirms that human beings, too, have a natural right to promote each of the basic goods comprising their telos. Natural human rights can be regarded both as negative claim rights held in rem against any individual who would prevent them from doing so, and as liberty rights (see below). TPHP can be read as a statement of the conditions under which our natural rights over-ride those of other organisms.

Liberty rights can be defined as actions that may be freely chosen and that one is not prohibited from performing. As Fagan (2003) puts it, a liberty right is "a right to do as one pleases precisely because one is not under an obligation, grounded in others' claim rights, to refrain from so acting".

Some animal liberationists accord liberty rights to animals as well as human beings. For instance, Regan (1988, pp. 78, 84-85) believes that all normal mammals aged one year or more possess preference autonomy: they have preferences and are capable of initiating actions with a view to satisfying them. According to Regan, this kind of autonomy entails that they have lives that they have matter to them, so they qualify as subjects-of-a-life and thereby acquire claim rights against moral agents, as well as liberty rights to act on their intentions. However, because Regan's preference autonomy presupposes self-consciousness and a sense of the future (1988, p. 81), relatively few animals appear to qualify for the rights he proposes.

Other defenders of animal rights lower the bar, grounding them in their possession of sentience (Rollin, 1992; Warren, 1997; Francione, 1996) or the possession of beliefs and desires (Feinberg, 1996). As well as ascribing claim rights to animals, some of these authors regard them as having liberty rights to act in pursuit of their interests.

From a biocentric perspective, the question of whether animals can have liberty rights turns on how narrowly one defines "freedom". Aristotle's remark that "children and animals have a share in voluntary action, but not in choice" (Nicomachean Ethics 3.2, 1111b8-9) suggests that animals are small "a" agents who could be eligible for liberty rights in a qualified sense. The account of basic animal goods that I proposed in chapter 5 narrows the object of these rights: it implies that animals have liberty rights to pursue only those ends that are actually in their interests, rather than ends in which happen to take an interest (e.g. the pleasure of an addictive drug).

Consistency demands that if we grant animals liberty rights to pursue the goods that comprise their telos then we must acknowledge that human beings have these rights too. Accordingly, TPHP affirms that human beings have liberty rights to perform acts that promote basic human goods. These rights are, however, limited by the ethical constraints within TPHP.

The argument from marginal cases (Dombrowski, 1997) is often advanced in support of extending rights from human moral patients to animals. Cohen (1986, pp. 865-866) has challenged this argument, on the grounds that rights-talk can only be meaningfully applied to creatures that are naturally capable of making claims:

Cohen seems to be arguing that to have a right to X, you need to be the sort of being that can make a claim to X. Human moral patients qualify because they have a rational nature (which allows lawyers to make claims on their behalf which they presumably would make, if they had the full use of their natural faculties), but animals are naturally incapable of claim-making, which means we cannot do it for them either. To make a claim on behalf of an animal is to do for it what it would do if it were a moral agent - in other words, if it were not itself.

Cohen's reasoning can be faulted on two grounds. First, he appears to overlook the distinction between making a claim and having a claim. Feinberg's (1974, pp. 43-44) definition of "claim right", cited above, uses the term "have a claim", not "make a claim". It can also be argued that the notion of having a claim is epistemologically more fundamental, as it sets bounds on the kinds of things that an agent can meaningfully make a claim to.

Second, making a proxy claim on behalf of an animal need not be construed as doing for it what it would do if it were a moral agent. It could also be reasonably construed as an attempt to secure some basic good that the animal has a claim to, simply because it has an interest in realising it.

I conclude that there is no reason in principle for excluding animals from the domain of natural rights. (I will touch upon contractual rights below.)

Broadly, there are two approaches to the question of how rights can be philosophically justified. These are variously referred to as: (i) the welfare or interest theory and (ii) the will or choice theory. According to the first theory, rights are required to secure, protect and promote important interests of the rights-bearer. (One representative of this theory is Finnis (1980), who argues that human rights are justifiable on the grounds of their instrumental value for securing the necessary conditions of human well-being: the basic human goods.) On the interest account, there is no problem in principle with the idea of extending rights to human moral patients, as they have their own interests and welfare. Indeed, any non-human individuals with interests could be regarded as having rights. On the account I am proposing, that would include all organisms.

The second and more exclusive theory makes choice central to the concept of a right. Rights are grounded in the rights-bearer's capacity for freedom. This theory easily accounts for active rights, but it has been criticised for being incompatible with rights for moral patients such as children. It is, however, possible on the 'will' theory for children to have proxy rights, exercised by trustees or representatives (Archard, 2002).

Most defenders of animal rights have focused their attention on the interest theory of rights, as it does not require moral agency for the exercise of a right, but only the possession of interests. For example, Feinberg (1974) rejects the view that moral agency is required for possessing rights, but contends, reasonably, that only beings with interests qualify as rights-bearers, on the grounds that (i) beings that have no interests cannot be represented by legal proxies, and (ii) rights-bearers must be capable of benefiting from something, and beings that have no interests cannot be said to benefit from anything. The point where Feinberg parts company with biocentrists is that he envisages interests in subjective terms, arguing that they presuppose the possession of beliefs and desires. Hence sentient animals qualify as possible rights-bearers, but plants do not. Plants have no interests; they merely have needs.

Feinberg's argument appears to confuse the concepts of having an interest in something and taking an interest in it: the latter is only possible for sentient animals, but as we argued in chapter 1, we can legitimately speak of other organisms as having an interest in anything that serves their intrinsic ends. To say that they merely have needs - in the same way as a car needs petrol, for instance - is to gloss over the fundamental distinction between beings having and lacking intrinsic finality.

If, then, we wish to uphold a concept of "interest-based" rights for animals, logical consistency requires us to extend it to every living thing.

A few philosophical defenders of animal rights (Melden, 1988; Watson, 1979) have adopted the choice theory of rights, arguing that animals such as chimps, dogs and whales are candidates for rights because they exhibit some degree of moral agency. However, even these authors would concede that the vast majority of animals do not qualify as moral agents.

Philosophical attacks on the notion of animal rights usually focus on their lack of a capacity for language and moral agency. Such attacks on animal rights presuppose a choice theory of rights.

