What a 23% longer torso looks like

September 20, 2009

Just checking: no-one’s bored of brachiosaurs yet, are they?

Thought not.  Right, then, here we go!

Greg Paul’s (1988) study of the two “Brachiosaurus” species — the paper that proposed the subgenus Giraffatitan for the African species — noted that the trunk is proportionally longer in Brachiosaurus than in Giraffatitan due to the greater length of its dorsal centra. Paul (p. 7) stated that the difference is “25%-30%” on the basis of his figure 2.

Having seen the dorsal vertebrae of the type specimens of both species, my gut reaction was that the difference was nowhere near this great, so I recalculated it for myself (Taylor 2009:table 3).  Dorsal column length is the sum of the “functional length” of the centra of the dorsal vertebrae, where functional length is the length of the centrum not counting the condyle (which of course is nestled in the preceding vertebra’s cotyle when the column is articulated).  For Brachiosaurus, Riggs (1904) did not give this measurement, but did give total heights, and using these for scale I was able to measure the functional lengths from his plate LXXII.  For Giraffatitan, Janensch’s (1950:44) superbly comprehensive table supplied measurements for D4 and D8; for D11 and D12 I was able to determine the length by measuring from Janensch’s (1950:fig. 62) figure, knowing the height from his table; and for D5-D7, D9 and D10, I interpolated linearly between the measurements that I had.  Summing the functional lengths of D6-D12, I got 226 cm for Brachiosaurus and 183 cm for Giraffatitan.  So Brachiosaurus is 226/183 = 1.23 times as long as Giraffatitan — in other words, 23% longer, which is pretty much what Greg Paul said.  So I learned something there.

(Yes, brachiosaurs probably had 12 dorsals.)

So: is a 23% longer torso a big deal?  Back when I was trying to answer that question for myself, I figured it would help to take an image of a familiar animal and stretch it — so here is a horse, stolen from here and stretched:

Horse (top); and evil mutant horse with 23% longer torso (bottom).

Horse (top); and evil mutant horse with 23% longer torso (bottom).

To me, that second picture is wrong enough to hurt my eyes a little; your mileage may vary, but I suspect those among you who love horses will feel ill when you look at it.  This image was one of the reasons — one of many — that I concluded that generic separation was unavoidable.

But here’s an odd thing: tonight, for this blog post, I did the same thing to a human body, expecting it to seem even more horrible in light of how familiar we are with our own bodies.  Here it is:


Flayed Homo sapiens in orthograde anatomical position, from Vesalius (1543) "Tertia Musculorum Tabula". Modified from Wilson (2006:fig. 1). Left, as drawn; right, with torso elongated by 23%.

To my surprise, the elongated human doesn’t look appallingly wrong to me.  It doesn’t look right, of course, but it seems within the realms of, for example, what might appear as a representation of a human body in the early issues of Fantastic Four.  I am not sure what to make of that fact.  I don’t believe I have a more finely tuned sense of horse anatomy than human anatomy: it might be that I am more used to badly drawn humans than badly drawn horses; or that there is more variation in human proportions than in horse proportions; or maybe weirdness just looks less weird when it’s upright than when it’s horizontal.  I’ll be interested to hear in the comments whether the Long Horse or the Long Human looks most wrong to readers.

(By the way, I casually talk about the type specimens of both “Brachiosaurus” species: while the situation is simple in the case of Brachiosaurus altithorax, whose holotype is FMNH P25107, things are more complex in the case of Giraffatitan brancai.  Janensch nominated “Skelett S” as the holotype of his new species “Brachiosaurusbrancai, but that turned out to be a chimera, composed of the two skeletons which he subsequently designated SI and SII — but Janensch never designated one of these as the type, and so far as I’ve been able to determine, neither has anyone else done so.  SI is represented by cranial elements and the first seven cervicals, but that’s all; SII is a much larger animal and is represented by most of the skeleton, and has been informally treated as though it were the type specimen most of the while, so I formally proposed HMN SII as the lectotype of the species (Taylor 2009:788) — just a bit of housekeeping.)

Here’s our old friend, the 8th cervical vertebra of HMN II, in a rare posterodorsal aspect, showing just how thin and, well, lamina-like the spinopostzygapophyseal laminae are.  All that space in between them?  Filled with diverticula, mostly.  Amazing.


Giraffatitan brancai lectotype HMN SII, 8th cervical vertebra, in posterodorsal view

Meanwhile some good news:

Remember the good news and bad news about the all-dinosaurs special volume of The Anatomical Record?  Well, since we posted that, the entire issue has been made open access!  Fantastic stuff there: details from D. Schachne of the Wiley-Blackwell Communications Team.  It’s not clear why the articles were all paywalled when originally posted, but all’s well that ends well.

And finally …

There’s been a gratifying amount of discussion in the comments on recent articles.  It can be hard to keep track of, but it helped a lot when I found an RSS feed for comments, which is what I now use.  For anyone else who wants it, it’s at https://svpow.wordpress.com/comments/feed/



30 Responses to “What a 23% longer torso looks like”

  1. Jake Says:

    The horse picture looks a lot worse than the human. I wonder some of the difference is in the nature of the original picture, the original man is a lot less like a real man than the original horse is like a real horse, if you did the same experiment on a different source you might get a different reaction.

  2. Mike Keesey Says:

    Interesting perceptual experiment. I have to say that the original human figure seems to me to have a somewhat short and rather thick torso. If you start with someone with more average proportions, you might get a stranger effect. (Also, we might be thrown off by the fact that he’s been flayed alive!)

  3. William Miller Says:

    That horse looks like some sort of bizarro dachshund/horse hybrid, but the human looks … fairly okay (besides lacking skin). If I *look*, I can tell that the torso/leg proportions are wrong, but it doesn’t feel wrong the way the horse does. I think this is because human proportions are variable anyway – even the elongated human doesn’t have as extreme a height/width ratio as, say, some NBA players.

  4. dobermunk Says:

    Another aspect that comes into play here are the relative degrees of realism within the images. The horse, while painted, is far more realistic. As representation approaches realism you experience deviation and ‘incorrectness’ as the uncanny valley.

    This is certainly at least playing a part in the perception. I wonder if it also has to do with the difference between vertical vs horizontal torso. I could imagine that the horizontal triggers a heightened awareness of structural credibility as effected by gravity.

  5. dobermunk Says:

    Had to return to report how wonderful I find this type of reporting. It makes the issues feel immediate and intriguing. Thanks!

  6. neil Says:

    I also wonder if the horizontal vs. vertical orientation plays a role. Likewise the realism of the depictions and other issues that have been raised –

    The expanded-torso human actually reminds me a bit of David — as I recall Michalangelo intentionally enlarged the upper body so the statue would appear properly proportioned when viewed from below.

    Would be interesting to see with just skeletons….

  7. Nathan Myers Says:

    American Mustangs are a breed that evolved in only two or three centuries from horses that escaped from Spanish explorers, and thrived in the north and western United States until extermination efforts by the U.S. government. (Initially this was part of a genocide project; now it is to reduce competition for range-fed cattle.) Those not heavily interbred with domestic stock are said to be proportionally shorter, shoulder to withers, but I haven’t found a reference to how much shorter. Comparison by analyses of photographs shouldn’t be too hard; people like photographing their horses.


    Instead of stretching horse pictures and eyeballing them, in other words, you might actually compare photos of live horses. Evidently the wild condition of horses is shorter than we’re used to, rather than longer.

  8. Heinrich Mallison Says:

    I guess there’s a bunch of reasons for the stretchuman looking fine, all already named. Personally, as a ballroom dancer, I KNOW humans have an incredible amount of trunk to limb and arm to leg ratios variability. Quadrupeds can’t afford that.

  9. Mark Witton Says:

    Artists play with human proportions all the time. In figure drawing, if memory serves, ‘average’ people are said to be six or seven head-heights tall. However, to make someone look more important – when depicting Gods or royalty, say – classical artists may depict them as a more ‘perfect’ eight head-heights tall, with the body and legs elongated to give them that extra height. Features are often exaggerated to characiture people, too: we see oddly proportioned figures all the time. Hence, the distortion of Vesalius’ image may stem from that: we’re used to seeing our own form shifted about by artists. That said, it would be interesting to see how Vesalius’ image matches up to the head-height count: as alluded to by Mike, the figure does look quite stocky to begin with. Try the same 23 per cent distortion on a size 0 fashion model: then we’ll see if it looks right. What am I saying? Size 0 doesn’t look right to begin with.

    Speaking of clothing, the distorted image does remind me of these people who walk around with the crotch of their jeans hanging around their knees. Such an inability to dress yourself not only means we can all see what brand of underwear you’re wearing, but also make you look to have a very long torso and stunted legs. In all seriousness, we’re probably more used to seeing people with proportion-distorting clothes on than their actual form: maybe this is a factor too?

  10. dobermunk Says:

    LoL! Now we just need Vesalius fashion prints on our undies and we’re ready to go!

  11. Nima Says:

    Hey guys, ever considered doing this torso stretching thing with Mark Hallett’s skeletal of Rapetosaurus?

    I’ve seen the actual skeleton and Mark Hallett’s version makes the neck and the torso FAR TOO SHORT.

    And while that stretched horse does look weird, I agree with Mike that the stretched human is within the realm of possibility… for a species, humans seem to have a LOT of variation in torso length, I’ve noticed this a LOT, some of which has to do with ethnic background/ancestry… for example the whole Masai vs. Bantu comparison.

  12. davidmaas Says:

    @Nima – like this?

  13. Easy. That horse does not exist; but we’ve all seen that human, albeit with his skin on.

  14. Vertebrat Says:

    for a species, humans seem to have a LOT of variation in torso length

    This paper suggests otherwise, surprisingly:

    We find that humans show low levels of within-population body height variation in comparison to body length variation in other animals. Humans do not, however, show distinctive levels of within-population body mass variation, nor of among-population body height or mass variation.

  15. David Marjanović Says:

    To my surprise, the elongated human doesn’t look appallingly wrong to me. It doesn’t look right, of course, but it seems within the realms of, for example, what might appear as a representation of a human body in the early issues of Fantastic Four.

    Bah. It looks a lot like my brother. I, on the other hand, am very similar to the left one, except I’m considerably thinner even with my skin on.

    And yes, this takes into account how far down the legs the arms reach.

  16. Brian engh Says:

    Longer human torsos look sexy. All your favourite super-models have long slender torsos. We like it because it indicates healthy good stuff for making strong jungle-babies. That said, if you were to elongate the already slender torso of a supermodel that additional 23% she would probably look like a terrifying space-monster.

    I agree with the people who have noted that the human selected for elongation was on the stocky side. Also, just looking at the drawing I feel like his head and shoulders are disproportionately large, which would make an increase in torso length look more proportionate. Perhaps this comparison would’ve worked a bit better if both of the images were photographs, or at least both illustrations drawn to a similar level of realism.

    Haha, I think the main reason the horse looks so bad to us is because the elongation of that curvature makes it look remarkably phallic. It really reminds me of one of Dali’s “soft forms” propped up on four stilt legs.

    Ewwww… It’s all subconscious-y!

  17. David Marjanović Says:

    Haha, I think the main reason the horse looks so bad to us is because the elongation of that curvature makes it look remarkably phallic.

    Speak for yourself. Not everyone has the same unconscious associations.

  18. Anonymous Says:

    Why is it when I try to picture a 23% longer brachiosaur I keep coming up with the fictional “moerisaurid” created by Thomas Tapir? Ironically, it is supposed to be brachiosaur (or maybe just macronarian) derived itself.

  19. Graham King Says:

    Thanks for this thought-provoking article. plus.. Rapetosaurus! Anthropometry! Such useful tidbits given in comments are part of the reason for the joy I find here.

    Overall height may vary relatively little in any given human population, but proportions (head size, neck length, torso length, limb length, robustness v. gracility) surely do vary plenty; within ‘mongrel’ populations at least. No offense is meant by that; I count myself as a mongrel, gladly! I just mean that, say, the British population is of very mingled ancestry compared to some national/regional populations which have experienced considerably less admixture over long periods (I think).

