Monday, November 03, 2008

Animal Minds

The recent post at Atomic Nerds (linked by Steve below) plucks a number of threads that run through this blog. If you have animals or hunt animals---and especially if you have animals that hunt animals---give it a read.

As a falconer who hunts (with a team of dog and hawk) a variety of common and familiar quarries, the notion of what animals might know comes up a lot. Every hunter ought to think about it.

I agree completely with the premises of the post, that animals (perhaps most of them) prove some sense of self when we look closely enough, and that we haven't looked very closely at most of them.

But two points gave me pause. One is related here:

"A wild dog learning to hunt simply cannot rely on learning to capture prey by small, digestible, simple sequences that then eventually link up into a complete behavior; the way a prey animal behaves is FAR too chaotic and unpredictable to rely on that kind of learning, because the sequence would never be repeated in the same way, would rarely even be begun in the same way. Thus, the animal must make simple plans based on the rapidly changing circumstances, and be able to think in a flexible enough way to try to solve problems as they arise rather than repeating stereotyped sequences of behavior and varying them slightly."

The point is that humans teach dogs to hunt by simplifying the lessons, breaking them down into small parts and stringing them together. Wild dogs wouldn't have this luxury, because their prey behave unpredictably. Thus, they must generalize and reason their way through.

This is an argument for intelligence in dogs, and is true I think as far as it goes, but I think it falls short on its assessment of prey.

As a falconer, I've been a participant to perhaps 20 thousand pursuit sequences involving at least one predator (more commonly two of two species) and a variety of prey animals. I arrive at the figure by multiplying the number of prey my hawks catch each year by 4, which is a fair assessment of their attack success (25%).

In my experience, which must be something like what wild predators face (since our prey is the same), prey are anything but unpredictable. In fact, I think if prey animals were any less predictable in their habits and escape sequences, most predators would starve within a week.

I don't think this invalidates the point that dogs are smart. It may in fact support it. But I wonder what it tells us about the prey, and whether or not they are so different than their predators, all of whom are famously "creatures of habit"?

Which brings me to my next small point of contention, or maybe just a point to ponder:

"One of the problems with judging if another animal has something like 'a sense of self' is that their minds are highly likely to be RADICALLY different from ours, to the point where they may not even really be comprehensible to us, as our minds are heavily rooted in our own perceptions, instincts, and frames of references."

Again, from my lay-experience as a falconer, I assume precisely the opposite. My basic assumption when training a hawk and a dog, and indeed when anticipating the behavior of our prey, is that we are all "of one mind." We will all come to similar conclusions based on similar circumstances. In fact, I don't think training would be possible were this not so. What basis would we have for understanding each other?

I think we clearly overestimate our training ability as humans---our animals are not nearly so well trained as they are simply correct in anticipating our behavior (provided we are consistent). Training a hawk is really just another way of training oneself to be consistent. Most animal trainers would agree with that, I think.

So there are my observations. Anyone want to weigh in? LabRat---what am I missing here?

38 comments:

Anonymous said...

It occurs to me that learning, for a wild dog or a hawk is different than learning for a human; the acuity of a hawk's vision or a dog's sense of smell is overwhelming when compared to ours. We should not think about how we learn, (through concept formation and dull physical sensory apparatus) and apply it to creatures whose other senses are so much sharper than ours. I mean, who needs stinkin concept formation when you can tell whether a dime is heads or tails from a quarter mile away? (OK, you'll never have air-conditioning...or even a swamp cooler! But neither will you worry about the differences between 'redistribution of wealth' and a 'fair tax'.) Just because animals learn doesn't mean they form concepts. I am prepared to have to re-think this, because I see hawks doing things I never would have guessed them capable of doing. You can see an animal cycle through known behaviors, but you can reinforce for creativity, and then things get real interesting. Also, I believe you are right about the predictability of prey species. Most of them behave the same way, day in and day out. Then there are those damn swing voters- 24 hours before an election and they still don't know who they are going to vote for- these are the giants that determine our political fate? Do they determine genetic fate as well? (Awful thought. Think I'll go have a glass of wine.)
Thanks for a great post!

Bridget

smartdogs said...

I think that this touches on the issue of implicit vs explicit learning.

Implicit (or tacit) knowledge includes things like physical skill sets, habits, culture and mental maps that we process mostly in the background - therefore making our access to them quick, fluid/adaptable and readily available. It is experiential knowledge and must be learned first hand. Most of what dogs know they know implicitly. There are therefore the easist skills to teach them.

Explicit knowledge is strictly ordered and rule-bounded. Things like grammar, balancing a checkbook, learning to play the cello (though once mastered this then becomes implicit!) It can be described and transmitted to others second hand. Teaching a dog to heel would be a form of transmitting explicit knowledge. It takes a lot more work than working on implicit skills - and it doesn't stick like implicit skills do. It requires practice to maintain - and HUGE amounts of practice to turn into implicit knowledge (as with guide or assistance dogs).

Matt Mullenix said...

I don't know if this is here nor there, but I think it supports the notion that we are not so so different as thinking animals: One of the most surprising (and strangely comforting) things I've learned by working with hawks and dogs is that they make mistakes.

Despite the measurably better eyesight of a hawk, it will still mistake a crow for a starling at 200 yards---a 85 gram kestrel will flap his wings and bate at a distant bird that could swallow him whole. That's a mistake I've seen repeated dozens of times, and one I've seen overcome with experience. Kestrels learn not to bate at crows.

Dogs err also, in my experience most commonly when they act in the heat of the moment with too little information. Hey---that sounds familiar, doesn't it?

To my eyes, this suggests we are all victims of imperfect and incomplete sensory input, and that who we are stems directly from our responses to that fact. We exist somewhere in between the physical world and our impressions of it. For no animal are those exactly the same.

What remains the same, despite our imperfect understanding, is our shared environment. The same rain falls on us, the same sun shines. Therein, somewhere, lies the truth we all learn to trust. It's what we have in common, and it's the basis of our demonstrably successful ability to communicate across species.

We may indeed all be radically different in mind. We can't know eachs others' minds directly. But this is as true of me and my wife as it is of me and my dog.

The simplest assumption we make everyday is that our spouses and colleagues think approximately the same way we do. We test this constantly through speech and action and are shocked on the rare occassions it proves incorrect.

I say we do the same with our animals. I know I do with mine. And I think they do the same with me. It's all we have to go on, but it's generally enough.

Daniel Newby said...

"A wild dog learning to hunt simply cannot rely on learning to capture prey by small, digestible, simple sequences that then eventually link up into a complete behavior; ..."

Then why does the play of puppies and kittens looks exactly like practicing a set of simple behaviors over and over? Pounce, pounce, pounce, run, run, stalk, pounce, pounce, ...

Anonymous said...

The debate on other animal intelligence as compared to ours(the human animal) is one of those issues that will go on forever. Of course there are huge differences one must try to be realistic about, as well as realizing that some animal perceptions(like the ones based on scent, for which we humans are virtual cripples--me especially as I have NO sense of smell at all!) are light years ahead of ours--yet there is common ground. Those who want to insist that animals are all just predictable, instinctive machines, are usually the people that have had the least hands-on experience with other animals--whether training them or closely observing them in the wild. And the ones that have the most success in training and living with animals tend to be the ones that grant them the respect of fellow understanding and at least some reasoning ability. Most "Instinctive machine" type scientists have poorly trained dogs, if they have a dog at all! My experience has been, that as soon as you come to a definite conclusion about some other animal's behaviour, said animal will then proceed to contradict all previous conclusions! And I cannot agree that "prey" animals behave so predictably--I have seen enormous variation, especially in Whitetail Deer, and it depends on how hard and what types of hunting pressure they are experiencing as a population. Humans can act stupidly predictable in a life-threatening, panicky situation, to where if this was your only observation of humans, you would not think they were very "reasoning"! People trying to escape a burning building act very much like sheep or cattle in a similar situation, for example. Prey animals often get short-changed in the smarts department, but I have seen rabbits, deer, and certainly coons, foxes, and coyotes(the last three being predators themselves, though) pull off some inovative stunts while trying to evade capture/death that from a human would be considered brilliant! A fascinating subject to ramble on about forever!....L.B.

Matt Mullenix said...

LB: please don't misunderstand. I am not saying prey animals are automatons incapable of surprising behavior. I have been amazed and surprised so many times by various birds and rabbits, etc., I could never conclude that.

However, it is not the case that prey animals (or any animals) regularly behave in a way that completely defies expectation. If they were so unpredictable, there would be no point in trying to catch them. The fact that they can be caught on a regular basis by regular, run-o-the-mill predators suggests they are behaving in a predicable fashion.

This is to take nothing away from them in terms of the physical and even mental feats of escape every animal can pull off when pursued. As I mentioned, my hawks and dogs rarely catch more than 25% of what they pursue or ambush; it's a numbers game, and they have to keep at it. It is just not not a random numbers game...

smartdogs said...

"The simplest assumption we make everyday is that our spouses and colleagues think approximately the same way we do. We test this constantly through speech and action and are shocked on the rare occassions it proves incorrect."

I kinda hate to do this but I think I need to step in here and offer a word of caution. My job is working with dogs that have behavior problems. Often severe ones. And one place that these problems start is when people *wrongly* anthopomorphize with respect to their dogs.

Your dog definately has a very different view of the world than you do. If you're observant and mindful (which, Matt, I suspect you are) it's not terribly difficult to figure out where those areas of dissonance exist and work your way around them... but a frightening (or lucrative, depending on your viewpoint) number of people have an astonishingly difficult time doing this.

I have a scary long list of stories about people in both conscious and unconscious denial about how their dog perceives and understands the world.

But - this does not poke holes in the idea that *dogs* are quite capable of putting themselves in *our* shoes. In fact, I see a lot of cases where the dog is far better at it than the humans he live with are.

Anonymous said...

