"TEMPERAMENT"
AND "INSTINCT" TESTING,
PANACEA OR NONSENSE?
Jim Engel
April, 1988
Canine temperament tests have been the subject of extensive discussion,
debate and promotion. Many have been devised and used, with a great diversity
in popularity and validity. Much confusion has existed about the purpose of
such tests, what the results say about the character of the dogs tested, and
what place they can or should play in working dog breeding programs.
More recently, tests which purport to certify the "instinct" of
a dog for a specific working function have become popular. The principle example
of this is the "Herding Instinct Test" which leads to the
"Herding Instinct Certificate." The chief reason for their popularity
seems to be the desire to certify a dog as a "true working dog"
without the time and effort of serious training.
The originators of such tests have largely been people concerned with the
generally perceived decline of the working attributes of various breeds,
seeking a palatable remedy. Successful trainers have by and large remained
aloof, mostly passing off such tests a simple minded nonsense, unworthy of
notice by serious people. Breeders with an established working line have
generally taken a posture of haughty disdain. Some kennels without any
particular working image have shown guarded enthusiasm.
In order to provide focus, this discussion will be limited to those
breeds, such as the Rottweiler, the Belgian herding dogs and the Bouvier, whose
modern working functions are primarily searching, tracking, protection and
herding, that is, the protective heritage working breeds.
When these breeds were founded, the common belief was that in order to
identify correct character for purposes of breeding selection one must train a
dog for his intended purpose and then evaluate how well he functions in his
work. Two generations of KNPV, Ring and Schutzhund trainers lived and bred
according to this credo. Now it would seem that some, not necessarily the
creators and promoters, believe that this need for extensive training and then
testing has been to some extent negated by the development of these tests
specifically for the untrained dog.
Thus the fundamental question becomes whether it is possible to observe
and test a young dog so as to accurately predict if proper training will
produce working excellence, and whether those attributes commonly perceived as
contributing to "temperament" are useful and sufficient predictors of
ultimate working potential.
For the purpose of discussion, let us define "working
character" as those elements of genetically predetermined emotional and
mental propensities in a dog which, when proper training is applied, result in
a good or excellent working functionality. Let us further agree that producing
a line of dogs consistently strong in working character (and, of course,
correct structure and type) is the only relevant issue.
My thesis is that working character as defined here is in fundamental
ways not measurable by the temperament tests as applied in North America and
that, depending on their specific nature, similar comments apply to most of the
European tests. The fact of the matter is that such tests tend to be calibrated
to the typical "show dog" as the norm, and that true excellence in
character can be a serious disadvantage.
So, just what is temperament? Something that dogs do well in a temperament
test have lots of? And if so, what does this have to do with inherent
"working potential?" Are temperament and working potential the same
thing? Are they even very closely related? My purpose here is not so much to
convince you that I know the answers to these questions, but rather to provoke
some really serious thought and discussion.
A dog who is active, friendly, boisterous, outgoing and lacking in
suspicion will be perceived by most of us as having a nice, pleasant
temperament, and will likely do very well in most temperament tests,
particularly the ATTS type test. The problem is that such dogs, while perhaps
personable and pleasant to be around, can nevertheless be seriously deficient
as working dogs. Such a dog can be stupid, grossly self centered or lacking in
sufficient courage or aggressiveness. Tests which certify such dogs as of
"good temperament", which is perceived as in some way equivalent to
"correct working character," pose a grave danger when taken
seriously.
What I find objectionable is not the testing, but the formalization, the
giving out of certificates on the basis of an opinion based on a very brief and
stylized test. Such an opinion, when provided as a private commentary by a
truly experienced trainer, can be beneficial and useful. But when it is
portrayed as a proof of working character, as a breeding criteria, then serious
abuse is virtually inevitable.
Certainly experienced trainers can, in an hour or so, get a pretty good
idea of the potential of an adult dog. They are able to predict which dogs are
likely to succeed and which ones are not, and, perhaps most importantly, which
ones could be made to succeed but are simply not worth the trouble. In making
such an evaluation, a trainer will devise his tests as he goes along, according
to what he sees, and try it again the next day in a marginal case.
I believe as surely as the sun will rise in the east tomorrow morning men
such as Coen Semler, the famous Dutch police trainer, could keep a Bouvier for
a week and tell you whether he was police dog material, and that you could take
it to the bank. The problem is that no one conducting formal temperament tests
today, to my knowledge, is remotely comparable to Mr. Semler in terms of
knowledge and experience.
The subject of temperament testing has become an awkward one for me,
since I was among the earliest supporters of the ATTS (American Temperament
Test Society) tests. So many years ago the idea of a straightforward series of
tests, administered in a standardized way by relatively inexperienced people,
which would provide an assessment of temperament and thus a guide to breeding
better dogs, was most attractive. We thought that it couldn't do any harm, and
could do a great deal of good by bringing the concept of character criteria to the
attention of the American public. I guess we were looking for a short cut, just
wanted it to be true.
Much of this promise has been realized, but deep and fundamental doubts
about the validity and wisdom of the process have grown persistently. Over the
years I have become increasingly uneasy, concerned that such tests are
perceived as having much more significance than experienced trainers know to be
possible. It has become apparent that those conducting the tests are all too
often lacking in knowledge and experience, simply don't understand and
appreciate working dog character in sufficient depth to render a meaningful
opinion.
The ATTS was founded by Alfons Ertelt twelve or fourteen years ago.
Although I did not know Mr. Ertelt well, I had many occasions to converse with
him before his untimely death and came to have great respect for his sincerity,
persistence and knowledge. Thus it is not without careful, even excruciating,
reflection that I propose that the course he charted must be altered.
Although it is perhaps presumptuous to speak for another, I have the
strong impression that he regarded the ATTS test as a starting point, as a
means of bring the matter of character before the public. Alfons was among the
founders of the Schutzhund movement in this country, and was enthusiastic about
the evolution of specialized tests for the various types of dogs. I believe
that he shared the belief that the ultimate goal was something equivalent to
the Ring or Schutzhund as the breeding qualification test for the protective
heritage breeds, that the ATTS test was a means, an evolutionary step, a tool,
not an ultimate objective.
The ATTS style test was never seriously expected to serve as an adequate
breeding eligibility test for working dogs, and its portrayal as such is a
serious perversion, simply does not stand up to scrutiny. Sure, a lot of the
dogs who fail are truly inadequate, but in looking back over the years it has
become clear that many truly good dogs would not be perceived as doing well
because of the nature of the tests, and because so many of the testers are
simply not at a level to recognize and appreciate a really good dog. Beyond
screening out dogs of almost obvious inadequacy, passing such a test in most
instances means very little. Today's reality is that the dog really has to do
virtually nothing but show a little interest and not panic. I am convinced that
a carefully administered sedative would allow virtually any dog to pass.
An excellent example is the agitation test at the end, where a person in
unusual garb first acts in a more or less drunken manner and then moves toward
and threatens the dog. I have made a habit of asking the owner about the
experience of the dog, and found that virtually every dog that looks impressive
has been previously agitated, that is, taught that an aggressive response is
acceptable and desirable.
Bouviers in particular are apt to just stand there, to not take the
situation very seriously. It is not the least bit unusual for dogs who turn out
to be quite good to take four or five sessions before they understand what is
expected of them and react aggressively. The fact that this slowness to react
is much more pronounced in dogs raised in the typical companion home than in
those who remain with experienced trainers. This shows how much of the reaction
of the dog is shaped by conditioning (environment), and how much the
fundamental, intrinsic character of a dog is covered up by his up bringing.
(This of course should give those who think that their dog will instinctively
protect them because they love them something to think about.)
So, in light of all of this, why did such tests come into existence? The
reasons are many and varied, but the general pattern seems to have been the
desire to do something in the face of a widely perceived decline in character
among our working breeds. We were, and are, in very serious trouble, and we had
to start somewhere. Much as we may have wanted to, we could not simply take gun
in hand, pirate the AKC headquarters and refuse to register pups whose parents
lacked a Schutzhund or Ring title. The problem was, and is, one of creating a
culture, an understanding of the importance of working character and aptitude,
a desire on the part of the public for the pleasure of living with a truly good
working dog.
Although I have not been to a lot of tests in recent years, the over all
quality of the dogs seems to be somewhat improved, which means that people have
come to understand that if old Fido is apt to look bad he is best left at home.
I believe that a large number of people have been made aware of temperament,
and have become able to recognize a reasonably wide spectrum of undesirable
characteristics. This is surely progress, but are these dogs not being tested
also not being bred?
Sure the temperament test was a simplistic approach, but the real purpose
was education, of creating the culture. To a large extent the things such
testing can do have been done, and the question now has become one of either
adapting the tests to the current situation, or eliminating them.
Subsequent to the death of Alfons there has apparently been a great deal
of amateur tinkering with the rules and procedures. It seems that any dog who
can be induced to touch the umbrella within two minutes, regardless of the handler's
actions short of throwing the dog in the right direction, is to be passed. In
the protection test only those dogs which panic, make a desperate attempt to
flee, fail.
Actually, the demands on the TT judge are much greater than those on the
Schutzhund or Ring judge. The sport rules are fairly well spelled out, and
either the dog does it or he does not. The judge merely has to determine if the
exercise was performed. On the courage test the dog has to bite, has to
release, and has to stay near the helper, not return to the handler. If he does
these things, he should pass. Sure, he can lose lots of points on procedure,
but pass or fail is fairly well defined.
In the working trial a dog is either able to stand up to the stick or is
driven off, either follows the track or does not, is impartial to gun shots or
is not. The working dog tradition is simple and direct, if a bitch or a dog
fails to stand up under the stick, fails to persist in spite of vigorous
opposition, then any pups they produce are not working dogs, are not Bouviers
or German Shepherds. The judge simply observes and makes an official record of
the test according to straightforward, concrete criteria.
But the temperament test is another thing entirely. It's a matter of
observing the dog and then deciding if his actions indicate good temperament, a
much more subtle and abstract business. What should a dog do when a stranger
opens an umbrella in his face? I think the dog who jumps back six or eight
feet, goes into a crouch and waits for the bastard to make a sufficiently
aggressive move to deserve being bitten shows great character and training. Why
the hell should the dog go up and touch the umbrella? Would you want your dog
to go up and touch the first six foot rattle snake he ever sees? Courage is
doing what you have to do, jumping into situations you don't understand is
stupidity.
Sure, you can tell something by what a dog does when a neutral stranger
approaches. If he makes an unprovoked attack or crouches and hides something is
wrong, but we don't really need a temperament test to know that. And I just
can't twist around disinterest in rattled stones in a bucket to a defect in
character, can't think of anything a dog can do in such a situation to warrant
"failure" in a temperament test, except perhaps run and hide.
In spite of the immense, virtually impossible, demands made of the judge,
relatively meager criteria for the post exist. It seems to me that one who
would evaluate working dog temperament should train at least two dogs to an
advanced Schutzhund or Ring title, of each breed he is certified to evaluate.
Since the job of the Temperament Tester is so much more difficult than that of
the Ring or Schutzhund judge, it seems reasonable that the qualification
criteria should be much more stringent. Many of the judges are sincere and
honest, but are by no conceivable standard qualified to quantify the character
attributes of a working dog.
In spite of the relatively undemanding nature of the ATTS test, the
Shepherd and Dobe clubs have had to come up with tests significantly less
rigorous because so many of the show dogs are blown away. At a time when breed
specific tests, more demanding according to the heritage of the breed, should
be coming into existence what we tend to see is a watering down of a
generalized all breed test.
Perhaps the silliest aspect of this whole thing is that you actually see
suggestions in print that people should not train for such tests. Any place in
the world where serious stock is put in the result of a test every breeder
will, and should, do everything he can prepare his dogs, so that they will not
be at a disadvantage. Any test based on the idea that people "on their
honor" will not train for it is so incredibly stupid that it bears no
further comment.
The danger is that such tests could do a great deal of harm by
certifying, at least in the minds of the public, very poor dogs as having
"good temperament" and failing good dogs because they walk around a
plastic sheet or grating on the ground or because an inexperienced person was
not able to evaluate a dog's reaction properly. The simple fact is that there
is no real evidence for their validity, and a lot of experience to indicate
that they very often provide terribly misleading results. Over the past years,
it has become increasingly apparent that establishing working character
requires hard work and a working test, there are simply no short cuts.
