[Rick Marken 2019-05-27_12:56:59]
Well, CSGNet has gotten a nice peek into my personal life over the last couple days, with my posts of pictures of Bill and myself in honor of the anniversary of Bill’s death and my accidental posting of an email to a personal friend that had nothing to do with Control Theory. So I thought I would post a little more personal stuff that is related to Control Theory and that I happened to discover this weekend while doing some spring cleaning.Â
What I found were the first correspondences between Bill and myself. They had been stored in plastic boxes and had suffered some water damage because the boxes were stored in an outdoor shed that had developed a roof leak. So many of the papers were destroyed but, luckily those first correspondences survived. Although I can’t find the very first letter I sent to Bill. So the following set of PDFs start with Bill’s reply to that first letter I sent to this mysterious genius who I had never heard of. The pointers to the PDFs are here:
https://www.dropbox.com/s/ykundvkz2brmzsj/1.Powers%20Feb%201979.pdf?dl=0
https://www.dropbox.com/s/adtgtysogbtdc54/2.Powers%20Nov%201979.pdf?dl=0
https://www.dropbox.com/s/6ijom1eehz46hxd/3.Powers%20April%201980.pdf?dl=0
https://www.dropbox.com/s/43jo6cyzx7zcdfz/4.Powers%20Nov%201980.pdf?dl=0
https://www.dropbox.com/s/ru8c152ycl5vb7m/5.Powers%20April%201982.pdf?dl=0
All but two of Bill’s replies are handwritten, but I believe they are quite legible. The most interesting one to me is his reply to my April letter (in the file “3.Powers April 1980.pdf”. In that letter Bill adds a derivation (as “A simple case”) that explains why I got the results I got in my first published paper on what was then called CS psychology. What I showed is that the cursor in a tracking task can’t be considered the cause of the outputs that keep the cursor on target. I think I never really gave a good explanation of why that is the case and at the time I clearly did not understand (or carefully read) Bill’s explanation at the end of the derivation. But at the end of that derivation is this:
That last line explains it all: “One of the “stimuli” is inside.” What a great line!! The “stimulus” inside is, of course, the reference signal! The cursor perception is just one cause of output; the other is the reference signal . If it were possible to hold the reference exactly constant there would indeed be a correlation between cursor and output (handle movement in this case), which is what I had found in my modeling. So I had to add “noise” to have the model perform like the subjects in my experiments. But I felt like this was cheating, in a way. But now, 40 years later, I finally understand what Bill was talking about; the source of noise is slight variations in the reference signal that exist even when you try to maintain a fixed reference signal.Â
I wish I had been able to understand this back then. But for anyone interested in doing PCT research, it seems to me that this derivation of Bill’s can be the basis for some interesting research on that other “stimulus” contribution to behavior that has been consistently ignored in psychology: the reference signal!Â
Oh, and I think these might be a nice addition to Bill’s archives. What do you think, Allie?
Best
Rick
···
Richard S. MarkenÂ
"Perfection is achieved not when you have nothing more to add, but when you
have nothing left to take away.�
                --Antoine de Saint-Exupery