Animal Rights and the Wrongness of Killing

8. Preference Utilitarianism and the Replaceability Thesis

8.2 The Practical Replaceability Thesis

Book cover: Reason and Religion by Rem B. Edwards

The 'theoretical replaceability thesis' allows for the replacement of one self-conscious human being with another of equal utility on the condition that all other circumstances are identical. This consequence appears shocking to some. I suggest that this startling result appears counterintuitive for two reasons. Firstly, in practice the killing of one normal, adult human being cannot be compensated by the production of another. For the human being killed, the potential future happiness brought to the lives of those close to the person cannot be replaced, nor can the sense of enormous loss be compensated, by the production of a new human being. Rational, autonomous human beings are not like plastic dolls; that if one gets broken another will do just as well.

Secondly, there are no conceivable circumstances in which an organized society would countenance the killing of rational, self-conscious human beings as a prerequisite to bringing into existence another such being and as a way of maximizing utility. Even if it were practically possible for a single act of replacement to maintain or increase utility, a social practice set up along such lines and allowing such motives as an excusable defence for killing rational, self-conscious human beings would rapidly lead to social disintegration. Such a practice could not possibly survive because of the intolerable strain placed on people's need for security. These comments similarly apply to non-human societies, but to a lesser extent, in which the adult members are self-conscious.

Let us call the alternative thesis that it is practically feasible for a society of self-conscious individuals to establish and maintain such a practice, with utilitarian warrant, the 'practical replaceability thesis'. It is only because the practical replaceability thesis looks absurd that the theoretical replaceability thesis seems counterintuitive.[17] But replaceability as a means of maintaining or maximizing utility can only be the prerogative of very special beings in incredibly unusual special circumstances. These beings must be a kind of deity with only limited power in some respects, but very powerful and intelligent in others. They are the kind of being who is able to kill rational, self-conscious individuals without negatively affecting the happiness of other beings by, for example, increasing their fear of being similarly killed. Furthermore, these deities must have the capability of creating new self-conscious beings with the same or greater utilities than the ones killed, but can only create such new beings on the condition that they kill an identical number of pre-existing individuals. Once we realize that, in the real world, such special conditions could not possibly be met, our disquiet over the theoretical replaceability thesis ought to be settled.

However, for Singer, this cannot be the whole story. While rejecting the application of the practical replaceability thesis to merely conscious animals, he none the less accepts the application of the theoretical replaceability thesis to these same creatures. Why this disparity with self-conscious creatures? I suspect that this incongruity is due to the fact that the application of the practical replaceability thesis to self-conscious beings seems so much more shocking than the application of this thesis to merely conscious animals. In this case, I think we need to put aside our initial intuitive impressions and favour a coherent and consistent approach to replaceability. Consider how difficult it is to justify the application of the practical replaceability thesis to merely conscious beings.[18] Now consider how much more difficult it is to justify its application to self-conscious beings. The implausibility of the practical replaceability thesis when applied to both merely conscious and self-conscious animals should strengthen our confidence in accepting the theoretical replaceability thesis in both cases without fear of dire consequences.

Singer [1979c: 102] points out that his acceptance of the application of the theoretical replaceability thesis to merely conscious beings and his rejection of its application to self-conscious beings is consonant with the much greater seriousness he places on the killing of the latter compared with the former. This greater seriousness of killing self-conscious beings cannot, he maintains, be adequately accounted for on a 'classical' or 'mental state' version of utilitarianism, such as the one that I have been advocating here. Singer [1979c: 78–81] concludes that we need to adopt 'preference utilitarianism' for such beings.

Is it true that the classical 'mental state' view cannot account for the greater moral gravity of killing self-conscious beings? Before addressing this question, I want to make a related point concerning Singer's form of utilitarianism. Singer [1979c: 12f, 80] adopts 'preference utilitarianism' not only to account for the extra seriousness of killing self-conscious individuals, but because he believes this to be the form of utilitarianism that follows from the universalisation of our interests. Such universalisation, he maintains, is the meta-ethical prerequisite for moral thinking. This second reason of consistency with meta-ethical principles allows Singer to escape comfortably the charge of making an ad hoc revision to square his theory with considered judgements about the seriousness of killing self-conscious vis-á-vis merely conscious individuals.

But why, then, does Singer retain a classical view for merely conscious beings? His stated reason that it is required by 'the universal aspect of ethical judgements' [1979c: 85] sounds very odd following his rejection of the classical 'mental state' view precisely because it does not follow from the universalisation of our interests. His retention of the classical view also cannot be grounded in a rejection of the view that merely conscious animals have interests. He steadfastly maintains that merely conscious animals do have interests. And that they do is absolutely essential to Singer's application [1979c: ch. 3] of the utilitarian principle, the principle of equal consideration of interests, to animals.

It may be the case that Singer has retained the classical view for merely conscious animals because to have also adopted a 'preference utilitarianism' in regard to them would have made the rejection of the application of the theoretical replaceability thesis to self-conscious individuals harder to defend. If we were asked to maximize the satisfaction of rational preferences for all beings, merely conscious and self-conscious alike, then the desire for life is simply one amongst many. And if the satisfactions of these other desires are replaceable, then it is not intuitively obvious why the desire for life should be singled out for special treatment as being irreplaceable. This reasoning may explain why Singer has tenaciously avoided the adoption of the preference view for merely conscious beings, as is demanded by his meta-ethical theory.[19]

Book cover: Enlightenment Now: The Case for Reason, Science, Humanism, and Progress by Steven Pinker

Let us return now to the question I posed above: Is it true that the classical 'mental state' view fails to account for the greater moral gravity of killing self-conscious beings? According to Singer [1979c: 78–81], on a classical view, the only difference in seriousness between the killing of a self-conscious being and the killing of a merely conscious being is due to the extra loss in utility resulting from the side-effects of killing a self-conscious individual. On this view, if I kill a self-conscious being and it became known to other self-conscious beings, then they would worry that their lives were also in danger, thereby reducing their happiness. As Singer points out, if the killing is committed in complete secrecy, then even this extra reason for not killing self-conscious individuals vanishes.

