"... there have been many experiments running rats through all kinds
of mazes, and so on -- with little clear result. But in 1937 a man named
Young did a very interesting one.
[ long description of "rats in corridor maze experiment" deleted ]
[ Young basically was very careful with controls on how the rats
found their way about. He controlled for scent, ambient light,
different noise of rat footsteps in different parts of the maze,
and so on ]
I looked into the subsequent history of this research. The next experiment
and the one after that, never referred to Mr Young. They never used any of
his criteria of putting the corridor on sand, or being very careful. They
just went right on running rats in the same old way, and paid no attention
to the great discoveries of Mr Young, and his papers are not referred to,
because he didn't discover anything about the rats. In fact he discovered
*all* the things you have to do, to discover something about rats. But
not paying attention to experiments like that is a characteristic of cargo
cult science."
[ob. cit. page 315 et seq.]
I can think of several possibile explanations for this situation.
1. Feynman was wrong
2. Rat experiments are no longer performed or even of interest
3. Rat experiments don't need to have these kinds of controls, because
they are designed to measure attributes not affected by these things.
4. Rat psychologists have crummy experimental procedures and hence
worthless results.
Can anyone offer some insights into which of these might be the case?
Perhaps someone who has worked or is currently active in the field?
--
Peter van der Linden lin...@Eng.sun.com 415 336-6206
Elvis has now left the building.
> [ Young basically was very careful with controls on how the rats,
> found their way about. He controlled for scent, ambient light,
> different noise of rat footsteps in different parts of the maze, and
> so on ]
[Feynman complains that there was no follow-up of this analysis of
defects in the older maze-running studies.]
>I can think of several possibile explanations for this situation.
>
> 1. Feynman was wrong
> 2. Rat experiments are no longer performed or even of interest
> 3. Rat experiments don't need to have these kinds of controls, because
> they are designed to measure attributes not affected by these things.
> 4. Rat psychologists have crummy experimental procedures and hence
> worthless results.
>
>Can anyone offer some insights into which of these might be the case?
>Perhaps someone who has worked or is currently active in the field?
What happened around 1937 was that
5. B. F. Skinner developed ways to control all those external
variables by enclosing the experiment in a sealed, soundproof,
lightproof, etc., box. The results were reliably reproducible, and a
great deal was learned. The boxes were soon names "Skinner Boxes" and
became the new paradigm for studying animal learning. Skinner and
many others switched to pigeons, for various reasons, but others
continued to use rats.
When I was undergraduate in the late '40s, I hung around that lab and
helped with some switching and sequencing stuff to make the
experiments more convenient. I don't remember the name of Young, but
it was folklore that the change *was* because someone had found that
rats appeared to be able to navigate by distant cues, e.g., the
appearance of the ceiling, so that the traditional open-topped maze
experiments might be flawed.
Marvin Minsky
>In that most excellent book "Surely You're Joking, Mr Feynman",
>Nobel-laureate Richard Feynman says the following astonishing thing:
> "... there have been many experiments running rats through all kinds
> of mazes, and so on -- with little clear result. But in 1937 a man named
> Young..." [discovered important controls for running rats in open top
> mazes that subsequent researchers ignored.]
Middle of post omitted....
>I can think of several possibile explanations for this situation.
> 1. Feynman was wrong
> 2. Rat experiments are no longer performed or even of interest
> 3. Rat experiments don't need to have these kinds of controls, because
> they are designed to measure attributes not affected by these things.
> 4. Rat psychologists have crummy experimental procedures and hence
> worthless results.
>Can anyone offer some insights into which of these might be the case?
>Perhaps someone who has worked or is currently active in the field?
Yes, they are all correct if rewritten:
1. Feynman was partly wrong
2. Some kinds of rat experiments are no longer performed or even
of interest.
3. Most current rat experiments don't need to have these kinds of
controls, because they are designed to measure variables not
affected by these things. [They need different kinds of controls.]
4. Some rat psychologists have crummy experimental procedures and
worthless results, but manage to get their data passed by
lazy journal reviewers anyway.
--Doug Fitts
Department of Psychology
Univeristy of Washington
Seattle, Washington
dfi...@u.washington.edu
THAT explains something for me.
When I was back in the UK, I knew of a folk group called Skinners Rats -
I always wondered how they came up with their names, as none of them were
called Skinner. I assume at least one of them had some dealings with habit
pulling.
Meanwhile, back to your scheduled programming...
