On Topic (Really!): Parrotspeak / Parrotvision

By Ctein

As I indicated in the comments to my last column, Elmo, my parrot, doesn't use words the way humans do. His listening comprehension is huge and grammatical, but his "speaking vocabulary" is only modestly more sophisticated than a cat's or a dog's. I'm not saying he couldn't learn to speak purposefully; we haven't tried to teach him to. I'm just saying he doesn't.

What I have noticed, though, is that his nonverbal communication is extensive, subtle, and very complex. It's not just about body postures and positions. He does a lot of very elaborate feather flicking and ruffling and repositioning that is clearly intended to convey information, most of which I don't understand. It's far, far more sophisticated than just fluffing oneself up or bristling like a cat does. It looks more like the way patterns change on cuttlefish. Paula and I both have a strong sense that Elmo is "saying" things that we don't understand.

I have a hypothesis why, and here is where it gets photographically interesting: Bird vision is very different from ours. Birds see in four frequency bands rather than three. The bands are more sharply defined than ours (so they can discriminate more distinct colors than we can), and one of the bands is in the near ultraviolet. As best as we can tell, birds process that information in ways similar to the way we do. For example, we confuse yellow with green plus red, because monochromatic yellow stimulates the long and medium wavelength cones the same way green plus red does. Birds have a second color, violet, where that should also occur: a proper mix of blue and ultraviolet should look the same to them as pure violet if they are visually interpreting color in an analogous way. Experiments show that, indeed, they can't tell the difference.

Consequently, birds don't look, to each other, the way they look to us. Budgies, for example, look to us to have dark blue cheek patches and pink feet. In the ultraviolet those are brilliant white (thank you, David Attenborough!), while the rest of the bird is very dark. In a few cases, researchers have UV-photographed wild species where they thought the male and females looked identical. In the ultraviolet they don't look anything alike; possibly all birds are sexually dimorphic, even if we can't see it.

Elmo's emotional state and what he's communicating is different in these three photos.
His "body language" looks very similar to us, but how does it look to him?

So, back to Elmo. Elmo looks uniformly neutral in color, save for a red tail. The "sunscreen" powder on the face is near white, the beak is near black, and the feathers are various shades of medium gray with silvery fringes. I have a strong hunch that's not at all the way Elmo looks in four-band "parrotvision."

More significantly, I have an intuition that the subtle changes in brightness I see when Elmo flicks and ruffles his feathers produce substantial brightness and color changes in parrotvision. I think there's communication going on there that I can't even perceive.

I could test this hypothesis with the right equipment. The Fuji UVIR camera, one of the UV-transmitting lenses, and a set of UV blocking and transmitting filters would certainly let me see what Elmo sees in the UV, and might even let me do a false-color simulation of what he sees full-spectrum.

I don't have that equipment, and I can't afford it. I used to have close enough friends at Fuji that I'd feel comfortable asking them for the loan of that kind of gear without an obvious review being in the offing, but I don't any longer; they've moved on or retired.

So, anyone here have an idea how I might get my hands on the gear? Somebody who knows somebody, perhaps? I figure it can't hurt to ask; you folks are collectively far more resourceful than me.

It might produce some interesting science. It might even produce some interesting art. It might produce nothing of note. Results are never certain, in art or science. It would be fun to find out.

Ctein
(Photos © 2009 by David Dyer-Bennet)

Comments

[Ben Rosengart]

Perhaps you should get in touch with Irene Pepperberg?

[misha]

My favorite breed is the Norweigan Blue.

Amazingly intelligent. They can play dead better than any other animal.

[jr cline]

Great idea. As a parrot companion myself, I'd be interested in seeing your results if the equipment appears from some where.

[Alex Vesey]

Are you up for grant writing? There are a number of organizations that would be willing to fund such a project. You need not purchase the equipment--just buy the time to use the required devices in a research institution.

For example, if you knew of a Japanese research organ that has the items you require, you could approach the Japanese Society for the Promotion of Science. This is just an example--there are other such opportunities.

Alex

[Roger Bradbury]

So the pixel array in a parrotvision digital camera would have pixels with four different coloured filters, instead of three. TV and monitor screens would create colours from red, green, blue and near ultraviolet.

Would this last colour be a sort of Norwegian Blue?

[steve kalman]

Ben's post is spot on. She has an excellent book recounting her research with an African Grey. ISBN: 0061672475

[Randolph]

Hmmm...the Museum of Vertebrate Zoology at Berkeley might be interested.

