Retinal physiology and the demise of the pure percept

Rooming with 2 philosophy majors warps the mind even if it was 50 years ago.  Conundrums raised back then still hang around.  It was the heyday of Bertrand Russell before he became a crank.  One idea being bandied about back then was the ‘pure percept’ — a sensation produced by the periphery  before the brain got to mucking about with it.   My memory about the concept was a bit foggy so who better to ask than two philosophers I knew.

The first was my nephew, a Rhodes in philosophy, now an attorney with a Yale degree.  I got this back when I asked —

I would be delighted to be able to tell you that my two bachelors’ degrees in philosophy — from the leading faculties on either side of the Atlantic — leave me more than prepared to answer your question. Unfortunately, it would appear I wasn’t that diligent. I focused on moral and political philosophy, and although the idea of a “pure precept” rings a bell, I can’t claim to have a concrete grasp on what that phrase means, much less a commanding one.

 Just shows what a Yale degree does to the mind.

So I asked a classmate, now an emeritus prof. of philosophy and got this back
This pp nonsense was concocted because Empiricists [Es]–inc. Russell, in his more empiricistic moods–believed that the existence of pp was a necessary condition for empirical knowledge. /Why? –>
1. From Plato to Descartes, philosophers often held that genuine Knowledge [K] requires beliefs that are “indubitable” [=beyond any possible doubt]; that is, a belief counts as K only if it [or at least its ultimate source] is beyond doubt. If there were no such indubitable source for belief, skepticism would win: no genuine K, because no beliefs are beyond doubt. “Pure percepts” were supposed to provide the indubitable source for empirical K.
2. Empirical K must originate in sensory data [=percepts] that can’t be wrong, because they simply copy external reality w/o any cognitive “shopping” [as in Photoshop]. In order to avoid any possible ‘error’, percepts must be pure in that they involve no interpretation [= error-prone cognitive manipulation].
{Those Es who contend  that all K derives from our senses tend to ignore mathematical and other allegedly a priori K, which does not “copy” the sensible world.} In sum, pp are sensory data prior to [=unmediated by] any cognitive processing.

So it seems as though the concept is no longer taken seriously.  To drive a stake through its heart it’s time to talk about the retina.

It lies in the back of our eyes, and is organized rather counter-intuitively.  The photoreceptors (the pixels of the camera if you wish) are the last retinal elements to be hit by light, which must pass through the many other layers of the retina to get to them.

We have a lot of them — at least 100,000,000 of one type (rods).  The nerve cells sending impulses back to the brain, are called ganglion cells, and there are about 1,000,000 in each eye.  Between the them are bipolar cells and amacrine cells which organize the information falling on the photoreceptors.

All this happens in something only .2 milliMeters thick.

The organization of information results in retinal ganglion cells responding to different types of stimuli.  How do we know?  Impale the ganglion cell with an electrode while still in the retina, and try out various visual stimuli to see what it responds to.

Various authorities put the number of retinal ganglion cell types in the mouse at 11, 12, 14, 19 and 22.  Each responds to a given type of stimulus. Here are a few examples:

The X-type ganglion cell responds linearly to brightness

Y cells respond to movement in a particular direction,

Blue-ON transmits the mean spectral luminance (color distribution) along the spectrum from blue to green.

From an evolutionary point of view, it would be very useful to detect motion.  Some retinal ganglion cells being responding before they should. How do we know this?  It’s easy (but tedious) to map the area of visual space a ganglion cell responds to — this is called its receptive field.  The responses of some anticipate the incursion of a moving stimulus — clearly this must be the way they are hooked to photoreceptors via the intermediate cells.

Just think about the way photoreceptors at the back of the spherical eye are excited by something moving in a straight line in visual space.  It certainly isn’t a straight line on the retinal surfaced.  Somehow the elements of the retina are performing this calculation and predicting where something moving in a straight line will be next.  Why  couldn’t the brain bedoing this?  Because it can be seen in isolated retinas with no brain attached.

Now for something even more amazing.  Each type of ganglion cell (and I’ve just discussed a few) tiles the retina. This means that every patch of the retina has a ganglion cell responding to each type of visual stimulus.  So everything hitting every area of the retina is being analyzed 11, 12, 14, 19 or 22 different ways simultaneously.

So much for the pure percept: it works for a digital camera, but not the retina.  There is an immense amount of computation of the visual input going right there, before anything gets back to the brain.

If you wish to read more about this — an excellent review is available, but it’s quite technical and not for someone coming to neuroanatomy and neurophysiology for the first time.  [ Neuron vol. 76 pp. 266 – 280 ’12 ]

Post a comment or leave a trackback: Trackback URL.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: