‘Animal Liberation’: An Exchange

The following statement concerning Peter Singer’s book Animal Liberation (New York Review/Random House, second edition, 1990), was sent to The New York Review following a review by Mr. Singer of several books on the treatment of animals in the April 9 issue.

To the Editors:

The objectives of both editions of Singer’s book are clearly stated in the preface of the 1975 edition, which is reprinted in the 1990 version. He wishes to convince us that “animals suffer from the tyranny of human beings” (p. iii) and to persuade us to end this “oppression and exploitation” of animals by extending to them “the basic moral principle of equal consideration of interests” (p. ii). To achieve these goals, he uses the techniques of propagandists, but he masquerades them in the guise of responsible scholarship. His methods include quoting authorities out of context, misquoting them, and quoting from obscure sources that are at best difficult to find and frequently impossible to check. Because of space limitations, we shall restrict our comments to his chapter on the use of animals in biomedical research, and present only three examples of the misrepresentations that it contains.

Singer selects research projects that can be exploited for maximal emotional impact and portrays them as the norm. About 25 percent of the pages in this chapter are devoted to criticizing studies on animal behavior and drug addiction—yet such research constitutes a very small fraction of the total use of animals in research and testing. He is particularly critical of the work of Harry Harlow and others who studied the effects of rearing infant monkeys in isolation. In attempting to discredit the value of and need for such studies, Singer quotes (p. 32) a British psychiatrist, John Bowlby,¹ who wrote:

The evidence has been reviewed at some length because much of it is still little known and the issue of whether deprivation causes psychiatric disturbance is still discussed as though it were an open question. It is submitted that the evidence is now such that it leaves no room for doubt regarding the general proposition—that the prolonged deprivation of the young child of maternal care may have grave and far-reaching effects on his character and so on the whole of his future life. Although it is a proposition exactly similar in form to those regarding the evil after-effects of rubella in foetal life or deprivation of vitamin D in infancy, there is a curious resistance to accepting it.

Singer’s quote deleted the important qualifier in the first sentence (underlined) without using ellipsis dots to indicate the deletion, and he neglected to include the last sentence (also underlined). He states that even though Bowlby’s conclusions were made in 1951, before Harlow began his research, “This did not deter Harlow and his colleagues from devising and carrying out their monkey experiments” (p. 32). However, when stated correctly (i.e. as he wrote them), Bowlby’s conclusions support doing the kind of studies that Harlow and others conducted in order to test the validity of the clinical evidence in an animal model. Such tests would (and did) convince skeptical physicians of the dire consequences of inadequate maternal attention.

It is significant, but not surprising, that Singer neglects to mention any of the benefits that have derived from the studies of Harlow and his associates. These benefits include improved methods of care for premature infants so that they thrive and thus can be removed from incubators earlier, and the acquisition of important insights into helping children who have problems socializing with their peers.²

In another attempt to discredit animal psychology research, Singer selectively quotes (p. 47) from an article by Steven Maier,³ which examines the usefulness of learned helplessness in animals as a model of depression in humans. The complete statement of Maier (beginning in the middle of a paragraph, where Singer’s quote begins) follows, and the sections that Singer deleted are underlined:

It can be argued that there is not enough agreement about the characteristics, neurobiology, induction, and prevention/cure of depression to make such comparison meaningful. Indeed, it has been argued that depression itself would not meet a rigorous application of the above criteria (McKinney, 1974). That is, depression might be sufficiently heterogenous in behavioral characteristics, neurobiology, causation, and prevention, that a given collection of depressed individuals might not closely match another. It might seem that a consideration of subtypes would resolve the issue, but even subtypes are probably not unitary in nature. Even a subtype of depression is a clinically defined syndrome or collection of events, and there is no strong reason to believe that such a collection will have a single cause. There may be many “routes” to what is labelled as depression, and they may not reduce to a single type even on a conceptual level. It would thus appear unlikely that learned helplessness is a model of depression in any general sense. However, animal “models” seem useful precisely because they do not duplicate the full clinical phenomenon but can allow the study of a single “route” in isolation [emphasis added].

Singer then states (p. 47), “Although Maier tries to salvage something from this dismaying conclusion by saying that learned helplessness may constitute a model not of depression but of ‘stress and coping,’ he has effectively admitted that more than thirty years of animal experimentation have been a waste of time and of substantial amounts of taxpayers’ money, quite apart from the immense amount of acute physical pain that they have caused.”

Thus, Singer completely misrepresents the sense of the statement from which he selectively quoted. Maier was in fact saying that although none of the available models of learned helplessness in animals duplicates exactly the condition(s) of depression in humans, the models are nevertheless useful for studying components of depression and how the condition(s) may develop.

One final example will further serve to illustrate Singer’s modus operandi. In a section on drug testing, Singer states (p. 53): “In fact, even when the test [in animals] is carried out on a medical product, it is most probably not going to do anything to improve our health. Scientists working for the British Department of Health and Social Security examined drugs marketed in Britain between 1971 and 1981. New drugs, they found,

have largely been introduced into therapeutic areas already heavily oversubscribed …for conditions which are common, largely chronic and occur principally in the affluent Western Society. Innovation is therefore largely directed towards commercial returns rather than therapeutic need.⁴

The first and second parts of this seemingly condemnatory quote are separated by more than six pages of text in the article cited. The section after the ellipsis is taken from the Discussion and Conclusions, in which the bulk of the text is summarized. By juxtaposing the two disparate statements as if they came from a single sentence, Singer makes the report seem to be much more critical of drug marketing than the authors intended (as is readily discernable by anyone reading the entire report. This sort of “scholarship” illustrates why writers are taught to be very careful in the use of ellipsis in direct quotes). It is significant that Singer omitted the sentence following the second part of his quote, vis., “The pharmaceutical industry like any other major industry is of necessity motivated by the need to be profitable.”⁵ Singer again twists the words of others to make them fit his version of the truth. Although most new drugs may be directed at alleviating the more common ailments of Western society, the process ensures that better pharmaceuticals are continually made available.

These are but three examples of many that we could cite to show how Singer misrepresents the truth about biomedical research and drug and product-safety testing. In addition, many of the sources that he cites could not be found (e.g., his ref. 62 and the first part of 87) or verified (e.g., 63, 105, 109). The number of errors in his citations (e.g., incorrect volume numbers, dates, journal titles) are also remarkable for a “scholar’s” work.

In the preface to the second edition of his book, Singer comments that the first edition is often referred to as the “bible” of the animal liberation movement. He states: “I don’t believe in bibles; no book has a monopoly on truth” (p. viii). This statement is most appropriate for both editions of Animal Liberation. Singer also proclaims: “The strength of the case for Animal Liberation is its ethical commitment; we occupy the high moral ground and to abandon it is to play into the hands of those who oppose us” (p. xiii). Singer’s ideas about what constitutes “the high moral ground” are evidently quite different from those of most people, as are his views on the moral equivalence of animals to humans.

Peter Singer is widely regarded as being the leading philosopher and moralist of the animal rights/liberation movement. Nevertheless, it is clear from the three examples of many we could have presented that he was obliged to grossly misrepresent the truth about animal research to support his thesis. We believe that this is a clear illustration of the moral bankruptcy and intellectual impoverishment of the animal liberation philosophy.

Charles S. Nicoll, Ph.D.
Professor of Physiology

Sharon M. Russell, Ph.D.
Research Physiologist and Lecturer

Audrey Lau, B.A.
Department of Integrative Biology
University of California
Berkeley, California

Peter Singer replies:

Charles Nicoll and Sharon Russell have previously published, in a journal that circulates among experimental biologists, an emotive call to other scientists to become “players” and not “spectators” in the defense of animal research against the “irrational zealots” of the animal liberation movement.⁶ Now, with the assistance of an academically more junior colleague, they appear to have subjected the second chapter of Animal Liberation to an unusually searching examination that has included checking all the footnotes. Again, their language is strong. They say that I use the “techniques of propagandists,” that my scholarship is a “masquerade,” and that I “grossly misrepresent the the truth about animal research.” What did they find that justifies such an attack?

Nicoll et al. turn first to a passage I quoted from John Bowlby. They are correct to say that I should have added ellipsis dots to indicate that I had not included the full first sentence. But even with the remainder of the first sentence, or the additional sentences to which Nicoll, Russell, and Lau refer, my point still stands. Harlow’s series of deprivation experiments on monkeys, conducted over many years with apparent indifference to the welfare of the monkeys involved, led to no new information about the effects of deprivation of maternal care on human infants. Bowlby himself was convinced, before Harlow’s experiments had begun, that the evidence left no room for doubt. Nicoll et al. think that Harlow’s experiments were nevertheless justifiable, not because they provided new knowledge, but because of the need to test the validity of the evidence in an animal model in order to convince skeptical physicians of the dire consequences of inadequate maternal attention.

To enable readers to judge this claim, I include here one of the (unchallenged) quotations in which Harlow describes his own work:

For the past ten years we have studied the effects of partial social isolation by raising monkeys from birth onwards in bare wire cages…. These monkeys suffer total maternal deprivation…. More recently we have initiated a series of studies on the effect of total social isolation by rearing monkeys from a few hours after birth until 3, 6, or 9 months in [a] stainless steel chamber. During the prescribed sentence in this apparatus the monkey has no contact with any animal, human or sub-human.

Harlow found that this form of isolation “reduces these animals to a social-emotional level in which the primary social responsiveness is fear.”⁷ Together with a colleague, Stephen Suomi, he then continued to experiment for several more years, as I describe in Animal Liberation. The experiments involved varying the dimensions of the isolation chambers to resemble what Harlow and Suomi called a “well of despair” and a “tunnel of terror.” Harlow and Suomi also developed a “rape rack” (again their own term) so that they could find out how female monkeys, reared in isolation themselves, performed as mothers.

Given the nature of Harlow’s experimentation, and Bowlby’s prior statement that the evidence of the dangers of maternal deprivation left “no room for doubt,” what are we to think of Nicoll, Russell, and Lau’s claim that Harlow’s experiments were justified in order to test the validity of this evidence in an animal model and thus to convince skeptical physicians of the harmful effects of maternal deprivation? I regard it as both intellectually deficient and morally callous. It is intellectually deficient because not even the worst orphanages of the 1950s ever put human infants in stainless steel chambers for several months after birth, isolating them from all contact with human beings or any other animal. Nor did they devise “wells of despair” or “tunnels of terror” to keep them in. So any “skeptical physician” who doubted the validity of the clinical evidence assembled by Bowlby would presumably also be skeptical enough to ask what on earth Harlow’s “animal model” has to do with the real-life situations in which human infants are suffering from maternal deprivation.

It is significant that although Nicoll et al. castigate me for omitting a sentence they consider to be relevant, they themselves, in providing a fuller quotation from Bowlby, omit an important sentence. Immediately after the last sentence that Nicoll, Russell, and Lau quote, Bowlby adds:

Indeed, there are still psychiatrists in all countries who challenge these conclusions, though it is to be remarked that few of them have had training in child psychiatry or experience of work in child-guidance clinics.

Had Nicoll et al. quoted this sentence, the weakness of their case for the justifiability of Harlow’s experiments would have been readily apparent. For Bowlby is saying that those who are most qualified in the field are already convinced of the psychological need for maternal care. The omitted sentence reinforces my suggestion that there were other ways in which these skeptical physicians could be convinced, without wreaking emotional destruction on generations of monkeys. The skeptical physicians could have been briefed on the available evidence by those with qualifications and experience in child psychiatry and child guidance. If “skeptical physicians” who deal with human patients are really more impressed by studies of monkeys under conditions of total isolation from all other beings than they are by the evidence of observations of human infants, that merely shows the excessive reliance that they place in the idea of an “animal model.” Any benefits that flowed from Harlow’s work could have been obtained, almost certainly for less cost and in a shorter time—quite apart from the cost to the monkeys—by other methods.

That Nicoll, Russell, and Lau should seek to justify Harlow’s experiments in this way is morally callous, because it is saying, in effect, that fifteen years of experiments that inflicted appalling suffering on monkeys were justified, because that was the most convenient way of convincing skeptical physicians of the validity of evidence that already existed. Even if this were true, it would omit any consideration of the price paid by the monkeys. I find it profoundly disturbing that Nicoll et al. see fit to defend Harlow’s experiments without making a single comment about the suffering they involved. They make no attempt to dissociate themselves even from the most extreme and prolonged forms of suffering that he inflicted on monkeys. It is precisely such moral blindness among those who carry out experiments on animals that shows the importance of curbing the virtually absolute power of American scientists over the nonhuman subjects of their experiments.

Nicoll et al.’s second example concerns my use of a passage in which Steven Maier, who has built his career on animal experimentation, concedes that exposing animals to inescapable electric shock has not produced a satisfactory animal model of depression in humans. My critics contend that my omission of a few sentences (represented by ellipsis dots) “completely misrepresents” Maier’s statement. But I informed my readers about what Maier said in these sentences. Following the quotation, in reference to the omitted sentences, and some subsequent ones, I wrote that Maier “tries to salvage something from this dismaying conclusion by saying that learned helplessness may constitute a model, not of depression, but of ‘stress and coping’.” I then added my own comment that, in saying this, Maier had “effectively admitted” that thirty years of exposing animals to inescapable electric shock has been a waste of time and taxpayers’ money, quite apart from the immense physical pain that it has caused. My use of the words “effectively admitted” made it clear that this was my own view of the effect of what Maier said, and not a literal account of his own words. Of course, Maier himself does not admit that thirty years of extremely painful animal experimentation was a waste of time—it would be difficult for him to do so, given that at the time he wrote the paper from which I quote, he had himself, with the assistance of the taxpayer, been inducing “learned helplessness” in animals. Nevertheless it remains my view that in the omitted sentences, Maier was making a desperate but implausible attempt to salvage something from his own recognition of the failure of a long line of extremely painful experimentation on animals.

