21 August 2008

Magpie Self-Reflection

Filed under: Language, Psychology — ktismatics @ 6:39 am

I’ve posted about mirror self-recognition before, both to illustrate Tomasello’s usage-based theory of language acquisition as a social role-taking skill and to present empirical evidence conflicting with Lacan’s theory of the Imaginary. Now researchers in Germany have demonstrated experimentally that magpies can recognize themselves in a mirror.

The experimental protocol is very similar to the one used by developmental psychologists to demonstrate self-recognition in human infants. A bit of colored paint is placed on the magpie’s feathers in a place where the bird can’t see it. The bird is placed in front of the mirror. If after seeing its own reflection the magpie begins attempting to remove the paint from itself, it can be inferred that the bird recognizes itself in the mirror. And this is precisely what the researchers discovered. This link to the study summary contains video clips of the magpies viewing themselves and attempting to remove the paint. Watching Gerti’s vigorous yet futile efforts to peck, scratch, and rub the offending mark off her feathers offers compelling testimony that she understands that it’s herself she sees in the mirror.

Researchers chose magpies for their investigation because other research had found magpies and related species (crows, ravens, jays) to be unusually intelligent. They have unusually large brains relative to their body mass. They’ve been observed in the wild modifying found objects for use as tools for obtaining food. And they’ve shown exceptional social intelligence, being particularly adept at picking up social cues from others. Magpies who rob other birds’ stashes of food take more trouble when hiding their own, a behavior suggesting that they expect their fellows to share the same larcenous intent they themselves have. When a predator encroaches on a magpie flock’s territory, for example, the first magpie to spot it will take up a position near but out of reach of the predator, emit its warning cry, and point its beak in the direction of the predator. When other magpies heed the warning and arrive at the scene, they too will point their beaks. If they pointed in the same spatial direction the original magpie was pointing, they would be showing imitative behavior (impressive enough in its own right for a bird brain). But it turns out the other magpies, arrayed at varying angles relative to the first bird, all point their beaks toward the predator, despite the fact that their eyes, placed on either side of the head, make it impossible to see the predator while pointing at it. By warning one another through pointing, magpies demonstrate that they understand the communicative intent of their own and their associates’ gestures.

In order to understand the other’s pointing gesture, a creature must be able to regard the other as similar to itself, with similar kinds of intentionalities. To follow the pointing finger or beak, the creature must be able to see the world from the other’s perspective, following the other’s line of sight to the object of attention. This intentional social role-taking ability is regarded as an essential prerequisite to language acquisition in humans: eventually the physical pointing is replaced by intentionally gazing directed toward the object of joint attention, and eventually by speech, in which the words are used as symbolic pointers directing the hearer’s attention to the referential object.

Social role-taking is also regarded as a prerequisite to recognizing one’s own image in a mirror. If the other points in my direction, I need to be able to understand that I am the object of the other’s attention. I thus need to objectify myself in order to understand the other’s communicative intent. By regarding myself as an object of the other’s attention, I’m also able to regard myself from outside myself, as if I were looking at myself from the other’s perspective. This self-objectifying ability is presumably what’s required if I’m going to recognize myself in a mirror — an act that requires me to see myself from a distance, as others see me.

Most higher primates, dolphins, and elephants can do it. Some breeds of dog — a subspecies of wolf bred for sociability with humans — can do it. Monkeys and other mammal species apparently cannot. Remarkably, it turns out that at least one species of bird can do it.



  1. This isn’t particularly significant, but I was reading through Metzinger’s Being No One and was reminded of this post. Specifically, at one point Metzinger talks about a disorder called prosopagnosia where individuals lose the ability to recognize faces – including their own in some cases. Visually, all the information is there, but they can’t make the connection that it’s a self-reflection in the mirror. But of course, self-consciousness doesn’t suddenly disappear; they may lose that visual sense of self, but they can still be aware of their symbolic position. Anyways, I thought it was interesting and figured you might too, considering your recent attempts to tackle Lacan with some empirical evidence (which I fully support!).


    Comment by Nick — 18 September 2008 @ 8:53 pm

  2. Nick, I’ve enjoyed reading your blogs — clearly written observations on stimulating topics. Yes, that’s interesting about the prosopagnosia. One could suppose that the ability to recognize faces would have a temporary catalytic effect on the emergence of self-awareness, such that it’s no longer necessary once this awareness has been achieved and consolidated developmentally. I’ve not seen research on congenitally blind people, but I suspect their sense of self doesn’t differ dramatically from that of sighted people. If Tomasello is right, then self-recognition and symbolic language acquisition both rely on the development of a more general social role-taking ability, so being either blind or deaf would affect only the particular sensory manifestations of this general underlying cognitive ability.

    I’m not familiar with Metzinger’s book, but looking at others’ comments it seems that he wants to discount the subjective sense of self altogether as a sort of mistaken identity — that what we think of as self is a kind of brain activity rather than an inner state or entity. It also sounds like a very dense book. Have you posted on Metzinger before? Empirical psychology abandoned self-report data about inner subjective states as being inaccessible to intersubjective validation, but self-report is linguistic and as such is an intersubjective data source like any other verbalization. And we do understand what someone else means when she says I think, I feel, I believe, I imagine, etc. This sort of data wouldn’t pass muster as arche-fossil, but I suppose the attempt to study mind from a perspective beyond thought is intrinsically absurd.


    Comment by ktismatics — 19 September 2008 @ 12:59 am

  3. Thanks Ktsimatics, I’ve been really enjoying your posts on Lacan and Chiesa lately too, even though my mind has been focused on non-Lacanian topics the last little while.

    I’m only about half-way through Metzinger’s book yet (it’s over 600 pages), so I’d preface all my comments about him by saying they’re uncertain at best. That’s particularly relevant because the first half of the book is focused on consciousness, with the second half being devoted to self-consciousness and subjectivity. So the really interesting questions only get answered later. As for density, I’m not finding it terribly dense. He’s very clear and he provides a lot of useful examples from actual psychological studies. A minimal knowledge of phenomenology and analytic philosophy of mind are more than adequate I’d say.

    But to answer your other questions, he’s definitely looking to bring subjectivity into the ambit of neuroscience. So, one of the key questions he wants to answer is how does a first-person perspective emerge – what is it that makes consciousness have a centered perspective in the first place? His general approach so far (and I’m in the process of writing a post about this) seems to basically attempt to update Kant’s transcendental project with neuroscientific knowledge. So instead of conditions for one, single type of consciousness, Metzinger looks at 10 different ‘constraints’ that combine together to form a variety of different types of consciousnesses. In normal, everyday experience, we rely on all 10 constraints to function properly, but in certain neurological disorders some constraints are missing (which Metzinger then uses as evidence that such a constraint must be an independent constraint). So, for example, the constraint responsible for embedded holism (e.g. the experience of a unified book within a unified room within a unified building, and so on) is sometimes missing. In those cases, there is still an overall holism, but individual objects can’t be picked out.

    Now I’m not sure if he’s going to continue this method when he takes on subjectivity (I presume he will, but I’m not certain). The closest he’s gotten so far is the constraint of ‘perspective’ – consciousness is always centered on some particular perspective. This isn’t, however, self-consciousness yet, but merely an attribute of normal consciousness. In order to be aware that it is “I” that perceive, and not just some unified sense of “perceiving this scene”, that requires more, which is what he wants to tackle in the second half.

    As for your suggestion about self-report data, Metzinger uses phenomenological description as the basis for discerning the constraints that comprise consciousness. But he then analyzes the constraints according to a number of other different levels – representational, computational, functional and neurological too. Basically, if consciousness appears in a certain way, we need to take that fact seriously, and therefore use science to find its neurological basis. So there doesn’t seem to be any shying away from introspection on his part.

    I gotta get back to work, but I do hope to write some more about Metzinger in the near future. He’s really interesting, but I just wish the sheer size of his book wasn’t so intimidating to readers (it scared me off for a while haha).


    Comment by Nick — 19 September 2008 @ 10:43 am

  4. Thanks Nick, I’ll continue following your posts on Metzinger.


    Comment by ktismatics — 20 September 2008 @ 6:39 am

RSS feed for comments on this post. TrackBack URI

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

Blog at WordPress.com.

%d bloggers like this: