Intraspecific communication is of crucial importance for survival and reproduction in the majority of animal species. Much of this communication takes place by means of ‘displays’, conspicuous, stereotyped and species-specific postures, movements and vocalizations that are specifically adapted to serve as a signal to another member of the species (cf. Tinbergen, 1952). Because of these characteristics, social displays provide interesting material for the study of behavioural development. First, displays are especially suitable for the study of the development of complex stereotyped motor patterns that may be influenced by social experience. Second, the study of display development can provide a better understanding of the ontogeny of social behaviour in general and of the immediate causation of adult social behaviour (Kruijt, 1964). Third, displays are believed to be derived in the course of evolution from intention movements (such as to flee, to attack, to preen) as a result of ritualization and emancipation (Tinbergen, 1952; see below). Because phylogeny is modified ontogeny, changes in ontogeny may reflect changes that have occurred in evolution. Consequently, the study of the ontogeny of displays may also provide insight into the evolution of displays.

Despite these interesting properties of displays, their ontogeny has hardly been analysed quantitatively. A notable exception is the development of bird song. Because of its relation to learning, the discovery of the neural systems involved in song, and the possibility of manipulating feedback by deafening the bird, the study of bird song has become one of the most flourishing fields in ethology.

Young organisms cannot do many of the things that adults can, and some of these things are cognitive in nature. Very young humans cannot understand or produce speech very well. Young animals cannot recognize predators or collect food at the adult level of proficiency. Young humans, however, develop the ability to understand and produce speech at a rate that is scarcely believable, and young animals can learn to forage and avoid predators, as well as form social attachments, learn to communicate, and learn to orient themselves in the world, sometimes by means that are not available to adults. This chapter offers a highly selective review of cognitive development in animals, concentrating on a few questions. The two fundamental questions are, do animals have ‘cognitions’, and do they change ontogenetically? The next question concerns the nature of cognitive development. Cognitive development clearly involves a multitude of changes in behaviour and the nervous system. This discussion will emphasize that there is an important difference between the acquisition of information by naive animals and developmental change in cognitive mechanisms. Several examples, drawn from foraging, aversion learning, and spatial behaviour illustrate this distinction, and allow for some discussion in passing of developmental changes in the brain that are correlated with cognitive development.

I begin this chapter with a brief discussion of what I mean by a behavior system, and then use the dustbathing, hunger, aggression, and sex systems of chickens to illustrate how such systems develop. Using these examples plus comparable information from investigations of mammalian behavior, I next consider the questions of whether there are any general differences between the development of perceptual and motor mechanisms and between social and non-social behavior systems. In the context of social behavior systems, I review some ways in which early experience can have far-reaching effects. Finally, I look at the development of interactions among behavior systems, and ask whether any new principles are necessary to understand this complex process.

Behavior systems

A definition of a behavior system and its components is given in Chapter 1, and a depiction of the concept is presented there in Figure 1.1 (p. 6). A more extensive discussion of this concept and many of the other issues considered in this chapter can be found in Hogan (1988). In brief, a behavior system consists of an organization of its components: perceptual, motor, and central behavior mechanisms. Each of these components is also organized. The study of development comprises: (1) describing the changes in both the organization of the components themselves and the organization of the system as whole (i.e. the connections among the behavior mechanisms), and (2) investigating the causes of those changes.

Avian song has received extensive attention, both as a natural behavior that is easily studied and can be related to ecology or to principles of natural selection, and as a neuroethological preparation in which it is possible to determine the neural basis for a complex motor activity in a vertebrate. This chapter is a survey of some of the central findings in both domains. It concentrates on the development of song and of brain regions that are responsible for song. It attempts to relate the behavioral and neurobiological findings in this system to more general issues of the nature of early perceptual and motor development and juvenile learning. In addition, it attempts to extend ideas on behavioral development to the neurobiological substrate for this behavior.

