From brain resonance to intersubjectivity:
Are we by now bridging the gap?
Université Strasbourg II & LPPA, Collège de France
Encyclopaideia, XII (23), 97116, 2008, ISSN 1590492X 97
Cognitive neuroscience established the existence of patterns of neural functioning in brain that could be construed as causal correlates of behaviours. Far from being limited to sensorimotor connexions inside one isolated brain, the research stretches out to larger scale behaviours, which imply interacting organisms. Overcoming the traditional isolated brain model, a new ‘social neuroscience’ is beginning to emerge today. It will not be limited to a naturalization of the individual mind but will also seek to naturalize social ‘spirit’, including the institutions, primarily language, of culture. This ambitious project relies on the bringing to light of ‘resonant systems’ by means of registering cerebral activity in one subject or more, and in the future a whole group of subjects performing a task at the same time. Such systems are supposed to correlate homological regions of the brains of different individuals in the performance of nonlinguistic activities. However, on that basis it is tempting to extrapolate in the direction of a general theory of the communication of mental states, – inclusive by language. In this paper, we evaluate the chances of this enterprise.
Keywords: mirror neurons resonating systems social acts social neuroscience
Le recenti neuroscienze cognitive hanno dimostrato con successo l’esistenza di pattern del funzionamento neurale in specifiche regioni del cervello che potrebbero essere interpretati come le correlazioni causali di una lunga lista di interazioni comportamentali. Ben lontano dall’essere limitate alle connessioni sensomotorie inputoutput astrattamente considerate dentro un cervello isolato, il nuovo campo della ricerca pone l’accento in larga scala sui comportamenti che implicano organismi interattivi. Riconoscere come atto un movimento corporeo percepito, imitare qualcuno, percepire un’emozione su di un viso, empatizzare con stati dolorosi di sofferenza, sentire se stessi esclusi e essere angosciati da ciò, ecc, tali stati mentali o condizioni non possono che essere descritte come interazionali o di carattere interpersonale. Allo stato attuale ognuna di queste condizioni mentali è già stata localizzata in distinti e ricorrenti modelli di attivazione dei circuiti neurali. E procedendo su questa traccia verso altri effetti impercettibilmente contagiosi o influenze reciproche, dovute dal fatto di stare vicini o congiuntamente attivi con altre persone, queste attitudini piuttosto straordinarie come leggere il labiale o comprendere le intenzioni inespresse di qual cuno, dimostrano di non essere senza un loro personale tocco materiale nel cervello. Questo progetto ambizioso si appoggia alla scoperta dei “sistemi risonanti” (resonant systems) grazie alle tecniche sperimentali impiegate per registrare l’attività cerebrale in uno o più soggetti, e, in un prossimo futuro, un intero gruppo di soggetti che compie un compito allo stesso tempo.
Si suppone che tali sistemi correlino regioni omologhe del cervello di individui differenti nel compimento di attività nonlinguistiche, quali atti portati a termine e compresi, intenzioni formate e percepite, dolori sofferti e osservati, emozioni esperite e riconosciute. Tuttavia, basandosi su questi contesti essenzialmente nonlinguistici, si sta tentando di estrapolare una teoria generale della comunicazione degli stati mentali.
Il fatto che i risultati delle ricerche sono sempre più numerose suggerisce ad alcuni di annunciare la nascita di una nuova “neuroscienza sociale”. Questa nuova neuroscienza sociale, superando il modello tradizionale di cervello isolato, non sarà limitato a una neutralizzazione della mente individuale, ma cercherà anche di naturalizzare lo “spirito” sociale, includendolo sotto le strutture della cultura, per primo il linguaggio. Mentre si rinnova la vecchia ambizione psicofisiologica del XIX secolo, i ricercatori empirici di oggi sembrano inconsapevoli del criticismo lasciato senza risposta di queste stesse ambizioni avanzate dai fenomenologi dell’essenza degli atti sociali. Con questo ci riferiamo ad Adolf Reinach, nella sua discussione sulla proposta di Theodor Lipps di una genesi naturalistica dei doveri sociali basata sull’Einfühlung, la capacità di proiettare se stessi in un altro.
Mettendo da parte ogni domanda sul fondamento delle loro argomentazioni, queste vorrebbero procedere e rendere naturale in termini di biologia del cervello, le scienze umane, includendo i loro più apparentemente irriducibili dipartimenti, come le scienze sociali, includendo quindi gli studi della conversazione e del discorso, radicati, come sono, nelle scelte arbitrarie degli agenti personali. Tuttavia, accettando l’esistenza dei sistemi risonanti nel cervello, le presunte basi delle corte distanze casuali, influenze contagiose nelle cure quotidiane, non ci dispensano dal prendere seriamente i profondi radicamenti dell’individuazione personale. Un’individuazione per cui la libera creazione di volontari, arbitrari, nonnaturali collegamenti tra persona e persona diventa un tentativo di ricompensazione con il fine di rendere la vita sociale possibile. Collegamenti esemplificati in convenzionali istituzioni della cultura che richiedono la comunicazione intersoggettiva attraverso atti verbali, che sono allo stesso tempo visibili gesti della bocca e essenzialmente scommesse non garantite che giacciono sulla affidabilità di ciascheduno. La struttura logica delle regole costitutive di tali interazioni comunicative, costruite come giochi linguistici, sono state studiate da Daniel Vanderveken, nel suo recente ampliamento della teoria di John Searle sugli atti linguistici. Cioè non che la realtà sociale possa stare per aria, o risiedere su un’impalcatura di formali, nella fattispecie, apriori impedimenti messi sulla categorizzazione di enunciazioni quali atti del tal tipo logico o del tal altro. Lontano da ciò, i pattern di attività nel cervello di un osservatore dovuto alla diretta unilaterale, contingente influenza di un agente individuale non possono essere la sua base ontologica. Ciò deve ancora essere fondato. Il nostro suggerimento è che potrebbe essere cercato preferibilmente nel reciproco carattere anticipatorio degli atti sociali sulla scia di una intersoggettivamente sincrona emulazione interna di una Lebenswelt che è degna di esistere non solo per me, ma anche per un’intera comunità.
Parole chiave: neuroni specchio sistemi di risonanza atti sociali neuroscienze sociali
Recent cognitive neuroscience successfully established the existence of patterns of neural functioning in specific regions of brain tissue that could be construed as the causal correlates of quite impressive a list of behaviours. Far from being limited to inputoutput sensorimotor connexions abstractly considered inside one isolated brain, the new domain of research stretches out to larger scale behaviours which imply interacting organisms. Overcoming the traditional isolated brain model, a new ‘social neuroscience’ is beginning to emerge today. It will not be limited to a naturalization of the individual mind but will also seek to naturalize social ‘spirit’, including there under the institutions, primarily language, of culture.
This ambitious project relies on the bringing to light of ‘resonant systems’ by means of experimentation techniques employed to register cerebral activity in one subject or more, and in the near future a whole group of subjects performing a task at the same time. Such systems are supposed to correlate homological regions of the brains of different individuals in the performance of non linguistic activities, such as: actions accomplished and comprehended, intentions formed and perceived, pains suffered and observed, emotions experienced and recognised. But on the basis of these nonlinguistic contexts it is tempting to extrapolate in the direction of a general theory of the communication of mental states – inclusive by language.
Mirror neurons: What do they amount to?
In 1992, Giacomo Rizzolatti and his equip of the Institute of Physiology of Parma University discovered a new class of visuomotor neurons in the frontallateral cortex of the monkey (Di Pellegrino et al., 1992)1. These neurons are characterized by the fact that they are activated in a similar way by the execution of a grasping act and also in a visual stimulus condition consisting in the observation of a grasping act made in front of the monkey by another agent. Such bimodal activity was interpreted as providing a neural correlate for the perceptual recognition by the monkey in the observed behaviour of another agent of ‘the same action’ as one of its own ecological repertoire. The actions in question were, in fact, only parts of the feeding movements of the animal, such as picking a raisin between thumb and index, seizing a banana with full hand, wringing something, putting something in the mouth, etc.
