The above Pickles cartoon will resonate with those of us who have canine companions. (The whole week starting on June 24 features this storyline.) But do dogs really know that they have done something wrong and do they have guilty looks?
In the book Moral Origins: The Evolution of Virtue, Altruism, and Shame (2012) that argues that group selection played a major role in how morality evolved, the author Christopher Boehm says that we are fooling ourselves in thinking that dogs have consciences and show guilt and remorse. He describes himself as a fanatical dog lover and says that dogs are “friendly, affectionate, loyal, empathetic, eager for approval, and, if their owners are in trouble, often protective and self-sacrificing.” But he says that however much we may think so, they do not seem to have a rule-internalizing conscience and sense of shame.
[T]he idea that after-the-fact punishment can produce a positive shift in the dog’s behavior, just as it does with humans, is quite erroneous. Any professional dog trainer will tell you that you must punish your canine pet right in the commission of the deviant act – or at most within just six-tenths of a second after the dog’s unappreciated deed is done. Otherwise, apparently your dog will be confused because it will see you, a person it is closely bonded to, being hostile or hurting it for no good reason. (p. 22)
Clearly this assumes that aversive training, where one punishes the dog for any errors, is a good thing. I am not sure what the punishment consists of on the spectrum of scolding, though light smacks, to more severe physical punishment.
I do not approve of corporal punishment for children or dogs. Wagging my finger and speaking in a stern voice is as far as I go. I have no idea of this is effective with dogs, or even children, come to that.
UPDATE: As I said in the comment, Baxter the Wonder Dog almost never does anything that merits a reprimand and on the rare occasion that he does, he looks so cute that one simply does not have the heart to admonish him except in the gentlest way, like ruffling the top of his head and saying “Why did you do that?”
If an animal has been punished, would it not learn the human’s expression and body language that might signal impending suffering? I’d expect an animal to be very attuned to that. I’ve observed this in horses many times, including a horse that would threaten to kick anyone who came near it holding a riding crop.
We might interpret such behaviors as “guilt.”
It’s my impression, from experience and watching PBS, that dogs (and other animals) are very good at reading us. Not just tone of voice, or body language, but facial expressions as well. They know when we’re mad at them, although it may not always be clear why. And they respond in a way which suggests what we might call guilt. But many humans are reluctant to admit that non-humans might share some of our feelings. Weird. What’s the opposite of “anthropomorphize”?
A few years ago I saw a swallow attack* a sparrow which had recently killed all the swallow’s chicks**. Dog forbid we ascribe anything as sophisticated as “rage” to that action.
*Not a sustained attack. Just a top speed beeline for where the sparrow was sitting.
**I blame myself. I let the sparrows nest in an old swallow nest in our carport one summer. The next summer, the swallows beefed the same nest up and used it. I guess the male sparrow felt territorial or summat. Never seen or heard of that before.
I would agree with the author that the answer to your question is “no”. “Guilt” and “shame” are human constructs. They make sense to humans. They are not going to make sense to non-human animals.
Now, humans try and understand things on their own terms, of course, but a dog can’t really know or care about that. Dogs live in packs and their highest drive is to be a good pack member. To that end, they will sacrifice chances at sex and food, primary drivers for animals, for the sake of being a good member of the pack. This makes them amenable to human companionship: they replace us with the pack, and will work for the good of us instead of themselves. But that does not magically transform them into animals that descended from primates, with all their very different behaviors.
Part of experiencing “guilt” and “shame” has to do with understanding time passing, and being able to remember what happened in the past. Dogs really don’t grasp that… you may have noticed that your dog is thrilled to see you no matter how much time has elapsed. In some cases it can even be just walking around the yard. If your dog is SUPER thirsty, but you only give the dog half a cup of water, it can’t see into the future and understand that it’s not enough water. The dog will simply drink it and still be thirsty. And as the author has pointed out, unless your “correction” (HOPEFULLY NOT HITTING THE DOG) happens nearly simultaneously, the dog will not understand why the correction happened. This is why, in operant conditioning, one must use a good reinforcer AS the animal is performing the desired behavior. Your timing is really critical. In the case of operant conditioning of course one is using only positive reinforcement; one does not punish in this system of behavior modification.
But a sense of guilt or shame does not come into it at all, for the dog. A human might imagine (something else dogs can’t really do) how *they* would feel in the same situation, but frankly, human behavior is not the “best” or the “top” behavior. It’s just the behavior that humans happen to have. It does not therefore follow that every single animal on the planet feels exactly the same way about things that humans do.
@kestrel -- I see no evidence that humans learn guilt and shame through anything other than operant condition.
We want to privilege human emotions; denying that other animals also feel is essential to our human supremacy.
How do you know when a person internally feels any emotion they’re expressing? Some people can shed tears on demand. Since we can’t tell, we make the easiest assumption: assume the person is feeling the emotion they display unless we have reason to think otherwise.
