Interpretation And Explanation In The Study Of Animal Behavior

Volume I, Interpretation, Intentionality, And Communication

Ph.D. Bekoff

• Publisher: Routledge
• Year: 2021
• ISBN:
  9780429713651
• Edition: 1
• Topic: Filosofía
• Subtopic: Historia y teoría filosóficas

People have long been fascinated, not just by the behaviour of non-human animals, but by the problem of how this behaviour is to be interpreted and explained. This is one of two volumes of original essays on the cognitive and emotional dimensions of non-human minds and the relationship of natural minds to behaviour. The essays also address questions concerning the meaning and significance of consciousness; animal intelligence, awareness and emotions; behavioural plasticity, flexibility and constraints on understanding animal minds; and the structure of explanation in the study of behaviour.

II Method, Analysis, and Critical Experiment

DOI: 10.4324/9780429042799-8

Introduction

DOI: 10.4324/9780429042799-9

How do we come to know what we know? What is good evidence? How much of what we learn is constructed in the first place by our presuppositions or by the methods that we use to collect data? How much of what we learn is "really out there?" While the collection of behavioral data is often simple, its proper analysis and explanation can be very difficult. Early ethological studies were typically highly descriptive and nonquantitative, with little evidence of hypothesis testing. Perhaps this is what led de Solla Price (1960: 1830) to write that "biology is a system that proceeds from biochemistry to the associated subjects of neurophysiology and genetics. All else, as they used to say of the non-physical sciences, is stamp-collecting."

The chapters in this section go a long way towards making it clear that ethological research is not mere stamp-collecting. Some essays consider case studies, while others discuss the applications of various techniques to the study of animal behavior. They all begin with the foundational assumption, articulated by Charles Otis Whitman (1898/1985) and Oskar Heinroth (1910/1985), and later championed by Konrad Lorenz (1981), that behavior is something that an organism "has" as well as something that an organism "does:" Behavior is a measurable phenotype that evolves (Barlow 1977; Bekoff 1977; also see Gordon Burghardt & John Gittleman's chapter in this section) and is influenced by immediate environmental variables.

In their case study of the role of the Mauthner cell in avoidance and antipredatory behavior, particularly in zebra fish, Randolf DiDomenico & Robert Eaton show how rigorous study can inform arguments about the causation of this deceptively "simple" motor response. They use the term "neural positivism" to refer to "attempts to define neural functions by what fails to happen in the absence of neural structures, and by what occurs when neural structures are artificially stimulated." Based on their critical analysis of what is learned using the three main ways of studying relationships between the nervous system and behavior correlation between the activity of a structure and observed output; stimulation and the production of a behavioral response; and lesioning or preventing the activity of a structure and observing whether a given response is eliminated - they reject the notion that a univocal (one function, one structure) causal paradigm can be applied to the nervous system. They show that even with respect to very "simple" behaviors in "simple" organisms, a given end is not always achieved by fixed means. This has important consequences for experimental design in neuroethology.

Jane Packard and her colleagues also consider methodological and analytical issues in their discussion of how artificial intelligence (Al) can be applied to the study of animal behavior. Although there is a great deal of popular and scientific interest in Ai, there is little consensus about its utility in explaining mind and behavior (see for example Dreyfus 1979; Searle 1980, 1990; Minsky 1986; Hamad 1989; Penrose 1989; Churchland & Churchland 1990). Packard et al.'s essay is a good introduction for the uninitiated and a good review for the more sophisticated reader. Starting from the assumption that the four basic components of behavioral processes are perception, action, motivation, and learning, and that a detailed understanding of how stimuli and responses are paired "is central to all analyses of behavior," they argue that Al can be valuable in helping us to study and to understand complex behavioral patterns. They are optimistic that communication among scientists in different disciplines will improve with the application of Al to the analysis of behavior.

Comparative analyses of behavior are central to furthering our understanding of the evolution of behavioral phenotypes but, until recently, rigorous analyses using analytical techniques were restricted to comparative analyses of morphological phenotypes and often suffered from misapplications. Gordon Burghardt & John Gittleman provide a valuable discussion of the historical roots of the comparative study of behavior (see also Richard Burkhardt in section I of this volume) and show how different taxonomic methods stemming from phenetics and cladistics can be used in comparative ethological studies. Burghardt & Gittleman also discuss new methods for distinguishing and quantifying phylogenetic and adaptive effects and "outline how some new comparative methods might provide a rigorous means to incorporate the early Darwinians' dream of a comparative psychology of mental characteristics into modern evolutionary biology."

Donald Kroodsma emphasizes the fundamental role of experiment and the importance of experimental design in his essay on bird-song and the study of dialect variation. As he points out, "An experiment that is inadequately designed is severely limited in what it can tell us about the world." He convincingly argues that many studies of bird-song dialects have not succeeded in testing the hypotheses that they were intended to test. Rigorous concern with methodology may slow or reduce publication but, as Kroodsma rightly concludes, our goal should be to attain knowledge about animal behavior rather than to publish papers (see also Bekoff 1989).

Steven Lima's chapter on vigilance (antipredatory) behavior in birds is also skeptical about some commonly accepted results. Individual birds cannot scan for potential predators and feed at the same time. This has led to the "many eyes" hypothesis which holds that by living in a group an individual can benefit from others who scan while it feeds and feed while it scans. The area of vigilance also is one in which cognitive ethologists can find plenty of room for research. For example, does an individual monitor its behavior based on an awareness of what others have done, are doing, or are likely to do? A few data strongly suggest that it does. If this is the case, we can ask how this information processed.

While some will feel that Lima and Kroodsma overstate the case that perhaps very little is actually known about their areas of expertise, none could deny that the applications of rigorous thought that characterizes their two chapters will be valuable for their own areas of inquiry and also for the field of animal behavior in general. Proper experiments and more data are needed in these and other instances.

The final two chapters in this section also show how methodological and analytical procedures can influence our acquisition of knowledge. Walter Koenig & Ronald Mumme develop this thought in response to a challenge to functional explanations of helping behavior. This challenge suggests that helping "originated and is currently maintained nonadaptively as a result of its tight linkage with the clearly adaptive behaviors associated with normal parental care." On this view, helping is an unselected consequence of what is selected. Although the authors do not consider Sober's (1984) distinction between selection for (the causal concept) and selection of (the effects of a selection process), it is relevant here. The critic of functional explanations of helping behavior is suggesting that there is "selection of" this behavior but not "selection for" it. Koenig & Mumme reply that if four levels of analysis are distinguished (evolutionary origins, functional consequences, ontogenetic processes, and mechanisms [including cognitive and physiological processes]) there are ample data to support the notion that helping has current adaptive value. However many studies fail to control for confounding variables such as the "group-living effect" (perhaps helper effects are merely a consequence of living in groups) and Koenig & Mumme are aware of the fact that more rigorous analyses of data are needed as are more carefully stated hypotheses.

Gail Michener, in her analysis of the behavioral ecology and reproductive biology of (mainly Richardson's) ground squirrels, echos many of the concerns of the other chapters. She notes that, because researchers employ various methods of data collection and analysis, results are often difficult to compare. She shows how "date-based" analyses (which involves pooling data collected on the same date regardless of the reproductive status of the individuals comprising the sample) and "event-based" analyses (which involves pooling data from individuals of the same reproductive status without regard to the date on which the data were collected) produce different results. Thus, when comparing the results of different studies, one must examine in detail how the results were obtained.

The chapters in this section make clear that the study of behavior involves more than "stamp-collecting" (serious philatilists might say the same about "stamp-collecting"). If studies are really to give us information about animal behavior and if the results are to be used for meaningful comparative analyses, then we cannot forsake a detailed concern with method, analysis, and critical experiment in the interest of speedy publication.

LITERATURE CITED

• Barlow, G.B. 1977. Modal action patterns. In: How Animals Communicate (ed. by T.A. Sebeok), pp. 94-125). Bloomington, Indiana: Indiana University Press.
• Bekoff, M. 1977. Quantitative studies of three areas of classical ethology: Social dominance, behavioral taxonomy, and behavioral variability. In: Quantitative Methods in the Study of Animal Behavior (ed. by B.A. Hazlett), pp. 1-46. New York: Academic Press.
• —. 1989. Assessing publication impact. Bioscience 39, 586.
• Churchland, P.M. & Churchland, P.S. 1990. Could a machine think? Scientific American 262, 232-237.
• Dreyfus, H.L. 1979. What Computers Can't Do, 2nd Edition. New York: Harper and Row.
• Hamad, S. 1989. Minds, machines and Searle. Journal of Experimental and Theoretical Artificial Intelligence 1, 5-25.
• Heinroth, O. 1910/1985. Contributions to the biology, especially the ethology and psychology of the Anatidae. In: Foundations of Comparative Ethology (ed. by G. Burghardt, transl. by D. Gove & C.J. Mellor), pp. 246-301. New York: Van Nostrand Reinhold Company Incorporated.
• Lorenz, K.Z. 1981. The Foundations of Ethology. New York: Springer-Verlag.
• Minsky, M. 1986. The Society of Mind. New York: Simon and Schuster.
• Penrose, R. 1989. The Emperor's New Mind: Concerning Computers, Minds, and the Laws of Physics. Oxford: Oxford University Press.
• Searle, J.R. 1980. M...