I’m working on my Intro for APDT’s symposium on Canine Cognition, and one of the hot topics right now is why dogs seem to be better able to interpret a pointing gesture than are wolves or chimps. Various research projects (see below) have shown that dogs go directly toward food hidden under one of two objects (both scented with food) if a person points toward it, while wolves and chimps do not. Some have speculated that at least 10,000 years of co-evolution (probably more) have resulted in a genetically-mediated ability of dogs to inherently understand human communicative gestures.
I’ve always wondered about the results of these studies, because in my experience, you have to teach dogs to look in the direction you are pointing. When they are young, it seems to me that they pay no attention to the direction you are indicating, and simply go to your hand and sniff it. Granted, it is pretty easy to teach them to follow a point, but you DO have to teach them to do it. That’s not the conclusion of some of the researchers, and the results of these studies have been cited a gazillion times as evidence that our relationship with dogs has created a natural selection process that has made dogs able to communicate with humans more effectively than other species.
In addition, researchers (Brian Hare, Michael Tomasello and others) have compared the responses of adult dogs and “puppies” with that of chimpanzees and wolves. The conclusions have been that even young puppies are able to follow the direction of a pointing arm, while chimps or wolves of any age are not (thus, there must be a genetic component to the behavior).
I’ve seen some of the videos of the research, and I have to tell you I think the issue is a bit more complicated. First off, the young wolves, even though they have been reared by humans, appeared to be squiggly and fidgety and not able to focus on anything. Just trying to hold them in position seemed to be almost impossible. I’ve worked with a few high percentage wolf-dogs, and I can tell you that the young ones are like ADHD dogs and can barely stay still for a second. The ones in the video I watched looked frightened and completely unable to focus. A good friend and colleague tells me that adult wolves appear to be much more focused on humans and their gestures than the young ones, so perhaps there is a developmental component to the behavior.
Secondly, you’ll see that the video below shows chimpanzees, adult dogs and “puppies” being tested.The research does seem clear that as smart as they are, chimps really don’t ‘get’ what a pointed arm and finger mean. But you’ll note in the video below that the adult dogs did very well, while the comment about the pups is “Even six-month old puppies catch on pretty quickly…”. That suggests to me that there was learning involved, and that the behavior is not 100% innate, which fits with my experience with dogs. I’ll look up the actual data on “puppies” (shouldn’t it be ‘juveniles?’) before I leave for Atlanta. Brian Hare will be there too, and I very much look forward to meeting and talking with him.
Perhaps a good explanation at this point on the issue is that there appears to be an innate tendency in dogs to be predisposed to learn to follow a pointing gesture. This is basically a nature AND nurture argument: that dogs are indeed different than other species because of their co-evolution with us, but that the issue is more complex than a simple innate ability to understand what a point means. Their responsiveness might also relate to an ability to view to humans as cooperative beings who are ‘on their side’ . . . Note that chimps tend to be extremely competitive over food, and perhaps that might be a factor in their inability to understand that a person is trying to tell them where the food is. Brian Hare is speaking Friday about differences in emotional reactivity in different species, sounds fascinating. He’s one of the few researchers who has been doing work on domestic dogs for years, and it’s great that he’ll be at the conference.
I could go on, but I’ll just raise this issue now, and take it back up after hearing the talks of Clive Wynne and Monique Udell, both of whom have been doing their own research on this topic of pointing gestures and how they are interpreted. If I understand it correctly, they too have been somewhat skeptical of some of the conclusions drawn by other researchers.
Here’s a video from YouTube of some of the work . . .[Note that the researcher in this clip is actually presenting 3 signals (at least): Head Turn, Eye Gaze and Arm/Finger point. That is by design; the research worked on sorting out dog’s responses to all 3, and found that dogs will target the container with food even if the experimenter keeps her head still and just moves her eyes toward it. Added 10/13]
Alexandra says
Trish, this is such a cool post (and video)! I agree with you – in my limited experience dogs do have to be taught to follow a point but they will learn easily enough. Do you know if any similar research been done on bonobos? I have read that studies have shown that bonobos are much better cooperators than chimps, particularly around food (as described in Vanessa Woods’s book, Bonobo Handshake). I wonder if there would be a difference in the pointing study?
Tracy says
Perhaps a willingness to label dogs’ ability to understand pointing as “innate” is related to our willingness as a species to expect dogs to just “get” what we’re trying to tell them (and if they don’t, they’re stubborn or mad at us or trying to be alpha, etc). Dogs do seem to be pretty uniquely attuned to humans, and if you start by assuming they innately “get” us, then when dogs don’t do what we ask, it’s not our fault.
Hm. That sounded much more negative than I intended. It just seems to me that the researchers may be coming in with the same preconception as many dog owners, and perhaps it is this preconception that causes so much dog-human miscommunication.
Kat says
“Perhaps a good explanation at this point on the issue is that there appears to be an innate tendency in dogs to be predisposed to learn to follow a pointing gesture. This is basically a nature AND nurture argument: that dogs are indeed different than other species because of their co-evolution with us, but that the issue is more complex than a simple innate ability to understand what a point means. ”
You’ve captured exactly what I think is happening. It is so easy for dogs to learn what a pointing finger means that it looks like they’re born knowing. Are there any studies showing whether human babies have an innate understanding or do they simply learn almost instantly as well?
mungobrick says
Purely anecdotal…but Daisy seems to be an extremely intelligent dog and at 20 months has not been taught what a pointing gesture means. She NEVER looks where I point, just at my finger. If I say “where’s…?” (where’s Nick (my husband), where’s Angus (her boxer pal), where’s your bunny), she’ll look around and knows what she’s looking for, but pointing is meaningless for her.
hornblower says
As a sidenote to pointing:
As we all know, pointers & setters & many other sporting breeds point with their noses. Many sporting & hunting dog owners have suggested that it works better to point with ournoses when we want a dog to look at something. I know of at least one person who reported on a gundog email list trying this first with a pointy “nose extender” (made out of paper IIRC) attached to his (the human’s) face to really make it clear to the dog that ‘yes, I’m pointing at that & I want you to look at it’….. LOL
But I think that probably a nose extender is not really necessary.
I think that rather than using our hands, our faces may be more communicative to dogs. I wouldn’t be surprised if dogs learned to look at where we point with our fingers/hands BECAUSE we tend to demonstrate it by using our heads at the same time. Eventually, the cue for the behaviour will be just the hand, but it was because they followed our ‘nose’ to where we were looking that they learned the behaviour.
Alexandra W says
If my dog was born with an inherent ability to understand pointing gestures, I’ve yet to see it. I know, I know, an anecdote doesn’t contradict peer-reviewed science, but…
I’ve been trying to teach my beagle to play a “find it!” game. I’m no whiz trainer, so I’ve been muddling along by taking bite-sized pieces of stake and placing them at some distance from him, where he could see them, and then releasing him to go get it, and eventually working up to putting the steak around a corner. And I’d release him, and he’d look for it (not sniff… I have the world’s worst beagle!) and out of some silly human impulse to help, I’d point at it.
And he’d walk over and start sniffing and licking at my hand, which… smelled like steak.
Actually, I don’t know what that story indicates, other than that my dog and I aren’t very good at the “find it!” game yet.
Jane says
I agree with Hornblower. I think dogs initially learn to follow a pointing finger because our faces and eyes key them to. Later on, a pointing finger is enough by itself. In the video above the tester is always turning her face and eyes in the direction of the point.
Catherine says
My experience echoes the “predisposition to learn what pointing means” explanation. My redbone coonhound (adopted at age 3) didn’t know what pointing meant when I got him, but usually follows a point now. I believe he learned it by playing Find It, because I often tossed a piece of kibble and he would visually follow the motion of my hand. If I was tossing in a wide area (e.g. the whole yard vs. just the few feet in front of us), and he didn’t see the toss motion, he would sometimes look at me, and if I then pointed, he would follow the point the same way he would have followed the tossing motion. This seemed to be a learned behavior, but one that he learned through observation rather than deliberate training. So I think it was the motion of the hand, rather than the final position of the hand or pointing shape of the hand, that signified “treat is over there.” My impression is that he does not yet fully generalize this to a non-moving, pointing hand or finger. I’ll have to test & observe this more closely.
In the video of the dog shell game, the person pointing is also gesturing with head and eyes. The chocolate lab puppy appears to look up to the human’s face (2:16) before deciding which way to go. Although the next segment shows isolating the eyes, it would have been interesting to isolate the pointing gesture, and have the person’s face completely hidden. As is, this seems to support the hypothesis that dogs are predisposed to read body language, especially eyes, but not that they are predisposed to understand pointing.
Amy says
I’m not a scientist, dog trainer, or possessive of any other remotely qualified title to really speak on this post BUT I do have a spirited Pembroke welsh corgi and have read your book The Other End of the Leash. I often think of a story you told in your book about these types of gestures and dogs. You said that dogs are more aware of our body language than we are (which for me is totally true) and even YOU get tripped up from time to time. (This has been a comforting thought on more than one occasion) For instance, one of your dogs DIDN
Roberta Beach says
Oh, Alexandra, your Beagle sounds hilarious! I found it interesting that the chimp guy was so shocked a DOG would be able to follow a point at all much less way better than the chimp. I also liked the demonstration of just using eyes. I do CGC pretty consistently with varying rescue dogs. One thing my instructor emphasizes is getting the dog to look at you, always be looking at you….but it definitely is a learned response….unless I do have steak on my nose :). I have mostly hounds, too. I was contacted by Beagle Brigade about Shiloh once. He was like your Beagle, A. – not food motivated but young. The others I have are food motivated but too old for the program….but I am still on the look out for the perfect young one.
Caroline Smith says
The treat test to me is inconclusive. It seems to me that disobeying a command while a persons eyes are closed is a learned behavior. I’ll only look at my dog through the corners of my eyes (my head facing forward) and correct him if he goes for the treat, but I wouldn’t correct him if my eyes are closed since I can’t see him. So, he learns that eyes open means that I’m paying attention (no matter which where my head is facing), but eyes closed he can get the treat and I won’t do anything about it.
Cindy M says
People with autism do not inherently follow a pointing hand, arm or finger, either. In human development, this is known as a 3-point gaze shift. Autistic children don’t do it naturally either. I think dogs are just looking for any kind of movement or cue from the people, especially if there is a history of some reinforcer if the hand throws food.
Kevin A says
I have two shelties that have seemed to easily understand hand gestures and look to me in many situations that only require me to direct them with a point or hand gesture alone. I never practiced or “trained” them, I just started doing it from the time they were young and they just easily caught on. They are two now and will look for my hand signals to do many things. They also seem to really be watching me intently when I am wearing sun glasses so I make it a point to raise my glasses when I am talking to them so they can see my eyes. Just a lay persons observations.
CharlieDog says
Hey! You’re coming to Atlanta?? Is the public going to be allowed in? I’d love to come here you speak. Are dogs going to be allowed too?
jackie d says
Have owned three dogs two of them as a kid. I don’t think the beagle ever learned what pointing meant, but I can’t remember trying to teach her. Both the other two dogs picked up what pointing meant coincidentally to learning other commands: the labX generalised from anticipating which way a thrown ball would go. My current colliexspaniel generalised from me using a pointed finger as the hand signal for ‘go to bed’ – because I pointed in the direction of the bed regardless of where we standing.
John says
The test set up is very basic and allows for a ton of other signals to influence where the dog is going. Just the fact that the human/woman is also moving her head and looking at the cup skews the results.
In my experience, as Patricia pointed out, my dogs are more likely to come and sniff or lick your hand when you are pointing somewhere, unless you trained them or at least unknowingly conditioned them to follow your gestures. Again this gesture might involve you pointing AND turning your head AND looking towards the goal.
Beckmann says
I have seen this entire documentary and I was quite fascinated by it. My first impression on the chimps
LorieK says
I’d be careful about using that comparison between young wolf-dog pups and ADHD dogs on speed (“the young ones are like ADHD dogs on speed”). There’s a reason that stimulants are used in the treatment of ADHD — it actually enhances a person’s ability to focus. As the mom of an ADHD child, I can tell you that sedatives would be more inclined to produce a totally wired and unfocused result.
Steve says
I believe it was Stanley Cohen who observed that his dogs immediately understood pointing when he put a conical party hat on as as nose/muzzle on himself and used it to point like dogs point with their nose. He was experimenting with pointing with his built in nose/face direction and the dogs didn’t get the direction reliably until he enhanced his profile with the party hat. I would suspect that the learning curve on dogs figuring out pointing with hands is simply figuring out that our hand corresponds to dog pointing with their nose. Similar to dogs having to translate our body language as Turid Rugaas found in using calming signals.
trisha says
To John: You are absolutely right about the fact that there are other signals being given besides the pointing (eye gaze, head turn) but it’s my fault I didn’t mention that, not the researcher’s. She is well aware of all those signals, and the work includes sorting each of those signals separately (it seems that dogs respond even to eyes shifted to the left or right without the head turning or an arm movement). The clip I included just happens to show her doing all 3 at once… again, my fault for not mentioning that. I’m going to amend the post to mention that, thanks for your good comment.
And to LorieK: Oh heavens, was my brain turned off! Lorie is absolutely right about stimulants acting to focus those with ADHD; I’ve had 3 or 4 clients over the years with dogs who responded to Ritalin for example by becoming focused and attentive rather than unable to concentrate for even a second at a time.
Jade says
I would love to see if there’s a difference between when the handler is looking where she is pointing or if merely point alone would work? I think the eye contact with the cup makes all the difference.
Or is it possible that dogs have a better sense of smell than chimps and wolves? I’m asking this because I don’t know.
John says
Trisha, I guess overall the video or experiment still shows that dogs have a unique (or rare) ability to discern subtle difference in gestures/facial expressions.
The next, task, ask you pointed out is to find out how much of this is nature and how much nurture (starting from puppyhood).
Again, as you pointed out, this might well be (with so many things, if not all behaviors, human or canine…) a combination of nature AND nurture.
Jeff Line says
I have trouble seeing what is so profound about dog’s reaction to pointing. It seems to me that as a result of the generations involved in domestication, often occurring in a working together context, dogs are inclined (avoiding nature/nurture argument here) to look to humans for resources. Whether its a hand command to sit, a point or my favorite, raising your hand to cue a dog to go down, they tend to watch what we are doing. I would even argue that the most disconnected dog remains acutely aware of what the human it is apparently disconnected from is doing. Other species have had no reason to develop that skill.
I found the blindfolded demonstration to be fraught with confounds. What if the dog has found young females to be most likely to give treats in the past or the dog doesn’t like men? There were too many variables to actually serve as a demonstration or test of anything.
Dena (Izzee's Mom) says
The “Signal Retrieve” in obedience is an exercise where the dog is to go fetch one of three gloves, the one the handler indicates with a hand signal (basically pointing with the whole hand). I was taught to begin training this exercise by filling a glove with treats, tossing it just in front of the dog, and feeding them treats out of the bag when they bring it back to me. (The glove is velcroed shut, so they need me to get the treats out for them.) This gradually becomes a hand signal toward a glove that is already out there. But I have seen many dogs take a long time to “get” this notion without the motion of the glove. Some, of course, catch on much more quickly, and some trainers are better than others at working a dog through the necessary steps.
Beth says
I have never had a dog who inherently understood a point, and in fact one of the things that tends to re-shock whenever a new puppy comes home is how little we are able to communicate with the tiny ones. Calling names brings no response, pointing, waving. Hand-clapping seems to universally get puppies’ attentions (and adult dogs love it too). But pointing? Has to be learned. (Most) dogs learn it quickly, true, but I’ve yet to see one who understood it right out of the gate.
Lynn says
I agree that dogs may have an inherent predisposition to learning what “pointing” means – or what other hand gestures mean, for that matter. Pointing is perhaps a bit more abstract than “sit”, and in my experience it’s been a heckuva lot more difficult for any of my dogs to master.
I’ve been trying to teach my 1.5-year-old Border collie what I mean by pointing – “there, that’s where your stick (or ball) went, go get it”. It certainly wasn’t hardwired in him, and I’m still not certain that he “gets it”. And it’s not that he’s dim – in Monday’s obedience class, we taught him to “put his toys away” in ten minutes flat. On the other hand, on many occasions I watch him, clearly frustrated that we dumb humans don’t “get it” when he’s so clearly pointing for us – with his nose. Sitting outdoors at my in-laws, where a football is on the middle of the table by the pool, and the dog stares, first at me, then at my husband, then at the ball, then at my husband… and finally gives up and switches alternating his gaze between me and then the ball. Or if I go outdoors with him, and he’s trotting by my side, only to suddenly come still – and stare at the corner of the house, then at me, then at the house, back and forth. Of course! It’s the lacrosse stick with the ball in its cradle, propped up against the house. It’s just that many times we humans are too dumb to “get” the point and act on the request.