Oh, Mr. Mullenix, I am not disagreeing, just elaborating! I think humans AND (other) animals have a base set of behaviours that vary according to circumstances/culture. In the case of prey animals, a lot of them never get the chance to learn new tactics. It seems to take close calls and a certain amount of luck to stimulate the learning process for ALL of us! Two good examples I have witnessed are #1: animals that will "tree" when pursued by hounds(their natural instinctive response to canine predators),after a close call by being wounded or seriously frightened by the accompanying hunter, will sometimes learn to "clear out" as a hunter approaches the tree, clearly deciding to risk the dogs over the hunter(and often escaping!) Another example(#2) are deer(Whitetails) that have been pursued and trailed A LOT by dogs, and their incredibly savvy responses in foiling pursuit, that you will not see from deer that live in areas where pursuit by dogs is rare or nonexistant. Such inexperienced deer will often behave in a panicky, stereotypical manner that can definetly be taken advantage of by a hunter. But are these deer less intelligent, or just inexperienced? I think they have the CAPABILITY to learn, but just don't get the chance. But if the new type of predation and pursuit continues, you can bet SOME of the critters will survive and learn! And the predators/hunters must adapt in turn....L.B.

Matt Mullenix said...

Daniel: good point! I think there is definitely something to that.

Smartdogs: I agree that many human trainers are blockheads. Most are carried along by the good will of their dogs. :-) But I still don't think we are so different on the inside. Human understanding may be slowed down by layers of language and conceptual knowledge, where dogs just call it as they see it. ...When we do likewise, we communcate better.

LB: also agreed that experience matters a lot to both predators and prey. And when either one is overmatched by the other's wisdom, everything will go south.

LabRat said...

I'll be back with more later- for obvious reasons, my mind is scattered to the four winds today.

Much of the posting I've been doing on this is as much the process of thinking about it as it has been the conclusions, and one conclusion I've come to is that that line comparing chaining small behaviors together to the way dogs hunt... suffice it to say what I was TRYING to say is that the school of thought I see referred to in this thread as the "instinctive machine" doesn't make a great deal of sense. I didn't mean to imply that prey aren't predictable, but that the process of hunting is nonetheless fluid enough that you do need a mind much more flexible than this learning model suggests. And in this much, we obviously agree!

Daniel- play is the process of practicing those simple components, much like you need to build the muscle memory of throwing and catching to play baseball, but learning the game itself- and more importantly, how to play it WELL- is several orders of complexity above those basic motions.

My better half is prodding me to get moving so we can get to the polls and get lunch, so I'll break off here, but Matt, if you haven't read my followup yet, do- I think you and I are thinking along similar lines, and also that you may be ahead of me.

Daniel Newby said...

LabRat - Brains seem to learn a sequence of actions by storing it as a complete program, which can later be activated by simply "pushing its button". The marvelous thing is that a learned program can include pushing the buttons for other programs, and even cues about when to make decisions and what the allowed answers are. Quite complicated behaviors can be built up that operate without the need for much cognition. They are learned as progressive elaborations upon simpler behaviors.

Consider human sleepwalking, where incredibly complex behaviors are performed for no goal, with no apparent consciousness, and often at great risk. Sleepwalkers have even been known to drive their cars around town! If a human can do that with their executive function almost turned off, I have no trouble imagining a coyote hunting down a rabbit with little resort to logical deduction.

LabRat said...

Daniel- I think we're talking about two different things; I never meant to imply that a coyote reasons his way through a hunt.

To borrow an example that was brought up discussing this elsewhere, an athlete "in the zone" spends very little effort on conscious executive planning, but that doesn't mean that his game isn't heavily dependent on his awareness of himself, his opponents, his teammates, and the perspectives (self-of-others) of all of them.

I'm trying to say that there's a huge difference between being self-aware and being self-conscious. What I'm trying to figure out is where the lines are and what is necessary for what.

Matt Mullenix said...

Hi LabRat. Thanks. I just checked out the update.

"Humans are constantly awash in self-consciousness...Dogs, if self-aware, are highly unlikely to be similarly self-conscious- if for no other reason than being that wrapped up in yourself requires a great deal of abstract thinking and mental energy. We can make that assumption with confidence..."

Only Gregg and Russ (who know my dog) will nod at this, but as I was reading the above quote I was thinking of her.

If any dog I've known is both self-aware AND self-conscious, in the neurotic sense, it's her. She is a mess of conflicting impulses, displacement activities, high anxiety and a guilt complex. Poor thing is a basket case, raised mostly in a pen in the immediate aftermath of hurricane Katrina.

Gregg says "she a thinker," meaning, a worrier. It's a problem he sees occassionally in his dog training business and attributes to intelligent and sensitive animals that are, for exactly that reason, sometimes difficult to manage.

For such a wallflower, she is a wonderful and dynamic hunter, daring or delicate as needed at the moment. Altogether a different animal in the field.

I don't know how much abstract thinking Rina does. For that matter, I don't know how to quantify how much of it I do. In the field, my impression is that we tend to come to the same conclusions about things---the location of a bird after a chase, or the likelihood of a flush after a thorough search---and at the same time.

At home I know she watches me for clues to whether it's time to hunt (whether I change into my brush pants after work, or if I rumage through the back of the truck). Does a general conclusion (we're hunting!) drawn from specific evidence (he's got his boots on!) indicate abstract thought?

As for mental energy, she wastes more of that just worrying about whether she's supposed to sit on the couch than I think her skinny frame can sustain. With her eyes darting and her toungue flashing over her lip, she will pace back and forth until I sit down. I don't know where she gets the energy for that. It's exahusting just to watch her.

I realize none of this can be substantiated to everyone's satisfaction. But it's fun to think about. My feeling, restated, is that falconry and dog training works best for me when I assume my animals are seeing things the same way I do. This may well be a faulty or anthropomorphic assumption, but my experience suggests I keep on with it.

Mike Spies said...

Matt -- you wrote:

"The point is that humans teach dogs to hunt by simplifying the lessons, breaking them down into small parts and stringing them together. Wild dogs wouldn't have this luxury, because their prey behave unpredictably. Thus, they must generalize and reason their way through..."

In many years for developing, training and hunting bird dogs, I would dispute this comment. Because we do not TEACH dogs to hunt, instead we train them tp co-operate.

Bird dogs learn to hunt through exposure to game - the more frequently, the better. I call this development as opposed to 'training' because the BIRDS teach the dog, and our job is to quietly facilitate the process.

This is why genetics is so important - the dog must come from the womb with the tools to learn how to hunt.

I would say that, in bird dogs at least, the components are 50% genetics, 45% exposure, and about 5% actual training.

Matt Mullenix said...

Mike,

Actually I was trying to summarize LabRat's argument there.

My own point (perhaps not well expressed) was that I see the situation differently. My belief is that dogs consider things in about the same way we do--allowing for a different toolkit (in some ways inhanced, in others not)of sensory inputs.

But about that, as Darwin said, the difference would be one of degree, not of kind.

What goes on behind a dog's nose (and between her ears) I think is about the same thing that goes on behind and between ours.

The alternate theory---that all animals must by virtue of their unique physiology or phylogeny think in a radically different way than us---is difficult to prove. Who can say what is the internal experience of another creature?

The simplest assumption, the one requiring no claims to special knowledge, is that our mental processes are similar and that this accounts for the reason we understand each other. After all, all of us evolved here; we all share most of the same DNA; we are all contemporaries of one another. The similarities, to my mind, outnumber the differences.

Matt Mullenix said...

...also, I agree completely that dogs (and hawks) learn best and most through exposure. Their "training" is mostly about us and for our benefit. My philosophy with my animals is to train as little as possible and hunt as much as possible. I give them great credit for filling in the gaps on their own.

Anonymous said...

Dang, but I love discussions like this! It's a good thing for you guys all I can do is "hunt-and-peck" on a keyboard, and that my computer time at work(my only access) is strictly limited! Although, I realize the "sleepwalking" comparison to a coyote hunting was only for discussions' sake, I must say it is a bit of an awkward comparison if you've dealt with coyotes much, one of the cleverest, most adaptable, most "aware" critters out there! Sure, a certain amount of their hunting will be purely instinctual, but you can bet they vary that and innovate as necessary. Not to throw a monkey wrench into this "scientific" discussion, and I admit that it is definetely not a subject taken seriously by many, but I do wonder sometimes about the whole "mental telepathy" thing--I have seen and heard of others' experiences that at least make one wonder....I think anyone who trains animals has to ponder this at times. Perhaps there could be specific scents released with particular thoughts, and animals are going by this? I don't know, but it is purty eerie(and great!) when you are just THINKING about something, and your dog/cat/horse/goat/ferret/etc. just does it! I do apologize for bringing this up........L.B.

Matt Mullenix said...

L.B.: No apologies. The level of communication possible between dogs, hawks and people working at a shared task is "magical" in a sense, and prompts all sorts of possible explanations. You wouldn't be the first to wonder if somthing "extra sensory" might be involved.

And in an way, it must be so. If you believe the mind exists somewhere in between our sensory experience of the world and our reactions to it, then every individual floats on a sea of extra sensory perceptions.

Memory is an extra sensory perception.

Belief is extra sensory.

Insofar as my animals remember and believe things (I think it's clear they must), and so far as I share those memories and beliefs, our communication is informed by them. What results is just the kind of instantaneous conclusion that can seem like ESP or magic or any of the other supernatural explanations people have long attributed to our connection to animals.

I think it is not magic nor inexplicable. I just think we think alike.

mdmnm said...

Re: "mental telepathy"

I've talked to plenty of other hunters and myself have had the strange feeling of being watched while in the field ("bearanoia"). Sometimes you figure out what was giving you the hairy eyeball, sometimes you don't. Conversely, I've been able to make feeding or resting deer nervous by staring at them intently, particularly when I've been interested in trying to shoot one of them. Nowadays if I'm trying to sneak up on an animal, I try to make sure not to look at it directly or for too long unless I'm a few hundred yards away. I think this gets more important as you get closer. It doesn't always happen, but something goes on beyond sight, sound, or scent that both humans and animals can pick up on.

Anonymous said...

Indeed, there are enough hunter stories around about animals behaving radically diffrent during hunting season from the rest of the year, that it makes you wonder if they keep a calendar in the woods.....Another point I'd like to bring up(which I sorta have already)is the diffrent conclusions involving animal(other than human)intelligence/behaviour that serious scientists and us lay people come up with. And although non-professional animal owners CAN be ridiculous with some of their assumptions(as stated by smartdogs), so can the scientists! One blaring example occurred some years back and was a National headline in papers and magazines, so some of you may have also seen it and shook your heads sadly, as I did, at the incredible stupidity of a group of scientists' conclusions that were "ground-breaking" in the annals of animal behaviour, and for which they wanted due credit, on the experiments they'd done on discovering the why and wherefore of Dogs Barking. Their conclusion was that it was simply an inane release mechanism with no real meaning or purpose! And it had no communication value even with other dogs! ANYONE with even the most basic experience with dogs recognizes happy barks, lonely barks, angry barks, terrified barks, and can usually deduce who or what is around from the infinite variations of their dog's barking! I wondered HOW these scientists could have come to such an idiotic conclusion, until I found out the experiment was conducted on caged beagles kept in a laboratory! Yeah, I bet those wretched little beagles WERE barking rather inanely and pointlessly!!!!...L.B.

Matt Mullenix said...

LB: when scientific conclusions border on the absurd I take it for granted that the conclusions were drawn in a vacuum. The need, logical and inherent in the sciences, to reduce variables and biases, leads to some laughably reductionist thinking.

But science is not to be blamed for this. Science needs reductionism and controlled variables in order to provide anything of use to us.

Science becomes ridiculous when it attempts to atomize phenomena that are irreducible. The mind may be one such thing, as it is contigengent on so many other things----How can a mind exist outside its body? How does the body exist outside its environment?

I say it can't, which makes the mind's entire context relevant and necessary. That context, when one looks fairly at it, becomes the Universe.

We have a tradition for understanding universal and irreducible phenomena, but that tradition's primary tools are faith and humility.

Science, in contrast, cannot operate on faith. As for humility, we require too much of science now to allow it to be humble. It is charged with solving all our problems, especially the ones it caused.

Rebecca K. O'Connor said...

Although asked to chime in...there are a multitude of great discussions here, so I'm not really sure how I want to chime...

As far as choosing to function on the belief that our animals see the way we do and their self awareness...I love to imagine that and sometimes interact with that in mind, but when called to address behavior or forced to adjust behavior of my own charges, I have to go back to what I can substantiate. I am constantly reminding myself when working with animals to eliminate constructs, emotions, preconceived conclusions and simply look at the behavior, because behavior is pure truth as long as you eliminate any internal speculation. It doesn't make a difference HOW they experience, it matters that they DO experience.

A behavior that repeats has been reinforced. I can't tell you why an animal does something the first time, but I can likely deduct why the animal (person) does it repetitively. There is always an antecedent and a consequence to the behavior. If I want to manage the behavior, change it or eliminate it, I must either change the trigger or change whether or not it's rewarding.

Successful hunting is a learned behavior no matter whether you are wild or domestic, no matter if you use talons or a shotgun. Applied behavior analysis works no matter the species, no matter the behavior. You try something... and likely your genetics lead to a propensity to try a certain range of behaviors...one of those behaviors works and so you do it again.

Arguing that learning demonstrates self-awareness doesn't jive with me. However, the ability to quickly absorb lessons based on minimal reinforcement and exhibit creative behaviors to get reinforcement isn't about being self-aware. Does environmental interactivity really make a being self-aware?

I would further argue that if you are going to argue that learning in animals demonstrates self-awareness that superstitious behaviors are indicative of a lack of self-awarenesses. The variety of superstitious behaviors I have seen in shows and in falconry that are linked to accidental or incidental reinforcement are amazing. Self-awareness might allow the animal to overcome this superstitious behavior, recognizing it as such. Animals don't seem to ask themselves...why would this be linked to that...the way a child would ask as she becomes more self-aware, "why would holding this blanket really stop me from having nightmares?" --of course humans do engage in superstitious behaviors, but have the ability to reason their way out of them.

That said -- I think there is a whole lot more going on with animals than we usually guess. Probably self-aware maybe even self-conscious. I just don't agree with this as a jumping off place for the argument.

I'll tell you what though...I'm pretty sure I would much rather be gifted with the ability know what my dog or hawk was thinking that just about any human being. :-)

Rebecca K. O'Connor said...
This comment has been removed by the author.
LabRat said...

My argument was not so much that being environmentally interactive demonstrates the possibility of self-awareness- I think that's demonstrated much more persuasively by things like this experiment- but rather an evolutionary argument for self-awareness as such a useful tool for a complex social animal that its existence is likely rather than unlikely.

