Aplasic phantoms and the mirror neuron system: An ...

Phenom Cogn Sci DOI 10.1007/s11097-009-9138-2

Aplasic phantoms and the mirror neuron system: An enactive, developmental perspective Rachel Wood & Susan A. J. Stuart

# Springer Science + Business Media B.V. 2009

Abstract Phantom limb experiences demonstrate an unexpected degree of fragility inherent in our self-perceptions. This is perhaps most extreme when congenitally absent limbs are experienced as phantoms. Aplasic phantoms highlight fundamental questions about the physiological bases of self-experience and the ontogeny of a physical, embodied sense of the self. Some of the most intriguing of these questions concern the role of mirror neurons in supporting the development of self–other mappings and hence the emergence of phantom experiences of congenitally absent limbs. In this paper, we will examine the hypothesis that aplasic phantom limb experience is the result of an ontogenetic interplay between body schemas and mirror neuron activity and that this interplay is founded on embedding in a social context. Phantom limb experience has been associated with the persistence of subjective experience of a part of the body after deafferentation through surgical or traumatic removal. We maintain that limited association is inconsistent with the extent to which phantom limb experience is reported by aplasic individuals. Keywords Aplasia . Phantom limb experience . Body schema . Enactivism . Mirror neuron system . Embodiment

Introduction Phantom limb experiences among individuals who suffer the loss of a limb through trauma or disease provide an intriguing illustration of an inherent uncertainty at the S. A. J. Stuart (*) University of Glasgow, 11 University Gardens, Glasgow G12 8QH, UK e-mail: [email protected] R. Wood Dipartimento di Neuroscienze - Sezione di Fisiologia, Universita’ di Parma, Via Volturno 39, I-43100 Parma, Italy e-mail: [email protected] e-mail: [email protected] R. Wood Centre for Computational Neuroscience and Robotics, University of Sussex, Brighton BN1 9QG, UK

R. Wood, S.A.J. Stuart

center of our apparently veridical self-experience. We know that human perceptual systems can be easily fooled by visual illusions and other “tricks” that subtly misrepresent a state of affairs in the world. Yet still, we hold on to the idea that our inherent notions of self are immune to perceptual legerdemain. Thus, I may be vulnerable to an assortment of visual, auditory, and proprioceptive illusions but still feel that no such chicanery can trick me about my own physical form. Even manipulations such as the infamous “rubber hand illusion” (RHI; Botvinick and Cohen 1998) can only make me feel as if the fake hand on the table might belong to me if what I see happening to the rubber hand matches what I feel happening to my own hand. As soon as these conditions are broken, my normal self-perception reasserts itself. In fact, quite stringent conditions must be met in order to bring about the RHI, the most important of which is temporal synchronicity of sensory inputs. Second only to temporal synchronicity is the need for coherence with existing visual and proprioceptive information about the body (Tsakiris and Haggard 2005). The rubber hand can be, in some limited sense, incorporated into the subject’s ongoing sense of itself only so long as the seen stimulation of the fake hand and the felt stimulation of the real hand are synchronized and the rubber hand is also in a position which is congruent with rest of the subject’s body. The RHI helps to demonstrate the importance of coherence in the subject’s experience of her physical self, an alien object can be incorporated into the body schema if the object is in the right place and sensory feedback comes at the right time. I can be tricked into temporarily owning a fake hand but only if the claim does not violate the coherence of my lived body schema. So the RHI which might, at first sight, seem to be evidence for some instability in our embodied notions of self actually attests to their fidelity, showing as it does that a carefully orchestrated set of circumstances is required for the illusion to be sustained. By contrast, however, phantom limb experiences demonstrate that our self-perceptions might be rather more fragile than we like to think. Here, it is not manipulation or subterfuge that creates an illusion, but rather, the body itself generating sensations and percepts apparently pertaining to missing parts. This is illustrated in its most extreme form in the case of aplasia when congenitally absent limbs are experienced as phantoms. Much of the data on phantoms of congenitally absent limbs has been obtained through self-reports from aplasic individuals [the experiments conducted by Brugger et al. (2000) and Gazzola et al. (2007) being notable exceptions]. According to such reports, upper and lower limb phantoms are experienced and, in the extremely rare case of tetra-amelia in which neither upper nor lower limbs develop fully, both arm and leg phantoms may be experienced by the same person (Price 2006). Phantoms may be extremely vivid and present almost all of the time or be quite fleeting experiences which occur every few months. The phantom may represent the entire missing limb (a complete forearm for instance) with all phantom parts being equally vivid or the phantom may, for example, be limited to the vague sensation of a hand with undifferentiated, “finger-like” projections (Price 2006). Phantoms are reported both by users and nonusers of prosthetic limbs (Saadah and Melzack 1994; Melzack et al. 1997; Price 2006). In some instances, the emergence of a phantom limb appears to be linked to an injury or medical procedure affecting the limb stump, and yet in others, the first phantom experience appears not to be linked to any obvious physiological event. Melzack et al. (1997) document the phantom

Aplasic phantoms: An enactive, developmental perspective

experience of two aplasic adolescents, both missing an arm below the elbow and who both report bad weather or rain as being among the triggers for their phantom experiences [p. 1610]. The role of attention in eliciting phantoms appears to be ambiguous. Some subjects report that their phantom disappears when they attend to their limb stump or think about the phantom itself, but others say that the phantom emerges when they think about or bump their stump against some object [ibid.]. The existence of aplasic phantoms and the strong heterogeneity of aplasic phantom experience highlight fundamental questions about the physiological bases of selfexperience and the ontogeny of a physical, embodied sense of being a self. How can we understand the role played by coherence in the generation of aplasic phantoms and what does this tell us about embodiment and self-experience?

Self as perspective If we consider phantom limbs as a form of distortion or disturbance of the sense of self, then it is useful to spend a little time considering notions of selfhood and how these might contribute to our understanding of how aplasic phantoms come about. There is in fact no generalized, theoretical consensus about what the notion of “self” means and there exist a great many competing definitions of the term (Zahavi 2005). Some authors go so far as to deny that such things as selves exist in any substantial or forensic sense (Stuart 2006), others even preferring to posit representations to account for the sense of perspectival ownership or phenomenological “mineness” of experience (Metzinger 2003a, b). On this view, complex representational processes generate the experience of being a self from the top down, the self is thus a byproduct of cognition or, rather, there is no self, only representational content. This is a more extreme position even than the standard materialist account in which the experience of being a self amounts to no more than the net product of brain activity. The materialist position is possibly best exemplified in Churchland’s analysis wherein I feel like a self because my brain makes it so; the activity of my brain is coherent and continuous, thus so is my sense of myself (Churchland 1981). The account offered by eliminative materialism, being rooted in a neuromechanistic level of analysis has plenty to contribute to the question of where the feeling of selfhood comes from but, by definition, says less about that what it is comprised of. By contrast, phenomenological approaches attempt to couple investigation of the content of experience with exploration of what it means to be a self. Merleau-Ponty describes two components of selfhood: ipseity realized by embodied interaction with the world and temporality referring to the continuity of experience in time. The coupling of “aboutness” and temporal coherence provides a model of self-experience, which goes beyond the notion that a concatenation of brain processes is responsible for the feeling of perspectival ownership at the heart of selfexperience. The prereflective bodily self is extended spatially through preattentional, real-time representations that are generated by the interplay of proprioceptive, somatosensory, and vestibular inputs. Since spatial extension necessarily implies temporal extension and the organism’s spatiality is, for the most part, transparent, which is to say that it possesses a nonself-conscious immediacy, so, too, in the prereflective mode, is the agent’s temporality. Embodied sensory immersion in a


temporally continuous experiential world implies a point of view, a locus on which the flow of experience is centered. Nevertheless this ceaseless welling up of time is not a simple fact to which I am passively subjected, for I can find a remedy against it in itself, as happens in a decision which binds me or in the act of establishing a concept. It withholds me from what I was about to become, and at the same time provides me with the means of grasping myself at a distance and establishing my own reality as myself (Merleau-Ponty 1962, p. 380). The phenomenological conception of self exemplified by Merleau-Ponty and, more recently, Zahavi and others (see, for example, Gallagher 2005 and Legrand 2006) might be characterized as emanating from the inside and projecting out. These approaches to selfhood focus on the phenomenal content of experience from a locus of perspective, that is, the self as a phenomenological point of view. Conversely, this focus can also be inverted to informative effect; thus, the experience of self can also be understood as shaped by the outside projecting in. On this view, self is brought into relief by the medium in which it is immersed, and phenomenological experience arises at the interface between the experiencing individual and the world of things as they are encountered.1 Put differently, self is made manifest in the context of its relationship with other.2 Merleau-Ponty defines ipseity in terms of embodiment and “being-in-the-world” (Merleau-Ponty 1962, p. 408) concepts that serve to emphasize the necessary coupling between the self as a locus of experience and the world of phenomena to be experienced. [T]his view makes it clear that self-awareness is not to be understood as an awareness of an isolated, worldless self, nor is the self located and hidden in the head. To be self-aware is not to interrupt the experiential interaction with the world in order to turn one’s gaze inwards; on the contrary, self-awareness is always the self-awareness of a world-immersed self. The self is present to itself precisely and indeed only when it is engaged in the world (Zahavi 2005, p. 10). Merleau-Ponty talks of the world of objects as bringing forth the body to conscious experience; the body is experienced through interaction with things, 1

We might understand this by employing the metaphor of the cell membrane whose tensegrity, or tensional integrity, is maintained by compression and pressure from each side of the membrane’s surface. Its embodied immersion has the membrane in sharp relief at the interface. Though here we might be accused of moving towards a philosophy of immanence, of the necessary balance of the constituted/ constitutive forces, rather than remaining strictly within the phenomenological conception. 2 We refer here only to the physical world of objects, but would argue that the capacity to think reflexively of oneself as a self is only possible in socially embedded agents, and these are agents which naturally possess an endogenous intersubjectivity (Gallagher 2007), that is, the practical knowledge of oneself that guides from the inside out. It is through the combination of a socially embedded self-awareness and prereflective bodily self-awareness—derived through active felt agency—that a subjective conception of ourselves by ourselves (first-person) and an intersubjective conception by others of us as persons (secondperson) is entailed. It is this notion of personhood which extends the self temporally into a world of morally culpable agents and morally vulnerable patients. Thus, it is that agency is ontically and epistemically prior to self, and it is agency and reflective awareness which entails personhood as judged from a first-person (subjective), second-person (empathetic/intersubjective), and third-person (objective) perspective.


which, in turn, are perceived in terms of the individual’s intentions towards them. On this view, immersion in a world of manipulable objects calls forth “habitual intentions” or ways of being-in-the-world by which the self experiences both body and world. Thus, a world of graspable things is experienced by the individual in terms of affordance. At the same time, the individual experiences her embodied self via her capacity to act on such things. In the case of someone who has lost a limb, the world evokes those same habitual intentions but now there is a discontinuity between the available affordances and the capacity to act. The world remains full of manipulable things but the subject lacks an effector and somehow this mismatch must be reconciled. Thus, phantom limb experience might be understood as “called forth” by the world of objects, a relic of the habitual mode of being-in-the-world of a self or agent that has yet to remap the relationship between embodiment and affordance [ibid.]. This body–environment circuit of motor intentionality is constitutive of what Merleau-Ponty calls the “intentional arc” subtending the life of consciousness, which integrates sensibility and motility, perception and action (Merleau-Ponty 1962, p. 136). The intentional arc and being-in-theworld overall are neither purely first-person (subjective) nor purely thirdperson (objective), neither mental nor physical. They are existential structures prior to and more fundamental than these abstractions (Thompson 2005, p. 410).