If dogs could follow a human gesture so easily (either inherently or through training), then why don’t the handlers point with their hands at “double lift” sheepdog trials? This is where the dog is sent out to gather one group of sheep, and the handler will typically make sure the dog sees the sheep before sending them out. The dog brings them through the first fetch panel, and then is instructed by the handler to leave the group and to go gather a second group which, unbeknownst to the dog, has been released in a corner of the field. It’s my understanding (Trisha, correct me if I’m wrong!) that the handler usually signals to the dog where to start to look by whistling to the dog to “flank” the first group of sheep either in a clockwise or in a counterclockwise direction to get it started moving in the right way, before whistling to it to “look back”. They do NOT whistle to get the dog’s attention and then point off in the direction they want the dog to take. Even at a distance of 300 yards, the dog ought to be able to discriminate between a handler’s pointing to the far right corner of the field vs pointing to the far left corner. Anyway, I find it hard to believe that if pointing to dogs to indicate directions worked so well, that this wouldn’t have become an essential tool in working sheep.
Michele says
Others have also made the predisposition to learn argument (Pam Reid’s review paper in Behavioural Processes; Juliane Kaminski’s talk at Canine Science Forum in Vienna this year). I tend to agree with it, as well.
While a healthy dose of skepticism is always in good order, I am also skeptical of those who have been highly critical of work that has been replicated in labs around the world. Conflicting results may be due to differences in methods. Some have suggested (boldly, in my opinion) that their results negate others’ results. I’m not sure how helpful the public back-and-forth has been to the field of canine cognition and behavior, as most of us, whether researchers or practitioners, are interested in more than pointing and in examining the same questions with other methods.
Roberta from Vancouver Island says
I found it interesting in the video that the primates are behind glass, separated from the people, while the dog was out and about in the room with the experimenter. Does this suggest anything to anyone else? That perhaps the primates have a much more controlled way of interacting with people, while dogs live us all the time, and are constantly watching and learning from us?
Rita & Guro says
In my opinion the dog is much smarter than the monkey. The dog is also better at paying attention to the human. I love how dogs can really concentrate. My dog has been taught that pointing involves me trying to help or show her something. She also understands many words and word combinations, like her name + give. She can be in a deep sleep and jump up at the words: What shall we GIVE THE DOG for dinner today? and act like Where is it? What am I getting??
Ellen Pepin says
Dogs are very good at interpreting human body language. They get all sorts of clues about another dog by looking them in the eye, so following human eye movements does not seem much of a stretch to me. My two dogs follow certain commands, such as “stay, better when I use a voice command and a hand signal together. I am teaching my collie to obey the command “come” with the use of voice and a hand signal. When she is in the yard and highly distracted by squirrels, the neighbor’s six kids, and other dogs, she is doing better with the voice and hand signal than with voice alone. She is not obeying 100% of the time, but the percentage of the times she does is rising.
Wes from Indianapolis says
I think we have to be very careful with terms here. A “point” is not a “point”. What I mean is the orientation and distance between dog and human subjects as well as the object’s position are critical. When I point and lean toward the ground right by my foot many of the dogs I’ve worked with will sniff the ground to find the implied goodie. But if I point to something in the distance away from both of us I get a hand sniff or confused walking around much of the time. I can’t say much about the success I’ve seen and innate vs learned, but I can say spacial parameters make a big difference.
Marianna says
My cats have also learned as easily as my dogs to follow a point. They could also follow my eyes looking at an object and then looking at them and back. I did find once my cats were older, they became less interested.
DebC says
I find this all very fascinating –
but as far as a dog knowing ‘inately’ what a point means, I find that a little tough to swallow.
Between the two dogs we have, one knows exactly what a ‘point’ means, and he ‘follows’ the finger; the other one is clueless, and tends to just go to the hand.
Guess which one is the Border collie? lol
D says
My competition obedience BC was taught to follow a point (under the guidance of a good trainer, who was careful to tell me to look at the item as I pointed – not to look at my dog). She learned it very quickly, but it was taught, I don’t think she initially understood it. My younger BC has not been taught to follow gestures but he seems to “get it,” however, I honestly think that he has learned it from observing the older dog interacting with me. (which goes back to one of your previous posts… do dogs learn from dogs?)
So…one of my questions about the research is how much they know about the subject dogs’ backgrounds. Dogs may have somehow learned the behavior prior to the experiment, even though their owners didn’t realize they were being taught.
Kat says
Since this subject came up I’ve been paying attention to just how often I point for Ranger. He’s one of those that would make it look innate because he responds to the point even when I’m pointing at something that isn’t close. I can also see how he learned it without anyone thinking they were teaching him. Today he was visiting my classroom and one of the exercises we do is to let each child ask for one of his tricks and reward him with a treat. Good thing he’s about as bomb proof as they come because getting to have a big dog do tricks for them gets the kids very excited and in my really casual classroom this often means kids out of their seats pressing in and begging for their turn. I’ll cue him which kid he should be attending to with a point sometimes a very subtle one of no more than the flick of a finger. When I want him to move from behind or beside me to sit in front of me I draw a line with my finger tracing the path. Since good things (from Ranger’s perspective anyway) almost always happen when he goes to where I point he’s highly motivated. I’m a human, I point to convey information. He’s a dog, getting information from the human maximizes good things in his life so figuring out all the ways I convey information to him is a good strategy. What I find interesting is not the question of whether following a point is learned or innate but that dogs are so good at interpreting the behavior and gestures of another species. The average dog is very good at interpreting human but the average human is not good at interpreting dog.
brandy says
I have heard a bit about this research before. What I’d read then was that dogs can be taught what pointing means very quickly and easily, but cats basically cannot catch on to it no matter how much you teach them.
It’s interesting and entirely unsurprising that eye gaze is so important – humans babies learn to interpret eye gaze very early on, before they can talk, and dogs are one species that seems to be quite adept and interpreting human eye gaze.
Anecdote: I’ve found that rats are also very good with eye gaze. When I owned rats in the past, they always knew when I was looking at the cage. If it was around mealtime, they’d just be hanging out until I happened to look at them, then they’d suddenly jump up and start throwing fits to show how hungry they were. And I had one rat in particular, whom I had to give medicine mixed with baby food to every morning… I’d give it to her on my bed, which she wouldn’t jump off of, BUT she’d found a place in a corner where she could climb down. As long as I was watching her she’d behave and eat her medicine, but the instant I looked away she’d take off for the corner to try and get to the floor.
em says
Fascinating topic. I’ll echo many of the posts above when I say that Otis came to us as an eighteen-month old who demonstrated no ability to interpret a pointing finger. He learned it as it was done in the video, coupled with motion and eye contact, very quickly (within than one day), implying that he had no innate understanding of the point, but that he did have the ability to easily learn the gesture.
