- Email: [email protected]
How the brain understands intention: Different neural circuits identify the componential features of motor and prior intentions
Consciousness and Cognition Consciousness and Cognition 15 (2006) 64–74 www.elsevier.com/locate/concog
How the brain understands intention: Different neural circuits identify the componential features of motor and prior intentions Cristina Becchio *, Mauro Adenzato, Bruno G. Bara Center for Cognitive Science, Department of Psychology, University of Turin, Turin, Italy Received 19 October 2004 Available online 2 June 2005
Abstract In this paper we present theoretical and experimental evidence for a set of mechanisms by which intention is understood. We propose that three basic aspects are involved in the understanding of intention. The first aspect to consider is intention recognition, i.e., the process by which we recognize other peopleÕs intentions, distinguishing among different types. The second aspect concerns the attribution of intention to its author: the existence of shared neural representations provides a parsimonious explanation of how we recognize other peopleÕs intentions (what they are doing), but in and of itself, is not sufficient to determine who the agent is. Once the intention has been recognized and attributed to an agent, the reasons for, and the aim of, the intention are to be considered. Hence, the third aspect concerns the aim-intention motivating the execution of a certain action. We discuss the neural basis of these three theoretical aspects suggesting a conceptual synthesis. Ó 2005 Elsevier Inc. All rights reserved. Keywords: Motor intention; Prior intention; Social cognition; Shared neural representation; Agency; Prefrontal cortex
*
Corresponding author. Fax: +39 011 815 90 39 E-mail address: [email protected] (C. Becchio).
1053-8100/$ - see front matter Ó 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.concog.2005.03.006
C. Becchio et al. / Consciousness and Cognition 15 (2006) 64–74
65
1. Introduction People normally distinguish between ‘‘intentional’’ and ‘‘unintentional’’ behaviour, and this distinction applies not only to self-initiated actions, but also to other peopleÕs actions. This is true as early as the second year of life (Carpenter, Akhtar, & Tomasello, 1998; Csibra, 2003). By observing people acting, we can usually say what they doing and what their goals are (Baldwin & Baird, 2001). We can frequently even imagine the reason why they are acting. It is perhaps easy to take this capacity of ours for granted, until we consider that no such things as goals, intentions, reasons actually can be either seen, heard or smelled in the objective world. The topic of this paper is intentionality. In particular, we present theoretical and experimental evidence for a set of neural mechanisms that, from the observation of action, enable the understanding of intention. We highlight, moreover, how three basic aspects are involved in the understanding of intention: intention recognition, agent attribution, and aim-representation (see Table 1).
2. Recognising intention Accumulative empirical evidence suggests that recognising the intentions of others is based, at least in part, on the same mechanisms underlying the formation of oneÕs own motor intention (Frith, 2002). The idea is that the same cortical areas that are activated when we execute an action are also activated when we observe other people performing a similar action. In this section we limit our analysis to the case of action observation, but of course not all examples of intention recognition are motor based. For example, intentions may be understood on the basis of eye gaze and emotional expression (Baldwin, 2000; Castiello, 2003; Jellema, Baker, Wicker, & Perrett, 2000). And also a silence can convey an intention (Bara, 2005). The discovery made approximately 10 years ago of a particular class of motor neurons in a sector of the ventral premotor cortex of monkeys, called F5, provided the first convincing physiological evidence for a direct match between action execution and action perception (Gallese, Fadiga, Fogassi, & Rizzolatti, 1996; Rizzolatti, Fadiga, Gallese, & Fogassi, 1996). In fact, quite unexpectedly, the study of motor functions showed that not only does the F5 area accommodate purely motor neurons, but it also houses neurons that fire when a recorded monkey observes another monkey, or even an experimenter, performing a similar action. These neurons were designated as ‘‘mirror neurons.’’
Table 1 Processes contributing to the understanding of intention Recognition
Agent attribution
Aim-representation
What? Enables identification of the intention from the observation of action, distinguishing among different types of motor intention
Who? Enables attribution of the intention to its author, distinguishing between self and others
Why? Enables identification of the aims of an intended action
66
C. Becchio et al. / Consciousness and Cognition 15 (2006) 64–74
Subsequent experiments showed that mirror neurons are not restricted to the premotor cortex, but are also found in other areas of the monkey brain, notably, in the posterior parietal cortex (Gallese, Fogassi, Fadiga, & Rizzolatti, 2002), and in homologous areas in the human brain (Fadiga, Craighero, Buccino, & Rizzolatti, 2002; Gre`zes, Armony, Rowe, & Passingham, 2003; Hamzei et al., 2003). Kohler et al. (2002) recently demonstrated that a fairly high percentage of mirror neurons also fires when an action is merely heard. These neurons were defined ‘‘audio-visual mirror neurons.’’ As Jackson and Decety (2004) noted, the existence of a system matching executed and perceived actions offers a parsimonious explanation of how we recognise other peopleÕs intended actions— i.e., by a direct mapping of the visual representation of the observed action onto our own motor representation of the same action. Hence, the same neural representation is used for intending actions as well as recognising other peopleÕs motor intention (Garbarini & Adenzato, 2004). The role of intentionality in triggering motor resonance phenomena is demonstrated by recent neuroimaging studies. Using an adaptation of the apparent motion paradigm, originally developed by Shiffrar and Freyd (1990), Stevens, Fonlupt, Shiffrar, and Decety (2000) presented participants in a PET scanner with a human model moving in different positions. The left primary motor cortex and the parietal cortex were found to be selectively activated when participants perceived possible human movement paths. No selective activation of these areas was found during conditions of biomechanically impossible movement paths. Tai, Scherfler, Brooks, Sawamoto, and Castiello (2004) used a factorial design to directly measure how neural responses associated with observation of manual grasping action are modulated by the biological nature of the model. The study confirmed that the mirror property of the premotor cortex is biologically tuned: to activate the mirror system, the action must be part of the motor repertoire of the subject watching it. 2.1. Shared neural representations One theoretical consequence of the equivalence between perception and action is that we must hypothesise the existence of a level of neural representations that are neutral in terms of modality of activation and agent. Distinguishing between these two aspects of neutrality allows us to redefine the notion of shared neural representations (introduced by Decety & Grezes, 1999) at two different levels. On the intraindividual level, neural representations are shared in the sense that they are activated in different action modalities. Being neutral in terms of modality, the representations may be shared by different modalities. As Jeannerod (2003) noted, however, this is not the only sense in which neural representations are shared. They are active when a person executes an action, but also when he sees or hears another individual performing the same action. Hence, neural representations are shared not only by the same structures for different types of action, i.e., executed and observed action, but also by different brains (see Table 2). When two agents interact socially with one another, the activation of mirror networks creates a shared neural representation—i.e., representations simultaneously activated in the brains of two agents. Given that self-generated actions and other peopleÕs actions are mapped onto the same neural substratum, the same representations are activated in both agents. The existence of shared neural representations enables us to settle the question of recognition, but it poses the problem of action/intention attribution (Becchio & Bertone, 2004): if the same
C. Becchio et al. / Consciousness and Cognition 15 (2006) 64–74
67
Table 2 Shared neural representations Modality
Agent
The same representation is activated in different action modalities, e.g., when an individual executes an action, observes another individual performing the same action, or simply imagines doing the action. The activation concerns the intended action, regardless of the modality through which it is inferred The representation of the action is agent-free: the action can be ascribed either to the self or another person
Different modalities share the same neural representation
Different brains share the same representation
areas are activated while representing oneÕs own action and during observation of another personÕs action, how do we distinguish between ourselves and others? The activation of shared neural representations allows us to identify the intention (e.g., to discriminate between the intention of grasping an object and the intention of touching it), but is not sufficient per se to attribute the intention to an agent.
