INTENTIONAL CONSTRAINTS ON PERCEIVING AND ACTING:
AN ADVANCED WORKSHOP
UNIVERSITY OF PORTSMOUTH.
SATURDAY-SUNDAY, 5-6TH OCTOBER, 1996


Intentionality is regarded as a distinctive feature of humans and animals: our activities are 'about' a distal object (e.g., a goal, a plan, or a problem to be solved). Cognitive psychology attempts to address this issue by assuming representation. The issue of intentionality is also implicit yet fundamental in Gibson's work. Indeed, ecological psychologists have subsequently been seeking non-representational solutions for their paradigmatic problem: the perceptual control of action in an intentional context.

Robert Shaw's sabbatical visit to Portsmouth provided an excellent reason for organizing an initial meeting to review and explore some of the central issues concerning the role of intentionality in ecological psychology. The issues include:

The Portsmouth meeting was intended to the scene for further meetings. Two further symposia on intentionality are already programmed for the ICPA conference in Toronto in July 1997. Our plan is to gather together contributions to this series of meetings, and related material, for publication.

Organizers: Endre Kadar (kadare@psy.port.ac.uk), Paul Morris (morrisp@psy.port.ac.uk) Vasu Reddy (reddyv@psy.port.ac.uk) & Alan Costall (costalla@psy.port.ac.uk).

ABSTRACTS:
DIMENSIONS OF THE PROBLEM OF INTENTIONALITY

Endre Kadar
Department of Psychology
King Henry Building
University of Portsmouth, PO1 2DY
kadare@psyc.port.ac.uk

The problem of intentionality, explicitly or implicitly, has been an important focus of philosophical and scientific debate from the earliest times. The aim of this paper is to make fundamental aspects of intentionality explicit and provide a basis for a constructive discussion by outlining a theoretical space of a few dimensions: ontological, epistemological, axiological, pragmatic, and methodological. Applying this framework, similarities and differences of various theoretical approaches can clearly be demonstrated. In particular, different interpretations of ecological psychology can be used to show the value of the proposed dimensions.

ILL-POSED VERSUS WELL-POSED QUESTIONS ABOUT INTENTIONAL BEHAVIOUR

Bob Shaw
Center for the Ecological Study of Perception and Action
Department of Psychology
Box U-20, University of Connecticut
Storrs, CT 06269-1020, USA
reshaw@uconnvm.uconn.edu

Three fundamentally different theoretical approaches to the problem of goal-directed systems arise from three different ways of posing the question of how goals relate to causes. Mechanists ask how systems that appear to be goal-determined can be shown to be causally determined? This question is ill-posed if, as teleologists claim, goals impose a kind of 'backward causation' on systems. Teleologists ask how the goals of systems can select the causes needed to achieve those goals? But this question is likewise ill-posed if, as the mechanists claim, a future final state cannot act backwards in time to determine efficient causes. A compromise position asks how a goal may act as a constraint on the current state (memory or perception) of the system from initial state to final state? This question, too, remains ill-posed so long as it fails to ask how memory or perception, presumed to be informational, can specify the forces required to move the body mass in the stipulated goal-directed manner. Intentional dynamics, a method from ecological psychology, attempts to make this third question well-posed and hence capable of scientific investigation.

WHAT DO INFANTS WANT TO DO?

Julie C. Rutkowska
Division of Psychology
University of Sussex
Falmer
Brighton BN1 9QN
julier@cogs.susx.ac.uk

My contribution views 'intention' as an example of traditional between-the-ears re-presentational intentionality, and adopts the revisionist stance that 'intention', like other intentional terms, is an unsatisfactory primitive for psychological explanation. Issues of 'when', 'whether' or 'what' intention is present are best replaced by asking: What mechanisms underlie the generation of recurrent activities and repeated attainment of (more or less) stable states of the S--E system that observers identify with behaviour patterns and goals/outcomes?

I shall focus on some implications for alternative, action-based accounts of abilities of different ways of talking about constraints between cognitive and physical-motor processes, using the development of infant prehension as a linking example. Classical computational concepts have more to offer our understanding of embodied, embedded systems than their traditionally centralized focus suggests. However, they may be better (if not ideal) for describing stable organizations than for getting at the processes through which they are attained. Unless supplemented by mechanism accounts that are more amenable to psychologically meaningful decomposition, notions of affordance--effectivity mappings may prove too molar to clarify the structure and development of action, and serve to reinforce an unwelcome form of subject--environment dualism. Exploring CNS--body--environment systems through dynamical systems concepts may hold most promise, but key developmental applications are currently hindered by invoking traditional goal-like notions to explain how action is established and controlled. Whether/how this might be avoided poses a key future question.

INTENTIONALITY IN (EN)ACTION.