Also, have you done a literature search? A quick poke at Google Scholar revealed a few articles, but the researchers apparently used spectrophotometers. In any event, I wonder if some of the researchers in the field might be interested in funding some photographic work. There's an Ecology of Vision laboratory in Bristol that has done some work with parrots--they might have funding, or know researchers closer to home.

[Martin Doonan]

That put other ideas in my mind: what do parrots (or indeed other brightly coloured creatures) look like in PredatorVision? And given that ParrotVision is tuned to UV, do they look different to each other in and out of sunlight?
Oh yes, and would the Norwegian Blue have evolved differently for life in the fjords? (Sorry, couldn't help myself.)

[Jim Kofron]

Assuming that you can't get ahold of the Fuji, you could try either the Sigma SD10 or SD14---removing the dust protector/IR block gives you a camera with full spectral sensitivity from UV to IR. From there, you just need a lens (M42 mount) that is decent in the UV, and an appropriate filter that cuts visible and IR. I used to have most of that stuff for the SD10, but I sold it off. I've got a M42 adapter that I can loan you if you want to give this a try.

The nice thing about this setup is that you can convert back and forth from a 'regular' camera to a UV or IR sensitive camera with relative ease...

[Oliver Francks]

It's a parrot.Parrots are not by any stretch of the imagination one of the more intelligent species that inhabit this universe. They have failed entirely to invent anything etc. Anthropomorphising your parrot may make you feel better about keeping it prisoner but has no real grounding in reality.
If the parrot were communicating, however, it would surely be just saying "Let me go. I'd rather take my chances in the hostile world than put up with this tedious life."

[Roger Moore]

You might try to do without the fancy gear. Bjorn Rorslett has an excellent page on "All You Ever Wanted to Know About Digital UV and IR Photography, But Could Not Afford to Ask", where he discusses some cheaper ways of achieving decent UV results. In particular, he mentions that some of the older Nikon cameras (D2H and D70) have usable UV sensitivity, and he discusses which old lenses have good transmission. My impression is that the hardest part is finding a good UV filter.

[Pak]

Googling quickly (and you may have done this already). There's this tutorial here where someone uses filters to do this:

http://www.naturfotograf.com/index2.html

It's interesting that his photos of flowers do seem to show IR / UV response -- he hasn't put much in terms of emperical testing but the photos do show some response of UV.

And there are various tutorials on the web (and indeed services) where you can modify your camera to have IR response by removing the IR cut filter.

However, none of these resources are likely to generate what Elmo *really* sees :)))

Pak

[Richard]

I guess that begs the question of whether we see color the way it should be seen, what is natural color. On a separate note I have been told that animals are not able to see a flash as it is too fast for them. I tell people in the studio their dog is reacting to the sound of the flash rather then the light.

[Daniel Fealko]

Ctein,

Have you considered proposing the idea to KickStarter? They claim that "Kickstarter is a funding platform for artists, designers, filmmakers, musicians, journalists, inventors, explorers..." It might be worth the effort of a proposal. It might also be a way for TOP readers to help fund your investigation/exploration.

I have no experience with this organization, but thought I'd send their link your way in case it's of use.

Dan

[Luke J. Slater]

Bjørn Rørslett at naturfotograf.com has some cool ideas for home-made UV setups. But I suspect that it would be hard photographing a living creature at close range with black filter(s) over your lens.

[YS]

Alex was really neat. He'll be missed.

PS. Sorry Ctein, can't help you with the UV equipment there.

[Jeff]

"I could test this hypothesis with the right equipment."

Well, I would suggest that you don't use a Leica.

[V.I. Voltz]

The first thing to do is ask someone like Justin Marshall from The University of Queensland: http://www.uq.edu.au/uqexperts/profile.php?staff_id=6007 if someone has already worked on this. Given the popularity of african greys as experimental animals for behavioural studies, I'd be surprised if the work has not already been done.

If it has not, you should publish your findings.

Voltz

[fjf]

I wonder if Elmo ponders if he should spend more time teaching you and Paula the details of hair flicking and ruffling and repositioning. Is it possible he senses that the two of you are communicating in ways he does not yet understand?

I am not sure if this is a serious question or not. The parrot side of me seems to think it is. The human side of me has a "beakish" sense of humour and thinks not.

[Mark M]

@Richard: I was just reading a book review in the NY Times ( http://bit.ly/7K0on ) that suggested dogs actually have a higher “flicker fusion” rate that humans, which would suggest that a flash is not at all too fast for them to see.

@Ctein if you think parrotvision in cool, consider shrimpvision. The purple spot mantis shrimp can see IR, UV and also distinguish the direction of polarized light. ( A little toward the end of this post: http://www.photo-mark.com/notes/2009/apr/13/abyss-uncertainty/)

[Andrea B.]