In the third and last instance of alleged misquotation, I drew on a British Department of Health and Social Security report to show that even animal tests on products labeled “medical” are “most probably not going to do anything to improve our health.” Nicoll, Russell, and Lau take me to task because six pages separate the first and second parts of my quotation from this document. If this separation did, as they assert, distort the meaning of the authors, it would merit criticism. But it does not. Here is the first paragraph from which I quoted, in full:

The 204 new chemical entities marketed are listed by therapeutic category in Table 2. New chemical entities have largely been introduced into therapeutic areas already heavily over-subscribed.

And here is the entire paragraph from which I took the second part of my quotation:

Thirdly, despite this increase in the number of Product License approvals, the number of new chemical entities licensed each year has remained in the region of 20 per year. The bulk of the new chemical entities licensed have been limited to a relatively small number of therapeutic groups, for example, non-steroidal anti-inflammatory agents, corticosteroid, ß-adrenergic receptor blocking agents, antidepressants, benzodiazepines, penicillins, cephalosporins, oral hypoglycaemic agents, anti-hypertensives, in fact for conditions which are common, largely chronic and occur principally in the affluent western society. Innovation is therefore directed towards commercial returns, rather than therapeutic need. The pharmaceutical industry like any other major industry is of necessity motivated by the need to be profitable.

My combination of passages from these two paragraphs is more concise and easier to read than the entire two paragraphs. But does it create a different impression than the two paragraphs themselves? And if I had included the last sentence of the second paragraph, would that have weakened, or rather strengthened, my case that even when animal tests are carried out on medical products, the product is most probably not going to do anything to improve health? Does my compression justify the conclusion that “Singer twists the words of others to make them fit his version of the truth”? Note it is Nicoll, Russell, and Lau—not the authors of the report—who say that the market process “ensures that better pharmaceuticals are continually made available.” The report displays no such complacency about the workings of the market. On all these questions, I am happy to leave the answers to the reader, who now has the information needed to judge my standards of scholarship, and the fairness of Nicoll, Russell, and Lau’s critique.

I have now shown, I believe, that none of the three examples presented by Nicoll, Russell, and Lau justifies what they say about my scholarship. In addition to these three examples, they claim that they could cite many others. Since they give no details of these many others, such a claim is easy to make. But the three they chose were their own choice. I can only presume that they chose the most serious examples they could find. They also mention two footnoted sources that they could not find, and three that they could not verify. That in a total of 138 footnotes, five should be difficult to locate or verify is surely not unusual. One of these, for example, refers to a magazine published in Durban, South Africa, another to a regional Californian newspaper, and a third to a transcript supplied to me in 1974 by the American television network WNET-13. I am not sure to what lengths Nicoll et al. went to try to obtain these documents. My critics also refer to a “remarkable” number of errors, without giving a single example. That, again, is an easy claim to make, but a difficult one to refute.

The second chapter of the revised edition of Animal Liberation consists of sixty-nine pages of text and eleven pages of footnotes. The text includes descriptions of more than fifty experiments on animals, some of them involving more than a thousand animals each. Most of these experiments were carried out in the United States in the past decade or two. The descriptions often include quotations from the experimenters’ own publications. They include accounts of the slow poisoning of beagles, the infliction of severe and inescapable electric shock on dogs, and the lethal irradiation of monkeys who, while vomiting and suffering diarrhea from radiation sickness, were nevertheless forced by electric shock to keep running on a treadmill until they died. Given this, it is noteworthy that Nicoll, Russell, and Lau do not object to a single description of an experiment. Nor do they challenge the fairness of even one of the quotations I have taken from the many journal articles in which experimenters describe the suffering they themselves have inflicted on animals. These descriptions and quotations were virtually all taken from articles published by the scientists themselves in standard scientific journals, easily available to a research team with access to the library of the University of California. (None of the footnotes to these articles are among the five that Nicoll et al. found difficult to locate or verify.) The descriptions of experiments constitute the core of the chapter. Doubtless if Nicoll and his colleagues had been able to find any errors in this material they would have proclaimed their discovery as further evidence of my “propagandist” techniques. That they say nothing against my accounts of the nature of experimentation on animals can therefore be taken as a reluctant endorsement of the accuracy of this damning material. For that I thank them.