3. Attributing intention The fact that agents in normal conditions are indeed able to correctly attribute actions to their authors implies the need for a specific causal process that disambiguates representations by articulating who the agent is—i.e., a ‘‘Who’’ system (Georgieff & Jeannerod, 1998), specifically dedicated to action attribution. Various neural mechanisms have been proposed to explain how such a system might operate. One hypothesised mechanism is the monitoring of signals arising from body movements, i.e., comparison between the control signals that contribute to generating a movement and signals arising out of its execution. There is firm neuroimaging evidence that the inferior parietal cortex and the insula are crucial components of this mechanism, which is specifically involved in perceiving the spatial features of movements (Chaminade & Decety, 2002; Decety, Chaminade, Gre`zes, & Meltzoff, 2002; Farrer et al., 2003; for a review see also Jackson & Decety, 2004). Convergent evidence from neuropsychological research show that patients with parietal lesion fail to discriminate between their own actions and actions performed by an experimenter (Sirigu, Daprati, Pradat-Diehl, Franck, & Jeannerod, 1999). Interestingly, in a recent fMRI study Leube et al. (2003) found that right posterior superior sulcus activation positively correlates with a temporal delay introduced on-line between an action and its visual feedback. The spatio-temporal congruence of the features of a movement certainly represents a strong index for determining the experience of agency. This mechanism, however, does not account for the frequent instances in which an intention is generated but the corresponding action is not executed. A paradigmatic situation where this occurs is motor imagery (i.e., imagining doing an action), in which there are no reafferent signals and no proprioceptive signals. Yet, attribution of action is clearly made (Jeannerod, 2003). Jeannerod (2003) hypothesises that in situations such as these, correct attribution depends on the existence of non-overlapping areas that are specifically activated for the self and others. The diagram in Fig. 1 is a tentative illustration of this process. It represents a motor cognitive situation
68
C. Becchio et al. / Consciousness and Cognition 15 (2006) 64–74
Fig. 1. The diagram depicts the interactions of two agents (A and B) observing each other. Each agent builds a representation of his or her own intentions/actions and of the intentions/actions of the other agent. Representations of self-generated actions and observed actions tend to overlap. Adapted from Jeannerod (2003).
involving two people. Agent A generates a representation of a self-generated action, i.e., a motor intention. If it is then executed, it will become a signal for agent B, such that agent B will form a representation of the action she sees. The interaction of the two agents should depend on the interaction of the representations of the observed and executed actions within each of the two brains. In fact with an individual, the two representations are closely situated and partially overlap. Determining who is acting—the other or the self—should be based on the non-overlapping part. Motor imagery studies provide a crucial opportunity to test these hypotheses. In a PET experiment, Ruby and Decety (2001) asked participants to imagine actions from different perspectives. In the first-person perspective, participants were told to imagine themselves executing a given
C. Becchio et al. / Consciousness and Cognition 15 (2006) 64–74
69
action. In the third-person perspective, the instruction was to imagine that they were watching somebody else doing that same action. In the first-person perspective, specific activation was observed in the inferior parietal lobule in the left hemisphere. By contrast, activation was found in a symmetrical area of the right hemisphere for the third-person perspective. Other areas were activated in both conditions. It is interesting to note that in both experimental conditions, participants were required to imagine an action. The representations activated in both situations were thus activated from within. The fact that a difference in activation was nevertheless observed in the right inferior parietal lobe shows that the inferior parietal lobe is not simply involved in associating actions and their sensory consequences (Farrer et al., 2003), but also in distinguishing the self from others. In a subsequent PET study, Ruby and Decety (2003) investigated the neural correlates of the self/other distinction at the conceptual level by comparing the neural networks involved in answering the same set of questions from either a first- or a third-person perspective. Participants were selected from a group of medical students and were presented with written sentences related to health sciences. They were then instructed to make truthful judgements about the statements, according either to their own perspective or according to what they thought a layman might think. Variations in rCBF between the two conditions demonstrated that the brain areas distinguishing the cerebral correlates of first- and third-person perspective-taking at the conceptual level are similar to those already detected at the motor level, i.e., the right inferior parietal lobe and the frontopolar and somatosensory cortices. Furthermore, the role of the inferior parietal lobe in first-person/third-person perspective distinction is confirmed by neuropsychological and psychopathological evidence. Mesulam (1981) reported the case of a patient with an abscess of the right parieto-occipital region who suffered from delusion of external control. Schizophrenic patients experiencing passivity phenomena believe their thoughts and actions to be those of external, or alien, entities. Using a PET methodology on such patients, Spence et al. (1997) found hyperactivation of the right inferior parietal cortex, and experience passivity as compared to healthy subjects during the performance of freely selected movements.
4. Aim-representation: Distinguishing between private and social aims The processes described thus far concern the recognition and attribution of motor intention. Both of these aspects contribute to what Searle (1983) calls intention-in-action, i.e., the mental and causal component of an action. An intention-in-action is the cause of an agentÕs movement. Yet, it should be noted that the causal domain of the intention-in-action extends only as far as the bodily movement of an action (Searle, 1983, p. 95). To cover the overall conditions of an action we need to analyse the prior intention (Searle, 1983) that orients the action as a whole. As Jacob and Jeannerod (2005) observe, the mirror system is well designed for representing the agentÕs motor intention, but not her prior intention to execute an action. In contrast to intention-in-action, which represents conditions of satisfaction of an act in progress, the prior intention is formed in advance, representing goal states that are in some way very distal to the chain of events that lead to their fulfillment. If action is a ‘‘causal and intentional transaction between the mind and the world’’ (Searle, 1983), prior intention can be said to initiate
70
C. Becchio et al. / Consciousness and Cognition 15 (2006) 64–74
Fig. 2. Graphic representation of the relation between action, intention-in-action, prior intention, and intended-aim.
