A fundamental requirement of the comparative method is that the features compared across a sample should be homologous. Robson-Brown (Chapter 2) provides an overview of attempts to define homology and their application in socioecology. Some definitions require that a common feature in two species must derive from a common ancestor in order to be homologous ('taxic' definitions), whereas others require only that features are indistinguishable ('operational' definitions). For the types of analyses contained in this book, the former definition is not only extremely restrictive, in that the necessary work on the detailed development, structure and evolutionary history of the features concerned has rarely been done, but it would also render most of the analyses invalid. Many working in comparative biology agree that only when the presence of similar features in different taxa results from separate evolutionary events can these taxa provide independent data points for the testing of informative correlation or association (see Chapter 3). Clearly, this requirement is mutually exclusive with the more restrictive definition of homology, and acceptance of both would rule out the use of the comparative method. Indeed, the separate evolution of closely similar features in several species or taxa can provide a useful basis for the deduction of the adaptive or functional cause of their evolution, as many examples in this book illustrate. This is a version of the Method of Agreement.
Van Schaik et al. (Chapter 8) present data showing that male infanticide has not been observed in primate species in which communal infant care is well developed. The forms of communal care involved are not identical in all the sample species; for example, the caretakers other than the mother vary. Also, this type ofbehaviour must have evolved more than once, given its distribution across the primate phylogeny. Therefore, 'communal care' does not meet the conditions for the stricter, 'taxic' definition of homology. However, the basis of the use of the comparative method in van Schaik et al.'s study is that similar behaviour evolves for similar functional reasons, and it is not necessary for either factor to be absolutely identical across the sample for the results to enhance evolutionary understanding. The Method of Agreement is a flexible tool that requires reasonable, but not rigid, application.
Nevertheless, care must be taken in deciding whether or not such co-categorisation of non-identical features is reasonable. Kappeler (Chapter 10) provides an extensive discussion of whether lemur social structures are similar enough to those of anthropoids for their comparison to provide a useful test of the generality of theories originally developed for anthropoids. He examines whether lemur species can be divided into similar categories to anthropoids on the basis of variation in four major aspects of female-female relationships: philopatry, nepotism, tolerance and despotism. Female-female relationships in lemurs, he concludes, display a number of different features, and different combinations of features from those of anthropoids. Reasons for this lack of comparability are suggested, including fundamental grade differences between the groups such as the lesser visual acuity of lemurs. This is clearly an example of too many concurrent dissimilarities alongside other similarities, rendering the comparative method unusable, and further theoretical groundwork is needed first. In other words, the conditions are not met for the Method of Agreement, the Method of Differences, or the Joint Method.
Was this article helpful?