The answer to this question consolidates the points made in this essay. In §6 above, I argued how the normal life of a self-conscious being has much greater utility than the normal life of a merely conscious being. This is because the capacity for self-consciousness brings with it a whole host of new desires, including desires for the future. So, on the classical view that I propose, the killing of a normal, self-conscious being is much worse than the killing of a normal, merely conscious being, because, other things being equal, it leads to a much greater loss of utility. This loss of utility is greater not only because of the additional side-effects of the killing; that is, the effects on the utilities of those left alive. Most importantly, the greater loss is also direct; from the greater loss of utility of the self-conscious individual themself.

The greater loss of direct utility from the killing of a self-conscious being happens whether the killing occurs in secret or not. In addition, the rules in practice that human beings and other self-conscious social animals adopt on mutually beneficial grounds morally prohibits the killing of fellow social members moreso than merely conscious animals. The normative theory behind the rules in practice that deem such killings more serious is consistent with a 'mental state' version of utilitarianism.[20] For these reasons, I consider the 'mental state' view to account naturally and easily for the greater moral approbation we give to the killing of self-conscious beings.

I further suggest that the view on killing self-conscious beings, including humans, and other animals proposed here avoids the difficulties in Singer's position without sacrificing its advantages. The view proposed is embedded within a unified utilitarian framework that is appealing because of its coherence and simplicity. It contains a single maximizing principle that avoids the problems of earlier views and sports a single theory of value; namely the classical 'mental state' theory.

In contrast, Singer has adopted two maximizing principles; the 'total' view for merely conscious beings and the 'prior existence' view for self-conscious individuals, both of which are encumbered with serious difficulties. He has also adopted two theories of value; the classical view for merely conscious beings and the preference view for self-conscious beings. This complication necessitates the development of a theory that will explain how these two types of value are to be balanced in cases where they conflict with each other. A clear example of such a conflict of values is where we are forced to choose between saving one self-conscious human life and many merely conscious animal lives.

Furthermore, the theory that I have advocated here explains in a coherent and systematic way the increasing seriousness of killing an animal as we ascend the ladder of neurophysiological complexity. It explains naturally the much greater seriousness we attribute to the killing of a self-conscious individual compared with the killing of a merely conscious being. Also accounted for is the asymmetry between our obligation not to kill an adult human being and the lack of obligation to bear children.[21]

A few brief remarks are warranted concerning the implications of the 'mixed' view maximizing principle for our obligation to optimize non-human animal populations. An optimum population size, according to this principle of maximization, is one in which mixed utility is maximized.[22] Ought we manipulate, using the technology we have available, the population size of each natural, non-human species in order to achieve some desired optimum? The short answer to this question is that there appears to be no conceivable means of manipulating natural ecosystems in a way that will increase mixed utility. Firstly, as far as we can tell, the mixed utility of a species' population cannot be increased by reducing the population size. (The exceptions here are cases of epidemic population growth that threatens the survival of other species within the same ecosystem.)

Secondly, to increase the population size of some species requires the importation of extra food resources from outside the ecosystem. These extra food resources, though, are better utilized in feeding severely undernourished human populations. Disadvantaged human populations possess a much higher potential for an increase in utility compared with non-human populations. It may be thought that it would be best, then, to kill all non-human animals in order to make extra room for an increase in human populations. However, this option is not feasible because of the already existing enormous difficulties we encounter in human populations with inequitable food distribution, rapid resource depletion, environmental pollution and global warming.[23] It seems, then, that it is not only in the interests of natural, non-human populations that we leave them substantially alone, but it is also in our human interests.


  1. [17] I have also argued against the practical replaceability thesis in its application to normal, adult human beings in my Allan [2015: §8].
  1. [18] For rebuttals of the application of the practical replaceability thesis to merely conscious animals, see Singer [1976: 169–74, 1979c: 100f, 1980a: 249f, 1980b: 332–4].
  1. [19] In personal correspondence with me, Peter Singer indicated that he is now inclined to apply 'preference utilitarianism' to merely conscious animals as well.
  1. [20] I refer the reader to my Allan [2015: §3] for the development of such a view.
  1. [21] On this last point, see my Allan [2015: §6–§8], where I deal with the problem of asymmetry in some considerable detail.
  1. [22] See my Allan [2015: §5]
  1. [23] See, for example, Meadows et al [1974].

Copyright © 2015

You will be interested in
Book cover: The Little Book of Humanism: Universal lessons on finding purpose, meaning and joy by Andrew Copson and Alice Roberts
Book cover: Philosophy of Mind: A Contemporary Introduction by John Heil
The God Delusion by Richard Dawkins
Book cover: The Invention of Science: A New History of the Scientific Revolution by David Wootton
Book cover: A Treatise of Human Nature by David Hume
Book cover: The Point of View of the Universe: Sidgwick and Contemporary Ethics by Peter Singer and Katarzyna De Lazari-Radek

Share This

  • twitter
  • facebook
  • linkedin
  • googleplus
  • gmail
  • delicious
  • reddit
  • digg
  • newsvine
  • posterous
  • friendfeed
  • googlebookmarks
  • yahoobookmarks
  • yahoobuzz
  • orkut
  • stumbleupon
  • diigo
  • mixx
  • technorati
  • netvibes
  • myspace
  • slashdot
  • blogger
  • tumblr
  • email
Short URL: