Message Number: YG860 | New FHL Archives Search
From: RRC
Date: 2001-03-07 11:34:00 UTC
Subject: Re: Another Behavior Question (Bullying)

It happens every time; just when something of great personal interest
comes up, I have to take a break for a while (In this case, I'll be
going to Massachusetts, then a week of research on the east coast). I
can hardly give this subject (or Brett) the time and effort it deserves.
I will make a few comments, but will have to hold off on the really deep
stuff until after I return (assuming someone doesn't kill me for the
length of this post).

First, I have to admit I feel more than just a little discomfort
performing a serious critique on work which I think is one of the better
books on ferrets. Fara Shimbo did her homework, obviously worked
extremely hard in obtaining, reading and understanding her references,
and I think she did a pretty good job of communicating that knowledge to
the reader. From her style of argument and her frequent admissions of
poor data, I have the feeling she made a dedicated attempt to be as
factual and accurate as possible. Certainly her book, nearly a decade
old, is as accurate or more accurate than most non-vet ferret books
currently in print. While I have factual and philosophical disagreements
with some of the content and conclusions presented (some of which may be
due to the books publication date and the subsequent increase in
knowledge), the book is still in my top 5 list of non-vet ferret tomes.
I find it unfortunate that a generalized discussion of behavior has
become a pointed discussion of her opinion, because I respect her work
and the efforts she has made bettering ferret lives. My fear lies in the
misinterpretation of "scientific critique" as a value judgment of the
person or the work as a whole (Not you, Brett, nor most on this list,
but the people on this list are not the ones I worry about). That is
decidedly not the case. So, with that caveat in mind, here is my limited reply.

When I made the comment that Shimbo's observations were filtered through
her personal world view, implying she saw what she wanted to see, I was
speaking from her own work, as she wrote it. I don't know her, have
never met her, I have high opinions of her, but I was speaking
specifically on her comments on ferret social structure. The
descriptions of the various groupings, as well as the experiments,
EXCLUDE groups or individuals where dominance behavior might have been
shown. I don't think this was done with malice, nor on purpose.
Behavioral studies are hard to do, and those which involve extensive
human contact, control and interaction are among the hardest to
interpret. I think Shimbo clearly shows this design flaw when she states
(on p. 46):

"My ferrets are allowed to run loose in the house in three
shifts—because of their social faults."

Shimbo then describes the various group interactions, which are a broad
definition of the "social faults" causing group separation. These
"social faults" can be summed up in various aggressive behaviors,
subsumed within fighting. What Shimbo has done is to create social
groups where internal conflict is at a minimum. This lack of conflict
could exist because a group is essentially equal in social stature,
because whatever dominance hierarchies that exist are well entrenched
and difficult to visualize, or for a number of other reasons. Shimbo
documents the "social faults" in detail, carefully explaining which
ferret is liked (loved), disliked (hated) and the various interactions
between them. Again, invariably, those negative interactions are
typically one of fighting. Indeed, at one point, Shimbo mixes all
groups, which resulted in a day of 'social faulting", or rather,
aggressive interactions (fighting). This type of intraspecies aggression
is exactly what you would expect if a group of animals were attempting
to create a hierarchical social dominance structure. In essence, with
each opportunity to observe the types of behaviors which result in
social stratification, physical isolation is used to prevent such
interactions. You simply cannot observe the creation of dominance
hierarchies when they are not allow to occur.

What Shimbo has unintentionally done is redefine aggressive behaviors
that potentially define dominance hierarchies as "social faults,"
removing them from the equation. Later, when social structure is being
analyzed, evidence of dominance hierarchies is lacking, so she concludes
it wasn't there. But was it not there to begin with, or was it simply
excluded from observation? In this case, it is very difficult to
discover dominance hierarchies when such effort has been done to prevent
their expression. Without expression, there can be little observation,
and without observation, there can be little data. Shimbo made the
correct observations, but the interpretation is flawed because the
observational data is biased. The cold fusion effect.

