Even pigeons can recognize "bad" art

i-482042ec4bfabedefc91eff610dc35ac-dn17383-2_300.jpg

. . . at least according to a Japanese researcher, who trained them to differentiate "bad" and "good" children's art.

According to New Scientist,

This isn't Watanabe's first efforts to teach art appreciation to pigeons. In 1995, he and two colleagues published a paper showing that pigeons could learn to discriminate Picasso paintings from Monets - work that earned him that year's Ig Nobel prize. New Scientist plays no role in selecting winners, but Watanabe's latest study make a strong case for another award.

Zing.

Of course, Watanabe first had to determine what was "good" or "bad" art. He basically used grades to do this: "A" paintings, like the panda painting at the top of this post, were deemed "good" art. (I have to quibble with this: although this painting is skillful for a child, it is by no means "good art" by any measure I'd use).

Here's an example for "bad" art (C or D grade art):

i-d11202e5ae9789bdc329b182869393f3-dn17383-3_300.jpg

I think it's more interesting than the "good" art, honestly!

What we really have here is pigeons being taught to apply K-12 art teachers' standards, which are probably based largely on skill with the media, whether the subject matter is recognizable, and degree of realism. That's different from distinguishing "bad" vs. "good" art.

What do you suppose these pigeons would do with a Pollock or a Picasso? We don't know, because although Watanabe taught pigeons in his previous study to distinguish Picassos from Monets, he didn't ask whether the pigeons thought either was "good" art. All we know is that pigeons were able to discriminate the A-level art from C- or D-level art to a statistically significant degree.

Watanabe says,

The results suggest that the pigeons used both color and pattern cues for the discrimination and show that non-human animals, such as pigeons, can be trained to discriminate abstract visual stimuli, such as pictures and may also have the ability to learn the concept of "beauty" as defined by humans.

Yikes. Pigeons can appreciate beauty? I think that's a massive overstatement. Human beings (aside from parents, of course) can generally agree on whether a child's artwork is skilled or unskilled, which is what these pigeons are doing. But we apply a very different standard when we ask whether something is beautiful. What is the relation between beauty in art and the quality of the art? Specifically, can "good" art be ugly? Can beautiful art be "bad"? Can ugly art, paradoxically enough, be beautiful?

I think a valid case can be made for "yes" to all three questions, as strange as that sounds. And unfortunately, I don't think pigeons are going to help us move forward in our understanding of either "beauty" or "good art."

Watanabe does say, "The experiments demonstrated the ability of discrimination, not the ability to enjoy painting." But here's the problem with that disclaimer. "Enjoying" art is yet a different concept again from "appreciating" art, which is still different from "discriminating" art. I do not "enjoy" The Scream, nor anything by Picasso or Pollock. On the other hand, I enjoy the paintings of Hopper very much. But I can appreciate that Picasso, Munch, and Pollock created works with meaningful aesthetic content, and that they required skill to create - just as much as a Hopper. Based on the information we have, pigeons can discriminate between a Pollock and a Hopper, and that's about as far as it goes.

I don't like that this paper conflates so many different aspects of the human response to art. Plus, I think it confuses the situation with regard to animals, and whether they have the capacity to create, appreciate, and enjoy art - which is an open question. If we really want to know if animals have those capacities, or learn what animals perceive as "beautiful," we should probably be asking chimps, elephants, dolphins, parrots and ravens - not pigeons. At least that's my opinion.

Via BoingBoing

More like this

This seems similar to previous research in pattern recognition in pigeons. Basically, behaviorists such as Herrnstein have taught pigeons to "recognize concepts" such as "person" or "tree" by reinforcing pecking that agreed with the human-defined categories. Pecking at photos that had people (or trees) in them was differentially reinforced--in particular, pecking at the area of the picture that contained the people was reinforced. Pigeons reliably discriminated--that is, they reliably pecked in agreement with the human-defined categories.

That is what they are doing here. The good and bad art are human-defined categories; that the pigeons are able to reliably agree may simply mean that human appreciation of art is not nearly as special as we may think it is.

I note that you quote Watanabe as saying "may also have the ability to learn the concept of "beauty" as defined by humans", but immediately object to the notion that "[p]igeons can appreciate beauty." The two are not at all [necessarily] the same, of course. To be able to discriminate X from Y does not imply an appreciation for the difference. The trick, of course, is in discerning what exactly an appreciation entails.

Not having yet read the paper, I cannot meaningfully comment on your last paragraph. It is entirely possible, given your equating Watanabe's statement with "appreciation", that it is not the paper that is doing the conflation. It could be that pigeons can discriminate between this particular operationalization of good and bad art; that does not mean they necessarily appreciate it. Of course, it is also possible that we humans like to exaggerate our own appreciation of art, and we simply don't like being reminded that what we are doing could be accomplished by a pigeon, with sufficient training.