a transaction, by representing the end or aim of the action before the action is undertaken. (Fig. 2). The same behaviour, performed by the same agent can be initiated for different aims: enlarging the temporal horizon, by shifting from intention-in-action to prior intention, allows us to focus on what the agent aims at by its action. This aspect of intention—aim-intention, in Tuomela (in press) terminology—determines the mode of the intention (Tuomela, 2002). Consider the case of an agent, John, lightening a candle. John may light a candle because the electricity has gone out, and he wants to read a book, or because he is planning a romantic evening with Mary, or to celebrate Independence Day. In the first case, John is acting in the pursuit of a merely private goal, driven by an I-mode intention; in the latter case, lighting a candle satisfies a shared we-attitude (Tuomela, 2002). The action is the same, and so is the author. What changes is the ‘‘mode’’ of intention: ‘‘I-mode’’ means acting and having an attitude privately, as a private person, whereas ‘‘we-mode’’ means having it as a group member. The instance of John planning a romantic evening with Mary may represent a sort of ‘‘intermediate case’’: although John is not acting for a shared social reason, he is oriented toward future social interaction. Walter et al. (2004) have launched a new series of brain mapping experiments, with the objective of distinguishing between merely private aim-intention and social aim-intention. In a first fMRI experiment, participants were asked to read short comic strips that depicted an unfolding story plot. The participantsÕ task was to choose the most logical story ending from three answer pictures. Story content was either physical, i.e., unintentional (a ball blown by the wind breaking several bottles) or intentional. In turn, intentional strips pertained to three conceptual categories, depicting either the intentional action of a single agent (changing a broken bulb to read a book) or of two agents acting independently (an agent building a doghouse while another agent set up a tent to camp), or social interaction between two people communicating through gestures (requesting another person to pass a bottle by pointing to it). The most interesting result was a significant increase in activation associated with the social interaction condition. Seeing two agents communicating resulted in significant activation in the medial prefrontal cortex, especially in the anterior paracingulate cortex. The fact that Walter et al. (2004) found no paracingulate activation during the reading of comic strips depicting either one agent acting or two agents acting independently of each other suggested that the observed neural activation was not due simply to the intentional content or to the number of agents represented in stories. Indeed, the activation of the anterior paracingulate cortex required two socially interacting agents. In this first experiment, social interaction always involved two communicating agents. Communicative intention is necessarily social, for it involves taking other people into account, as part of oneÕs reasons for acting (Tuomela & Bonnevier-Tuomela, 1997). In contrast to private intention, which can be realised by an isolated person, a communicative intention can occur only during social interaction (Bosco, Bucciarelli, & Bara, 2004). According to Cognitive Pragmatics (Bara, 2005), however, communicative intentions represent a ‘‘special’’ sort of social intention, consisting
C. Becchio et al. / Consciousness and Cognition 15 (2006) 64–74
71
not only of the intention to communicate meaning, but also of the intention that this first intention should be recognised by the addressee. Thus, one question raised by the results of Walter et al.Õs (2004) first experiment was as follows: to what extent can the paracingulate cortex activation be attributed to the specificity of the communicative interaction1? Would a social interaction not involving communicative intention have resulted in the same activation? Walter et al. (2004) designed a second experiment to further investigate the nature of the observed anterior paracingulate cortex activation. This experiment introduced a new conceptual category: prospective social intentionality, which the authors referred to as the intention of a single agent whose private action is oriented towards future social interaction (e.g., John preparing a romantic evening with Mary). The reasoning underlying the experiment was that if anterior paracingulate cortex activation was due to the presence of two agents actually interacting—or, more specifically, communicating—no activation should be observed in the prospective social intentionality condition. Conversely, if the activation depended on the social nature of the aim an agent pursued, anterior paracingulate cortex activation should still be observed. The results confirmed this second prediction: pattern of activation similar to that observed in the communicative intentionality condition was also found in the prospective social intentionality condition. As expected, based on the results from the previous experiment, no paracingulate activation was detected in the intentional conditions that presented agents acting for a private aim. Both of the above described experiments used passive, off-line tasks, in which participants merely observed social interaction without taking part in it. Analogous conclusions for the special role of the anterior paracingulate cortex derive, however, from some recent studies in which participants interacted directly. In an fMRI experiment conducted by McCabe, Houser, Ryan, Smith, and Trouard (2001), participants in a scanner played standard, two-party ‘‘trust and reciprocity’’ games. Participants were matched with either a human or a computer counterpart following a probabilistic strategy and were visually informed of what type of counterpart they would be playing with before beginning the game. Participants then received cooperation scores based on behavioural data and were divided into two groups: cooperators and non-cooperators. In the cooperator group, regions of the prefrontal cortex, including the anterior paracingulate cortex, were more activated when participants played with a human than when they played with a computer. There were no significant activation differences between human and the computer conditions in the non-cooperator group. Both Walter et al. (2004) and McCabe et al. (2001) used cooperative social interaction. A similar activation of the anterior paracingulate cortex was found by Gallagher, Jack, Roepstorff, and Frith (2002) in a PET experiment involving competitive interaction. Volunteers played a computerised version of the childrenÕs game ‘‘stone, paper, scissors’’ against an opponent producing a random sequence. In one condition, volunteers believed they were playing against the experimenter, and in the comparison condition, they believed they were playing against a computer. Yet, the only difference between the two conditions was participant attitude. Only one brain region was more active when volunteers believed they were interacting with an intentional agent, i.e., the anterior paracingulate cortex (see Table 3).
1 The question was moreover warranted by the fact that Kampe, Frith, and Frith (2003) found similar activation by using communicative intention conveyed both by direct eye gaze and by hearing oneÕs own name pronounced.
72
C. Becchio et al. / Consciousness and Cognition 15 (2006) 64–74
Table 3 Anterior paracingulate cortex: Conditions for activation Type of interaction
Number of agents
Modality of interaction
Time of interaction
Cooperative or competitive
One agent (prospective) or more (actual)
Participated or observed
Present or future
5. Conclusions This paper has examined three basic aspects of understanding intention: intention recognition, agent attribution, and aim-representation. On the basis of theoretical and neuroscientific evidence, we suggest that these aspects constitute fundamental structures in a complex functional and anatomical architecture. The questions raised by experimental investigation can be traced only in part back to traditional philosophical issues (e.g., the distinction of prior intention from intention-in-action). Others, rather, are new and evolve out of the empirical investigation process. This is the case of agentattribution: as we have seen, the problem arises out of the discovery of a recognition level that is based on agent-free mechanisms. The data available at present allow us to hypothesise a solution in terms of intention-in-action. More specifically, shared neural representations—which are neutral with respect to both the modality and the agent—are disambiguated by the ‘‘Who system’’ operating on them. With respect to previous evidence, the interest of new findings such as Walter et al. (2004) consist in the fact that they open the way to experimental investigation of aim-intention and social intention, showing that the anterior paracingulate cortex is activated in representing social aims, independently from interaction type (cooperative vs. competitive), time (present vs. future), and modality (participated vs. observed). Acknowledgment We would like to thank Prof. Raimo Tuomela for valuable comments to an early version of this manuscript. This research was financially supported by MURST of Italy (cofin 2003, protocol no. 2003119330_008). References Baldwin, D. A. (2000). Interpersonal understanding fuels knowledge acquisition. Current Directions in Psychological Science, 9, 40–45. Baldwin, D. A., & Baird, J. A. (2001). Discerning intentions in dynamic human action. Trends in Cognitive Sciences, 4, 171–178. Bara, B. G. (2005). Cognitive pragmatics. Cambridge, MA: MIT Press. Becchio, C., & Bertone, C. (2004). Wittgenstein running: Neural mechanisms of collective intentionality and we-mode. Consciousness and Cognition, 13, 123–133. Bosco, F. M., Bucciarelli, M., & Bara, B. G. (2004). The fundamental context categories in understanding communicative intention. Journal of Pragmatics, 36, 467–488.
