Target Paper for BBS
Title: A critical review of reports and theories of phantom limbs amongst congenitally limb-deficient subjects and a proposed theory for the developmental origins of body image.
Author: Elfed Huw Price
Address: Wellcome Unit for the History of Medicine, University of Oxford, 45-47 Banbury Road, Oxford, United Kingdom. OX2 6PE
Email: huw.price@magd.ox.ac.uk
Article keywords: aplasic, body image, body schema, congenital, mirror neurons, phantom limb.
Word Counts:
Abstract 1: 90
Abstract 2: 210
Main text (including tables): 4,434
References: 699
Entire: 5554
Abstract 1
Confirmed reports of phantom limbs amongst subjects congenitally lacking one or several limbs (aplasics) have generally been interpreted as strong evidence that mental/cerebral representations of body image are genetically hardwired. In this paper it is argued that the same data may also be interpreted as supporting an alternative hypothesis in which body image is learnt from sensory input. Reports of phantom limbs amongst aplasic subjects dating back over 160 years are reviewed and systematised within a single comprehensive hypothetical framework in which body image is learnt from the body itself.
Abstract 2
Reports of phantom limbs amongst aplasic subjects (i.e. those congenitally lacking one or several limbs) have been widely taken as evidence that body image is, at least partially, ‘hard-wired’ in the brain, and that sensory input from the body is not necessary for the formation of body image. Although alternative explanations of aplasic phantoms have been put forward, these have only been on a case by case basis, and as such they have been unsuccessful in providing a general theory for the development of body image. This paper collates previous published reports of aplasic phantoms and attempts to systematise these findings within a comprehensive theoretical framework.
The article argues that the 'hard-wired' theory is neither practical nor necessary in order to explain phantom limb phenomena. An alternative assessment of the data is presented, in which the formation of aplasic phantoms is interpreted as a natural consequence of the processes underlying the development of body image, learnt during pre-natal development and early childhood. In this process it is suggested that pre-natal movement, cross-cortical connections and mirror neurons each play a prominent role. This model unites several previously disparate theories and offers a viable solution to several longstanding phantom limb puzzles and highlights potentially rewarding avenues that may be worthy of further inquiry.
1 Introduction
It is now well established that the majority of amputees continue to experience bodily sensations as if the absent limbs were still present. These ‘phantom’ sensations occur in all modalities (including movement, temperature, touch and pain) and can be ‘triggered’ by a variety of stimuli, including pressure on the stump, emotional distress, and even changes in the weather (Ramachandran & Hirstein 1998).
To date, attempts to present a comprehensive explanation for the developmental basis of phantom limbs (and, by association, neural representations of the body) have remained inconclusive (Grouios 1998). Despite substantial attention and investigation, the neurophysiological basis for these ‘phantom limbs’ remains unresolved. The ‘traditional’ standpoint, dominant until recent years, is that positional awareness of the body, otherwise known as the ‘body image’ relies on afferent sensory input (e.g. Head & Holmes 1912; Simmel 1962). More recently, the importance of input has been emphasised by neural network theories (Churchland 1995) and the developmental basis of other neural systems such as those responsible for vision and language (e.g. Sengpiel et al 1999; Pinker 1994).
Studies of the incidence of phantom limbs amongst childhood amputees appear to show that sensory input is of direct influence on the development of body image (Simmel 1962; Melzack et al. 1997). The data (presented in fig. 1 below) show that the chances of developing a phantom limb following amputation increase steadily with age at amputation. The implication is that body image is learnt during this period, becoming steadily more established until it stabilises at around ten years of age.
Figure 1: Incidence of phantoms amongst childhood amputees.
Yet, even though there are strong theoretical and empirical reasons to believe that sensory input and proprioceptive feedback play an important role in the development of body image, sensory based theories have been unable to account for a number of anomalous phantom phenomena. Of these, reports of phantom limbs amongst children congenitally lacking one or several limbs (aplasics) have been particularly problematic. On first glance, the occurrence of phantom sensations despite the lifelong absence of sensory input and feedback from the body part concerned may appear to be inexplicable by a sensory based theory.
Although a number of cases of aplasic phantoms were documented by the physiologist Gabriel Gustav Valentin as early as 1836, it is only within the last half century that the subject has seriously been investigated. As late as the 1962, Simmel was able to overlook the existence of such phantoms; yet, since then, a substantial barrage of studies (see table 1) have provided convincing evidence that aplasic phantoms can and do occur albeit to a much lower extent than amongst adult amputees. Melzack et al (1997) report an incidence of just under 20%; some studies have reported comparable but lower frequencies than this, whilst in several others, reports of aplasic phantoms are entirely absent (Vetter & Weinstein 1967; Saadah & Melzack 1994).
Table 1. All known reported cases of congenital phantoms with details: (by study).
|
Study |
Author |
Year |
Number of new cases |
|
A |
Valentin |
1836 i |
4 |
|
B |
Valentin |
1844 i |
1 |
|
C |
Mickorey |
1954 |
1 |
|
D |
Weinstein & Sersen |
1961 |
4 |
|
E |
Poeck |
1964 |
1 |
|
F |
Weinstein et al. a |
1964 |
13 |
|
G |
Burchard |
1965 |
1 |
|
H |
Ramachandran |
1993 |
1 |
|
I |
Saadah & Melzack |
1994 |
4 |
|
J |
Grouios |
1996 |
1 |
|
K |
Melzack et al. |
1997 |
15 |
|
L |
Wilkins et al. |
1998 |
2 |
|
M |
Brugger et al. |
2000 |
1 |
i
These dates are according to Vetter and Weinstein (1967). The earliest copy of Valentin’s work that was available for this study was dated from 1848.a
This study presents the cases from the 1961 paper again, not included in this count.Undernote: Only 'true' congenital phantoms are included. Reports of infant amputations are excluded (e.g. Weinstein & Serson 1961 case #5: Sohn 1914 : André et al. 2001).
This evidence of apparent development of body image in the absence of sensory data has led a number of experts to infer that neural representations of the body are underwritten by a primordial, genetically encoded, hard-wired blueprint (Melzack et al. 1997; Gallagher et al. 1998; Ramachandran & Hirstein 1998; Brugger et al. 2000). Yet, whilst the ‘hard-wiring’ hypothesis offers a resolution to the aplasic phantom mystery, this resolution is itself problematic in that it raises further complications; it presupposes that body form is genetically encoded twice - once for the development of the body and once for a neural ‘body image’. Furthermore, it introduces the need for a means to accurately connect these two pre-determined body forms. In demanding such an array of technical adaptations, the hard-wiring hypothesis arguably raises as many problems as it solves. It also raises unanswered questions, such as what are the evolutionary benefits of hard wiring? And, if body image is hard wired, why are congenital phantoms not more common?
Given the problems associated with the hard wiring hypothesis, it may be wise to evaluate any alternative explanations that avoid these difficulties. In the following sections the reports of aplasic phantoms are collated and systematised within a single comprehensive theory that attempts to explain all phantom phenomena (including aplasic phantoms) as originating from sensory input, avoiding the complications associated with the hard-wiring hypothesis.
2 Re-evaluating the aplasic data
In this section the implications of several disparate theories and research findings already in circulation are discussed and applied to the development of body image in order to provide an alternative, sensory-based, explanation for aplasic phantoms.
In order to explain how phantom limbs can be present in aplasic subjects several authors have hypothesised that aplasic phantoms may be learnt from the intact matching contralateral limb (e.g. Burchard 1965; Grouios 1996). This suggestion has been corroborated by more recent studies of phantom phenomenon amongst amputees for some of whom sensations of movement and touch from the healthy opposite limb are not only felt in that limb, but are also experienced simultaneously as a contralateral phantom. (Ramachandran & Hirstein 1998; Melzack et al. 1997). This has led Ramachandran and Hirstein (1998) to propose that the neural representations of paired limbs, situated in alternate cerebral hemispheres, are interconnected via cross-cortical connections. Moreover, since they note that neither pain nor temperature appear to be referred to the opposing phantom limb in this way, they suggest that cross-cortical connections may be absent for these specific modalities.
Ramachandran and Hirstein’s hypothesis gains further credibility when conjoined with the findings of a study of cerebral and behavioural development in human embryos. Before the connections may be seen, it is first necessary to give an outline of the study.
Using ultrasound, Hepper et al. (1998) recorded preferentially lateralised arm movements at 10 weeks, before any signs of asymmetry can be seen in the brain itself. The authors propose that, at this stage of development, the contractions are most likely to be due to spontaneous activation of the muscles themselves, rather than neural or cerebral triggers. They conclude that:
"differential motor behaviour precedes and determines neural development."
(Hepper et al. page 533; italics added).
If valid, this proposal has significant implications for the development of body image. Neural representations of the body may derive from sensory and proprioceptive information originating in the limbs, and prompted by from movements initiated by the muscles themselves. It is suggested that this occurs in utero at a stage when cortical sensitivity to somatic input would presumably be particularly marked.
Despite differences in approach and focus, the findings outlined above can be combined into a single theory. Embryonic arm movement is predominantly one sided (usually right-handed) and serves as a stimuli for neural development (Hepper et al. 1998). A primitive body image could conceivably be developed, in utero, as a result of such movements. However, if a lateralised bias to body image were to be avoided, the input from the right arm (in right-handers) would need to feed into development of a bilateral body image for both left and right sides of the body.
The supposition that this stage of development occurs in utero is complements Ramachandran and Hirstein’s cross-cortical hypothesis: the absence of cross-cortical referrals of pain and temperature is a direct reflection of an environment such as the uterus, in which painful stimuli and changes in temperature are negligible.
Hence body image may be learnt bilaterally in utero. If so, one would predict that aplasic phantoms would i) match an opposite limb and ii) be present from birth.
Table 2: Aplasic phantom data, sorted according to paired limbs and age of onset.
|
Group |
Case |
I.D. |
Missing parts/ abnormality |
Phantom description |
AOO |
Paired? |
|
I |
1 |
A1 |
h (side unreported) |
palm & 5 fingers |
M |
Y |
|
2 |
F3 |
L fa |
L fa, h & "some" fingers |
M |
Y |
|
|
3 |
F2 |
R h |
entire |
M |
Y |
|
|
4 |
F4 |
L h |
entire |
M |
Y |
|
|
5 |
F7 |
R fa |
R h |
M |
Y |
|
|
6 |
F8 |
L a |
fingers at stump |
M |
Y |
|
|
7 |
F9 |
L a |
entire |
M |
Y |
|
|
8 |
F10 |
L fa |
palm & middle finger |
M |
Y |
|
|
II |
9 |
F1 |
R a |
entire |
7 |
Y |
|
10 |
F6 |
L l (below k) |
calf and shin |
9 |
Y |
|
|
11 |
F11 |
L fa |
L fa stump phantom |
7 |
Y |
|
|
12 |
F12 |
L fa |
short L fa + h |
4-6 |
Y |
|
|
13 |
I1 |
L fa |
entire |
16 |
Y |
|
|
14 |
I4 |
deformed arms; short R la |
R L |
6 |
Y |
|
|
15 |
K1 |
L fa, (weak R a) |
L thumb & 3 fingers |
4-5 |
Y |
|
|
16 |
K2 |
L fa |
entire |
8-9 |
Y |
|
|
17 |
K3 |
L fa |
L h & 4 fingers |
9 |
Y |
|
|
18 |
K4 |
L h |
all fingers |
6-7 |
Y |
|
|
19 |
K6 |
L h |
L h & fingers |
7 |
Y |
|
|
20 |
K7 |
L fa |
entire |
10 |
Y |
|
|
21 |
K8 |
R fa |
R h & fingers |
12 |
Y |
|
|
22 |
K10 |
small malformed R h |
R fingers |
12-13 |
Y |
|
|
23 |
K11 |
L fa |
L fa & h, no fingers |
12 |
Y |
|
|
24 |
K13 |
L fa |
entire |
8-10 |
Y |
|
|
25 |
K14 |
L a |
L fa, h & fingers |
4-5 |
Y |
|
|
III |
26 |
F14 |
L fa; R h; B l (below K) |
L wrist, thumb & palm |
M |
N |
|
27 |
H1 |
B a |
B arms |
M |
N |
|
|
28 |
M1 |
B fa & B l |
B fa, h, l, 1st & 5th toes |
M |
N |
|
|
IV |
29 |
E1 |
B fa |
B h & fingers |
6 |
N |
|
30 |
F5 |
B a |
L a stump (2 inches) |
5-6 |
N |
|
|
31 |
F16 |
B a |
extended stumps |
7 |
N |
|
|
32 |
F17 |
B a, shortened R ll |
L a entire; R a stump; R ll |
4 |
N |
|
|
33 |
F18 |
L fa; R a; R ll; shortened L ll |
L fa stump |
12 |
N |
|
|
34 |
I2 |
B fa (h’s present at elbows) B ll (B f present at k) |
B ll ("fill his prostheses") |
26 c |
N |
|
|
35 |
I3 |
Deformed arms. B ll (B f present at k) |
B ll (which appear when prostheses are removed) |
30 c |
N |
|
|
36 |
K5 |
L a; R fa (R h present) |
L a & L h entire |
9 |
N |
|
|
37 |
K9 |
L fa; 2 fingers of R h |
L fa & L h entire. |
3-4b |
N |
|
|
38 |
K12 |
L fa, R fa |
L fa, L h & fingers |
12 |
N |
|
|
39 |
K15 |
small malformed R h; shortened R fa; L fa (no L h); R l (f present) |
R h |
12-14 |
N |
Table legends and annotations
a
partially amputated at 3 years old.b
phantom appeared at some point during first 5 yearsc
Phantom was induced by physical traumaAbbreviations:
AOO = age of onset: M = as long as they could remember: L = left, R = right, B = both: a = arm, f = foot, fa = forearm, h = hand, k = knee, l = leg, ll = lower leg.
The I.D. code is in the form, "report" (using letters designated in table 1) + "case number" (as designated in the original report referred to). The cases reported by Weinstein and Sersen in 1961 are repeated in the report by Weinstein, Sersen and Vetter from 1964 using alternative case numberings. For the purposes of the present paper, case numbers from the latter study will be used.
Unfortunately the details from a number of cases were not reported (or available) in sufficient detail to meet the requirements relevant to testing the proposed hypothesis. Generally either age of onset and/or whether the phantom had a paired healthy limb or not were not specifically recorded. In consequence, the following studies/cases have been excluded: Valentin (1836) Cases 2 – 4; Valentin (1844); Mickorey (1954); Weinstein and Sersen (1964) case 15; Burchard (1965); Grouios (1996); Wilkins et al (1998). The author has no reason to believe that the absence of data from these studies will bias the investigation.
As can be seen from table 2, the majority of cases (25 out of 39) can indeed be explained via cross-cortical connections. However, it is equally clear that a many phantoms are not matched by paired limbs (groups III and IV).
Many subjects report phantom limb sensations as starting several years after birth (groups II and IV). The emergence of phantom sensations during childhood, (and, in some cases, as adults), implies that other influences are at work in the formation of body image than those that occur in utero, be they sensory based or hard-wired. Hence, it is possible that late-onset phantoms have a different source of origin than those who report experiencing phantom limbs for "as long as they can remember". Rather than learning their body images in utero from somatic input, the late onset unmatched phantoms may have been induced by other forms of sensory input, especially through observing other people and possibly via prolonged use of prostheses.
The importance of the role of visual input in the formation and maintenance of body image has been put to effective therapeutic effect in Ramachandran’s ‘mirror box’ treatment. By using this specially designed mirror box, the paired healthy arm is reflected in such a way that it fills the space that is felt to be occupied by the phantom arm. Subjects have found that moving the healthy limb and observing its reflection has enabled them to control the position and posture of the contralateral phantom (Ramachandran & Hirstein, 1998).
The possibility that prostheses may play a role in the formation of phantoms is emphasised by research which demonstrates that under certain conditions it is possible to incorporate inanimate objects into our conceptions of self (e.g. Ramachandran et al. 1998; Pavani et al. 2000).
On the basis of the proposals outlined above it is possible to propose two stages of development of body image: firstly, a pre-natal stage during which bilateral aspects of body image are ‘learnt’. Secondly, a post-natal stage, in which there is a consolidation of individual (lateralised) limbs during the first decade of life.
Accordingly, these stages correspond to two classes (or subtypes) of congenital phantom limbs; those which match the existing opposite limb and which are present from the earliest memories and those which emerge due to visual (or other) input later in life.
The differentiation of aplasics phantoms into sub-groups is generally compatible with the data. In Group I, every instance of phantom limb may be understood by cross-cortical connections during development in utero. Group II may originate from a combination of cross-cortical connections in utero and visual input. Group IV from visual input/prosthetic influences alone. Out of 39 cases, there are only three exceptions which continue to defy explanation: the phantom thumb and palm in a subject lacking both hands (case 26), and a case of bilateral aplasic phantoms arms (case 27), and the quadrilateral aplasia and phantoms described in case 28. These anomalies are addressed later in the paper.
3. Summary: The Quadripartite (four part) Hypothesis: a comprehensive framework for both phantom limb phenomena and for body image.
input from the body parts themselves hence avoiding the genomic redundancy and other complications that accompany the hard-wiring hypothesis (i.e. body image is learnt).
a) tangible interaction with the world, including touch and proprioception
b) visual input
c) prosthesis usage
4. Mirror neurons: the visual basis of body image.
For the quadripartite hypothesis (QH) to be tenable, some neural mechanism is required to explain how limbs could be learnt through observation. The strongest candidate for this mechanism is the mirror neuron system. It has been established that mirror neurons in monkeys (situated in area F5 in the ventral premotor cortex) are activated both when a monkey performs an action and also when it observes the same action being enacted by an experimenter (Gallese & Goldman 1998). Whilst the processes underlying the development of this form of ‘somatic empathy’ have yet to be uncovered, it seems likely that mirror neurons could play a crucial role in the development of body image, and especially so in cases of aplasic phantoms (Brugger et al. 2000).
5. Explaining the anomalies
Despite the explanatory power of the QH, three cases remain to be explained. (Case no.’s 26 - 28 in table 2, cf.). It is to these anomalies that we now turn.
5.1 Case 26:
Case 26 was reported by Weinstein et al. (1964). Whilst most of the subject’s phantom are described as matching the opposite limb (down to the wrist) an additional, unpaired, phantom palm and thumb were also present. Yet, if the mirror neuron hypothesis is accepted, the phantoms presented are not difficult to explain. Experiments on monkeys have shown that mirror neurons are especially sensitive to grasping postures, which by definition include thumbs and palms (Gallese & Goldman 1998). It follows from this that the phantom arm – which was matched by a corresponding paired (although handless) limb could have been learnt in utero via cross-cortical connections, whilst the thumb and palm were added later through visual input in the subject’s early childhood.
5.2 Case 27:
Ramachandran (1993) briefly describes a 20 year old woman (D.B.) with bilateral phantom arms. The case report notes the prominent role the phantom arms played in gesticulatory movement during conversation. In a discussion of the case Ramachandran and Blakeslee (1998: p.41) state that the subject reported experiencing the phantoms from as far back in her childhood as she could remember.
This report of complete bilateral aplasic phantom limbs that date back as far as memory permits offers a more substantial challenge to the quadripartite theory. Ramachandran and Hirstein have interpreted these phantoms as evidence for a genetically hard-wired basis for body image (Ramachandran 1993; Ramachandran & Hirstein 1998). As the case clearly cannot be explained by cross-cortical connections, nor by somatic input; could these phantoms have been ‘learnt’ visually so early in life?
As mentioned above, D.B.’s phantoms were particularly associated with gesticulation. Yet, as we know from cross cultural studies and the congenitally blind, gesticulatory body language is not simply hard-wired; it needs to be learnt. (e.g. Axtell 1991; Apple 1972).
According to Rizzolatti and Arbib (1998), mirror neurons and their corresponding hand movements may have played a crucial role in the evolution of language. Language requires a degree of shared understanding that is implicit in the activation of mirror neurons; moreover, it has been suggested that the cortical area F5 in the monkey brain is homologous to Broca’s area in humans (which is involved in speech) (Gallese & Goldman 1998). These theories imply that there may be a direct link between gesticulation - a form of body ‘language’ - and language proper.
Returning to the case under discussion, it is possible that D.B.’s gesticulating phantoms were learnt, like speech, early during childhood. This would be fully consistent with her report that she experienced has them for as long as she can remember. (How many people can remember speaking their first word?) Hence the case of D.B. is not in itself sufficient proof to establish the existence of a hard wired body image.
5.3 Case 28:
The strongest challenge to the QH comes from the final anomaly - a case of tetra-amelia described by Brugger et al. (2000). This case stands out from all the others on several counts. The woman, A.Z., who was 44 years old at the time of the study, was born without forearms and legs, yet reported experiencing vivid phantoms of both missing forearms, including hands and fingers, and both legs, including feet and first and fifth toes, from as long as she could remember. She did not use prostheses.
The veracity of A.Z.’s phantoms seems beyond doubt – she was an intelligent adult and her subjective accounts of phantom sensations were substantiated by both functional magnetic resonance imaging (FMRI) and transcranial magnetic stimulation studies.
Whilst mirror neurons might have contributed to the development of A.Z.’s upper limbs (as Brugger et al. note) mirror neurons are specifically associated with upper limbs and hands: it seems unlikely that they could be responsible for the phantom legs. In the two other cases of bilateral leg phantoms (cases 34 and 35), both are associated with the use or removal of prosthetic legs (Saadah & Melzack 1994). In contrast, A.Z. does not use prostheses. Hence, if the account presented by Brugger et al is both accurate and complete then this is the one and only case in which the evidence appears inexplicable without recourse to some kind of genetically determined, hard-wired body image, along with all the complications that this would entail. As such, this would potentially be a landmark study, and worthy of much greater attention. However, if regarded in light of the Quadripartite Hypothesis certain details reported by Brugger et al. take on a different cast, and an alternative interpretation presents itself.
Firstly, the FMRI studies indicated that phantom movements were associated with bilateral activation in both the dorsal premotor and posterior parietal cortices, whilst there was no activation of the primary motor cortices. Secondly, A.Z. rated pain and temperature in her phantom limbs as absent. According to the QH, both of these findings are indicative of body image in the uterine stage of development.
According to Brugger et al., A.Z. was born without forearms and legs, there was no history of abnormality in her family, and that "the reasons for her tetra-amelia are unknown". Given the significance of this case it must be queried whether that A.Z. was, in fact, subject to congenital aplasia. The study neither discusses nor precludes the possibility that A.Z. lost her limbs somehow in late pregnancy or early infancy (due to a botched childbirth, for instance). Other evidence in support of this possibility is provided by A.Z.’s comment that her phantom limb sensations were sometimes enhanced by genital stimulation. This phenomena that has been widely described in amputees for whom it has been interpreted as occurring due to the proximity of genitals to limbs on the somatosensory map. It has not previously been reported in any case of aplasic phantoms.
Hence, if interpreted as an early amputee rather than aplasic case, aspects of A.Z.’s case report provide evidence for the QH. Furthermore, their imaging studies suggest that the primordial bilateral body image, learnt in utero, its represented in the premotor and posterior parietal cortices, whereas lateralised body image is represented in the primary motor cortex. This is a possibility that may deserve further investigation.
6. Additional support for the quadripartite hypothesis from a non-aplasic case
Finally, further support for the quadripartite hypothesis can be drawn from a non-congenital case that is otherwise difficult to explain. La Croix et al. (1992) documented the case of a subject (K.G.) with multiple phantoms. K.G. was born with her right leg 10 cm shorter than her left. At age six the right leg below the knee was amputated for medical reasons. When interviewed several years after the operation she reported that, following the amputation, she had developed three phantom limbs: one in the image of the former limb; one was a ‘phantom foot’ at the terminus of the stump; and one was the full length of the matching leg.
The first holds no problems for the standard sensory theories as it could have been learnt from the shortened limb prior to amputation. The second is also compatible - stump phantoms are not uncommon, and the process (known as ‘telescoping’) is thought to result from the disproportionate areas of the brain dedicated to representations of the hands and feet relative to the limbs; this disproportion is itself explicable through sensory based theories. (Simmel 1962; Ramachandran & Hirstein 1998). However, Le Croix and colleagues interpreted the third of these phantoms (the ‘extended’ leg) as evidence for a genetically hard-wired body image, on the basis that there was no prior experience from any such leg. In contrast, the quadripartite theory is able to explain this phantom as a learnt phenomenon, reliant on cross-cortical connections. The quadripartite explanation is arguably the stronger in this case, as the hard-wired theory does not explain why the pre-programmed body image represents the longer left leg, rather than (or in addition to) the shorter right one.
7. Conclusion
In this paper, it has been argued that reports of phantom limbs amongst aplasics do not necessarily indicate that body image can develop without sensory input. Neither are such cases evidence of an "innate body schema" (Gallagher et al. 1998) nor of a hardwired, genetically determined neural body image (Melzack et al. 1997; Ramachandran & Hirstein 1998; Brugger et al. 2000). A quadripartite (four part) hypothesis is presented which offers a means of explaining the development of body image by multiple factors, including spontaneous muscle activity in utero, and post-natal sensory feedback from voluntary muscle activation, visual input and social interaction.
The quadripartite hypothesis not intended as a proof in itself, but rather as offering an alternative framework to those which claim the existence of a hard-wired body image. At present, neither the sensory nor the hard-wiring hypothesis can be ascertained with such certainty.
It seems unlikely that either theory will be ‘proved’ correct or false by individual cases, but rather that the resolution of the dispute must await further basic research on the physiological processes underlying development. If even only partially correct, the quadripartite hypothesis may have implications for current and future research into and understanding of pre and post-natal development of body image, sensory modalities, language and mirror neurons.
References:
André, J. M., Paysant, J. Martinet, N. & Beis, J. M. (2001) Classification et mécanismes des perceptions et illusions corporelles des amputés. Annales de Réadaptation et de Médicine Physique 44: 13-18
Apple, M. M. (1972) Kinesic training for blind persons: A vital means of communication. New Outlook for the Blind 66 (7): 201-208
Axtell, R. (1991). Gestures: The do's and taboos of body language around the world. New York: John Wiley and Sons.
Brugger, P., Kollias, S. K., Müri, 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: 6167-6172
Burchard, J. M. (1965) Zur Frage der Natur von Phantomlebnisse. Arch. Psychiatr. Z. Ges. Neurol. 207: 360-377 (cited by Vetter & Weinstein 1967, and by Brugger et al 2000, cf.)
Churchland, P.M. (1995) The engine of reason, the seat of the soul: a philosophical journey into the brain. MIT Press.
Gallagher S., Butterworth G.E., Lew A. & Cole J. (1998) Hand-Mouth Coordination, Congenital Absence of Limb, and Evidence for Innate Body Images.Brain and Cognition 38: 53-65
Gallese, V. & Goldman, A. (1998) Mirror neurons and the simulation theory of mind-reading. Trends in Cognitive Sciences 2: 493-501
Grouios, G. (1996) Phantom limb perceptuomotor ‘memories’ in a congenital limb child. Medical Science Research 24: 503-504
Grouios, G. (1998) Phantom limb sensations: a review. Medical Science Research 26: 219-223
Head, H. & Holmes, G. (1911-2) ‘Sensory Disturbances from Cerebral Lesions’ Brain 34: 102-254
Hepper P. G., McCartney G. R. & Shannon, E. A. (1998) Lateralised behaviour in first trimester human foetuses. Neuropsychologia 36: 531-534
La Croix, R., Melzack, R., Smith, D. & Mitchell, N. (1992) Multiple phantom limbs in a child. Cortex 28: 503-7
Melzack, R., Israel, R., Lacroix, R., Schultz, G. (1997) Phantom Limbs in people with congenital limb deficiency or amputation in early childhood. Brain 120: 1603-1620
Mikorey, M. (1952) Psychosomatische Betrachtungen über Phantome und Doppelgänger. Münchener Medizinische Wochenschrift 94: 44-46 (cited by Vetter & Weinstein 1967, cf.)
Pavani, F, Spence, C. & Driver, J. (2000) Visual capture of touch: Out-of-the-body experiences with rubber gloves. Psychological Science. 11(5): 353-359
Pinker, S. (1994) The Language instinct: the new science of language and mind. Allen Lane.
Poeck, K. (1965) Phantoms following amputation in early childhood and in congenital absence of limbs. Cortex 1: 269-275 (- aplasic phantom previously reported in German ibid. (1963) Zur Psychophysiologie der Phantomliebernisse. Nervenarzt 34: 241-256
Ramachandran, V. S. (1993) Behavioural and magnetoencephalographic correlates of plasticity in the adult human brain. Proceedings of the National Academy of Sciences of the United States of America 90: 10413-20
Ramachandran, V. S. & Blakeslee, S. (1998) Phantoms in the Brain (Fourth Estate, London)
Ramachandran, V. S. & Hirstein, W. (1998) The perception of phantom limbs. Brain 121: 1603-1630
Ramachandran, V. S., Hirstein, W. & Rogers-Ramachandran, D. (1998) Phantom limbs, body image, and neural plasticity. IBRO (International Brain Research Organization) News 26: 10-11
Rizzolatti, G. & Arbib, M. A. (1998) Language within our grasp. Trends in Neurosciences 21: 188-194
Saadah, E. S. M. & Melzack, R. (1994) Phantom limb experiences in congenital limb-deficient adults. Cortex 30: 479-485
Sengpiel F, Stawinski P, Bonhoeffer T (1999). Influence of experience on orientation maps in cat visual cortex. Nature Neuroscience 2(8): 727-732.
Simmell, M. L. (1962) The reality of phantom sensations. Social Research 29: 337-356
Sohn, L. (1914) The psychic complex in congenital deformity. New York Medical Journal 100: 959-961
Valentin, G. G. (1836) Ueber die subjectiven Gefüle von Personnen, welche mit mangelhaften Extremitäten geboren sind. Reprium Anat. Physiol. 1: 328-337 (as cited by Vetter & Weinstein 1967)
Valentin, G. G. (1848) Lehrbuch der Physiologie des Menschen 2: 716
Vetter, R. J. & Weinstein, S. (1967) The history of the phantom in congenitally absent limbs. Neuropsychologia 5: 335-338
Weinstein, S. & Serensen, 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 sensations in cases of congenital aplasia. Cortex 1: 276-290
Wilkins, K. L., McGrath, P. J., Finley, G. A. & Katz, J. (1998) Phantom limb sensations and phantom limb pain in child and adolescent amputees. Pain 78: 7-12
Acknowledgements
I would like to thank the MRC and Wellcome Trust for funding, the University of Oxford for institutional backing, and both Robert Österbauer and especially Jacki Rosenfeld for their suggestions and support.