 |
| Jason Braithwaite |
 |
| Hayley Dewe |
The rubber-hand illusion (RHI) involves experimentally inducing an anomalous body experience in observers (Botvinick & Cohen, 1998). Typically observers report everything from mild sensations such as their real hand getting cold, to more striking experiences where they 'think' the rubber hand is connected to their actual body. This protocol is a particularly useful way to explore the underlying neurocognitive processes of embodiment with implications for scientific and philosophical theories of self-consciousness.
The RHI protocol has been used to examine populations who are known to be predisposed to aberrant experiences of the bodily self (e.g., patients with schizophrenia / eating disorders: Burrack & Brugger, 2005; Peled et al., 2000; Mussap & Salton, 2006; Thakkar et al., 2011). The aim here has been to assess any apparent biases in multi-sensory integration that may underlie spontaneous breakdowns of embodiment and account for many of the positive hallucinatory experiences reported by such groups. In this discussion we want to focus on two issues; (i) the over-arching metaphorical account for the RHI in certain patient groups and; (ii) the actual empirical evidence recruited in support of that account.
In the literature the predominant account appears to be that the RHI reflects a general 'malleability' in body representation. Central to this view is the idea that weaker representations of the body-image are simply easier to override and this weakness in embodiment underlies a host of spontaneous anomalous body experiences reported by patients with schizophrenia, eating disorders, etc. (Burrack & Brugger, 2005; Peled et al., 2000; Mussap & Salton, 2006; Thakkar et al., 2011).
Therefore, the easier and faster it is to induce the illusion, and the stronger it is rated, then the more 'malleable' the body-image is said to be. However, while conceptually seductive and appealing, this notion of malleability is at best nebulous, and at most little more than metaphorical. In addition, this poses a conundrum. If the body-image is weakly represented in those prone to anomalous bodily experiences, how can it be that the RHI is said to be easier and faster to induce when it engages with and recruits the same impaired multi-sensory integration processes? Surely, if the same weakened processes that establish and maintain embodiment are recruited in the RHI, then arguably it would be harder to induce the RHI? In other words, impairments predisposing patient populations to report spontaneous anomalous bodily experiences, would surely also deaden the impact of the incoming visuo-tactile stimulation from the RHI procedure?
It also follows that it is not clear how the illusion can be 'stronger' and more convincing in certain patient populations, while at the same time, their own more permanent body-image is said to be 'weaker'. How can this be? The general concept of malleability suggests that the very processes the RHI seeks to tap into may well be the very processes that are hampered, biased, or impaired - which could conceivably impact on the induction of the RHI. At a fundamental level, there is some friction here between these ideas and while not irreconcilable, they await further clarification. We are not arguing here that the notion of malleability cannot be rescued or refined, just that it is under-specified and perhaps to such an extent that its intuitive appeal distracts us from its true explanatory power.
In addition to these conceptual issues, the empirical demonstrations that individuals with disorders of embodiment actually show stronger RHIs or that they can be induced faster isnot well founded. For example, in the first study reported by Peled et al. (2000), patients with schizophrenia were compared to controls on the RHI. However, there were no objective measures of the illusion and there was no control condition within the experiment (see also Mussap & Salton, 2006 for similar limitations). It is well known that patients with schizophrenia, reporting active hallucinations, are also incredibly prone to response biases and the power of suggestion. Their increased endorsements of the questionnaires in those studies can be easily explained by these factors - and thus, may not reflect the presence of quicker or stronger illusions at all. Indeed, a methodologically superior study (Thakkar et al., 2011) showed that patients with schizophrenia did indeed endorse all items on an RHI questionnaire more for both the main illusion and control conditions - consistent with the notion that suggestibility is indeed playing a role in the RHI.
Furthermore, MacLachlan, et al. (2003), have already shown that from a number of body-illusions, the RHI is particularly susceptible to suggestion. A close examination of these studies shows that the fundamental assumption of 'malleability' is not only a nebulous concept, but it also remains to be empirically demonstrated in relation to those prone to aberrant states of body consciousness. The problem is further compounded by a recent study from our own laboratory on non-clinical individuals who were predisposed to anomalous body experiences (Braithwaite et al., 2014). We found that hallucinators took longer to declare the specific presence of the illusion relative to controls. We also identified objective autonomic indicators (electrodermal activity) that appear to be related to the competition in neural processes sub-serving such aberrations of self-consciousness.
Our findings suggest an opposing interpretation, where there may well be more noise in the mechanisms responsible for setting up multi-sensory contingencies supporting embodiment. Our tentative suggestion was that these findings were in line with a "dysconnection" account - where aberrant connectivity between and within networks supporting embodiment were impacting on the conscious experience of the bodily self. Whether such dysconnection merely impacts on the setting up of representations, or also the strength of them when established (or both) remains to be seen. Even if we view malleability as a consequence of dysconnection, the usefulness of this metaphor may now have expired, and neuroscientists may want to cultivate more helpful mechanistic frameworks which are capable of generating clearer questions that are also backed up by empirical findings of a higher quality.