Additionally, there were at least five fatal flaws in Shimbo's
experimental design. This in not something I would normally discuss in
public forum. I mention it only because the point was made that an
observational experiment yielded specific results contradicting a
position I support. Since I will not attack the observational skills of
the person reporting the findings, I am left with the task of showing
that even with accurate observations, because the experimental design
was flawed, the conclusions are in doubt. This is not a reflection of
the intelligence or skill of Shimbo; outside of coming up with really
original ideas, the hardest thing you can do is create a flawless
experimental design. Shimbo has done no worse than anyone else in that
regard—and much better than many highly respected scientists. Besides, I
doubt if Shimbo actually attempted to create a research design. I think
she was only interested in looking to see if she could spot dominance
behaviors, and wasn't worried much about the details.

The first fatal flaw: Groups were separated because of social faults,
and intergroup aggression was noted when they were mixed. Shimbo
interpreted this as a type of exclusionary aggression, which it probably
was. However, social hierarchies can require considerable time to
develop, ranging from hours to weeks or longer, depending on the
individuals. The external expression of exclusionary fighting is similar
to that of dominance fighting; only the ferrets seem to know for sure.
We know because (in nature) one individual leaves, or they form a stable
group. Because the groups were never allowed to form territories or
establish hierarchical structures, we cannot define which of the two
were taking place. Also, it is possible that exclusionary fighting could
evolve, with time, into dominance fighting. In other words, we can call
the behavior exclusionary, but because the experiment was stopped before
other possibilities could be observed, the point is left unproved. This
flaw is inherent throughout all observations, where conflicting groups
were segregated because of "social faults'.

The second fatal flaw: The experiment was carried out on solitary
carnivores with the assumption that ferret dominance behaviors would
parallel those shown by social carnivores. Shimbo was looking for the
types of interactions shown in social carnivores and did not find them.
There is no evidence presented, theoretical nor observational, that
suggests there are parallels between the two. It may well be that the
lack of deference during feeding by ferrets is reflective of their
progenitor's solitary lifeway; they never needed rules regarding
feeding, so they never evolved them. The flaw is that the observations
were not made using established polecat behaviors. For example, if two
ferrets care less who is eating or drinking at the same time, does that
reflect a lack of social structure, or is it due to their domestication
from a species which were same sex exclusionists and feeding rituals
were never needed? Most descriptions of dominance behaviors use the
social carnivore paradigm, but ferrets are nonsocial carnivores and
evolved a different set of rules.

The third fatal flaw: The experiment was carried out with groups of
animals that had already established some sort of rules for getting
along. This is sort of like observing couples that have been married for
20 years and using the data to explain behaviors of newlyweds. The
behaviors may be identical, but muted in the older individuals, making
observation difficult. There is a stability to hierarchical
relationships; once created, there may be little observable difference.
That doesn't mean hierarchical social grouping does not exist; all it
means is it is so well established that the subtle signs are not easily observable.

The fourth fatal flaw: The experiment failed to account for the human
presence. This was not a set of experiments where the experimenter was
an outside observer. Shimbo was an internal factor, separating groups,
preventing some behaviors, rewarding others. This was not a group left
to sort out its own behaviors, but one that sorted out limited
behaviors, some which were suggested or rewarded by Shimbo. That is not
an accurate reflection of ferret behavior. The potential for
unrecognized conditioning is so high any result is suspect.

The fifth fatal flaw: The experiment was carried out on an heterogeneous
group of animals, ranging from whole to neutered, young to old, some
abused, some abandoned, and many possessing remnant neurotic behaviors
from prior ownership. How do we know such a group will display typical
behavior of ANY type? Some ferrets, as described in Shimbo's book,
appear to be so fearfully neurotic that it would be unusual to expect
ANY normal behavior. Depending on the age of neutering, age dependent
behaviors may have been learned, or not. The results may be different in
groups composed of sexually intact individuals compared to early or late
neuters. Ferrets are intelligent, altricial mammals, so a lot of correct
behavior is learned from adults and siblings. The degree of expression
of some behaviors are influenced by hormone levels. Ferrets are easily
conditioned, so some behaviors may reflect human desire rather than what
is normal for the ferret. You cannot assume such factors have no bearing
on dominance behavior, unless you can show the effects of environment
are not a factor.

I guess I've made my point and further critique is unnecessary. I
apologize to Fara Shimbo, or anyone else offended by these critical
remarks; they are not meant as a personal attack, but rather as a means
to demonstrating the importance and difficulty of observing and
interpreting behavior. I have always felt it particularly mean-spirited
for someone to "walk behind" an author who has done a pretty good job,
nitpicking tiny flaws to pieces. As a result, I have rarely commented on
any ferret book (save Dustman's book with the doctored photo of the
ferret stalking a chicken, printed by Fancy Publications, owner of
Ferrets and Ferrets USA), and I feel bad for these comments which have
broken that policy.

Maybe Shimbo is correct and I am not. As I've explained before, I think
ferret dominance structures are wacky and very hard to interpret. Some
types of dominance structures are very easy to spot; only alpha wolves
breed. Others are harder, like in some matrilineal monkeys. Polecats
evolved as same-sex exclusionists, and as adults they display few
dominance behaviors compared to some other carnivores. But ferrets are
domesticated polecats who have considerable retention of juvenile
behaviors (behavioral neotony), which DO include a considerable number
of dominance behaviors.

As for some of the traits which are not exclusionary but are classic
dominance behaviors, who hasn't seen ferrets in a stare-off? The
dominant ferret always wins. That is a typical dominance behavior.
Bumping or pushing is another one (not during play, but during dispute).
I have watched my dominant ferrets (Carbone and Tui seem to be
codominant, although I think Tui has a slight edge) eat first,
investigate new objects first, and gang up on misbehaving ferrets. Many
time, when they go to bed, other ferrets will sleep around them. There
are lots of other things I could detail, but this post is already way
too long. I think ferrets form hierarchical dominance structures, not as
well defined as in social carnivores, but still with clear social
ranking. I think it is an example of behavioral neotony, and helps
explain why ferrets act gregarious even though they evolved as the
original social outcasts.

I profusely apologize for the length of this rebuttal.

Bob C




From: "Brett Middleton" <brettm@a...>

RRC <rrc961@m...> wrote:
> Regarding Fara Shimbo, well, I hate to disagree but I must.
[snip]
> Third, dominance behaviors are not best defined by fighting as much
> as they are by submission. When you put down a bowl of food, who eats
> first? Who waits for the food, even waiting until the others leave?
> Those are the types of behaviors which better signal dominance
> hierarchies than simple fighting.

I just reviewed the pertinent chapter of her behavior book, and this
is, in fact, her point. Though she admitted to not having enough data
to draw solid conclusions, she couldn't *find* any patterns of
submissive or deferential behavior, even when she did some experiments
such as withdrawing food for 24 hours and then providing food in a
container that could only be accessed by one ferret at a time. She
also tried withholding water and then providing a single water bottle.
Thus, her take was that fuzzy A does not attack fuzzy B because A wants
B to be more deferential, but because A wants B to *not exist*.

Mind you, I'm not swallowing this theory whole, but she *does* seem to
have some of her ducks lined up, unlike many ferret owners who casually
talk about dominance relationships among their fuzzies without really
knowing anything about the subject. Of course, I have to give a great
deal of weight to opposing views expressed by posters such as yourself
and Sukie, especially since I have no observations of my own to
contribute. (Someday maybe I'll be out of rental housing. If you hear
a mysterious booming noise, that'll be the fuzzy population exploding
in my vicinity.)

As for the "cold fusion" effect: I don't know Ms. Shimbo, nor do I
know anything about her background and qualifications other than the
hints she's dropped in her books, so I have no basis for judging in her
case. However, I'm certainly familiar with the phenomenon, on both a
personal and professional level. B-)

Brett

*SLMW 1.0* My wisdom is legendary. Um, or is that "mythical"? ...