The Chronicle: Colloquy Live Transcript

How Smart Is Fido?

Thursday, April 14, at 2 p.m., U.S. Eastern time

Animal behaviorists have largely ignored dogs, choosing instead to study closer human relatives, like chimpanzees and gorillas. Until recently, dogs were not thought, at least in scientific circles, to be highly intelligent. New research out of Hungary, however, suggests that in adapting to our environment, dogs have acquired a remarkable number of mental traits that closely resemble our own. Among other things, they can predict social events, provide and request information, cooperate, and imitate human actions.

But are the Hungarian scientists giving too much credit to canines? To what extent are the animals simply associating events or objects with one another, as Pavlov's dogs did, rather than actually "thinking" out problems? Is the Hungarians' research limited by their use of dogs that have been bred as pets?

» Clever Canines (4/15/2005)

Adam Miklósi, a research fellow in the department of ethology at Loránd Eötvös University, in Hungary, is part of the team whose research suggests that dogs may have evolved to be particularly attuned to understanding human communication. He has also studied social behavior in fish and rats.

A transcript of the chat follows.

Rich Monastersky (Moderator):
Welcome to The Chronicle's live chat with Adam Miklósi, a research fellow in the department of ethology at Loránd Eötvös University, in Hungary. Mr. Miklósi has studied social behavior in dogs, as well as in rats and fish. He is part of a research team led by Vilmos Csányi, the chairman of the ethology department and the author of "If Dogs Could Talk: Exploring the Canine Mind," which was published in January by North Point Press.

Welcome Mr. Miklósi

Question from Judith Grant, Ohio University:
Thank you so much for doing this wonderful research. I wanted to comment that I not only agree with your findings about dogs (based on my ownership of companion animals), but have observed the same thing in particular horses with whom a special bond. My horse can figure out how to open gates after watching me do it, can open gates to let other horses out. Furthermore, I have on several occasions modeled behavior to him that he can then repeat - specifically, I have climbed onto a platform that he was afraid to stand on. He watched me carefully and then did it himself. I think animals learn by relationship - to each other and to us.

Adam Miklósi:
I agree with you that in many respects horses share some traits with dogs (and of course us). To some extent this is not a surprise because during their domestication, horses have been selected for their ability to interact with huamns. Your observation that your horse learns by observing humans is very interesting. This has not been described by researchers but I am sure would be worth investigating in the future. We know that dogs learn also by observation but interestingly this method has been used very rarely in dog training.

Question from Karen Tibbetts, Rockford College:
Have dogs developed "intelligence" as they have increased their interaction with humans? i.e., my 3rd generation Westie seems much more "human" than my 1st, as do other dogs I observe.

Adam Miklósi:
Intelligence is not a good word but I know what you mean. Certainly, dogs (as other animals) change their behaviour over many generations. As present days' westies do not look like the "original" ones, the same could be also true for behaviour. I am not sure whether big changes can take place in a few generations when it comes to "cognition". In addition you should not forget that you have also changed over the years. So it might be that now you can see signs of "intelligence" in your westie that escaped your attention earlier. Only objective testing and observing could answer such questions.

Question from Chris Brown Mahoney; University of MN, Carlson School of Mgmt:
Is there any evidence of a difference in "intelligence" between breeds OR difference between dogs of the same breed?

I ask this because I have had Pulis for 30 yrs & while I am used to their alertness & problem solving, many who meet them for the 1st time think they are very smart. It seems that way to me, as well. I also notice what seems to be differences in intelligence between the 5 pulis I have had.....Thanks for any insights!

Chris

Adam Miklósi:
I like to use the word intelligence when it comes to compare individuals within a breed. Of course one should define what is meant by this term, for example one could speak about "social intelligence" (how one deals with social problems) or "mathematical intelligence" (how one can solve problems by the means of mathematics. Some may be good in one but bad in the other etc, and their performance will depend on many things like genetics, experience, enviromental factors, etc. As in humans there are, of course, differences in intelligence among individuals. However note that in spite of the claims of many, there are no good "intelliegence test" for dogs.