Self and other At first sight, this analysis has little to contribute to the understanding of aplasic phantom experience; in the case of congenital limb absence, there is no history of embodied interaction with the world to account for the phantom, that is, there is no history for the absent body part. The aplasic infant develops her capacities to act with whatever effectors are available. Her mapping of action to object is altered though her embodied self is still founded in the coupling between the givenness of her first-person phenomenological experience and her immersion in a world of interactive possibilities. Yet, in trying to understand the roots of aplasic phantom experience, it may perhaps be useful to remember that the external world of things includes other people and that these are entities that are, more or less, like ourselves. Recent discoveries in neuroscience have helped to illuminate some of the processes by which we recognize and interpret the actions of other conspecific agents or selves, and these findings may have some bearing on how the embodied self is, in turn, shaped through its interaction with others. Mirror neurons are hypothesized to play a fundamental role in intersubjectivity, allowing an individual to understand the intentions of others by observing their actions (Rizzolatti et al. 2001). Mirror neurons, networks of which have been identified in the inferior frontal gyrus (region F5), the inferior parietal lobule (Di Pellegrino et al. 1992; Gallese et al. 1996; Rizzolatti et al. 1996), and the cortex of the superior temporal sulcus (Perrett et al. 1989, 1990) in macaques and human subjects, are active when we engage in certain goal-directed actions and when we


perceive similar actions in others (usually conspecifics). Mirror neuron networks can thus be said to link perception and action, providing a mechanism by which the observed behavior of others is “mirrored” at the neuronal level. It is proposed that mirror mechanisms underpin intention understanding (and thus social3 behavior) by means of motor “resonance” whereby perceived actions provoke concomitant activity in the motor cortex of the observer, as if the observer herself were acting. “In order to be triggered by visual stimuli, mirror neurons require an interaction between a biological effector (hand or mouth) and an object” (Rizzolatti and Craighero 2004, p. 170). In the rostral part of the inferior parietal lobule, area PF of Von Economo (1929), a third of the neurons are somatosensory, just over one tenth are visual, and the other 56% are bimodal (somatosensory and visual) neurons. “Virtually all mirror neurons show congruence between the visual actions they respond to and the motor responses they code” (Fadiga and Craighero 2007, p. 102). It is claimed that: Each time an individual sees an action done by another individual, neurons that represent that action are activated in the observer’s premotor cortex. This automatically induced, motor representation of the observed action corresponds to that which is spontaneously generated during active action and whose outcome is known to the acting individual. Thus, the mirror system transforms visual information into knowledge (Rizzolatti and Craighero 2004, p. 172). Mirroring, thus, can only occur for actions that are in the behavioral repertoire of the observer and so we might expect impaired intention understanding and concomitant social deficits in individuals born with a reduced or altered behavioral repertoire. Developmental aplasias such as tetra-amelia (characterized by the absence of both upper and lower limbs) would be good candidate syndromes in which to observe such deficits induced by dysfunctional mirror systems. However, the literature does not support the notion of impaired action understanding (or systematic social deficits) in cases of congenital limb absence. In addition, studies of phantom limb in aplasic limb absence indicate the presence of neural structures to map nonexistent body parts and motor responses (Brugger et al. 2000) providing further reason to think carefully about how we conceptualize the role of mirror mechanisms in aplasic phantoms. Rizzolatti et al. (2001) ask “What are the neural mechanisms that underlie action understanding?” where, by action understanding, they mean “the capacity to achieve the internal description of an action and to use it to organise appropriate future behaviour” [p. 661]. They propose a “direct-matching hypothesis” which states that another’s action is understood when the visual representation of the observed action is mapped on to our own motor representation of the same action.

3 For the sake of clarity, let us state simply that our definition of “social” is broad and not the one favored by, for example, some linguists and sociologists as being specific to human beings and including a set of social structures, norms, institutions, culture, language, and so on. “Social” in our sense is more broadly biological, applying to living organisms that interact in a collective coexistence. See De Jaegher and Di Paolo (2007) for an operational, enactive definition of the social which is compatible with our use of the word.


Gallagher (2006) mentions the many people, including de Vignemont (2004), Gallese (2005), Hurley (2005), and Jeannerod and Pacherie (2004), who have claimed that “representations in those brain areas that are activated when I perform an action and when I perceive another person perform the action, are neutral in regard to determining agency” [p. 2], and those who are now concerned with finding the “Who system” (Georgieff and Jeannerod 1998) which makes agent determination possible. But their quest is misguided. There is a givenness of myself that is immediately manifested in my experience (Henry 1963) without which conscious experience would be meaningless (Merleau-Ponty 1962). This givenness might correspond with what Sheets-Johnstone (2003) describes as the “energic qualities” that are present in my goal-directed agency, and which make it clear to me, without the need for any secondary system, that it is me who is the agent. Without the “energic qualities,” one could imagine the feeling of mineness or agency being absent and, subsequently, the issue of agent determination not arising. At a reflective level, one can imagine concluding, it is not me who is acting, so it must be you, but at a prereflective level, this is simply unnecessary; it is happening at a “completely neurological and sub-personal” level, “intentions in almost all cases come already clothed in agency” (Gallagher 2006, p. 3).

The given body The self-givenness4 of action is, at least in part, enabled by the possession of a body schema, that is, the system of motor capacities, abilities, and habits that enable movement and the maintenance of posture and which function without the necessity of perceptual monitoring. The body schema “is not a perception, a belief, or an attitude … [but] … a system of motor functions that operates below the level of selfreferential intentionality, although it can enter into and support intentional activity. It involves a set of tacit performances, preconscious, subpersonal processes that play a dynamic role in governing posture and movement” (Gallagher and Meltzoff 1996, p. 6). Gallagher and Meltzoff frame their analysis around the supposition that the body schema, in some form at least, is innately specified. Furthermore, they suggest that the evidence from developmental psychology supports the notion of this innate schema as providing a mechanism to account for both early imitation and aplasic phantoms “the evidence raises the possibility that the basic framework of a body schema is innate” (Gallagher and Meltzoff 1996, p. 7). The body schema as described by Gallagher and Meltzoff (they are careful to differentiate between body schema and body image, a distinction that they point out is not maintained in many discussions of aplasic phantoms) is very much predicated on action, it is not a perception of the body but rather an experience of it; so, in fact, here it seems that Gallagher and Meltzoff’s own characterization of the concept mitigates against their assumption of innateness. The infant (aplasic or otherwise) is 4 Perhaps we might understand self-givenness in terms of Husserl’s concept of “eidetic intuition”: the direct givenness which “refers to the acts in which ‘objects show up in person’” (Depraz et al. 2003, p. 45) and which primarily reveals itself as a perceptual and imaginative act concerned with disclosing an essence [ibid., p.55]. Self-givenness is concerned with the revelation of the tight experiential coupling between body and ownership of the experience.