There were a lot of things like that in the early days of caring for Otis. He didn’t understand throwing, either. He’d bounce eagerly when shown a ball, but when we threw it, he’d just continue to look at us until he heard it hit the ground, then whip around, spot it, and bound gleefully after it, giving every indication that he was pleasantly surprised to see it suddenly across the yard. We considered the possibility of brain damage. But he learned to understand throwing in very short order, too, (though he still doesn’t fall for the ‘fake out’ throwing gesture…if he doesn’t see it leave our hands, he assumes we still have it). Again, it seems that he is not lacking in any innate capacity, but was much more likely the result of an unconventional upbringing.
We struggled VERY hard to teach him verbal communication (it took three months to learn his name, almost six to learn the word for ‘sit’), but he was acutely sensitive to changes in our posture, the direction of our gaze, and motion of our hands and feet. He was able to learn all his basic commands very rapidly, cuing off gestures, but for nearly six months it was very difficult for him to understand spoken words, or even differentiate between emotional tones of voice. Once he caught on, though, he began learning at an exponential rate, picking up accidental words like “delicious” and “medicine”, in addition to a normal range of verbal communication. During his learning phase, we could see human speech transforming for him from meaning nothing, to indicating that ‘something’s up..look at the human for more info’, to imparting emotional information, to the final stage of specific words imparting particular meaning. The process has lead me to re-evaluate my understanding of ‘innate’ ability and behavior in dogs, particularly as it concerns communication.
I have become more skeptical of evolutionary explanations for learned canine behaviors like following a point or making protracted eye contact (Otis had to be taught that, too), though I admit that the possibility that nature and human breeders have selected not for behavior but for the ability to learn certain behaviors intrigues me. I’d love to hear more!
Heddie Leger says
This is absolutely the most intriguing topic. Those who are versed in dog training know very well that dogs have their own language that predominantly consists of body language consisting of metasignals that humans often miss. We have become fairly adept at decoding the canine body language and the anticipated meaning. We are not as good at decoding what their barks and different tonations mean. It is entirely possible that our pointing, eye signals, and head turns are no different to the dog than when another dog provides a physical body signal to communicate any number of nonverbal messages. Being verbal communicators we have to relearn by careful observation the art and science of body language. The pointing experiments and studies are attemping to prove what we already know…dogs are experts at reading body language. The next question that comes to mind is how do they learn it? Are they hard wired to certain things or are they just very adept at picking up signals quickly. It is possible that individual dogs have a certain ability more than others to learn in this capacity, just like some humans are better at some things than others.
AnneJ says
My dogs don’t follow a point from a distance very well, not until taught at least, but they do learn to follow a close up point such as was demonstrated in the video really quickly. Little puppies will just look at a finger and sniff if they think you might have food. They do seem to follow our eyes though. They will follow the eyes of another dog too- such as if one dog is looking keenly at something all the others will look the same way to see what is there. I can’t imagine that chimps in the wild don’t do the same thing. Did anyone consider that maybe the chimp was choosing opposite to the tester because he thought she was pointing to the one she had claimed as hers, so he’d take the other one?
trisha says
To Lynn: Great question about the use of pointing in Double Lift Herding Dog Trials (where the dog brings one group of sheep and is then told to “Look Back” for another. The second group is usually on a different side of the field, so the dog is expected to run, for example, clockwise to get one group, then counter clockwise to get the other. It is indeed true that the dogs could be taught a pointed arm means the sheep are in “that” direction, but I suspect no one does because we all want our dogs to keep their eyes on the sheep and away from us as much as possible. That’s an interesting contradiction in Field Trials, in which the retrievers are given a “Stop and Turn around and Look at Me” signal (often the dog is swimming at this point) and then the handler points them to the right or left. I watched that and thought “Gosh, why don’t they teach the dog a whistle to go right or left?” I suggested that to the handlers and they thought is was a lousy idea. They said “But we want our dogs to turn and look at us!”
To Michelle: Thanks for mentioning Dr. Pam Reid’s article in Beh’l Processes. I thought it was excellent too. By the way, she is also speaking at APDT on what ‘dominance’ is and isn’t. Should be a great talk. Regarding your wondering if the public ‘back and forth’ about the meaning of the results of pointing studies has been useful, I would argue that it is. Yes, absolutely I agree that the issue is a lot more complex than some would like it to be, and that there are many interesting issues regarding cognition that get missed in some of the arguments, but isn’t this how science usually goes? One result leads to different interpretations by different people, which leads to more studies that at first make the issue even more complicated if not confusing, but which eventually lead to us truly progressing in our understanding. I do agree that thoughtful skepticism is one thing, but blanket statements of ‘right’ and ‘wrong’ are not useful — just goes to remind us that the process of science, if done correctly, is indeed objective, but scientists, being human, never really are.
To Wes: You bring up a good point (ha!) — and I absolutely agree that the distance between the end of our arm or finger and the object of interest is of great importance. I suspect that might be one of the factors that makes dogs so quick to look toward a container with food in it but not follow the line of our pointing arm when we use it in ‘real life.’ In the research the containers are very close together, and the person is seated on the ground. Usually when we point ‘in real life’ we are standing and the object we are pointing to is much farther away.
