Virtual out of body experiences

In 2005, Olaf, his PhD student Bigna Lenggenhager, and I embarked on
a series of virtual-reality experiments. Our first goal was to turn the
OBE into a fully replicable phenomenon in healthy subjects. Proper research
required that we be able to investigate and repeat out-of-body experiences
in the lab. The guiding question was whether there could be
an integrated kind of bodily self-consciousness that is a phenomenal
confabulation. In short, could one experience a hallucinated and a bodily
self at the same time, a full-body analog of the rubber-hand illusion?
Here is an example of one of our early experimental protocols using
the paraphernalia of virtual reality: a head-mounted display (HMD)
consisting of goggles that showed two separate images to each eye, creating
the three-dimensional illusion of being in a virtual room. Subjects
were able to see their own backs,
which were filmed from a distance of
2 meters and projected into the threedimensional
space in front of them
with the help of a 3D-encoder. When I
acted as the subject of the experiment,
I felt as if I had been transposed
into a 3D-version of René Magritte’s
painting La reproduction interdite.
Suddenly I saw myself from the back,
standing in front of me.
While I was looking at my own
back as seen in the head-mounted display,
Bigna Lenggenhager was stroking
my back, while the camera was
recording this action. As I watched my own back being stroked, I immediately had an awkward feeling: I felt
subtly drawn toward my virtual body in front of me, and I tried to “slip
into” it. This was as far as things went.
Our studies became more systematic. All of our subjects would be
shown their own backs being stroked (this was the “own-body condition”)
and in a subsequent test would be shown either the back of a
mannequin (the “fake-body condition”) or a large rectangular slab
(which didn’t look like a body at all, the “object condition”) being
stroked. An additional condition was the degree of synchronicity between
the seen and the felt stroking, which could be varied by projecting
the camera image into cyberspace with a certain time lag.
Afterwards, an independent measure for the strength of the illusion
was introduced. They were blindfolded, moved around and disoriented,
as in a game of blindman’s buff, and then asked to return to their initial
position.
At the end of the experiment, the subjects were asked to fill out a
questionnaire about their experiences. Results showed that for the synchronous
conditions in which they were observing either their own
body or a mannequin, they often felt as though the virtual figure was
their own body, actually identifying with and “jumping into” it. This impression
was less likely to occur in the case of the wooden slab, as well as
in all of the asynchronous conditions. The synchronous experiments
also showed a significantly larger shift by the subjects toward the projected
real or fake body than did the asynchronous control conditions.
In other control conditions, subjects observed a screen without a body
in it and were then displaced (visual scene), or were simply displaced
only. These data suggest that locating the “self” in the case of conflicting
visual and somatosensory input is as prone to error as was reported for
a body part in the rubber-hand illusion.
Here is what I call the “embedding principle”: The bodily self is phenomenally
represented as inhabiting a volume in space, whereas the seeing
self is an extensionless point—namely, the center of projection for our
visuospatial perspective, the geometrical origin of our perspectival visual
model of reality. Normally this point of origin (behind the eyes, as if a little
person were looking out of them as one looks out a window) is within the
volume defined by the felt bodily self. Yet, as our experiments demonstrated, seeing and bodily self can be separated, and the fundamental
sense of selfhood is found at the location of the visual body representation.