The extrapolation from monkey to man and the demonstration of the existence of a mirror system with humans has been confirmed first by using transcranial stimulation, then thanks to several series of experiments with brain imagery (Rizzolatti et al., 1996; Grafton et al., 1996; Grèzes et al., 1998). At first, PET scan, a technology which limitations only permitted a gross localization of ‘a mirror system’ recruiting coordinate activities in the frontallateral and parietal areas during the observation and execution of grasping an object. Later on, fMRI imagery confirmed the somatotopic organization of this pattern of brain activity along the somatosensory and motor Rolando cortices, as a proof of its dealing with an action observation – action execution functional link (Buccino et al., 2001). What one has become accustomed to call today ‘the’ mirror system, in fact, covers several cerebral areas in the form of a network, more specifically a network bearing on quite specific parts of the premotor dorsal and ventral cortex and the upper temporal sulcus. According to the research protocols employed, these areas are differently described: sometimes as coding the repertory of actions of the individual, sometimes as a network underlying motor imitation and learning, sometimes as the neural foundations of empathy, of verbal communication and of culture.
Mirror systems: What can they (actually) do?
Some dimensions of the possible role of mirror neurons were enhanced by the following. First, in the monkey the anticipatory character of action was traced back to mirror neurons which selectively fire during the end part of the trajectory of an observed hand catching an object when this catching occurs behind a screen. These neurons are supposed to endow the agent with predictive capabilities for the targeting of objects by hand movements. Second, in man, learning is known to modulate the understanding of the other’s actions by specifying the motor repertoire of the observer. Such influence was allegedly confirmed in mirror neuron brain imagery (fMRI): viewing videos of ballet movements induces higher brain activations in dancers than viewing capoeira movements (CalvoMerino, et al., 2005). However, we do not know if what is actually a street fighting technique reshuffled in tourist attraction can even be perceived as a dance style. Again, the possible role of a mirror system in the recovery of the intention of an observed action was evidenced by comparing the pattern of brain activity induced in the observer by a hand taking a cup of tea and the pattern of brain activity induced by the viewing of a dressed table before and after tea (Iacoboni, et al., 2005). In this experiment, the methodology of image subtraction in fMRI brain imagery may have biased towards a shallow theory of intention as the mere result of subtracting an action from its context. Recently, the enhancing of covert differences in brain activities through the technique of transcranial magnetic stimulation highlighted the motor nature of the neural correlate of pain perception. Viewing an alien hand being painfully pricked with a syringe and the subject feeling the pain in one’s own hand: such a condition proved to induce a selective inhibition of the corticospinal motor representation of the observer’s hand muscle homologous to the observed subject’s lesioned muscle and no comparable effect in the rest of the observer’s hand map (Avenanti, et al., 2005). Summing up so disconnected, even if impressive, apercus on brain functioning, it remains difficult to form a coherent image.
It is all the more needed to avoid any uncritical extrapolation from the monkey’s gesture in taking a seed and putting it in his mouth to the totality of human relations with others. Even the proposition advanced by M. Arbib and Rizzolatti, concerning the possibility of a direct relation between the mirror system and the appearance of language, due to the proximity of the premotor area containing the mirror neurons and the Broca area (Rizzolatti & Arbib, 1998; Stamenov & Gallese, eds. 2002)2, is the subject of an intense debate. It should be added that this mechanism is certainly not the only one to function in the resonant mode. For example, another network relatively distinct from the first is implied in the perception and the observation of the other’s emotions (Adolphs, 2002; Keysers et al., 2004; Wicker et al., 2003; Grèzes & De Gelder, 2005), and these two networks do no doubt interact with each other, as well as with yet other networks in the emotional life of the individual. A condition which puts in jeopardy the unity some would like to lend to the mirror system.
