David Marr's book Vision

Date:
1979-1980
Reference:
PP/CRI/H/6/18
Catalogue details

Licence: In copyright

Credit: David Marr's book <i>Vision</i>. Source: Wellcome Collection.

    7/20
    THE SALK INSTITUTE Dr. David Marr page 2 that you are happening to overhear them. For this reason, if for no other, there is no need for you to pay especial attention to them. In attacking any scientific problem no progress can be made unless certain difficulties are firmly put on one side. One cannot always justify the choice of what to leave aside and it is a bore to be continually asked to do so. However from time to time it is wise to stand back from the subject and recognize explicitly what has been ignored. A person with a different viewpoint will naturally tend to do this without prompting but one is under no obligation to counter his objections and queries at every turn, or one would never make any progress at all. We would never have solved the genetic code if we had continually worried about exactly how protein fold themselves up. However I think 1 should record one general impression. I have the continual feeling that one is missing an important idea which is really just round the corner. This idea is to do with the addressing problem. It should be possible, I suspect, to make a rather strong statement about what kind of addressing must be there. (Just as, on quite general grounds, one can show that for natural selection to work, inheritance must be particulate and not blending.) I think if one could make such a statement it would have important consequences for studying the problem experimentally. It would also, hopefully, justify your present general approach to these matters, since I find it remarkable that you can make a plausible case for so many different tokens. This surely implies some special way of relating them. I do hope you can fit in a visit here some time during the Spring. Let me know fairly soon as engagements are apt to accumulate as time goes on. I promise not to nag about all the problems in the attached comments since after a while nothing useful can be said. I'm sure we can find more rewarding topics to discuss. Incidentally I've been writing to Tommy about his generalization of our Theorom 2 which I believe he has not got quite right (the question is: given a band of width W, upper limit w u , lower limit <*> , where W = (w u - u e ) is not an integral multiple of w e , can you obtain tne complete signal from equally spaced samples, spaced II/W apart? I claim not.) As far as the visual system is concerned this looks rather irrelevant but we'd better get it right if we are to publish it. 1 was glad to see that Donald Michie had nice things to say about your approach in a recent review in Nature. I feel sure that it's going to have a big impact, especially when the book comes out. Have they given you tenure yet? It's really a scandal that they hadn't done it before. Happy Christmas! i/Wvi I FHCC/bml F. H. C. Crick
    digitised image 1
    digitised image 2
    digitised image 3