To All: Absolutely, as many of you have wisely said, dogs are both predisposed and taught by us from an early age that we humans are not only fonts of information, but are also willing to share lots of it. Dogs look to us to solve problems, get the food scrap out from under the couch for them, encourage them to clean up the food we just dropped on the floor and give them most of all of their food. Chimps, on the other hand, are much more competitive over food, and even if they’ve been fed by humans, are less likely to think of us a cohorts in food gathering. And very few people sit with adult chimpanzees and share popcorn on the couch. And the reasons for the glass between the chimp and researcher is that chimps can be extremely aggressive. Very few people want to be around an adult chimp when it gets frustrated . . .
em says
It also occurred to me while watching the video that the chimp researcher, shocked that a mere dog could understand something that a chimp struggled with, leapt rather precipitously to the conclusion that dogs have specifically evolved to perform a specific behavior (following a point), rather than drawing the similarly plausible (in my view) conclusion that he might have shortchanged at least one aspect of canine cognitive capacity. ‘Intelligence’ is a slippery term, especially when one considers that, as this research shows, not all animals respond equally well to all tests. Based just on the ‘point’ test, one might conclude that dogs are brighter than chimps, but tests for other sorts of cognition reveal quite different results. There seems to be a temptation to define intelligence as ‘greater’ or ‘lesser’ and expect consistent results across a full gamut of ‘intellectual’ activities, but it doesn’t seem to me that all such activities (communication and logical deduction, for example) are driven by the same cognitive processes. It also seems tempting to dismiss unexpected behavior as deriving from a lack of capacity rather than a lack of inclination. I’d be particularly interested to see more about the testing on wolves and other wild canids for that reason. It does not surprise me that squirmy young wolves wouldn’t pick up human gestures as quickly as calmer, more focused dogs, but it would surprise me if wolves couldn’t learn the signal at all, under any circumstances.
em says
Another thought. Sorry! I’ve never observed that my cats can follow a pointed finger, but I think that anyone who’s every owned a cat can attest to how well they follow a gaze. In my house, this usually manifests in a near obsessive preoccupation with whatever a human is staring at and/or a deliberate effort to interject themselves between the gazer and the object. Books, newspapers, computer screens…things that are treated as uninteresting when alone, become irresistable when I look at them for long enough. I had one cat in particular who would jump on top of the television and bat at the moving images, but only when my husband or I was watching it. A TV left on in another room never elicited the slightest interest, but sit down to watch television, and she’d go bananas.
I think this is primarily about soliciting attention, but nothing GETS my cats’ attention like pointedly staring at something. If the object seems even slightly interesting, my cats poke at it. If not, they poke at ME, meowing, rubbing, putting themselves physically between me and it, seemingly trying to get me to ‘stop looking at that stupid thing.’ I have no sciencey explanation for this behavior, but I have observed them doing it among each other as well, most notably as they ‘tag team’ a mouse. I’ve never seen the interrupting behavior among them–maybe because cats never stare at something other cats find uninteresting? Anyways it strikes me that lots of animals are good at noticing and following a human gaze. Horses, sheep, rabbits, rats, deer, etc. all seem to notice and react when a person’s eyes are on them and at least some of them can follow a gaze directed at something else. Can any of them follow a pointing gesture toward a desired resource? I guess my question is not so much “why can dogs interpret a point readily?’, but ‘why CAN’T a chimp? ‘
Kat says
As others have mentioned, dogs point. They also hunt cooperatively. One day at the park Ranger and a Husky pal were hunting Labradors. It was clearly a game and it was accepted as such by all participants including the “prey” but it was fascinating to watch as Ranger and the Husky picked out their prey (there were probably 20 dogs and six of them were Labs–no idea why they were only hunting Labs and not any other dogs) and point the specific Labrador they’d be hunting with their noses then somehow (never did figure out how they agreed who would do what) decide which one would hide in the bushes in ambush and which would herd the prey in. One would drive the Lab into the ambush and the other would leap out and pounce on it. The dogs would wrestle and the Lab would be released to go play and Ranger and his pal would pick out another one to hunt. Sometimes it would be the same Lab other times a different one. It was fascinating. In light of the whole discussion of pointing though it struck me as relevant that they were pointing to their “prey.” Since they do point when working together as part of their communication it’s not that surprising that they pick up very quickly that humans point their fingers to convey information. I seem to remember that chimps do not point to convey information.
jackied says
Nature vs Nurture: our unsocialised dog (we got him at 12 months) learned facial and hand gestures very quickly, but spoken commands took a lot longer. In fact it is still difficult to teach him a spoken cue until he has learned the behaviour with a hand cue.
KB says
I have a Lab who is presently in a leg cast due to tendon surgery so I’m playing all sorts of brain games with her. One is object discrimination. I find that I can “help” her by turning my head and looking at the object that I want her to choose. However, pointing at it doesn’t work very well at all. That surprised me because I’ve also trained her to retrieve outdoors using pointing gestures. I place 2 identical training dummies on her left and right, and I point to which one to retrieve (I don’t remember how long it took her to learn this – but it wasn’t long). She’s almost 100% correct in that game. However, her understanding of pointing did not transfer to a game where she’s lying down with two different objects on her left and right. I point, and she says “huh?”. But, she responds automatically if I stare at one of them.
I love this post! Hearing about the real experiments on this topic (rather than anecdotes like mine) is fascinating.
em says
Kat,
I’ve seen that hunting behavior, too. Though he’s willing to dabble in tooth fencing and wrestling (play fighting)with his dog friends, Otis’ preferred games have always been ‘play hunting’. Most of the time he ‘works’ solo, but with some of his friends, most strikingly with other danes, he will stage the same sort of elaborate ‘hunt’ that you describe. Usually they target known dogs as they first approach across the field. The hunters sweep out from one another, eyes pinning the ‘victim’ with a deadly focus, one dog more directly in line with the victim while the other ‘takes point’ off to the side. They then stalk (often crouching all the way to the ground as their ‘prey’ approaches), and finally one dog will charge directly for the victim, driving his hapless prey directly into a whomping T-Bone slam from the point guard. Interestingly, Otis will switch positions when hunting with another male, but when he plays with his great dane girlfriend, he ALWAYS takes point, letting her bolt the ‘prey’ while waits, then lowers the hammer, so to speak. It’s not just a size thing, Otis is noticably bigger than some of his male huntin’ buddies, but pretty close to the same size as his girlfriend. It’s thrilling to watch, in any case.
For just a few moments, it gives me a window another part of my dog’s psyche-he stops being the lovable doofus who dreads stepping out in the rain and channels some fierce wild thing. Then, at the last second he pulls up, greeting his ‘prey’ with a solid-but-not-too-solid chest whump and a few minutes of wrestling before bounding off to set up the next game.
J. says
I just love posts like this – I spent the last couple of days chewing it over in my mind. I am so used to Tara finding what we’re looking for because I point to it, I wasn’t even sure whether it was something she always did or an ability she developed.
But I’ve come to the conclusion it developed over time and I think it’s because dogs are hardwired to look in the same direction as their leader (whoever that may be). Didn’t you once post something about a herding problem caused by the fact that you weren’t looking/facing in the right direction?
Then I think that dogs figure out that we, with our un-subtle body language, enhance a look by pointing and learn to use the point instead.
I am quite sure dogs themselves point to things by looking at them. I used to keep the treats in a tin on a shelf in the kitchen. When my elderly ridgeback thought she’d done something that deserved a treat, she would sit down, stare at me and when she caught my eye, she would pointedly! glance at the tin and then back at me without moving her head :).
kecks says
Still had not have time to read through all the great comments; but just wanted to let you know, that I had the pleasure to speak to Tomassello himself and he assured us that the pupps in the test were “as young as we could get them”. Don’t know if this helps.
BTW I was taught that there is no really “100% innate” behaviour?! Besides breathing or some other very basic reflex things… every more or less complex behaviour (like “follow a pointing hand”) has to be trained using a innate disposition to get the behaviour very quickly (like human young ones learn to talk only and only if someone is talking to them on a regular basis).
kecks says
okay, just finished reading the original post. sorry for me talking about the disposition thing – i jumped to this conclusion much too early instead of finishing the post. sorry!
Kat says
em,
That’s a great description! Ranger used to have a Dane pal at the park that he’d play hunt with. They played that game so much they got known as the R & D smack down team. And for some reason they used to hunt Labs and Lab mixes too (wonder what it is about Labradors that make them good “prey”). When Ranger and Dig would play hunt Dig was always in ambush and Ranger the herding breed always drove the prey in. Ranger was the undisputed brains of the operation. He’d set Dig where he wanted him and then drive the prey into him. Often they’d hunt the same Lab multiple times and you could see the Lab trying multiple strategies to escape. One actually found a strategy that worked. It was a smaller Lab and rather than trying to dodge or escape (strategies that had failed in previous efforts) ran full tilt straight at Dig, dropped low at the last second and came out the other side running flat out. Sadly Dig got less reliable as he got older and can’t come to the park anymore.
With his Husky pal Ranger was taking turns and he wasn’t necessarily the undisputed brains; she might hide and pounce two or three times then he’d hide and pounce a couple of times. And whoever was hiding looked to me like they were choosing their own ambush site rather than being placed there. Either way it’s a fascinating glimpse into a part of dog culture we don’t often see.
JJ says
My first agility instructor encouraged the class to “play as many pointing games as possible”. She thought it would help with agility. I don’t know if pointing games help with agility or not, but ever since I got that advice, I started to make a conscious effort to work with Duke on pointing. I have realized how helpful it is on a day to day basis to have Duke understand my pointing behaviors. I do lots of pointing all the time (on purpose) and Duke seems to know what I’m “saying”. My “point” is: whether it is an innate behavior or not, if you can teach your dog to understand a human point, you may be glad of it.
Since reading this post, I’ve started wondering what the limits of “the point” are in my family. How close to me does Duke have to be to understand the finger? How close to the object or location does Duke have to be to understand the information?
I don’t have answers to these questions, but I’ve been playing with it a bit and been pleasantly surprised with the results.
Mary Straus says
I’d be interested in seeing how chimps or wolves who were raised in a house with people did with the same test. Dogs might do better simply because they spend more time with us. If chimps and/or wolves raised with people do better on the test, that would indicate the results could be more nurture than nature.
Aly says
I was recently thinking about weather dogs innately know what pointing means too. Mainly because I was in the process of teaching my new dog what pointing means (normally that I’ve dropped a cookie and she should go get it.) She learned quite fast, but it was definitely something I had to teach her.
Deanna in OR says
Iit’s interesting how my Border Collie quickly learned to “point” for me, using her eyes (much like the human researcher’s eyes in the video). I will play with her with a toy for a while, then put it up high (on top of a tall bookcase or the fireplace mantle). Not being willing to stop just yet, she will often sit in front of me, looking at my face, then at the toy, then back at my face, then at the toy.
I didn’t teach her to do this, and she doesn’t get rewarded for it. But she has done it since she was quite young. (Often right after jumping up on the couch and “Kissing” my ear, trying, I think, to sweet-talk me into continuing the game.)
It may be easier for a dog to learn to interpret a “pointing gaze” in another being if they have this innate ability themselves.
JKFL says
I think this is a fascinating research topic. The researchers have probably controlled for the following, but I will ask anyway as I am curious.
Isn’t is possible that even though both containers are scented with food, the dog (with its exquisite sense of smell) can smell the (fresh) treat?
Did the dog reliably go to a container that did not contain food, when the researcher pointed at it?
Camilli Chamone says
As a genetic researcher and a dog lover person, I agree there many probably dark points in this research.
I
Lindsay says
I find the comments about cats interesting. I would think that cats would also be able to understanding pointing since they have lived and depended on humans for so many generations as well. My cats do not get it, though. Maybe it’s because cats depend on us less. Maybe it’s because we spend less time interacting with our cats. I have a cat that is obsessed with food and you’d think it would be easy to teach him to follow a point, but instead he just fixates on my hand. I work with a lot of dogs, and even the dogs that are not trained at all seem to understand my pointing gestures, whether it’s pointing out food or suggesting they walk in a certain direction.