The problem comes when you want to compare breeds because there are some genetic differences which can seriously hinder performance in a task or thinking about a problem. For example, terriers like to dig. So their first idea about solving any problem is to start with digging a hole. This means that I do not like to compare intelligence of breeds but say how different they are in this or that situation.

Question from Paul, small 4-yr Texas university:
I ame fascinated with your research. Besides the Mudi, were Hungarian Vizslas used in your study.

Adam Miklósi:
Mudies were used only in some of the experimenents,( not many are living in Budapest). We work also with vizslas of course. In the case of most experiments, we used as many different breeds as we can because we would like to find out what is special about dogs in general and not what kind of breed differences there are.

Question from Dr. Philip R. Breeze, Kutztown University:
Can any of this be reliably extrapolated to dolphins?

Adam Miklósi:
If one thinks in a comparative persepctive then I would say yes. The BIG question in this field of reseach is what kind of cognitive abilities animals (species) have and how can one compare this to human cognition. During evolution, different species face different task to solve (or they die out). For example, wolves had to evolve the ability to hunt in packs because otherwise they could not get food in winter. It is fascinating to investigate what kind of problems dolphins can solve (and cannot solve). Especially in the case of dolphins, it is very interesting that they seem to be able to solve problems in captivity that we have no idea what they use for in the wild. For example out in the sea, there are not many objects yet they are very skillful with them in captivity. The problem is that dolphins are very expensive to keep and do research on; some of the experiments we did with the dogs have been also done with dolphins by Louis Herman in Hawaii, but I think we can find our more about the cognitive evolution of dogs within a shorter time, and dolphin cognition will remain a mystery for longer.

Question from Sonia Dutton, The Metropolitan Museum of Art:
A couple of days ago, my new puppy got his first haircut. Being a poodle, he was transformed from poof of wild hair to elegant silhouette. Almost instantaneously the dog whom virtually everyone on the street paid attention to for a "cuteness" factor now occasionally gets an audible whisper. While I have no interest in cultivating the former, it made me think: can dogs perceive how people feel about their appearance? And if so, how are they affected?

Adam Miklósi:
We do not know for sure, but in general I would say dogs are not interested in their appearance. In humans we know that appearance is very important in male/female choice but this is not the case in dogs. Dogs seem to habituate very fast to their mirror images.

One test to your question could be to see whether your dog looks longer in a mirror after getting his haircut.

On the other hand they might learn that after getting haircuts people behave differently. However I have not heard about a dog who would "force" the owner to get his hair cut !(just joking)

Question from Chase Billingsley Kimmel, Hanover College:
After reading "Clever Canines - Did domestication make dogs smarter?" by Woodard, it seems that Mr. Hare has associated a form of “intelligence” with “friendliness.” This seems to suggest that the friendlier a dog is, the more potential he has for being intelligent because he is more attentive to his caregiver. But when looking at various species of domesticated dogs this does not seem to be the case, for example, the Doberman pinscher is very intelligent but not so friendly and the typical Basset hound is friendly but not very intelligent. But one might say, “What actually is the definition of friendly?” What would be your explanation of this? It could also be observed that friendliness might be related to low intelligence in some breeds. Some people believe that Dalmatians are one of the bravest and most heroic breeds of dogs. People might think this because they will go into a burning building with their firefighter owners to save people. Is this obedience or just plain stupidity?

Stanley Coren has conducted Canine IQ tests to determine breed intelligence. Twelve different areas of intelligence are scored (e.g. problem solving, social learning, attention, etc) and then time and score are totaled to determine the intelligence of a breed. What are your criteria for determining intelligence? What is your definition of intelligence and how do you measure it? There are various types of factors that can affect what we think of as intelligence, such as personality, obedience, instincts, the ability to pay attention, etc. Are any of these factors addressed in your assessment of intelligence?

Adam Miklósi:
The relationship between "intelligence" and friendliness is very complex. Dr Hare argues that dogs can display their intelligence with humans because they are not aggressive toward them or do not feel fear. In general it is very problematic to apply human terms to other species without a definition. In our secientific papers we never use the word intelligence. I also do not believe in intelligence tests comparing dog breeds; such questions could be only asked with relation to dogs of the same breed. There is no such thing as "intelligence" you must always say what tests you have used to measure this or that cognitive ability. If you want a general defintion it would be like flexibility of problem-solving but this could be affected by previous experience, evolutionary heritage and learning.

