Archive for October, 2011


Follow up to vision research

A friend of mine showed me this blog post, and it’s pretty awesome. I’ve known about these techniques for a while (upside faces look weird because the face “gestalt” is not registered, but this post brings them home pretty nicely

The point is that learning to draw well is actually learning to undo your natural biases — the biases that let us function so easily in everyday situations. We see a door as a rectangle, a sliver, a square and everything in between. At least these objects makes these shapes ON THE RETINA, but that’s now how we see them. We see them as a rectangular door, just from many perspectives. Same with

My hypothesis is that judgments about morality and what we have reason to do are similar. We have built in biases that make certain actions stand out to us in certain ways which makes it easy to navigate society, but unless we become “painters” of morality and practical reason, then we are at the mercy of these shortcuts. We must learn to see around them.

Also from my last post, still the thing that is most interesting to me is that we couldn’t watch TV if our brains created our perception of the world differently. What more proof could you need that TV makes you fat other than the act of watching engages the opposite visual system than the ACTING visual system. Almost by definition watching TV can’t involve acting. Only perceiving at its most passive.


Amazing Book on Perception

I just finished “Sight Unseen” by Melvyn Goodale and David Milner, two researchers of percetion.

This book is just really good. It’s short and accessible but awesome in the way it will change your thinking.

The main idea is that we have two perceptual systems, one for action and one for representation. The deepness in which these themes are pursued is truly impressive.

Dee Fletcher, the main research subject of the book, lost the ability to see objects as objects. Her visual experience is hazy and objects seem to “run together,” but she can see color and texture. HOWEVER, she is still able to act like normal human beings. She can mail a letter in a small slot, grasp objects of a variety of shapes and so on. The evidence amassed over the course of the book is impressive. The representational system is closely tied to consciousness and is not much use for action. If we see an object, we can pick it up easily, quickly, and with the right amount of visua-motor adjustment. If we wait a few seconds while our eyes our closed and SIMULATE picking up the same object, we’re slower and less adroit with our motions, the hypothesis being that information for our motor system clears out after a short period of time, leaving us with our perceptual system, which is not built for action.

The idea is strengthened by patients who have damaged visual systems for action. These people can see everything crystal clear, but have trouble and uncertainty ACTING. If such people wait a few seconds to engage with an object, their performance improves, because rather than engaging their ordinarily superior (but now broken) motor-vision system, they engage their work (but second-rate) representational-vision system.

This is used to explain a wealth of ordinary and fascinating data. First, why do young kids draw the same pictures of houses and cars, but master artists can introduce perspective and the appearance of reality? The reason is that our eyes work “behind our backs” to create object recognition over time. A black shirt in sunlight reflects more light than a white shirt in a dim room, but the former still looks black and the latter still looks white. That’s our brain helping us and compensating for what we “see” (see is in quotes because I’m using that word to refer crudely to just the light that hits our retina. Real perception is much, much more than just light hitting the retina). So, young children draw things as the OBJECT is. We recognize a door from all sorts of positions, even though a door’s shape on our retina can be trapezoidal, parallelogram, or square depending on the perspective. But throughout it all we recognize a door, and that’s why kids draw the door as the object the brain cues to us — the rectangular straight on object. Master artists have in a sense mastered the ILLUSION and clues our brain uses to help us see perspective.

Also, why is it hard to mime actions. Well, because of what I said before. We’re quite bad at mimicking the FLUIDITY of action that is guided by our motor system. Painters have to learn to recreate what our REPRESENTATIONAL visiionsystem does for us, but mimes must learn to duplicate actions normally controlled by our ACTIONAL visual system. That’s why miming is really hard.

Something that really blew me away, is that TV would be impossible without our representational visual system. All the lower mammals have an actional visual system, but only we have a representational system which works mainly by relative distances and sizes. Thus, everything on a TV looks right, because its compared to everything else on the TV. Godzilla looks big (even though he’s small) and 3-d (even though he is on a 2-d screen).

The MOST interesting thing though was that our representational system can be fooled by illusions but our actional visual system CANNOT. When looking at a line illusion (with the arrows) we can’t help but SEE the lines as different lengths, but if we try and reach out to grab the lines, our fingers will stretch the right distance. They won’t be fooled by the illusion. What this means is that acting in the world reveals the truth of the world in a special way. ONLY through action can we work around some illusions. The world is revealed to us in a special way when we TAKE action. Is there some relationship to the wall street protests, i.e. when we ACT on our political beliefs rather than just think about them, the world maybe revealed to us in a special way. I’m fascinated by that thought.


Efficiency Loss and a great CBO report

This CBO report  is really good on scoring various policy proposals and how much money they will save.

Pg. 211 is what I’m interested in these days. Municipal bonds are tax exempt (no one pays taxes on the interest one receives from them) and so they are bought a lot by investors. But, we can incentivize those investors to the same amount by making them pay taxes and then give a tax credit back. This would gain the U.S. treasury a small sum of money AND have the same effect as the current tax free status.