M. Jeannerod, C. Farrer, N. Franck, P. Fourneret, A. Posada, E. Daprati,
N. Georgieff.
Institut des Sciences Cognitives,
67 Boulevard Pinel, 69675, Bron, France.

Abstract.
The ability to attribute an action to its proper agent and to understand its meaning when it is produced by someone else are basic aspects of human social communication. Several psychiatric symptoms, such as those of schizophrenia, relate to a dysfunction of the awareness of one's own action as well as of recognition of actions performed by others. Such syndromes thus offer a framework for studying the determinants of the sense of agency, which ultimately allows to correctly attribute actions to their veridical source. The Chapter will report a series of experiments in normal subjects and schizophrenic patients dealing with the recognition of actions. The basic paradigm used in these experiments was to present the subject with simple actions which may or may not correspond to those they currently execute. Systematic distortions have been introduced, such that the threshold for accepting an action as one’s own could be determined. In normal subjects, this threshold is relatively high, indicating the existence of a specific mode of processing for action signals, independent from visual processing used in other perceptual activities. In schizophrenic patients, this threshold is further increased, with a strong tendency to self-attribute actions which do not correspond to those they have performed. The results reveal a clear distinction between patient groups with and without hallucinations and/or delusions of influence. Influenced patients show a higher rate of self-attributions. These results point to schizophrenia and related disorders as a paradigmatic alteration of a "Who ?" system for action monitoring and self-consciousness.

1. Introduction. The sense of agency.
The mechanism by which one becomes aware of one's own actions, and one can distinguish them from those of other people, is a critical one, for a number of reasons. First, the ability to recognize oneself as the agent of an action is one way by which the self builds as an entity independent from the external world. Second, the ability to attribute an action to its proper agent is a prerequisite for establishing social communication. Ultimately, functions such as understanding the meaning of an action and inferring the intention of its agent may be determined by these more elementary abilities.
Being aware of one’s own actions is not a straightforward process. Following up Gallagher (2000), awareness of action may involve at least two components: indeed, an agent may be aware of the fact that he is moving (the sense of ownership) without being necessarily aware that he has a causal role in controlling the movement (the sense of agency). Although in many circumstances, the two components coincide, there are situations where they dissociate. This is the case of movements caused by external forces, like reflex responses, for example, where the subject feels being the owner of the movement without feeling that he is the author of the response. In experimental conditions, movements caused by brain stimulation may enter this definition. Finally, pathological conditions offer many examples of abnormal movements due to impairments of the motor system, and the cause of which lays outside a conscious sense of agency. To illustrate this difference between sense of ownership and sense of agency it may be interesting to briefly refer to classical discussions among physiologists in the XIXth century. Some of them, like Alexander Bain, thought that the conscious sensation arising from a movement had its sole origin in the efferent discharges to the muscles, in the “outgoing stream of nervous energy”. Others, however, expressed a more balanced view. Duchenne de Boulogne, for example, distinguished between what he called “muscular consciousness”, originating from voluntary muscular contractions, and “muscular sense”, the set of sensations generated by the displacement of the limb. He conjectured that muscular consciousness could exist independently of muscular sensations, in patients with complete anaesthesia of one limb, for example. In a similar vein, Lewes introspectively considered that the complex experience arising from a voluntary movement was the sum of both the “sense of effort” and the “sense of effect” (for references, see Jeannerod, 1993). According to the above definition, the sense of effort and the sense of effect would superimpose with Gallagher’s sense of agency and sense of ownership, respectively.
Another point concerns the contribution of the sense of agency to the ability to attribute an action to its proper agent. As can be inferred from the previous paragraph, the sense of agency should be based on the presence, during the execution of a movement, of internal signals, or cues related to the intention of the agent. According to a widely accepted hypothesis, these cues contribute to building an internal model which represents the action to be performed in order to reach the desired goal. During execution of the movement, the end result is compared with the internal model by way of the peripheral signals (e.g., visual, proprioceptive) arising from the moving limb and from its interactions with the external environment. This comparison process allows for determining the degree of completion of the movement and its degree of match with the desired goal (e.g. Jeannerod, 1990; Wolpert et al, 1995). The term of “central monitoring” has been used to designate the ensemble of processes by which an agent becomes aware of his causal role in an action (see Frith, 1992). Thus, the presence of internal cues should normally induce a judgement of self attribution. By contrast, in their absence, no sense of agency should be experienced by the subject, and the movement should be attributed to another agent. This discussion reminds the classical theories on the role of corollary discharges in distinguishing self-generated events from events arising from the external world (Sperry, 1950). These central to central discharges have been thought to arise from the action generation system and to inhibit (or cancel out) the effects of a self produced action on the sensory systems, such that this action cannot be attributed to an external origin. Note that the corollary discharge concept does not imply that the action is executed: In the case of paralysis of the effector, the corollary discharge is nonetheless present and its effects can be “felt” by the subject, by way of an illusory displacement of the visual world, for example.
The distinction between self generated movements and movements produced by other agents, and the corresponding attribution judgements, however, is not always that simple. Which renders this distinction difficult is the existence of situations where an agent is not actually executing an action, but where representations of actions are nevertheless present in his brain. Consider first the frequent case where the action is imagined or intended, but not executed. Here, the available information for self attributing the action is entirely internal to the self, based on the internal model which has been built in correspondence to the agent’s intention. But, even if the goal of the movement remains a virtual one, and no possibility exists for matching the end result with the internal model, the above internal cues which are responsible for the sense of agency should nevertheless be present and the action should be correctly attributed to the self. Another situation, which bears a close resemblance to that of an intended or imagined action, is observation of an action performed by someone else. In that case, it can be conjectured that the observer also builds a representation of the action he sees, which he will use for understanding its meaning, inferring the agent’s intentions and, ultimately, establishing social communication. In principle, because this representation is built from external information and no internal signals are generated, no sense of agency should arise and the action should be correctly attributed to the external agent.
 There is a potential confound between these two situations, however, for both conceptual and empirical reasons. First, conceptually, recent theories on represented actions (e.g. intended, imagined or observed) have insisted on the idea that these actions undergo the same processing as effectively executed actions, except that they are not executed. Accordingly, many of the properties of motor images have been tentatively explained by the possibility that they would be “quasi-actions” blocked from execution by an additional, inhibitory mechanism (Jeannerod, 1994). Concerning observed actions, the theory goes on saying that they would activate within the observer’s brain the same mechanisms that would be activated, were that action intended or imagined by that observer. In turn, this representation in the brain of the observed movements influences the interpretation made by the observer. In other words, the observer would use an implicit strategy of putting himself “in the shoes of the agent”. This theory (see Gallese and Goldman, 1998 for review), which posits that actions performed by others can be understood by an observer to the extent that they can be “simulated” by that observer, would represent the basis for a broad spectrum of cognitive functions, starting from understanding others’ actions, up to learning by observation and imitation.
Second, this concept of simulation is now used, in neural terms, as an explanatory framework for the finding that both forms of motor representations (imagined and observed actions) involve a subliminal activation of the motor system (Jeannerod, 1999, Jeannerod and Frak, 1999). Brain mapping experiments (using PET or fMRI) show activation of a cortical and subcortical network during motor imagery and action observation. This network involves structures directly concerned with motor execution, such as motor cortex, dorsal and ventral premotor cortex, lateral cerebellum, basal ganglia ; it also involves areas concerned with action planning, such as dorsolateral prefrontal cortex and posterior parietal cortex. A recent meta-analysis of these data (Grèzes and Decety, 2000) reveals that the degree of overlap between different modalities of representations varies from one cortical area to another. Concerning primary motor cortex itself, fMRI studies unambiguously demonstrate that pixels activated during contraction of a muscle are also activated during imagery of a movement involving the same muscle (Roth et al, 1996). No such activation has yet been found during action observation. However, the involvement of primary motor pathways in action observation was demonstrated using a direct measurement of corticospinal excitability by transcranial magnetic stimulation (TMS) of motor cortex (Fadiga et al, 1995). In premotor cortex and SMA, the overlap between imagined and observed actions is almost complete, as it is also in posterior parietal cortex. By contrast, action observation largely involves inferotemporal cortex, which is not the case for action imagination. Although the cortical networks pertaining to each form of covert (or represented) actions do not entirely overlap with each other, there are large cortical zones which are common to both. This relative similarity of neurophysiological mechanisms accounts for both the fact that actions can normally be attributed to their veridical author, and that action attribution remains a fragile process. Indeed, there are in everyday life ambiguous situations where the cues for the sense of agency become degraded and which obviously require a subtle mechanism for signaling the origin of an action. As the experiments below will show, situations can be created where normal subjects fail to recognize their own actions and misattribute to themselves actions performed by another agent. This may also be the case of situations created by interactions between two or more individuals (e.g., joint attention, matched actions, or mutual imitation), or situations pertaining to the domain of man-machine interactions (e.g., telemanipulation, virtual reality systems, etc). A neural mechanism for explaining these difficulties will be presented at the end of the Chapter.
Pathological conditions offer many examples of misattributions: a typical case is that of schizophrenia. Among the wide range of manifestations characterizing this disease, the so-called “first-rank symptoms” have been considered critical for its diagnosis. According to Schneider (1955), these symptoms refer to a state where patients interpret their own thoughts or actions as due to alien forces or to other people and feel being controlled or influenced by others. First-rank symptoms might reflect the disruption of a mechanism which normally generates consciousness of one’s own actions and thoughts and allows their correct attribution to their author. Thus a study of attribution behavior in schizophrenic patients would not only help understanding the factors responsible for misattribution in the patients, but also shed light on this critical function in normal life.
The pattern of misattributions due to agency disturbances in schizophrenic patients is twofold. Patients may first attribute to others rather than to themselves their own actions or thoughts (underattributions); second, patients may attribute to themselves the actions or thoughts of others (overattributions). According to the French psychiatrist Pierre Janet (1937), these false attributions reflected the existence in each individual of a representation of others’ actions and thoughts, in addition to the representation of one’s own actions and thoughts: false attributions were thus due to an imbalance between these two representations. A typical example of underattributions is hallucinations. Hallucinating schizophrenic patients may show a tendency to incorporate external events in their own experience, or to interpret environmental cues as specifically directed to themselves. Accordingly, they may misattribute their own intentions or actions to external agents. During auditory hallucinations, the patient will hear voices that are typically experienced as coming from a powerful entity trying to monitor and control his own behavior. The voices are often comments where the patient is addressed in the third person, and which include commands and directions for action (Chadwick and Birchwood, 1994). The patient may declare that he or she is being acted upon by an alien force, as if his thoughts or acts were controlled by an external agent. The so called mimetic behavior observed at the acute stage of psychosis also relates to this category.
The reverse pattern of misattribution can also be observed. Overattributions were early described by Janet (1937): what this author called “excess of appropriation” corresponded for the patient to the illusion that actions of others are in fact initiated or performed by him/her and that he/she is influencing other people. In this case, patients are convinced that their intentions or actions can affect external events, for example, that they can influence the thought and the actions of other people. Accordingly, they tend to misattribute the occurrence of external events to themselves. The consequence of this misinterpretation would be that external events are seen as the result expected from their own actions. This type of errors by overattribution is an exaggeration of what can be observed in normal subjects who, according to Nielsen (1963), also attribute to themselves actions performed by others when they are presented in ambiguous conditions.
In the next sections, several possible explanations for these attribution impairments will be explored. The problem will be to determine whether they relate to a general cognitive deficit, or to a more specific problem dealing with action recognition per se. Under the light of the mechanisms described in the previous section, the hypothesis has been developed that schizophrenics would fail in monitoring their willed intentions, including those related to the expression of thought. In other words, these patients would fail to attribute elements from which they form their goals and plans, to their real origin. The consequence of this failure in self monitoring would make them unable to disentangle actions arising from the external world, from those generated as a consequence of their own cognitive functioning. This hypothesis will be further developed and integrated within a different framework, which includes a prominent, and perhaps neglected, aspect of schizophrenia, namely its intersubjective dimension.

2. Can the attribution deficit in schizophrenia be attributed to an unspecific cognitive impairment ?
Investigators using neuropsychological methods for testing cognitive abilities in schizophrenic patients have established the deficience in explicit and conscious modalities of processing information as a dominant characteristic of the disease (see Kuperberg and Heckers, 2000, for review). Accordingly, such patients have consistently been shown to poorly perform in cognitive tasks where working memory is likely to be involved, like learning abstract sequences (Dominey and Georgieff, 1997), management of rules (Laws, 1999), attention tasks (Bernard et al, 1997), processing of context (Servan-Schreiber et al, 1996), or semantic processing (Kuperberg et al, 1998). This behavioral evidence, however, is not entirely supported by neuroimaging data. Because in normal subjects working memory has often been associated with activation of the frontal lobes (Koechlin et al, 1999; Prabhakaran et al, 2000), one should expect that schizophrenic subjects should not show this pattern of cortical activity. As a matter of fact, fMRI studies in patients do not systematically show frontal lobe deactivation during working memory tasks (Stevens et al, 1998; Manoach et al, 1999). It has been claimed that the difference in frontal activation with respect to normal subjects is too modest to support the idea of a frontal lobe dysfunction (Zakzanis and Heinrich, 1999). Future work in this field will have to focus on broader task-related neural networks linking cortical (Fletcher, 1999) and subcortical areas (Andreasen, 1999). Dysfunction of these networks, including defective inhibition or disinhibition might be responsible for the working memory impairment.
It remains that such an impairment would probably affect the ability to correctly interpret the origin of an action. Recognition of a self-produced action implies that its goal and potential consequences are anticipated by the agent. Anticipation, in turn, requires short-term storage of the parameters encoded during the early stages of action execution, or even prior to execution. Self attribution of the action will rely on the degree of match between these parameters and the final appearance of the action. This type of predictive behavior is at work in nearly all human everyday activities.
Examining the ability of schizophrenic patients to anticipate forthcoming events and to monitor their own actions (and, by extension, those of other people), therefore seems to be a legitimate enterprise. One possible experimental approach to this problem consists in monitoring responses of subjects in tasks where they can manipulate advanced information either provided to them or acquired through learning. The ability to produce anticipatory responses, which relies on conscious processing of the relevant information should reflect the elementary process which ultimately allows predictive behavior in natural situations.

2.1. An experimental study of anticipation in schizophrenia.
Anticipation was studied using a simple situation where subjects had to learn a sequence of colors appearing on a computer screen: A subject who knows that the colors appear in a fixed order learns the sequence voluntarily, simply by watching the computer screen and repeating the color names verbally. If, during this process of learning, the subject is instructed to press a specific key each time a given color appears on the screen, the time to press the key sharply decreases after the complete sequence has been consciously disclosed and verbalized. As a matter of fact, keys are now pressed before the next color is shown.
A group of 20 schizophrenic patients, and a group of matched control subjects ran experiments where stimuli were presented as temporal sequences (Posada et al). Patients (6 women and 14 men) had a mean age of 36,4 years ? 7,9 and a mean educational level of 10,8 years ? 2,2. They were classified as paranoid (n=10), undifferentiated (n=5), residual (n=4) and disorganized (n=1) according to DSM-IV criteria. All patients were clinically stable at the time of evaluation and testing. Schizophrenic symptoms were assessed by the Scale for Assessment of Positive Symptoms (SAPS) (Andreasen, 1984) and the Scale for Assessment of Negative Symptoms (SANS) (Andreasen, 1983). The total SAPS score was 28,8 ? 15,2, and the total SANS score was 36,5 ? 22,4.
All subjects were first tested in a condition where they knew the existence of the sequence and had to learn it ; subsequently, after they had acquired the sequence, they were tested for their ability to anticipate their responses with respect to the appearance of the stimuli. Although patients were found to be able to acquire the sequence almost normally, they proved to be impaired in using their explicit knowledge to produce anticipatory responses
Subjects were seated in front of a computer screen where sequences of colored (yellow, green or blue) rectangles were displayed. Whenever a rectangle appeared, subjects were instructed to push the corresponding color button on a button box. Response time was recorded. For each trial, a feedback was given after the subject’s response. In the first condition (Sequence learning), the subject was instructed to try to disclose the sequence and to report it verbally to the experimenter. Whenever a subject could not find the sequence after 7 blocks, he/she was helped by the experimenter until he/she discovered the sequence. In the Anticipation conditions, the repetitive sequences were explicitly given to the subject before the experiment. The subject received the instruction to try to push the correct buttons before the next element of the sequence appeared on the screen.
In the first condition of the experiment (Learning), the number of blocks necessary to disclose and to explicitly report the sequence was determined. Subjects who failed to disclose the sequence after 7 blocks are reported as “failures”. Schizophrenic subjects needed 3,4?1,6 blocks (n=14; 6 failures) to find the temporal sequence. Controls needed 2,6?1,61 blocks (n=18; 2 failures) to find the temporal sequence. No significant difference was found in the number of blocks between groups for the temporal sequences (t-test, p=0,2).
No significant difference in RTs between groups was found. In block 1, when subjects completely ignored the sequence, the two groups have nearly identical RTs, showing that schizophrenic subjects did not differ from normal subjects for what concerns pure reaction times to visual stimuli. In the following three blocks (4-6) where subjects knew the sequence and received the instruction to anticipate, control subjects showed a highly significant decrease in RTs with respect to schizophrenic subjects. The RTs in control subjects dropped down to 280 ms, but only to 680 ms for the schizophrenic group. In a final, random sequence block (R), where no anticipation was possible, RTs grew for both groups, but significantly less so in schizophrenics than in controls.
In the other condition (Anticipation), where subjects explicitly received the sequence information at the beginning of  the test, the response pattern was very similar between groups, but RTs were higher for schizophrenic subjects. Control subjects had mean RTs between 100-200 ms and schizophrenic subjects between 600-700 ms, except in the easiest condition, where the RTs were around 350 ms. In both groups, response times remained stationary during the blocks with instruction to anticipate (blocks 1-7). The ANOVA analysis showed a highly significant difference between groups but no interaction between groups and blocks. Finally, RTs increased for the random sequence block (R). Although this increase was much more marked in the control group, the fact that the schizophrenic patients also increased their response times for the random sequence clearly indicates that they were still presenting some degree of anticipation.
The group mean of the percentages of anticipation (e.g., the mean difference between RTs for the anticipation blocks and RTs in the random sequence block) is shown in Figure 1. There was a significant reduction of the capacity of anticipation in schizophrenic subjects. Control subjects showed percentages of anticipation between 70-90% and schizophrenic subjects, between 30-50% only.

2.2. A working memory impairment ?
The above results reveal a highly specific deficit in anticipatory behavior in the group of schizophrenic patients. This deficit cannot be the consequence of impairments in elementary perceptual or motor functions. The patients had no difficulty performing the basic sensorimotor task of associating a color displayed on the computer screen with the corresponding color button; they also showed normal values of reaction times when they responded to presentation of colors before knowing the sequence. The number of errors remained low, not significantly different from that of control subjects. In addition, all patients were able, more or less rapidly, to learn and to remember the target sequences in each of the experimental conditions. Although control subjects typically required two blocks to acquire the sequence, patients needed three blocks. Failures (inability to acquire the sequence after 7 blocks) were about three times more frequent in patients than in controls: however, those who had not been able to disclose the sequence by themselves, when properly trained by the experimenter, were able to memorize it. These results indicate that, although patients learned more slowly than controls, they had retention abilities compatible with task execution.
The major impairment in the schizophrenic patients in dealing with the learned color sequences appeared when the instruction to respond in anticipation to the colors was introduced. In this condition, patients could only partly reduce their response times in comparison with their performance prior to learning. Whereas control subjects produced response times in the range of 100-200 ms (i.e., well below typical reaction times), patients were barely able to perform better than 600-700 ms. Note, however, that some degree of anticipation persisted in patients, as demonstrated by the fact that, in all conditions, their responses times were shorter than their purely reactive response times in the random sequence blocks. Thus, the patients in this experiment, although they were strongly impaired in using the explicit knowledge they had available about the sequences to anticipate the occurrence of the next event, had retained the ability to understand and to perform the task.
To anticipate an incoming event is a cognitive process which requires the knowledge of regularities in the temporal unfolding of external events. In the particular process of anticipating the next incoming color, subjects must use simultaneously the notion of a repetitive sequence, the sequence information stored in memory, the instruction to anticipate, and the perception of the color of the stimulus. All these components are integrated to produce a motor representation which activate the motor command to push the correct button. This integrative process, which has received considerable interest in neuropsychology, fits the definition of the “working memory” which operates at the interface between memory, attention and perception (Baddeley, 1998).

2.3. The behavioral and clinical consequences of lack of anticipation
A deficit in using available information to correctly anticipate an incoming event likely represents an explanatory framework for some of the pathological aspects of schizophrenic behavior. Such a deficit may be particularly deleterious in the domain of action, where a lack of anticipation may create situations where the consequences of actions are not properly evaluated. When planning (or intending) to move his/her hand to the right, for example, a normal individual anticipates to see and feel it moving in that direction. If the anticipation process is impaired, the direction in which the hand is seen to move may appear to be unrelated to the desired direction. This might be the sort of situations schizophrenic patients are faced with: in the above example, the patient may feel that his/her hand was displaced by an external agent, or even that it was an alien hand. When asked whom this hand belongs to, or who was the author of this movement, the patient may adopt different strategies: he may attribute the hand and/or the movement to an external agent, or alternatively he may use a default strategy by attributing the movement to himself. Both strategies correspond to frequently observed clinical symptoms in certain categories of schizophrenic patients. In other studies to be described below we have repeatedly shown that such patients tend to over-attribute to themselves movements performed by other agents (Daprati et al, 1997) and have increased thresholds for detecting movements differing from those they have actually performed (Franck et al, 2001). Lack of anticipation would thus preclude the normal match of actions with their internal representations, with the consequence that self performed actions would not be recognized and would be misattributed. This hypothesis appears to be complementary with the more general theory of Gray et al (1991) who suggested that such patients may fail to use, in processing new information, stored regularities about the external world: hence their inability to anticipate forthcoming events (see also Frith et al, 2000).
It may appear tempting to generalize this reasoning to other schizophrenic symptoms. An abnormal time integration has been early proposed to participate in the production of schizophrenic symptoms (Minkowski, 1927). Impairment of patients in matching executed and represented actions could lead to the production of positive symptoms, such as incongruous actions or hallucinations, as well as to negative symptoms like impossibility to act, apathy, or lack of social interactions. It remains, however, that defective working memory and defective anticipation cannot represent a sufficient explanation for schizophrenic symptoms: it has also been found in other, non psychotic, categories of patients (e.g., patients with frontal lobe lesions), who show no attribution problems.