Question from Bob, Professor Emeritus, State University System:
Which are smarter--labradors or college deans?

Adam Miklósi:
In my view that is a wrong type of question. In biology any type of "intelligence", "cleverness" or similar things do not exists in themselves but only with relation of the environment. So first you should define the environment in which you study them, and only then can I say (after testing) who is smarter (e.g. how would college deans hunt a bird?)

Rich Monastersky (Moderator):
We're midway through out time with Dr. Miklosi, so don't wait to send in your questions.

Question from Marilyn Cooper, Michigan Technological University:
Marc Hauser (Willd Minds) argues that different animals employ different "mental tools," and that behavior that looks the same in different animals may not involve the same kind of thinking. Have you considered this hypothesis, and how would you say it relates to your work with dogs?

Adam Miklósi:
I think that is the correct approach. During domestication dogs got the "mental tools" for how to behave/"manipulate" etc humans, and they use them well because we are contributing a lot to their survival (and doing it without force). For dogs this means that they have abilities for understanding many forms of our gestural and to some extent verbal communication, observe our behaviour and learn from it, etc. Other animals (e.g. wolves) do not do this even if they have the chance

Question from Leslie, small liberal arts univ:
I've heard that dogs have learned to make eye contact with humans whereas wolves do not do this. If this is true, have dogs learned that by making eye contact they can get a human's attention and then communicate with that human? My dog will come to me from another room, place herself in front of me and, standing still, stare at me. When I get up she runs to the door but if I don't follow she comes back, and does the same stare. She's taught me that this means she needs to go outside!

Adam Miklósi:
I think you are right. Dog-human communication goes both ways. We teach them and they teach us, or they learn from us and we learn from them. That is wonderful because it works so smoothly, and although it can also be done with some other animals (e.g. dolphins) with dogs it is very natural, and no special training is needed for either of the parties.

Question from Rich Monastersky:
The article, "Clever Canines," talks about research showing that foxes could be domesticated in as little as 18 generations (50 years). In that short time, the foxes demonstrated a marked increase in their ability to extract information from human cues. Does this research suggest that dog domestication might have happened quite quickly? And if so, might a research project with wolves be able to produce domesticated animals particularly tuned to human communication in just a few decades?

Adam Miklósi:
The fox experiment shows that you can get changes is social behaviour very fast. However foxes have not reached the "level" of dogs, there are still important differences. In principle one could try to domesticate wolves but there might be differences depending on what kind of wolf population one starts from. It is believed that our dogs originated from Asian wolves that are very different from the European ones.

It is more intresting whether one can domesticate dogs "further"!

Question from Ryan Singer, in Chicago:
From the article:

<< Michael J. Owren, assistant professor of psychology at Cornell University, says Mr. Csányi's team may be underestimating the flexibility of associative learning, the most basic kind of learning that comes not from "thinking" out the problem, but simply by associating events or objects with one another, as Pavlov demonstrated with his dogs. >>

I wonder how "associating events or objects with one other" differs from "thinking" in the above. It seems that in order for two objects to be associated in a mind, each object must be represented as a concept (it must be something that can be differentiated from everything else from the perspective of the experiencer) and the activation of one concept should give rise to the other (if they are indeed associated). This sounds a lot like thinking to me.

Adam Miklósi:
I do not like this debate (in science) about "associate" versus "cognition". Actually we avoid saying anything about the inner processes because there are no proofs for it. There is no mental or cognitive ability without learning, so one cannot debate the existance of "associations" but we have no test for it, and neither can one do control experiments. The dog is not a rat despite of what many think. So I like to speak about "abilities" that are special in dogs, some of which, of course, might emerge by association learning.

Question from Dr. Alka Chandna, University of Western Ontario:
I was interested to read about the work of Manuel Berdoy, an animal behaviorist at Oxford University. Berdoy released 75 lab-born rats into a farmyard and documented the quick recovery of wild behaviors in the rats once they were released. The rats found water, food and hiding holes almost immediately. They started to establish social hierarchies within days, and it was only a few weeks before they had established an extensive pattern of paths across the colony. Although the rats had spent their whole lives being fed on pellets, the females immediately prepared for pregnancy by foraging and storing appropriate food. I wonder whether Dr. Miklósi's work with rats would offer similar insights into rat behavior and cognitive ability.

Thanks very much!

Adam Miklósi:
Actually, I had only a very limited experience with rats testing their ability to learn by observation of others. On the other hand, if Berdoy had released the rats in the wild, as opposed to the farmyard, none or may be a few would survive there. This is not about cognitive abilities but about experience and genetics because laboratory rats have been selected against living in the wild: e.g. that should not avoid predators (as it would be normal). This way humans (resarchers) can work with them.

Question from Bruce Friedrich, Grinnell College:
I see that you are also studying rats and fish--that's extremely interesting. I've read that fish have memories, learn from one another, and can do better on some cognition tests than dogs, and that rats can dream and play and learn, also. It's all so interesting. How would you compare the cognitive capacities of dogs, rats, and fish? What other interesting little tidbits do you have about rats and fish, from your behavioral research with them? Where, in English, might we find some of your studies? Thanks so much--very, very exciting stuff. Thank you for doing it.

Adam Miklósi:
I would never compare dogs, rats and fish (actually there is no such thing as a general "fish", but there are individuals belonging to certain fish species). These animals live in different environments so the abilities and challenges are very difficult. When you ask about comparison you want an objective answer but this is impossible. Could you tell me a problem that would be similar for "fish", rats and dogs. Our group has published over 50 papers on dogs, I have another 15 on various animals. They are all in English, and you can get some of them by searching on the web, or looking at our webpage (or write me by email)

Question from Bob Louisell, Professor Emeritus, St. Cloud State University:
Some say humans can be distinguished from animals because of their facility with language. Others say mathematics; still others say morality. Do you have any evidence of moral behavior having evolved in dogs?

Adam Miklósi:
No or very little. I am not a philosopher but I think of "morality" as a kind of social aggreement among people. This would mean that we should have such kind of aggreement with our dogs which seem to be unlikely. But there are behavioural traits both in humans and dogs that can be the basis of "moral" behaviour, for example "helping" or "cooperating". I think only human social system (with the means of communication, etc) can form such abilities in complex "moral" behaviour.

Question from Debbie Huerta, Colgate University:
I certainly concur with Judith's comments and also thank you for your research and papers about dogs. We have had dog's (companions) who initiate interesting behaviors after making a connection between a certain behavior and a desirable outcome. For example, our standard poodle would go to a window and bark with great enthusiasm. Our labrador would join her abandoning her food, bone, toy, spot on couch, etc. Said object or place would then be taken by the poodle. Perhaps it is unfairly easy for a poodle to deceive a labrador, however, I wondered if you or members of your team have had the opportunity to observe or study this sort of behavior? Thank you.

Adam Miklósi:
In our class on "animal cognition" this is one of the examples I use, because I can explain that this behaviour ("deceit") can be the outcome of both "learning" and "cognition." However at the moment, it is very difficlult to find out what really goes on here. Such behaviour is also very difficult to test in the laboratory or even at the home of the owners. Neither dogs nor humans "lie" on command but I agree it is a very interesting topic. An interesting question would be whether your poodle would do it also at other places or with other dogs...?

Question from Cecile McKee, U. of Arizona:
I'm raising a dog the belongs to an organization that breeds and trains dogs to help people with various disabilities. The experience has helped me appreciate individual variation across dogs in new ways. A couple of questions have already hinted at variation issues. Could you comment more on this, both across individuals in a breed and across breeds?

Next, most organizations that train service dogs have a high proportion of dogs that don't make it through the testing. Some of this is of course health issues (a dog that's going to sometimes pull a wheelchair cannot have hip trouble), but a lot of it relates to training and 'attitude' (like gentleness or willingness to 'work'). Can you speculate on what such organizations might do to improve training? Or what puppy raisers might do to improve dogs' chances to do well in the later training?

Adam Miklósi:
1. with regard to breed, the situation is easier because there are certain limits both physical and behavioural for many breeds. As you would not use a small poodle for helping people in wheelchairs, you prefer dog breeds with ceratin characteristics (e.g. retreival abilities) for you goals. The question of individuals is more difficult. At the moment there is no test for young puppies telling us how a dog will behave in the future. In addition, they will experience a lot (and we will not know about most of this because we are not there). Then comes the training experience etc. So I think large variation is the normal situation in most cases. And this brings me to the second point because I have always wondered why people think that dog should pass the training at 100%. At our university we have 160 biology students, and only 10 or 20 of them will be a biologist. Are we doing a bad job as teachers? I think not. Being a biologist or helping a disabled person are very complex things. It is inevitable that only a minority can do it. Of course there is also room for improvement with very early training and selection of the puppies as training proceeds. But there will be no perfect solution partly also because the task for these dogs is also very different depending on the future owner (actually we also train dogs for disabled here in Hungary and have the same problems)

Question from Linda Sommer, U of Indianapolis:
Have you done any work with dogs that develop a disability and how that affects their interaction with their environment? I grew up around a blind dog who was quite remarkable at adapting, and as an adult had a dog who went deaf. Both learned to cope quite well without human intervention. That seemed quite an indiction of intelligence to me.

Adam Miklósi:
Actually, not but I think this kind of adaptation to loss of senses is based on the flexibility of the nervous system and not "intelligence" as such. The problem is that, in wild animals, such individuals would not survive for long but in captivity they have a better chance. In Budapest in the Zoo we had a blind male wolf living with his son, and the younger wolf brough the meat to his father regularly.

Question from Mary, medium public comprehensive college:
Certain breeds are noted for certain traits--hunting, retrieving, pointing, herding, digging. Does your research have any implications supporting or questioning this characteristic?

Adam Miklósi:
We try to avoid compapring breeds. In the mind of dog-loving people, breed" is a well-defined category, but in reality, it is not. This is because people think in pictures (the photos of a breed) and not actually in terms of real animals and their behaviour. You can find in all breeds individuals who might dig, retrieve etc. But in some cases, it is very difficult to train a dog for doing such and such a behaviour whilst in other cases it is easier. So breed difference is not a black or white issue!

Question from Ellen Dannin, Wayne State University Law School:
Jó napot! I saw a positive review of your book in Science News this week. What are the intriguing questions you see out there with regard to canine intelligence?

And I have a dog I got when he was 4. I thought was not very bright. So I never tried to train him. But I noticed one day he was trying a trick I had taught my other dog. He learned by watching his canine companion. It helped that treats were involved.

Each of my dogs is smart in a different way. Figuring out their strengths is a lot of fun.

Adam Miklósi:
First of all it is not my book but my professor's Vilmos Csányi. There are many questions to answer. Actually, I prefer smaller questions to bigger one. We are conducting studies to find out how much really (on an everyday basis, without special training) dogs understand about human's visual and verbal communication. We are running studies to find out what dogs can learn from each other or from humans (just as your example shows). We are looking into issues how the relationship of the dog with different members of the family influences their behaviour. We are also studying dog barking to see whether there is information for other dogs or humans. Köszönöm a kérdését, üdvözlettel

Question from Tom Boyer, non-instituted:
Have you ever heard of a dog immitating vocal exercizes? As a singer, I sometimes do this descending ooh sound, down an octave. I had a jack russell that copied me,and even could be prompted by a pitch pipe to do the exercize, and pretty in tune too!

Adam Miklósi:
I know of stories (anecdotes) about such "singing" dogs, and we have also some around that would start "calling" or howling after listening to some musical instrument. Actually, this is one experiment that we are running now, so I hope to report on this in the next few months.

Rich Monastersky (Moderator):
That's about all we have time for. I want to thank Mr. Miklosi for staying up late (in Budapest) and answering these interesting questions. And thanks also go to all of you who participated in today's discussion.