Why should a dog recognise itself in a mirror?

[Darat]

I was watching a TV programme that, in passing brought up the idea that only humans (and perhaps some of the apes) recognise themselves in a mirror and how dogs don't. The way they concluded that dogs don't is that they don't seem to correlate the image in the mirror to themselves e.g. put a red dot on a person's cheek and show them a mirror and they know that dot is on their cheek.

But today when I was out walking my dog I was watching her greet other dogs and it was quite apparent that dogs don't just rely on vision to identify other dogs. So couldn't another explanation for the apparent lack of an ability in dogs to recognise themselves in a mirror simply be that for a dog the required information, for such an identification, is simply not present?

[Fiona]

Seems reasonable to me, though I do not know how you could test it: a smell mirror?

[jasonpatterson]

A mirror was used because it's an easy test. Your criticism of it is one that was made fairly early after it was devised, as well. Attempts to make scent tests for dogs have been made, but I'm unaware of any of these that did also have fundamental problems. (That doesn't mean that they don't exist, only that I don't know about them. )

[nimzov]

I agree with the OP.

I think that in many tasks, such as finding food, vision is secondary to smell with dogs.

When my dog finds a piece of bread on the sidewalk, it is usually by smelling it rather than seeing it.

[sphenisc]

Originally Posted by Darat

I was watching a TV programme that, in passing brought up the idea that only humans (and perhaps some of the apes) recognise themselves in a mirror and how dogs don't. The way they concluded that dogs don't is that they don't seem to correlate the image in the mirror to themselves e.g. put a red dot on a person's cheek and show them a mirror and they know that dot is on their cheek.

But today when I was out walking my dog I was watching her greet other dogs and it was quite apparent that dogs don't just rely on vision to identify other dogs. So couldn't another explanation for the apparent lack of an ability in dogs to recognise themselves in a mirror simply be that for a dog the required information, for such an identification, is simply not present?

I not sure I understand your argument. Your explanation is fine, but you don't say what their explanation was or how yours differs from it.

[Darat]

Originally Posted by sphenisc

I not sure I understand your argument. Your explanation is fine, but you don't say what their explanation was or how yours differs from it.

What they were saying was that dogs can't recognise themselves the way humans do and that indicates some lack of an "I" - in other-words when presented with an image they can't connect that image to themselves as we can do.

[Professor Yaffle]

I think the point Darat is making is that the mirror test is usually taken as an indication of self awareness. Dogs fail the test and are therefore seen as less self aware than animals that pass the test. But because dogs rely more on scent than vision for identification purposes, this inference may be invalid.

ETA - oops, was writing as you were posting Darat.

[sphenisc]

Ahh... this is like the misinterpretation of the Turing test.

"Passing the test implies a sense of self."

doesn't mean that

"Not passing the test implies a lack of sense of self."

[Wolrab]

Dogs can't figure out how to sniff their image's butt.

[Gagglegnash]

Hi

I remember seeing a TV show once that mentioned that a human's brain has a huge visual cortex, but a dog's brain is mostly olfactory cortex.

The show's writers said that, in fact, the dogs' olfactory and visual cortices were pretty much overlapping, so dogs may actually, "see," scents. I took that to mean that while their eyes only pick up sepia tone stuff, they may get colors like, "mostly horse meat, but had fish yesterday, and rolled in a decaying opossum within the last week."

I agree with the OP. The test is flawed because it makes assumptions of facts neither supported by research nor currently in evidence.

[Bikewer]

That's the explanation normally given by trainers and such; puppies will generally react to a mirror image as if it's another dog, but they rapidly learn there's no distinct scent, and thereafter ignore such reflections.
Now I suppose you could extrapolate this to a failure of recognizing "self"...But only visually.

[Olowkow]

I'm pretty sure it was the most recent "Are We Alone" podcast, with Seth Shostak where his guest was discussing pigs, and their intelligence. Pigs are one of the few animals who can use a mirror properly. As I remember it, they can find food behind something by viewing it in a mirror, not going after the mirror, but the actual food.

[Denver]

I wonder what opinion or awareness a dog has of their shadow.

[JoeTheJuggler]

Originally Posted by Darat

What they were saying was that dogs can't recognise themselves the way humans do and that indicates some lack of an "I" - in other-words when presented with an image they can't connect that image to themselves as we can do.

If "they're" saying that, it's not so accurate. From what I've heard, ethologists hold the mirror test as proof of the presence of self-awareness, but do not hold it as proof of the absence of self-awareness in animals that don't "pass" the test.

[Professor Yaffle]

Are there any other animals that rely more on smell than vision that DO pass the mirror test?

[JoeTheJuggler]

Originally Posted by Professor Yaffle

Are there any other animals that rely more on smell than vision that DO pass the mirror test?

I don't think so, but some cetaceans pass it, and you could argue that they rely as much on hearing for identification as dogs do on smell.

At any rate, while I think the test is good evidence of the presence of self-awareness, it ought not be used to argue for the absence of self-awareness.

[sphenisc]

Originally Posted by JoeTheJuggler

If "they're" saying that, it's not so accurate. From what I've heard, ethologists hold the mirror test as proof of the presence of self-awareness, but do not hold it as proof of the absence of self-awareness in animals that don't "pass" the test.

That guy in the mirror already said that.

[JoeTheJuggler]

Originally Posted by sphenisc

That guy in the mirror already said that.

Not only that, you also said it more succinctly than I did!


But I guess I'll have to reflect on this some more. . .

[Vortigern99]

My opinion on this, as an amateur zoologist and dog- (and cat-) lover, is that dogs don't react to a reflected image past the puppy stage because they recognize it for what it is -- a reflection, or at least, some kind of illusion -- and have no impetus to respond to it. If it cannot touch/affect them and they cannot touch/affect it, it is beyond their interest.

As to the "red dot on the face" test, I'd be interested in learning how the researchers achieved this test with a dog. Dogs cannot see color, it's said, so the animals would be unlikely to notice a spot of paint on their face, unless they could feel it. And, even if the animals did see the spot in the mirror, do we expect them to touch it, as human beings do, as an indication of self-awareness, with their paws? I'm not sure I buy that a dog would care if it had a paint spot on its face, again unless it could feel it, which would invalidate the mirror test.

On a side note, I've seen a youtube video of pugs watching themselves on a youtube video on a laptop. The dogs sat and watched themselves reacting (by tilting their heads) to their master's voice, asking them if they wanted to "go for a ride", as if they understood they were watching a video of animals rather than real animals. It's hard to say, of course, what was going on in their minds, but they did appear to recognize they were watching themselves.

[Professor Yaffle]

With most animals, when they see the spot (doesn't have to be colour, it could be a very dark spot on a light coloured dog or vice versa) try to rub it off themselves in some way - if they pass the mirror test.

I did just notice this in the wikipedia article on the subject:

Quote:

Adaptations of the mirror test have been made in other modalities, such as scent. For instance, biologist Marc Bekoff developed a paradigm using dog urine for testing self-awareness in canines.

http://en.wikipedia.org/wiki/Mirror_test


ETA:

Quote:

The simplest opposing argument is that failing the mirror test does not prove a lack of self-recognition. By its very nature, the test is biased in favor of visually oriented species with stereoscopic vision. Dogs, for example, are notorious for failing the mirror test. However, dogs also have relatively poor visual acuity, instead relying predominantly on their senses of smell and hearing. They don't recognize themselves in a mirror, but multiple studies have demonstrated that canines reliably recognize their own scent.

http://everything2.com/node/1967156

[roger]

Besides the other objections, it's a pretty big cognitive leap to understand that a reflection is a reflection of a real world object. On of my dogs has managed to get trapped in a corner (no, nothing in the way, he just had his head in a corner) - I don't expect him to puzzle out that a mirror is reflecting light such that the image is a representation of a real world object. That says nothing about whether he has self awareness.

But I also think 'self awareness' is nearly as poorly defined and understood as 'free will'.

[Professor Yaffle]

From a Psychology Today article by Mark Bekoff

Quote:

How do animals differentiate themselves from others? Many studies of self-awareness have used mirrors to assess how visual cues are used. They've been effective for captive primates, dolphins and elephants. Although mirror-like visual images are absent in most field situations, it's possible that individuals learn something about themselves from their reflections in water. But we also need to know more about the role of senses other than vision in studies of self-awareness because some animals for example, rodents who can distinguish among individuals don't seem to respond to visual images. Odors and sounds are very important in the worlds of many animals. Many mammals differentiate between their own and others' urine and glandular secretions, and many birds know their own and others' songs. Moving Jethro's "yellow snow" from place to place allowed me to learn that Jethro made fine discriminations between his own and others' urine (http://www.psychologytoday.com/blog/animal-emotions/200906/hidden...). Perhaps a sense of self relies on a composite signal that results from integrating information from different senses.

[Brian-M]

Originally Posted by Vortigern99

Dogs cannot see color, it's said, so the animals would be unlikely to notice a spot of paint on their face, unless they could feel it.

It's not completely accurate to say that dogs cannot see color...

http://psychlops.psy.uconn.edu/eric/...dogvision.html

Quote:

A number of studies have been done to investigate the color vision of dogs, and the results have been conflicting. However, more recent, better controlled studies indicate that dogs do possess and use color vision, but not to the same degree that humans do. The photoreceptor used for color vision is the cone, and there are cones present in the canine retina. However, they are present in low numbers, comprising less than 10% of the total photoreceptor population in the central area of the retina, as opposed to the human retina which consists of nearly 100% cones in the fovea. Two distinct type of cones appear to be present in the canine retina. One type is maximally sensitive to light in the wavelength that appears violet to people, and the other type is maximally sensitive to light in the wavelength that appears yellow-green to people. Thus, it appears that the visual spectrum of color in dogs is divided into two hues; one in the violet and blue-violet range, probably appearing as blue, and the other in the greenish-yellow, yellow, and red range, which is probably seen as yellow. Light that appears blue-green to people probably appears as white or shades of gray to dogs. Dogs are unable to differentiate colors that appear as green, yellow-green, orange or red to people, and are unable to differentiate greenish-blue from gray. This is similar to people who are red-green color blind. However, one study indicates that dogs are better able to differentiate between subtle shades of gray than people, which would be advantageous in increasing visual discrimination in low light conditions, where insufficient light is available to stimulate cones.

[JFrankA]

I hope this adds to the discussion, just a little memory this topic sparked, for what it's worth. I remember seeing a television show about a mirror test done on dolphins. The show concluded that dolphins recognizes a mirror image for what it is: just a reflection. They go on to say that the dolphins seem to be using the mirror to groom themselves. It was quite a while ago, and I don't even remember the entrie program, just that bit of it.


ETA: Of course, I should just look it up on the internet before I open my big mouth...err...fingers....

http://www.thefreelibrary.com/Dolphi...es.-a075496753

[Molinaro]

I seem to recall it being suggested that most dogs don't take an interest in images on TV because there is no smell to associate with the images. I imagine it's the same thing with images in a mirror. They have no smell, hence they are not of any interest.

[nimzov]

My dog will watch TV for a minute or two if he sees animals that move or if he ears sound of animals. When he was younger he stepped right in front of the tv to "investigate".

[Taarkin]

It's weird that they usually don't react to mirror TV images at all (unless it's the bunny from Monty Python and the Holy Grail, that guy gets my dog in a tizzy). If I detected something with a secondary sense that didn't show up on my primary sense, I would still investigate.

[Modified]

Originally Posted by Molinaro

I seem to recall it being suggested that most dogs don't take an interest in images on TV because there is no smell to associate with the images. I imagine it's the same thing with images in a mirror. They have no smell, hence they are not of any interest.

They take plenty of interest in things on the other side of the window that they can't smell. They can also recognize other dogs (distinguish friend from stranger) through a window.

[Modified]

All the dogs I've had would refuse to look in a mirror. If you put them in front of one, they would look left, right, up, down - anywhere but straight ahead at the mirror. So they seem to know it is not reality and they seem somewhat disturbed by it, but at least they don't think another dog is in there. A parakeet, on the other hand, will spend a lifetime pecking at that other parakeet and never figure it out.

[Modified]

A quick google search suggests that a refresh rate of 60 Hz is too slow for dogs and that they require 70-80 Hz to see the TV as smooth motion rather than flickering images. The newer 120 and 240 Hz flat screens should be more dog-friendly.

[Thabiguy]

I think there's perhaps too much generalization here. I've seen dogs that ignore mirrors and TV, and I've also seen dogs trying to interact with their reflection or images on TV (particularly of other dogs). - I would say it depends a lot on the individual dog, and my guess is just that smarter dogs realize that the picture simply isn't real, so they know better than to waste their time with it - much the same way most humans won't attempt to eat a picture of icecream.

I also think people may be reading too much into "primary" and "secondary" senses. It's not that some senses are somehow intrinsically privileged and more trusted than others, and it's especially not that dogs would use smell analogically to how humans use sight. - Both humans and dogs will use sight to localize/identify distant objects or avoid obstacles, both humans and dogs will use voice and hearing for signalling over distance, both humans and dogs will use smell and taste to assess their food. - Dogs just smell a lot better than we do, so in some situations, it's easier for them to use smell where it would be easier for humans to use another sense. But any of the senses can be used as primary information source when the situation calls for it, in both humans and dogs.

[Modified]

Originally Posted by Thabiguy

I think there's perhaps too much generalization here. I've seen dogs that ignore mirrors and TV, and I've also seen dogs trying to interact with their reflection or images on TV (particularly of other dogs). - I would say it depends a lot on the individual dog, and my guess is just that smarter dogs realize that the picture simply isn't real, so they know better than to waste their time with it - much the same way most humans won't attempt to eat a picture of icecream.

I've heard of exceptions, but I've had five dogs in my life, each of a different breed or mix, and all ignored TV and refused to look in a mirror (not just ignored the images in mirrors, but avoided looking at them period). I've also had six cats, and all would react to TV if there were cats or furry animals on, and all refused to look at a mirror, though not as strongly refused as the dogs.

[Thabiguy]

Modified, if you still have a dog, and a large mirror, here's an interesting experiment you can do: position the mirror so that it allows to see "around the corner", and try communicating with your dog visually through the mirror (many possibilities here - show it its favorite toy, a snack, gesture it to sit, etc.). - Just be careful about calling the dog to yourself, it could easily run into the mirror, break it and/or get hurt. It might be safer to tie the dog to something, or place the mirror sufficiently high above the floor. - Once you get that going, think about various ways the experiment can be modified...

[kritter]

One time I put my cat on a mirror I placed on the rug, she freaked out and ran off. Not sure if it was the fact she saw herself, thought it was a drop, or just didn't like being put onto this new and weird flat shiny thing on the ground.....

[roger]

Originally Posted by Thabiguy

Modified, if you still have a dog, and a large mirror, here's an interesting experiment you can do: position the mirror so that it allows to see "around the corner", and try communicating with your dog visually through the mirror (many possibilities here - show it its favorite toy, a snack, gesture it to sit, etc.). - Just be careful about calling the dog to yourself, it could easily run into the mirror, break it and/or get hurt. It might be safer to tie the dog to something, or place the mirror sufficiently high above the floor. - Once you get that going, think about various ways the experiment can be modified...

When I'm driving my dogs will communicate with me through the mirrors. We have a bunch of mirrors in one bathroom and one of my dogs will look at me in the mirror, then look at the real me with a bit of a startle, then look in the mirror again, etc. They seem to get that it's really me in the mirror. Oh, and a few times my dog Simon got very angry at the dog he saw in the rear view mirror (it was my other dog Koa with her head hanging out the window).

[Lucian]

Originally Posted by roger

When I'm driving my dogs will communicate with me through the mirrors. We have a bunch of mirrors in one bathroom and one of my dogs will look at me in the mirror, then look at the real me with a bit of a startle, then look in the mirror again, etc. They seem to get that it's really me in the mirror. Oh, and a few times my dog Simon got very angry at the dog he saw in the rear view mirror (it was my other dog Koa with her head hanging out the window).

My cat also seems to recognize my reflection in the mirror, though he's disinterested in his own. He likes sinks: he likes drinking water from the sink, and since he's freakishly water-resistance, he enjoys a bit of a frolic in water. He also enjoys getting petted while he's having his happy sink time. Sometimes he will look up at my reflection and "murr" at me

[roger]

Okay, and when is the last time your dog groomed himself if he saw an object on part of his body that he could see? I don't really ever recall my dogs fussing over whether there was some red lint on their paws. Why would they fuss if they saw some on their forehead?

The real test would be to put some chicken up there, except of course they could smell it without the mirror, and probably not recognize it in the mirror anyway.

[GreNME]

Originally Posted by roger

When I'm driving my dogs will communicate with me through the mirrors. We have a bunch of mirrors in one bathroom and one of my dogs will look at me in the mirror, then look at the real me with a bit of a startle, then look in the mirror again, etc. They seem to get that it's really me in the mirror. Oh, and a few times my dog Simon got very angry at the dog he saw in the rear view mirror (it was my other dog Koa with her head hanging out the window).

This correlates with my own experience (and understanding of dog behavior). Dogs, depending on breed (large and giant breeds show this more often), have a distinct sense of "self" but not in the same type of context that we humans have. Basically, it seems to be not in an individualistic sense. The following video illustrates some of what I'm saying (watch it to the end):

YouTube Video This video is not hosted by the JREF. The JREF can not be held responsible for the suitability or legality of this material. By clicking the link below you agree to view content from an external website.

I AGREE

I wish I had the full program on this, because it had more excellent points on this subject of dog behavior.

Also worth noting: different breeds of dogs are going to have different expressions of behavior with regard to visual stimuli. Larger breeds will in general have more feedback, but sighthounds are going to be more interactive with stimuli because of the greater development of their vision centers and their use of sight in their prey drive. Whether that translates to reacting to a mirror, however, depends on factors like the comfort of the dog in a given environment and its level of sophistication in reacting to human items in a given environment (in other words, a dog is more likely to exhibit a behavior at home than in a strange location).

Also, I believe the general consensus is that dogs can see in color, but it's not as vivid as we see or there may be some colorblindness. However, in terms of contrast, some breeds of dogs have exceptional sight, where they can see things that would require weak binoculars for us to see. Essentially, they don't see things like we do, so our understanding of what they see when they look at something is somewhat limited in scope.

[JoeTheJuggler]

Originally Posted by Vortigern99

My opinion on this, as an amateur zoologist and dog- (and cat-) lover, is that dogs don't react to a reflected image past the puppy stage because they recognize it for what it is -- a reflection, or at least, some kind of illusion -- and have no impetus to respond to it. If it cannot touch/affect them and they cannot touch/affect it, it is beyond their interest.

I don't think the mirror test examines what happens after the animal has been exposed to the mirror for a long time, but just their initial reactions. In my own experience, dogs and cats first react to a mirror as if they're seeing another animal. They frequently will look behind the mirror to see where that "other" animal is. After a while, they learn that it's not real--same as with images on TV. ETA: The point is, that at no time do we see evidence that the dog understand the reflection is itself. Yes, it's wrong to conclude from this that the dog has no self-awareness, since it could merely just be something else that accounts for its behavior, like an inability to understand reflective surfaces.

Originally Posted by Vortigern99

As to the "red dot on the face" test, I'd be interested in learning how the researchers achieved this test with a dog. Dogs cannot see color, it's said, so the animals would be unlikely to notice a spot of paint on their face, unless they could feel it. And, even if the animals did see the spot in the mirror, do we expect them to touch it, as human beings do, as an indication of self-awareness, with their paws? I'm not sure I buy that a dog would care if it had a paint spot on its face, again unless it could feel it, which would invalidate the mirror test.

The lack of color perception is certainly not a barrier to doing this test on dogs. Also controls are used to eliminate the ability to feel the spot.

I don't know what we "expect" anything in these tests. We run the tests to see what happens. I don't think hands are required to demonstrate self awareness in the mirror test. Here's a description of how the test was done on dolphins. I think the evidence of self-awareness was in how the dolphin would (or would not) orient itself to see the mark in the mirror.

But again, what sphensic said early on was right on target:

Originally Posted by sphenisc

Ahh... this is like the misinterpretation of the Turing test.

"Passing the test implies a sense of self."

doesn't mean that

"Not passing the test implies a lack of sense of self."

So about all we can say about doing the mirror test on dogs is that we don't see in these tests any evidence of self-awareness. (This is not the same as saying we see evidence of the absence of self-awareness.)

[Amapola]

These experiments are certainly very interesting but to my way of thinking not very important. After all, why should we expect other animals to behave the way humans do? For all we know, dogs are smugly watching us relate to our mirror image and thinking "Fools!".

Each creature is adapted to its environment in the best way it can be. One way is not "better" or "worse" than another way. Just imagine throwing away all your tools and clothes, and trying to survive in the desert the way a rattlesnake does.

I don't think an animal is more valuable, or less valuable, because it is more (or less) like a human. I think we need to value them for what they are, not for what we perceive they should be. It is certainly interesting to conjecture on the comparative abilities of animals but I hope it is not the way of defining our relationship with them. I would also question our methods of testing since they would be necessarily biased towards creatures that are more similar to humans.