My dog, while recovering from anaesthetic, defacated inside. I cleaned up the mess without rebuking it for something I knew it could not help, but to me, with my ignorant human perspective, it looked very ashamed. I could ascribe other emotions to it, but do I really need to?
Came here looking for a Baxter tWD pic…
Geoffrey @6: Wonder Dogs never do anything to be ashamed of.
@Marcus: I think you’ve taken my point backwards… I am suggesting that what animals feel is what they feel and what humans feel is what humans feel. Assuming that human feelings are the gold standard, and that all other animals have to or do feel exactly the way we do, is to me the wrong idea. Maybe goat’s feelings are the best ones! Maybe the way Great Horned owls live is the best way! What I see people doing is trying to turn an animal into a pretend human so they can value it more -- I think the animal is valuable exactly the way it is, there is no need to try and project my own feelings on it to make it more valuable or important.
There is also danger in turning an animal into a pretend human. If you don’t understand its nature and behavior and try and treat it like it’s a human, you can even end up causing its death. For example kangaroo rats are highly territorial, and if you are keeping them confined for some reason and think they are “lonely” and put several kangaroo rats together, in the morning you will wake up to one battered and bloody survivor, who has finally succeeded in driving the intruders out. Even though a human would feel lonely a kangaroo rat feels differently and we need to honor that.
Rob @#7 is right.
Baxter the Wonder Dog almost never does anything that merits a reprimand and on the rare occasion that he does, he looks so cute that one simply does not have the heart to admonish him except in the gentlest way, like ruffling the top of his head and saying “Why did you do that?”
But I should have posted a photo of him and have updated the post to do so.
I grew up with dogs and have a fair amount of experience with them. Later in life I’ve become more of a cat person but I’ve been around dogs for almost half my life now. The last few years or so a close friend has been digging into behavioral psychology and how it applies to dogs so I’ve had peripheral exposure to that.
From what I’ve seen and with the opportunity to interpret based on some loosely translated behavioral theories, I don’t really think dogs feel guilty. They just feel hurt when a person who they’ve bonded with closely seems to be very mad at them. People can interpret that as guilt but a more accurate interpretation would be to compare it to a child who’s just been scolded. You can’t look at the child and tell whether the scolding is justified or not*.
Honestly I think a much more interesting question is how do you form a really good bond with an animal or child that you can use to promote good behavior. Someone who favors a more hardass view of the world might suggest Baxter is spoiled from the description but it sounds a lot more to me like he just has good friends. And that naturally leads to wanting to be happy together with them.
Oh, and one thing my friend relayed is that punishment is always counterproductive. If you want anyone of whatever species to behave well around you the best way to encourage that is to be their friend. Being the person who punishes them doesn’t do as much to deter in any species as we seem to collectively think. It seems that punishing a person or creature reinforces not good behavior but that you are the one who the person or creature will feel awful with after the fact. Events preceeding the punishment still largely stand on their own. My interpretation of that is that yes, you can get behavioral change through punishment at least with humans, but you’re not getting that by a direct route. You’re getting change in behavior because you’re degrading the life experience of your target. It’s a threat. Less like “I feel bad for what I did, I will try to do better” and much more “Maybe I should carry pepper spray in case I get mugged.”
Almost all of what I’ve typed here is opinion, either mine or someone else’s. Nobody seems to have all the answers on this topic. Think critically.
* Yeah, I know some people think they can do that. They’re overly self-important assholes. And they’re wrong.
There is plenty of evidence to show that positive rewards are the most effective way of training any animal, even fish. Dogs can certainly learn what is wrong in your household this way, you ignore the behaviour you wish to cease and ask the dog to do something you can reward instead. At this very moment I am continuing my training of Jake who is inclined to bark at too many things, he was being very good this morning until my neighbour decided to test his PA system and (not unreasonably to my mind) Jake responded by barking, so we are practising ‘get on’ a dog bed that happens to be the same size as the vet’s scale whenever Jake hears my neighbour’s voice -- which doesn’t normally set him off, but is after the PA incident earlier. When he gets all four paws on the bed he gets a treat. I am also not distracting him from quieter expressions as his previous owners taught him not to growl (f^cking idiots) and we are trying to teach him that growling is ok and even useful to him … aah he’s just settled down of hiis own accord and will probably go to sleep if nothing else hapens to set him off. Jake has a look that you could percieve as guilty when he is doing something he knows we don’t like, say bringing a stick or stone into the house, I could speculate that it means he is aware this is against the house rules, but wants to do it any way, but I don’t know what it really means. I do know that he responds well to reward based training as have my other dogs to varying degrees.
Thanks, Jazzlet, I intended to complain that everyone talks about punishment as a method to train dogs and nobody about rewards.
As for guilt and shame, asking the question whether dogs have a sense of it is pretty useless as long as you don’t define what you mean with them. It’s not hard to imagine a definition that totally would apply to dogs.