There is a Sticky over on the photography forum Nikongear where I listed a lot of info about UV-capable cameras, lenses and filters.
You need not be a member to view it at
http://nikongear.com/smf/index.php?topic=8355.0
The camera info is a bit Nikon-biased given that it is a Nikon forum ;-) .

The best UV-Pass filter currently known is the Baader 2" made by Baader Planetarium in Germany. (It is available here in the U.S. at stores that sell astrological equipment and filters.)

Most other UV-Pass filters have an IR "bump" that leaks enough IR to cause contamination in the UV photo, but the Baader UV-Pass filter leaks almost no IR.

Given the longish exposures needed for UV photos, your biggest problem will be illumination. There are UV flashes available and UV LED flashlights, but one has to be careful with the damaging effects of UV. It might not be a good idea to use artificial UV illumination on a living Elmo ! Of especial concern would be potential damage to the eyes. However, strong sunlight will work just fine for UV photography, so placing Elmo by a bright window will probably work.

[Andrea B.]

Additional Note: Robert DeCandido in New York City has done some work on UV vision/appearance in birds. You can probably contact him through his website at
http://www.birdingbob.com/index.html

[James]

Ascribing to Elmo an aspiration to escape a tedious life and take his chances out in the big bad world (or any conscious state we can imagine or sympathize with) is equally to anthropomorphize him.

For a fascinating foray into animal perception, see Richard Dawkins's writing on bat echolocation. Bats use ultrasound in a manner analogous to our use of light - to perceive and map three-dimensional space:

'The bat's [sonic] sensation of her mate may be no more different from my visual sensation of a flamingo, than my visual sensation of a flamingo is different from a flamingo's visual sensation of a flamingo.'

And the essay by Thomas Nagel, 'What is it like to be a bat?' is essential. The human imagination only stretches so far.

On the other hand, false-colour simulations of various animals' visual fields are something I'd like to see more of. Good luck with your investigations, Ctein.

[Daniel Fealko]

@Andrea --

"However, strong sunlight will work just fine for UV photography, so placing Elmo by a bright window will probably work."

I believe window glass filters out UVB and transmits UVA, unless your home windows use film to filter this too. I know one has to account for attenuation in the UV band through glass as a result of a Physics thesis I performed many years ago. I had to go with quartz so as not to attenuate the UV of interest.

Maybe Elmo would prefer just to be out in the sun, assuming his wings are clipped.

[Andrea B.]

Daniel writes: I believe window glass filters out UVB and transmits UVA, unless your home windows use film to filter this too.
Daniel's comment serves as a good reminder to test your UV equipment on known UV-patterned subjects using the chosen method of illumination before performing your actual experimental work.

And, indeed, as he writes, some window glass may filter out some or all UV. But my older (1965 ?) windows in my house do not filter much UV, and I have made many UV photographs using my natural window light. As UV exposures tend to be longer, it is nice to sometimes shoot floral subjects indoors where the breezes can't toss them about and turn the 1" second exposure into a motion study. :-)

My favorite UV-patterned subject used for testing UV equipment is the Black-eyed Susan (Rudbeckia hirta) which features the well-known UV-dark 'bullseye center' and UV-bright petal edges. Other folks use chemical powders of known UV reflectance or coated swatches.

Most recent camera lenses are constructed from glasses and coatings that will block UV and IR light from passing. Dedicated UV-Pass lenses generally feature quartz glass and no coatings. However there are many older non-coated 'normal' lenses that will also pass enough UV to make basic UV exposures possible --- although the exposure time can be quite long. I've made some UV exposures with the plain vanilla Nikon 50mm f/1.2 AIS lens on my modified D200 (internal sensor filter removed). The old Nikon EL-Nikkor line passes UV fairly readily -- especially the 63mm f/2.8.

[ctein]

Dear Martin,

Good questions! Mammalian predators only see in two colors, except for the primate strain that human beings are part of, so it probably doesn't have much bearing on that. Avian predators, though, have similar vision to Elmo's. (Which brings up yet another question, which is whether the obligate carnivore avians see similar wave bands to the herbivores and opportunistic omnivores. I have not a clue.)

UV light is extremely heavily scattered (Raleigh scattering goes as the fourth power of frequency) so there would be a high UV component in direct sunlight and in the shade.

----------------------------------------

Dear Pak,

Yes, flowers look extremely different in the ultraviolet, because many insects see ultraviolet. Many otherwise ordinary looking flowers practically have a landing grid painted on the petals in ultraviolet, for the benefit of pollinating insects.

