Do Dogs Have a Theory of Mind?

And my previous reply seems kind of snotty.

Conceivably, one can come up with a set of conditions to deny ToM in dogs. In fact, any scientific theory (in the classical sense of observable repeatable data) has what is called falsifiability. Conditions under which the theory could be proven wrong. For example, if Newton throws and apple and it fails to hit the ground, the "law" of gravity could be in trouble. That theoretical condition is known as a test of faslifiability.

However, those who wish to avoid anthropromorphism in analyzing dogs are guilty of possibly doing that very thing by defining ToM by human standards. That is, by saying that dogs could only have ToM if it exists by how humans "do it." Just because your dog doesn't text you on your i-phone or blackberry doesn't mean the dog doesn't have ToM.

In other words, does a dog have ToM? And by what definition? Dogs do not necessarily think like humans do nor do we always have the same priorities. Such as the mirror thingy. Whether a dog interacts with a mirror or not does not prove or disprove ToM. Again, we are judging their ToM in mirrors because of how we judge ourselves. Anthro-ing like a big dog, so to speak.

I have personally seen ToM between my dog and neighbor dogs. A stranger cat was walking in front of a neighbor's house. My dog barked and the border collies caddy-corner to us moved to positions that would allow them to see where the cat would be a moment from his barking. No, I did not have any white lab coats or cameras going. But it was a communication. The BC's determined what the "intruder" status was as well as a projected or future location. So, a) there is Shadow's desire to express a viewpoint, from whatever motivation, b) the other dogs' ability to hear and value Shadow's "perspective" and extrapolate it for their own benefit. Which fits in with what I know of dogs. Dogs are about resources aquisitions and protection thereof. The ToM a dog would have serves the needs of a dog, not necessarily what a human needs it for or views it as.

So, while we are trying to define falsifiability for canine ToM, we must also determine what are the reasons for the tests of falsifiability we choose and whether they are accurate and germaine, or simply reflect our human ego and arrogance. How could a dog possibly be as smart as a man? (Actually, in some cases, anecdotally, I think dogs are smarter. )

ron2

Conceivably, one can come up with a set of conditions to deny ToM in dogs. In fact, any scientific theory (in the classical sense of observable repeatable data) has what is called falsifiability. Conditions under which the theory could be proven wrong.

However, those who wish to avoid anthropromorphism in analyzing dogs are guilty of possibly doing that very thing by defining ToM by human standards. That is, by saying that dogs could only have ToM if it exists by how humans "do it."

I have personally seen ToM between my dog and neighbor dogs. A stranger cat was walking in front of a neighbor's house. My dog barked and the border collies caddy-corner to us moved to positions that would allow them to see where the cat would be a moment from his barking. No, I did not have any white lab coats or cameras going. But it was a communication. The BC's determined what the "intruder" status was as well as a projected or future location. So, a) there is Shadow's desire to express a viewpoint, from whatever motivation, b) the other dogs' ability to hear and value Shadow's "perspective" and extrapolate it for their own benefit. Which fits in with what I know of dogs.

Dogs are about resources aquisitions and protection thereof. The ToM a dog would have serves the needs of a dog, not necessarily what a human needs it for or views it as.

So, while we are trying to define falsifiability for canine ToM, we must also determine what are the reasons for the tests of falsifiability we choose and whether they are accurate and germaine, or simply reflect our human ego. How could a dog possibly be as smart as a man? (Actually, in some cases, anecdotally, I think dogs are smarter. )

Hi, Ron,

Thanks for the comment.

I agree that in some ways dogs are smarter than us.

As for whether they have a "Theory of Mind," that's a theoretical construct developed by philosophers and cognitive scientists. It's used to differentiate between the ways humans and animals think. It's not a construct I necessarily agree with 100%, but that's its purpose.

In actual fact, the ability to entertain another being's point of view can be seen on fMRI studies of the human brain. And the areas of the brain that light up aren't found in the brains of other animals (except perhaps some cetaceans).

