Conceptual representations in goal-directed decision making by ProQuest


More Info
									Cognitive, Affective, & Behavioral Neuroscience
2008, 8 (4), 418-428

                                   Conceptual representations in
                                   goal-directed decision making
                                                  Nicholas shea aNd KristiNe Krug
                                                   University of Oxford, Oxford, England

                                                           PhiliPPe N. tobler
                                             University of Cambridge, Cambridge, England

                Emerging evidence suggests that the long-established distinction between habit-based and goal-directed
             decision-making mechanisms can also be sustained in humans. Although the habit-based system has been
             extensively studied in humans, the goal-directed system is less well characterized. This review brings to
             that task the distinction between conceptual and nonconceptual representational mechanisms. Conceptual
             representations are structured out of semantic constituents (concepts)—the use of which requires an ability
             to perform some language-like syntactic processing. Decision making—as investigated by neuroscience and
             psychology—is normally studied in isolation from questions about concepts as studied in philosophy and
             cognitive psychology. We ask what role concepts play in the “goal-directed” decision-making system. We
             argue that one fruitful way of studying this system in humans is to investigate the extent to which it deploys
             conceptual representations.

1. Introduction                                                         inferences sufficient for concept possession, because in-
   Many animals have at least two systems for produc-                   ferences may be made by making transitions in thought
ing instrumental behavior: the habit system and the goal-               between nonconceptual representations. By concept, we
directed system (Dickinson & Balleine, 2002). These                     mean a constituent of a mental representation: In order
multiple action and learning systems are the scope of                   to possess concepts, a thinker must have internal repre-
modern learning theory (Balleine & Dickinson, 1998a),                   sentations with a semantically constituent structure—for
which goes beyond previous models (Dickinson 1980).                     example, the familiar subject–predicate structure with
There is growing evidence that the systems responsible for              which some individual is picked out and some property
human instrumental behavior and decision making divide                  predicated of that individual. (For more detail, see Sec-
along similar lines (Gottfried, O’Doherty, & Dolan, 2003;               tion 3 below.) The commonsensical idea that people often
Valentin, Dickinson, & O’Doherty, 2007). However, it is                 reason with concepts has been vindicated and refined by a
often found that in many domains, human psychology is                   long tradition of work in both cognitive psychology (Mur-
more sophisticated than that of other animals. This may                 phy, 2004) and philosophy (Evans, 1982; Millikan, 2000;
also be true of the mechanisms of decision making. In                   Peacocke, 1992). This review applies some of those in-
particular, the human capacity for parsing and produc-                  sights, asking what role concepts have in generating goal-
ing the complex syntactic structures of language may well               directed human behavior.
have an impact on how we formulate plans and decide                        The article is structured as follows. We start in Sec-
between available options. More specifically, language                  tion 2 by characterizing the difference between the habit-
mechanisms allow humans to have and think with con-                     based decision-making system and the goal-directed
cepts. This is not to say that something like human lan-                decision-making system. In Section 3, we introduce the
guage is necessary for concept possession but, rather,                  distinction between conceptual and nonconceptual rep-
that thinking with internal representations that mirror the             resentations, as clarified by work in philosophy. In Sec-
constituent structure of natural language is sufficient for             tion 4, we ask whether the goal-directed decision-making
concept possession. The term concept is used for a variety              system in nonhuman animals makes use of conceptual
of phenomena. Some authors use it for the capacity to re-               representations. Finally, in Section 5, we suggest how that
spond in the same way to a variety of different stimuli—an              particular question can be addressed in humans, arguing
ability that, as we use the term, could be mediated by a                that there is preliminary evidence that the operation of
nonconceptual representation. Nor is the capacity to make               the dedicated decision-making system in humans that is

                                                  N. Shea,

Copyright 2008 Psychonomic Society, Inc.                            418
                                                                                   ConCepts in DeCision Making           419

homologous to the goal-directed system studied in other          erable experience of the failure to obtain reward, enabling
animals does—in humans at least—make use of concep-              the system to adjust its action values and thus its preferred
tual representations.                                            course of action. Encoding of the action–outcome contin-
                                                                 gency, as evidenced by immediate and high sensitivity to
2. Habit-Based Versus Goal-Directed                              reward devaluation, is the signature of the goal-directed
Decision Making                                                  system (Dickinson & Balleine, 2002).
   There is substantial evidence that in many animals, there        A distinction between the goal-directed system and
are two independent mechanisms for selecting actions.            the habit system is most frequently drawn empirically, in
Modeling work explains how we might benefit from hav-            terms of encoding action–outcome contingency (Dickin-
ing two relatively independent decision-making mecha-            son, 1980), but various other contrast
To top