I've seen the "palms facing forward" quote in a few news reports about last week's Dmanisi postcrania paper. It's pretty nonsensical when you see it devoid of context. Consider Bruce Bower's Science News article:
However, the arms of Dmanisi hominids appear more like those of australopithecines, an earlier line of hominids. For instance, unlike people, the new specimens have upper arms that are straight rather than slightly curved, their shoulders are relatively narrow, and their palms are oriented forward rather than inward.
This is quite a vision, isn't it? How exactly is that humerus curved, again? And they stand with their palms forward? What?
OK, so it's tough to give a description of humeral torsion while making it sound important. Your humerus has two ends. The proximal end, called the head, attaches to your shoulder; the distal end is part of your elbow joint. When you are born, the head of your humerus faces toward the back (posteriorly). As you grow up, the humerus twists, so that the head faces inward toward your body (medially). The amount of twist is called the torsion; it is measured relative to the cross-section of the distal end of the humerus.
Nobody really knows what purpose is served by this twisting growth pattern. Presumably, the twisting adjusts for a change in the orientation of the shoulder joint, although that growth pattern has yet to be documented. But humans are more twisted than apes, and low humeral torsion is the key link that people are pointing out between Dmanisi and Homo floresiensis. So the articles are forced to describe it somehow. As a paleoanthropologist, I'm used to describing skeletal changes in punchy ways. Humeral torsion is a challenge -- without a really clear explanation of its function, it is hard to describe it in concrete, memorable terms.
Where does the "palms-forward" interpretation come from? We can trace it to Daniel Lieberman's commentary:
In modern humans, the elbow joint is typically rotated relative to the shoulder joint, so that the forearm naturally hangs with the palms facing inwards; but the new Dmanisi humeri lack torsion, so their palms would have been oriented more forwards. Lack of humeral torsion, a highly plastic and variable feature, suggests something different about the shoulder in these specimens.
Now, I'm sure that most of my readers will be scratching their heads over this one. People carry their hands palm-inward not because the humerus is twisted, but because the radius is habitually rotated across the ulna. That's the same reason why my hands are currently palm-downward on the computer keyboard. The humeral torsion is entirely irrelevant to the palm position when the arms are "naturally hanging" -- I can assure you, all of my children walk with their palms facing inward, despite the fact that their adult humeral torsion hasn't developed.
And of course, if humeral torsion is really about the orientation of the shoulder joint, as Lieberman suggests, then it really has no importance to the function of the elbow at all -- different torsion values would maintain the same lower arm mechanics with different shoulder orientations.
Still, neither the function nor adaptive value of humeral torsion are obvious. As Lieberman mentions, the trait is variable -- Larson and colleagues (2007) reported ranges in recent human populations extending from less than 110° to more than 170°. The value for the adult Dmanisi D4507 humerus is 110°, at the very lowest end of the modern human range; the value for the subadult D2680 is 104°. Humeral torsion continues to increase until age 16 in living people, although most change occurs before age 8 (Edelson 2000).
Larson et al. (2007) suggest that low humeral torsion is related to a short clavicle -- the idea being that the shoulder joint (glenoid fossa) was anteriorly (forward) placed, and the head of the humerus therefore had to face more posteriorly. I'm not sure that explains the low torsion at Dmanisi, since the Dmanisi clavicles aren't especially short -- like the long bones, they are right in the middle of the modern human range. But they might have had an anteriorly-facing glenoid fossa even if their clavicles weren't short, and given the low humeral torsion I suppose they probably did.
None of this means that the Dmanisi people or any other early hominids stood with their palms forward. Paleoanthropologists usually do a really good job of describing anatomy in down-to-earth terms, but humeral torsion seems to be a challenge!
Edelson G. 2000. The development of humeral head retroversion. J Shoulder Elbow Surg 9:316-318. doi:10.1067/mse.2000.106085
Lieberman DE. 2007. Homing in on early Homo. Nature 449:291-292. doi:10.1038/449291a
Larson SG, Jungers WL, Morwood MJ, Sutikna T, Saptomo EW, Duw RA, Djubiantono T. 2007. Homo floresiensis and the evolution of the hominin shoulder. J Hum Evol (in press) doi:10.1016/j.jhevol.2007.06.003