The difference between the interest and choice theories may appear relatively inconsequential if rights alone are being considered, but it becomes much more serious if a creature lacking rights is denied moral status. This would imply that we have no direct moral duties towards animals and other organisms. For instance, contractarians such as Narveson (1977), Carruthers (1992) and Leahy (1994) have argued that because animals cannot enter into moral contracts, they have no moral status of their own, and hence no rights:

An evaluation of the merits of a contractarian account of rights would be beyond the scope of this thesis. Suffice to say that its ethical starting point appears to neglect a morally relevant fact about organisms: they have interests. In chapter 5, I argued that all ethical argumentation has to start somewhere, and proposed that the proper sphere of moral action is the promotion of the welfare of individuals, by acting in their interests. (The alternative notion that behaving ethically is about respecting the choices of individuals, or more broadly letting them pursue whatever they take an interest in, was rejected in chapter 1 as too inherently subjective to ground objective moral claims, especially when the "good" being sought turned out to be something harmful to the individual.) I concluded that any individual possessing interests - including biological interests - was worthy of respect. Thus even non-sentient organisms qualify as morally significant others, towards whom we have (prima facie) duties.

Even contractarian accounts are not totally hostile to the notion of animal rights: if rational human agents can choose to award proxy rights to human moral patients, there is nothing to prevent them, if they choose, from assigning proxy rights to animals and other organisms. Legal enactment of proxy rights for non-human organisms (e.g. the right to legal defence of their interests by a court-appointed lawyer; the administration of a trust fund for the purpose of protecting and maintaining their habitat) would serve a practical purpose, as a counter-balance to our all-too-human tendency to overlook the interests of other life-forms - often to our own long-term detriment.

Surprisingly, not all proponents of sentientism or biocentric individualism consider "rights-talk" to be philosophically helpful. The long-standing view that "rights" can be re-expressed as duties is widely influential:

Thus some philosophers who support animal liberation, such as Singer (1987), affirm the inherent value of sentient beings but deny the usefulness of ascribing them rights, preferring to treat of their interests. Others (Midgley, 1984) eschew the concept of "right" as too obscure and too closely tied to a legal context to be usefully applied to animals.

If we accept that our duties towards others are grounded in their interests, then from a moral standpoint the ascription of rights to other organisms may appear redundant. On the other hand, the legal recognition of rights can confer a massive practical advantage in the real world, as the status of animals and other organisms requires a substantive defence against the encroachments of human beings, who are all too ready to put their own interests first. The legal affirmation of qualified rights for other creatures, coupled with a system whereby a legal advocate could make proxy claims on their behalf against any proposed developments adversely affecting their interests, could serve to redress this imbalance on a practical level.

Curiously, some philosophers advocate the granting of legal rights for organisms as a way of protecting their interests, but deny them moral rights. Paul Taylor (1986, chapter 6) regards moral rights as claims made by moral agents who can make choices, have respect for themselves and for others, observe moral norms, assert their rights, and complain if these rights are being violated - criteria which would exclude non-human life-forms. Sztybel (2001) considers Taylor's arguments to be question-begging, insofar as they stem from an anthropocentric, agent-centred model of how rights are exercised, and argues that mammals, at least, qualify as "autonomous self-managers" and hence rights-bearers. Additionally, he criticises the choice theory of rights (whose shortcomings I examined above) for its narrow focus on autonomy as the supreme good.

Finally, other writers (Donovan, 1990), while not rejecting the rights perspective, consider it too thin to ground our moral duties towards animals, and appeal instead to our capacity to sympathise and emotionally bond with animals as a proper basis for argumentation about the moral status of animals. On epistemic grounds, Donovan is generally correct: we do not acquire the knowledge that beating an animal is wrong from being told that its rights are violated, but by witnessing the mistreatment of animals and identifying with them.

There are, however, certain kinds of mistreatment whose wrongfulness may be more appropriately expressed in the language of "rights-talk". For instance, the frustration of animals' first-order desires is often condemned using the language of "animal rights", and the "liberation" of oppressed animals such as battery hens is frequently advocated. The ascription of rights to animals is thus philosophically useful if it makes us more clearly aware of ways in which we may wrong them.

Defenders of rights for animals and other organisms may regard these rights as either absolute or qualified. However, general agreement exists that these rights may not be over-ridden simply because of beneficial social consequences for human beings. Francione (1996) differentiates between what he calls animal welfare theories and animal rights theories on this basis:

I have argued that TPHP is compatible with limited animal rights. It is not vulnerable to Francione's criticisms of animal welfare theory, because it justifies actions not in terms of their consequences but in terms of their intrinsic "whatness": the harm-inflicting act must inherently tend to promote a basic human good for some individual. Merely having good consequences that "outweigh" the bad ones does not suffice to make an act morally justifiable.

Although TPHP is compatible with animal rights theories, its built-in axiom of self-preference principle entails that these rights are qualified rather tham absolute: human beings have a right to promote their telos too. Thus TPHP condones certain human practices that other animal rights theorists would object to.

TPHP has two advantages over other rights theories: it offers clearer moral guidelines about the circumstances in which human beings may put their own interests first, and it strikes a via media between self-denying moral rigour and excessive laxity.

Regan's (1988) theory tends towards rigour. It would allow subjects-of-a-life (mentally normal mammals aged one year or more) to over-ride each other's prima facie rights for three reasons: to prevent comparable harm from being inflicted on a greater number of individuals (the mini-ride principle); to prevent a considerably greater harm from being visited upon a few individuals (the worse-off principle); and to avoid being made worse-off by other individuals, provided that we do not treat them as mere resources when inflicting harm on them (the liberty principle). Despite these concessions to human interests, Regan's animal rights position could be criticised for being too strong, insofar as he envisages all subjects-of-a-life as having equal inherent value (although he admits that some are harmed more by death than others) and his theory does not guarantee the right of human beings to pursue certain kinds of goods (e.g. basic human goods).

Mary Anne Warren (1987) extends the concept of a right to all sentient individuals, but denies Regan's assertion that all animals have equal rights. Instead, she advocates a "weak animal rights" position which accords greater rights to human beings and allows them to over-ride the rights of animals. On her "sliding scale of moral status", some sentient animals (e.g. apes and dolphins) would have relatively strong rights that could only be over-ridden for very grave reasons, while others (e.g. insects) would have much weaker rights that could be over-ridden for a variety of social considerations. For these lower animals, Warren's notion of rights appears to be ethically impotent. Rights, as Angus Taylor puts it, are supposed to function as "a protective shield around an individual" (1999, pp. 18-19). Because it offers so little protection to simple-minded animals, Warren's account seems closer to what Francione calls a "welfare" theory than a rights theory.

