Published in

Behavioral and Brain Sciences 16, 1993, 39-41


Intentionality, mind and folk psychology

Winand H Dittrich & Stephen E G Lea

Department of Psychology, University of Exeter, Washington Singer Labs, Exeter EX4 4QG, England.

Intentionality and theory of mind: Gopnik tries to demonstrate that empirical evidence from developmental psychology supports the view of a "theory theory" in which common sense beliefs are constructed to explain ourselves and others. Often it is suggested that the intentional stance is primarily a first-person experience subsequently generalized to other minds. Contrary to this view Gopnik asserts that children construct the idea of intentionality in order to understand their own experiences and the behaviour and language of others. Although there is disagreement among developmental psychologists when exactly the cognitive construct is truly mastered, there is general agreement that a dramatic shift in childrens' cognitive ability, in particular in its usage of representations, occurs at the age of about four years. However, a necessary differentiation seems to be neglected by Gopnik. In agreement with other researchers (e.g. Shields 1978, Trevarthen & Humbley 1978, Bruner 1977) it can be assumed that there are fundamental differences between person knowledge and object knowledge even though both types of knowledge are dependent on the child's behaviour. The differences concern the way of knowledge acquisition and its content. Person knowledge is mainly acquired through a dialogue or communicative process and consists of the knowledge of the self and others as psychological agents with intentions, beliefs, emotions and the ability to communicate. In contrast, object knowledge is acquired through acting upon an object and consists of a kind of 'naive' physics. In presenting evidence for his assumption it is misleading when Gopnik gives empirical findings from both domains without mentioning the different informational processes involved.

Furthermore, focusing the informational processing routes possibly involved we would like to argue that his main thesis that the idea of intentionality as a cognitive construct is the most plausible interpretation of the empirical findings lacks support at least for two reasons: one methodological and one structural.

Methodologically, the idea of intentionality is given on the one side the status of a theoretical construct in order to understand itself and others. That is the core idea of the 'theory theory' concept. However, it is impossible to proof a hypothetical construct empirically but furthermore it remains unclear what the possible criterions are 'how the idea of intentionality is developing in childrens' behaviour'. Thus, Gopnik's attempt on the other side to answer 'the question whether our beliefs about intentionality are theoretical, enculturated or derive from a more direct first-person apprehension' on a purely empirical basis seems highly problematic from the beginning on.

Considering the structural basis of the findings Gopnik is presenting, the following analogy seems helpful. Even if, to give an example, person permanence and object permanence (e.g. Jackson, Campos & Fischer 1978) seem to develop more or less with one another, this synchronous development does not necessarily imply, as Gopnik is assuming in the case of self experience and that of others, that both processes are based on the same underlying process. Gopnik is assuming that the cognitive construct of intentionality is this common underlying process. First of all, the pure parallelness, as in the case of permanence perception, is no real evidence as it lacks the explanatory power and secondly, contrary to Gopnik's constructive theory of mind interpretation, the empirical findings presented can be interpreted in quite a different way, namely through the internal relationship of the processes. We would like to present two alternative interpretations which focus on different aspects. The first interpretation is based on first-order knowledge or what Gopnik called Searle's view or the simulation view. Certainly, it can be argued that the development of an adequate knowledge of its self requeires a specific representational system which has only developed at the age of 3-4 years and can than be used to understand the other mind problem. Thus arguing that what Gopnik described as theory theory construct is nothing else than the application or generalization of first-person knowledge in the domain of third-person knowledge or as Rene Spitz (1957) puts it: "The beginning awareness of the self is predicated on the awareness of the 'other'" (p.130). In terms of the simulation view it is the child's attempt to simulate the other's mind by using the world of its self. The identical relationship between the development of an understanding of self and other can be understood as a result of the developmental process in which an understanding of self is used to inform an understanding of others, and vice versa. How children think of the self and how they conceptualize their experiences are intimately related issues. Whether self-understanding and other-understanding develop synchronously is indeed an empirical question. These possible views were falsely rejected because Gopnik has a concept of intentionality in which intentionality is not the result of a psychological state but the result of the interpretation ( 'cognitive construct' of intentionality) of a psychological state based on the experience to this state. Gopnik's view (sect.8) seems to be highly similar to the so called James/Lange theory of emotion in which emotional states are interpreted as a result of their experienced consequences or feelings - the same is true for Goldman's second functionalist model. Before discussing the concept of intentionality in more detail we are pointing at a second interpretation, which is not in contrast to the first one rather focusing on a different aspect. In this interpretation the shift in children's behaviour which is correctly highlighted by Gopnik can be interpreted as synchronization between their representational abilities and their ability to express themselves adequately. Before this age period there is a so called "developmental dilemma" (Heckhausen 1983) in which representational ability and acting or communicating seems discrepant. This would explain why children could solve some task, which are mainly based on perceptual competence and not others which involve some competence in performance. It would also explain why representations of the self can be shown at the age of about two years, i.e. one and a half year before this behavioural shift, but the generalization to represent others seems not yet possible because the appropiate emotional basis is not fully developed. Thus the development of the representation system of the self and others is much more related as Gopnik assumed. In any case the two interpretations seem to strenghten the point that the question what theory of mind children develop seems less important than the investigation of the nature of the cognitive processes involved. Otherwise Gopnik's approach cannot answer the question about the nature of the children's representations. What is the evidence that children and adults have the same understanding of mental states or linguistic terms about psychological states even if they use the same words? We would like to argue that this problem can only be solved if the basis of our intentionality, namely the structure of objects or actors themselves, is taken into consideration. Most important, however, it should be noted that in extension of Gopnik's list of evidence which is mainly based on examples about cognitive development the emerging sense of self as intentional agent is also seen in infant's competence motivation as precursor of achievement motivation (Heckhausen 1982) or the infant's ability of self-regulation (Kopp 1982). It is assumed here that in addition to the pure intellectual aspects information processing within the motivational, emotional and communicative domains seem equally important for the development of conceptual knowledge about the physical and the mental world of infants.