C. Becchio et al. / Consciousness and Cognition 15 (2006) 64–74
73
Carpenter, M., Akhtar, N., & Tomasello, M. (1998). Fourteen- to 18-month-old infants differentially imitate intentional and accidental actions. Infant Behavior and Development, 21, 315–330. Castiello, U. (2003). Understanding other peopleÕs actions: Intention and attention. Journal of Experimental Psychology: Human Perception and Performance, 29, 416–430. Chaminade, T., & Decety, J. (2002). Leader or follower. Involvement of the inferior parietal lobule in agency. NeuroReport, 13, 1975–1978. Csibra, G. (2003). Teleological and referential understanding of action in infancy. Philosophical Transactions of the Royal Society of London, Series B-Biological Sciences, 358, 447–458. Decety, J., Chaminade, T., Gre`zes, J., & Meltzoff, A. N. (2002). A PET exploration of the neural mechanisms involved in reciprocal imitation. Neuroimage, 15, 265–272. Decety, J., & Grezes, J. (1999). Neural mechanisms subserving perception of human actions. Trends in Cognitive Sciences, 3, 172–178. Fadiga, L., Craighero, L., Buccino, G., & Rizzolatti, G. (2002). Speech listening specifically modulates the excitability of tongue muscles: A TMS study. European Journal of Neuroscience, 15, 399–402. Farrer, C., Franck, N., Georgieff, N., Frith, C. D., Decety, J., & Jeannerod, M. (2003). Modulating the experience of agency: A PET study. Neuroimage, 18, 324–333. Frith, C. (2002). Attention to action and awareness of other minds. Consciousness and Cognition, 11, 481–487. Gallagher, H., Jack, A. I., Roepstorff, A., & Frith, C. D. (2002). Imaging the intentional stance in a competitive game. Neuroimage, 16, 814–821. Gallese, V., Fadiga, L., Fogassi, L., & Rizzolatti, G. (1996). Action recognition in the premotor cortex. Brain, 119, 593–609. Gallese, V., Fogassi, L., Fadiga, L., & Rizzolatti, G. (2002). Action representation in the inferior parietal lobule. In W. Prinz & B. Hommel (Eds.). Attention and Performance (Vol. 19, pp. 247–266). New York: Oxford University Press. Garbarini, F., & Adenzato, M. (2004). At the root of embodied cognition: Cognitive science meets neurophysiology. Brain and Cognition, 56, 100–106. Georgieff, N., & Jeannerod, M. (1998). Beyond consciousness of external reality: A ‘‘Who’’ system for consciousness of action and self-consciousness. Consciousness and Cognition, 7, 465–477. Gre`zes, J., Armony, J. L., Rowe, J., & Passingham, R. E. (2003). Activations related to mirror and canonical neurons in the human brain: An fMRI study. Neuroimage, 18, 928–937. Hamzei, F., Rijntjes, M., Dettmers, C., Glauche, V., Weiller, C., & Buchel, C. (2003). The human action recognition system and its relationship to BrocaÕs area: An fMRI study. Neuroimage, 19, 637–644. Jackson, P. L., & Decety, J. (2004). Motor cognition: A new paradigm to study self-other interactions. Current Opinion in Neurobiology, 14, 259–263. Jacob, P., & Jeannerod, M. (2005). The motor theory of social cognition: A critique. Trends in Cognitive Sciences, 9, 21–25. Jeannerod, M. (2003). The mechanisms of self-recognition in humans. Behavioural Brain Research, 142, 1–15. Jellema, T., Baker, C. I., Wicker, B., & Perrett, D. I. (2000). Neural representation for the perception of the intentionality of actions. Brain and Cognition, 44, 280–302. Kampe, K. K. W., Frith, C. D., & Frith, U. (2003). ‘‘Hey John’’: Signals conveying communicative intention toward the self activate brain regions associated with ‘‘mentalizing’’, regardless of modality. Journal of Neuroscience, 23, 5258–5263. Kohler, E., Keysers, C., Umilta`, M. A., Fogassi, L., Gallese, V., & Rizzolatti, G. (2002). Hearing sounds, understanding actions: Action representation in mirror neurons. Science, 297, 846–848. Leube, D. T., Knoblich, G., Erb, M., Grodd, W., Bartels, M., & Kircher, T. T. J. (2003). The neural correlates of perceiving oneÕs own movements. Neuroimage, 20, 2084–2090. McCabe, K., Houser, D., Ryan, L., Smith, V., & Trouard, T. (2001). A functional imaging study of cooperation in twopersons reciprocal exchange. Proceedings of the National Academy of Sciences, USA, 98, 11832–11835. Mesulam, M. M. (1981). Dissociative states with abnormal temporal lobe EEG. Multiple personality and the illusion of possession. Archives of Neurology, 38, 176–181. Rizzolatti, G., Fadiga, L., Gallese, V., & Fogassi, L. (1996). Premotor cortex and the recognition of motor actions. Cognitive Brain Research, 3, 131–141.
74
C. Becchio et al. / Consciousness and Cognition 15 (2006) 64–74
Ruby, P., & Decety, J. (2001). Effect of subjective perspective taking during simulation of action: A PET investigation of agency. Nature Neuroscience, 4, 546–550. Ruby, P., & Decety, J. (2003). What you believe versus what you think they believe: A neuroimaging study of conceptual perspective-taking. European Journal of Neuroscience, 17, 2475–2480. Searle, J. (1983). Intentionality: An essay in the philosophy of mind. New York: Cambridge University Press. Shiffrar, M., & Freyd, J. J. (1990). Apparent motion of the human body. Psychological Science, 1, 257–264. Sirigu, A., Daprati, E., Pradat-Diehl, P., Franck, N., & Jeannerod, M. (1999). Perception of self-generated movement following left parietal lesion. Brain, 122, 1867–1874. Spence, S. A., Brooks, D. J., Hirsch, S. R., Liddle, P. F., Meehan, J., & Grasby, P. M. (1997). A PET study of voluntary movement in schizophrenic patients experiencing passivity phenomena (delusions of alien control). Brain, 120, 1997–2011. Stevens, J. A., Fonlupt, P., Shiffrar, M., & Decety, J. (2000). New aspects of motion perception: Selective neural encoding of apparent human movements. NeuroReport, 1, 109–115. Tai, Y. F., Scherfler, C., Brooks, D. J., Sawamoto, N., & Castiello, U. (2004). The human premotor cortex is ÔmirrorÕ only for biological actions. Current Biology, 14, 117–120. Tuomela, R. (in press). We-intentions revisited. Philosophical Studies. Tuomela, R. (2002). The philosophy of social practices: A collective acceptance view. Cambridge: Cambridge University Press. Tuomela, R., & Bonnevier-Tuomela, M. (1997). From social imitation to teamwork. In G. Holmstro¨m-Hintikka & R. Tuomela (Eds.), Contemporary action theory, Vol. II: Social action (pp. 1–47). Dordrecht and Boston: Kluwer Academic Publishers. Walter, H., Adenzato, M., Ciaramidaro, A., Enrici, I., Pia, L., & Bara, B. G. (2004). Understanding intentions in social interaction: The role of the anterior paracingulate cortex. Journal of Cognitive Neuroscience, 16, 1854–1863.
How the brain understands intention: Different neural circuits identify the componential features of motor and prior intentions Cristina Becchio *, Mauro Adenzato, Bruno G. Bara Center for Cognitive Science, Department of Psychology, University of Turin, Turin, Italy Received 19 October 2004 Available online 2 June 2005
Abstract In this paper we present theoretical and experimental evidence for a set of mechanisms by which intention is understood. We propose that three basic aspects are involved in the understanding of intention. The first aspect to consider is intention recognition, i.e., the process by which we recognize other peopleÕs intentions, distinguishing among different types. The second aspect concerns the attribution of intention to its author: the existence of shared neural representations provides a parsimonious explanation of how we recognize other peopleÕs intentions (what they are doing), but in and of itself, is not sufficient to determine who the agent is. Once the intention has been recognized and attributed to an agent, the reasons for, and the aim of, the intention are to be considered. Hence, the third aspect concerns the aim-intention motivating the execution of a certain action. We discuss the neural basis of these three theoretical aspects suggesting a conceptual synthesis. Ó 2005 Elsevier Inc. All rights reserved. Keywords: Motor intention; Prior intention; Social cognition; Shared neural representation; Agency; Prefrontal cortex
*
Corresponding author. Fax: +39 011 815 90 39 E-mail address: [email protected] (C. Becchio).
1053-8100/$ - see front matter Ó 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.concog.2005.03.006
C. Becchio et al. / Consciousness and Cognition 15 (2006) 64–74
65
1. Introduction People normally distinguish between ‘‘intentional’’ and ‘‘unintentional’’ behaviour, and this distinction applies not only to self-initiated actions, but also to other peopleÕs actions. This is true as early as the second year of life (Carpenter, Akhtar, & Tomasello, 1998; Csibra, 2003). By observing people acting, we can usually say what they doing and what their goals are (Baldwin & Baird, 2001). We can frequently even imagine the reason why they are acting. It is perhaps easy to take this capacity of ours for granted, until we consider that no such things as goals, intentions, reasons actually can be either seen, heard or smelled in the objective world. The topic of this paper is intentionality. In particular, we present theoretical and experimental evidence for a set of neural mechanisms that, from the observation of action, enable the understanding of intention. We highlight, moreover, how three basic aspects are involved in the understanding of intention: intention recognition, agent attribution, and aim-representation (see Table 1).
2. Recognising intention Accumulative empirical evidence suggests that recognising the intentions of others is based, at least in part, on the same mechanisms underlying the formation of oneÕs own motor intention (Frith, 2002). The idea is that the same cortical areas that are activated when we execute an action are also activated when we observe other people performing a similar action. In this section we limit our analysis to the case of action observation, but of course not all examples of intention recognition are motor based. For example, intentions may be understood on the basis of eye gaze and emotional expression (Baldwin, 2000; Castiello, 2003; Jellema, Baker, Wicker, & Perrett, 2000). And also a silence can convey an intention (Bara, 2005). The discovery made approximately 10 years ago of a particular class of motor neurons in a sector of the ventral premotor cortex of monkeys, called F5, provided the first convincing physiological evidence for a direct match between action execution and action perception (Gallese, Fadiga, Fogassi, & Rizzolatti, 1996; Rizzolatti, Fadiga, Gallese, & Fogassi, 1996). In fact, quite unexpectedly, the study of motor functions showed that not only does the F5 area accommodate purely motor neurons, but it also houses neurons that fire when a recorded monkey observes another monkey, or even an experimenter, performing a similar action. These neurons were designated as ‘‘mirror neurons.’’
Table 1 Processes contributing to the understanding of intention Recognition
Agent attribution
Aim-representation
What? Enables identification of the intention from the observation of action, distinguishing among different types of motor intention
Who? Enables attribution of the intention to its author, distinguishing between self and others
Why? Enables identification of the aims of an intended action
66
C. Becchio et al. / Consciousness and Cognition 15 (2006) 64–74
Subsequent experiments showed that mirror neurons are not restricted to the premotor cortex, but are also found in other areas of the monkey brain, notably, in the posterior parietal cortex (Gallese, Fogassi, Fadiga, & Rizzolatti, 2002), and in homologous areas in the human brain (Fadiga, Craighero, Buccino, & Rizzolatti, 2002; Gre`zes, Armony, Rowe, & Passingham, 2003; Hamzei et al., 2003). Kohler et al. (2002) recently demonstrated that a fairly high percentage of mirror neurons also fires when an action is merely heard. These neurons were defined ‘‘audio-visual mirror neurons.’’ As Jackson and Decety (2004) noted, the existence of a system matching executed and perceived actions offers a parsimonious explanation of how we recognise other peopleÕs intended actions— i.e., by a direct mapping of the visual representation of the observed action onto our own motor representation of the same action. Hence, the same neural representation is used for intending actions as well as recognising other peopleÕs motor intention (Garbarini & Adenzato, 2004). The role of intentionality in triggering motor resonance phenomena is demonstrated by recent neuroimaging studies. Using an adaptation of the apparent motion paradigm, originally developed by Shiffrar and Freyd (1990), Stevens, Fonlupt, Shiffrar, and Decety (2000) presented participants in a PET scanner with a human model moving in different positions. The left primary motor cortex and the parietal cortex were found to be selectively activated when participants perceived possible human movement paths. No selective activation of these areas was found during conditions of biomechanically impossible movement paths. Tai, Scherfler, Brooks, Sawamoto, and Castiello (2004) used a factorial design to directly measure how neural responses associated with observation of manual grasping action are modulated by the biological nature of the model. The study confirmed that the mirror property of the premotor cortex is biologically tuned: to activate the mirror system, the action must be part of the motor repertoire of the subject watching it. 2.1. Shared neural representations One theoretical consequence of the equivalence between perception and action is that we must hypothesise the existence of a level of neural representations that are neutral in terms of modality of activation and agent. Distinguishing between these two aspects of neutrality allows us to redefine the notion of shared neural representations (introduced by Decety & Grezes, 1999) at two different levels. On the intraindividual level, neural representations are shared in the sense that they are activated in different action modalities. Being neutral in terms of modality, the representations may be shared by different modalities. As Jeannerod (2003) noted, however, this is not the only sense in which neural representations are shared. They are active when a person executes an action, but also when he sees or hears another individual performing the same action. Hence, neural representations are shared not only by the same structures for different types of action, i.e., executed and observed action, but also by different brains (see Table 2). When two agents interact socially with one another, the activation of mirror networks creates a shared neural representation—i.e., representations simultaneously activated in the brains of two agents. Given that self-generated actions and other peopleÕs actions are mapped onto the same neural substratum, the same representations are activated in both agents. The existence of shared neural representations enables us to settle the question of recognition, but it poses the problem of action/intention attribution (Becchio & Bertone, 2004): if the same
C. Becchio et al. / Consciousness and Cognition 15 (2006) 64–74
67
Table 2 Shared neural representations Modality
Agent
The same representation is activated in different action modalities, e.g., when an individual executes an action, observes another individual performing the same action, or simply imagines doing the action. The activation concerns the intended action, regardless of the modality through which it is inferred The representation of the action is agent-free: the action can be ascribed either to the self or another person
Different modalities share the same neural representation
Different brains share the same representation
areas are activated while representing oneÕs own action and during observation of another personÕs action, how do we distinguish between ourselves and others? The activation of shared neural representations allows us to identify the intention (e.g., to discriminate between the intention of grasping an object and the intention of touching it), but is not sufficient per se to attribute the intention to an agent.
