In the last post, I described a bit about how the brain codes the peripersonal space around the hand.
But to expand before a recap, let’s say we’re able to record single neurons from all over the macaque’s brain. What we would find when the monkey looks at a banana from across the room would be cells selective for bits of curvature and for the specific yellow hue of the banana, all firing in primary visual cortex. As you move forward in the brain from primary visual cortex to secondary areas , more abstract representations of shape are formed by cell assemblies. The area of space occupied by the banana will excite cells that repond to objects at that location in the visual field. Location information is coded along the so-called dorsal stream (the “where” stream of processing), while the “identity” of the banana will be represented by activity in inferotemporal regions, in the ventral or “what” stream of processing. Computations for reaching occur in the “where” stream; indeed, Goodale and Milner have called this pathway the “how” stream, to emphasize the nature of the activity therein. A great dorsal/ventral stream graphic is here.
If the banana was within reaching distance of the monkey though, populations of cells responsive to the hand’s position in space would fire. These cell assemblies are located in the postcentral gyrus and other superior and inferior parietal lobe areas. Crucially, if the monkey has been using a rake for 5 minutes to retrieve food, the cells that normally fire for the object when it is within reaching distance will now fire even though the banana is the same (far) distance away.
To recap the last post, there are ways to fool the human peripersonal space system to think that a rubber arm is a real arm by simultaneously stroking the real hidden arm and the in-view rubber arm. We know the illusion affects behavior because in neglect patients we see extinction of left-hand touch when the rubber right hand is touched or nearly touched. The rubber hand illusion is strong; when the procedure was performed on me I really felt as if my hand and arm under the table were the rubber arm I was viewing on top of the table.
In unilateral visual neglect, many patients will fail to respond to objects in the left side of space. If presented with small lines and asked to make a mark across each line, they will “cancel” lines on the right side of the page but not on the left. If you ask a neglect patient to bisect a line using a pencil, the patient will reliably mark the line to the right of the actual center; these line bisection errors are because of the patient’s inability to properly attend to the left side of space (and/or the left side of objects, by the way).
A clear distinction in behavior emerges, though, when you ask a neglect patient to bisect in extrapersonal space (maybe 150 cm away) as opposed to peripersonal space (maybe 50 cm away), by indicating the center point of a line from far away with a laser pointer. In extrapersonal space, bisection performance can be more accurate than in peripersonal space. The opposite pattern is also seen in some cases, indicating even more clearly a neural dissociation in the representation of near and far space, confirmed in neurologically intact people with PET.
Bisection in neurologically intact people isn’t perfect, either. People like you and me will show “pseudoneglect” in line bisection tasks — we’ll bisect to the left of the center of the line, rather than rightward as with the patients. McCourt and Garlinghouse (2000, Cortex full .pdf available here) show this effect and so do Bjoertomt, Cowey and Walsh (2002, full .pdf here). Critically, though, pseudoneglect is also sensitive to the position in space in which the bisection is to occur. Pseudoneglect is more prominent in peripersonal space than in extrapersonal space.
So my suggestion is to engender the rubber arm illusion in normal participants in a two-stage process; first, for a few moments with a normal sized arm, then for a few minutes with an arm twice as long. Why not go for the gusto!
There are ways to assure the effect is felt, by using the original Botvinick and Cohen questionnaire. We can also measure the illusion somewhat more directly (as B & C did and others have since them) by looking at proprioceptive drift. In general, the position of the hand that is projected onto the rubber arm is mislocated in the direction of the rubber arm rather reliably. In this experiment, proprioceptive drift should be seen not horizontally but along the length axis of the arm.
So, if one can engender the feeling that an arm is very long, you’ve then changed the boundary between peripersonal and extrapersonal space. Now immediately use the “elongated” arm in line bisection tasks. Where (within the same subject, most powerfully) you once showed pseudoneglect disappearing as the participant performed the task in extrapersonal space, now the pseudoneglect should reappear!
By the same logic, one could foreshorten arm length as well with the rubber hand illusion, so that former peripersonal space could become extrapersonal space, again measureable by shifting line bisection performance.
Some body distortions such as limb foreshortening happen in patients experiencing autoscopy or out-of-body experiences (OBEs), and guess where the locus of the effect is? Right you are, inferior parietal cortices, generally speaking. See all the great work of Olaf Blanke for confirmation.
In summary, this set of experimental predictions is one more attempt, among many out there at the moment, to characterize the way the self is embodied. Researchers have learned a great deal about the mechanisms by which the nervous system codes interactions of vision and touch, and under what circumstances the visual system can override touch. This proposed experiment is along the latter line of investigation, and I’m trying to provide a benchmark by which we might see “how far is too far” in the brain’s elaboration of the body in space.
by a guy who has filled his site with, ahem, interesting animations…