>Marvin Minsky
Jon `Wot No Curry?' Brawn
: In that most excellent book "Surely You're Joking, Mr Feynman",
: Nobel-laureate Richard Feynman says the following astonishing thing:
: I looked into the subsequent history of this research. The next experiment
: and the one after that, never referred to Mr Young. They never used any of
: his criteria of putting the corridor on sand, or being very careful. They
: just went right on running rats in the same old way, and paid no attention
: to the great discoveries of Mr Young ... [and so on]
I think there are two points in the acceptance of that experiment of Young:
1) Though Feynman did not concern much, the purpose of Young's experiment
was to train (or to know if he can train) rats to choose a door on the location
RELATIVELY third to a door through which the rat entered the maze. That is,
it was a problem of RELATIVE positioning of 'goal'. (Psychologists would call
this kind of situation as CONDITIONAL discrimination..) Before he put the corridor
on sand, the rats went to a door on an ABSOLUTE location. Rats did not understand
that they were required to choose in that relative way, and/or they only followed
the RECENT success.
Most of people engaged in rats' maze running at that time might not be much
interested in the RELATIVE discrimination and still remained in the ABSOLUTE
discrimination world, as Young's rats were.
In article <1993Apr11.0...@news.media.mit.edu>
min...@media.mit.edu (Marvin Minsky) writes:
: 5. B. F. Skinner developed ways to control all those external
:variables by enclosing the experiment in a sealed, soundproof,
:lightproof, etc., box. The results were reliably reproducible, and a
:great deal was learned. The boxes were soon names "Skinner Boxes" and
:became the new paradigm for studying animal learning.
And I think such a RELATIVE discrimination problem may be still a
contemporary concern of "Skinner Box" researchers, though they may not
use a runway maze. As in the addressing methods in computer programming,
ABSOLUTE addressing may be basic but primitive, and RELATIVE addressing
may have a much more important and efficient role in learning studies both
in rats and machines.
2) The second point is what Feynman seemed to like to stress: the search
of stimulus which controls rats' ABSOLUTE positioning. I don't know the
reason why other researchers at that time were not interested in Young's
control effort for this point, because I feel psychologists would be much
interested in such a search (though I have not heard of any psychologist
who put a "Skinner Box" on sand..) and it is really a problem of scientific
dignity.
BY THE WAY, I am much interested in Dr. Minsky's evaluation on Skinner
and his followers' works, since he wrote in the same article:
:When I was undergraduate in the late '40s, I hung around that lab and
:helped with some switching and sequencing stuff to make the
:experiments more convenient.
I feel I have read something related in his book, but forgot what he exactly
wrote about Skinner's approach in the book. Not only because I am interested
in computer simulations of rats's behavior in "Skinner Box", but also because
Skinner passed in 1990 and there can be interesting interactions between
learning 'theories' of AI Minsky and pigeon Skinner, Dr. Minsky's view on
Skinner's approach would be of interest to many. (I hope.)
Arata Kubota
E-mail aku...@amphora.ignis.org 7666...@compuserve.com
~~~~~ *+* dedit ventos ad ulteriora noscenda *+* ~~~~~
I am interested in listening to them. Any information on their titles
will be appreciated. "How many years have to pass before they learn"?
Gee. Then you'd think they'd care about HOW the rat found the right door,
wouldn't you? Was the rat actually memorizing the turns in the maze, or was it
using external cues? And that, of course, requires removing the sort of queues
that Young was dealing with. Arata's paragraph above is yet another example of
pseudocience in action (give an invalid rationalization for why the
experimenters didn't bother with useful controls, rather than improving the
controls).
--------------------------------------------------------------------------------
Carl J Lydick | INTERnet: CA...@SOL1.GPS.CALTECH.EDU | NSI/HEPnet: SOL1::CARL
Disclaimer: Hey, I understand VAXen and VMS. That's what I get paid for. My
understanding of astronomy is purely at the amateur level (or below). So
unless what I'm saying is directly related to VAX/VMS, don't hold me or my
organization responsible for it. If it IS related to VAX/VMS, you can try to
hold me responsible for it, but my organization had nothing to do with it.
> ...several possible explanations...
> 1. Feynman was wrong
> 2. Rat experiments are no longer performed or even of interest
> 3. Rat experiments don't need to have these kinds of controls, because
> they are designed to measure attributes not affected by these things.
> 4. Rat psychologists have crummy experimental procedures and hence
> worthless results.
All good guesses. The answer is partly 4, partly modified 3. Maze
running turns out to have several drawbacks, one of which is situation
confounds, but another is just it's hard to automate. There are other
ways to measure behavior that are usually better: cheaper, cleaner,
more powerful, less species specific.
But mostly 5: Feynman's criticism is apt, his example accurate, and
his conclusion correct. Much of what gets called science isn't. The
results won't reproduce. And the social sciences have more than their
share. However bickering over whose field has the most is pointless.
It's just as much a waste of time and money whether it's cold fusion
or planaria learning via injected DNA.
-- Jon
--
Jon Krueger j...@ingres.com
Janet "I'm going home now" Christian
--
Janet Christian jchri...@indetech.com
My boss agrees with everything I say - well, actually, he thinks I'm working.
"Never argue with a Scorpio - it's frustrating and you'll lose, anyway..."