Wednesday 7 July 2010

To cernuate or not to cernuate... (cernuation III)

It would appear that my previous explanation of cernuation left something to be desired, so I will try again. When I coined the word, I was thinking about a mode of locomotion invented by the people responsible for 'The Future is Wild' (TFIW). The squibbons, descendants of cephalopods, have taken to the trees and swing from branch to branch from their tentacles. That could be just what a tentacle might be good at, as it mostly involves tensile forces; tentacles are not well fit to withstand compressive forces needed for walking.

Brachiation is what comes to mind when gibbons or monkeys swing from branch to branch: they hang from one arm while the other swings to grab another branch. While they do so the body stays largely upright, meaning the head is always above the feet.

The way squibbons do this, cernuation, has similarities: the animal swings suspended from one or more arms while other arms move forward to grab a new branch. But the movement is like an inverted somersault, so, while one swing sees the head upwards, the next has it downwards. I presented a fragment of 'TFIW' showing a squibbon moving that way, but the lighting was not very even so the movement may not have been well visible. This post rectifies that.

Click to enlarge; copyright Gert van Dijk

Above you see a rough model of a cernuator. It is only a simple ZBrush model, not meant as a proper animal (I hope my ZBrush friends will forgive me). The animal has very long eye stalks, long arms and long legs, and the head is kept downwards so the eyes stay near the centre of the body; the body is even bent to make room. Having the eyes near the centre was borrowed from the squibbon; this particular concept is stressed in the book 'TFIW'. The idea is probably that this eye position minimises the vertical distance the eyes travel over during cernuation. We'll get back to that.

Click to enlarge; copyright Gert van Dijk

This image shows part of the movement cycle, starting when the body hangs vertically from the arms, and ending where the animal can grasp another branch with its feet.

Click to enlarge; copyright Gert van Dijk

The next diagram starts at the same stage, and ends at the point where the hands are ready to grasp a branch. Note that the animal is upside down during this part of the movement, and moves with its back towards the direction of swing.

Click to enlarge; copyright Gert van Dijk

And finally the animal swings from its hands right up to the point where the cycle began. Notice that the body is now right side up again, and the back faces, well, backwards. That's cernuation.

As I wrote before, I admire the ingenuity that went into its design. The more I think about it though, the more I start doubting how well it would work. There is no mechanical problem at all in swinging like this. Instead, the problem is one of motor control and of visual perception. It is amazing that animals like gibbons actually manage to travel through woods at high speeds. Just imagine how much more difficult their task would be if their image of the world not just zoomed forwards as well as bobbing up and down, as it does in brachiation, but also rotated constantly as it must in cernuation. Have a look at the animal's head above: it is actually turned upside down at one point. How do cernuating animals manage to pick out the next branch?

Click to enlarge; copyright Gert van Dijk

That must have been the reason why the squibbon's designers situated its eyes near the centre of mass. The picture above shows all stages in the movement, with a black line indicating the position of the eyes. Compare that to the position of the feet or the hands (blue line), and you will see that the centre of the body moves less vertically than any part far away from the centre. You may now also understand why the squibbon's eyes stick out sideways: if not, it would not be able to look forwards at times because its body would be in the way!

Click to enlarge; copyright Gert van Dijk

While this eye position minimises vertical displacement, it does nothing to solve the problem of the eye rotating, and with it the animal's view of the world. Perhaps that can be rectified. Let's suppose the head can be rotated by about 180 degrees. Have a look at animal A shown here. Hanging from its hands like this it will view the world as being the right side up. If it were to change the position of its head as in B, it would see the world upside down, right? That is not at all useful, unless of course the body itself would be upside down. Obvioulsy that is the case while cernuating, so if we take animal B, freeze it and rotate it, head and all, you get animal C. The trick would be for the animal to rotate its head from one position to the other quickly when the point of contact changes. During the swing, the animal could then always keep the head in the same vertical position so it could see what it is doing! I do not know whether squibbons were supposed to this by rotation of the eye stalks, but why not...

Mind you, I think brachiation is by far the easier solution, and I do no intend to fill Furahan forests with screeching cernuators. Still, it is interesting to think about, isn't it?

This is probably the point where the rift between those who say 'yes' and the vast majority of mankind is revealed...


Evan Black said...

Still, it is interesting to think about, isn't it?

This is probably the point where the rift between those who say 'yes' and the vast majority of mankind is revealed...

Yes!!! I've always thought 'cernuation' was interesting, but may be too complex to happen very often statistically speaking. Perhaps squibbons would actually just resort to brachiating, dedicating their other limbs to something else, like carrying loads. Nereus has brachiators, but as with many of my species, it has some unusual center-of-gravity issues that may make occasional cernuation a viable option.

j. w. bjerk said...

The useful perception of the world from a cernuator's rotating perspective would certainly not be something *my* brain could handle, but i don't think it's impossible or implausible

Various animals take in a lot of weird and high speed sensory information, and manage to turn it into something useful. In my layman's opinion i don't think its a greater mental/sensory feat than a bat turning minute differences in the timing of echos into a picture of 3D space, for instance.

A cernuator would just need a brain wired to understand a rotating world.

Luciano N. Ribeiro said...

Maybe some extra eyes on the back would be useful for these cernuators :)

Sigmund Nastrazzurro said...

SN, for Metalraptor:

Because my computer will not let me post on your blog once again, I thought I would leave my comments regarding the plausibility of cernuation. In living animals, information taken in by the sensory organs oftentimes has to go through quite a bit of adjustment before it can be interpreted by the brain. One often-cited fact is that our eyes actually interpret the world upside down because of the way our nerves are wired, and the image received is later flipped the proper way up by the brain. It shouldn't be that hard to build a system (maybe tied in with gravity sensing organs like statoliths?) that screen the image received by the eyes and make sure its flipped the right way up. Or perhaps a solution similar to mantis shrimp, which I have heard at one point contain some of the sensory equipment that we would have in our brain in the eyes themselves, to get around the problem of sticking the necessary nerve tissue in a limited-size arthropod carapace. In our hypothetical cernuator, some of this tissue would be devoted to the sole purpose of orienting the world the "right way up". The reason we probably don't see any cernuators on Earth today is that all of our brachiating organisms are vertebrates, whose structures are built around the concept of bone, a structure that is built to resist compression forces much better than tension forces. Stick a terrestrial mollusk (perhaps similar to my suggested proto-terrasquid, with several normal legs derived from tentacles in the "Walking With Tentacles" fashion and several more normal tentacles for manipulation), and you might end up with a cernuator.

-- Metalraptor
Ask not for whom the allosauroid hunts, he hunts for thee.

Sigmund Nastrazzurro said...

J.W.: I, too, think cernuation is possible, particularly with the visual adjustments I described here. I just do not think it is likely: my predilection for brachiation stems from a hunch that, once you have two legs to swing from, brachiation is easy to start evolving, and afterwards there is no need for the more complex system that is cernuation.

Luciano's and Metalraptor's ideas both deal with other ways of handling visual information. However, do not underestimate the cost of high-resolution visual processing. If you were to add eyes to Earth vertebrates and have them process the visual information as well as is done for their extant eyes, you are looking at very considerable increase in nervous tissue, and that is very expensive in metabolic terms.
There is probably a reason that visual systems are specialised: a cat's visual sharpness/resolution is a lot worse than ours, but it can see much better in dark conditions. We do neither we nor cats have excellent colour vision, high resolution as well as excellent dark adaptation? It may be too expensive to have it all.

Luciano's and J.W.'s solutions might benefit from 'switchable processing'. In Luciano's case, the information of one set of eyes could be given preference at will over that of the other eyes' input. In J.W.'s case, retinal output should be switchable, so that it reaches a particular part of the visual cortex at one body position and another part at another position. I have never heard of such a system; instead, specific retinal regions map to specific cortical regions, which is how the brain can tell (together with information of the rotation of the eyes in the head and the position of the head in space) where perceived objects are in space. Lacking actual examples of 'switchable' wiring patterns, I doubt you can take neurobiology this far.

Finally, the retinal image is indeed upside down, just like a camera's. But this is not a problem that has to be solved. The cortical map of the world is also upside down as well as immensely distorted. That is not a problem either: as long as there are ways to map visual information to perceived 3D space, there is no problem. The motor cortical map is also upside down and completely distorted, and right and left are widely separated. No problem, as this does not impair motor control.

The idea that the upside down projection is a problem may historically stem from the idea that there is a 'me' looking at these brain processes. Instead, we ARE those brain processes.

j. w. bjerk said...

Sure, i agree that cernuation is not very probable, and the visual processing is probably expensive.

Andy said...

Something struck me while I was looking at the posted images on the cernuation loop and reading your text. Whilst the rotation causes problems design wise I think that in practice the actual vertical displacement of the centrally located eye stalks would be limited. The animal is highly unlikely to be at full stretch between the branches (and so accentuate the vertical movement of the head/eye position while cernuating) and much more likely to be swinging with limbs bent to a certain degree, which would maintain the 'head' at a reasonably steady position. Full stretch would probably only be used in fight or flight mode. I think it's definitely possible.

JoshT said...

I love it when you invent new terms!

Louis Corbett said...

in the "adding oditty:alien plants" blog, you said that you couln't find a good use for eyes in plants. well, i have a theory: it is possible that the "eyes" are actually a form of mouth, maybe a sponge-like organ that sucks up neutrience, like a fly's proboscis.

Square Cow said...

Why don't we have a eye on each hand and then only look out of the hand that is hanging onto the branch. That eye (or pair of eyes) would be still so the visual information would be easy to process.

The process works like this. The squibbon would hang onto a branch with two hands. While looking through still eyes it would reach up to a new branch. Then it would start looking out of the eyes on the new branch, and let go with the old hand. Then it would repeat.

Sigmund Nastrazzurro said...

Square cow: that would work wonderfully in theory. Eyes on Earth are always located near the brain, in whichever animal you chose that has a well-developed visual system. If that 'law' is universal, that the eyes on the hands would need to be equipped with a nearby brain. The ensemble might become too heavy 9and might also be prone to damage). I like the out-of-the-box nature of your idea.
(But not many people will see it here, posted on post over two years old.)

Unknown said...

I don't think the brain self-correcting would be too unlikely, as I have actually experienced that. I remember when I was fairly young and laying sideways on a sofa and looking out, but suddenly realizing that I was looking at the world sideways but still my brain interpreted the information so the world looked upright.
An animal that has evolved to translate this information really quick wouldn't really mind being upside down... Or maybe an animal that actually doesn't get motion blur.
It could "blink" really fast during a swing, creating a film-effect. (the blurring would be a problem I suppose... Hm.

The movement itself would be kinda dangerous in the sense that you would have difficulty stopping yourself from swinging forward -> hurting yourself on other nearby obstacles. Also swinging on something like that might stress a branch more than transferring your weight on it like a monkey might do.

Oh well, ramble bramble. I hope I'll be able to check the response somehow.

Unknown said...

Derp, that's what you get when you don't see your entire message while typing it. Not sure if it's possible to eliminate motion blur, since I'm not too sure what causes that! Maybe it's the brain being unable to parse all the information received quickly enough -> blur?