Now on ScienceBlogs: Open Lab: Time is Ticking!

Seed Media Group

Collective Imagination

« Business Books | Main | The Neuroscience of Admiration »

The Hollow Mask Illusion

Posted on: April 14, 2009 11:19 AM, by Jonah Lehrer

Wired Science reports on a fascinating finding: schizophrenics have trouble seeing the hollow mask illusion, in which people perceive the concave inside of a mask as an actual face. The reason we're vulnerable to this illusion is that our expectations of what we'll see - we're used to seeing real faces - profoundly influence our actual sensations. Schizophrenics, however, seem to have trouble with modulating their perceptions, which might explain why they persistently believe in delusions and far-fetched conspiracy theories. (Interestingly, the only other group of people who don't see the hollow mask illusion are those drunk on alcohol or high on cocaine.)

The polite term for this mental ability is "top-down processing," a term that describes the way cortical brain layers project down and influence (corrupt, some might say) our actual sensation. After the inputs of the eye enter the brain, they are immediately sent along two separate pathways, one of which is fast and one of which is slow. The fast pathway quickly transmits a coarse and blurry picture to our prefrontal cortex. Meanwhile, the slow pathway takes a meandering route through the visual cortex, which begins meticulously analyzing and refining the lines of light. The slow image arrives in the pre-frontal cortex about 50 milliseconds after the fast image.

Why does our mind see everything twice? Because our visual cortex needs help. After the prefrontal cortex receives its imprecise picture, the "top" of our brain quickly decides what the "bottom" has seen, and begins doctoring the sensory data. The end result is we see a face where there are only concave shadows.

Sure enough, this latest experiment saw differences in activation in the frontoparietal network in schizophrenics as they looked at the illusion, suggesting that the lack of top-down processing was responsible. The irony, of course, is that the absence of top-down processing doesn't simply lead to a more literal view of the outside world. It could also be responsible for schizophrenic hallucinations. T.S. Eliot was right: "Humankind cannot bear very much reality."

Share on: Stumbleupon Reddit Email + More

Trackbacks

Trackback URL for this entry: http://scienceblogs.com/mt/pings/106495

Comments (11)

1

T.S. Eliot was right: "Humankind cannot bear very much reality."

Or maybe it's just that the two systems are vulnerable to different kinds of errors. Most of time, when working together, each cancels the errors of the other. But you can find corner cases where top-down forces an error on bottom-up, or vice versa.

It's not that humans can't bear too much reality - it's that neither system alone can perceive very much of reality.

Sorry, a cute joke ruined by over-analysis. :->

Posted by: Ray Ingles | April 14, 2009 12:26 PM

2

So the $64k question is, "Is there any way to artificially stimulate more top-down processing in any subject?"

Posted by: Russell Kilbane | April 14, 2009 12:34 PM

3

I've actually been writing a series of posts on my blog about top-down processing in visual systems. There is still one upcoming post, but the most recent one (with links to the previous parts) can be found here. I find it interesting just how little people generally acknowledge our post-sensory processing.

Posted by: Mozglubov | April 14, 2009 12:40 PM

4

Fascinating. Seems like the brain has something of a 'hallucination engine' that's always at work but kept in check by PFC. Goya has a nice painting that gets at this idea: 'The sleep of reason produces monsters'.

Posted by: jbc | April 14, 2009 2:10 PM

5

Perhaps the hollow mask illusion illuminates an example of a human tendency toward anthropomorphism? :)

But anyway, that research is very interesting. And it seems to confirm Gerald Edelman's theory of reentry.

Posted by: Lexi | April 14, 2009 2:45 PM

6

Being told by my computer that I'm not schizophrenic is the nicest thing that has happened to me today.

Yeah, going to bed on a high note...

Posted by: Fertanish | April 14, 2009 11:02 PM

7

It's been known for a long time that schizophrenics are immune to the illusion, but this is the first time anyone's tried to discover the neural correlates of it.

Interestingly, cannabis has also been shown to make people immune to the illusion (in a slightly different form).

The methodology used in this study was actually a little more complicated than in a typical fMRI experiment - researchers in London are always playing around with sophisticated fMRI techniques. I think I ought to blog about it now...

Posted by: Neuroskeptic | April 15, 2009 5:45 AM

8

Mozglubov, in figure 1 from your blog posting, I saw a second illusion. There were faint dark quarter circles in the corners of the white square, as if the white square was not quite opaque and the hidden quadrants of the black circles were just barely visible through it. (If people start seeing these dark corners after having been told of them, that would be another illustration of top-down processing.)

Posted by: David Canzi | April 15, 2009 12:06 PM

9

I believe LSD illustrates something similar: Humans cannot deal with the 'realness' of reality.

Posted by: Rhodora Online | April 20, 2009 1:56 PM

10

While I was sitting a meeting, a colleague showed me a picture on his blackberry. It was a cross-section of a candy bar with a bite taken out of it. "Jesus," he said. The hollow mask illusion probably goes a long way to explaining miracles such as these.

Posted by: Ryan Shewcraft | April 20, 2009 2:17 PM

11

I've got a bit more about this illusion here...

Posted by: Neuroskeptic | April 23, 2009 9:58 AM

Post a Comment

(Email is required for authentication purposes only. On some blogs, comments are moderated for spam, so your comment may not appear immediately.)

ScienceBlogs

Search ScienceBlogs:

Go to:

Advertisement
Enter to win a free copy of The Monty Hall Problem
Visit the Collective Imagination blog
Advertisement
Collective Imagination

© 2006-2009 Seed Media Group LLC. ScienceBlogs is a registered trademark of Seed Media Group. All rights reserved.

Sites by Seed Media Group: Seed Media Group | ScienceBlogs | SEEDMAGAZINE.COM