Overview of singing behavior

Acquisition of song – the perceptual phase

Several key findings are central to understanding avian song acquisition and performance. First, many features of song are learned. Thus, for example, many of the sounds comprising a canary's song closely resemble songs heard by the bird as a juvenile, and are distinct from sounds produced by other canaries (Marler & Waser, 1977, Waser & Marler, 1977). Zebra finches form a song by splicing elements of the songs of several individuals that they heard as juveniles, apparently favoring adults which had fed or interacted with them (Williams, 1990a). Similarly, nightingales form an elaborate song repertoire by acquiring and retaining ‘packages’ of sounds as a juvenile (Hultsch & Todt, 1989b), and swamp sparrows form a song by selecting from songs heard when young (Marler & Peters, 1988a).

I still recall a pronouncement from my earliest days of graduate training in linguistics that ‘phonemes are the building blocks of language’. I have forgotten who said it – possibly a number of linguists issued similar pronouncements and I am only recollecting a composite. In any case, the statement is believed by many individuals who study language. In this chapter I shall show that the phoneme is not the elemental unit best suited for understanding the development of spoken language.

The term phoneme refers to the smallest difference that can differentiate two spoken words. In English, /p/ and /t/ are phonemes because they distinguish ‘pie’ from ‘tie’, /s/ and /z/ are phonemes because they distinguish ‘hiss’ from ‘his’. It makes intuitive sense to suppose that these sound units are in fact the building blocks of language because words and grammatical markers are made of such sounds, and phrases and sentences are made of words. But this logic serves us poorly if our intent is to understand what human developments produce linguistic capacity and determine the form of linguistic behaviors. These are ontogenetic questions, and in asking them our concern cannot reside with elemental units of a behavior not yet acquired. Rather, we must concentrate on developmental mechanisms which facilitate or enable behaviors that – as the human child ultimately discovers – are decomposable into those units. From a developmental perspective, then, the phoneme is unavoidably a posteriori and therefore incapable of building any of the child's earlier behaviors.

There is now a considerable literature concerning the phenomenon known as sexual imprinting and the mechanisms underlying it (for reviews see Bateson, 1966; Immelmann and Suomi, 1981; Kruijt, 1985), However, recent findings by Immelmann, Lassek, Pröve & Bischof (1991) and by Kruijt & Meeuwissen (1991) suggest that earlier concepts of imprintinglike learning have to be revised. In this chapter I will analyse this new evidence and discuss its implications for some of the presumed characteristics of imprinting, such as the existence of a sensitive period and the stability of preferences. Further, I will consider some important questions such as stimulus selection and the reasons for stability of preferences. Many of the ideas I will present here are speculative and have little experimental backing. However, they may help us discard some of the old ideas concerning imprinting and so allow for the generation of new ones.

I will start with a brief description of the findings which prompted this chapter. Then I will propose an interpretation of these findings in terms of a two-stage process. The period in early development where information about the appearance of the parents is stored is called ‘acquisition phase’ here. Subsequently, there is a ‘consolidation process’ which takes place when the animal becomes sexually mature. In the final section of this chapter, I summarize the main features of the two-stage process and try to evaluate how the ideas presented here can be generalized to other learning paradigms.

This chapter is concerned with mental representations, and their development by learning via association. By representation, I mean to imply structures that permit recognition and identification of a stimulus without necessarily entering the conceptual domain. I hope to show that associative mechanisms can construct and employ representations of this type in such a way as to permit explanation of phenomena such as latent inhibition, perceptual learning, and the trade-off between context-specificity and generalisation. To this end, an outline model of associative learning, that concerns itself with the formation of associations between elements representing motivationally neutral stimuli, is developed in the course of the chapter. The emphasis, however, will be on the representational assumptions contained within the model and their consequences. The chapter starts by examining an elemental approach to stimulus representation, which proves surprisingly powerful in an associative context. Nevertheless, there are drawbacks to an elemental approach, which the latter half of the chapter attempts to solve by invoking a configural modification of an elemental account.

Introduction

The modern concept of an association, and hence of an associative learning system, probably stems from the work of the British empiricists (e.g. Locke, Hume). At the heart of this approach is the essential idea that one stimulus can bring about some recollection of another by virtue of the fact that the two stimuli were associated at some time in the past.