Intentionality and knowledge: In a recent study (Dittrich & Lea 1992) we developed a novel task in order to investigate the perception of intentionality and found among others that the intentional content of a scene was mainly effected by the total number and the directedness of its elements. Evidence was presented that the identification of intentionality or meaning is a process of cognition but grounded in perceptual input characteristics. It is based upon the comparison between two sources of information - a visually offered one, and a stored one. If the process of comparison of the object itself and its conceptualized representation yields equivalence, semantic recognition can take place. Just this is the essence of comprehension of intentionality and it is evident that specific cognitive operations (e.g. 'hybrid representations') have to be developed or have to be available before comprehension can take place. One of these 'hybrid' operations has to involve the relationship of the self to others or objects in the world. Therefore, it seems pointless ,also, to look for a special definition or construct of intentional information. It receives its specific nature through the verification of perceptive and stored information as signs for something, but nor through the direct pick-up of intention from the stimulus, nor through a specific nature of the cognitive process, e.g. in the comprehension of the specific meaning of signals or as a specific 'idea of intentionality'. Concerning the mechanisms of the perceptual processes, we assume that humans can use cues or clues to interpret the visual scene in building semantic (de dicto) representations. Perception of intentionality is possible because of the specific integration of different features and therefore in order to understand the role of intentionality our approach focused on the analysis of the cognitive processes which are involved. It is argued that the traditional linear sequential model of object representation with a linguistic stage at the end should be extended towards a model of parallel processing of different visual aspects. In any case the key feature ot these operations are concepts, and in this respect we agree with Goldman that folk psychology can be related to the psychology of concepts, in particular under developmental aspects.

Intentionality and mind: Goldman raises an important question when he is asking how people ascribe mental states to themselves. We agree with Goldman and Gopnik in their attempt to integrate psychological facts given only phenomenologically, which are strongly related to our everyday life and thus have ecological validity, into scientific psychological theories. However, we disagree with the role which is attributed to the phenomenology of the psyche. In contrast to the two authors, which try to establish the primacy of psychological experience compared to psychological states in order to understand our cognitive processes - Gopnik through the developmental history of the theory theory and Goldman supports a prominent role of phenomenology in our mental concepts - we suppose that the role of phenomenology is a pure descriptive one in order to provide a body of psychological realistic facts for which psychology may give scientific explanations. Certainly, in some cases scientific psychology and common sense may be congruent. Thus, we would like to argue that phenomenology is undoubtly of great value but cannot become itself part of scientific theories with their explanatory power. Scientific theories and theory theories are pointing at quite different levels of understanding and therefore they are immune to either combination or rivalry.

In this sense, the immediacy of the theory theory approach seems to be attractive on the one side as it asserts to be a purely empirical approach, but is shortsighted on the other side as it may fail to understand the psychological processes which are involved and which are not in all cases directly accessible. Understanding requires a theoretical framework as how to get access to and integrate the empirical phenomena. Therefore, failures in the childrens' performance may be as informative as the successes on which Gopnik is focusing.

And only in this sense Gopnik's expertise analogy seems to hold, as e.g. golf experts know a lot about playing golf but this knowledge seems not sufficient to understand what the basis of their particular success or failure is. That is what sport psychology is about by confirming or rejecting and substituting the theory theory of the golf player. In general, however, as discussed above the expertise analogy seem to be misleading in particular if one considers evidence that expertise and immediacy is not as strongly related as Gopnik claims and experts' theories may even interfere with information processing resulting in poor judgements (e.g. Mischel, 1968, Rosenhan 1973).

Goldman's attempt to understand common sense mental representations by using an analogy from visual perception seems problematic for several reasons.

First of all it is unclear why the visual mode of processing is priviliged to provide the basis of the analogy. Even if one agrees that the visual modality is a prominent one the question remains what qualifies it as basis of a general model. In particular, as Goldman is later refering to other sense modalities, trying to justify his emphasis in qualitative properties in cognition. At this stage, it became obvious that Goldman's analogies are rather arbitrary and lack substantial reason either from phenomenology or from the mode of cognitive processing itself. He tries to justify his emphasis of qualitative properties with the occurence of qualitative sensations. However, qualitative sensations are exceptional and prominent only in the gustatory modality, as we have four distinct qualities of taste. The acceptance of the Weber/Fechner law in most other modalities convincingly demonstrates the absence of qualitative sensations in these senses.