born with a system of motor functions and, crucially, with a history of prenatal, in utero motor behavior. The uterine medium is sufficiently dissimilar from the post birth environment as to require significant relearning of proprioceptive contingencies on the part of the infant, processes which can be seen as further shaping and directing the ontogeny of a lived body schema. Either the body schema depicted by Gallagher and Meltzoff is meant to refer to this history of uterine movement in which case the schema is clearly not innate but a product of fetal experience or it simply comprises a set of motor capacities or possibilities for action within the infant’s nascent behavioral domain. It is not clear in what sense this set of morphological affordances can be usefully distinguished from the body itself. It is obviously quite possible to draw a clear ontological distinction between notions of what the body is and what it can do but then how is the capacity for action represented as a schema. One answer might be to argue that the body schema resides in the neural implementation of the motor system, that the capacity to act is thus represented in neural networks. This level of description intuitively seems not to do justice to Gallagher and Meltzoff’s original notion of the body schema as a structure involving “performances” underlying intentional activity. Thus, we come back to a characterization that links motor capabilities (possibilities for action) with a history of being in a world (abilities, habits). This richer concept seems more plausible as a means of sustaining the notion of body schema as the locus of linkages between mere movement and intentional action. Merleau-Ponty also distinguishes body image and body schema, arguing that the attempt to account for phantoms by reference to body image generates no new insight while the body image is held to be constitutive of the physical experience of self rather than “the residue of habitual cenesthesis” (Merleau-Ponty 1962, p. 86). Going still further, he rejects the reduction of body schema to the sum total of “associations established during experience,” arguing that it is better characterized “as an attitude directed towards a certain existing or possible task” [ibid.], that is, a way of being-in-the-world. For Merleau-Ponty: The body schema functions as if it were an “innate complex” (p. 84), that is, as strongly and pervasively as if it were innate, but, as an acquired habit with a developmental history, it is not innate. It follows that the existence of a phantom limb is based on a history of sensory inputs, and the continuation of sensory inputs at the stump. Sensory impulses “establish and maintain its place, prevent it from being abolished, and cause it still to count in the organism.” They are the sine qua non by which we “build up the phantom” (1962, p. 86) (Gallagher and Meltzoff 1996, p. 6). Gallagher and Meltzoff reject this view of the acquired body schema, citing evidence that, in one study of 30 aplasics, 17% experienced phantoms (Weinstein and Sersen 1961)5. Their argument centers on the idea that, if the body schema develops, then it should be true to the physical form of the individual and aplasics should not experience phantoms. Thus, in the nonaplasic infant, there is a 5

Melzack et al. (1997) found that 20% of congenital aplasics in their questionnaire study to have experienced phantoms (n=76). Saadah and Melzack (1994) report a phantom incidence of around 10% in another questionnaire study of 100 aplasics.


rudimentary, innate body schema6 ready to be elaborated through action, action interpretation, and imitation. In the aplasic case, the infant has the same rudimentary body schema and this innate structure is sufficient to overcome the lack of visual, tactile, and proprioceptive awareness for the missing limb(s) and to produce the phantom. But surely, we might ask exactly the opposite question: if phantom experience proves the innateness of the body schema, then why are aplasic phantoms not more common? One answer to this question might be that the phantom represents the failure of an innate schema to adapt to the actual form of the body, and thus, a failure to incorporate the absence of missing limbs. Thus, the majority of aplasics have body schemas which accurately reflect the individual’s particular physical form, but in some cases, the missing limbs are still “present” in some rudimentary configuration in the body schema. But this solution seems to beg the question of what develops and when. If the account is now that the body schema is innate and that it is ontogeny which fits it to the infant, then it is not clear what special merit this explanation for the aplasic phantom has over and above an alternative view in which the body schema develops from the outset. In order to make a case for the ontogeny of the body schema as a system of motor capabilities, it is crucial to take account of the evidence pertaining to the infant’s history of embodied action prior to birth. In an ultrasound study of fetal movement kinematics, Zoia et al. (2007) found an unexpected degree of action planning in the movements of 22-week-old fetuses. Kinematic patterns for bringing the hand to the mouth (for thumb-sucking) and for bringing the hand to the head (for touching) were found to be quite different both in terms of peak velocity and movement duration. By 22 weeks, hand reaches become straighter and more accurately aimed with acceleration and deceleration phases of the movement predicated on the size and sensitivity of the target. These findings provide strong evidence that the neonate comes into the world with a nontrivial history of embodied action and, while it is true that the medium in which it is moving has quite different characteristics from the environment in which it will soon find itself, the noteworthy aspect is that the infant already has an experiential history with its own body [ibid.]. We can see further evidence that the neonates’ experience of motion comes from studies of biological motion perception in newborn infants. Two-day-old infants show robust, orientation-dependent sensitivity to biological motion when exposed to point-light displays of dots moving in a coherent pattern predicated on walking. In a study by Simion et al., infants consistently preferred patterns based on a hen walking above random displays. Such data could be explained by recourse to an innate “biological motion detector module” designed (perhaps by evolution) to differentiate between animate and inanimate objects in motion (Simion et al. 2008). Alternatively, we might also propose that it is fetal experience with vestibular and proprioceptive motion cues which shapes the preference for coherent motion patterns and that this very early experience helps to calibrate shared visuomotor circuits involved in mirror 6

This notion of an innate body schema challenges the earlier acquired postural model favored by, amongst others, Piaget (1962), Merleau-Ponty (1962), Wallon (1947, 1965), and Simmel (1966). In their model, the infant’s experience is entirely interoceptive with the external perceptual abilities, which mark the exteroceptive domain, only developing after a matter of months.


neuron systems, a point also raised by Simion et al. These findings contribute to an understanding of the newborn infant as a fully embodied agent with a ready history of coupling to an environment. On this view, the body schema develops with the infant and the aplasic phantom cannot be explained as the result of an “innate” structure that fails to map the body accurately. Caution is required here; there is a risk that, in seeking to question one perspective on the causes of aplasic phantoms, we may inadvertently perpetuate a much more insidious set of assumptions about developmental causation in general. Aplasic phantoms can be interpreted as evidence to support the notion of an “innate” body schema, though here we question that view by examining evidence for the ontogeny of the body schema and, by extension, the phantom. This process inevitably entails discussion of “acquired” behaviors and characteristics. However, that should not be taken as tacit endorsement of the conventional innate–acquired dualism in developmental theory. Oyama provides a constructivist understanding of ontogeny which is highly compatible with enactive approaches to cognition (Varela et al. 1991). The concept of enaction whereby knowledge is indelibly tied to the praxis of action is valuable here as we are seeking to build an account in which the aplasic phantom develops through a history of embodied interaction in the (social) world. Oyama argues strenuously against the “dichotomization” of development and the prevalence of the “gene as information” metaphor (1985) [these sentiments are memorably echoed by Thelan and Smith in their rejection of the structure versus process “Balkanization of developmental phenomena” (1994, p. 38)]. In ontogeny, there are multiple sources of what might be termed “information” (a.k.a. causes) both within and without the organism; no one of these sources can be usefully accorded causal privilege over another. Development occurs as the unfolding of a process; otherwise understood as the dynamics of a self-organizing system. Any “cause” exists as such only in the wider context of a process enacted in time. The notion of developmental “information” thus becomes obsolete regardless of whether it is considered to be located internally in genetic material or externally in some aspect of an environment. By extension, internal–external, innate–acquired dichotomies are similarly rendered redundant in the context of a dynamic process of self-organization by which a system constructs itself through time [ibid.]. Our own position with regard to the innate–acquired dichotomy accords most closely with Oyama’s. Phantom ontogeny Aplasic phantoms seem most often to appear in early childhood or even later: [I]n the majority of cases of aplasic phantoms the onset of the phantom takes place relatively late. In cases where specific ages are provided to indicate the onset of the aplasic phantom, the age of onset ranges from 4 to 30 years, with the majority of subjects experiencing the onset of the phantom between the ages of 5 and 8 years (see Poeck 1964; Saadah and Melzack 1994; Weinstein and Sersen 1961, and Weinstein et al. 1964) (Gallagher and Meltzoff 1996, p. 8). In a review of 39 cases of aplasic phantoms, Price (2006) reports 11 in which the phantom has been experienced for as long as the subject could remember; however,


it is clearly not possible to establish from this description how early in life the phantom became apparent. The most obvious reasons for the lack of data early onset phantoms are either (1) that younger children are simply incapable of accurately recognizing and distinguishing a phantom from a nonphantom experience or (2) that they are incapable of reporting on any phantom experiences they might have. Yet, it is significant that there are no reports in the literature of infants with congenital aplasia trying to use absent limbs or in any other way indicating that they experience somatosensory feedback from their missing limbs. Interestingly, Melzack et al. (1997) report that phantoms emerge later in congenital aplasics than in amputees; according to their data, the average age of phantom onset in congenital aplasia was 9 years (n=15) compared to 2.5 years following amputation (n=26). They also found that, the older a child was at amputation, the shorter the time to the emergence of a phantom. There are clear methodological problems associated with attempting to resolve these questions; indeed, it seems unlikely that such issues could ever be resolved with absolute certainty. However, if we accept the extant literature on the age of onset for phantoms, it again appears likely that some developmental interaction underlies their eventual emergence. We can add to this phenomenological picture of phantom ontogeny with new physiological data about the neural correlates of phantom experience. In a case study of a woman with tetra-amelia, Brugger et al. (2000) used transcranial magnetic stimulation (TMS) of the sensorimotor cortex to generate contralateral phantom hand and finger sensations. Movements of the subject’s existing upper arm structures produced activations in areas of cortex deprived of afferences or efferences (silent areas) and these activations mapped to phantom hand sensations. In addition, phantom sensations were obtained in the absence of concomitant motor-evoked potentials (MEPs), a result which contrasts with data obtained from traumatic amputees. The subject of this study A.Z., born with neither forearms nor legs, reported having had vivid phantom limb experiences for “as long as she could remember” [ibid., p. 6168]. Her phantoms comprised forearms with hands and fingers and legs with feet plus first and fifth toes. Interestingly, A.Z., who uses a wheelchair but not prosthetics, reported that when manipulating objects with an arm stump she would feel the phantom fingers but lose awareness of the forearm, whereas when contact with the object stopped, the fingers would return to their distal position. TMS was used to map A.Z.’s sensorimotor cortex with MEPs obtained from the deltoid muscles. In 13 out of 18 stimulation sites where a response was elicited, a coincidence of deltoid MEPs and phantom sensations were reported; four stimulation sites elicited phantom movement sensations without corresponding MEPs. Phantom movement sensations were experienced exclusively in the limb contralateral to the stimulated hemisphere. These were described as slow movements of one or several fingers or of the whole hand. In contrast with TMS-elicited contralateral phantom movements, functional magnetic resonance imaging (fMRI) data showed volitional phantom movements to produce bilateral activations, indicating a difference between externally triggered and self-generated phantoms. Phantom finger and hand movements failed to produce activation of primary motor cortex; however, bilateral activations of the dorsal premotor cortex and superior posterior parietal cortex were observed.


Brugger et al. conclude: “These data indicate that body parts that have never been physically developed can be represented in sensory and motor cortical areas” [ibid., p. 6167]. While they acknowledge that these findings could be taken as evidence in support of an innate body schema, they also invoke shared neural networks for action preparation and observation (i.e., mirror neurons) to account for the finding that absent limbs are mapped in the brain. In the absence of a physical substrate for the execution of an action, habitual perception of conspecifics moving extremities could still activate networks mediating a visuomotor limb representation. This activation may give rise to phantom sensations in at least a minority of individuals with limb aplasia [ibid., p. 6172]. The aplasic individual inhabits a world of graspable, manipulable objects and of conspecifics who execute their intentional actions though the means available to them (including the “exercise” of effectors the aplasic lacks). The aplasic person is able to perceive others moving their limbs and, possibly, mirror their action neuronally which would also provide some low-level somatosensory experience. Significantly, the aplasic individual is also able to attend to the goals of others’ actions, that is, not just the “how” of action but also to the “why.” By these means, it is possible to postulate that mirror mechanism activity might reach a threshold such that it signals the presence of a limb even where no actual limb exists. It is important to emphasize the strongly embodied sense in which mirror neurons are postulated to contribute to this aspect of phantom generation. Action observation–execution mapping takes the form of motor resonance. Observation of an action evokes neural activity concomitant with execution of the action, the neural structures which produce this response are located in the brain of the observer but the correspondence they enact is entirely founded on the body and its capacity to act. By extension, the obvious difficulty with this account arises from the necessity for mirrored actions to be in the behavioral repertoire of the observer. In other words, there is no mirror mechanism response for actions which the observer has no personal experience of performing. But a solution to this might be found in the history of observing the species-typical action primitives of conspecifics which has an ontogenetic priming effect on mirror neurons, such that, for instance, hand actions, are remapped onto the effector systems available to the aplasic person. This analysis suggests that mirror mechanisms should be relatively flexible with regard to the mapping of body parts to actions with the means of execution being of considerably less importance than the goal. Thus, the correspondence between actions should lie not in the effector systems employed but in their goals. So, the use of a hand to bring a cup to the mouth and drink can be (goal-)equivalent to the use of a foot to lift the cup to the mouth in order to drink. This notion is, to at least some extent, supported by recent experimental work by Catmur et al. (2008) who found that, in morphologically typical adults, mirror responses could be obtained for hand to foot mappings. In this fMRI study, two groups of participants were trained to execute movements in response to the actions of a model. The control group was trained to respond with a congruent effector (hand to hand and foot to foot) while the other group responded with an incongruent effector (foot to hand). Catmur et al. found countermirror activity in the incongruent


condition. Thus, brain areas that responded most strongly to observation of hand actions in the control group responded more to foot actions in the incongruent group. This far from trivial finding demonstrates that motor resonance is inherently plastic and it further indicates that the means of execution does not predicate the response of the mirror mechanism. The results of this study confirm the predictions of experience-based theories that postulate that the mirror system consists of links between neural populations coding for sensory and motor action representations, and that these are forged through correlated experience of seeing and doing actions (Heyes 2001; Keysers and Perrett 2004; Brass and Heyes 2005). These theories further suggest that the processes that create these links are stimulus-general associative learning mechanisms (Catmur et al. 2008, p. 1213). Thus, if observational input were sufficient to stimulate the aplasic mirror neurons and to establish rudimentary mappings between disparate effector systems, then a history of interaction with morphologically typical conspecifics might well be sufficient to develop the mirror mechanism to such an extent that it produces an altered body schema and even eventually generates phantom limb experiences. This is not to neglect the essential correspondence, that is, the “mirror” in the system. This correspondence between self and other cannot be established simply through observation. In that case, the motor system would play no part in the process of self– other mapping or in any consequent action understanding. It is necessary for observation and action to be coupled through a history of interaction and this, we propose, is the social component in the ontogeny of the body schema. By extension, this is also the level at which the body schema interacts with the mirror mechanism and, therefore, the level from which phantom experience arises. Evidence in support of this view comes from an fMRI study by Gazzola et al. (2007) in which they examine aplasics’ mirror system responses to hand actions. Two aplasic and 16 typically developed participants were scanned while they watched videorecordings of hands manipulating objects while they themselves simultaneously manipulated objects using various effectors, including mouth, toes, and for typically developed subjects, hands. The aplasic participants activated regions associated with mouth and foot execution in response to observation of hand actions, that is, they mapped feet to hands, thus generating a goal-level correspondence between actions performed with different effectors. These results highlight once again the question of how the correspondence between observed and executed actions is realized: is the association between an observed action and an equivalent goal or is it based simply on the corresponding means of achieving the goal? Our stated position is that the association occurs at the level of goal equivalence. This notion of correspondence as founded on the goal rather than the means of an action should, however, not be taken to mean that neuronal activation is “in itself” a “representation of” or indeed “thought” of a goal or intended outcome. The claim of goal rather than means equivalence is borne out by work carried out on the mirror systems of monkeys. The monkey mirror system comprises “broadly congruent” (bcMN) and “strictly congruent” (scMN) mirror neurons where scMNs respond for observed actions with the same goal


and the same effector as the executed action and bcMNs also respond when the action is performed using a different effector. Thus: [T]he idea that the observation of an action also recruits motor programs of actions with corresponding goals but differing means endows the observer with the flexibility of mapping the observed action onto the behavioural alternative that is most suited under his present circumstances (Gazzola et al. 2007, p. 1239). This flexibility also helps to account for the lack of any deficit in action understanding that we might expect to see if the altered morphology of aplasia precluded the use of motor resonance-based mirror mechanisms for acts performed by typically developed conspecifics. In fact, even more impressively, it helps to account for any systematic deficit in perspective taking or other social cognitive function. Price (2006) argues for a developmental account of body image formation and aplasic phantom experience. Price does not distinguish body image from body schema but uses image to refer to the putative innate structure by which the body is mapped. In this paper and elsewhere, this would be referred to as the body schema. One of the fundamental claims Price makes is that, prior to birth, an in utero “primitive bilateral body image [schema]” is formed on the basis of spontaneous fetal movement. Post birth, the infant’s visual and sensory systems are brought together through the mirror neuron system which can also facilitate the incorporation of prostheses into the body schema, thus contributing to the development of experiential phantoms. “This hypothesis is capable of explaining all cases of aplasic phantoms without recourse to an independent genetically determined neural representation of self” [ibid., p. 320]. Price proposes that, to evaluate the developmental account of aplasic phantom experience, it would be necessary to exclude opportunities for the individual to acquire visuomotor mappings for the absent limbs and suggests studies of phantom limb experiences among congenitally blind subjects as a test of the validity of the ontogenetic perspective. Studies of phantom limb incidence amongst the congenitally blind could provide a valuable test of the validity of the hypothesis presented in this paper. Drawings and models of human figures made by blind children indicate that they hold a mental representation of human form in which body parts critical to their exploration of the world—such as hands and arms—are exaggerated in size (Critchley 1953; Kinsbourne & Lempert 1980). This phenomenon alone suggests that their representations of the body are built upon somatic sensory experience. A phantom limb, or limbs, experienced by a congenitally blind subject with bilateral aplasia and no history of prosthesis usage would be the strongest possible evidence for the existence of an autonomous hard-wired body image. On a lower explanatory level, unilateral aplasic phantom in a congenitally blind subject would indicate that a process other than visual input is involved, be it due to a bilateral representation of body image, or genetic hard-wiring [ibid.]. However, as both aplasia and congenital blindness are individually rare and together even rarer, it is unlikely that enough people with both conditions could be found to make this a viable experimental approach. It would also be unlikely, if they were to occur together, that the individual would be of an age and ability to report the experience of phantoms before they were to have a prosthetic limb fitted.


Conclusions In this examination of aplasic phantom limb experience, we have sought to offer an alternative to the view that an innate body schema is necessary to account for phantoms of congenitally absent limbs. We have presented an account in which a history of embodied action (both pre-birth and post birth) plays a constructive role in the ontogeny of the body schema. The developmental aspects of this account, particularly with respect to the role of embedding in a social world, fit well within the enactive approach (Varela et al. 1991; Thompson 2007). On this view, the aplasic phantom does not emerge as a vestigial fragment of a rudimentary, innate structure that has failed to adapt to the lack of certain limbs, rather, and much more positively, the aplasic phantom can be viewed almost as a developmental acquisition, a social construction which reflects the ontogenetic interaction between the body schema, as an element of the prereflective, basic self, and the role of the mirror neurons in mediating the bidirectional relationship of self and other. Recent experimental evidence shows that action observation–execution mappings are inherently plastic and that it is the goal rather than the means of attainment that is central to mirror mechanism activity (Catmur et al. 2008). In addition, aplasic subjects have been shown to map observed hand actions to executed foot actions, demonstrating that effector mapping is flexible (Gazzola et al. 2007). These experiments help to explain how it is that aplasic subjects show no deficit in action understanding despite their inability to replicate the means of attainment of the goal in an observed action. By extension, they also provide some insight into evidence to show neural mappings for congenitally absent limbs (Brugger et al. 2000). Previously, such evidence has been taken to indicate that the neural mapping of body parts, that is, the formation of the body schema, is innate, hard-wired, and at least, at a rudimentary level developmentally isolated from the growth or failure to grow of limbs and other elements of morphological structure. A view which by extension entails that such neural structures must, therefore, not be experience dependent; the fetus’ history of spontaneous movement in utero being irrelevant to the neural structures mapping its body. In contrast, here we have presented an account that places a strong emphasis on a constructive continuity of experience between prenatal and postnatal life and which endeavors to take account of the developmental trajectory traversed by the aplasic infant. On this view, the fact that a neural mapping can be demonstrated when its target, a particular limb, is absent does not mean that we need to couch our questions about the neural system which underpin that mapping in terms of innateness versus acquisition. There are a number of other possibilities and it seems most useful to start from a position that can take into account the ontogeny of a system, thus we can ask questions such as, at what point in fetal development did the aplasia occur? If neural structures are in any way affected by afferent feedback from the target sites, then such inputs will be available from the earliest stages of embryogenesis and will continue until development is diverted in aplasia. This interpretation of an “altered” trajectory is, at least, one which can be contrasted with the view of aplasia as a complete absence of development. The claim that phantoms of congenitally absent limbs do not require explanation in terms of an innate body schema is complementary with the notion that the aplasic


phantom can be understood ontogenetically, as a fundamentally social acquisition. Thus, just as nonaplasic phantom limb experience might be understood as “called forth” by the world of graspable, manipulable objects (Merleau-Ponty 1962), so the aplasic phantom could be enacted through the medium of intersubjectivity and the visual and somatosensory experience of others’ actions in a world of graspable, manipulable objects. The experience of interacting with conspecifics, of joint action often featuring the coupling of hands and other effectors in the service of shared goals, could provide the stimulus for the aplasic individual’s incorporation of absent limbs into her own body schema. Thus, the aplasic individual might be said to acquire a somatosensory mapping of self to other that is richer and more complex than that of her typically developed counterpart. On this view, her body schema explicitly accounts for the other within the most basic structures of herself; her mirror system could provide the substrate for this flexible negotiation of dual modes of action. The aplasic phantom might, therefore, be seen to exemplify the notion of “differently abled” in its most positive sense, signifying the capacity to extend one’s own embodiment to take experiential account of the other.

Acknowledgements The authors would like to thank Vittorio Gallese, Alessandra Umilta and Ezequiel Di Paolo for helpful discussions on the themes of this paper. RW’s work on this paper was supported by the EU Marie Curie - Research Training Network 035975 “DISCOS - Disorders and coherence of the embodied self”.

References Botvinick, M., & Cohen, J. D. (1998). Rubber hand ‘feels’ what eyes see. Nature, 391, 756. Brass, M., & Heyes, C. (2005). Imitation: Is cognitive neuroscience solving the correspondence problem? Trends in Cognitive Sciences, 9, 489–495. Brugger, P., Kollias, S. S., Muri, R. M., Crelier, G., Hepp-Reymond, M.-C., & Regard, M. (2000). Beyond re-membering: Phantom sensations of congenitally absent limbs. Proceedings of the National Academy of Sciences of the United States of America, 97(11), 6167–6172. Catmur, H., Gilmeister, H., Bird, G., Liepelt, R., Brass, M., & Heyes, C. (2008). Through the looking glass: Counter-mirror activation following incompatible sensorimotor learning. European Journal of Neuroscience, 28, 1208–1215. Churchland, P. (1981). Eliminative materialism and the propositional attitudes. Journal of Philosophy, 78 (2), 67–90. Critchley, M. (1953). Tactile thought, with special reference to the blind. Brain, 76, 19–35. Depraz, N., Varela, F., & Vermersch, P. (2003). On becoming aware: Steps to a phenomenological pragmatics. Amsterdam: John Benjamins (Advances in Consciousness Research Series). De Jaegher, H., & Di Paolo, E. (2007). Participatory sense-making: An enactive approach to social cognition. Phenomenology and the Cognitive Sciences, 6(4), 485–507. de Vignemont, F. (2004). The co-consciousness hypothesis. Phenomenology and the Cognitive Sciences, 3 (1), 97–114. Di Pellegrino, G., Fadiga, L., Fogassi, L., Gallese, V., & Rizzolatti, G. (1992). Understanding motor events: A neurophysiological study. Experimental Brain Research, 91, 176–180. Fadiga, L., & Craighero, L. (2007). On electrophysiological data on mirror neurons and motor representations. In S. Bråten (Ed.), On being moved: From mirror neurons to empathy. Amsterdam, The Netherlands: John Benjamins Publishing Company. Gallagher, S. (2005). How the body shapes the mind. Oxford: Oxford University Press. Gallagher, S. (2006) “Perceiving others in action/La perception d’autrui en action”, Fondements cognitifs del’interaction avec autrui, Collége de France (22 February 2006).

Aplasic phantoms: An enactive, developmental perspective Gallagher, S. (2007). Moral agency, self-consciousness, and practical wisdom. Journal of Consciousness Studies, 14(5–6), 199–223. Gallagher, S., & Meltzoff, A. (1996). The earliest sense of self and others: Merleau Ponty and recent developmental studies. Philosophical Psychology, 9(2), 211–234. Gallese, V. (2005). ‘Being like me’: Self–other identity, mirror neurons and empathy. In S. Hurley & N. Chater (Eds.), Perspectives on imitation I (pp. 101–118). Cambridge: MIT. Gallese, V., Fadiga, L., Fogassi, L., & Rizzolatti, G. (1996). Action recognition in the premotor cortex. Brain, 119, 593–609. Gazzola, V., van der Worp, H., Mulder, T., Wicker, B., Rizzolatti, G., & Keysers, K. (2007). Aplasics born without hands mirror the goal of hand actions with their feet. Current Biology, 17(14), 1235–1240. Georgieff, N., & Jeannerod, M. (1998). Beyond consciousness of external events: A ‘Who’ system for consciousness of action and self-consciousness. Consciousness and Cognition, 7, 465–477. Henry, M. (1963). L’Essence de la manifestation/The essence of manifestation. The Hague: Nijhoff. 1973. Heyes, C. (2001). Causes and consequences of imitation. Trends in Cognitive Sciences, 5, 253–261. Hurley, S. L. (2005). Active perception and perceiving action: The shared circuits model. In T. Gendler & J. Hawthorne (Eds.), Perceptual experience. New York: Oxford University Press. Jeannerod, M., & Pacherie, E. (2004). Agency, simulation, and self-identification. Mind and Language, 19 (2), 113–146. Keysers, C., & Perrett, D. I. (2004). Demystifying social cognition: A Hebbian perspective. Trends in Cognitive Sciences, 8, 501–507. Kinsbourne, M., & Lempert, H. (1980). Human Figure representation by blind children. The Journal of General Psychology, 102, 33–37. Legrand, D. (2006). The bodily self: The sensori-motor roots of pre-reflexive self-consciousness. Phenomenology and the Cognitive Sciences, 5, 89–118. Melzack, R., Israel, R., Lacroix, R., & Schultz, G. (1997). Phantom limbs in people with congenital deficiency or amputation in early childhood. Brain, 120, 1603–1620. Merleau-Ponty, M. (1962). Phenomenology of perception. New York: The Humanities Press. Trans. Colin Smith, London: Routledge and Kegan Paul. Metzinger, T. (2003a). Phenomenal transparency and cognitive self-reference. Phenomenology and the Cognitive Sciences, 2, 353–393. Metzinger, T. (2003b). Being no one. The self-model theory of subjectivity. Cambridge: MIT. Oyama, S. (1985). The ontogeny of information: Developmental systems and evolution. Cambridge: Cambridge University Press. Perrett, D. I., Harries, M. H., Bevan, R., Thomas, S., Benson, P. J., Mistlin, A. J., et al. (1989). Frameworks of analysis for the neural representation of animate objects and actions. Journal of Experimental Biology, 146, 87–113. Perrett, D. I., Mistlin, A. J., Harries, M. H., & Chitty, A. J. (1990). Understanding the visual appearance and consequence of hand actions. In M. A. Goodale (Ed.), Vision and action: The control of grasping (pp. 163–342). Norwood: Ablex. Piaget, J. (1962). Play, dreams, and imitation in childhood. London: Routledge and Kegan Paul. Trans. Gattegno & Hodgson. Poeck, K. (1964). Phantoms following amputation in early childhood and in congenital absence of limbs. Cortex, 1, 269–275. Price, E. H. (2006). A critical review of congenital phantom limb cases and a developmental theory for the basis of body image. Consciousness and Cognition, 15, 310–322. Rizzolatti, G., & Craighero, L. (2004). The mirror-neuron system. Annual Review of Neuroscience, 27, 169–192. Rizzolatti, G., Fadiga, L., Fogassi, L., & Gallese, V. (1996). Premotor cortex and the recognition of motor actions. Cognitive Brain Research, 3, 131–141. Rizzolatti, G., Fogassi, L., & Gallese, V. (2001). Neurophysiological mechanisms underlying the understanding and imitation of action. Nature Reviews Neuroscience, 2, 661–670. Saadah, E. S. M., & Melzack, R. (1994). Phantom limb experiences in congenital limb-deficient adults. Cortex, 30, 479–485. Sheets-Johnstone, M. (2003). Kinesthetic memory. Theoria et Historia Scientiarum, 7, 69–92. Simion, F., Regolin, L., & Bulf, H. (2008). A predisposition for biological motion in the newborn baby. Proceedings of the National Academy of Sciences of the United States of America, 105(2), 809–813. Simmel, M. L. (1966). Developmental aspects of the body scheme. Child Development, 37, 83–95. Stuart, S. A. J. (2006). “Extended body, extended mind: the self as prosthesis”, in screen consciousness: Mind, cinema and world. Amsterdam: Rodopi.

R. Wood, S.A.J. Stuart Thelen, E., & Smith, L. (1994). A dynamic systems approach to the development of cognition and action. Cambridge: Bradford Books. Thompson, E. (2005). Sensorimotor subjectivity and the enactive approach to experience. Phenomenology and Cognitive Science, 4(4), 407–427. Thompson, E. (2007). Mind in life. Biology: Phenomenology and the sciences of mind. Cambridge: Harvard University Press. Tsakiris, M., & Haggard, P. (2005). The rubber hand illusion revisited: visuotactile integration and selfattribution. Journal of Experimental Psychology: Human Perception and Performance, 31, 80–91. Varela, F. J., Thompson, E., & Rosch, E. (1991). The embodied mind: Cognitive science and human experience. Cambridge: MIT. Von Economo, C. (1929). The cytoarchitectonics of the human cerebral cortex. London: Oxford University Press. Wallon, H. (1947) “Les origines de la pensee chez l’enfant”/“The Origins of Thought in the Child”. In The World of Henri Wallon, trans. Michael Vale. Jason Aronson (1984). Wallon, H. (1965) “L’evolution psychologique de l’enfant”/“The Psychological Development of the Child”. In The World of Henri Wallon, trans. Michael Vale. Jason Aronson (1984). Weinstein, S., & Sersen, E. A. (1961). Phantoms in cases of congenital absence of limbs. Neurology, 11, 905–911. Weinstein, S., Sersen, E. A., & Vetter, R. J. (1964). Phantoms and somatic sensation in cases of congenital aplasia. Cortex, 1, 276–290. Zahavi, D. (2005) Being someone. Psyche, 11/5, 1–20. Retrieved February 3, 2009, from http://psyche.cs. monash.edu.au/. Zoia, S., Blason, L., D’Ottavio, G., Bulgheroni, M., Pezzetta, E., Scabar, A., et al. (2007). Evidence of early development of action planning in the human foetus: a kinematic study. Experimental Brain Research, 176, 217–226.