Recently I asked the question, “Can Dogs Tell Stories?” I didn’t mean stories in the sense of fictions we tell children to help them go to sleep at night, or novels about sexy spies, but stories as internal narratives that describe the “who, what and where” of an event in the past. In other words, can a dog have thoughts like “The last time I came to this dog park a big, dark dog ran up to me and play bowed and we had so much fun running around together.”
Readers, as usual, had some truly thoughtful answers to the question, and if you haven’t read them I encourage you to do so. It turns out that many others have been pondering this question for quite a while, and I thought you’d be interested in what they have to say.
Not surprisingly, given the complexity of this issue, there is not a definitive yes or no answer, based on the research that’s been done. Scientists don’t agree by any means, but… we are getting closer to an answer.
The technical term for an internal narrative of a past event is “episodic memory,” or an autobiographical event that is remembered as happening to one’s self. It has been argued (primarily by Tulvig, for those of you who would like to pursue this), that three things are required to be able to have episodic memories: 1) “Conscious recollection,” (titled “autonoetic awareness” — hey, don’t blame me for some of these terms) or an awareness of remembering a past event, 2) a subjective sense of time and 3) a sense of self (as in, these things happened to me). Tulvig and others argue that animals are not capable of episodic memory, that the ability to construct mental sequences of events is uniquely human, requiring self awareness and language at the very least. However, for a long time it was argued that animals had no capacity for long term memory, a belief long disproved by innumerable studies on a variety of animals from Bonobos to Blue Jays.
Some neuroscientists are questioning the arguments of others than animals are incapable of recalling, in journalistic fashion, the who, what and when of an event. The challenge is to investigate if animals have episodic memory without the luxury of being able to ask them. To summarize a coma-inducing stack of research reports[1], Neurobiologist Richard Morris suggests that there are at several reasons to suggest that some animals might posses an “episodic-like memory system.” First, several studies have shown that rats, mice and some species of monkeys are able to solve tasks that suggest the existence of episodic memory. In addition, both similar brain structures (especially the hippocampus) and similar physiologies related to the storage, retention and recall of memory are found in non-human mammals, suggesting that both structure and function could lead to similar abilities.
A review article by E. Dere et. al. (2006, Neuroscience & Biobehavioral Reviews 30) also argues that evidence suggests that animals can create mental narratives. In one study cited in the paper (Menzel 1999), a language-trained chimpanzee observed a caretaker hiding a food beyond the fence of the enclosure, out of reach of the chimp. Sixteen hours later, the chimp recruited a different caretaker, who did not know where the food had been hidden, indicated the kind of food hidden and directed the caretaker to the food itself.
In another study on Scrub Jays (a highly social bird that lives in the southeastern US), the birds were given both preferred and non-preferred food to cache somewhere in their enclosure (this species commonly stores nuts in tree bark for the winter months, much like squirrels bury acorns in the ground in fall). They had learned that after 28 hours, the preferred food spoiled if left where they’d put it but the non-preferred food did not. The question was: when allowed back into the enclosure with the hidden food, which food would they go to? Sure enough, the jays choose to first retrieve the preferred food if allowed back less than 28 hours after it had been hidden, but the non-preferred (and still palatable) food if more time had gone by.
Some argue that this is indeed good evidence for episodic memory, showing that the jays recalled “what, when and where,” and used this informative to guide their future behavior. However, not all animals have proven equally adept at solving the same problem. Rats, given the same choices, did not respond in the same way, and appeared unable to modify their foraging behavior based on past experience of whether food had had time to degrade or not. Other studies have found that mammals are often not particular adept at performing in ways that suggest episodic memory, so it is important not to cherry pick through the data to confirm what one wants to believe. If you’d like to read about more studies that flirt with answers about how close animals can get to a human-like narrative description of the past, read the review article by Dere et al that I mentioned earlier. It’s not bedtime reading–I had to read it in two sessions because half way through my mind began to smoke–but it you’re interested it’s a great resource. (Thanks to Julie Hecht for directing me to it!)
Here’s one more set of facts that might shed light on the topic: The importance of the hippocampus in episodic memory. All mammals have a hippocampus, the structure in the brain most responsible not just for memory, but in humans, for episodic memory. We know that people who have had damage to their hippocampus’s can remember many facts, but can’t remember narratives. In other words, they lose the ability to remember events as a series of related ‘who, what and where’s.’ In some experiments, rodents with a damaged hippocampus do not perform well on tasks that require a memory of past events, although they are able to use their “short term” memories to solve problems.
Dere and co-authors conclude that the evidence is strong that “animals” (which ones?) “are endowed with episodic memory.” They argue that current research suggests the ability to form mental narratives is yet another example of an ability with an evolutionary history, and that perhaps we should be not be asking “yes/no” questions, but rather looking at cognitive continuities. I find this approach most reasonable, so here’s my best guess at an answer: Dogs can’t “tell stories” as richly as humans, given their restricted cognitive abilities and lack of language, but they can probably think about the ‘who, what and where” of past events in a way that is at least somewhat similar than not to human narratives. And that’ll just have to do until we have more information, or someone teaches their dog to talk.
[1] Here’s a representative quote from Morris’s 2001 paper: Put together (see Figure 5), the input specificity of transient or lasting synaptic change is determined by the pattern of glutamatergic synaptic activation that, in addition to causing transient potentiation, also sets synaptic tags.
MEANWHILE, back on the farm: I came home from Denver to a massive thunderstorm and cloudburst, and another amazing inch of rain fell just yesterday. It was good to be home, although my trip to the Institute of Human Animal Connection at Denver University was an absolute joy. More about that in another post, because it deserves a lot of attention.
This weekend is supposed to be mostly sunny but chilly, so hopefully I can get in more work in the garden. Willie and I can’t wait to work sheep again, the clinic we attended last weekend was great for both of us. (More on that in a later post too. And a then there’s that case study I want to talk to you about, and another video… so much to talk about, so little time!)
Next week I’ll be in Milwaukee for an evening speech, the first time I’ve given Lost in Translation as an evening talk versus a full day seminar. (FYI, I’m hoping to be doing the full-day version in Chicago next August, details to come soon.) The week after I’ll be doing the Texas Tango, one night in Austin (sorry, it’s full at 300 but I think there is a waiting list…) and San Antonio on Nov. 14th (still room!).
Here’s a favorite recent photo of Mr. Willie boy, happy and relaxed after his lesson at the Shannahan clinic. He’d just had his lesson and is watching the next dog work sheep. You can’t see the other dog working very well, but part of what I love about the photo is watching a dog, completely off leash, lie down quietly and watch other dogs work. So very grown up of them. I posted this photo on Facebook but just couldn’t resist posting it here too…
And here’s a fall color photo I took a few weeks ago, just to keep all those luscious hues fresh in my mind:
Jennifer Hamilton says
I just realized an example from one of my dogs that might qualify for episodic memory…
Our tennis ball obsessed dog is only allowed to have two tennis balls in our home at one time, otherwise she goes absolutely crazy trying to keep all of them under her “control” as they roll around on the hard floors. About once a week, we would notice that there was only one tennis ball in play (we assumed the other was lost somewhere on our property or under some piece of furniture in the house), at which point we would offer our dog a second tennis ball to meet her two tennis ball quota. We did not think much of this for the next 6 months. We were also aware of our OCD dog’s need to cruise the perimeter of the property once each night before settling down to bed with her two balls. Over time, we thought her perimeter check was a bit odd because we noticed in the dark shadows that she was stopping every so often to stick her head in a bush or behind a log as if to check on something…and it was the same routine and the same spots every night.
One night, we followed about 50 feet behind her in the dark, checking to see what she was doing. In each place she stopped, there was a tennis ball hidden deep in a bush, log or dirt hole. All of the tennis balls that had gone missing over the six months had been carefully placed in hiding spots throughout our property without our awareness of it.
We now refer to this evening routine as “checking on her babies”. We watch her go around the entire property in the dark, touching each hidden ball with her nose, and then moving on to the next until all “babies” are accounted for. Every so often, she moves one of her two indoor tennis balls from the house to her outdoor collection (always in private as we never see her place them.)
So I do imagine in her head, there is a narrative that says, “Before I go to bed, I must go check on each of the tennis balls I have placed outside over the past several years to make sure they are safe and exactly where I left them.” Her behavior seems to include a who, what and where…including sense of self and long-term memory.
Beth with the Corgis says
I think the problem I have here is with the anthropomorphization of the description. If indeed dogs have episodic memory of the type you describe (“The last time I came to this dog park a big, dark dog ran up to me and play bowed and we had so much fun running around together.”) then of course it must also be true that Rover can say “Last time I shredded the garbage Mom came home and got angry.”
You can’t have one without the other, and thus we open a whole can of worms that generally contradicts all we know about good dog training.
I think there are some things dogs memorize MUCH better than we do. We recently took our dogs on a hike that Jack has been on 2 times before (and Maddie once). Am I convinced that Jack recognized the place as soon as we got out of the car? Yes, definitely. And do I think he remembered the trail, where it crossed the road and re-entered the woods, and which way we looped around the pond, and why we weren’t going to continue in THAT direction after crossing a bridge to visit a young woman and her tiny son, but instead were going back in THIS direction to continue on? Absolutely, no doubt in my mind.
But do I think there was a human sort of “And then we took another trail in the woods, but this one was shadier, and last time we had lunch before the hike but this time we ate at home, and then mom and dad had tea before the drive home” sort of narrative involved? No, not really.
I think it’s important that you pointed out that the jays had a certain sort of response that indicated they understood the possible sequence of events while rats did NOT. I do believe that there are animals besides humans that can have running narratives, because we have evidence of some animals actively teaching others of their kind new behaviors and that necessitates a concept of self vs other, an idea of a desired outcome, an idea that your own knowledge would be useful to someone else but is not currently possessed by that someone else. And all that presumes an initial sense of story of what happened and what will happen next. And many social birds (but not rats) engage in active teaching in their native environments. But my understanding is that wolves and their canid cousins are not an animal that has been observed doing much active teaching; passive learning occurs when one watches another do something interesting, but I have not heard much about active teaching (and would be interested to see literature if there is some out there because I may be wrong).
I think if there is episodic memory it probably involves more clusters of images and smells as opposed to narrative/story. Indeed, the fact that dogs act just as guilty if someone else strews the garbage as they do when they themselves strew it would seem to indicate that they don’t really have a narrative of what happened at all. They have some sort of mental flash of garbage on floor=angry mother, but no overarching concept of what connects those things together, no beginning-to-end understanding of how action leads to reaction. And without that, you don’t have a story.
Trisha says
Very thoughtful comments so far, thanks both Jennifer and Beth with Corgis. There’s no question in my mind that we don’t know the answer yet, but it does seem reasonable to suspect that dogs can form some kind of narrative in their heads, although I can’t imagine it is as nuanced and richly connected as ours. Just for the heck of it: One could say that Jennifer’s dog’s checking on her hidden balls could indeed happen without an internal narrative (could certainly all be done unconsciously, just as people who are sound asleep can drive). But one could also argue that “I went to the park and had fun with a dog” is a very, very different mental construct than “I understand why I am being punished for something I did 5 hours ago.” I sympathize with the concern that owners might take an ability to understand ‘who/what/where’ as an excuse to punish a dog for something it did a long time ago, but I would argue that it is an entirely different matter for a dog to link 2 narratives together. (I ripped up the garbage and it was fun, and I got yelled when my owner came home when there was garbage in teh kitchen.) Just because a dog could potentially have both thoughts does not, in my opinion, mean the dog could ink the two together (I”m getting punished for something I did hours ago.) Make sense?
Cindy says
I just love the picture of Willie respectfully honoring the other dog’s work. Our herding trainer has had my gsd Stosh and I do that on a regular basis- Stosh lies quietly while one of her dogs works the sheep all around us, then passes them off and we have our go. Stosh seems to really the teamwork.
Beth with the Corgis says
I agree that story-telling, like everything else, likely had evolutionary development, and it’s incredibly unlikely (from an evolutionary view) that such a skill just sprang up, fully and richly formed, in humans and does not exist elsewhere.
I suppose I disagree on how far along that evolutionary path dogs lie.
“Dogs can’t “tell stories” as richly as humans, given their restricted cognitive abilities and lack of language, but they can probably think about the ‘who, what and where” of past events in a way that is at least somewhat similar than not to human narratives.” is something that I don’t think we have an evidence of, one way or the other. WHENEVER humans tell stories they have an audience in mind. Anyone who has ever written anything or thought of a clever quip has an idea of who that story is targeted towards. Even when we tell stories to ourselves, we do it in terms of two parties: speaker and listener. When we tell ourselves stories to make sense of what happened, we are using one part of our brain or personality to try to shape the behavior of another part of ourselves. Storytelling is a way of communicating concepts to others (as opposed to asking others to help us with needs, a different and simpler form of communication), and the animals I’ve mentioned that seem to tell stories also go to great lengths (in animal terms) to tell events to others.
While it is possible dogs can do that to, it is not at all necessary to explain any of their behavior and we have no real evidence of them doing so. Brief episodic memory, if it exists in dogs, is not the same thing as what we mean by “story” in even its simplest terms. “Can dogs recount a very simple chain of events?” Well, clearly yes and we can demonstrate that with experiments. “Do dogs attempt to create a meaningful narrative of that chain of events?” The scientific part of my brain really wants some sort of proof of that before making that very large leap.
Back to the garbage: I agree with your thought that the dog does not link the two together, but we are not meant to punish a dog for something it did even 45 seconds ago (let alone 5 hours) because it’s already forgotten what it did, supposedly. I have trouble reconciling that with a level of thinking that allows a dog to say something akin to “We went to this dog park and I met a big dog and I play bowed and then we had a game of chase.” And I think a lot of other owners would too, which is why I think that going down that path is wrought with potential trouble.
Here, let me give you an example of something that can easily be seen as a narrative. A couple that lives near us does a lot of rescue and had a dog-aggressive Aussie mix that they walked. Jack has very good aggressive-dog sensors, and he did not want to go within 20 feet of them when they had this particular dog out in their pack that they walked. He would approach them if that dog was not present and got along well with some of their other dogs.
Well, we went to a CGC class and lo and behold, the female of the couple was in the class with one of her other dogs. Bear in mind that when out walking, if the Aussie was not in the group Jack would approach them.
Jack was not pleased to see this woman and for the first three classes would not go near her, even though he knew the other dog. It is easy to imagine him thinking through something like “That’s the woman with the scary dog; I’d better stay away” but the whole thing falls apart when you realize that he only avoided her in other situations when the scary dog was with her. We were in a smallish room, he could clearly see the Aussie was not there, and the dog who was with her was one he normally liked.
After several classes he got over his hesitation and started greeting her and her dog warmly.
So while “There’s that woman with the scary dog. Ok, I’ve finally convinced myself the scary dog is not here” story sounds more pleasing to my ears, the way he behaved honestly is explained BETTER by associative learning; he associated her with the scary dog, seeing her triggered the fear despite all evidence that the scary dog was not around, whatever triggers OUTSIDE countered that fear when the dog was absent did not apparently exist INSIDE, but after several episodes of seeing her with nothing happening his fear diminished over time due to repeat exposure. If indeed he had a story that “Lady is only scary when scary dog is with her” he would not have needed to relearn that in his new environment, because that is the beauty of stories: they help us translate one set of behaviors to other people and to other situations.
Occam’s razor, so to speak. Even we clever humans do not tell stories about the vast majority of our existence. Much of what we do is also explained by associative learning, rote memory, muscle memory, etc. I would not ever say that no dog is ever capable of having or telling a story, but if it does occur I think it’s not only very rare, but something that only certain of the cleverest dogs are really capable of. The idea of story involves a pretty advanced understanding of self.
Beth with the Corgis says
Another thought about the evolution of story: birds, monkeys, prairie dogs, and a few other very intelligent animals live in large social groups and/or proximal clusters that include kin (and so their own genes). These animals are exposed to threats both from the sky and the ground. Some of the ground threats can climb trees, tunnel into holes, or what have you and others can’t.
These animals have developed very specific warning calls that vary from predator to predator with the explicit intent of telling their genetic relatives what danger is approaching and where it is approaching from and what it might do when it arrives (so “hawk” warning cry might prompt birds to take to low shrubs while “martin” might encourage them to go higher up or take to the sky). It is easy to see that animals who developed these warning calls that were threat-specific would be more likely to have surviving offspring, if the offspring could understand the warning. Anyone who has fed chickadees has noticed that the number of “dee-dee-dees” is not the same for “human” as it is for “human with dogs” or “cat”.
We are seeing the beginnings of language, and with language comes the idea that sounds denote ideas, and also the desire to communicate those ideas to others. Ideas= primitive stories of the situation.
I would be interested to see if there has been research showing that dogs have developed some sort of language that is designed to communicate particular ideas, as opposed to just getting attention. I know they can understand words WE use, but I’m speaking of going the other way, recognizing that you have knowledge and want to transmit that thought to someone else. I seem to recall studies of dogs barks saying that initial analysis showed that there is no chickadee-type intent to communicate different ideas, but there might be more out there now.
The reason is that the more I’ve thought it through, the more I realize that we can’t separate the human idea of story (even basic stories told by pre-speaking toddlers) from the idea of speaker vs audience. So much like the tree falling in the empty forest, the question is can we define “story” without the idea of “audience”?
liz says
Though I don’t relish adding another question to the pile, can conditioned response be tied (at all?) to episodic memory? I gather that conditioned response can be evoked in many species w/out requiring the thought that occurs in substantially more advanced species. But can automatic responses become part of a larger internal narrative involving thought? Considering the Scrub Jays, maybe, if their food preference is conditioned- Food X was tasty and satisfying and the best and happy-making, while Food Y evoked no special notes or feelings despite filling a belly- then is there a suggestion that animals can wrap an internal narrative around their best conditioned responses in order to receive more of them? (Or on the flip side, the worst conditioned responses of fear or anger could be tied to story in order to be avoided….)
Like so many dog issues, it’s especially interesting for me to consider the implications of dogs’ cognitive abilities on several levels. Whether dogs carry and use stories to make sense of home life is one thing, while the possible use of stories to makes sense of a shelter environment brings about other implications. To me, one of the most redeeming aspects of time in an animal control or shelter environment is observing dogs who cope quite well, who seemingly have no ‘chips on their shoulders’ or excessive stress due to lack of comfort or control over events. These dogs are paragons of adaptability, and for whatever reason (oh if I only knew!) they are able to take everything in stride. Still, even with dogs who exhibit comfortable body language/disposition, it is the human storytelling capacity that can bring sadness into the equation. And when factoring in the dogs who struggle in such an environment, I think I’d rather not compound their emotional challenges by believing that they understand the “who/what/where” of the scenario. So in a kind of contorted way, from an animal welfare standpoint I believe my bias is that dogs have no such ability, while from an owner’s standpoint I want to believe that dogs are capable of profound greatness! This is all to say that both overestimating and underestimating sucks, and I would love for science to be more definitive asap 😉
LisaH says
I believe dogs have the ability to develop some form of episodic memory that isn’t soley based on association or conditioning or reinforcement or patterning behavior. One example, once, and only once, I walked a trail near my home (but not on my property) w/a neighbor and for more than a year afterwards my dog would pause at that trailhead & look to me as if to check if we could take it. He does the same w/a neighbor he likes whose yard he has played in – pauses at the end of her driveway to check if we can go there. He knows the names of at least 50 toys, learns them within minutes and years later recalls each one correctly and will find it when asked. If its hidden from him where he can’t get to it, he does not bring an alternative toy back as he knows its not the one. He is also an excellent problem solver and definitely is able to generalize from one situation to another, and this is across time and different situations. In comparison, my other dog’s behaviors overall seem far more conditioned & patterned, and she does not seem to generalize well either.
em says
Interesting discussion. Again, I find that my own answer to this fascinating set of questions very much depends on the definition of “narrative” and renewed frustations over my inability to know what is happening inside my dog’s heads. Like LisaH, I have two dogs- one who is unusually good at generalization, pattern recognition, and problem solving and who demonstrates a particularly good memory and sense of object permanence, and a second dog whose behaviors seem much more ‘knee-jerk’, usually being either conditioned or scattered and impulsive rather than thoughtful. If I had to predict that one or the other would be more capable of narrative, I would guess that my thinker, Otis, would be the likely candidate, but oddly enough, I have anecdata about BOTH dogs to add to the pile.
A couple of thoughts, just to add more fuel to the fire:
First, the Sandy story, inspired by Beth w/Corgi’s hypothetical example of the dog being punished for getting into the garbage. Warning:this is exactly the type of very common anecdote that inspires the frustrating tendency of owners to attribute “guilt” or malicious intent to their dogs.
Both my dogs are very trustworthy when left home alone. Sandy has never been crated, Otis was crated only for the first few months with us. This has never been a problem, but once in a blue moon, one of the dogs will go scavenging for food left unattended. We don’t typically leave much out, but about once a year, Otis will snag a loaf of bread off the dining room table (he did this before the era of Sandy) and help himself. While I can’t be certain of the culprit, two or three times since Sandy came to live with us a year and a half ago, one or both of the dogs has rootled through a knapsack on the floor, pulled out a small plastic baggie of dog biscuits, ripped it open and ate them. This is no big deal. Not only have I never punished the dogs for this behavior, I typically react with mild amusement. Since they have never done anything dangerous or destructive, I don’t believe that they could be picking up on any unintentional punishment in the form of supressed anger, worry, or frustration.
But I always know when there has been an ‘indiscretion’, because when I return home, Sandy won’t meet me at the door as she always does otherwise. Otis doesn’t appear to have a care in the world on these occasions, behaving normally, but Sandy will actually hide, out of my sight. I’ll find her, skulking around the corner, and she’ll hang her head and tuck her tail- I’ll call her and she’ll walk very slowly to me, tongue flicking and wagging her very low tail rapidly in short, tense strokes. I’ll pat and speak kindly to her, and she’ll slowly perk up and become comfortable again, but it takes a while- if I let her out to go pee in the backyard, rather than running straight back as usual, she’ll sit in the middle of the yard, staring at the door until I call her. These episodes clearly cause her serious anxiety and distress. *note-she will behave normally toward Otis at these times, just not toward me or my husband.