Jo says
My dog is a rescue Border collie and I honestly can’t remember if I taught him what pointing meant or if he already knew before I got him – if I did teach him he must have learnt really fast as I don’t recall training it at all. I do believe following a point is something puppies have to learn but I also think dogs are predisposed to watch us for cues and act on them when they recognise one. It makes sense that a dog would innately understand that when someone deliberately looks at something there is some reason for it, and would naturally want to investigate what that person was looking at – however I would have thought wolves would act in the same way (perhaps they do, but only towards other wolves?).
My cats understand pointing now but I did have to teach it to them, and it took a long time. I can point at a treat and they will look for it in that direction, or at a chair/bed and they will jump onto it, point at the floor and they will jump down off a counter, or point to another room and they will go to it. People are often surprised that cats can learn to do this but I think it’s really just a matter of how long you take to train them and how much patience you have.
I think the difference between dogs and cats is mostly that cats lack that predisposition to watch their owners for cues and try to follow them. They can learn to follow commands but it takes a lot longer to train them to do so than for dogs. Even though they’ve also been domesticated for a long time, they are much more independant than dogs and are naturally solitary animals. They can survive very well without human help so I suppose their survival never depended on obedience to humans. I think cats are actually better at learning alone than dogs: my cats can open closed doors by jumping up to the handle and turning it while pushing or pulling at the door. I never trained this behaviour, it’s something they learned by observing us opening doors. My dog has never learned this, despite being larger and physically stronger than my cats (it’s a real effort for a 3kg cat to open the heavy, fireproof kitchen door but they still manage it!).
Karen says
Just ran across this post & thought I’d add this: children have to be taught to look at what is being pointed to also. When you point at something, babies will look at your finger, then look up at your face, and then you can coach them to look at what is being pointed at. So it is not an innate behavior in humans, either. I think maybe one of the differences in dogs and chimps is that dogs have lived in a cooperative relationship with humans for thousands of years and have been selected for that cooperativeness. The relationship between a human and a dog is more of a partnership than the relationship of a chimp and a human.
Karen says
P.S. My dog will stare at me and then stare at the back door when she wants out. She doesn’t jump or bark, she just stares at me until I look at her and then she stares at the door. I then get up and let her out. So who has trained who?
Dean says
Before he died at the age of seven a couple of years back from liver cancer, my rottweiller could go to where i pointed, if say we were on a walk and i saw something he might be interested in i would point in its direction and say ”whats that over there son, go and have a look” i actually used to speak to him in full sentaences as opposed to just one word commands, which he also seemed to understand spokkily well, and he would go to where i was pointing to investigate, in fact he did a lot of things i wouldnt have thought a dog could do from just words i said or gestures i made. one thing i brought him up to do, was to walk off a lead, everywhere, even if it was crowded. i did used to comments like ”a dog like that should be on lead” and so on, but he was that well behaved that i didnt deem it neccessary. if we were out together and i wanted to cross the road i’d say (word for word most of the time) ”howay son, lets cross over” and he would. one time creeped me out in particular, i had took him to where i was staying (as at this time he was living with my brother) and i put him some food and water down. now he was has asleep on the bed, and when he woke up, he went over to eat some of his biscuits. he’d always been a noisy eater, so this particular moment and because i was used to just talking to him in full sentences, just as he was about to dive in to his biscuits, i said ”nanook, you’ll have to keep it down cuz its late m8” and suddenly he stopped and then started to slowly and quetly take out one biscuit at a time and eat each one really gently, now i never taught him to this, i didnt even expect him to know what the hell i said, but he did what i had just said. and it was the coolest thing ever. its a shame he’s no longer with us, as i think he wouldve made a brilliant study case for you in dog intelligence. i never brought him up in any special way, i just gave a hell of a lot of love, and spoke to him like he was a mate or something, and more often than not it was like he understood what i said. did i just get lucky or something?
E Jones says
I have a springer, Bruce, a welsh-english cross. He’s our first dog so we set out knowing little, but he’s bright and training up as a disability help for my son. Age six months he grasped that if I point he needs to go pick up what I’m pointing at, and at age 1yr he’s pointing back. He looks at the ball or biscuit he can’t reach and points at it with his paw.
I didn’t know dogs could do that!
Cindy says
What I found interesting is my fox seems to understand pointing more quickly and easily than my dogs do. Just something I have noticed.
Sally S. says
Cat person here! (I speak several feline dialects, in factđ). I completely agree that cats generally donât respond to a pointed finger by following the point, but cats understand eye pointing, for sure. They do it themselves, and they understand when itâs done to/at them. The chances of them complying with the request are slender, though. Cats are almost exclusively motivated by self-interest!
Shannon says
I know this is a super old article, I apologize! But I just got a dog, and I’ve been observing this in her. We got her at 6 months old; she had spent her whole life in a shelter prior. She doesn’t understand pointing yet; I’ve been teaching her for the last 2 months, and she is just starting to sort of get it. Prior to that, she would mob your finger.
What maybe makes this more interesting is that I now also have a DNA analysis of her. She’s a terrier mix, and has 0.0% wolf alleles in her genetic makeup. To me, knowing that makes the conclusions in that video seem even more absurd.
Anyway, thank you for writing this article. I remember watching that documentary and being really frustrated at how wrong parts of it were.
Fredericus Rex says
Just stumbled on this fascinating article.while doing an internet search on the topic. Thank you for your reporting.
BTW, in Bulgari, at least in the areas with which I was familiar, pointing was done with the elbow, so fingers actually pointed in the opposite direction indicated. I wonder if dogs in Bulgaria learn to respond appropriately. Cross cultural research on this topic would be interesting.