3. Attributing intention The fact that agents in normal conditions are indeed able to correctly attribute actions to their authors implies the need for a specific causal process that disambiguates representations by articulating who the agent is—i.e., a ‘‘Who’’ system (Georgieff & Jeannerod, 1998), specifically dedicated to action attribution. Various neural mechanisms have been proposed to explain how such a system might operate. One hypothesised mechanism is the monitoring of signals arising from body movements, i.e., comparison between the control signals that contribute to generating a movement and signals arising out of its execution. There is firm neuroimaging evidence that the inferior parietal cortex and the insula are crucial components of this mechanism, which is specifically involved in perceiving the spatial features of movements (Chaminade & Decety, 2002; Decety, Chaminade, Gre`zes, & Meltzoff, 2002; Farrer et al., 2003; for a review see also Jackson & Decety, 2004). Convergent evidence from neuropsychological research show that patients with parietal lesion fail to discriminate between their own actions and actions performed by an experimenter (Sirigu, Daprati, Pradat-Diehl, Franck, & Jeannerod, 1999). Interestingly, in a recent fMRI study Leube et al. (2003) found that right posterior superior sulcus activation positively correlates with a temporal delay introduced on-line between an action and its visual feedback. The spatio-temporal congruence of the features of a movement certainly represents a strong index for determining the experience of agency. This mechanism, however, does not account for the frequent instances in which an intention is generated but the corresponding action is not executed. A paradigmatic situation where this occurs is motor imagery (i.e., imagining doing an action), in which there are no reafferent signals and no proprioceptive signals. Yet, attribution of action is clearly made (Jeannerod, 2003). Jeannerod (2003) hypothesises that in situations such as these, correct attribution depends on the existence of non-overlapping areas that are specifically activated for the self and others. The diagram in Fig. 1 is a tentative illustration of this process. It represents a motor cognitive situation
68
C. Becchio et al. / Consciousness and Cognition 15 (2006) 64–74
Fig. 1. The diagram depicts the interactions of two agents (A and B) observing each other. Each agent builds a representation of his or her own intentions/actions and of the intentions/actions of the other agent. Representations of self-generated actions and observed actions tend to overlap. Adapted from Jeannerod (2003).
involving two people. Agent A generates a representation of a self-generated action, i.e., a motor intention. If it is then executed, it will become a signal for agent B, such that agent B will form a representation of the action she sees. The interaction of the two agents should depend on the interaction of the representations of the observed and executed actions within each of the two brains. In fact with an individual, the two representations are closely situated and partially overlap. Determining who is acting—the other or the self—should be based on the non-overlapping part. Motor imagery studies provide a crucial opportunity to test these hypotheses. In a PET experiment, Ruby and Decety (2001) asked participants to imagine actions from different perspectives. In the first-person perspective, participants were told to imagine themselves executing a given
C. Becchio et al. / Consciousness and Cognition 15 (2006) 64–74
69
action. In the third-person perspective, the instruction was to imagine that they were watching somebody else doing that same action. In the first-person perspective, specific activation was observed in the inferior parietal lobule in the left hemisphere. By contrast, activation was found in a symmetrical area of the right hemisphere for the third-person perspective. Other areas were activated in both conditions. It is interesting to note that in both experimental conditions, participants were required to imagine an action. The representations activated in both situations were thus activated from within. The fact that a difference in activation was nevertheless observed in the right inferior parietal lobe shows that the inferior parietal lobe is not simply involved in associating actions and their sensory consequences (Farrer et al., 2003), but also in distinguishing the self from others. In a subsequent PET study, Ruby and Decety (2003) investigated the neural correlates of the self/other distinction at the conceptual level by comparing the neural networks involved in answering the same set of questions from either a first- or a third-person perspective. Participants were selected from a group of medical students and were presented with written sentences related to health sciences. They were then instructed to make truthful judgements about the statements, according either to their own perspective or according to what they thought a layman might think. Variations in rCBF between the two conditions demonstrated that the brain areas distinguishing the cerebral correlates of first- and third-person perspective-taking at the conceptual level are similar to those already detected at the motor level, i.e., the right inferior parietal lobe and the frontopolar and somatosensory cortices. Furthermore, the role of the inferior parietal lobe in first-person/third-person perspective distinction is confirmed by neuropsychological and psychopathological evidence. Mesulam (1981) reported the case of a patient with an abscess of the right parieto-occipital region who suffered from delusion of external control. Schizophrenic patients experiencing passivity phenomena believe their thoughts and actions to be those of external, or alien, entities. Using a PET methodology on such patients, Spence et al. (1997) found hyperactivation of the right inferior parietal cortex, and experience passivity as compared to healthy subjects during the performance of freely selected movements.
4. Aim-representation: Distinguishing between private and social aims The processes described thus far concern the recognition and attribution of motor intention. Both of these aspects contribute to what Searle (1983) calls intention-in-action, i.e., the mental and causal component of an action. An intention-in-action is the cause of an agentÕs movement. Yet, it should be noted that the causal domain of the intention-in-action extends only as far as the bodily movement of an action (Searle, 1983, p. 95). To cover the overall conditions of an action we need to analyse the prior intention (Searle, 1983) that orients the action as a whole. As Jacob and Jeannerod (2005) observe, the mirror system is well designed for representing the agentÕs motor intention, but not her prior intention to execute an action. In contrast to intention-in-action, which represents conditions of satisfaction of an act in progress, the prior intention is formed in advance, representing goal states that are in some way very distal to the chain of events that lead to their fulfillment. If action is a ‘‘causal and intentional transaction between the mind and the world’’ (Searle, 1983), prior intention can be said to initiate
70
C. Becchio et al. / Consciousness and Cognition 15 (2006) 64–74
Fig. 2. Graphic representation of the relation between action, intention-in-action, prior intention, and intended-aim.
a transaction, by representing the end or aim of the action before the action is undertaken. (Fig. 2). The same behaviour, performed by the same agent can be initiated for different aims: enlarging the temporal horizon, by shifting from intention-in-action to prior intention, allows us to focus on what the agent aims at by its action. This aspect of intention—aim-intention, in Tuomela (in press) terminology—determines the mode of the intention (Tuomela, 2002). Consider the case of an agent, John, lightening a candle. John may light a candle because the electricity has gone out, and he wants to read a book, or because he is planning a romantic evening with Mary, or to celebrate Independence Day. In the first case, John is acting in the pursuit of a merely private goal, driven by an I-mode intention; in the latter case, lighting a candle satisfies a shared we-attitude (Tuomela, 2002). The action is the same, and so is the author. What changes is the ‘‘mode’’ of intention: ‘‘I-mode’’ means acting and having an attitude privately, as a private person, whereas ‘‘we-mode’’ means having it as a group member. The instance of John planning a romantic evening with Mary may represent a sort of ‘‘intermediate case’’: although John is not acting for a shared social reason, he is oriented toward future social interaction. Walter et al. (2004) have launched a new series of brain mapping experiments, with the objective of distinguishing between merely private aim-intention and social aim-intention. In a first fMRI experiment, participants were asked to read short comic strips that depicted an unfolding story plot. The participantsÕ task was to choose the most logical story ending from three answer pictures. Story content was either physical, i.e., unintentional (a ball blown by the wind breaking several bottles) or intentional. In turn, intentional strips pertained to three conceptual categories, depicting either the intentional action of a single agent (changing a broken bulb to read a book) or of two agents acting independently (an agent building a doghouse while another agent set up a tent to camp), or social interaction between two people communicating through gestures (requesting another person to pass a bottle by pointing to it). The most interesting result was a significant increase in activation associated with the social interaction condition. Seeing two agents communicating resulted in significant activation in the medial prefrontal cortex, especially in the anterior paracingulate cortex. The fact that Walter et al. (2004) found no paracingulate activation during the reading of comic strips depicting either one agent acting or two agents acting independently of each other suggested that the observed neural activation was not due simply to the intentional content or to the number of agents represented in stories. Indeed, the activation of the anterior paracingulate cortex required two socially interacting agents. In this first experiment, social interaction always involved two communicating agents. Communicative intention is necessarily social, for it involves taking other people into account, as part of oneÕs reasons for acting (Tuomela & Bonnevier-Tuomela, 1997). In contrast to private intention, which can be realised by an isolated person, a communicative intention can occur only during social interaction (Bosco, Bucciarelli, & Bara, 2004). According to Cognitive Pragmatics (Bara, 2005), however, communicative intentions represent a ‘‘special’’ sort of social intention, consisting
C. Becchio et al. / Consciousness and Cognition 15 (2006) 64–74
71
not only of the intention to communicate meaning, but also of the intention that this first intention should be recognised by the addressee. Thus, one question raised by the results of Walter et al.Õs (2004) first experiment was as follows: to what extent can the paracingulate cortex activation be attributed to the specificity of the communicative interaction1? Would a social interaction not involving communicative intention have resulted in the same activation? Walter et al. (2004) designed a second experiment to further investigate the nature of the observed anterior paracingulate cortex activation. This experiment introduced a new conceptual category: prospective social intentionality, which the authors referred to as the intention of a single agent whose private action is oriented towards future social interaction (e.g., John preparing a romantic evening with Mary). The reasoning underlying the experiment was that if anterior paracingulate cortex activation was due to the presence of two agents actually interacting—or, more specifically, communicating—no activation should be observed in the prospective social intentionality condition. Conversely, if the activation depended on the social nature of the aim an agent pursued, anterior paracingulate cortex activation should still be observed. The results confirmed this second prediction: pattern of activation similar to that observed in the communicative intentionality condition was also found in the prospective social intentionality condition. As expected, based on the results from the previous experiment, no paracingulate activation was detected in the intentional conditions that presented agents acting for a private aim. Both of the above described experiments used passive, off-line tasks, in which participants merely observed social interaction without taking part in it. Analogous conclusions for the special role of the anterior paracingulate cortex derive, however, from some recent studies in which participants interacted directly. In an fMRI experiment conducted by McCabe, Houser, Ryan, Smith, and Trouard (2001), participants in a scanner played standard, two-party ‘‘trust and reciprocity’’ games. Participants were matched with either a human or a computer counterpart following a probabilistic strategy and were visually informed of what type of counterpart they would be playing with before beginning the game. Participants then received cooperation scores based on behavioural data and were divided into two groups: cooperators and non-cooperators. In the cooperator group, regions of the prefrontal cortex, including the anterior paracingulate cortex, were more activated when participants played with a human than when they played with a computer. There were no significant activation differences between human and the computer conditions in the non-cooperator group. Both Walter et al. (2004) and McCabe et al. (2001) used cooperative social interaction. A similar activation of the anterior paracingulate cortex was found by Gallagher, Jack, Roepstorff, and Frith (2002) in a PET experiment involving competitive interaction. Volunteers played a computerised version of the childrenÕs game ‘‘stone, paper, scissors’’ against an opponent producing a random sequence. In one condition, volunteers believed they were playing against the experimenter, and in the comparison condition, they believed they were playing against a computer. Yet, the only difference between the two conditions was participant attitude. Only one brain region was more active when volunteers believed they were interacting with an intentional agent, i.e., the anterior paracingulate cortex (see Table 3).
1 The question was moreover warranted by the fact that Kampe, Frith, and Frith (2003) found similar activation by using communicative intention conveyed both by direct eye gaze and by hearing oneÕs own name pronounced.
72
C. Becchio et al. / Consciousness and Cognition 15 (2006) 64–74
Table 3 Anterior paracingulate cortex: Conditions for activation Type of interaction
Number of agents
Modality of interaction
Time of interaction
Cooperative or competitive
One agent (prospective) or more (actual)
Participated or observed
Present or future
5. Conclusions This paper has examined three basic aspects of understanding intention: intention recognition, agent attribution, and aim-representation. On the basis of theoretical and neuroscientific evidence, we suggest that these aspects constitute fundamental structures in a complex functional and anatomical architecture. The questions raised by experimental investigation can be traced only in part back to traditional philosophical issues (e.g., the distinction of prior intention from intention-in-action). Others, rather, are new and evolve out of the empirical investigation process. This is the case of agentattribution: as we have seen, the problem arises out of the discovery of a recognition level that is based on agent-free mechanisms. The data available at present allow us to hypothesise a solution in terms of intention-in-action. More specifically, shared neural representations—which are neutral with respect to both the modality and the agent—are disambiguated by the ‘‘Who system’’ operating on them. With respect to previous evidence, the interest of new findings such as Walter et al. (2004) consist in the fact that they open the way to experimental investigation of aim-intention and social intention, showing that the anterior paracingulate cortex is activated in representing social aims, independently from interaction type (cooperative vs. competitive), time (present vs. future), and modality (participated vs. observed). Acknowledgment We would like to thank Prof. Raimo Tuomela for valuable comments to an early version of this manuscript. This research was financially supported by MURST of Italy (cofin 2003, protocol no. 2003119330_008). References Baldwin, D. A. (2000). Interpersonal understanding fuels knowledge acquisition. Current Directions in Psychological Science, 9, 40–45. Baldwin, D. A., & Baird, J. A. (2001). Discerning intentions in dynamic human action. Trends in Cognitive Sciences, 4, 171–178. Bara, B. G. (2005). Cognitive pragmatics. Cambridge, MA: MIT Press. Becchio, C., & Bertone, C. (2004). Wittgenstein running: Neural mechanisms of collective intentionality and we-mode. Consciousness and Cognition, 13, 123–133. Bosco, F. M., Bucciarelli, M., & Bara, B. G. (2004). The fundamental context categories in understanding communicative intention. Journal of Pragmatics, 36, 467–488.