I brought up hunting because dogs, like humans, hunt in groups- and this is one of many complex group behaviors in which self-awareness in its most basic sense, which is the recognition that you are an individual with a unique perspective and so are the other members of your group, who have different perspectives, would vastly simplify learning. Things that vastly simplify complex behavior tend to be innovated multiple times independently in evolution.

In the case of a dog hunting with a hawk and a human, the dog can adapt to new situations, new hawks, new prey, and new humans much faster if it realizes basic things like the hawk being able to see much further and better than it, the human being similarly gifted but also being utterly noseblind. This, I argue, suggests some basic sense of self.

(Forgive me if I'm ignoring anyone or oversimplifying, I've got one king-hell headache...)

Rebecca K. O'Connor said...

I would still argue that is learning not self awareness.

Antecedent: Dogs smells a grouse
Behavior: Dog wags tail
Consequence: the human walks past the grouse.

Okay that didn't work.

or

Antecedent: Dog smells grouse
Behavior: Dog points
Consequence: Human praises dog, sets up shot (and hopefully hits the grouse which adds more reinforcement)

That wouldn't require self awareness, but reinforcement. The dog doesn't know he can smell better than the human, he just figures out the reinforcing behavior.

Even in a pack, the dog doesn't necessarily know he has a different perpective, just that his behavior is rewarded or punished. Wouldn't he behave accordingly regardless of self-awareness?

Matt Mullenix said...

Rebecca thanks much for adding to this discussion. This entire thread has been fabulous and very helpful to me as a trainer.

You write: "...when called to address behavior or forced to adjust behavior of my own charges, I have to go back to what I can substantiate. I am constantly reminding myself when working with animals to eliminate constructs, emotions, preconceived conclusions and simply look at the behavior, because behavior is pure truth as long as you eliminate any internal speculation."

That's the view I hoped you would articulate. It's the crux of behaviorism, and although I have great respect for it, I've come to realize it differs from my own belief. I am unable to eliminate my emotions nor all my preconceptions; some are quite resilient to change. I do evaluate constantly, and react accordingly to behavior as I see it. But on top of that I pile the evidence of my emotions and preconceptions.

I can't help myself. My animals affect me, and I respond to that. Mainly I fear that by refusing to respond, I am denying them one avenue of communication.

My brother, a physician, once told me that part of his training is to be sensitive to his own emotional reactions to his patients---if they make him mad, for example, it might be an important clue to a specific diagnosis. That factoid resonated with me. I realized I do the same thing with my animals.

I cannot reduce the experience of working with my team to its constituent parts or behaviors. When trying to do that I run into something like the Uncertainty Principle, where tight concentration on one aspect leaves others slipping away from me.

I am better off with the gestalt. I try to take it all in, including my own reactions and preconceptions (which I trust are not irrational but based on some acquired wisdom). I try not to dismiss anything as relevant.

Obviously, both approaches can work. I've seen your team in action. :-) You must be doing something right!

I am at heart just a lumper, not a splitter.

Mike Spies said...

Dogs have a 'sense of self' - that is, they realize that they are not 'all dogs', but an individual dog.

In a group (pack) they learn how each of the other dogs vary in reaction to their behavior. In even a small group this becomes a fairly complex set of information.

They seek to gain advantages over the other dogs in the group. Individuals display this behavior to varying degrees and are sensitive to the differences between other packmembers' reactions to this behavior.

They deliberately draw attention to themselves.

They may tease other dogs or play 'canine jokes' on their packmates.

Sound at all familiar? People and dogs share a lot of common behavior traits. At core dogs and people behaviors are not that different.

This opinion is not based on any 'scientific' experiment, but on observation during years of living with dogs 7 X 24.

John Scanlon FCD said...

Rebecca said
"The dog doesn't know he can smell better than the human, he just figures out the reinforcing behavior.
"Even in a pack, the dog doesn't necessarily know he has a different perpective, just that his behavior is rewarded or punished. Wouldn't he behave accordingly regardless of self-awareness?"

Saying that the dog 'figures out the reinforcing behavior' or 'knows that his behaviour is rewarded or punished' strikes me as the wrong 'level' of description. Conditioning is not a product of 'knowing' and 'figuring out' in the sense of conscious cognition, but occurs in sea slugs and people with total short-term memory loss or in permanent vegetative states. Human minds are very diverse, and your mind can vary a lot from one day (or minute) to another, so for some of us, some of the time, it is relatively easy to understand how some animals think. Talking about it, however, is harder.

Eric Edwards said...

Matt wrote: “I think we clearly overestimate our training ability as humans---our animals are not nearly so well trained as they are simply correct in anticipating our behavior (provided we are consistent). Training a hawk is really just another way of training oneself to be consistent. Most animal trainers would agree with that, I think.”

It’s been a while since I’ve trained animals or trainers professionally but during that time I trained more than a few of each. The best trainers are the ones that keep it simple, that think of training in simple terms and understand that animals (and people) are basically motivated by one of two things, gain pleasure or avoid pain. Everything we do (and that our animals do) in our daily lives revolve around these two motivators in varying degrees.

I completely agree that consistency is one of the keys to training along with clear communication, the trick is to get the animal to understand what you’re asking of it and to reinforce that behavior. Do it enough times and you have a trained behavior.

I’m not sure we overestimate our training ability, if anything the opposite might be true, most of us don’t live up to our abilities as trainers. Personally, I think I have the ability to train my birds and dogs to a much greater degree but I’m not looking for a hawk, or dog, that performs like a machine, I like a little unpredictability in my falconry. In my shows on the other hand I didn’t want any although I usually got my share.

Rebecca K. O'Connor said...

Matt--

Do I anthropomorphize? Do I make training decisions based on what I "think" might be going on in my animals head? Do I speculate about their thoughts and desires? You betcha. My falconry isn't my job and I can't avoid the fact that I'm human and that's my lens. And it's supposed to be fun, damnit. And whenever this gets brought up, you always make it sound like being emotionless is something I constantly do. That is highly improbable, captain.

Like your brother the doctor I listen to my emotions when they act up but then go back to the science. If you brother feels anger toward a patient and the need to analyze why, I somehow doubt he guesses what his patient is thinking and feeling and then uses it to diagnose him. He asks (which your not going get from you animals), he looks at the chart and he does diagnostic tests. In short he goes back to the science, which is exactly what I do.

I'm wrong a lot when I anthropomorphize and guess at my animals feelings and thoughts. If life, loss or a paycheck are on the line, you can bet I'm going to constructionalize the behavior and sort it out in a more scientific manner. For the hundred thousandth time, do animals have emotions, desires, neurosis, emotional history? of course!! Can I quantify it and utilize it in communicating through training with animals? Its a crap shoot. And there are times when a crap shoot isn't good enough.


And John, you are absolutely right! I should have said the dog "reacts" to the positive reinforcement. I was trying to avoid putting it that way, because I hate it when people assume I think animals are stimulus/response machines. I know there's a lot more going on in there than reactions. However, the way I stated it was absolutely incorrect.

Matt Mullenix said...

Rebecca I just needed you to be a redheaded straw-man so I could retreat into my romanticism and razz the heartless scientific world.

I admit it.

Anthropomorphism is inevitable. Your argument cautioning against it, from our past conversations, stems from folks simply getting it wrong. In other words, assuming their animals have emotions and opinions is not their problem---they've just made the wrong assumptions or at the wrong time.

I think your job as a trainer has been as much to correct human misconceptions as animals' misbehaviors. Right?

The furthest I get toward scientific treatment of my training is taking notes. I take fewer than I used to, but I develop some kind of working database on all my animals. This helps me correct for my own failures and inconsistencies. But once in the field, I rely solely on immediate input, or as close to that as I can manage.

There the assumptions of similar thinking among my animals serves me pretty well. At least, we get the job done and avert disaster most of the time. If we are making errors in our assumption of each others' bahavior and motivation, they must not be mistakes that matter much in the field.

I know you do much the same as I do in the hunt. After hours post-processing with blog posts or beer (or both) is another sport altogether!

Rebecca K. O'Connor said...

But do you really think the animals are trying to figure out your thoughts and motivations? Or are they just responding to their environment in a way that promises success -- as Eric put it, gaining pleasure or avoiding discomfort. And are they just reading your behavior? Maybe you all do well in the field because at least one side of the team isn't trying to emote and ponder. :-g

I tell clients with parrots that they are really missing out by training their parrots as if they think and emote like them. Maybe they do, but when you constantly work on this assumption, there is no opportunity to learn from your animal's differences and strengths. Think of all the things you could learn from a dog, parrot, hawk's ability to read a human companion.

I know my parrots break my behavior down and recognize the tiniest signs. My grey says goodbye the moment I think about leaving, even if I haven't picked up keys or grabbed my shoes or something else obvious. It could be the way I furrow my brows, hold me shoulders, run my fingers through my hair...I have no idea, but the parrot sure does. And he's not guessing what I'm thinking. He knows that that is a precusor to the next behavior - picking up my keys - which leads to the next behavior - walking out the door. What an amazing little human behaviorist. Think of all the diabolical uses of that level of reading body language. I want to be him when I grow up! But I can't learn how if I don't break down what he does...

LabRat said...

I think we're now talking about two different things- the fundamental question I was asking is if whether a basic sense of self-awareness in dogs is a logical one. The one you seem to be answering is whether or not it's a useful one- and as the thread of conversation is proving, those questions may very well have two different answers.

The study in the article I linked to features dogs not only learning by observation- in other words, learning how to solve a problem with no reinforcement whatsoever until they try it for themselves- but changing the way they solved the problem based on what they observed about the dog they saw. The dog they saw was trained to solve the problem by manipulating something with her paw. If they saw her simply manipulating it with her paw, they copied her in that fashion. If they saw her manipulating it with her paw and a ball in her mouth, they manipulated the object with their mouth- suggesting they were inferring something about her point of view just from the observation.

You are absolutely correct that a strict behavioral standpoint is extremely useful in training and in preventing yourself from gumming up the works with all sorts of human baggage, but I still maintain that a total lack of self-awareness is not the most logical assumption to make when simply considering dog minds rather than actively working to shape them.

Matt Mullenix said...

Rebecca wrote: "But do you really think the animals are trying to figure out your thoughts and motivations? Or are they just responding to their environment in a way that promises success -- as Eric put it, gaining pleasure or avoiding discomfort."

That's a good question, and the answer is "probably not." Eric's point is well taken; animals are pretty self-centered, most of the time, just like us.

My feeling is that rather than trying to puzzle out my thoughts, my animals probably assume they already know whatever they need to know about me and the other team members. They've learned this from observation alone.

I've decided I generally make the same assumption about them.

Here's a concrete example, although not one I claim to be airtight: Ernie leans forward and mantles over his kills when I am closest to him, but he stands tall and bats his wings or charges at Rina when she is nearest.

He seems to think he can bully her away (he's right), but with me he seems to think there's not much point in arguing. Right again!

Now I can't know what's going on in his head, but his behavior is consistent and perfectly logical. It's the same conclusion I'd probably come to if our positions where switched.

It's this sort of thing, repeated in dozens of different variations between both the animals, that brings me to my working assumption that they must think pretty much the way I do about things. Again, that may be totally erroneous or even silly, but we rarely misunderstand each other.

Matt Mullenix said...

Labrat notes: "the fundamental question I was asking is if whether a basic sense of self-awareness in dogs is a logical one. The one you seem to be answering is whether or not it's a useful one- and as the thread of conversation is proving, those questions may very well have two different answers."

I would say that dog (and other animal) self-awareness is certainly logical, if indeed it is logical that we have it. How else, unless we are talking about supernatural/divine causes, could it be that humans are fundamentally different in this one way, when we are so demonstrably similar in every other way?

That's the basis of my view about them: the simplest explanation is that our mental faculties are more or less as similar to animals' as are our physical faculties---which are different mainly in degree and not kind.

Every animal is pretty closely related to every other one, if your frame of reference is the Universe.

Rebecca K. O'Connor said...

LabRat-- Matt drug me off the topic. It wasn't my fault, I swear. I was trying to argue that although I believe animals are self aware, I do not believe that the way they learn is an indicator of this.

There are lots of examples of modeling behavior in other species though, one of the most famous being the tits in the UK that learned rip off the top of milk bottles to drink the cream by following the examples of other birds. I really don't think a tit is anywhere near as self aware as a dog and they are just as capable of modeling behavior.

I would also argue that there absolutely WAS reinforcement involved in what the dog experiment. A behavior won't repeat if it isn't reinforced, period. Reinforcement can be a very minimal stimulation though. You flip a light switch wanting light and the light goes on, that's reinforcement. It doesn't have to be food and overt praise. It may just be a fun behavior to do-- that's still reinforcing. There's plenty of things that are self-reinforcing. Or the humans in the experiment might have exhibited signs of pleasure -- to a dog, a smile can be reinforcing.

So I just don't buy that the ability to learn behaviors creatively equals self awareness.

LabRat said...

There are lots of examples of modeling behavior in other species though, one of the most famous being the tits in the UK that learned rip off the top of milk bottles to drink the cream by following the examples of other birds. I really don't think a tit is anywhere near as self aware as a dog and they are just as capable of modeling behavior.

Are you sure about that? Magpies, another passerine, have passed the mirror test, and tits actually have some pretty complex social behaviors of their own.

I guess my bottom-line question is this; if you don't think modeling behavior is a strong indicator of self-awareness (let alone changing the way they model based on an apparent guess about state of mind, as the dogs in the one study did), then what do you think this is based on, other than a recognition that since another member of your group is an individual like yourself, therefore copying them to solve the same problem is likely to work?

I've been arguing that self-awareness is the likeliest root of these complex social behaviors because it's an elegant solution to a number of problems at once. That means it should be like other widespread examples of convergent evolution, and multiply innovated across multiple species and families that have to solve the same kinds of problems.

If you don't think it is, then what is? Right now you're just arguing for a null hypothesis, but what's a likelier answer?

I would also argue that there absolutely WAS reinforcement involved in what the dog experiment.

I didn't mean to imply that there wasn't any reinforcement at all, but rather that reinforcement had nothing to do with the dog's initial decision to try the same solution it saw another dog use, unless you're arguing that the dog found the sight itself reinforcing, which doesn't exactly argue against self-awareness not being involved.

Matt Mullenix said...

Labrat writes: "I've been arguing that self-awareness is the likeliest root of these complex social behaviors because it's an elegant solution to a number of problems at once. That means it should be like other widespread examples of convergent evolution, and multiply innovated across multiple species and families that have to solve the same kinds of problems."

My question is why should it even be THAT complicated? Why covergent and not basal to all who have it today? Why not say self-awareness (or for that matter, Mind) simply is what it is, and is common to many animals?

Maybe the mind came first, and the form followed? To me, that's the more conservative and more elegant theory.