A group of neuroscientists and a philosopher have devised a
series of novel experiments using virtual reality that could
shed light on decades of clinical data pointing to cognitive
and perceptual mechanisms involved in humans' concept of self.
Their results, published August 23 in Science Magazine, show
that a person's sense of self can be manipulated using conflicting
multisensory bodily input, indicating that spatial unity and
bodily self-consciousness depend on brain mechanisms and can
be explored experimentally.
The "I" one thinks of as "myself" is inextricably attached
to one's bodily location. In patients with certain neurological
conditions this sense of spatial unity can break down, causing
disturbing sensations such as out-of-body experiences in which
the global self is localized outside one's body limits (often
Previous experiments have shown that people may attribute fake
body parts to their own bodies. In the "Rubber Hand Illusion",
a person's unseen hand is stroked synchronously with a visible
fake hand, and then the person is asked to point to his own
hand. Subjects invariably err in the direction of the fake hand,
attributing it to their own bodies. Because the attribution
does not involve the whole body, the sense of global bodily
self-consciousness is not affected. EPFL Professor Olaf Blanke,
graduate students Bigna Lenggenhager and Tej Tadi, and philosopher
Thomas Metzinger hypothesized that the same approach could be
used to study the concept of global bodily self consciousness
by using a single, coherent body representation instead of just
a body part.
Working with EPFL computer engineers, the researchers designed
a series of simple virtual reality experiments in which a subject
saw a projection of a three-dimensional representation of his
own body, the body of a dummy, or a simple object directly in
front of him. The subject then saw the back of the image being
stroked with a paintbrush, either in or out of sync with someone
stroking his own back. Immediately after, the subject was blindfolded
and backed up, and then asked to return to his original position.
Subjects whose backs were stroked synchronously with the virtual
image of himself or the human dummy consistently overshot their
position in the direction of the image; but subjects who saw
no virtual image or a simple object did not. The synchronously
stroked subjects went farther in the direction of the virtual
image than those who were stroked out of sync.
According to the researchers, several subjects reported feeling
"weird" but none actually reported the disembodiment classically
described in an out-of-body experience. They knew the body image
was not theirs. Despite this, the subjects still localized their
selves to a position outside their own bodies, indicating that
the brain is compiling a sense of spatial unity from an integration
of visual, somatosensory and cognitive input, in which the visual
appears to dominate.
In a departure from decades, if not centuries, of philosophical
arguments that self-consciousness is a uniquely human trait
related to language, memory and the capacity of self-referral,
Blanke and colleagues' research indicates that humans' sense
of the embodied self depends on brain mechanisms at the temporo-parietal
junction. Experimentally creating illusions of the global self
using virtual reality technology could open up avenues for investigating
the neurobiological, functional and representational aspects
of the embodied self, potentially in other primates as well
as in humans.
Blanke's group plans future research using this approach to
investigate a spectrum of disturbed body perceptions ranging
from body-related hallucinations to full-blown out-of-body experiences
in patients with brain damage or psychiatric illnesses.