Neither Regan's approach nor Warren's can tell us which kinds of human goods we are legitimately entitled to pursue when over-riding the rights of non-human individuals (animals and other organisms). One philosophical advantage of TPHP is that it not only affirms people's right to pursue basic human goods, but allows them to over-ride the claim rights of other organisms when pursuing these goods, if certain conditions are satisfied. People's rights to promote basic human goods are both negative claim rights held in rem (against anyone who would prevent them from doing so) and liberty rights (as actions that they may not be prohibited from performing).

For Francione (1996), the fundamental right animals possess is the right not to be used as a means to human ends. Similarly, Regan (1988, pp. 258-263) contends that animals have a basic right to be treated respectfully, which in turn entails a general, prima facie right not to be harmed. Rollin (1992, p. 83) argues that animals have but one "absolute, invariable and inalienable right" - the right to be considered morally.

I would argue that these proposed "fundamental rights" are better construed as definitions of the status of animals, for two reasons. First, these "fundamental rights" appear to be neither claim rights nor liberty rights. They cannot be liberty rights, as they do not describe actions that may be freely chosen. Nor can they be claims, because claims require some specific content: one has a claim "to something", as Feinberg (1974, pp. 43-44) puts it. Typically one has a claim to a good such as food or land, or the unimpeded pursuit of a good. Claims to "respect", "moral consideration" and immunity from being "used as a means to human ends" seem too nebulous to qualify as proper claims.

Second, if the natural law account of ethics being defended in this thesis is correct, we should define the most fundamental claims of morally significant individuals in relation to their telos. Generally, the fundamental claim right of animals and other living creatures should be their right to seek or pursue the basic goods that comprise their telos. Although this right is negative insofar as it does not entitle them to anything from us, it has a specific content which is determined by an organism's telos.

The proposals of Francione, Regan and Rollin contain merit, but they make more sense if they are viewed as definitions of animals' moral status. Animals are not in the same category as "things": they are not mere resources. (Neither, I would argue, are other organisms.) Because they have interests, they (and other living things) warrant moral consideration.

So far, we have only discussed telos rights, which pertain to the realisation of basic goods, but other rights have also been proposed for animals. I will examine two plausible candidates here: the right to the use of their own bodies, and the right not to be intentionally made to suffer.

Suconik (2000) proposes that nonhuman animals have a right to the "property", i.e., the physical bodies, that are their own. As he puts it, "Everything that constitutes the cat is the cat's own property." However, the notion that one's body can be regarded as one's property seems to create more ethical problems than it solves: it implies that one may trade away one's life and/or liberty, if one so chooses. Additionally, one important difference between one's property and one's body is that the former is alienable from oneself, while the latter is not.

There is a way to rescue Suconik's proposal. Instead of saying that animals own their bodies, it would be better to say that animals have a right to have their bodily integrity respected, while they are still alive. This kind of claim right has a certain plausibility: it accords well with traditional moral norms regarding animals, e.g. the Noachide Code of Judaism, which absolutely prohibits tearing the limbs from a living animal.

Pace Suconik, giving animals this right would not make it immoral for us to slaughter them for food. What it would entail is that animals, while they are alive, have a prior claim to the integrity of their own body parts and so may not be dismembered or deliberately subjected to irreparable bodily injury while alive. (I have added the word "irreparable" to exclude cases where animals lose body parts that they can grow back later.) There seems to be no good philosophical reason why this right could not be extended to plants too.

The right to bodily integrity need not be absolute: if it were, the neutering or spaying of companion animals would be unjustifiable. However, this right could be regarded as a very strong prima facie right, along the lines of a priority claim: no other individual has a greater claim to an animal's body parts than the animal itself. In the Baby Fae case, a baboon was cruelly mutilated - its beating heart was literally ripped out - to save a child's life, but the intuition that the baboon had a prior claim to its heart remains a strong one. The case of neutering a companion animal is different, as no other individual is claiming the animal's reproductive organs. Rather, the concern is to protect the community and the local ecosystem from uncontrolled breeding. (I shall discuss this below in relation to the Justifiable Defence of Ecosystems Principle,or JDEP.)

Recognising this kind of priority claim would also entail that any individual who killed an animal in order to harvest its body parts would be violating its rights under clause (vi)(c) of TPHP. I shall say more about this below in the section on xenotransplantation.

Would recognition of such a priority claim be compatible with a self-loving, telos-promoting ethic, such as TPHP? I believe so. While the killing of animals has at times been historically necessary for human survival, the same could never be said for animal mutilation. On the other hand, Note 2 in clause (vi), which says that other organisms' basic interests can never take precedence over ours, now requires an exception for cases where the notion of a prior claim applies.

Another right that is often urged for animals is the right to not be intentionally made to suffer. Here, we need to distinguish between deliberately making animals suffer as an end in itself and making them suffer as a means to an end. In chapter 5, I argued that an act of the former kind contravenes our most basic obligations to sentient creatures and can never be justified. TPHP can accommodate this moral insight by granting animals an absolute right not to be subjected to cruel treatment for its own sake, under clause (vi) part (iii).

On the other hand, making animals suffer as a means to an end is clearly justified in some circumstances, as the above case of the man-eating shark shows. Thus the right not to be intentionally subjected to cruel treatment cannot be absolute, but it could be viewed as a strong prima facie claim right. As we saw in the above section on TPHP and cruelty to animals, TPHP does not give a definitive answer of how strong this claim right is, but it can set upper and lower bounds on its strength. The right not to be subjected to cruel treatment is not trivial: it is strong enough to over-ride cruel human practices that do not promote a basic human good, as well as practices for which there is an alternative, practicable, cruelty-free way of realising the basic good. On the other hand, the right cannot be legitimately invoked against cruel past or present practices that have contributed in a significant way to human survival (e.g. hunting for food). I will discuss below its application to vivisection.

The practices of land-clearing, hunting and fishing, meat-eating, family planning, animal experimentation and xenotransplantation serve as useful illustrations of the kinds of practices TPHP would tolerate, question and disallow.

Land clearing is recognised as the biggest threat to wildlife in Australia: clearing of remnant (mature) vegetation destroys 190 million trees and kills at least 100 million native mammals, birds and reptiles each year in Queensland alone. 94% of all land clearing in Queensland is carried out to make new paddocks for cattle grazing; the rest is cleared for infrastructure, housing and crops such as sugar cane (Australian Conservation Foundation, 2003; Green Left Weekly, November 26, 2003, World Wildlife Fund, 2003).

Every list of basic human goods includes the good of "life" or "health". Food and shelter are essential parts of this good. It might be argued that because cattle grazing and sugar cane fall under the umbrella of "food", they would be allowed by TPP, while housing would be covered by "shelter". However, "food" is a very broad category. Our right to eat does not entitle us to eat wastefully. It is generally accepted that animal husbandry places much greater demands on the environment, for every kilogram of protein produced, than crop-growing. A proponent of TPP might ask whether Queensland should phase out its cattle industry, and engage in less destructive forms of food production. Land clearing for sugar cane is even harder to justify; many Australians are obese because they consume far too much sugar. Finally, medium or high-density housing would also reduce the need for new land.

On an individual level, matters are more complicated. Farmers have demanded that they should receive adequate compensation for revenue foregone, before they can stop clearing land. From a TPP perspective, the farmers are (reasonably) upholding their right to engage in the basic human good of work. The Queensland government should therefore be commended for its recent decision by to provide $150 million to assist landholders affected by the controls and to provide incentives to protect bushland, thereby enabling it to phase out broadscale land clearing of remnant bushland by the end of 2006 (World Wildlife Fund, 2004).

In Australia, suburban sprawl is a typical feature of city development. The harm done to living things is considerable: the clearing of land to build new homes destroys large numbers of soil organisms, and the construction of houses "demands a great deal of scarce material and energy for building, and considerable energy for operation" (Miner and Stomberg, 1998). TPHP would entitle us to clear land to realise basic human goods, but would also require us to minimise the harm done thereby. The idea that city-dwellers should adopt medium to high-density housing may strike many Australians as abhorrent, but it is commonplace in Europe and certainly compatible with an enjoyable lifestyle. Miner and Stomberg (1998) also suggest: building where possible on degraded land and then reclaiming it, rather than on undeveloped land; building or buying housing close to our daily work destination; and designing more compact, energy-efficient homes.

Natural law theories of ethics classify procreation as a basic human good, which contributes to human flourishing. While many people may choose to forego some of the basic goods, procreation is a good that almost everyone wants to realise at some stage. Most parents regard their families not only as their greatest source of joy in their life, but also as their greatest accomplishment in life. We need compelling reasons, then, before we can tell parents that it would actually be morally wrong (and not just inadvisable) for them to create another life. Citing a poorly measured, uncertain, long-term danger to the biosphere will hardly do. Because of massive uncertainties about the future, the danger should be clear and no more than two generations away, to count as a reason for family limitation.

TPHP is not intended to prescribe "ideal" courses of action, but simply to help us distinguish right actions from wrong ones. Consequently, TPHP cannot tell us how many children a couple should have. Nevertheless, it can impose ethical constraints on family size.

As we saw above, the decision to have another baby certainly satisfies the first three conditions of TPHP. The fourth condition does not rule out procreation: it only requires that parents take whatever steps are practicable to minimise the environmental consequences of their actions. The sixth condition is likely to be violated only in extreme circumstances. If, for instance, parents are barely able to feed their existing children, then their duties to care for them are of over-riding importance, as clause (vi)(b) stipulates. Even for a couple with no children who are facing starvation due to famine, the decision to create a child would fall foul of clause (vi)(c), as the child's basic subsistence rights would be denied. It was also argued that since a child's right to life was a natural one, the life of any child - even a wrongfully created one - has to be treated as sacrosanct.

The critical condition of TPHP is the fifth, which implies that not only the act of procreation, but also the long-term project of planning for a family, should be undertaken in a way that minimises environmental harm.

I shal confine my discussion of the environmental harms caused by population increases to three key areas: global warming and deforestation (both of which pose a threat to biodiversity), and the total ecological footprint of humanity. In addition, local environmental factors may limit the range of a couple's ethical choices when planning a family.

A recent report in Nature by Thomas et al. (2004), warns that global warming taking place between now and 2050 will doom to eventual extinction at least 9% and possibly a third to a half of all land animal and plant species. The report's authors urgently recommend cutting greenhouse gases and storing the main one, carbon dioxide, to save as many species as possible. Cutting emissions requires a cut in consumption. One way to achieve this would be to reduce the number of consumers (people).

Deforestation also threatens the world's biodiversity hotspots. According to NASA's Earth Observatory, "[i]f the current rate of deforestation continues, the world's rain forests will vanish within 100 years - causing unknown effects on global climate and eliminating the majority of plant and animal species on the planet" (Urquhart, Chomentowski, Skole and Barber, 2004). Reasons for deforestation include slash-and-burn agriculture by peasant farmers, capital-intensive construction of large-scale cattle ranches, commercial logging operations, and the construction of towns and dams. Population is one of several contributing factors.

The global carrying capacity of the earth's ecosphere is a matter of serious concern. Environmentalists warn that human consumption and waste production have overshot the earth's capacity to create new resources and absorb waste by a factor of 20%, and that we would need three Earths to support an American lifestyle for everyone (Wackernagel and Rees, 1995; Wackernagel et al., 2002).

The aggregate consumption of resources that threatens the planet is due to a combination of increasing consumption per person and increasing numbers of people. Of the two factors, affluent consumption is by far the more dangerous. A country's total CO2 emissions are much more strongly linked to its economic affluence than its population: the United States accounts for 24% of all global CO2 emissions, but has just 5% of the world's population, while India is responsible for 4% of emissions, but has 16% of the world's population (The Globalist, 2003). Similarly, the USA's per capita ecological footprint of 12.22 hectares per person dwarfs India's footprint of 1.06 hectares per person (NationMaster.com, 2004), giving the USA a much larger overall footprint than India.

