A central and largely unsolved problem in the brain sciences is to understand the functional architecture of the vertebrate nervous system. Many questions about this architecture revolve around the issue of action selection. Because it is a fundamental property of neurons to be selective with regard to the patterns of input activity to which they respond, claims that particular brain subsystems are specifically or preferentially involved in the selection of action, as distinct to other aspects of control, must meet more stringent requirements (see below). It is also by no means inevitable that the functional decomposition of the brain will contain specialist action-selection mechanisms (see Seth, this volume). Appropriate behavioural switching could be a global property of nervous system function, and of its embedding in a body and environment, that cannot be attributed to specific subcomponents of brain architecture. In other words, it is plausible that an animal may ‘flip’ from one integrated pattern of behavioural output to another without some identifiable internal ‘switch’ being thrown. On the other hand, theoretical arguments can be presented, based for instance on the benefits that accrue from modularity (Bryson, 2005; Prescott et al., 1999; Wagner and Altenberg, 1996), to suggest that biological control systems may include specialised action-selection components. Hence, one important debate in this field is whether there are specialised mechanisms for action selection in animal nervous systems, and, if so, where these might be found (see also Prescott, 2007, for an evolutionary perspective on this question).

Neural substrates for action selection in cortico-basal ganglia loops

Redgrave et al. (1999) have proposed that, to be considered as a candidate action-selection mechanism, a neural subsystem should exhibit the following properties. First, it should have inputs that carry information about both internal and external cues relevant to decision making. Second, there should be some internal mechanism that allows calculation of the urgency or ‘salience’ that should be attached to each available action. Third, there should be mechanisms that allow for the resolution of conflicts between competing actions based on their relative salience. Finally, the outputs of the system should be configured so as to allow the expression of winning actions whilst disallowing losers. There is now a growing consensus in the neuroscience literature that the basal ganglia – a group of functionally related structures found in the mid- and forebrain of all vertebrates – meet these criteria and therefore may represent an important neural action-selection substrate.

In nature, action selection is rarely purely an individual matter; rather, adaptive action selection usually involves a social context. As Seth demonstrates in Part I, the apparently irrational behaviour of an individual viewed in isolation can, in fact, be optimal when considered in a semi-social context that includes competing conspecifics. Importantly, the agents in Seth's model express no explicitly social behaviour – there are no direct costs or benefits associated with social interactions. Rather, agents in the model interact indirectly via their effects on resource distribution in the environment. This demonstrates the ubiquity of social phenomena in nature, which of course will, in general, lead to selective pressure for social adaptations.

The final section of our volume is dedicated to studies of action selection in an explicitly social context. All of the models here do include social actions for individuals to select among. Once we begin to consider societies, however, it also becomes apparent that we can study action selection at the level of the society itself, and how such selection emerges from individual behaviour. In this introduction, we discuss the evolutionary and theoretical background underlying social action selection whilst briefly highlighting some of the contributions of subsequent chapters regarding current controversies in this field.

Action selection is the task of deciding ‘what to do next’. As a general problem facing all autonomous entities – animals and artificial agents – it has exercised both the sciences concerned with understanding the biological bases of behaviour (e.g., ethology, neurobiology, psychology) and those concerned with building artefacts (e.g., artificial intelligence, artificial life, and robotics). The problem has two parts: what constitutes an action, and how are actions selected?

This volume is dedicated to advancing our understanding of the behavioural patterns and neural substrates that support action selection in animals, including humans. Its chapters investigate a wide range of issues, including (1) whether biological action selection is optimal, and if so what is optimised; (2) the neural substrates for action selection in the vertebrate brain; (3) the role of perceptual selection in decision making, and (4) the interaction between group and individual decision making. The mechanisms of action selection considered in these contexts include abstract neural circuits (e.g., Bogacz et al., this volume) through specific brain systems (e.g., Stafford and Gurney, this volume) to policy choices exercised by political parties (Laver et al., this volume.) Taken together, this research has broad implications across the natural, social, medical, and computing sciences.

Parts of this book were originally published as an issue of the Philosophical Transactions of the Royal Society B: Biological Sciences (Volume 362; Issue 1485) but have been modified and updated. Anil Seth's contribution to the book was supported by EPSRC Leadership Fellowship EP/G007543/1 and by a donation from the Dr Mortimer and Theresa Sackler Foundation. Tony Prescott's contribution was supported by the EU 7th Framework Programme via the projects BIOTACT (ICT-215910) and EFAA (ICT-270490). Joanna Bryson's effort was funded in part by the US Air Force Office of Scientific Research, Air Force Material Command, USAF, under grant number FA8655-10-1-3050.

When an animal does one thing rather than another, it is natural to ask ‘why?’ A common explanation is that the action is optimal with respect to some goal. For example, when observing the foraging behaviour of a shorebird, one may ask whether the intake of food is being maximised. This ‘normative’ view, a direct extension of Darwinian principles, has its more recent roots in behavioural ecology (Krebs and Davies, 1997) and optimal foraging theory (Stephens and Krebs, 1986). Adopting a normative perspective on action selection can be very useful in placing constraints on possible underlying mechanisms, for developing and comparing theoretical frameworks relating behaviour to mechanism, and for explaining instances of apparently irrational or suboptimal behaviour. The seven chapters within this section present new insights and modelling results relevant to each of these issues. They also connect with the other parts of this book in important ways. The constraints on underlying mechanisms are thoroughly explored by the computational neuroscience models described in Part II, and patterns of both rational and irrational social behaviour are encountered in a variety of forms in Part III.

Suboptimality and ‘matching’

Assessing animal behaviour from a normative perspective has particular value when observations deviate from predictions, because the scientist is now forced to consider the origin of the apparently suboptimal – or ‘irrational’ – behaviour. In one important example, many animals behave according to Herrnstein's (1961) ‘matching law’, in which responses are allocated in proportion to the reward obtained from each response. However, as both Houston et al. (this volume) and Seth (this volume) note, matching is not always optimal. One response to this observation is to propose that suboptimal matching arises as a side-effect of some underlying principle of behaviour, such as Thorndike's ‘law of effect’, which proposes that behaviour shifts towards alternatives that have higher immediate value (Thorndike, 1911). Another is given by the notion of ecological rationality – that cognitive mechanisms fit the demands of particular ecological niches and may deliver predictably suboptimal behaviour when operating outside these niches (Gigerenzer et al., 1999). In line with ecological rationality, Seth (this volume) shows that simple decision rules that lead to optimal foraging in competitive environments with multiple foragers also lead individual foragers to obey the matching law. The remainder of Seth's chapter describes a novel methodological approach that combines agent-based modelling and optimal/normative approaches via the use of genetic algorithms. The resulting approach, which is given the label ‘optimal agent-based modelling’, achieves the important task of integrating function and mechanism in explanations of action selection (McNamara and Houston, 2009).