    Finally, can you guys do something useful for a change and get to the real meaty matter of sauropods: when they were not standing tall and walking about.. that is, when they crouched, knelt and lay down?
    ‘Cos, Looking at Rapetosaurus brought home to me again how robust that pubic foot is. Surely it acted as a strut to take the weight off their (pedal) feet at times. How about the ischia too? Were both those points covered by tough callosities in life, fit to endure weight and abrasion when resting/rising? What about femur length relative to hip depth: to rest their belly on the ground, did the hind legs splay, or angle forwards (straight, or doubled-up ‘kneeling’?) or angle backwards?

    Most finally: could a sauropod sit up, resting on its ventral pelvic extremities and erect forelegs, with the backs of its hindlegs along the ground, soles of its hindfeet directed forwards? Rapetosaurus looks as if it could..

    OK, those are not vertebral questions really. Sorry! So.. do sauropod vertebrae give any clues to plausible life-postures: not just re spinal curvature within the sagittal plane (well-covered here), but re limb-deployment?

  20. Graham King Says:

    Finally, can you guys do something useful for a change and get to the real meaty matter

    !that was meant only jokingly BTW, no disrespect!

  21. Mike Taylor Says:

    Graham, your questions are excellent ones. Sadly, I don’t have even the beginnings of hint of a direction that might lead towards the outline of an answer. And I don’t know whether anyone else does, either.

  22. Graham King Says:

    What we need of course are some fossils of sitting or crouching sauropods preserved in vivo position.
    Maybe some that died in their sleep?
    However, how could one tell that any such prone or supine posture found in remains was a natural one adopted in life, and not one imposed after death?

    A mummified family group might serve.. say, a sauropod equivalent of the families tragically entombed in volcanic debris at Pompeii.
    Of course those could be trauma-induced sauropod postures, maybe not representative of their everyday life..

    Shucks! Guess I just need a time-machine.. again.

  23. badfrog101 Says:

    I had a friend once, who was 6’3″, and had a 29″ inseam. At 6’1″ I have a 34″ inseam. My friend also had (I believe I remember) a 38″ sleeve; mine is 36.” He looked like a giant on the legs of a dwarf, but was a champion boxer. Some basketball and football players are even more disproportionate. We can use our minds to survive or even thrive at different bodily proportions, whereas a horse only has one trick, and different proportions would destroy it’s running ability. A disproportionate horse in the wild would not grow up to reproduce.

  24. […] Giraffatitan brancai–that are almost exactly the same size in limb bone dimensions (although B.a. had a longer torso). But we know that brachiosaurids got bigger, as evidenced by the XV2 specimen of Giraffatitan, and […]

  25. […] Similarly, although the torso was therefore longer than Gilles had intended, it might have ended up correct, as careful comparison of the lengths of the Brachiosaurus and Giraffatitan dorsals suggests that the torso of the former was about 23% longer. […]

  26. […] in my 2009 paper I estimated with reasonable rigour that the torso of Brachiosaurus was probably about 23% longer than that of Giraffatitan, yielding 4.82 m rather than 3.92 that Janensch gave for Giraffatitan. On much less solid […]

  27. […] more on how torso length can affect the visual appearance and estimated mass of an animal, see this post and Taylor […]

  28. […] I deliberately picked the longest C9 in the AMNH collection, and the least-distorted C10. The first surprise for me was how well this C. supremus C9 hangs with D. carnegii in terms of proportions. That is one looooong Cam vert. In any other sauropod, it would probably be beautiful. But because it’s Camarasaurus it attained its length in the most lumpen possible way, with the diapophysis way up front, the neural spine apex way at the back, and in the middle just…more vertebra. Like a stretch limo made from a Ford Pinto, or Mike’s horrifying BOBA-horse. […]

  29. Brad Lichtenstein Says:

    Not sure when this was posted (am I just missing dating info for the posts?), but followed Matt’s link here today. I’m approx 6’4″, and always knew I had long arms and legs for my height. When I sit, I’m not much taller in a chair than most adults. I have a friend who was on the same rowing team who’s 6’8″, we wore jeans with the same inseam length. When I seat raced against him, it was the first time I felt the rigger’s geometry (and therefore, the oar it held) was set for me: I’d been rowing 4 years to that point. I had another friend who was my height, and we noticed one day that we saw eye to eye, yet his waist was something like 4-6″ lower than mine. We figured he was the other extreme.

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

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: