The parable of the Blind men and an elephant, a group of blind men (or blindfolded people or people in complete darkness) touch an elephant to learn what it is they are facing. Each person feels a single part such as a leg or the trunk or a tusk. The “blind men” then compare notes and discover that they are in complete disagreement. According to Wikipedia, the parable has been used to provide insight “into the relativism, opaqueness or inexpressible nature of truth, the behavior of experts in fields where there is a deficit or inaccessibility of information, the need for communication, and respect for different perspectives.”
But the parable also applies to a single person who has to figure out from touch alone what object it is they are confronting (and that they cannot see because they are blindfolded or otherwise prevented from seeing it). Imagine using a single finger to run along on unfamiliar shape and trying to figure out what it is—your brain is confronted with a task not unlike that of the group of blind men who need to integrate their various perspectives into one agreed “percept.”
A recent article in Attention, Perception, and Psychophysics, a journal of the Psychonomic Society, examined how people use “haptic” (touch) information to form an impression of the 3D structure of objects.
Researcher Hiroyuki Mitsudo departed from the known fact that the human vision system “fills in the blanks” when it is confronted with partial information. As Dr. Mitsudo explained, “what we see is more than what the retina receives. If we see a big animal (say, a bear) behind a tree, the light from the occluded body part does not reach to the retina. Nevertheless, we perceive a whole bear rather than the fragmented visible parts. Our visual system has mechanisms that actively compensate for lost information by using visual inputs from surrounding locations. Is this the case in touch? That is, is it possible to haptically recognize object parts that are not directly touched?”
In other words, can you recognize an elephant from touching its tusk and its trunk?
To investigate this question, participants had to place their hand on an array of small bars that were separated by a varying distance. The figure below shows the stimuli and their various configurations:
Crucially, participants touched the top part of bars, but not the bottom of the “trenches” formed between the bars. Participants touched the bars while keeping their hand flat, as shown in the figure below. The hand was hidden from their view behind a screen so people could not rely on their vision to do the task.
Participants were asked to estimate how deep the “trenches” were between the bars. Note that touching the top of the bars does not yield any direct information about the trenches at all: If one wanted to make any guesses about their depth, then it would be based on an inference that is constructed in the same way that we “see” a complete bear behind the trees.
It turned out that people apparently “felt” the trenches similar to the way we “see” the complete bear among the trees: The judged depths ranged from 20 mm to 65 mm, and the estimates were positively correlated with the physical separation between touched elements. In a nutshell, the further apart two things are, the deeper you believe the trench in between will be.
Dr. Mitsudo conducted several follow-up experiments that are particularly intriguing because he employed the stimuli from the first study as “distractors” in a memory task. Briefly, participants first acquired a memory for a vertical position by touching a stimulus panel. They then touched a “distractor” just like the stimulus above, and following that touch they reported the memorized vertical position by touching a response panel. The question of interest is whether the perceived depth of the “trenches” in the distractor stimulus might affect the recall of the vertical position.
To explain this in a bit more detail, the apparatus for the experiment (which was hidden from participants’ view) is shown below:
On each trial, the participant would touch the target panel and memorize its vertical position. The person would then touch the distractor before recording the response on the left by marking the vertical position of the target on the response panel.
The results showed that people’s responses were biased in the direction of the presumed depth of the trenches in the distractor stimulus: The further apart the bars of the distractor, the lower people thought the initial target had been. People’s memory was distorted by the implied depth of objects they touched after memorizing a stimulus.
These intriguing results suggest that people form a 3-D representation from touching a 2-D surface, and that this representation obeys regularities often found in the environment—similar to the way that we can “see” a bear behind the trees, we can “feel” the trenches in between things that touch our hand. And those representations are sufficiently powerful to alter our memories.