I'm not getting a clear picture of your example, but it sounds somewhat similar to behaviors seen when wolves or African wild dogs are hunting together. It's quite well-known that the human mind is hardwired to interpret animal behavior through humanlike mental thought processes. So I would encourage you to re-look at the scenario you've described, try to remove your personal bias (i.e., that the behavior "fits in with what I know about dogs";), and, just as an experiment, see if you can come up with an alternative explanation that doesn't require the dogs to be aware of each others' "knowledge states," but that only requires them to be able to tune in to each others' feelings and desires in that heightened moment.

Also, for your example to be relevant, scientifically, I think it would have be repeatable under varying circumstances with different sets of dogs, and you'd have to have videotaped evidence to show exactly where each dog was in relation to the others at the time it happened.

Thanks again. I enjoy your posts.

LCK

PS: I don't believe dogs are about the acquisition and protection of resources. That makes them sound like greedy little bastards who only think of themselves and their needs. That doesn't fit at all with what I know about dogs.

Were either of you watching ABC recently when they broadcast a story about a gentleman whose brain image is consistent with that of a serial murderer, with the parts that influence impulse control severely lacking?  His a functioning, compassionate, upstanding member of society. Possibly, the fact that he grew up in a nurturing environment without some of the variables that influence behavior toward the criminal side of impulsivity had a mitigating effect on his genetic makeup. 

P.S.  I don't think the acquisition and protection of resources makes dogs greedy little bastards any more than humans would be considered greedy for trying very hard to feed and shelter themselves.  There are, of course, extremes in both species.

I'm coming to this discussion late, but I feel like there is a crucial point that is being missed by some of the examples of supposed ToM in dogs. For example, one can say that a dog may do something when your eyes are closed that they wouldn't do when your eyes are opened. Or dogs communicating a need that humans then fulfil. I will say that my hare does the same things and there is no way he has a ToM. He doesn't even really know his own name. He has a very cool animal genius about him, but it makes sense how in interacts and communicates when you look at simple cause and effect. Example: when someone is looking at him this is significant to him not because they may be able to SEE him, but because if he runs they are LIKELY to chase him. When someone looks at him in a crouch with their muscles tense, they are even more likely to chase him. If someone looks away from him, they are unlikely to chase him. If someone looks away from him and leans away from him at the same time, they are even more unlikely to chase him. He certainly responds to eyes, regard, and the balance of weight in other animals, but I think that it is an extra leap to say he does so because he is imagining what they might want, for example. He responds because it means something to him, something more basic than what another being may be thinking.

I think we need to carefully consider this in relation to dogs as well. How can we say that the dog takes that extra step from communicating a need to someone and having that need met, or from noticing that if no one is looking at them they are free to do what they like, to being aware that these effects are the result of a human having their own feelings and desires? To me, that is an extra step that is not necessary, so therefore I don't buy it for now.

 To be honest, I don't see how any of your examples prove there is no ToM in dogs.  My guess is that there is, although not to the extent that humans possess it, and if we are always filtering through our own perception of what ToM is, then we will miss what there is of it in dogs. I am not alone in that belief.  There are some fairly heavy duty behaviorists who also acknowledge the possibilities.

Re the mirror thing, just an interesting observation that may or may not be relevant.  Yesterday, while on my way to the training center, I had both Sioux and Sequoyah in the truck with me.  Sequoyah was lying in her crate, but Sioux was on a bed directly behind me, but in front of Sequoyah's crate.  I pulled in to McD's to get a coffee, and as I was waiting in line, happened to glance into my rear view mirror.  Sioux was watching the mirror and as I glanced up, she shifted her glance to meet mine - in the mirror!  I opened my mouth into a "doggy happy face" and so did she.  Mind you, my face was not visible to her except in the mirror. Now, I know that she could be thinking that the mirror is alive and has a blond lady in it, but she could also think that the image is me.  Don't know, but clearly she interacted with my reflection on some level.  That's what I mean about "we don't know yet quite at what level" dogs' cognition operates.   I, for one, am not about to suggest that they have more ability than they have, but I'm not closing the door on finding out that they have more than most people are aware of.

Corvus, I am not sure under what conditions you determine that there is no ToM in your hare. Even LCK has pointed out that ToM is a way of defining a difference in how humans think from how dogs think, though that is as far as he goes and would, as one must in a double blind, not assume their is ToM until proven by repeatable results. But therein lies the rub. If ToM is admitted to be just a way that humans assert their intellectual superiority over dogs, then it is self-serving to say that dogs don't have ToM.

Will someone, even on a dare, try what I said from my friend, Lee? Point at your hare, cat, rat, whatever, and see what they do. At best, most of them will look at your finger. A dog will look where you are pointing to. That means, to me, that they are considering that you have a vantage point or knowledge that they can gain by looking at it from your perspective. Which means that they have the intuition that you are outside of themselves with a mind of your own, un-canine as it may be.

True, that's not definitive proof of ToM in dogs. And are we just splitting hairs? Of what value is it to find whether or not dogs have ToM as we define it? Let's say that I could set up a videocam and randomly introduce, through a 3rd party, stranger cats to walk around the neighborhood. And capture my dog and the neighbor dogs doing this reconaissance maneuver over and over again. Even with different directions of travel for the cat.

The argument meant to attempt falsifiability of ToM will then say that, even then, that's not ToM. It's just dogs reacting emotionally to an intruder and assuming positions by habit that give them the best view to everything.

And we're right back where we started. Trying to prove a sentience in dogs when, as LCK admitted, the ToM thingy is just our human way of saying we are smarter than dogs. All because the dog didn't say "I see what you mean" in the Queen's English.

I may think my dog has ToM and he is a problem solver. But that doesn't mean I expect him to solve a quadratic equation or find the second derivative of a multi-variable tensor equation. But his not being able to do those math functions (I think he can't do those but I am not sure that he can't) doesn't mean that he is not smart or does not possess intelligence or does not have anything losely defined as ToM. He understands sounds that I speak in english. He probably could have understood sounds in spanish or german if I had trained him using commands in those languages. But it doesn't mean that if I say "Como estat, Shadow?" that he will reply with "Muy bien. Y tu?" But just because we don't exactly share a spoken language doesn't mean that he doesn't have ToM.

So, we need a definition of ToM independent of human arrogance and conditions. Now, there's a challenge.

spiritdogs

To be honest, I don't see how any of your examples prove there is no ToM in dogs.  My guess is that there is, although not to the extent that humans possess it, and if we are always filtering through our own perception of what ToM is, then we will miss what there is of it in dogs.

ToM IS a human construct. If we're going to change what it is so other animals can have it too, then what's the point? I'm not disagreeing with you, but if we want to accurately describe what dogs have and need to invent a new term for it, then that's what we should do. Not apply a term that doesn't really fit. I don't think any of my examples prove anything and I wasn't trying to prove anything, just offer an example in which a simpler explanation may suffice. An understanding of cause and effect is not a ToM, whether it involves other individuals or not.

spiritdogs

Sioux was watching the mirror and as I glanced up, she shifted her glance to meet mine - in the mirror!  I opened my mouth into a "doggy happy face" and so did she.  Mind you, my face was not visible to her except in the mirror. Now, I know that she could be thinking that the mirror is alive and has a blond lady in it, but she could also think that the image is me.  Don't know, but clearly she interacted with my reflection on some level. 

I know someone who is doing research on this very thing. They call it using the mirror as an information-gathering tool. Apparently pigs do it.

ron2

Corvus, I am not sure under what conditions you determine that there is no ToM in your hare.

*shrug* I don't. I determine that there is a simpler explanation. He's damn good at body language. Other creatures in his world do things that are significant to him because they disrupt his sense of security. He interprets and reacts to those things because it alleviates his anxiety to do so.

Granted, I expect that the hare reacts instinctively, akin to the non-linear dog theory, in trying to maintain his own emotional equilibrium. I just wasn't sure that the predilection toward instinctual behavior ruled out ToM, nor am I aware, either, of anyone investigating this in hares, domestic or wild.