Umm... sorry. On reading my comment, my last paragraph sounds much too snarky. I do not mean it that way, I promise. I say "you", but of course you are saying what the vast majority would say, honestly and sincerely.

Anon, you have missed my point entirely. Recognizing concepts is one thing, but here's what the abstract of the paper says, and I quote in detail so there is no ambiguity:

"Humans have the unique ability to create art, but non-human animals may be able to discriminate âgoodâ art from âbadâ art. In this study, I investigated whether pigeons could be trained to discriminate between paintings that had been judged by humans as either âbadâ or âgoodâ. To do this, adult human observers first classified several childrenâs paintings as either âgoodâ (beautiful) or âbadâ (ugly). Using operant conditioning procedures, pigeons were then reinforced for pecking at âgoodâ paintings. After the pigeons learned the discrimination task, they were presented with novel pictures of both âgoodâ and âbadâ childrenâs paintings to test whether they had successfully learned to discriminate between these two stimulus categories. The results showed that pigeons could discriminate novel âgoodâ and âbadâ paintings. . ."

I posit in this post that Watanabe should not have called the art "good" or "bad". Those terms are far too subjective, value-laden, and complex to throw around casually. "skilled for a human child"/"unskilled for a human child" would have been an accurate description of the dichotomy in question. But by using "good" and "bad," Watanabe *deliberately* takes the paper out of the realm of discrimination of different qualities like texture and color and contrast, into discrimination of concepts that even human beings can't agree on - what is "good" and "bad" art.

You say "that the pigeons are able to reliably agree may simply mean that human appreciation of art is not nearly as special as we may think it is". No, the pigeons are able to reliably agree because they are recognizing basic visual features common to "skilled art for a human child". That really doesn't surprise me much. But it doesn't say anything about what makes art "good." In my opinion, the "good" panda painting is not "good" art. It's "skilled art for a human child" because it demonstrates control of media, realism, proportion, and other basic fundamental of art. It would not, however, have a place in a museum, nor I think, in most people's homes. Therefore I'd argue it is not "good" art at all. Whether or not the pigeons have any appreciation for what I would consider "good" art, I can't say using this data, because Watanabe didn't really test it - and it would be very hard indeed to test that, since not even adult human beings can reliably reach consensus on whether a significant proportion of art is "good".

You say, "I note that you quote Watanabe as saying "may also have the ability to learn the concept of "beauty" as defined by humans", but immediately object to the notion that "[p]igeons can appreciate beauty.""

What I object to here is that what Watanabe does is equate the concept of distinguishing "skilled art for a human child" with distinguishing "good art" with the potential for distinguishing "beauty." None of those three concepts is the same to a human being, so why should they be the same for a pigeon? I have no idea if a pigeon can appreciate beauty or not - I say at the end of the post that animal aesthetics is an open field - but it really doesn't matter, because Watanabe wasn't investigating "beauty." Now, if he had been using human faces with different degrees of symmetry, and trained pigeons to recognize more symmetrical human faces - which humans do reliably agree are more "beautiful" than asymmetrical human faces - then he would have started getting into the realm of "beauty," and I think he'd have a point. But I think using children's schoolwork just complicates the whole situation unnecessarily.

You say "To be able to discriminate X from Y does not imply an appreciation for the difference." Actually, I completely disagree. Re: the definition of appreciation (see http://www.thefreedictionary.com/appreciation, for example), appreciation = "sensitive recognition of good qualities, as in art" or "Awareness or delicate perception, especially of aesthetic qualities or values." "Discrimination" is listed as a synonym. I'm not talking about "appreciation" like "I appreciate your kindness;" that's a different use of the word.

You also say "Of course, it is also possible that we humans like to exaggerate our own appreciation of art, and we simply don't like being reminded that what we are doing could be accomplished by a pigeon, with sufficient training." I'm not sure what you mean by the first sentence, since I think most people would argue they have a fairly tough time discriminating "good" and "bad" art. At least that's what I hear whenever I go to the National Gallery and people are standing around going, "I just don't get it."

As for your second sentence, based on this study, pigeons CAN appreciate some of the fundamentals of art. That doesn't bother me in the slightest. What actually bothers me here is not the pigeons' ability to do anything, it's Watanabe's use of terms like "good," "bad", and "beauty", which are simply fraught with subjectivity, and as far as i'm concerned, hype his results far beyond what he actually has measured.

Perhaps this work will herald the use of pigeons to replace teachers in marking students' work. Imagine the savings at exam time!

Bob, you're just ASKING for an argument over whether modern art is good art! Or maybe it's the pigeon that's asking... ;)