Whatever is going on in Sandy’s head is not likely the direct result of conditioning-it happens very infrequently, and she has never been punished for this behavior. She is not taking cues from my attitude or behavior because she hides BEFORE I am even aware that anything has happened. (If she was punished by a former owner for a similar transgression, her reaction would represent generalization, not simple conditioning-not necessarily conscious, I suppose, but pretty impressive if it’s not). I don’t believe that Sandy has an objective concept of right and wrong- I don’t believe that she feels guilt, because I don’t believe that she feels that she has done something inherently bad or even that she has a concept of inherent badness, but I DO believe that in these instances, she is aware that 1) I don’t want her to help herself to cookies 2) She did and 3) I might be mad.
Narrative might be the simplest explanation in this instance. I don’t know whether it represents episodic memory, but it certainly seems to represent an ability to evaluate context that is likely to be conscious. When Sandy DOESN’T eat cookies or my lunch, set on the coffee table while I go refill my water glass, despite the fact that I am out of sight and she could easily do so before being caught, is she consciously thinking? If so, what? Does she have a complex enough internal narrative to tell herself, ‘don’t eat, mom will be mad.”? I can’t be sure, but I’d bet you dollars to donuts that if instead of going to the kitchen, I put on my shoes and went out the front door, my sandwich would be down her gullet faster than you could say Jack Robinson. Doesn’t that suggest that she is making a conscious decision based on context, reconciling two unconnected ideas- ‘mom will be mad if I eat’ and ‘mom’s gone’, drawing a conclusion, ‘it’s ok to eat!’ and governing her actions according to such a narrative? Again, hard to say for sure, but her reaction upon my return is where it gets really difficult to understand. My layperson’s impulse is to say that the simplest explanation is that Sandy is capable of narrative thinking- connecting the memory of her action- ‘I ate the cookies’, with her memory of her fear, ‘mom will be mad if I eat’, evaluating the present situation and predicting the future, ‘mom’s back. mom will be mad’.
Does this mean that Sandy is somehow morally culpable for her actions? Of course not. But thousands upon thousands of incidents like this one, experienced first hand by pet owners, and not adequately explained by behaviorists and other scientists, bolster that misunderstanding. Perhaps the best and most convincing way to moderate the tendency to anthropomorphize is not to shy away from the concept of dogs as narrative thinkers, but to explore and develop a nuanced scientific understanding of the actual cognition at work.
The issue with punishing a dog for stealing a loaf of bread or scattering the garbage long after the fact, or even correcting a dog 45 seconds later is not, I think, that the dog is incapable of remembering the incident, or even that they are incapable of understanding that an action carried out joyfully in the past is likely to have negative consequences in the present, (I have no doubt that Sandy thoroughly enjoys her purloined treats at the time)it’s that it is impossible to effectively communicate the connection between one and the other to the dog. Humans likely couldn’t do it either.
If for some reason two people were prohibited from using any language-based communication and one were tasked with teaching the other not to do something –let’s say, putting a cup down on a table–think how confusing it would be if instead of tapping the learner on the shoulder immediately and taking away the cup, the teacher waited until 45 seconds or two minutes or two hours after the fact before reacting. How would the learner ever know what was meant by the correction? It would take a long time, if it happened at all, and I imagine there would be a great deal of angst and frustration in the meantime, even despite both humans’ clear memories of events and well-developed capacity for narrative thinking.
I’ve already nattered on far too long, so I think that I’ll skip the Otis stories, though I will end with the first thought I had when I read about the Jays forgoing cached food that they predicted to be spoiled. I wondered about how I would recreate such an experiment with dogs, but the first thing that popped into my mind was an image of a labrador doing a head tilt with a thought bubble over his head reading, “Wait, food SPOILS? No WAY!” 🙂
Kat says
“You helped me get this food you can have some.” “You did not help me get this food you cannot have any.” These are two “stories” that I see played out at my house. Finna (dog) and The Great Catsby (cat) will work together to get into any bag of treats left unsecured. The Great Catsby will knock the bag onto the floor and Finna’s bigger teeth make short work of getting it open. When they’ve worked together they share the treats. If The Great Catsby has opened the bag on his own he will not share with Finna. If Finna managed to counter surf the bag herself she will not share with the cat. Sharing only takes place if the two worked together. Finna could easily take treats that Catsby has found on his own away from him but she does not. Catsby could steal treats that Finna found herself but he does not. They only share if they’ve both contributed to the effort of “hunting” the treats. It seems to me that for what I described to happen the cat and dog need to have a concept of yours, mine and ours and to have a memory that they worked together or alone and the conclusion driven by the memory of who “hunted” the treats determining who gets the treats. When I’m handing out treats Finna will take Catsby’s treat when she thinks she can get away with it and Catsby will steal Finna’s treat if he can. I’ll leave everyone to draw their own conclusions about my story.
Frances says
My immediate reaction to the story of the jays was to wonder about their sense of smell – if it was obvious at a distance by the smell that the preferred food was inedible, the choice to go straight to the less desirable cache would not require the same level of cogitation.
I would be interested to explore how much of our human tendency to tell stories is post hoc – a way of explaining what we have already done, or decided to do. I suspect dogs manage very well without needing to share the whys and wherefores, and therefore do not need this kind of story. In fact, I sometimes try to think like a dog – to simply experience what we are doing without that internal narrative that describes and places and sequences and explains. It becomes almost a form of meditation …
But I would so like to know how Sophy can walk across fields once, and still remember the route, including hidden stiles and bridges a year later – and to understand just what is in Poppy’s head when we go round the last few corners on the way to my sister’s house, and she knows we are nearly there!