Michael Wheeler
Christ Church
OXFORD, OX1 1DP

and
Department of Experimental Psychology
South Parks Road
OXFORD, OX1 3UD
michael.wheeler@psy.ox.ac.uk

In mainstream cognitive science and analytic philosophy of mind, intentionality is standardly defined as the property of 'being about the world' that is possessed by representational mental states. This notion of intentionality depends on a view according to which perceptual and cognitive activity are accurately characterized by a Cartesian subject- object dichotomy. In this talk, I shall articulate an opposing and more promising view, according to which the subject-object dichotomy is no longer thought of as a fundamental feature of (most) perceptual or cognitive activity. The view on offer is a development of ideas to be found in contemporary Continental Philosophy and the enactive approach to cognitive science (as proposed by, among others, Francisco Varela), and is in harmony with recent work in situated artificial intelligence. I shall argue that such a view deserves the attention of ecological psychologists.

PIGS IN INTERACTION: DIMENSIONS OF EXPRESSION AND THE EXPLANATION OF BEHAVIOUR.

Francoise Wemelsfelder
Genetics and Behavioural Sciences Department
Scottish Agricultural College
Bush Estate
Penicuik EH 26 0QE, Scotland
esa032@ed.sac.ac.uk

In my work on the measurement of subjective well-being in growing pigs, I have chosen an interactive paradigm as a starting-point. In interaction with a human experimenter, pigs appear as highly expressive and inquisitive animals. A wide variety of subtly different descriptive categories is required to do justice to the perceived richness of the expressive repertoire. The nature of these categories does not seem to warrant a mechanistic explanation of behavioural expression. On the basis of these findings, I will raise some points for further discussion on the notions of subjectivity, intentionality and affordance.

THE "OBSERVABLE SUBJECT": MAKING SENSE OF ANIMAL BEHAVIOUR

Paul Morris
Department of Psychology
King Henry Building
University of Portsmouth, PO1 2DY
morrisp@psyc.port.ac.uk

The use of anthropomorphic or any kind of psychologically rich description of animal behaviour has long been regarded within mainstream psychology as unscientific. It is alleged that to be scientific we must use neutral, mechanistic accounts of animal activity. However, there are examples within academic psychology of the use of psychologically meaningful description. Hebb (1946) reports that he and his colleagues, despite attempts to use mechanistic categories to describe behaviour, could only arrive at useful, meaningful descriptions of the animals in their laboratory by the use of rich psychological terminology. It is significant that these rich descriptions of personality and temperament were of benefit to newcomers in the laboratory in their interactions with the laboratory animals. For these descriptions to be of utility some consistency and pattern to the description would have been required. Hebb is tantalisingly brief and anecdotal in his accounts of how he and his colleagues described the behaviour of the laboratory animals. In our studies of interactions between animals and people we have systematically analysed peoples' descriptions of animal-human interactions. We have found great consistency in the terms used by observers in the descriptions of behavioural episodes. We have also examined the temporal patterns within the descriptions and have found that not only do people use similar terminology, but perceive psychological events and the order of events within episodes highly consistently. In our current study we are using a technique developed by Newtson (1973) in social psychology to give exact timings to the beginnings and ends of psychologically meaningful events within episodes.

Hebb, D.O. (1946). Emotion in man and animal: An analysis of the intuitive processes of recognition. Psychological Review, 53, 88-106.

Newtson, D. (1973). Attribution and the unit of perception of ongoing behavior. Journal of Personality and Social Psychology, 28, 28-38.

THE VARIABLE STRUCTURE OF INTENTIONAL BEHAVIOUR.

Michael Weir
Computer Science Dept
North Waugh
St. Andrew's University
Fife, KY16 9SS
mkw@dcs.st-and.ac.uk

Examples of accidental and intentional behaviour are given which contain the same mental goal state and associated reasoning. It is argued that representing the goal's causal element solely as a current state of the mind or brain fails to preserve the distinction between the two types of behaviour. The motivation for this representation is attributed to the doctrine of efficient causation. The fixed structure inherent in efficient causation is described through a General systems model. By contrast, the distinctive characteristics of Intentional Behaviour is taken to be that it shows directed variation in its structure or causal strategy while pursuing a goal. A theory of flexible action encapsulating such variation is given in the form of a systems methodology and exemplified.

DISCUSSANT'S COMMENTS

Dankert Vedeler
Department of Psychology
University of Trondheim
N-7034 Trondheim
Norway
Dankert@sv.ntnu.no

How should prospective control be conceived? And, how should the concept of goal be understood? These two questions are connected because they both address the issue of intentionality and the future. As I see it, future directedness implies a "discretization" of time, that is, a distinction between the present and the future. According to such a distinction, future directed intentionality requires representations (of goals) that make the future present. Representational intentionality is derived, and in order to be able to consider more basic forms of intentionality as future directed it is necessary to overcome the discretization of time, in a way similar to Kadar & Effken's (1994) proposal for overcoming the discretization of space.

This issue leads to the question of the development of intentionality. Michael Wheeler's reference to Heidegger's proposal that most human behaviour, and certainly all animal behaviour, is not intentional might be reformulated in terms of emergent intentions instead of preconceived intentions, thus avoiding representations of goals and the partitioning of time into present and future which the latter presupposes. Such a reformulation, however, raises, in turn, the problem of emergence.

Kadar, E., & Effken, J. (1994). Heideggerian meditations on an alternative ontology for ecological psychology: A response to Turvey's (1992) proposal. Ecological Psychology, 6, 297-341.

ISEP HOME PAGE