THE INEVITABLE BOND (Davis & Balfour 1992; Davis 1993) is a fine example of what a multi-author book should be, both cohesive and rich and if there's a gain to be made by bonding with one's animals, it's worth considering the question of just how pervasive that bonding can be.
1.1 THE INEVITABLE BOND (Davis & Balfour 1992; Davis 1993) is a thoughtful look at how we students of animal behaviour relate to our subjects. It is a fine example of what a multi-author book should be, both cohesive and rich. Davis and Balfour have given this book a coherence that other volumes often lack. They have provided an introductory paragraph to each chapter, and have carefully ordered similarly themed chapters together; these flow from one to the next beautifully. The result is that ideas present in one chapter are picked up and reiterate in others, becoming powerful motifs.
2.1 One of the strongest themes in THE INEVITABLE BOND is the use of anecdotes, not only as a heuristic source for new ideas, but as scientific information in its own right (see also Bekoff 1993, paragraph 6). I put this first, because many of the points I will raise are based entirely on anecdotes and have not been systematically investigated. Because so much of this book consists of anecdotes, anyone teaching animal behaviour should be able to find stories worth telling their students.
2.2 Fentress puts the case for anecdotes forcefully and convincingly (Ch. 4). He shows that deep insights into the nature of an animal's behaviour can be gained by "one-off" incidents, although he cautions that such observations should be later supported by more controlled and rigourous testing (also Duncan, Ch. 18). Echoes of Fentress's argument are found throughout the book. Virtually every chapter has vivid anecdotes illustrating some point of behaviour. Even Summerlee's chapter describing the neurophysiology of arousal (Ch. 8) ends with an informal observation of how he could HEAR whether his rabbits had been disturbed by listening to their physiological response played through a speaker hooked into their apparatus.
2.3 Even though casual observations are often striking and can be a gold mine for new ideas, there is practically no place to publish anecdotal information. Natural history seems, alas, to be lost amid charts and tables (Kortlandt 1990). In neglecting anecdotal information, we do ourselves a disservice in several ways. For one, we can end up avoiding the topics that attracted us to animal behaviour in the first place (Crowell-Davis, Ch. 20). Second, we lose a rich source of knowledge about behaviour that can serve as a platform for other research (Bullock 1993). Worst of all, because so many of these informal observations inspire and guide our research, ignoring anecdotes can fundamentally distort our understanding of the research.
2.4 Here's an example: how much of behavioural science is DEPENDENT on animal/human bonds? More than one might think. As a result of spending many years working with snakes, Burghardt (Bowers & Burghardt, Ch. 16) is able to elicit attack behaviours that other researchers can't. Boysen (Ch. 12) says that a stable social environment for her chimps is "crucial" for gathering data. Similarly, chimps may have to form not just a relationship, but a particular kind of relationship with the researchers. Chimps are sensitive to social surroundings and differentiate between people who are "playmates" and those who are "mentors." Some humans even seem to be more popular with the chimps than others, but again, this has not been systematically investigated (Oden & Thompson, Ch. 13). Wolves seem to be very sensitive to a person's "body language" (Fentress, Ch. 4); it's likely that some people will be more effective working with wolves than others. Finally, Burghardt describes how, of several graduate students who worked with bears, only ONE developed a relationship with the bears that allowed him to run certain tests (Ch. 23).
2.5 These aren't just illustrations of how hard it can be to work with animals. These examples clear the path to a fundamental epistemological problem: what constitutes good evidence? Good scientific evidence is supposed to be repeatable by anyone. But if only a rare individual is able to form a working relationship with the animal(s), what should we make of the evidence then? Even worse, what if we aren't TOLD that the data were dependent on a relationship between experimenter and animal, because that information was seen as "too anecdotal" and left out of the paper?
3.1 Animals can be remarkably subtle and sophisticated processors of the information that we transmit. One example of that is their ability to recognize us individually. As one might expect, mammals recognize individual people (and sometimes their signals); this is amply documented in chapters on apes (Boysen, Ch. 12), bears (Burghardt, Ch. 23), wolves (Fentress, Ch. 4), seals (Schusterman et al., Ch. 21), rats (Dewsbury, Ch. 3), and horses (Crowell-Davis, Ch. 20). More unexpectedly, reptiles (Bowers and Burghardt, Ch. 16) and octopuses (Mather, Ch. 15) can recognize individuals just as well as "higher" vertebrates and remember people for long periods of time (weeks at least).
3.2 Animals can also be very sensitive to the signals we give off. Crowell-Davis makes this point in discussing the well known story of Clever Hans (Ch. 20). The lesson normally taken from Clever Hans is that experimenters should avoid inadvertently cueing their subjects. This seems to be what the editors have in mind when they declare the Clever Hans effect to be outside of what they are attempting to document in THE INEVITABLE BOND (Davis & Balfour, Ch. 1). Crowell-Davis argues that a subtler point is often missed, namely, how good animals can be at interpreting the behaviour of other species. Clever Hans attended and reacted to signals from another species (humans) that even conspecifics (other people) weren't aware of. Even when people became aware that they were cueing Clever Hans, they could not voluntarily control those cues. This sensitivity to interspecific signals has been neglected, despite its importance in so much research. For example, Pepperberg (Ch. 11) describes how when she is working with her parrot, Alex, she and her colleagues must ACT as if the task is interesting to them, or Alex is likely not to "pay attention" and start preening. Even worse, we often have very little idea WHAT animals attend to. A human wearing gloves and standing quietly induces more fearful responses in pigs than a bar-handed human crouching, for example (Hemsworth et al., Ch. 17), and chickens are less stressed by being caught by a machine than by humans (Duncan, Ch. 18).