Angel says
Beth’s latest comment about animals giving different warning cries for different predators made me think of different barks in dogs. I work at a dog daycare, and the dogs can see outside. When the dogs are taking a break from play, most of them laying down resting, there’s always one or two that want to stay by the window so they can see outside. Several times I’ve observed a dog bark while on “lookout”, and none of the other dogs respond. When I went to check it out, there wasn’t anything worth barking at that I could see – no people walking by, no dogs or birds or other animals, no cars in the parking lot. And then later the same dog will bark, and *immediately*, every other dog in the group is up, running to the window, and on alert. And when I went to check it out, there was something or someone worthy of a bark out there (at least to the dogs).
So dogs have different barks to signify different meanings, as well. Do they teach these to their young, what they mean and how to make them themselves? Are dogs born knowing how to give different barks in different circumstances? It is clearly a form of communication. Is it a story they are telling? “There’s a strange dog outside walking around in our yard.” Or is it just an alert?
Beth with the Corgis says
em, I have some thoughts about the “guilt response” in dogs, though they are very vague. Right now I’m battening down all hatches for Hurricane Sandy and so only will post a quick response about the second part of your post, regarding a human trying to teach another not to put a cup on a table, and how that could be accomplished without language if there were a time delay between the cup being placed and the teacher responding.
And the way I see it, this is a perfect example of why humans have narrative, and how they use it. Because if you think about it, if you had a foreign house guest without a shared language, and you went into a room and found a cup where it did not belong some time after the guest placed it there, a simple series of actions (tap the guest, frown at the cup, pick it up and move it) would quickly convey the idea that you don’t want the cup there. Why? Because the person who put it there has a story about what happened earlier (I put a cup on a table) and can assume by your displeasure on finding it that the cup should not be on the table.
And even if the person did not make the backward association, he would probably make the forward association the next time she went to put the cup there (“Ahh, wait, I remember something about cup on table = displeased host! I will put it elsewhere”). Unless of course that person were your husband, and the place the cup belonged was the dishwasher, in which case he would never learn the lesson. But I digress… 🙂
Why is that? Because one of the things narratives allow us to do is develop associations between people’s reactions NOW and something that happened in the past, or in the present. It is probably one of the several reasons why that ability developed and spread so rapidly amongst human kind; it’s so darned useful. We can trace a consequence now with an action we made prior to now because when we recall the narrative, it moves the action back into the present.
Dogs, it seems, never make that association (or rarely). No matter how many times we correct something after the fact, they don’t make that association. The idea that “garbage on floor = unhappy human” does not enter into their head while they see the garbage on the floor (if people’s remote dog-watching video cameras are any indication). The idea that “garbage on floor = unhappy human” only comes into their heads when they see BOTH the human AND the garage at around the same time.
One would think that if they held the narrative in their heads that “garbage on the floor=unhappy human” it would prompt them to think of the unhappy human as soon as the first piece of garbage hit the floor. But that does not seem to be the case.
It speaks more to associations than narrative, in my mind.
As for the appeasement behavior that follows, it seems to me that some dogs are born appeasers and go into appeasement mode any time they see something out of place until they are quite sure that no one is actually angry. Other dogs are more “meh, whatever” about life and don’t go into this mode no matter the mess they’ve made until AFTER the person becomes angry. That’s more abbreviated version, though I have some examples but am out of time…
Bri says
Hi! I read “The Other End of the Leash” a couple months ago, and I’m having fun reading some of the blog posts. I own a keeshond mix who was a difficult boy to raise and have read a lot of training books to understand him. This post got me thinking of an incident that happened around a year ago.
We had moved into a new house and lived in it for three months, and then I bought a new vacuum. The first time I turned it on, Lucky was shut into a bedroom upstairs. As I vacuumed closer and closer, I heard him barking. For a second I thought it was my foster dog or something because it was not his normal (happy, excited) bark, but it definitely emanated from the bedroom and the foster was downstairs. I guessed he was afraid of the new vacuum and shouted over the din something like “Hey Lucky, It’s okay boy. I’m not gonna run you over with the vacuum!” And he stopped barking without me opening the door or speaking gently to him. He didn’t start it up again as I vacuumed all around the bedroom he was in. I always wondered what went through his head, and how he felt my shouting was reassuring. He has always nervously skittered around vacuums but never barked at them, neither before or since. It really struck me as odd, as I tend to not humanize my dogs or their mental capacities, though I love them dearly. Did he use imagination and narrative to understand I was out there and that the noise was likely a machine I had control over?
Frances says
I rather agree with Beth about associations rather than narrative. I was shocked when Poppy followed me into my neighbour’s bathroom, and calmly squatted to pee on the bathmat while I was similarly, if more appropriately, engaged myself. She was quite oblivious to anything I said – it was pouring with rain outside, I had not picked up on any signals she might have given, bathmats are sort of accepted in our house as better than the alternatives in an emergency, and it was not our house, which always increases the risk. She quite obviously could not see what my fuss was about. If there is any kind of mess on the floor at home, however, and I get down on my hands and knees with cleaning materials, both dogs vanish immediately. Mess+human+cleaning stuff=unhappy, grumbly human, to be avoided; it makes no difference whether the mess was caused by the dogs or by me dropping a bottle of shampoo, their reaction is exactly the same.
mungobrick says
I’m coming in after the fact here, but what an interesting topic. I am currently living with a house and yard only dog, because it is hunting season around our little town and she has gunshot phobia. She will not get out of the car at any of the walks we take her on during the rest of the year. She will not go out of our yard or anywhere in town on leash. She will get out of the car to go into the hardware store, the vet, the groomer, a house she has never been in before (with people she isn’t comfortable with, and she’s a very shy dog), and last week we took her to a dog park in a nearby city as an experiment. It was outside, a woods environment (like all of her walks around home), and she got right out of the car and pulled me to the dog park gate. She hasn’t been there for two years (she is three.) I’m not sure if she’s telling herself stories or not, I wish I knew what was going on in her brain, but I have to believe there is logic somewhere here. (Okay, I guess I mean I WISH there was logic here. This is an incredibly frustrating situation.)
Elizabeth
Laura says
All the comments are great and I’ll add my own thoughts to the mix, as well as a few questions.
Em, if you’ve never gotten mad at Sandy for eating the dog biscuts, why would she associate a reaction of Mom’s mad with you. Could her apeasment behaivor be a result of anxiety because you were away from the house? I’m just stabbing at air here and that’s an honest question. I think dogs have contextual memories. I know my dogs certainly do, given how much they are able to pattern when we’re out working. Admittedly, some of my dogs have been better at it than others, but they all know/knew how to do it. I think that’s why dogs pause at the ends of driveways/trails, to check if we want to go there like last time. I know Seamus likes to take me to the ellivator every time we have to go to a different floor of the office I work in, because I trained him to find the elivator pannel with the buttons on it and I used food to do it. He loves food, therefore he likes to go to the elivator as often as he can cause it represents fun and food. I think Tricia’s description of a dog remembering a dark dog who played with him or her at the dog park is also contextual. The dog recognizes the dog park, associates it with fun and gets excited, but I don’t think that an example like that is anthrpamorfising the dog. I think we’re turning the dog’s communicating the excitement into a story to communicate it to other people. I explain Seamus’s tendency to whine while driving in the car this way. It’s a story i’m relaying because people ask me, when they hear him start to whimper, “Oh no, is he ok? Is he scared, hungry, is he hurt? Did he have a bad experience in the car?” My answer is no to all of those questions. I’ve tried everything to figure out the answer to why he does this. I’ve thought it had to do with the time of day in which the whimpering in the car started, believing he did it because it was after work and we were on our way home and he was excited because he was hungry and wanted food. Then He started whining when we would go out in the morning, right after he’d had his breakfast, so that idea was out. Then I thought he perhaps had to relieve, and was anxious to get home to find some grass. Nope, he’d start up the whining even after he’d just relieved before getting into the car. I came to the conclusion that he is just excited to go somewhere and can’t keep it to himself. I’d like to stop this behaivor, as it can get on my nerves if he carries on, but I don’t know what to do. should I reward him for being quiet, and how should I do that? He’s made a game out of food rewards before, doing the undesirable behaivor and then doing what I want him to do so he can get the food reward, and I don’t want that to happen with this. Should I pet him, give him a toy? Perhaps a toy would work, he loves to cary/hold things, silly Retriever, and maybe that would keep him quiet? Any suggestions you all would have would be appreciated and I’m enjoying the comments.
Donna in VA says
OK, now you got me thinking. My older dog dreams a lot lately, so I searched on “episodic memory” and dreaming. Some studies are showing a relationship. See Matthew Wilson of MIT and his research (I have not read it yet, but there were a lot of references and I’ll be looking at them.) The dream serves to package experience into a storable memory.
I’m also reminded of Temple Grandin’s book. She said she doesn’t “think” in words as most of us do, she visualizes situations and solutions, a style that would be closer to how our dogs think, I would believe – in pictures, not in words. But a sequence, or “story”, in the way I understand it.
em says
“If you’ve never gotten mad at Sandy for eating the dog biscuits, why would she associate a reaction of Mom’s mad with you?”
This is precisely my point. Sandy’s reaction is NOT a negative association with a real experience. And if it’s not an association, what is it? My guess is extrapolation. I’ve never even forbidden her to eat the dog biscuits. I’ve never HAD to, because despite the fact that she clearly knows where they are and that they are good to eat, she has never made any move to touch the knapsack when I am at home. I don’t even bother to put it up out of her reach. (In my defense, the concept of ‘out of reach’ is significantly less practical to great dane owners, since their dogs can reach anything less than eight feet off the ground) Dogs don’t have to learn every behavior specifically; they ARE capable of generalizing to understand ideas like ‘don’t touch mom’s stuff’ -otherwise we would have to train them to ignore every individual item that we own, one by one. Despite not having any memory of my being mad about the dog biscuits, my suspicion is that Sandy is consciously aware that touching mom’s stuff is forbidden and fears that I WOULD be mad IF she did. This fear is entirely internal to her-not based on experience- she came to us as a very well-mannered, desperate-to-please adult, so I honestly can’t think of a single instance in which I have EVER been mad at at Sandy. I can further assure that her reaction (hiding, slinking, drooping) was something that I noticed significantly BEFORE I noticed missing dog biscuits, even on the very first occasion. I have never come home to a blithe dog and a stolen bag of treats. Despite being internal, I suspect that Sandy’s actions are conscious and dependent on an ongoing assessment of the situation rather than unconscious/instinctive.
I think that in this instance, Sandy exhibits a very adaptive behavior that can frequently be observed at the dog park when one of the dogs is very possessive about a toy. The other dogs don’t touch it, even when the ‘owner’ sets it down or moves away. This makes perfect sense, since violators are pursued and ‘punished’ by the owner, who rushes back to stake his claim. But if the owner of the toy LEAVES the field, the toy becomes fair game- the other dogs pounce on and gleefully play with it, more often than not. Every setting is different, but the dogs react to contextual clues to determine whether a dog is ‘gone’ and they seldom involve simple parameters like ‘out of sight’. A dog behind a bush is out of sight, but not gone, a dog headed for his human’s car is functionally ‘gone’ but not out of sight. If the ‘owner’ returns, the dogs will immediately drop the toy again and move away, often offering appeasement gestures. Dogs who had formed unconscious negative associations like toy=punishment I would expect to leave the toy alone even in the absence of the enforcer, at least for a while. I just don’t see any reason to believe that the dogs’ actions are unconscious when they appear so deliberate- i.e. watching carefully to see if the coast is clear before completely and immediately reversing their behavior.