Secondly, Goldman is stressing the difference between the semantic content and the structure of words. In order to match the two aspects or to match single exemplars he introduces a homunculus, called cognitizer. Introducing a cognitizer seems consequent because it is otherwise quite unclear how intentional aspects could be implemented in mental processes, which Goldman models in analogy to pure sensations. He seems to propose that indeed the structural aspects are bottom-up processes, whereas semantics is provided by a top-down process. In Goldman's approach there remains a gulf between the data base and the interpretations erected upon it. Although the pure empirical nature of his approach, the same is true for Gopnik, is emphasized several times, it can be argued that the data do not by themselves determine an interpretation, but leave a gap to be bridged by inferential processes of various kind.

In this sense intentionalistic description of mental processes is theoretical vis-a-vis the behavioural one. The question, which is not answered neither by Goldman nor by Gopnik remains whether there is a shared observational base for the intentional or behavioural descriptions. In explaining the physical reality folk physics and physics share the same data base. This seem not always the case in human psychology where third-person knowledge may interfere with the person's own understanding of its behaviour (first-person knowledge).

Goldman's attempt to establish a psychology of folk psychology is nothing else than an attempt to reestablish the principles of introspection, here applied to mentalese. In this context, it is highly astonishing that he claims that 'the best known psychological critique of introspective access is Nisbett & Wilson's (1977)'. We do not doubt the value of this paper, but it has to be noticed that the short history of scientific psychology is characterized by its attempt to emancipate itself from introspective and intuitive methodology. Gestalt Psychology, behaviourism and cognitive psychology, for example, all developed methods to overcome introspective psychology. In relation to consciousness it can be argued, as Velman (1991) convincingly exemplified, that consciousness may result from the coordination and integration of diverse cognitive processes such as perception, attention, learning, memory, and thinking but that quite often and regularly the sequence of cognitive processes is inaccesssible to conscious experience. At least for this reason, Goldman's justification of a cognitive science of folk psychology is not convincing. In Goldman and Gopnik's approach reality is the sum of the facts that are cognitively accessible to humans. Therefore, accessibility through direct observation or inference from observational facts (in most of Gopnik's examples) is an absolutely essential factor for their concept. However, the main progress in understanding our cognitive processes and mental states was achieved by experimental methodology which enabled analyses independent of the question of accessibility either in psychology or the neurosciences. Gopnik's approach to demonstrate the development of the theory of mind in children is interesting, because it demonstrates the original formation and change of intuitive concepts. However, this approach has to be completed by studying the development of the underlying processes and mechanisms. Only this knowledge will then enable us to understand the children's theory of mind entirely.


Preparation was supported by the Alexander von Humboldt-Foundation and the University of Exeter through a Feodor Lynen Fellowship held by W H Dittrich.


BRUNER, J. (1977). Early social interaction and language acquisition. In H.R.Schafer (Ed.), Studies in mother-infant interaction (pp. 271-290). NewYork: Academic.

DITTRICH, W., & LEA, S.E.G. (1992). Motion feature integration and the perception of intention (Deparment of Psychology, University of Exeter, BPRG Internal Report no. 91/05) [published as: Dittrich WH,& Lea SEG (1994) The perception of intentional motion. Perception 23 253-26]

HECKHAUSEN, H. (1982). The development of achievement motivation. In W.W.Hartup (Ed.), Review of child development research, Vol.6 (pp. 600-668). Chicago: University of Chicago Press.

HECKHAUSEN, H. (1983). Concern with one's own competence: Developmental shifts in person-environment interaction. In D.Magnussen & V.Allen (Eds.), Human development: An interactional perspective. NewYork: Academic.

JACKSON, E., CAMPOS, J.J.,& FISCHER, K.W. (1978). The question of decalage between object permanence and person permanence. Developmental Psychology, 14, 1-10.

KOPP, C.B. (1982). Antecendents of self-regulation: A developmental perspective. Developmental Psychology, 18, 199-214.

MISCHEL, W. (1968). Personality and assessment. NewYork: Wiley.

NISBETT, R.E.,& WILSON, T.D. (1977). Telling more than we can know: Verbal reports on mental processes. Psychological Review, 84, 231-259.

ROSENHAN, D.L. (1973). On being sane in insane places. Science, 179, 250-258.

SHIELDS, M.M. (1978). The child as psychologist: constructing the social world. In A.Lock (Ed.), Action, gesture and symbol (pp. 529-556). NewYork: Academic.

SPITZ, R. (1957). No and yes. On the genesis of human communication. NewYork: International University Press.

TREVARTHEN, C.,& HUBLEY, P. (1978). Secondary intersubjectivity: Confidence, confiding and acts of meaning in the first year. In A.Lock (Ed.), Action, gesture and symbol (pp. 183-229). NewYork: Academic.

VELMANS, M. (1991). Is human information processing conscious? Behavioral and Brain Sciences, 14, 651-726.