A new study in Nature (Liu, Daily, Ehrlich and Luck, 2003) documents the effect of affluence on biodiversity. According to the study, an international housing boom is resulting in reduction in biodiversity around the world, as housing units throughout the world are being built at a rate that outpaces population growth. The housing boom is largely being driven by a global trend toward smaller households - primarily due to lower fertility rates, higher divorce rates, higher per capita income, ageing populations and a decline in multi-generational family units. Each household takes up space, requires resources to construct, and fuel to heat and cool it.

"We had hoped to find that, where human population growth was slowing, biodiversity might be given some breathing room," said Stanford University ecologist Gretchen C. Daily, co-author of the study. "But instead, we've found that urban and suburban sprawl are accelerating faster than population growth is decelerating" (Schwartz, 2003).

More people and greater affluence need not entail greater consumption of the planet's resources. Technological change can reduce humanity's ecological footprint. Indeed, some scientists (Bailey, 2002; Ausubel, 2000, 2001) question Wackernagel's (2002) key assumptions and projections, arguing that humanity's load on the biosphere is actually likely to diminish in the future, thanks to a combination of:

Additionally, Ausubel observes that "a vegetarian diet of 3,000 primary calories a day ... doubles the land spared" (2001).

It is universally accepted that Earth's current rate of population increase - about 1.2 per cent - cannot be sustained indefinitely. However, it is not immoral per se for a human society to increase at a rate that would be unsustainable in the long term. If it were, the Aborigines who first arrived in Australia would have been morally obliged to practice ZPG, instead of populating the continent, which is absurd. Historically, ecologically unsustainable population increases are typically brief, and are generally followed by more sustainable patterns. Populations do not typically increase exponentially over long periods of time, but follow a sigmoid (S-shaped) curve, where the the rate of growth is initially low, reaches a maximum and then gradually decreases to zero again (Boyle, 1998). As it happens, the global fertility rate has halved in the past 40 years. The earth's rising population of 6.4 billion is expected to peak at 9 billion and eventually settle at about 8.4 billion (BBC, 2002).

Even if humanity's total load on the biosphere is set to diminish as optimists forecast, clause (v) of TPHP requires couples planning a family to minimise environmental harm and not just keep it manageable. Does this mean that a two-child family is morally obligatory for each and every couple? The answer is no, for two reasons.

First, it is possible to intelligently plan for a larger family in a way that places no extra load on the biosphere. For instance, the above-cited study by Liu, Daily, Ehrlich and Luck (2003) found that in the case of deforestation, it is the number of housing units, rather than the number of people, that poses a threat to biodiversity. A household with an extended family, where two or more children share a bedroom, can be as eco-friendly as a smaller household. As one of the authors (Liu) put it, "In larger households, the efficiency of resource consumption will be a lot higher because more people share things". Thus a sensible ethical requirement relating to condition (v) of TPHP would be that the family one plans for must place no greater load on the biosphere than a two-child family.

I have used a two-child family as the benchmark because, barring exceptional circumstances, it is generally a good thing if a society reproduces at replacement level. Population decline is not viable in the long-term, for a variety of political and social reasons.

Second, even if ZPG were a duty, it would surely apply to society as a whole rather than to each and every couple. The reason is that in any society practicing ZPG, many couples will reproduce at below replacement level. That means that a few couples can have larger families, and society can still achieve the goal of ZPG.

Thus TPHP's fifth condition does not mandate family size, but effectively requires couples planning a family to limit the environmental harm caused by their family to the level of a two-child family. Additionally, in countries where fertility rates have fallen below replacement level, a few couples may provide a counterbalance by having larger families.

The practice of sport hunting kills over 130 million animals every year in the United States alone, not counting fish and other aquatic organisms killed for sport (Taylor, 1999, p. 80; World Farm Animals Day, 2003). In addition, the practice of recreational "catch-and-release" fishing causes many fish to die, mainly because of stress resulting from fighting the hook and wounding caused by the hook itself (Williams and Schwartz, 1992). The mortality rate for fish caught has been estimated at 10 to 50%. Indeed, it has been argued that for some popular species of fish, this kind of fishing actually kills more fish than traditional angling, owing to the large numbers caught and released (FishNet USA, 2000).

Defences of hunting and fishing which appeal to human goods in the spiritual domain - such as communion with nature (Ortega, 1972, Rolston, 1988; cited in Taylor, A., pp. 81-83) - would cut no ice with a proponent of TPHP, as these categories of good have alternative instantiations, that can also be realised without killing. (Mountain climbers can commune with nature too.) Nor can recreational hunting and fishing be defended by appealing to the basic human good of play; there are other enjoyable forms of recreation which cause no harm to creatures. TPHP's condition (iv)(a) is not satisfied for these goods.

Ted Kerasote (1993, discussed in Taylor, 1999, p. 87) defends his own hunting of elk for food, on the grounds that it kills fewer animals than the millions of mice, rabbits, birds and insects that are routinely killed, either by farm machinery and pesticides or through habitat destruction, in order to feed what he calls "supermarket vegetarians". From a TPHP perspective, Kerasote's argument could be construed as an appeal to the basic good of caring for other life-forms. Intentions complicate the moral equation. Many moralists would argue that intentional killing (such as hunting) is worse than killing which is foreseen but not intended (such as the killing involved in agriculture). However, Kerasote might reply that his hunting has a dual intention: he wants to (a) kill a few animals in order to (b) reduce the much greater destruction of animal life as a result of agriculture. A proponent of TPHP might regard such an aim as morally laudable. On the other hand, as Angus Taylor (1999, p. 87) points out, there is no way that the planet's six-billion-plus inhabitants could feed themselves by hunting, so "[o]n the large scale, then, the choice would seem to be between supermarket vegetarianism and supermarket meat-eating". The raising of animals to be hunted would violate condition (iv)(b) of TPHP, as these animals would have to be raised on land that could feed a far greater number of people through crop cultivation - "a vegetation diet of 3,000 primary calories per day ... doubles the land spared" (Ausubel, 2001).

On the other hand, from a TPHP perspective, the practice of subsistence hunting by traditional communities such as the Inuit was certainly justifiable in the past. Pursuit of the basic good of life (i.e. searching for food) justified hunting in environments where it was the only way to earn a living.

The moral issue becomes more problematic with the arrival of the welfare state and the cash economy. Today's Inuit hunt with snowmobiles and rifles purchased through trading, not with dog sleds and harpoons. In fairness, it should be pointed out that there are still few opportunities for work where they live, and that they use snowmobiles because they have to travel further to obtain food since being resettled by their government in the 1950s. Are the Inuit now obliged to cease hunting and either accept welfare handouts or migrate en masse to the big cities for work opportunities?