3.4 Human ability to interpret animal signals is, at best, mixed. There are strong examples of our misinterpretation: dolphins' gaping ("smiling") is actually a dominance challenge (Estep & Hetts, Ch. 2); head nodding by some seal species is an aggressive signal that trainers have actually reinforced, hoping to use it in performance (Schusterman et al., Ch. 21). But Schusterman et al. make the counterpoint that sometimes the only way one can work with a species is by understanding and using their social cues. Two aggressive male elephant seals will rear up, with the taller being dominant; consequently, scientists can work fairly safely among male elephant seals, even though the animals weigh over 20 times what an human does, because we stand higher than seals can rear up (Ch. 21). Similarly, Burghardt describes how his graduate students had to fight off dominance challenges by their bear subjects (Ch. 23).
3.5 The interpretation of interspecific cues is a prime example of how experimenter effects can form a continuum (Davis 1993). As our closest phylogenetic relatives, chimps share a great deal of our behavioural repertoire and so recognize many of our cues (Boysen, Ch. 12). On the other hand, even though octopuses can recognize individual humans (Mather, Ch. 15), it's likely that their ability to decipher other human signals is limited; likewise, we have an impoverished ability to appreciate the "mindset" of an animal that might as well be out of science fiction. This may be why training octopuses is so difficult, and why they so often do things that seem specifically aimed at frustrating their keeper (personal observations).
4.1 The conventional wisdom seems to be that one should interact with experimental animals as little as possible. Several chapters detail how bonds can be minimized. Baum and Hiestand (Ch. 14) talk about automated testing boxes of the sort championed by B.F. Skinner. Boccia et al. (Ch. 9), however, note that handling effects are hard to eliminate, particularly when animals are kept and used in many experiments over the years. Obviously, in many instances, trying to reduce the number and intensity of interactions with the animal subjects is an appropriate thing to do, as it will reduce both handling effects and stress on the animal and researcher (e.g., Duncan, Ch. 18). Paradoxically, one way to do this is by developing some sort of relationship with the animals. Reinhardt details how, by developing a bond with their monkeys, they no longer had to fight to take blood samples (Ch. 10).
4.2 Such an approach -- working closely with animals with the express idea of forming relationships with them -- is far better represented in THE INEVITABLE BOND. Some scientists make a conscious commitment to work closely, almost intimately, with their animals, thereby exploiting the bond. Pepperberg (Ch. 11) has based her research program (effectively a decade-long and ongoing case study) on her interactions with one subject, her parrot Alex. This situation seems to be the norm in studies of primates, represented here by the work of Boysen (Ch. 12) and Oden and Thompson (Ch. 13). Other examples can be found in the chapters by Fentress (Ch. 4) and Burghardt (Ch. 23). Not all experimenters can manage this amount of commitment to their animals (e.g., Thompson 1976). Since I've already talked at some length about the use of anecdotes (see paragraph 2.4), and one only gets anecdotes by spending a lot of time with animals, it's no surprise that I think there is far more to be gained by exploiting animal/experimenter bonds than trying to avoid them. As this book's title suggests, avoiding them might be impossible anyway.
5.1 So if there's a gain to be made by bonding with one's animals, it's worth considering the question of just how pervasive bonding is. Mather's chapter on octopuses (Ch. 15) implicitly raises the issue: If we can bond with octopuses, couldn't we conceivably bond with APLYSIA or locusts or jellyfish? Obviously, there's no real answer to that question, merely impressions. Personally, while I like the sand crabs, crayfish, and other assorted "crunchies" I'm currently working with, it would stretch the word to its breaking point to say that I "bond" with them. I doubt that my interactions with them change, in any significant or unforeseeable way, their locomotion, which intrigues me so. But that's one of the beauties of a book like THE INEVITABLE BOND: it makes you consider those sorts of questions.
Bekoff, M. (1993) Should scientists bond with the animals they use? Why not? PSYCOLOQUY 4(37) human-animal-bond.2.
Bullock, T. 1993. Integrative systems research on the brain: resurgence and new opportunities. Annual Review of Neuroscience 16: 1-15.
Davis, H & Balfour, A.D. (1992) (Eds.) The Inevitable Bond. Cambridge University Press.
Davis, H. (1993) Precis of: The Inevitable Bond. PSYCOLOQUY 4(12) human-animal-bond.1.
Kortlandt, A. (1990) An old-timer's lament. Animal Behavior Society Newsletter, 35(1): 13.
Thompson, N.S. (1976) My descent from the monkey. In P.P.G. Bateson and P.H. Klopfer, eds. Perspectives in Ethology, Volume 2. Plenum Press, pp. 221-230.