 After invoking the well-known British philosopher, David Hume, in support of my views on animal consciousness and reasoning on another thread, I realized his opinions should be brought to bear here, as well.

 http://www.bartleby.com/37/3/13.html

Hume's philosophy laid out 250 years ago had an inestimable impact on modern scientific inquiry.  He was decidedly opposed to Descartes idea of animals as unconscious, unfeeling mechanisms akin to 'cell-phones' .  His work profoundly influenced Darwin some 100 years later.

 There is no quantum dog or electrical battery theory of group consciousness in Hume or Darwin, as opposed to LCK's professed philosophy of dog training.

Burl

It may be of interest to recall that Ted Kerasote's dog, Merle, in the book _Merle's Door_ did not recognize himself in a mirror _until_ Ted stood next to Merle, and then Merle recognized himself as the companion of Ted, Thereafter, Merle looked at the mirror to see himself w/ or w/o Ted around.

Regardless, recent thinking has it that the mirror test is probably not very useful in determining a ToM. I would think research methods of infant psychologists like Piaget would be useful in studying dog cognition.

 I do not know much of Piaget, but more and more dog cognitive research seems to be pointing at the legitimacy of making close comparisons between the human pre-linguistic toddler's mind, and that of our dogs'.

 Here is a child cognition researcher that interests me

You know, Burl. I have noticed there is not falsifiability tests for NDT. I explained one for Newton's Principia Mechanica. One could define one for reward/punishment training, such as the thread on Patricia McDonnell. One could say, if she had to re-home a dog, then +R failed, she failed, the whole skinner-type behaviorism theory failed. Which doesn't prove it failed but is a test of falsifiability, especially if the failure was repeatable with any dog in any situation, which it is not. Therefore, as a test of falsifiabilty, denoting Patricia's "failure" does not constitute a valid evidence of falsifiability.

 I never took seriously LCK’s feigned charge that Trish’s re-homing represented a failure of +R.  It was merely useful as a segue to talk of his superior NDT ‘theory’.  A ‘theory’ by the way, that he is noticeably unable to describe or defend here on this forum.

To re-home a dog is to recognize that they have unique personalities (temperament falls too short in explanation), and that s/he would be more compatible with a different host personality.  To believe otherwise, one would have to view a dog as an unconscious  mindless mechanical device, like a cell phone, which can be programmed to talk to and return predetermined responses.

This is, in fact, precisely LCK’s NDT theory of dog and it is obviously at odds with most normal dog hosts.  My theory is that these unconscious ‘canine cell phones’ of LCK do such a good job of fooling us into believing and responding as if they are sentient reasoning subjects that, if nothing else, they are really great actors!

AND, this innate superior acting ability of dogs would no doubt make a self-proclaimed gifted acting-school attendee like LCK quite jealous.



   

And humans have unthinking, autonomic responses to stimuli, as well. For example, in confrontation, we often get a surge of adrenalin as our body preapres itself for fight or flight. Either way, we are unthinkingly preparing for an incredible exertion. There are some factors of human sexual behavior based totally on pheromones that we don't get to "decide" with our, oh, so , magnificent intelligence.

I will hurt some more feelings. Our language is an advantage over other animals and it may inform our style of thinking. But the sounds of it are still a series of grunts, buzzes, and whistles. We don't sound that much different from a mountain gorilla. And I know someone can say a mountain gorilla can't communicate. That's fine, except for the doc I saw on mountain gorillas where this one momma was cradling her infant and singing her infant to sleep. It was a low, almost glottal sound but it had pitch and melody. It was soothing to both mother and infant.

Each animal has autonomic responses to the environment and those responses often lead to survival or are meant to. But I'm not sure it limits or is the key to intelligence.

My dog may not speak to me in the northern dialect of German but that doesn't mean he doesn't think and not having a common spoken language doesn't forgo ToM. I could be wrong but, to me, least hypothesis is to say that dogs have some thought, perhaps some ToM. It explains a lot, quite adequately, even if it is not totally accurate or hasn't accounted for all things. An apple still falls to the ground, whether it is due to a force of gravitation between objects of mass or is the result of "space-time" (einsteinian word meaning aether) being deformed by objects of mass.