~ pax \ Ctein
[ Please excuse any word-salad. MacSpeech in training! ]
======================================
-- Ctein's Online Gallery http://ctein.com
-- Digital Restorations http://photo-repair.com
======================================

[ctein]

Dear Richard,

Yes, that question is pointless because the phrase "natural color" is entirely devoid of meaning. "Color" is a psychophysical perception; it's a brain's response to an object's spectral reflectance, evaluated in the context of the surrounding spectral reflectances. It is a reaction derived from objectively measurable data, but the reaction has no objective reality.

Example: I could offer up to you three scenes that contained a yellow ball. In the first scene, the ball was reflecting spectral yellow. The ball in the second scene was reflecting red and green wavelengths and no yellow whatsoever. And in the third scene it was predominantly reflecting blue light. They would all look yellow to you and none of them would be optical illusions: you would be perceiving the color "correctly."

This is actually photographically relevant, especially to folks trying to truly understand color photography and how to do it well. People who say that they want film or a camera that portrays color "correctly" often (erroneously) think they're making a statement about objective reality. What they're really asking for is a camera that sees colors the way they do.

And that's staying within the human visible spectrum. Wander outside of that and all bets are off. For instance, there are deep sea shrimp that see only infrared. Many fabric dyes to look black to us are highly reflective in the infrared -- one of the things that leads to the infamous IR problems with poorly filtered cameras. George Orwell, please take note: the shirt that looks black to us looks white to these shrimp!

And I would not have any idea what colors look like to those shrimp that see something like a dozen different spectral bands!

~ pax \ Ctein
[ Please excuse any word-salad. MacSpeech in training! ]
======================================
-- Ctein's Online Gallery http://ctein.com 
-- Digital Restorations http://photo-repair.com 
======================================

[John Taylor]

hmm Mr Francks is seemingly unaware of the linguistic and cognitave research that has accumulated around parrots of late… heck even a perusal of a recent issue Nation Geographic magazine devoted to animal intelligence could be illuminating. :)

[daleX]

Oliver Franks wrote:
> Parrots are not by any stretch of the imagination one of the more intelligent species that inhabit this universe.

Can't speak for the universe, but for this planet, this statement would certainly arouse the disdain of the scientific establishment. The more intelligent parrots, like the more intelligent canines, have the intelligence (such as humans reckon it) of a 2 1/2 year old human child, putting them very much in the elite. A common misconception is that if a particular animal's brain is so much smaller than a human's, it's intelligence must be correspondingly less. In fact, the amount of fatty sheathing needed to shield neurons from one another increases non-linearly with brain mass.

> They have failed entirely to invent anything etc.

I live with a couple parrots who would take great exception to this thesis. Parrots live in a world of sound; look there to find their inventions.

> If the parrot were communicating, however, it would surely be just saying "Let me go. I'd rather take my chances in the hostile world than put up with this tedious life."

Unfortunately, even that door is closed for most pet parrots. Relatively few living captive parrots in non-tropical countries were taken directly from the wild. Instead, they are the offspring of mated breeding pairs, themselves often several generations removed from the wild. All even vaguely competent breeders socialize young birds, which normally results in their imprinting on human beings instead of their own kind. For this reason a human-socialized parrot is typically useless in a breeding situation, being sexually attracted to human beings of the opposite sex and entirely unresponsive to parrots of its own species. This is just one example of the many ways in which captive-bred, socialized parrots are unequipped to deal with life in the "wild".

In my experience parrots are much more tolerant of tedium than humans and derive comfort from predictable routines. But there is a level of boredom and sensory deprivation that passes even the tolerance of pet parrots, and this is unfortunately all too common. Combine that with the need of a flock animal for the security of knowing that other flock members (whether human or parrot) are within calling distance, and you indeed have a situation in which many pet parrots are enduring situations of sometimes life-long torment. Any competent parrot owner, such as certainly Ctein must be, knows enough to provide an environment that is rich in activity and interactivity.

[Half Sigma]

African Grey parrots are considered the most intelligent species of parrot.

As far as the morality of keeping them as pets... they're still animals. Humans keep animals as pets, it's what we do. Lions eat other animals, it's what Lions do.

[ctein]

Dear Half Sigma,

So you're claiming that whatever humans do isn't subject to moral examination?!

I 100% disagree with everything Oliver said. But that doesn't make asking about the ethical or moral implications of our actions meaningless.

pax / Ctein

[jeux 3d]

I believe that it is hard to shot photograph of a living creature at close range with black filter of your lens. You have done great work. Thank you for sharing pictures here. The Norweigan Blue becomes my favourite breed now.