The up to date stubborn question is what function or functions are those of mirror neuron ‘action simulation’, a rather favourite wording in the sector. Actually, despite that researchers seem not to wary much about the fact, too many responses are in the offing. Action simulation is deemed important in action understanding. Action simulation makes possible pain and expression of emotion understanding. Action simulation is necessary for prediction of future action outcomes. Action simulation permits perceptual categorization of movements as actions. Action simulation is a way of imitation and observational learning, etc. Particularly in this paper, we would like to learn if action simulation can possibly cope as well with social action as rulegoverned (not necessarily explicit or verbalized) interaction between truly personal agents, the qualified participants in human, institutionalized communities.
To put it bluntly, our guess is that the idea of tapping mirror neurons for the development of a new social neuroscience amounts to an attempted neural vindication of mimicry psychosociology. Let us be more specific. Recognizing a perceived bodily movement as an action exerted upon a thing, mimicking somebody, perceiving an emotion on a face, empathizing with pains being suffered, feeling oneself excluded and being distressed about that, etc.: for each one of these mental acts or attitudes distinct recurrent activation patterns in neural circuits have already been located in the brain. Notwithstanding the fact that such mental acts or attitudes are currently described as interactional or even interpersonal in essence we cannot but be stricken by the fact that they happen nowhere else than in the observer’s brain or mind. Refurbishing the forlorn ambitions of turn of XIXth to XXth century “Psychologie als Grundlage der sämtlichen Geisteswissenschaften” after Wundt (1896) and others, this social neuroscience would take the path of trying and naturalize in terms of brain biology the human sciences, including their apparently most irreducible departments, the social and linguistic ones, rooted as they are in arbitrary choices of signs, words and expressions as employed in verbal exchange by personal agents. However, acknowledging the existence of resonating systems in brain, the presumed basis of short distance casual interactions in daytoday business, does not dispense us from taking seriously the deeprootedness of personal, embodied individuation. And that constitutes a stumbling block in the way of a social neuroscience, as long as it misperceives the limits of personal individuation that perhaps it is in the illusion of having superseded.
Taking individuation seriously goes hand in hand with taking seriously the objectivity of the social, because the creating of voluntary, arbitrary, non natural links from person to person seems to be an attempt made by participants in human communities to compensate for their bodily and mental separateness. Links exemplified in conventional, law like institutions of culture require intersubjective communication through speech acts, that are at the same time visible buccal gestures made by human bodies and more fundamentally bets ultimately unwarranted laid on each other’s reliability. No neural correlates in perspective for such unrealities as these mutual exigencies, never to be completely fulfilled, not least because they ceaselessly renew themselves, which tightly streamline our daily transactions all the life long!
The logical empirical gap between neurons and the social
In order to carry through the programme of establishing not merely casual correlations between brains but genuinely social links (depending upon the competent use of language or the possibility of its use) on the basis of resonant phenomena in brains, a logical empirical gap has to be bridged. This gap opens up between 1) the empirical data on resonant systems and 2) the meaning structure of social acts: To (1) our only access is observational and inductive, through the correlate registering both of movements of bodies and of activation patterns in brain tissues. But to (2) we gain access in a uniquely reflective mode through logical analysis of the meaning constraints upon speech acts and language games in conversation. These two modes of approach seem to be incompatible with each other. Is it even possible for the sponsors of “A unifying view of the basis of social cognition” (Gallese, et al., 2004) to show us how one could possibly manage to bridge such a gap?
Mimicry: an action simulation?
At first sight, the most amenable behaviour to an explanation in terms of mirroring or resonant system is mimicry. Mimicry, an automatic, unconscious, unintentional tendency to movement contagion, clearly does not stand in need of language: watching someone yawn makes one yawn; watching someone laugh makes one laugh, etc. With typical social acts, the contrast could not be shaper. An order, for an example of speech act (not necessarily expressed in words), can be understood without there needing to be in the one who complies any tendency to imitate the one who issued it! Can, thus, mimicry be used as ‘a primitive form’ of action simulation bridging the gap between movement contagion and coordination of action in a group? The prospect seems rather bleak. Some cognitive scientists assume that a direct link between performed and observed actions is crucial for interpersonal action coordination and that mimicry might provide such link. But, granting that at least some social acts have to be directly accomplished, social acts generally speaking, are not so much direct as mostly indirect. Does the admission of an action simulation tendency succeed in coping with all those forms of behaviour for which the partisans of mirror systems sustain they provide the neural basis, especially when they are part and parcel of speech acts?
Now, we are told that subjects sitting side by side in rocking chairs unintentionally tend to rock at the same frequency when they are permitted to view each other, even if their chairs are preset with non synchronous rocking frequency (Schmidt & O’Brien, 1997; Richardson, et al., 2005). This trivial bodily tendency towards movement synchronisation was hypothesized to be precursory to joint action. Building up on top of such automatisms, true mimicry proved to be a perceivingperforming link mediated by motivation. Psychologists make themselves sure that their subjects acquire goals through subliminal priming. Once primed with affectively charged words: affiliate, friend, partner, together, etc. subjects significantly increase their mimicring while observing behavioural mannerisms, like face rubbing or foot shaking, in interviewers (Lakin & Chartrand, 2003). More, failed affiliation with a former interviewer increases mimicry of a new partner of communicative interaction. When a subject has such an unconscious desire to affiliate with an interlocutor, his tendency to mimicry him is strengthened. Motivation determines a stronger perceptionaction relationship, which determines a stronger rapport between partners. What mars such results, whatever their statistic reliability, is an uncritical usage of the term ‘affiliation’, which ambiguity blurs the difference between a mere affective bondage (liking or otherwise) and a fully fledged social link, most of which require no special ‘desire to create a rapport’ for their coming into effect.
Interpersonal action coordination on the wrong track
An apparently naïve, first question for a cognitive science tackling the social is: Are groups even able to acquire and can they learn to apply optimal, anticipatory strategies ‘as well as individuals’? Which kind of mechanism (if any) supports the interindividual coordination of actions with respect to a common goal? Without any suspicion concerning the heavily individualistic, solipsistic presupposition of their very mode of questioning, cognitive scientists go along gathering data. And, unsurprisingly, they find a lot of confirmatory data for their prejudiced conception of agency as fundamentally unique. Two agents performing a task alongside each other tend to internally represent the rule of action of their neighbour even when no interpersonal coordination is required. But, representing the other’s rule and acting on a different rule, rather logically, creates a conflict in the mind of both agents. As expected, this conflict slows down their reaction times relatively to solitary action condition (Sebanz, et al., 2005). This ‘shared representation’ theory (a misnomer) proves to be biased in favour of the separately representing mind. When two agents perform actions answering to incompatible rules, their motor cortices react as if each participant were inhibiting performing the other’s action. From 300 msec. on a brain wave characteristic of response inhibition develops a positive deflection (the opposite of activation curve) relative to baseline that is more accentuate than in the case of a solitary actor abstaining from acting. An argument in favour of similar mechanisms underlying the representations of each other’s rules of action in the brains of simultaneously, independently acting agents. Let us dispel once for all, the notion of an action at its richer when collective and not merely the addition of solitary agents (Sebanz, et al., 2003)!
Tracking tasks, for example, tracking a moving target with a cursor on a computer screen, are applied to generate conflicts between anticipatory and compensatory actions (Knoblich & Jordan, 2003). While individuals can so easily ‘resolve action conflicts internally’ – note the ante Wittgensteinan ingenuity! – solving action conflicts in groups has to be externally mediated: which implies recourse to such unpredictable means as appealing to an arbiter, meeting on a common ground, and agreeing upon collective goals. The additional demand in cognitive resources that it creates is accused of putting the issue in jeopardy: a natural explanation of the observed fact that joint performance is hardly ever successful. As we know from single agent action theory, these demands can be met by anticipating others’ actions and adjusting one’s own action planning accordingly. ‘Individuals’ (independent players who are in full command of the tracker) are the best at learning the anticipatory strategy. ‘Groups’ (two players, one of whom can only accelerate whereas the other can only break the tracker) differ in function of their receiving or not auditory feedback on each other’s key press. Groups whose participants know of the other’s actions tend to do nearly as well as individuals. Groups whose members lack any cues on each other’s actions keep trying the compensatory (worst) strategy. In sum, groups can coordinate their actions ‘as well as individuals’, but only if lateral information about others’ actions being performed is available. An online adjustment that is only conceivable during the facetoface of all participants, not in indirect (agreement) group coordination on a longer timespan (Foo, et al., 2000). Explanation: individuals once having acquired an internal model of each other’s action timing, this coactor’s action timing provides a context for the online individual planning of one’s own actions. But joint a priori control of the action would require much more: a shared action goal such as the successful performing of speech acts, i.e. the making of moves in a language game, or any social context of interaction implying the use of a linguistic and/or gestural system of communication.
Cognitive theory of action is developed under the model of the mind as representational system implemented in the brain of an individual agent. To such model, that only knows of the individual, solitary (not to say solipsistic) agent, the very notion of collective action is definitely alien. In their absolute pre Hegelian innocence, they not even conceive of its possibility! Conforming to good mainstream cognitive doctrine, all parameters of action are internal representations in the agent. Solitary action alternatives are represented in one cognitive system only: here we have the best, standard case. As soon as action alternatives are at his disposal, the agent can predict and simulate their future consequences on the basis of an internal model, a mode of control over his action that is anticipatory and not only reactive. In contradistinction, groups lack shared internal processes, language serving as a – rather poor – palliative. The fact that joint action alternatives cannot but be distributed among distinct cognitive systems remains an evident handicap of collective acting.
Mimicry, empathy and mirror neurons
Vittorio Gallese in his recent ‘embodied simulation’ theory (Gallese, 2005) tries in the line of his former ‘simulation theory of mindreading’ (Gallese & Goldman, 1998) at linking mimetic contagion to social acts through the concept of empathy founded on mirror neurons system in brain. We willingly construe this theory as a revival of Theodor Lipps’ ‘Einfühlung oder innere Nachahmung’ theory (1903) as a natural basis of the obligatory character of social acts (Lipps, 1909, pp. 237241). And such interpretation imposes itself on us all the more so that we realize the seriousness of a question that simply pointing at neural correlates does not bring to an end: what are the limits of empathy? Does the admission of an empathy instinct succeed in coping with all those resonant forms of behaviour for which mirror systems presumably provide the neural basis, especially when they are part and parcel of speech acts?
Comparing brain resonance and social act:
Speech acts in language games are no resonances! It is not to say that social reality might be floating in a vacuum, but its ontological basis has to be found in a lately acquired extension of the mutually anticipatory character of social acts, one which neurophysiology proves to be deeply ingrained in our organisms. And these acts, far from reducing to rulegoverned abstract moves in a language game, are summoned to existence in the wake of an intersubjective synchronic internal emulation by flesh and blood human beings – let’s not forget grey matter! – an emulation of a Lebenswelt which is worth to live in each with the other. But, surely, it’s too long stories to recount in one go. As a strategic shortcut move, we had better fall back on Adolf Reinach’s phenomenology of social acts trying to uncover “the a priori structure of social acts” on the basis of an analysis of promising, the meaning structure of which is supposed in any positive systems of law (Reinach, 1913). Typical of sociality in promising (promising to give somebody an object in your possession), is the creation by this very act of a new entity: a normative link that didn’t exist before and will cease to exist once the promise redeemed. Far from displaying a simple, symmetrical, instantaneous resonance, it generates the eventually compensated disequilibrium of an agreement extended beyond the face-to-face of partners, and through which you oblige yourself in giving, and the other obtains a right to demand. Note that your actual accomplishment is only part of a whole, the other part of which is situated in the future, at a date determined by the norms regulating the same act in the context of current social life.