lag time for sequence perception

[From Bruce Nevin (991101.1503 EDT)]

from B:CP p. 138:

"When [electrical] impulses were injected in the range of one to eight per
second, an endless variety of repetitive movements of the extremities
occurred, beginning a few seconds after stimulation began and persisting
for a few seconds after the last electrical impulse. [...] the kind of
action depended on where the stimulation was injected; one signal generally
caused one kind of behavior."

Does this suggest that lag time for sequence control is "a few seconds"?

Is there a more accurate measure of the time delay observed in this or
similar experiments?

Would this kind of data -- the delay between electrical stimulation of the
brain and consequent behavioral output -- be useful for verifying the
perceptual control hierarchy? Are such data available for different kinds
of behavioral outputs, hence different levels of the hierarchy?

  Bruce Nevin

[From Bill Powers (991101.2213 MDT)]

Bruce Nevin (991101.1503 EDT)--

Does this suggest that lag time for sequence control is "a few seconds"?

Not necessarily. When very low impulse rates are employed, the
signal-to-noise ratio is necessarily low, and human systems (as well
properly designed engineered systems) slow down under those circumstances.
I would guess that the lag for sequence control would be somewhere between
half a second and one second, where I refer to the minimum time required
for recognizing that a wrong element had occurred and starting to do
something about it. It would probably be possible to do a differential
test, in which the reaction time for starting to oppose a disturbance of a
lower level variable would be measured, and then the _difference_ in
reaction time would be measured when the same variable had to be judged as
the correct or incorrect next element of a sequence.

Is there a more accurate measure of the time delay observed in this or
similar experiments?

Not yet.

Would this kind of data -- the delay between electrical stimulation of the
brain and consequent behavioral output -- be useful for verifying the
perceptual control hierarchy? Are such data available for different kinds
of behavioral outputs, hence different levels of the hierarchy?

There is probably a huge literature on reaction time, mostly pretty old.
And measures of reaction time usually are confounded by mechanical lags
(that's why I went to the electromyogram to find the true reaction time). I
think the experiments we need have probably not been done, particularly not
with modern equipment. Good project for a PhD.

Best,

Bill P.