The problem with dismissing dogs’ capacity for narrative thinking by assuming that a constant conscious narrative of “garbage on floor= unhappy human” would inhibit garbage raids in the absence of humans is the fact that it rests on the further assumptions that a)that’s the narrative as the dog sees it and b) this narrative would dissuade the dog from action. Dogs are social predators but also highly opportunistic eaters in nature, and the ability to consciously assess the likely consquences of an action in context is very useful adaptive behavior for this kind of lifestyle. It’s just as likely that the narrative reads, ‘no human, no problem.’ It’s still a narrative. Kids have a pretty similar tendency to enjoy misbehavior until detection becomes imminent, I’ve been told. It’s a reflection, perhaps, of a limited ability to fully appreciate the long-term consequences of their actions, but not a reflection of their inability to engage in narrative thinking.
A dog shredding the trash may not be convinced that the human is coming back, or may not be capable of thinking about that possibility in terms more complex than, “not soon”, but a dog hiding from its owner before the owner has even opened the door is very likely to need to be capable of reflection on the events of the past to conclude that punishment is imminent. (It is also worth noting that the dog needn’t have been an active participant- if the first dog was present while a second dog shredded the trash, that alone could very well trigger the perfectly logical thought process, ‘X happened, I will be punished’. But I’ve never heard of a dog acting ‘guilty’ over a mess its owner made.
Having ‘claimed’ my stuff as off limits, it makes sense that Sandy would scrupulously observe that rule when I am at home and break it without a qualm when I am officially “gone”, because the reason she is restraining herself is not because of a sense that the action is wrong in itself or blind habit or negative association (since she does not HAVE an actual, concrete memory to associate), but because, based on an extrapolation of her past experiences, she predictively fears punishment when I am present or my presence is imminent (even despite the fact that I have never actually meted such punishment out), but does not fear punishment when I am absent. This by itself is a remarkable cognitive achievement, requiring a great deal of situational awareness, generalization, and the ability to predict the likely short-term future. I have a hard time believing that dogs can keep a running list of conditional responses- “don’t, ifs” or “ok, whens” or “next, this might happen” , or ‘that + this + some other thing = bad’ , some of which are very complex, in their brains to act on without being consciously aware of them. To me, conscious awareness of a decision-making process is narrative thinking. “Don’t do” can be an unconscious impulse. Flinching away from a conditioned punishment can be an unconscious impulse. Breaking the rules when and only when the enforcer is absent? That’s harder for me to swallow as being an unconscious impulse. Not all dogs display “guilt”, but the great majority tailor their actions to context, one way or another.
As to the simple association of trash + mom= punishment, I very much doubt it. (And I do think that it could easily be disproven with a simple experiment.) If I scatter trash onto the kitchen floor and then leave the house, I’ll hypothesize right now that Sandy will not exhibit anxious or appeasing behaviors when I return.The events that have triggered Sandy’s “guilt” reactions leave precious little physical evidence- it’s likely that I might not even have noticed the small ripped plastic baggie tucked into the corner of the dog bed and, being the proud posessor of a husband, an empty plate on the coffee table is not a thing unknown in my household. Neither by itself would trigger her unless she also had a memory of having broken the rules. When my uncle returned home from a trip to the grocery store with his dogs to find that black bears had broken in and trashed his kitchen in search of food, his dogs’ reactions was far from the ones they would have exhibited if they had done it themselves- they reacted to the smell of the intruder and ignored the mess. I must conclude that it is NOT the visual stimulus of the mess alone, (though, devil’s advocate, it could be the dog’s own smell on the mess, I suppose) that triggers dogs in these instances, though I think it quite plausible that the dog would not remember without a prompt-if the evidence were gone the dog might not react at all.
Marilyn Kircus says
I often dog-sit for a friend who has very different ideas about keeping dogs. Her dogs mostly live in their crates. She had a standard poodle that always acted like she was mine when I had her. And she only had to be in her crate at night at my house and I would play with her lots more and take her for walks. Once, I kept her about two months after the last time I’d had her. Almost as soon as she got in my car, she started an excited whine. Once at my house, she raced into the house, found one of the toys I kept for her and tossed it up into the air.
I read that as being able to remember being at my house and being free and getting to play, and remembering the fun she had there.
chloe says
My BC is not a big greeter of humans or dogs on our walks. She is now almost 7. What she does do which always totally amazes me is greet with great great display, any human or dog she visited even briefly during the first 6/8 months of her life. she acts exactly as a wolf pup would upon her pack returning from an absence. licking mouth of dog, tail wagging low, butt low to ground whole body wiggle. This is so weird to me because she completely ignores everyone else, I’d even say she does not like close contact with dogs.
To me that is clearly a story.
I petsit and dog walk for a living, and dogs and cats most definitely tell me stories not with language so much but behavior, their bodies show me. The thing is I can’t always translate what they are telling me. Also I have learn to be cautious about being sure I understand what they are telling me. For example shyness/cringing/ cowering/afraid of a particular object etc. Most of the time they were not abused at all, but at some point something scared them. So at times that is all I understand, ” I see that you are scared of this, ok, all good, lets move on”. I’ve learned to not put words in their stories but acknowledge there is a story never the less.
chloe says
Just remember the cutest thing my dog did when she was 4 months old. I had a housemate that Baruch really loved. He was in his room sleeping and we were getting ready to go on our first walk, she went to his shoe in the mud room smelled it and wagged her tail!!!