C. Becchio et al. / Consciousness and Cognition 15 (2006) 64–74
73
Carpenter, M., Akhtar, N., & Tomasello, M. (1998). Fourteen- to 18-month-old infants differentially imitate intentional and accidental actions. Infant Behavior and Development, 21, 315–330. Castiello, U. (2003). Understanding other peopleÕs actions: Intention and attention. Journal of Experimental Psychology: Human Perception and Performance, 29, 416–430. Chaminade, T., & Decety, J. (2002). Leader or follower. Involvement of the inferior parietal lobule in agency. NeuroReport, 13, 1975–1978. Csibra, G. (2003). Teleological and referential understanding of action in infancy. Philosophical Transactions of the Royal Society of London, Series B-Biological Sciences, 358, 447–458. Decety, J., Chaminade, T., Gre`zes, J., & Meltzoff, A. N. (2002). A PET exploration of the neural mechanisms involved in reciprocal imitation. Neuroimage, 15, 265–272. Decety, J., & Grezes, J. (1999). Neural mechanisms subserving perception of human actions. Trends in Cognitive Sciences, 3, 172–178. Fadiga, L., Craighero, L., Buccino, G., & Rizzolatti, G. (2002). Speech listening specifically modulates the excitability of tongue muscles: A TMS study. European Journal of Neuroscience, 15, 399–402. Farrer, C., Franck, N., Georgieff, N., Frith, C. D., Decety, J., & Jeannerod, M. (2003). Modulating the experience of agency: A PET study. Neuroimage, 18, 324–333. Frith, C. (2002). Attention to action and awareness of other minds. Consciousness and Cognition, 11, 481–487. Gallagher, H., Jack, A. I., Roepstorff, A., & Frith, C. D. (2002). Imaging the intentional stance in a competitive game. Neuroimage, 16, 814–821. Gallese, V., Fadiga, L., Fogassi, L., & Rizzolatti, G. (1996). Action recognition in the premotor cortex. Brain, 119, 593–609. Gallese, V., Fogassi, L., Fadiga, L., & Rizzolatti, G. (2002). Action representation in the inferior parietal lobule. In W. Prinz & B. Hommel (Eds.). Attention and Performance (Vol. 19, pp. 247–266). New York: Oxford University Press. Garbarini, F., & Adenzato, M. (2004). At the root of embodied cognition: Cognitive science meets neurophysiology. Brain and Cognition, 56, 100–106. Georgieff, N., & Jeannerod, M. (1998). Beyond consciousness of external reality: A ‘‘Who’’ system for consciousness of action and self-consciousness. Consciousness and Cognition, 7, 465–477. Gre`zes, J., Armony, J. L., Rowe, J., & Passingham, R. E. (2003). Activations related to mirror and canonical neurons in the human brain: An fMRI study. Neuroimage, 18, 928–937. Hamzei, F., Rijntjes, M., Dettmers, C., Glauche, V., Weiller, C., & Buchel, C. (2003). The human action recognition system and its relationship to BrocaÕs area: An fMRI study. Neuroimage, 19, 637–644. Jackson, P. L., & Decety, J. (2004). Motor cognition: A new paradigm to study self-other interactions. Current Opinion in Neurobiology, 14, 259–263. Jacob, P., & Jeannerod, M. (2005). The motor theory of social cognition: A critique. Trends in Cognitive Sciences, 9, 21–25. Jeannerod, M. (2003). The mechanisms of self-recognition in humans. Behavioural Brain Research, 142, 1–15. Jellema, T., Baker, C. I., Wicker, B., & Perrett, D. I. (2000). Neural representation for the perception of the intentionality of actions. Brain and Cognition, 44, 280–302. Kampe, K. K. W., Frith, C. D., & Frith, U. (2003). ‘‘Hey John’’: Signals conveying communicative intention toward the self activate brain regions associated with ‘‘mentalizing’’, regardless of modality. Journal of Neuroscience, 23, 5258–5263. Kohler, E., Keysers, C., Umilta`, M. A., Fogassi, L., Gallese, V., & Rizzolatti, G. (2002). Hearing sounds, understanding actions: Action representation in mirror neurons. Science, 297, 846–848. Leube, D. T., Knoblich, G., Erb, M., Grodd, W., Bartels, M., & Kircher, T. T. J. (2003). The neural correlates of perceiving oneÕs own movements. Neuroimage, 20, 2084–2090. McCabe, K., Houser, D., Ryan, L., Smith, V., & Trouard, T. (2001). A functional imaging study of cooperation in twopersons reciprocal exchange. Proceedings of the National Academy of Sciences, USA, 98, 11832–11835. Mesulam, M. M. (1981). Dissociative states with abnormal temporal lobe EEG. Multiple personality and the illusion of possession. Archives of Neurology, 38, 176–181. Rizzolatti, G., Fadiga, L., Gallese, V., & Fogassi, L. (1996). Premotor cortex and the recognition of motor actions. Cognitive Brain Research, 3, 131–141.
74
C. Becchio et al. / Consciousness and Cognition 15 (2006) 64–74
Ruby, P., & Decety, J. (2001). Effect of subjective perspective taking during simulation of action: A PET investigation of agency. Nature Neuroscience, 4, 546–550. Ruby, P., & Decety, J. (2003). What you believe versus what you think they believe: A neuroimaging study of conceptual perspective-taking. European Journal of Neuroscience, 17, 2475–2480. Searle, J. (1983). Intentionality: An essay in the philosophy of mind. New York: Cambridge University Press. Shiffrar, M., & Freyd, J. J. (1990). Apparent motion of the human body. Psychological Science, 1, 257–264. Sirigu, A., Daprati, E., Pradat-Diehl, P., Franck, N., & Jeannerod, M. (1999). Perception of self-generated movement following left parietal lesion. Brain, 122, 1867–1874. Spence, S. A., Brooks, D. J., Hirsch, S. R., Liddle, P. F., Meehan, J., & Grasby, P. M. (1997). A PET study of voluntary movement in schizophrenic patients experiencing passivity phenomena (delusions of alien control). Brain, 120, 1997–2011. Stevens, J. A., Fonlupt, P., Shiffrar, M., & Decety, J. (2000). New aspects of motion perception: Selective neural encoding of apparent human movements. NeuroReport, 1, 109–115. Tai, Y. F., Scherfler, C., Brooks, D. J., Sawamoto, N., & Castiello, U. (2004). The human premotor cortex is ÔmirrorÕ only for biological actions. Current Biology, 14, 117–120. Tuomela, R. (in press). We-intentions revisited. Philosophical Studies. Tuomela, R. (2002). The philosophy of social practices: A collective acceptance view. Cambridge: Cambridge University Press. Tuomela, R., & Bonnevier-Tuomela, M. (1997). From social imitation to teamwork. In G. Holmstro¨m-Hintikka & R. Tuomela (Eds.), Contemporary action theory, Vol. II: Social action (pp. 1–47). Dordrecht and Boston: Kluwer Academic Publishers. Walter, H., Adenzato, M., Ciaramidaro, A., Enrici, I., Pia, L., & Bara, B. G. (2004). Understanding intentions in social interaction: The role of the anterior paracingulate cortex. Journal of Cognitive Neuroscience, 16, 1854–1863.