Angus Taylor (1999, pp. 84-87), in a review of the philosophical literature surrounding Inuit hunting and fishing, points out that the demoralisation caused by welfare payments has brought numerous social problems in its wake (suicide, alcoholism and drug abuse). Presumably these problems would only be exacerbated by the culture shock and loneliness resulting from wholesale relocation of these communities to urban areas. An overnight ban on hunting would devastate entire communities. A proponent of TPHP could argue that continued hunting by the Inuit (but not by non-indigenous peoples) is justified, in order to protect the basic human goods of work and society. (Animal liberationists may not relish this conclusion, but it is difficult to resist, unless one assumes that survival is the only goal humans are entitled to pursue when their interests conflict with the flourishing of other living beings - a position criticised above.)

Taylor (1999, pp. 84-87) also suggests that as the Inuit territory of Nunavut integrates with Canada's economy, other jobs are likely to become available, enabling a gradual phasing out of the hunt, although he questions whether the loss of traditional culture would be a good thing. From a TPHP perspective, the loss of traditional culture would not be a sufficient reason to maintain the practice of hunting. Human beings cannot flourish without some kind of culture, but the continuation of each and every practice in their local culture is not a necessary condition for their thriving. All cultures are in flux, and a gradual abolition of hunting over two generations would not cause massive disruption.

According to FAO figures, the number of animals killed worldwide for food in 2002 was 51.2 billion, which included 305 million cattle, buffalo and calves; 813 million sheep and goats; 1.2 billion pigs; and 49 billion chickens, turkeys and geese. (These figures exclude non-slaughter deaths, which are generally not reported.) The number of mammals and birds killed for food in the USA exceeded 10 billion in 2002, if we include animals who died before reaching the slaughter-house. During a 77-year lifetime, a typical U.S resident is responsible for the deaths of 11 cows, 32 pigs and sheep, and 2660 turkeys, chickens and ducks (World Food Animal Day, 2003).

In the preceding chapter, it was argued that we have a prima facie duty not to harm living things, and that the killing of an animal was worse than the killing of a plant. Additionally, the raising of animals requires the destruction of far more plant life (per kilogram of protein produced) than crop cultivation: "1 kg of beef costs 16 kg of grain, where the other 15 kg are used to support the animal's activities and parts we cannot eat... Poultry is somewhat cheaper in energy consumption" (Lomborg, 2001, pp. 102, 373). Since a vegan (or at least a vegetarian) diet would spare large numbers of plants and animals, we have a prima facie obligation to abstain from meat. From a TPHP perspective, the question of whether we actually have this moral obligation depends on whether there are any basic human goods whose realisation requires us to eat meat.

Meat-eating is sometimes justified on the grounds of taste alone, by people who consider it to be an irreplaceable good. However, on the natural law-based account which I am defending here, the pleasant taste of meat could never serve as an adequate justification for eating meat. First, many natural law theorists exclude pleasure from their list of basic human goods, arguing that it is not a good in abstraction from the activity in which pleasure is taken (Murphy, 2002).

Second, even if it is allowed that the pleasure we take in realising basic human goods is itself a basic part of human flourishing, this proves nothing more than that we are entitled to enjoy our food, in pursuit of the basic good of life. Alternative instantiations - that is, pleasant-tasting foods that do not contain meat - make meat-eating unnecessary and hence unjustifiable according to condition (iv)(a) of TPHP.

In the last few years, defenders of meat-eating have attempted to rebut common charges made by vegans and vegetarians that a meat-based diet is both "unnatural" and unhealthy. Colin Campbell's China study is commonly cited (Natural Hygiene Network, 2002) to show that a meat-based diet is linked to higher rates of cancer and heart disease. More recently, a new meta-analysis of 28,000 vegetarians and 48,000 non-vegetarians from five landmark studies, including the Oxford Vegetarian Study, found that over a 10-year period, the vegetarians were 24 per cent less likely than non-vegetarians to die of ischaemic heart disease, including heart attacks (Fox, 2000).

On the other hand, metabolic evidence of human adaptation to a diet that contains more meat than our ape relatives, coupled with anthropological evidence (from both contemporary and historical studies of hunter-gatherer diets) that meat-eating played a vital part in human evolution (Cordain, 1998) certainly gives the lie to oft-repeated assertions by some vegetarians that herbivory is natural for human beings. In a similar vein, Nicholson (1998) faults clinical studies which claim that vegetarian diets are healthier than omnivorous diets for simplistically equating "omnivorous diets" with "Western diets", which are very unhealthy.

From a TPHP perspective, however, these data cannot justify meat-eating. Rather, the crucial question is whether one can enjoy a healthy diet without meat or fish. In an affluent society, this is certainly possible.

Not so easy to dismiss is the increasing anecdotal evidence that veganism - and possibly vegetarianism - is associated with a decline in physical well-being known as "failure to thrive" in some individuals, possibly because the human body finds it harder to extract iron, zinc and essential fatty acids from plant-based foods. Fish and certain meats (liver, kidney and brains) are also ideal sources of DHA. Nicholson (1998) argues that clinical studies to date are too methodologically flawed for us to conclude that vegan or vegetarian diets suit everyone, and that reliance on anecdotal evidence, while problematic, is rational when little other evidence is available. He also contends that "mechanisms of built-in structural bias operate in the vegan community to minimize/prevent awareness of failure to thrive or its relevance" and suggests that two reasons why we have heard little about "failure to thrive" from ex-vegetarians is that they are often morally ostracised by their peers and tend to be less passionate about their beliefs than committed vegetarians.

Because it is a self-loving ethical principle, TPHP would permit those individuals whose health declines under a vegetarian regime to consume meat and/or fish, as condition (iv)(a), which relates to alternative instantiations, would no longer hold for these individuals. However, condition (iv)(b) - harm minimisation - would still apply: the individuals affected would only be entitled to eat the minimum amount of meat and fish needed to restore full health.

From a TPHP perspective, the argument over the environmental cost of meat-eating has been bedevilled by asking the wrong questions.

Much ink has been spilt in arguments as to whether meat-eating imposes an unsustainable cost on our environment. It has been claimed that if farmers worldwide continue to sustain a 2% annual growth in grain yields per hectare, ten billion people would be able to enjoy a meat-rich diet by 2070, and still use less land for farming than we do now (Ausubel, 2000). On the other hand, other models assume more modest annual growth rates of around 1 to 1.5 % (MacKenzie, 2000; Lomborg, 2001).

From a TPHP perspective, however, the important question is not whether the environmental cost of meat-eating is ecologically sustainable, but whether it minimises environmental harm.

Another morally misleading question related to land use for meat production is whether animal husbandry is an optimal way of using land. In certain places like Mongolia or the Australian outback, where the land is unsuitable for crop cultivation but can be used for grazing, meat production may indeed be "optimal". However, from a TPHP perspective, the ethical question at issue is not: "What should we do with Mongolia?" but: "What kind of healthy eating minimises the harm we do to the environment and to other living things?" This question arises from condition (iv)(b) of TPHP.

Even proponents of sustainable meat-eating acknowledge that the universal adoption of a vegetarian diet would result in the least disruption to the biosphere (Ausubel, 2001). It was argued above that the best way to measure this disruption was to use an aggregate measure such as the Ecological Footprint Model, developed by Wackernagel and Rees (1995).

Drawing on this model, Miner and Stomberg (1998) argue for vegetarianism on environmental grounds:

To be fair, Miner and Stomberg (1998) acknowledge that there are alternative non-vegetarian practices (eating only wild game for meat, and dropping eggs and dairy from our diets) could reduce our ecological footprint by a similar degree. It seems that the optimal practice, which would accord with condition (iv)(b) of TPHP, would be to combine the best of both worlds by adopting a vegetarian diet, and eating only as many eggs and as much dairy food as we need for nutritional purposes.

Even if dietary and environmental arguments are less than compelling, it may be argued that the cruelty inflicted on animals in farming of any sort - be it factory farming or so-called "free-range farming" - is so massive in its scale and scope that we are duty-bound to avoid all animal products, until humane farming practices become available on a large scale.

TPHP's fourth condition permits practices that inflict harm on other creatures, only if (a) there are no alternatives, and (b) harm is minimised. The fact that most of us can live without meat shows that condition (a) is not met. The widespread neglect of animals on today's farms, coupled with the suffering that they endure at our hands, show that (b) is not satisfied either.

In an affluent society where alternatives to meat abound, all of us should eat a lot less meat than the standard Western diet contains, although "failure to thrive" may excuse certain individuals from the obligations of a vegan or vegetarian diet.

Additionally, the cruelties that are endemic in modern animal farming oblige us to minimise our consumption of animal products, simply to minimise animal suffering. That means we should attempt to live on a vegan or near-vegan diet.

The ideal of cruelty-free eating is, however, unattainable. In our interactions with other life-forms, we live by killing (Passmore, 1980). Vegans have enough to eat only because farmers kill numberless insect pests, and crop cultivation kills not only plants but also worms and soil microbes. The best we can do is reduce the number of lives we take.

The mouse is the most common subject for animal testing. Image courtesy of Americans for Medical Progress.

Exact figures are not available, but around 60 million animals are believed to be used in the USA each year for biomedical research and testing. The vast majority are rats and mice (World Farm Animals Day, 2003).

The controversy surrounding vivisection is of two kinds: ethical and scientific. Some oppose the practice as immoral, while others question the scientific rationale for animal experimentation.

Rather than subject the reader to tedious non-philosophical arguments relating to statistics, comparative anatomy and neurophysiology, and the history of medicine, I have decided, after sifting through the literature in favour of animal experimentation (Americans for Medical Progress, 2002; European Federation of Pharmaceutical Industries and Associations, 2002; The Research Defence Society, 2004), and against it (Physicians Committee for Responsible Medicine, 1998; Greek and Greek, 2000; Americans For Medical Advancement, 2002; The Medical Research Modernization Committee; and the British Union for the Abolition of Vivisection, 2004), as well as publications by "open" groups with members from both sides of the debate (The Boyd Group, 2003; the Nufffield Council on Bioethics, 2001), to summarise the valid factual claims made by both sides in an Appendix. To cut a long story short: animal research has resulted in significant medical breakthroughs that would otherwise not have occurred, but its overall impact on human mortality rates is modest: improved sanitation and a better diet due to increasing affluence have had a much greater impact. The search for animal models for human diseases is very much a hit-and-miss affair, and has occasionally led scientists down a few "blind alleys", but the practical knowledge gained has been indispensable for combating diseases. The testing of new medicines on animals is a notoriously fallible procedure, but it is much better than no testing at all. Unfortunately, none of the proposed cruelty-free alternatives can completely replace animal testing at present.

It should always be kept in mind that "animal experimentation" is not a monolithic entity and that animals can be used in science for a variety of purposes:

Albino rabbits are commonly used for eye and skin tests because they are easy to handle and they have a very limited ability to "cry away" substances from their eyes during experiments. Image courtesy of British Union for the Abolition of Vivisection.

Anyone who believes that animal have an absolute right to control their own bodies will have to argue that animal research is intrinsically wrong, and that abolition is a moral imperative. While I respect the integrity of those who adhere to this absolutist position, I have argued above that a consistent application of their position would make it impossible for humans to satisfy even their vital needs.

Of the remaining arguments against animal experimentation discussed above, the strongest is the argument that it violates animals' right to bodily integrity. If, as was suggested above, animals have a prior claim to the integrity of their own body parts, then they may not be dismembered or deliberately subjected to irreparable bodily injury while alive.

The prior claim that animals have to their body parts would also prevent us from killing them simply in order to use these body parts. This would rule out the first rationale for animal experimentation, as well as xenotransplantation.

I should acknowledge, as a possible weakness of TPHP, the fact that while it allows for the possibility of "prior claims" such as an animal's right to bodily integrity, the existence of such a right is not implied by TPP but has been appended to Note 2 in clause (vi), in a rather ad hoc manner. Without this right, no other condition in TPHP would absolutely prohibit xenotransplantation. Condition (vi)(d), which prohibits destroying persons as a means to an end, would not apply, as non-rational animals do not qualify as persons.

Since TPHP applies only to the advancement of basic human goods, it cannot be invoked to justify the ninth rationale listed above for animal research (saving other animals from disease). As the animals whose lives are saved are typically not those upon whom research is performed, only a consequentialist could defend this kind of research. (Consequentialism has already been shown to be incompatible with TPHP's first condition.)

