[From Bill Powers (960316.1115 MST)]
Hi, Bruce -
I've spent some more time looking at videos, this time spending an hour
watching rat 96001 in runs 5 and 6. It's amazing what you see after
extended viewing that was just a jumble the first time.
The main obvious thing that I overlooked the first time (seen but not
noticed, if you know what I mean) was that there is a pattern in the so-
called "bar-pressing" behavior for this rat. It is not bar-pressing. It
is more like scratching in dirt, or running. The forepaws alternate
rapidly in being placed on the bar and drawn back, with an average of 6
or 8 scratches taking place before the apparatus is triggered. As soon
as the apparatus is triggered, in nearly all cases, the rat goes to the
food dish and eats what is there. Only very occasionally does it
continue scratching and cause a second delivery before it checks the
dish.
Occasionally (perhaps eight or ten times in a run), the rat will scratch
6 or 8 times and go to the food dish even though the apparatus has not
triggered. In those cases, it takes a brief glance into the dish and
immediately goes back to the bar and scratches some more, until the
apparatus is triggered. Later in the second run (006), I notice more
instances where it would do a scratch sequence without triggering the
apparatus, pause, and then do another one until the trigger occurred,
_without checking the dish_.
In run 006 there was a brief break for exploration at about the 9-minute
mark, and more protracted ones later, the most protracted one starting
at about minute 23. The rat nosed around the upper corners and edges of
the cage, and appeared to be licking it. Only after a long active period
of doing this did it pause to groom and go back to the bar for a few
more food deliveries. My impression is that it needed water -- all those
dry pellets! If water deprivation does begin to be felt partway through
the run, there is no point in denying it, because the water deprivation
interferes with the eating just as much as brief periods of drinking
would.
Once in a while (two or three times) the apparatus clicked and the rat
went to the food dish, but immediately returned to the bar and pressed
some more until there was another click. My impression is that the food
delivery failed, perhaps an empty slot in the delivery disk (as I
imagine it), or a stuck pellet.
Early in the first run (005) I got the impression that immediately after
failure of delivery after a sratching sequence, the rat went into a
brief exploration, as if concluding that the scratching no longer
worked. But it quickly found its way back to the dish and bar and tried
again. I don't remember seeing this behavior in the second sequence
(006).
I think there is probably more to be learned from the videos than from
the record of contact closures. We need to develop a simple ethological
notation, so someone can watch at least some of the videos all the way
through noting the patterns and times as they occur. One pattern is
sticking the nose into the dish, sitting back, and chewing. Another is
scratching at the bar until there is a click. Another is scratching at
the bar for some number of strokes, checking the dish briefly (not
sticking the nose in) and going back to scratch the bar until the click
occurs. Another is exploring the region around the dish and bar. Another
is exploring the whole cage. And there are grooming and sleeping.
Perhaps these patterns can be grouped into cycles that can be
represented by a single keystroke at start and finish (or for short
cycles, just somewhere in the cycle). A person who has memorized the
symbol system could run the tape at real-time speed and the computer
could record the times at which each keystroke occurred, so all the
observer would have to do would be to watch the tape and hit the right
key. For error checking a second observer should probably repeat the
run, and discrepancies can be checked against the tape. Maybe a point-
and-click system would be even easier to use.
Of course we couldn't do this for every run and every rat. But I think
this needs to be done as a spot-check, to explain the record of contact-
closures. There is a whole lot going on that is not recorded.
The new video viewpoint should let us see even more.
I'll look again at the other two rats pretty soon.
ยทยทยท
----------------------------------
Bill