3. Dysfunction of a specific mechanism for recognizing action.
Whether a lack of anticipation can represent the final explanation for misattribution of actions in schizophrenic patients thus seems highly doubtful. Misattributions are far from being evenly distributed among the patients. Instead, this type of symptom is observed preferentially (if not exclusively) in one particular class of patients. For this reason, we will now investigate an alternative possibility, that of the disruption of a specific system for perceiving, recognizing and attributing actions.
Our argument is based on the existence of the already mentioned ‘Schneiderian symptoms’ in schizophrenic patients. These symptoms include insertion of thought, auditory-verbal hallucinations, delusion of reference, delusions of alien control. They represent false beliefs which lead to a feeling of depersonalisation by impairing the distinction between the self and the external world. The fact that these symptoms pertain to the realm of action is strongly supported by a set of clinical and experimental arguments based on the study of hallucinations. It has been known from some time that auditory verbal hallucinations in schizophrenic patients are related to the production of speech by the patient. Some hallucinated patients even show muscular activity in their laryngeal muscles (e.g., David, 1994). Thus, auditory verbal hallucinations represent a typical example of misattribution, where patients perceive their inner speech as voices arising from an external source. Experiments using neuroimaging techniques have greatly contributed to this problem by studying brain activation during incurring hallucinations, or during inner speech in patients predisposed to hallucinations and subjects experiencing no hallucinations. The results show that, during hallucinations (as signalled by the patients), brain metabolism is increased in the primary auditory cortex (Heschl gyrus) on the left side (Dierks et al, 1999), as well as in the basal ganglia (Silbersweig et al, 1995). By contrast, subjects predisposed to hallucinations show decreased activity in speech temporal areas during inner speech and auditory verbal imagery, as compared to other subjects (McGuire et al, 1996). The clearest result from these conflicting data is that, during verbal hallucinations, the auditory temporal areas remain active, which suggests that the nervous system in these patients behaves as if it were actually processing the speech of an external speaker. Patients perceive their own thinking as originating from the outside world. This explanation would be consistent with the idea (e.g., Frith et al, 2000) that the normal mechanism for attributing thought to its internal origin is a comparison between the executive commands leading to speech and the anticipated sensory consequences of these commands. Inner speech would normally be accompanied by a mechanism that decreases activity at the level of primary auditory cortex, perhaps via an inhibitory projection from the frontal lobe. Because in verbal hallucinations the inner speech is usually not uttered, this hypothesis would be consistent with the notion that the sense of agency must be functioning even in the absence of comparison with external reafferences, and that actions can be monitored at the level of their representation, not only at the level of their execution. As a matter of fact, in clinical practice the Schneiderian symptoms almost exclusively concern non executed actions. Hallucinations, once considered as a perception without an object, should therefore be re-evaluated as an action without a subject.
Other types of hallucinations, such as mental automatism, and delusions of alien control might also correspond to an impairment of recognition of action. Spence et al (1997) examined cortical activity in schizophrenic patients with experience of delusional control. During the scan, the patients were required to voluntarily move a joystick and to freely select the direction of the movement. Most of them reported vivid experiences of alien control when performing the motor task. Brain activation was found to be increased in a cortical network including the left premotor cortex and the right inferior parietal lobule and angular gyrus, at the level of areas 40 and 39. This right parietal hyperactivity in deluded subjects is particularly interesting : it is noteworthy that lesions at this level frequently result in altered awareness (neglect) for the contralateral limbs and space, and denial of the disease (anosognosia) ; conversely, transient hyperactivity (during epileptic fits for example) may produce impressions of an alien phantom limb (see Spence et al, 1997).
In an effort to quantify the degree of misattribution in schizophrenic subjects, we designed experimental situations where subjects had to produce agency judgements about hand movements that were shown to them and that corresponded, or not, to their own movements. These experiments are described below.

3.1. A pilot study of agency in normal and schizophrenic subjects.
In this section, we will present data from an experiment where the performance of groups of schizophrenic patients was compared to those of matched controls (Daprati et al, 1997). A situation was created where the subjects were shown movements of a hand of an uncertain origin, that is, a hand that could equally likely belong to them or to someone else, using a paradigm directly borrowed from the study of voluntary movement by Nielsen (1963). Subjects were instructed to explicitly determine whether they were the author of the hand movements they saw. In order to give such a response, they had to use all available cues to compare the current movement of their unseen hand with the movement that was displayed to them.
 Subjects. Sixty subjects participated in the study, including 30 schizophrenic patients and 30 normal control subjects. Among the patients, 18 received the diagnosis of paranoid schizophrenia, 7 of indifferentiated schizophrenia, and 5 of residual schizophrenia. Hallucinations were present in 13 patients. Delusion of control was reported by 7 patients. Delusion of control and hallucinations were both present in 6 cases. Psychiatric symptomatology was quantified according to Andreasen’s rating scales for positive (SAPS) and negative (SANS) symptoms (Andreasen, 1983, 1984, respectively). All patients were under medication during the period of the study.
Apparatus and procedure. The Subject’s hand and the Experimenter’s hand were filmed with two different cameras. By changing the position of a switch, one or the other hand could be briefly (5 seconds) displayed on the video screen seen by the subject. A mirror with an inclination of 30 degrees in the vertical plane was placed at 40 cm from a table, and at 35 cm from the subjects' frontal plane. Subjects positioned their right hand on the table, below the mirror. A camera filmed the subject's hand, the image of which appeared on a TV-screen located on a shelf, 15 cm over the subject's head, and was reflected by the mirror. Thus, looking at the mirror, subjects got the impression that they watched their own hand as through a window. A second camera placed in another part of the room, filmed the experimenter’s hand. The display allowed the experimenter to exactly match the image of her hand with that of the subject’s hand before the beginning of each trial. The experimenter’s and the subject’s hands were covered with identical gloves, in order to minimise the effects of  gross morphological differences.
The task for the subjects was to perform a requested movement with their right hand, and to monitor its execution by looking at the image in the mirror. At the beginning of each experimental trial, a blank screen was presented. An instruction to perform a movement was given and the subject and the experimenter had to execute the requested movement at an acoustic signal. Once the movement was performed and the screen had returned blank, a question was asked to the subject : “You have just seen the image of a moving hand. Was it your own hand ? Answer YES if you saw your own hand performing the movement you have been executing. Answer NO in any other case, that is if you doubt that it was your own hand or your own movement.”
One of four possible movements of the fingers was required in each trial: 1. extend thumb, 2. extend index, 3. extend index and middle finger, 4. open hand wide. One of three possible images of the hand could be presented to the subjects in each trial: 1. their own hand (condition: Subject), 2. the experimenter's hand performing the same movement (condition: Experimenter Same), 3. the experimenter's hand performing a different movement (condition: Experimenter Different). Twelve trials were run for each hand condition, that is 3 trials for each movement type. Altogether, 36 trials were run for each set of movements. Order of set presentation was counterbalanced across subjects, whereas movement type and hand presented were randomised in each set.
Results. A descriptive analysis of the results showed that hallucinating patients made more recognition errors (median value: 24) than non-hallucinating patients (11), and controls (5). Similarly, patients experiencing delusion of control made more recognition errors (median value: 24) than the rest of the patients (18), and controls (5). The present analysis focuses only on patients experiencing delusion of control. The total number of recognition errors was significantly different between groups (Kruskal-Wallis test). Errors occurred especially in the condition Experimenter Same, i.e., in the trials where the hand on the screen performed the same movement that was required from the subject but was not his/her own, the median for error rate was 5 in the control group, 17 in the non-delusional group and 23 in the delusional group, whereas virtually no errors occurred when subjects saw their own hand, or a hand performing a different movement. Differences between the three groups for the condition Experimenter Same showed a strong trend to significance (comparison controls vs non-delusional p=0.006; controls vs delusional p=0.001; non-delusional vs delusional p=0.07, Mann-Whitney U test). All these differences were clearly significant by using the criterion of hallucinations instead of delusions.
Discussion. In this experiment, normal control subjects were able to unambiguously determine whether the moving hand seen on the screen was theirs or not, in two conditions. First, when they saw their own hand (trials from the condition Subject), they correctly attributed the movement to themselves. Second, when they saw the experimenter's hand performing a movement which departed from the instruction they had received (condition Experimenter Different), they denied seeing their own hand. By contrast, their performance degraded in the condition Experimenter Same, that is, in trials where they saw the experimenter's hand performing the same movement as required by the instruction: in this condition, they misjudged the hand as theirs in about 30% of cases.
It is also in that particular condition that the rate of incorrect responses increased in schizophrenic patients. The error rate amounted to 77% in the group of patients with hallucinations or 80% in the group with delusional experiences, whereas in the non-hallucinating group, it was around 50%. The fact that all patients gave nearly correct responses in the other two conditions (the error rate remained within 1-7%) shows that the effect observed in trials Experimenter Same was not due to factors unrelated to the task, such as lack of attention. The explanation for this effect therefore should be found in a deficit of the mechanism which is normally used for controlling and recognizing one's own movements.
The pattern of responses that we recorded in our condition Experimenter Same (in both normal controls and schizophrenics) could be explained by the paucity of movement related cues available to the comparator mechanism. Because the subject's (invisible) hand and the experimenter's (visible) hand both executed the same movement, some mismatch occurred between the anticipated and the perceived final hand postures. The only available cues were the dynamic signals generated during the movements themselves. Slight differences in timing and kinematic pattern between the intended movement and that perceived by visual and kinaesthetic channels had to be used in order to give the correct response. A decrease in sensitivity of this mechanism would explain the greater difficulties met by the schizophrenic patients. Even normal subjects misjudged the ownership of the experimenter's hand in about 30% of trials. This finding suggests that the mechanism for recognizing actions and attributing them to their true origin operates with a relatively narrow safety margin: In conditions where the visual cues are degraded or ambiguous, it is barely sufficient for making correct judgements about the origin of action, although it remains compatible with correct determination of agency in everyday life.

3.2. A parametric study of agency in normal and schizophrenic subjects.
The above experiment by Daprati et al demonstrates that the clinical difficulty in identifying the origin of an action observed in patients with delusion of influence can be experimentally provoked. The procedure used in this experiment, however, did not allow to determine those cues, used by the normal subjects to give correct attribution responses, that were missing in the influenced schizophrenic patients. Another experiment was therefore designed to answer this question (Franck et al, 2001). Using a situation similar to that of Daprati et al, a realistic virtual hand was used, instead of the hand of an experimenter, which was superimposed to the subject’s hand. Not only did this device allowed more standard experimental conditions; it also allowed systematically distorting the movements of the virtual hand with respect to those of the subjects’ hand. The results fully support the existence of impaired attributions of action in schizophrenia, especially in influenced patients.
Subjects. Twenty-nine schizophrenic subjects and 29 normal controls participated in the study. Patients were selected according to the DSM IV criteria. The schizophrenic and control groups did not differ significantly for age, sex, laterality and educational level. During the experiment itself, 5 schizophrenic subjects revealed unable to correctly perform the task and, for this reason, were not further included in the study. None of them were influenced or hallucinated. Accordingly, all data reported below only bear on the 24 patients who completed the task. Fourteen of the 24 patients were hospitalized at the time of the experiment. Eleven patients met the criteria for paranoid schizophrenia, 2 for disorganized schizophrenia, 9 for undifferentiated schizophrenia and 2 for residual schizophrenia. The mean average disease duration was 9.8 years. All patients were under treatment with antipsychotic medication. All patients underwent clinical assessment with the already mentioned Scale for Assessment of Positive Symptoms (SAPS) and the Scale for Assessment of Negative Symptoms (SANS). In addition, a passivity phenomena sub-scale score was defined, which consisted in items 15 to 19 of the SAPS. This sub-scale allowed classifying the patients as influenced or non-influenced. At the time of testing, 6 patients presenting a passivity phenomena sub-scale score superior to 2 were classified as “influenced”. The remaining 18 “non-influenced” patients scored 2 or less at this sub-scale. Finally, the patients underwent neuropsychological testing to assess their spatial perception abilities. To this aim, the Birmingham Object Recognition Battery (BORB) (Riddoch and Humphrey, 1993) was used. The patients’ performances were within normal range.
Methods and Procedure. During the experiment an image of an electronically reconstructed hand was presented to the subjects. A specially designed program run on a PC computer synthesized pictures of a hand holding a joystick according to the real position of a joystick actually held by the subject and connected to the computer. This design allowed the dynamic representation of the movements of the joystick held by the subject with an intrinsic delay inferior to 30 ms. Temporal or angular biases could be introduced in this representation (see below), modifying the apparent direction or the degree of synchrony of the movement actually performed by the subject with respect to the movement displayed on the computer screen. The computer monitor on which the virtual image appeared was placed face down on a metallic support. A horizontal mirror, located 18 centimeters below the monitor screen, reflected the image. The joystick was placed below the mirror on a table supporting the apparatus, so that the subject’s hand holding the joystick was located approximately 18 cm below the mirror. Thus, when subjects looked at the mirror, they saw the image of a virtual hand moving a joystick just above their own unseen hand actually doing that.
Subjects held the joystick with their right hand, with their elbow resting on the table. The position of their forearm was adjusted so as to coincide with the direction of the virtual forearm seen in the mirror. Subjects were instructed to maintain fixed the position of their fingers on the joystick and to restrict their movements to the wrist joint. The task consisted in executing a series of simple movements with the joystick from the resting position, either in the straight ahead direction, to the right or to the left. Each trial started with a dark screen. A green spot was displayed for 1 second on the left, on the right or on the top of the screen. The image of the virtual hand then appeared for 2 seconds during which the subjects had to execute a movement of the joystick in the direction indicated by the position of the green spot. Immediately after the trial, subjects had to answer the question: “Did the movement you saw on the screen exactly correspond to that you have made with your hand”? They had to answer YES or NO.
Three categories of trials were used: 1. Neutral trials: movements of the virtual hand seen on the screen exactly replicated those made by the joystick. 2. Trials with angular bias: movements of the virtual hand were deviated by a given angular value with respect to those made by the joystick. Seven values of angular bias from 5° to 40°, either to the right or to the left were used. 3. Trials with a temporal bias: movements of the virtual hand were delayed by a given time with respect to those made by the joystick. Seven values of temporal bias from 50ms to 500ms were used. Each trial with a temporal bias was run 4 times for each of the three directions of movement (N=84); trials with an angular bias were run two times with a bias to the right and 2 times with a bias to the left for each of the three directions of movement (N=84). Finally, neutral trials were run 12 times. Each subject therefore executed a total of 180 trials. The order of presentation of the 180 trials was randomized before the participation of each subject. Identical trials could not be presented twice in a row.
Results. Verbal responses of the subjects were recorded. According to whether trials were with or without bias, subjects could potentially make 2 types of errors: YES responses in trials with a bias, and NO responses in neutral trials. The maximum number of errors was 12 for the neutral trials and 84 for the trials with an angular or a temporal bias. Presentation of the results below will focus on the YES responses, for the reason that this type of response reflects the subjects ability to recognize a movement as their own.
Control subjects and patients gave YES (correct) responses in nearly all neutral trials. The median value of erroneous NO responses was equally small (N=1) for the control subjects and for the influenced and non-influenced patients subgroups. The distribution of YES responses for the biased trials, although it clearly differed between groups as will be shown below, kept a relatively similar pattern across groups. In both control subjects and patients, the number of YES responses was higher for the smaller temporal and angular biases, and became lower as the biases increased. In other words, only the shape of the distribution differed between groups.
Influenced schizophrenic patients gave globally more YES responses than non-influenced schizophrenic patients and normal controls in both the trials with angular (median values, influenced patients: 56.5; non-influenced patients: 39; controls: 33) and temporal bias (median values, influenced patients: 53.5; non-influenced patients: 49.5; controls: 29). The Median test on YES responses revealed that the differences between groups were significant for both the trials with angular bias (Chi2=7.67, p=0.022) and with temporal bias (Chi2=20.49, p<0.0001). The Mann-Whitney U-tests on global scores of responses revealed that, whereas influenced patients produced significantly more errors than controls in both trials with angular and temporal biases, non-influenced patients significantly differed from controls in trials with temporal bias only. However, the two groups of patients differed significantly in their number of errors in the trials with an angular bias.
Figure 2 reveals a further important characteristic of these results, namely that the differences between the control group and the two patients groups varied as a function of the amplitude of the biases. It shows the number of YES responses for trials with an angular bias. Whereas non-influenced patients show a sharp decrease in erroneous YES responses (down to 50% of maximum number of errors) already for a bias between 15° and 20°, a value not very different from that of controls, influenced patients do not reach the same score until the bias increased to 30°- 40°. The results were different for the temporal bias. Whereas control subjects show a clear decrease in YES responses for a relatively small bias (100-150ms), both influenced and non-influenced patients follow a similar trend and do not show a decrease in the rate of YES responses until the bias reaches 300ms.
Pairwise comparisons for each class of trials were performed using the Mann-Whitney U-test. This comparison showed that both groups of patients produced significantly more errors than the control group in the trials with a temporal bias for delays longer than 100 ms. In the trials with an angular bias, the difference with the control group was significant for angles larger than 10° for the influenced patients and for the 30° and 40° angles only for the non-influenced patients
Discussion. In this experiment, control subjects still recognized as their own a movement delayed by up to 150 ms with respect to the movement they actually executed. Similarly, if normal subjects saw their movement rotated from its actual trajectory by about 15 degrees, they still accepted it as their own. This result shows that the accuracy for detecting the features of one’s own movement is limited, and that this limitation is far beyond perceptual thresholds of the visual system for detecting temporal gaps or angular deviations (see Fourneret and Jeannerod, 1998). These results clarify the findings of Daprati et al (1997) where normal subjects failed in about 30% of cases to recognize an alien hand as distinct from their own hand, when the two hands performed nearly synchronous movements. As delays between the two hands and movement trajectories were not measured in Daprati et al’s experiment, one can only speculate on those cues that were missing for the subjects to identify a hand as alien. However, the present results indicate that these misattributions were likely to occur when the delays between the two movements were below 150ms, or their trajectories deviated from each other by less than 15 degrees.
In the present experiment, the patients were clearly worse than controls at recognizing as distinct from their own, movements that were delayed or deviated. All 24 schizophrenic patients responded at chance when a time delay up to 300ms was introduced. For angular deviations, only the influenced patients responded at chance up to 30° whereas non-influenced patients presented an error rate comparable to normal controls, i.e., became aware of the angle around 15°. Defective perceptual or attentional factors seem to be ruled out by the fact that all patients (including the influenced ones) performed well in the BORB test, indicating that they had retained a normal ability to discriminate small angular differences. In the case of temporal biases, one could argue that schizophrenic patients are known to be slow in many tasks and that their reaction times are globally increased, a feature which can be categorized among the negative symptoms (Cadenhead et al, 1997). In fact, our temporal delay condition is quite different from a reaction time task (see below). Finally, the fact that the same patients performed differently in the two tasks is a good indication that they were not influenced by unspecific factors when giving the responses.

