In lieu of an abstract, here is a brief excerpt of the content:

9 Planning, Memory, and Decision Making Amanda Seed, Nicola Clayton, Peter Carruthers, Anthony Dickinson, Paul W. Glimcher, Onur Güntürkün, Robert R. Hampton, Alex Kacelnik, Murray Shanahan, Jeffrey R. Stevens, and Sabine Tebbich Introduction Animals make a variety of choices, and it is a fair start to assume that the psychological mechanisms underpinning their choices will be adaptive in the sense of maximizing their net gain of resources, reproductive opportunities, predator avoidance and ultimately their fitness.1 Even choices that initially appear simple can be complicated because adaptive decisions often involve tradeoffs along multiple dimensions. All other things being equal, we would expect an animal to delay consumption of an immediately available green apple until it is ripe but not so long that it gets moldy, but all things are rarely equal, and an ideally optimal consumer is only a theoretical abstraction. Implementing ideally optimal choices may be seen as being tuned to how the fruit matures: to current temperature and humidity, to present and future needs, to competition with other consumers, etc. These are the complex trade-offs that cognitive mechanisms implement. In this chapter we are motivated by the following sorts of questions: What sorts of information do animals use to make decisions, and what mechanisms underpin the actions of different animal species in different contexts? Can they act now to secure an outcome with value in the future, even if it has limited value in the present? Can they integrate disparate sources of information, and recognize when they do not have enough information to make a good choice? 1 It is customary to frame biological research around the notion that traits are designed to maximize fitness. This is the approach we adopt because the psychological abilities that are at the center of this contribution are of course biological traits. The many virtues and vices of this approach have been discussed extensively so we will not belabor them here. 122 A. Seed et al. To what extent do these decisions result from deliberation, as opposed to the sculpting effect of consequences, either over the course of evolution or the life history, and how can we tell the difference from studying animal brains and behavior? Although answers to these questions will vary for different species in different contexts, comparative study can help us to extract common evolutionary and psychological principles. Impressive progress toward answering these questions has been made in the past decades within several disciplines, but substantial disagreements about how to interpret findings persist, and there is plenty still to learn. Our aim has been to outline the state of the art, to clarify the points of disagreement, and to suggest future directions. Decision Making: Who Decides, How, and Why? What Is a Decision? Stevens (this volume) broadly defines a decision as the results of an evaluation of possible options. This definition encompasses the study of decision making from several different perspectives, with the fundamental difference between them being who (or what) decides (i.e., evaluates) the options to determine the resulting choice (e.g., natural selection, the mind, neural networks). From an evolutionary perspective, decision making or “choice” simply describes the selection of one among a set of possible targets. Similarly, in economics, the processes by which decisions are reached do not figure among the list of priorities in the study of choices and preferences: preferences and decisions are what the subject does, not what it thinks about it or how it achieves it. The psychological approach, however, is critically concerned with the process by which options are evaluated in the minds of individuals. Determining what counts as a “decision” from a psychological perspective is slightly more controversial. One possible definition of a decision is any process (at the psychological , algorithmic or neural level) that leads to choices and preferences. An alternative definition is that only a subclass of such processes qualifies as decision making. For example, Dickinson argues that if choices result from no additional mechanisms than those generating action when an option is presented without competition, then this process implies choice without decisions . Dickinson uses the sequential choice model (SCM) as an example (Shapiro et al. 2008). In this model, when the agent encounters a source of reward, its actions toward it tend to reflect the value of this source relative to its context. The strongest and most quantifiable measure of this value is the frequency distribution of latencies to respond to the stimulus. More valuable...

Share