It needs to be stressed that TPP is a self-preference principle: it implies that we may value our own well-being above that of other organisms. On the other hand, the first condition of TPHP stipulates that any act causing harm to living creatures must not only achieve a good end, but have an essential connection to our telos: the harm-causing act must be re-describable as a telos-promoting act.

The first and second rationales for animal experimentation are easily amenable to a re-description of the harm-causing act as a health-promoting act per se. (The first rationale, however, violates animals' right to bodily integrity.) In these cases, animals are killed for spare parts or used as factories to make medicinal products (such as insulin). Likewise, killing animal to make medicinal products (e.g. vaccines) is obviously a health-promoting act. However, the fourth rationale - animal testing - cannot be re-described in this way, as it does not promote human health as such. Taking medicine derived from an animal makes me healthy; vivisecting or inflicting pain on an animal does not. Thus it might seem as if animal testing is not permitted by TPHP. The other rationales for animal research appear even more problematic, in the light of TPHP's first condition.

On the other hand, the human telos comprises other basic human goods apart from health, including the good of knowledge for its own sake. Although its medical benefits have been greatly exaggerated (see Appendix), animal experimentation, as a special field of scientific investigation, could certainly be said to promote this basic human good.

However, the fact that animal research enables us to acquire special knowledge about animals (including ourselves) does not justify the practice, any more than the fact that animals have a special taste justifies eating them. There are many different ways of pursuing the good of knowledge, without destroying animal or plant life. The fourth condition of TPHP would therefore prohibit experimenting on animals for pure research, as alternative instantiations of this basic good (other avenues of research) are available, which inflict no harm on creatures.

Much of what we know about the human brain comes from neuroscience research on monkeys. Image courtesy of Research Defence Society, U. K..

Practical activity is another basic human good that could justify animal research. Practical activity encompasses activities required to realise some other basic human good. This would include the pursuit of practical knowledge, such as the cure for a fatal disease. To justify experimenting on animals in search of a cure for the disease, one might argue that:

(i) the research is not "optional", as the cure that may be discovered as a result of the research is necessary for human beings who have the disease to realise the basic good of life; hence

(ii) the question of whether alternative instantiations (other avenues of applied research) exist is not morally relevant here; however,

(iii) this knowledge can only be gained by performing experiments on animals; therefore

(iv) animal research is necessary in relation to the end pursued (the practical attempt to save human lives).

This syllogism would apply to the third rationale for animal research, listed above. A similar syllogism could be constructed to argue for the necessity of safety testing of new life-saving medicines on animals, which falls under the fourth rationale.

Even so, it needs to be acknowledged that in animal experimentation, unlike the other harms we inflict on animals, the connection between the harm-causing act and the good achieved is a remote, long-term one. Most experiments that inflict pain on animals do not result in new knowledge on our part.

An animal liberationist might object that the above justification of animal research in terms of practical activity proves too much. The goal of "practical research" could be used to justify the infliction of unlimited suffering on animals, so long as it might advance our knowledge of how to save human lives.

Here is where the sixth condition of TPHP comes into play. Even if the research is needed to realise a basic human good, the act of engaging in the research and inflicting suffering on animals which does not in any way benefit them individually will inevitably be profoundly morally desensitising for the researcher, and may even irrevocably destroy his/her moral capacity for empathy. The resulting loss of virtue would therefore preclude the act, under clause (vi)(a) of TPHP. (This argument gains suasive force if the suffering is extreme, prolonged or routinely inflicted, or if the animal is irreparably harmed.)

There will inevitably be some disagreements among advocates of a telos-promoting ethic as to how their key principles (TPP and TPHP) should be applied to specific cases. Some moralists will argue that any research is morally hazardous and should therefore never be attempted; others, that the goal of saving human lives is so pressing that we have a duty to attempt this research. Moralists in the latter group could propose that we minimise harm to animals as well as the risk of researchers becoming morally desensitised. For instance, we might enact laws prohibiting mutilation as well as the infliction of certain kinds of suffering on animals, and we might only allow people to work in animal research laboratories for brief periods, to minimise the risk of desensitisation.

It was argued above that the eighth rationale for animal experimentation (pure research) contravenes TPHP's fourth condition. The sixth rationale (education) is defensible only where the practical knowledge gained is vital. Trainee surgeons, for instance, have to practice on live animals in order to fine-tune their techniques (Americans for Medical Progress, 2002).

4. Can harm done to other living things sometimes be justified on ecological grounds?

There is one more kind of harm that has yet to be addressed: sometimes it is necessary to inflict harm on some living things, in order to save other living things. In addition to TPHP, I would like to propose an ethical principle regulating this harm: the Justifiable Defence of an Ecosystem Principle (JDEP).

JDEP, which I proposed above, offers us an ethically sound and realistic way to adjudicate between the competing interests of different species, while looking after ecosystems. The governing idea is not that we have duties to ecosystems as such, but that in addition to their competing short-term interests, nearly all of the organisms in an ecosystem have a long-term interest in keeping their ecosystem sustainable. The sustainability of an ecosystem is a convenient short-hand way of representing the combined interests of the individual organisms living in an ecosystem. If humans manage an ecosystem in accordance with JDEP, they are acting on behalf of these organisms.

One implication of the foregoing proposal is that it would allow what Varner (1998) calls therapeutic hunting, to cull populations of animals in an ecosystem that are growing at an unchecked rate. ( A recent example is the culling of koalas on Kangaroo Island in March 2004.)

Another consequence would be that we are permitted to kill introduced species if doing so will save an existing ecosystem or an endangered species. (Tamar and brush-tailed wallabies are native to Australia, but they are a threat to native trees on New Zealand's Kawau island. Plans are afoot to eradicate them from the island by 2005, after about 100 have been shipped back to Australia to replenish local breeding stocks.)

Additionally, we are forbidden to jeopardise ecosystems simply in order to satisfy the needs of companion animals. (While TPHP allows for the possibility of destroying an ecosystem in order to promote human interests, it says nothing about the interests of companion animals.)

Also consider issues such as: the management of wildlife parks; and sustainable agriculture.

Chapter 6 - What are we entitled to do, vis-a-vis other organisms?