In summary, the concept of a temperament test applicable for all breeds,
from the Cocker Spaniel to the German Shepherd, has proven unwieldily and not
really useful. I believe that the ATTS type test may have a long term
usefulness as a low level screening criteria for breeds intended primarily as
family pets, such as the Poodle or one of the unused hunting breeds, but in its
present form has little to offer the serious working dog breeder or trainer.
Another example of the sham title is the "Herding Instinct
Test", where someone who became interested in herding last month evaluates
a dog and, if he qualifies, issues him a "Herding Instinct
Certificate" and the pooch forevermore goes around as Ch. His Majesty of
Jojo HIC and the proud owner prattles on about how wonderful it is that the
working character of the breed is being maintained.
How does this Herding test work you ask? Well, there is a great big pen
with sheep (or ducks) and the dog is let loose with the sheep. The sheep run
around, the dog runs around (I think he gets extra points for barking), and
after a while the tester decides if the dog has herding instinct. I have never
been able to discover any specific criteria, so there is really nothing involved
but the tester's opinion. The beauty of it is you can make a lot of money,
since lots of people will bring their breeding stock in and pay the fee, in
that there is no training involved, and if they pass then, by God, they got
real working dogs too! The whole thing takes about ten minutes. Neat, right? Of
course, nobody ever mentions simply letting the sheep or ducks out of the pen
and seeing if old Fido's natural herding instinct will come through and allow
him to recover the loose animals!
Don't get me wrong, when some really knowledgeable people put together
some serious, breed specific tests that really make a dog put out I am all for
it. I understand a German Shepherd can pass a herding test instead of the
Schutzhund I in Germany, but very few do because it is so difficult. I hope
that a strong herding trial movement evolves, but am concerned about practical
issues such as availability of animals to work and real applications. I am
aware of a number of experimental programs for herd protection, mostly the
prevention of coyote predation on sheep, using central European breeds such as
the Komondor or the Anatolian Shepherd, but relatively little about actual use
of dogs for herd control. All of these serious efforts are to be encouraged,
but standards must be maintained. What we don't need is a proliferation of
Mickey Mouse titles.
I think shepherds and bouviers (including the German bouviers, sometimes
referred to as Rottweilers) are such great dogs mostly because they are derived
from herding stock. It's not really all that hard to breed a race of dogs
willing to bite at any opportunity, but the discrimination, intelligence and
control that makes these dogs so great comes, in my opinion, mostly from the
hundreds of years of selection for good herd protection and control character.
It is no accident that many of the pure herd control breeds, such as the
Sheltie, excel at obedience and that the duel purpose control/protection dogs,
such as the shepherds and cattle dogs, excel at the more complete tests, such
as the Ring trials and Schutzhund.
Herding instinct tests held to provide the novice an opinion by an expert
on whether his dog showed promise, whether it would be worthwhile to train,
would be most useful and commendable. But that's not what the HIC is being used
as. Lots of people are getting the HIC and taking the position that it proves
that their breeding stock is of correct character, which is flat out Bull. I
have a letter or article someplace by the president of the Terv club stating
that the breed is in great shape because 70% pass the HIC. Utter nonsense!
(Before the Terv people come to lynch me - and they will have to stand in line
- I ask them to note that I did not deny that the breed is in great shape, but
merely stated that the HIC cannot prove it one way or the other. All I know is
that everybody in Europe works the Malinois.)
What it comes down to is that Mickey Mouse is Mickey Mouse and the truth
is the truth. I think that in concept temperament evaluations, herding
potential tests and Schutzhund/Ring aptitude tests are fine, that getting some
insight from the comments of a senior person makes a lot of sense. As long as a
private expression of opinion is all that is involved, there is no objection.
But the use of such opinions as certification of character and aptitude, as
proof of breeding suitability, is utter nonsense.
Although they have served a purpose, the time has come to do away with
"titles" based on opinion about potential and concentrate on
objective evaluations of performance. The TT, the HIC and any other such
certifications should be abandoned. Emphasis should now be put on providing the
training opportunities leading to meaningful working titles as measures of
achievement and indications of correct working character and aptitude.