Newton said the universe was a clock. The brain is a machine too, and we should be able to understand it. I think it would give more insight into what makes people people.

What event led you into research?

When I got to college, I was pre-med, and I got fascinated by spending my life solving puzzles...I went to graduate school in molecular biology for a year, and hated it for various reasons. I realized that what physicians did was make critical decisions on inadequate data, and that seemed like fun, so I went to medical school. And when I went to medical school, I had never been exposed to psychology or to any aspect of neurobiology as an undergraduate. In my first year of medical school there was this fabulous course, the Harvard neurobiology course, which was taught by six faculty, half of whom had been president of Society for Neuroscience, or were members of the National Academy [of Sciences]. This was an "oh wow, this is what I want to do when I grow up." ... Just the idea that you could understand the brain as a machine, I thought, what a great thing. Then there was a war in Vietnam, and every physician with a Y chromosome had to go into a uniformed service. And I was lucky enough to do it in Bethesda. Bob Wurtz was starting his lab. I was his first post-doc.... I was committed enough to being a doctor that I went back to do a residency, and I still do both....

The threads that have run through my research career are: first of all, how does the brain generate eye movements; second, how does the brain generate visual attention, and how does the brain create spatial accuracy out of this fragmentary sensory input? Then, what does free choice look like at a neural level?

And as I said before, it's all dedicated to writing the physics of behavior. Newton said the universe was a clock. The brain is a machine too, and we should be able to understand it.

[If we do], I think it would give more insight into what makes people people. If you really undstand neural processing, you would undersand what goes wrong in mental illness and how to treat it. What is a hallucination? What are the voices that schizophrenics hear? What is the drive that makes a serial killer kill? If you know how a machine works, you can fix it.

Who has most inspired and/or influenced your work?

In terms of science and ideas, David Hubel and Torsten Wiesel. In terms of the idea that you have to have a quantitative approach to the nervous system, David A. Robinson.... He showed how important it was to use quantitative methods; that it wasn't good enough to record from [electrodes in] a monkey [brain] and say, look, this cell is responding. You actually had to pull out numbers ... and then it was important to make real models. And I guess Steven Kuffler... He was the professor who organized the neurobiology course at Harvard, and who, more than anybody else personified the enthusiasm of science. Steven told us that being a neurobiologist was the best thing you could possibly be.... There was this kind of ineffable enthusiasm he brought to this area.

Who awarded you your first grant and what was it for?

I worked most of my life for the government so I haven't had grants. The first grant I ever had was a Human Frontiers science program grant that started two years ago. [The Human Frontiers Foundation funds international collaborations.] My first grant at Columbia was a Whitehall Foundation grant. That is about looking at the mechanisms in the frontal cortex during free choice.

What was your best idea or theory?

The coolest idea that I actually predicted had to do with shifting receptor fields. I said that was the way it had to work. The fact that the visual systems interacts at the level of receptor fields with the motor system.... When you're planning a movement, it results in a shift of the receptor fields.

There's actually a profound result in that. You move your eyes, and there's a visual representation. If you move your eyes and had to wait for information to come back to the retina, there would be 100 milliseconds in which all of your motor targeting information would be wrong ... because it's behind the times. What this predicted mechanism does is make a zero latency representation in the cortex: Every time you move your eyes, there is a spatially accurate visual representation of the visual world that's still centered on the center of gaze. You don't have an explicit representation of space. What you've got is a retinotopic representation. But retinotopic representation is where things are now on the retina, not where they were when they arrived to evoke the response you were looking at.... [Without it,] you'd never be able to hit a baseball; you'd never be able to do anything.... My prediction was for things to shift and it worked out that way.

Which scientific idea (yours or others') do you regret the most?

In 1972, Bob Wurtz and I wrote a series of papers on the superior colliculus. The first had to do with visual processing, and the second had to do with attentional modulation. The third had to do with neurons that discharge before eye movements, and the fourth was the effect of lesions on monkeys' behavior.

For a number of reasons, we said that the superior colliculus was not driving eye movements, that it was exclusively involved in visual attention. That was wrong. The superior colliculus clearly drives eye movements to an extent. We were 100% wrong.

What are your scientific plans for the next 5 years?

The first thing I want to do is to extend this series of attentional paradigms [his recent work] to other areas ... to see if we can look and get evidence for some area that's figuring out the winner in this winner-take-all mechanism [that underlies visual attention in monkeys]. The second thing I want to do is to look more carefully at the interaction of eye movements and visual perception in visual circuits. We have some very preliminary data about that, and I want to follow that up. And the third thing has to do with ... when we're looking at this issue of spatial constancy, we discovered an intermediate memory mechanism that lasts for about 20 minutes. And we have very preliminary data, which is coming out in the Journal of Neurophysiology, which we call between-trial memories. Basically, if you teach an animal that a spot of light is there on the screen, and you keep flashing and it always appears on the same place, when it goes away, neurons in the motor system begin to behave as if it were still there. It takes a couple of minutes for this ghost of the response to go away.

It's like [this]: If there's something familiar that you interact with in a motor way - a doorknob, a radio knob in a car - and then the radio knob falls off, you often behave as if the radio knob is still there. That's it. It's this memory of a target for movement that's no longer there. And we found some evidence for it in the frontal eye field, and I want to look more for that.

Finally, there's another thing I want to look for if I have time, space, and money, and that's the cortical representation of eye position.

What advice would you give someone starting out in research; i.e. what are the qualities of a successful researcher?

To think outside the box - that's the first thing. Not to take anything for a given. And also to have a life. I think life being a scientist is wonderful, but it may be that having a life outside science is also good. It's critically important.

If you could work with any scientist (historical or current), who would it be?

I think historical would be William James. Not really working with him, but talking to him, because he wasn't an experimentalist, he was a synthesizer. He had all these insights about how the brain works, entirely from introspection, or introspection and a little bit of reaction-time studies.... He even said that one of the physiological effects of attention is to increase cerebral blood flow. He had no idea about that. He just said, gee, it ought to be so. Without an experiment.