4. Neural constraints on action recognition in the social context.
Experimental manipulations of the appearance of one’s own movements impair the correct self-attribution of these movements, an effect which is dramatically increased in schizophrenic patients. It seems likely to attribute these difficulties in detecting differences between self-produced and externally produced movements, in the absence of other obvious visual deficit, to an impairment of a specific neural system devoted to perception of actions produced by living organisms and singularly by human beings, and operating during interactions between several people.

4.1. Perception of biological movements.
One of the prerequisites for the functioning of such a system is that it can distinguish biological movements from those produced by mechanical devices. Biological movements involve specific kinematic properties and regularities which make them recognizable to other people (Johansson, 1977). This property is illustrated by recent psychophysical experiments. These experiments tested the perception of apparent motion from static pictures of an object presented at different locations. If the object is a human figure (for example, alternating static pictures of a model with his arm positioned behind his head and in front of his head) the time between the two static pictures will have to increase up to 400 ms in order for the observer to report a biologically possible trajectory (the arm moving around the head). Below this threshold, the observer will report the most direct trajectory (the arm across the head) (Shiffrar and Freyd, 1990, Ramachandran et al, 1998, Stevens et al, 2000). This is consistent with the present results showing that the attribution system operates with a temporal and spatial resolution relatively low with respect to the capabilities of the visual system.
The effect of introducing a temporal delay between the observed movement and the actual movement can be discussed first. As stated above, this impairment is different from slowness to respond; it expresses a difficulty in the perception of slight temporal differences. Recent findings in schizophrenic patients might represent a rationale for the difficulty met by the patients from the present study: such patients are unable to discriminate between two different velocities of a moving vertical grating (Chen et al, 1999) and they cannot discriminate fast motions of a moving dot (Schwartz et al, 1999). Such inabilities might not be unique to schizophrenic patients: indeed, similar results have been obtained in autism (Gepner et al, 1995). Thus, there is a possibility that psychotic patients in general are unable to efficiently integrate time cues in their perception of movement, which would account for our finding that all patients in the present study, irrespective of their clinical symptomatology, were impaired in detecting delays in their own movements. Although this impairment probably contributes to the high rate of misattributions observed here, it may not represent the core of the problem, mainly because it does not differentiate between influenced and non influenced patients.
The deficit in detecting movement direction might be a better determinant of difficulties in attribution met by these patients. One of the main findings of the experiments of Franck et al (2001) is that only influenced patients were impaired in attributing movements with angular biases. Coding the direction of a movement is indeed a critical condition for an agent to precisely reach an object in peripersonnal space. Perceiving the direction of a movement is also a useful information for an observer to understand the action of the agent of this movement: during a movement, the arm points to the goal of the action and its direction may reveal the intention of the agent. It is thus not surprising that a patient deprived of this information will misinterpret the intention displayed by others in their movements, and that this will have consequences on understanding interactions between people. In the present study, the fact that influenced patients tended to self attribute movements, the direction of which was distorted, illustrates this problem. In Daprati et al’s experiment (Daprati et al, 1997) also, influenced patients were worse than non influenced ones at differentiating the movements they executed with their own hand, from movements executed more or less synchronously by another hand. The discrepancies between movements performed by the two agents, although they were perceptible to normal subjects and, to a large extent, to non influenced patients, were far too small, as we know from the above parametric study, to be detected by influenced patients.
The situations used in our experiments privileged self attribution responses. Only one hand was present at a time to the patient: thus, the situation always referred to the patient as the agent of the action. If another agent were clearly involved in the situation, this might leave the possibility for the patient to produce the opposite type of response, i.e., a response of underattribution. Further experiments are presently underway for testing this possibility.

4.2. A neural hypothesis for action recognition: the ‘Who’ system.
We have therefore created the experimental conditions for the study of social cognition and its disorders in psychotic patients. In this concluding section, we briefly present a framework for integrating social cognition to the neural substrate. Our present conception of action recognition (Georgieff and Jeannerod, 1998, Jeannerod, 1999) is based on the existence of neural networks subserving the various forms of representation of an action. Accordingly, each representation entails a cortico-subcortical network including to a various extent activation of interconnected neural structures. Some of these networks have been described in an earlier section of this chapter. Although these ensembles are clearly distinct from one form of representation to another (e.g., the representation of a self-generated action vs the representation of an action observed or predicted from another agent), they partly overlap: posterior parietal and premotor areas, for example, are activated during both. When two agents socially interact with one another, this overlap creates “shared representations”, i.e., neural structures that are simultaneously activated in the brains of the two agents. In normal conditions, however, the existence of non-overlapping parts, as well the existence of possible differences in intensity of activation between the activated zones, allows each agent to discriminate between representations activated from within from those activated from outside, and to disentangle which belongs to him from that which belongs to the other. This process would thus be the basis for correctly attributing a representation (or the corresponding action) to the proper agent or, in other words, for answering the question of ‘Who’ is the author of an action. The flow chart of Figure 3 is a tentative illustration of the many interactions between two agents. Each agent builds in his brain a representation of both his own intended actions, using internal cues like his own beliefs and desires, and the potential actions of the other agent with whom he interacts. These partly overlapping representations are used by each agent to built a set of predictions and estimates about the social consequences of the represented actions, if and when they would be executed. Indeed, when an action comes to execution, it is perceived by the other agent as a set of social signals which confirm (or not) his predictions and possibly modifies his beliefs and desires.
 This conception allows making hypotheses about the nature of the dysfunction responsible for misattribution of actions by schizophrenic patients. Changes in the pattern of cortical connectivity could alter the shape of the networks corresponding to different representations, or the relative intensity of activation in the areas composing these networks. Several examples in relation with verbal hallucinations or alien control of movements have been described in the present chapter. Unfortunately, little is known on the functional aspects of cortical connectivity underlying the formation of these networks and, a fortiori, on their dysfunction in schizophrenia. Among the relevant studies, several have pointed to the prefrontal cortex as one of the possible sites for perturbed activation : not only is it hypoactive in many patients (Weinberger and Berman, 1996) ; its morphological aspect has also been shown to be modified on post-mortem examination (Goldman-Rakic and Selemon, 1997). Because prefrontal areas are known to normally exert an inhibitory control on other areas involved in various aspects of motor and sensorimotor processing, an alteration of this control in schizophrenic patients might result in aberrant representations for actions. Referring to the diagram in Figure 3, one of the two agents would become “schizophrenic” if, due to an alteration in the pattern of connectivity of the corresponding networks, the degree of overlap between the representations in his brain increased in such a way that the representations would become undistinguishable from each other. The pattern of misattribution in this patient would be a direct consequence of this alteration: for example, decreased self attribution if frontal inhibition were too strong, or increase if it were too weak.
 Other cognitive deficits, like reduced memory span and inability to anticipate forthcoming events, are also likely to contribute to the impairment in action recognition. However, they cannot, by far, represent the only explanation to this problem. It could be proposed that a lack of anticipation impairs the comparison of an executed action with its internal representation and may therefore render self-attribution of that action more difficult. This explanation, however, cannot hold for misattribution of non executed actions, which is the most frequently observed pattern in schizophrenic patients.

References

Andreasen, N.C. (1983). The scale for the assessment of negative symptoms (SANS). The University of Iowa, Iowa City, IA.

Andreasen, N.C. (1984). The scale for the assessment of positive symptoms (SAPS). The University of Iowa, Iowa City, IA.

Andreasen, N.C. (1999). A unitary model of schizophrenia. Archives of General Psychiatry, 56, 781-7.

Baddeley, A. (1998). Recent developments in working memory. Current Opinion in Neurobiology 8, 234-238.

Bernard, D., Lancon, C., & Bougerol, T. (1997). Attention models in evaluating schizophrenia. Encephale, 23(2), 113-8.

 Cadenhead K.S., Geyer, M.A., Butler, R.W., Perry, W., Sprock, J. & Braff, D.L. (1997) Information visual processing deficits of schizophrenia patients: relationship to clinical ratings, gender and medication status. Schizophrenia Research, 28:51-62

Chadwick, P., Birchwood, M. (1994). The omnipotence of voices. A cognitive approach to auditory hallucinations. British Journal of Psychiatry, 164, 190-201.

Chen, Y., Palafox, G.P., Nakayama, K., Levy, D.L., Matthysse, S., Holzman, P.S. (1999) Motion perception in schizophrenia. Archives of General Psychiatry, 56:149-154

Danion, J.M., Rizzo, L., & Bruant, A. (1999). Functional mechanisms underlying impaired recognition memory and conscious awareness in patients with schizophrenia. Archives of General Psychiatry, 56, 639-44.

Daprati, E., Franck, N., Georgieff, N., Proust, J., Pacherie, E., Dalery, J., & Jeannerod, M.  (1997). Looking for the agent: an investigation into consciousness of action and self-consciousness in schizophrenic patients. Cognition, 65, 71-86.

David, A.S. (1994). The neuropsychological origin of auditory hallucinations. In  David AS, Cutting JC (Eds) : The neuropsychology of schizophrenia. Lawrence Erlbaum, Hove, pp. 269-313.

Dierks, T., Linden, D.E.J., Jandl, M., Formisano, E., Goebel, R., Lanferman, H. & Singer, W. (1999) Activation of the Heschl's gyrus during auditory hallucniations. Neuron, 22: 615-621

Dominey, P., & Georgieff, N. (1997). Schizophrenic learn surface but not abstract structure in a serial reaction time task. Neuroreport, 8, 2877-2882.

Fadiga, L., Fogassi, L., Pavesi, G. & Rizzolatti, G. (1995). Motor faciulitation during action observation. A magnetic stimulation study. Journal of Neurophysiology, 73, 2608-2611.

Fleming, K., Goldberg, T.E., Binks, S., Randolph, C., Gold, J.M., & Weinberger, D.R.,  (1997). Visuospatial working memory in patients with schizophrenia. Biological Psychiatry,  41, 43-49.

Fletcher, P. (1999). Abnormal cingulate modulation of fronto-temporal connectivity in schizophrenia. Neuroimage, 9, 337-42.

Fourneret, P., Jeannerod, M. (1998). Limited conscious monitoring of motor performance in normal subjects. Neuropsychologia, 36, 1133-1140.

Franck, N., Farrer, C., Georgieff, N., Marie-Cardine, M., Daléry, J., d'Amato, T. & Jeannerod, M. (2001) Defective recognition of one's own actions in schizophrenia. American Journal of Psychiatry (in press)

Frith, C.D. (1987). The positive and negative symptoms of schizophrenia reflect impairments in the perception and initiation of action. Psychological Medicine, 17, 631-648.

Frith, C.D. (1992). The neuropsychology of schizophrenia. Lawrence Erlbaum, Hove.

Frith, C.D., Blakemore, S. & Wolpert, D.M. (2000). Explaining the symptoms of schizophrenia: abnormalities in the awareness of  action. Brain Research Reviews, 31, 357-63.

Gallagher, S. (2000) Philosophical conceptions of the self: implications for cognitive science. Trends in Cognitive Science, 4, 14-21.

Gallese V. & Goldman, A. (1998) Mirror neurons and the simulation theory of mind reading. Trends in Cognitive Science, 2: 493-501

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

Gepner, B., Mestre, D., Masson, G. & de Schonen, S. (1995) Postural effects of motion vision in young autistic children. Neuroreport, 6: 1211-1214

Goldman-Rakic, P.S. (1994). Working memory dysfunction in schizophrenia. Journal of Neuropsychiatry & Clinical Neurosciences, 6, 348-57.

 Goldman-Rakic, P.S. & Selemon, L.D. (1997) Functional and anatomical aspects of prefrontal pathology in schizophrenia. Schizophrenia Bulletin, 23, 437-458.

Gray J.A., Feldon J., Rawlins J.N.P., Hemsley, D.R. & Smith, A.D. (1991). The neuropsychology of schizophrenia. Behavioral and Brain Sciences, 14 : 1-84.

Grèzes, J. & Decety, J. Functional anatomy of execution, mental simulation, and verb generation of actions: A meta-analysis

Janet, P. (1937) Les troubles de la personnalité sociale. Annales Médico-Psychologiques, tome II, 149-200

Jeannerod, M. (1990) The representation of the goal of an action and its role in the control of goal-directed movements. In: Computational Neuroscience, E. Schwartz (Ed),. Cambridge: MIT Press, pp 352-365.

Jeannerod, M. (1993) A theory of representation driven actions. In: The perceived self: ecological and interpersonal sources of self-knowledge. U. Neisser (Ed), Cambridge, Cambridge University Press, pp 89-101.

 Jeannerod, M. (1994). The representing brain. Neural correlates of motor intention and imagery. Behavioral and Brain Sciences, 17, 187-245.

 Jeannerod, M. (1999) To act or not to act: perspectives on the representation of actions. Quarterly Journal of Experimental Psychology, A 52:1-29

 Jeannerod, M. & Frak, V. (1999) Mental imaging of motor activity in humans. Currrent Opinions in Neurobiology, 9:735-9

Johansson, G. (1977). Studies on visual perception of locomotion. Perception, 6, 365-376.
Koechlin, E., Basso, G., Pietrini, P., Panzer, S., & Grafman, J. (1999). The role of the anterior prefrontal cortex in human cognition. Nature, 399, 148-151.

Kuperberg, G. & Heckers, S. (2000). Schizophrenia and cognitive function. Current Opinion in Neurobiology, 10,  205-10.

Kuperberg, G.R., McGuire, P.K., & David, A.S. (1998). Reduced sensitivity to linguistic context in schizophrenic thought disorder: evidence from on-line monitoring for words in linguistically anomalous sentences. Journal of Abnormal Psychology, 107, 423-34.

Laws, K.R. (1999). A meta analytic review of Wisconsin card sort studies in schizophrenia: general intellectual deficit in disguise? Cognitive Neuropsychiatry, 4, 1-35.

Manoach, D.S., Press, D.Z., Thangaraj, V., Searl, M.M., Goff, H.E., Halpern, E., Saper, C.B., & Warach, S. (1999). Schizophrenic participants activate dorsolateral prefrontal cortex during a working memory task, as measured by fMRI. Biological Psychiatry, 45, 1128-37.

 McGuire, P. K., Silbersweig, D. A. & Frith, C. D. (1996) Functional neuroanatomy of verbal self-monitoring. Brain, 119, 907-917.

McGuire, P. K., Silbersweig, D. A., Murray, R. M., David, A. S., Frackowiak, R. S. J. & Frith, C. D. (1996). Functional anatomy of inner speech and auditory verbal imagery. Psychological Medicine, 26, 29-38.

Minkowski, E. (1927). La schizophrénie. Psychopathologie des schizoïdes et des schizophrènes.  Editions Payot & Rivages.

Nielsen, T. (1963) Volition: A new experimental approach. Scandinavian Journal of Psychology,  4:225-230

Riddoch, M.J. & Humphreys, G.W. (1993) BORB: Birmingham Object Recognition Battery. Hove, Lawrence Erlbaum Associates.

Posada, A., Franck, N., Georgieff, N. & Jeannerod, M. Anticipating incoming events: An impaired cognitive process in schizophrenia.

Prabhakaran, V., Narayanan,, K., Zhao, Z., & Gabrieli, D.E. (2000). Integration of diverse information in working memory within the frontal lobe. Nature Neuroscience, 3, 85-89.

Ramachandran, V.S., Armel, C., Foster, C. & Stoddard, R. (1998) Object recognition can drive motion perception. Nature, 395:852-853

Roth, M., Decety, J., Raybaudi, M., Massarelli, R., Delon-Martin, C., Segebarth, C., Morand, S., Gemignani, A., Décorps, M. & Jeannerod, M. (1996). Possible involvement of primary motor cortex in mentally simulated movement. A functional magnetic resonance imaging study. Neuroreport, 7, 1280-1284.

Schneider, K. (1955) Klinische Psychopathologie. Stuttgart, Thieme Verlag.

Schwartz, B.D., Maron, B.A., Evans, W.J. & Winstead, D.K. (1999) High velocity transient visual processing deficits diminish ability of patients with schizophrenia to recognize objects. Neuropsychiatry, Neuropsychology and Behavioural Neurology, 12:170-7

Servan-Schreiber, D., Cohen, J.D., & Steingard, S. (1996). Schizophrenic deficits in the processing of context. Archives of General Psychiatry, 53, 1105-1112.

Shiffrar, M. & Freyd, J. (1990) Apparent motion in the human body. Psychological Science, 1:257-264

Silbersweig, D.A., Stern, E., Frith, C., Cahill, C., Holmes, A., Grootoonk, S., Seaward, J., McKenna, P., Chua, S.E., Schnorr, L., Jones, T. & Frackowiak, R.S.J. (1995) A functional neuroanatomy of hallucinations in schizophrenia. Nature, 378, 176-179.

Spence, S.A., Brooks, D.J., Hirsch, S.R., Liddle, P.F., Meehan, J. & Grasby, P.M. (1997) A PET study of voluntary movements in schizophrenic patients experiencing passivity phenomena (delusions of alien control). Brain, 120, 1997-2011.

Sperry, R.W. (1950) Neural basis of the spontaneous optokinetic response produced by visual inversion. Journal of Comparative and Physiological Psychology, 43: 482-489.

Stevens, A.A., Goldman-Rakic, P.S., Gore, J.C., Fulbright, R.K., & Wexler, B.E. (1998).  Cortical dysfunction in schizophrenia during auditory word and tone working memory demonstrated by functional magnetic resonance imaging. Archives of General Psychiatry, 55, 1097-103.

Stevens, J.A., Fonlupt, P., Shiffrar, M. & Decety J. (2000) New aspects of motion perception: selective neural encoding of apparent human movements. NeuroReport, 11:109-115

Weinberger, D.R. & Berman, K.F. (1996) Prefrontal function in schizophrenia : confounds and controversies. Philosophical Transactions of the Royal Society, B, 351, 1495-1503.

Wexler, B.E., Stevens, A.A., Bowers, A.A., Sernyak, M.J., & Goldman-Rakic, P.S. (1998).  Word and tone working memory deficits in schizophrenia. Archives of General Psychiatry,  55, 1093-6

Wolpert, D.M., Ghahramani, Z. & Jordan, M.I. (1995) An internal model for sensorimotor integration. Science, 269:1880-1882

Zakzanis, K.K., & Heinrichs, R.W. (1999). Schizophrenia and the frontal brain: a quantitative review. Journal of international neuropsychological Society, 5(6), 556-66.
 
 

Figure legends

Figure 1
Percentage of anticipatory responses given by control subjects and schizophrenic patients (Posada et al). Subjects were tested in several versions of the color sequence experiments. T: the color sequence is presented sequentially; S: the colors appear sequentially in different spatial locations; Easy and Difficult refer to sequences with few or many elements, respectively).
 

Figure 2
A parametric study of attribution responses in control subjects and schizophrenic patients (Franck et al, 2001). Single movements performed by subjects are shown to them through a mirror. The appearance of the movement may be deviated from its actual course. Subjects are instructed to answer YES if they consider that the movement they saw corresponds to the one they executed. At 0° (no deviation), all subjects correctly respond YES. Control subjects tolerate deviations of up to 15° before they consistently reject the movement they see as corresponding to their actual movement. One group of schizophrenic patients (with delusions of influence) continue to give YES responses up to 30° deviations.
Figure 3
A flow chart of the interplay between two agents. For explanation, see text.