Shocking Details

[From Bruce Abbott (950608.1105 EST]

Bill Leach 950607.21:23 U.S. Eastern Time Zone --

[From Bruce Abbott (950607.1155 EST)]

Bruce I would suggest that it may be necessary to run the experiment
over again but with a ramped shock onset condition or the already
suggested (by Bill P.) idea to provide a warning signal sufficiently
prior the to the onset of the shock inorder to effectively answer the
challenges to your conclusions.

I admit that this is involving some guesses on my part but it seems that
if a ramp rate could be selected such that the onset of the shock is
perceived prior to the shock becomming "annoying" such that the rat could
"choose" to turn it off or not.

Bill's suggestion was intended to permit the rat to avoid the shocks rather
than to escape more efficiently from them. The experiment suggested by Bill
was in fact done in that study. The rats were trained on a Sidman avoidance
schedule. In the absence of responding on the avoidance lever, they
received a 0.5 second duration footshock every 5 seconds. However, each
lever-press reset the shock timer so that the next shock was always 20
seconds from the last press. By pressing the lever at least once within
this 20-second interval, the rats theoretically could avoid every shock. If
a further response did not occur within the 20 seconds, a shock was
delivered, and further shocks were then delivered at 5-second intervals
until another response occurred. The rats became very efficient controllers
of shock frequency on this schedule, receiving only a few shocks in an
hour's time. However, when given a choice between the avoidable-shock
condition and one in which shocks were unavoidable (but delivered in the
same number and pattern as the avoidable shocks, the rats showed no preference.

Now, a finding of no-preference could indicate either that the rats were
indifferent between the conditions or that they had not learned how to
switch from one condition to the other. To rule out the latter possibility,
I added a warning signal just prior to shock to shocks in _one_ of the two
conditions (avoidable or unavoidable). I knew from previous research that
rats prefer signaled over unsignaled shock; if the rats knew how to switch
conditions they would now be expected to do so if switching put them into
the condition offering the signaled shocks. They did. The rats chose
signaled avoidable shock over unsignaled unavoidable shock, and they chose
signaled unavoidable shock over unsignaled avoidable shock. They controlled
for signaled shock, but not for avoidable shock.

The assumption that a 1% duration shock is linearly equivalent to a 100%
duration shock in terms of disturbance is stretching credibility in my

I didn't make that assumption. What I said was that the rats in the
escapable condition (of the escapable versus inescapable shock experiment)
were able to cut the duration of shock from 15 seconds to around 0.15
seconds by means of their behavior, or to 1% of its uncontrolled value.

From personal experience with shocks (unfortunately for me...
considerable), the most memorable component was that I was shocked at
all. There are a couple that I know must have been of _VERY_ short
duration based upon the physical fact that I am still with us (first was
a solid connection to a relatively "hard" 300 VDC and circuit common and
the second was a very solid connection to two phases of a 450 VAC
distribution bus. My own anecdotal evidence is that as far as perception
is concerned, the intensity of the shock is a far greater concern than
the duration. Indeed, if the intensity is low enough, the shock
experience can actually have pleasent perceptions associated with it.

This is a serious misconception many people have about the nature of the
shocks used in experiments such as mine. Most of us have had the misfortune
of experiencing contact with line voltage (or worse), definitely a memorable
experience. We've heard of electric shock being used as a means of torture
(usually the culprits are depicted as Nazis or, of course, Mad Scientists).
It's an easy step to thinking that what experimental psychologists use when
they deliver shock to their hapless subjects is something akin to torture.
Not so.

The shock you experienced when you came into contect with both sides of a
300 VDC source most likely involved a fairly high current flow (limited only
by the current-sourcing capacity of the supply and your own bodily
resistance). Skin resistance drops as current flow through the surface of
the skin increases, a positive feedback relationship leading to a rapidly
increasing current flow over time. If there's enough current for a long
enough time it can kill you; more typically it can jolt your muscles into
extreme contraction. If you contact a power line with your palm, this can
cause your fingers to close around the wire and you won't be able to let go.
This is nothing like what is used in psychology experiments.

What is used in the laboratory as a source of "aversive stimulation" is
typically a current-regulated source set to around 1.0 milliamps (one
thousandth of an ampere). This level is sufficiently strong that a rat will
acquire a response that prevents or terminates the shock. It is clearly
aversive to the rat but far from traumatizing. Rats maintain their
appetites during the study, are not unusually susceptible to disease, and
behave when handled in the same ways rats with no shock experience behave.

The 1.0 mA level is close to the minimum value that will produce rapid,
reliable escape behavior. At 0.5 mA the escape latencies are longer as the
rats are likely to be spending more of their time exploring the chamber,
grooming, and so on, and thus tend to be some distance from the escape lever
when the shock occurs. At 0.25 mA they may not acquire the escape response
at all.

I have regularly administered 1.0 mA shocks to myself in order to confirm
that the shock grid is working properly. The first time you try this you
tend to overreact (a phenomenon called fear-potentiated startle) but you
quickly discover that it's not all that bad, although you clearly don't like
it and would rather not have the experience.

The idea some people have about the shocks administered in these studies has
a parallel in their idea of what goes on in appetitive (food) studies. The
picture conjured up here is that of concentration-camp victims, starved down
to skin and bone. A picture that is closer to the truth is that of someone
who hasn't eaten since noon yesterday: hungry, certainly willing to work
for some supper, but not in agonizing even-eat-the-shoeleather hunger of
18th century sailors stuck in the doldrums of the Sargasso Sea. The purpose
of both the food deprivation and shock levels is simply to provide the
motivation for the behaviors the investigator wishes to study. To use
higher levels would defeat the purpose of the study. A rat cringing in
terror in the corner of an operant chamber is hardly going to learn to press
a lever that prevents shocks from occurring, and one enduring
semi-starvation will probably be too busy searching for food to pay
attention and learn fine distinctions on a discrimination task.

Occasionally we do hear about experiments that used extreme conditions, but
these are the rare exceptions. Such studies are only undertaken after
considerable soul-searching to decide whether such treatment is ethically
justifiable, and only with the approval of a properly constituted, federally
mandated review board.

The first is that perceived luminous intensity for a flashing light
source (flashing much faster than the high level perceptual system can
handle) is that the perceived intensity is a function of peak intensity
and is almost completely independent of duty cycle. Of course none of
the mentioned parameters in this case are linear. I believe that
perceived intensity is a log function similar to sound intensity.

Perceived shock intensity is roughly a log function of physical intensity,
although there are a host of other factors involved (which have been
investigated; I won't get into them here).

The second is that most testing involving sensory input at a high level
(ie: approaching maximum dynamic input range) very quickly appear to
loose a major amount of their input gain. Thus, the perception of a 15
second steady state shock current might be that of a differentator (with
some lossey DC bypass).

The psychophysics of both shock intensity and duration have been
investigated. The intensity function is steeper than the duration function.
As rats can clearly differentiate shocks of different durations, the input
function is not that of a differentiator. It's a log function like that of