What is program control?

Back in 1994 and again in 2009, I see advanced the notion that functions at the proposed Program level merely detect whether or not a program is running. Here, in 2009, Bruce Abbott was summarizing extended off-list discussions with Bill, and Rick disagreed.

This has always seemed incoherent to me. A program is running, but the function of the level at which it is running is not to run programs but only to perceive whether or not a program is running. This is incoherent because the level which perceives whether or not a program is running must be a level above that level where the program functions live and are running.

OK, so a function at the Program level specifies a reference to initiate a sequence at the level below, and then depending on the returned perception at the end of that sequence it specifies the sequence-initiating reference value for one or another of a specified choice of sequences at the level below, until it receives the perception that was requested at the Principle level above that.

On Rick’s Mind Readings website at Control of Program Perception is a demonstration that is said to distinguish between sequence control and program control. What is meant by program control is perceiving when a given program is not running and acting (pressing the spacebar) to cause it to resume running. The ‘reference program’ is

  • “if the shape is circle, the next color is blue; else, the next color is red”

The two shapes are a circle and a square.

I have described how in my experience it is much easier to control two sequence perceptions concurrently:

  • circle-red-spacebar
  • square-blue-spacebar

The difference in the speed that subjects can manage is explained by the differences in complexity of the task. The ‘sequence’ task involves pairwise comparison of relative size of configurations. The supposed ‘program’ task involves two levels (sensation of red or blue color and configuration of square or circle shape)

Rick has acknowledged that this is a problem, which he understands as a problem of confounding variables, but he has not modified the demo or its description or taken it down.

Note that the ‘sequence’ task can be formulated as a program. In the demo instructions, we read

the reference sequence is :

“small”, “medium”, “large”.

Unstated is that what is to come after ‘large’ is ‘small’ again in a repeating cycle. This can be stated as if/then contingencies in various ways, e.g. quite simplistically:

loop
If small then medium
elif medium then large
elif large then small
/loop

We could claim that the subject is controlling a perception that such a program for the ‘sequence’ task is running with as much justification as the demo description has for its claim that the subject is controlling the relationships of shapes and colors for the ‘program’ task.

There is a more fundamental problem of what is meant by ‘program control’. As sketched in B:CP and elsewhere, the proposed Program level controls by sending reference signals to the Sequence level. Detecting whether or not a program is running is done at the level above the program level. If we assume the levels as they are usually named, this would be a demo of Principle control. The ‘principle’ perception that the subject adopts for the sake of the demo is ‘this program should be running’. At this higher level, the subject might control whether or not the program is running (there are doubts, as noted), but doing so is not ‘program control’, i.e. control at the proposed Program level.

Well, now, hold on (I hear a voice say). In order to know that the program is not running, the subject has to actually run the program in imagination and compare its imagined output with the actual perceptual input. So a compliant subject is controlling at the Program level (in imagination) in order to control a relationship (?) of the imagined program output color to the currently perceived color.

Eh. That’s not a demonstration of program control, it’s an assumption that the subject is controlling at the program level in order to detect a color mismatch.

But from the fact that the computer is running a program, and even given the fact that the subject’s action (spacebar) causes the desired program loop to run, it does not follow that the subject is controlling that program. The subject cannot perceive the program loop within the computer. The assumption is that the subject has replicated that program loop with perceptual control functions. Quod non erat demonstrandum.