His "mild" version of the AAT/H

He has been trying to make a claim for a "milder" version of the AAT/H for a while now. This is an extension of the internally inconsistent manner in which the idea has been presented by virtually all of its proponent's for many years, what I've dubbed its "ZING!ability". I've described this, and the issue of Kuliukas's attempt at defining the AAT/H, on my "General Problems" page.

And by the way, when Kuliukas talks about our ancestors wading, he's talking about wading sideways.

He also exhibits the by now standard AAT/H coyness about what animals he's talking about when it comes to characteristics, but as always the characteristics are those of seals, whales, and sirenia -- yet we somehow got these features via convergent evolution by wading, and only some occasional diving and swimming (and the wading is sideways, remember).

Like all the major AAT/H proponents, he engages in special pleading. Take a look at this claim from the sci.anthropology.paleo newsgroup as an example:

"Stop waffling and let's get back to the point: Maximally efficient human bipedalism, based on the inverted pendulum (straight legged) gait simply does not work in habitats that are not flat, firm and relatively vegetation free."
--Algis Kuliukas Jun 3, 2004 sci.anthropology.paleo

There's actually a bigger -- and perhaps more subtle -- problem with the statement: this is the use of the qualifier "maximally efficient". Most of the time bipedalism will not be, and should not be expected to be, "maximally efficient". Carrying or feeding, for instance -- two activities where we commonly see non-human primates use bipedalism -- would render the bipedalism less than "maximally efficient". So would walking on soft ground, uneven ground, wet ground, walking up or down even the slightest of slopes, walking through vegetation, or around rocks, roots, and snags -- and a moment's reflection should cause you to realize that all those conditions also occur in water, with the added problem that they often aren't as easy to spot and therefore compensate for.

In this way he takes an incontestably true statement and, by leaving out an apropos but inconvenient bit of information, turns it into something that needs ironic quote marks -- a "fact". And he uses this abbreviated "fact" to claim that "our bipedal locomotion is not suited for woodland, dense bush or tall grassland" (Algis Kuliukas 06-21-04 sci.anthropology.paleo). Let me ask you, have you ever walked through a woodland, through some brush, through some tall grass? Did you find it as enormously difficult as Kuliukas suggests? I've done it and it just wasn't so difficult; I can't be that unusual.

This is just one example of what has become the classic AAT/H proponent's habit of ignoring contrary info; another example from Kuliukas is what he did when he found the data on hair and swimming speeds on my site (mentioned on this site's Hair and "Hairlessness" page). In fact, this habit is so ingrained he seems to be unable to comprehend why on earth I would mention data that doesn't bolster "my" case. I found conflicting data, and therefore wrote up and posted the data both for and against the idea that reducing body hair would improve swimming speed. I did this because that's the way you do science -- it's the only honest and sensible way to do it -- but Kuliukas assumes that I did so "by mistake" (Apr 26 2004 sci.anthropology.paleo). That he would assume so tells us a lot, I'm afraid, about his own attitudes toward data -- it's projection on his part.

And he has a critique of my site online, which can be summed up as underwhelming -- for instance, criticism of my URL (the web address www.aquaticape.org) figures prominently. He also seems to be annoyed that when I described the dishonest leaving out of contrary data about hair and swimming, I didn't mention his name -- I thought I was being kind; he apparently considered that deceitful. Now I've mentioned his name in connection with that episode, but somehow I still don't think he'll be happy -- some people are just hard to please.

His master's thesis

You may see Kuliukas mention that apes are bipedal 92% of the time they are in water; he frequently says this or some variation of it, and it figures prominently on his web site. As is often the case with AAT/H proponents' claims, it's what they don't say that's key.

On the 92% figure, what often isn't seen, or isn't reported, is that this was 92% of 37 seconds total for 9 occasions over 3 days. This is a problem for several reasons, one of which is that he doesn't mention as prominently that those same bonobos were bipedal on land more than 4 times as much. Another of course is that 37 seconds for nine occasions isn't much wading at any one time, yet he obviously considers this to be an example of a major selection pressure. Now it's not that 37 seconds itself isn't interesting and not worthy of reporting -- it's just that Kuliukas has made rather too much of it, and also ignored several important questions he should've asked himself before he even got to the point of writing it up.

When I did my first write-ups (in several online forums and newsgroups) of Kuliukas's thesis I was doing a quick read of his paper and inadvertently gave him credit, perhaps, for doing a bit more work than he actually did. His observations were done over 3 days, and at first I thought he did an actual day's worth of observations each time, but actually his observations totaled only 5 hours over 3 days. Even that may be giving him more than his due -- he mentions the 5 hours, but then also says that the observations consisted of following the actions of 5 individuals for a half hour each and that "one hour's worth of observations were recorded of isolated, real or anticipated instances of contextual wading behaviour". Whether that hour's worth were from the 5 half hour periods or some other periods isn't clear, and I don't know where the 5 hour total came from.

It's interesting how he chooses to present his conclusions (this is indicative of how AAT/H conclusions are generally put forward, by the way). He points out that these bonobos were bipedal on land 1.16% of the time and bipedal in water 92% of the time. That certainly sounds impressive -- a lot more impressive than pointing out that during his observations they were bipedal on land over 4 times as often as they were in the water (in his observations, 161 secs vs. 37 secs -- not a lot in either case). In fact, they were bipedal on land more than 3 times the total amount of time they spent in water (1.16% of the time versus 0.37%). He apparently doesn't consider either of those other points especially significant, and certainly doesn't present either as a bar graph, as he does (twice) for his own preferred presentation.

You may notice that unlike Kuliukas, I tend to mention both the percentages and the actual time while he doesn't seem to very often. It's best to mention both, just as it's apropos to mention how much time they spent bipedally on land. Using only the statistic, as Kuliukas virtually always does, seems (and let's hope this isn't so) to be a way of making it sound a whole lot more impressive than it in fact is. It's always a red flag when someone uses a statistic -- a percentage or ratio, for instance (most commonly, in my experience) -- and doesn't mention the actual numbers, and statistics drawn from such a limited sample are a dubious method at best. If someone is claiming that water is ever so much more important than land when it comes to the evolution of bipedality, but doesn't mention that the apes he watched were bipedal on land far more than they were in water, that also raises one of those red flags.

Perhaps the most critical failing in his thesis is that when he looks at the amount of time in water that his subjects were bipedal he gets a number that's far greater than any other study; several times larger than the next largest percentage (which was about 24%; studies of other bonobo groups have shown less than that -- and some bonobo groups apparently always wade quadrupedally). This cries out for examination as to why, but he doesn't do that. When you do a study where one of your measures is way off the graph, it doesn't necessarily mean that the measurement is bogus, but you have to ask why -- why is it so far out of line with everything else. There are several possibilities I can think of but these were apparently never examined; Kuliukas didn't seem to ask himself this obvious question. For instance, is it because the bonobos were getting food? we know that food-getting is a major factor in bipedality; in fact Kuliukas is aware of this, or should be (in fact in one case he mentions the bonobo in question was bipedal on land just before entering the water -- he was begging for food). Is it due to some factor around the way the moat they waded in was constructed -- the slope of the sides or how slippery it was or anything like that? Was it something else? we don't know for sure and Kuliukas apparently just didn't bother looking at that even though that is a question that should immediately spring forth when you measure something and get a number that's so wildly out of place compared to other studies.

The same is true of the number gotten for bipedality on land in Kuliukas's study -- is it similar to what you'd expect in a wild population or is it under the value you might expect? For instance, Hunt's study shows a very large part of bipedality used in feeding, but in captivity the feeding is far different than in the wild, so we might well expect there to be far less bipedality used in feeding in captivity -- this would mean the amount of land-based bipedality could well be greatly underrepresented in a study like Kuliukas's. This too he seems to have left unexamined.

And of course Kuliukas doesn't seem to be getting the distinction between what he did with percentages and what Hunt did -- he just sees percentages mentioned in both cases and assumes they're the same (much as we see when AAT/H proponents see water mentioned in the context of human evolution and automatically assume it's evidence for the AAT/H position). Hunt used percentages to show what activities contributed to bipedality; Kuliukas used what medium was used, without regard to activity. In itself this isn't necessarily a bad move on Kuliukas's part (although he ignores the fact that his bonobos were engaging in food-getting and carrying behavior known to contribute to bipedalism), but the problems arise when he doesn't examine why his numbers for bipedality in water were so many times higher than any other study and when he doesn't see that even in a setting where bipedalism on land may have been less than normal and bipedalism may have been more than normal (neither is examined) -- even in that setting bipedalism on land was used 3 times more than in water. So in terms of what activities might have contributed more to bipedalism Kuliukas's own study shows a 75%/25% split in favor of land -- and for the reasons mentioned above this would seem to be biased toward the water end of the scale.

He also doesn't seem to see any effect on the fact that all the times these bonobos entered the water they were retrieving food that had been thrown into the water (by visitors; he wasn't allowed to and didn't do it) and so of course they needed use of their hands to pick it up. That would of course encourage the choice of bipedal posture (as it often does on land when food is otherwise out of reach).

This is typical of AAT/H research, in my experience; selective reporting of facts, ignoring unwanted data (and sometimes just making things up; this last, thankfully, is something I don't see in Kuliukas's thesis). Frankly, I think he took a minimum of observation and stretched it far beyond where it could reasonably be taken, using percentages to make it sound more significant than it was. I suspect he might better have spent his time in the library digging out the specifics from as many papers as possible. He also had a long section in there about some tests he did with people wading in a pool to test water resistance, in which he comes up with the idea that wading sideways might have been a primary method used by early hominids, and really, that strikes me as a touch odd.

Also, he's going on about the number of "data points" in his study compared to Hunt's much longer study -- Data points in observations of a subject are the periods of time one has looked at one's subjects and there is always some degree of arbitrariness to what constitutes a "data point", but Kuliukas stretches this well past the bursting point. He's is right about the usefulness of using video, when possible, to record observations of this type, but of course this is much easier in a zoo setting than in the wild. Not only are wild observations harder on the equipment (although I expect video equipment and tape is far better at handling this than it used to be) but the sight lines in a zoo setting are generally far less obstructed than in the wild, and since you have them in a confined space, you don't have to worry about moving around much with that equipment. There are also other problems with using video; one big problem with a video record is that you tend to miss more of the overall goings on because you are focused on a small portion of the overall view. Any photographer is aware of how photos and video are a more restrictive view than a set of eyeballs gives you; ironically the eyeball, because it's attached to a massive data-shifting apparatus -- the brain -- can focus in tighter on a subsection of what it views. The camera records what it sees -- all of it -- so it's good that way, but since what it can see is typically far more restricted, you have both good and bad points in using a camera for observations. But overall, especially for observations in a zoo setting, video is a good tool.

But Kuliukas's notion that you get so many more "data points" is at the least disingenuous. The way he does it he concludes that he has 14,400 "lines of data" which sounds pretty good compared to the two studies he compares his with, because they have "21,030 data items" and "30,072 lines of data" respectively. This makes it sound like he's done about half the observing of the second study and two thirds of the other one -- but look at the numbers: the Hunt study with the "21,030 data items" was 701 hours of observation compared to Kuliukas's 5 hours -- 140 times as much, which is about 100 times more than Kuliukas's "lines of data" comparison implies (or about 50 times in the case of the Videan and McGrew study -- 250 hours of observation vs. 5 hours). Even if you use lower "actual number of data items recorded in the database was 1,319" (after skipping "long periods of inactivity") you still have a number that's wildly misleading -- the real difference is about 12 times more than Kuliukas's "lines of data" comparison implies (or about 11 times in the case of the Videan and McGrew study). Is this a deliberate attempt to inflate the importance of his study, or simply an inadvertent misuse of statistics? I don't know and don't care, because it really doesn't matter why it's wildly misleading. Whether deliberately dishonest or simply foolish, it's just wildly misleading and this helps demonstrate the thinking (or lack of it) behind the research.

He seems to be considering each second of his recording a "data point" compared to each 2 minute sampling period in Hunt's study (and each 30 second sampling period in the Videan and McGrew study). This is silly, and ultimately very arbitrary -- why not consider each frame of video a "data point"? just as valid (which is to say "not valid") as using each second, and you get 450,000 "data points" in a 5-hour study (using PAL or SECAM; 539,460 "data points" in NTSC, which makes North American and Japanese research that much better, doesn't it :). So using a video record is often good but entails some trade-offs; but the bottom line vis a vis "data points" is that Kuliukas's claim to have so many more is based on a rather arbitrary and meaningless measure. I'm sure he will disagree with this.

Another interesting thing, Kuliukas has later asked (apparently after reaction to his thesis online) one of the zoo workers whether his observations were typical of the bonobos at that wildlife park, and the answer was yes. Interesting that those particular bonobos wade so seldom and for such short periods -- if they do indeed generally just duck in and out for a fraction of a second or a few seconds at most, as Kuliukas reported for his observation period. Because that's the other part; those particular bonobos seem to be rather unusual, as they seem to not wade at all if they can help it, and they wade so very little, unlike virtually all wild bonobos. Does that affect Kuliukas's conclusions from his observations? Not according to him, apparently, although it should, since they seem to be acting quite differently from wild bonobos -- to be expected in one way or another, of course. But to use observations from that short observation period of apparently unusually acting bonobos as if they were more representative of ape behavior than longer observations of wild bonobos is not a sensible method.

So here's the bottom line regarding Kuliukas's thesis: he took a total of 5 hours of observation over a period of 3 days. Notice that he virtually always uses percentages when he mentions this study, perhaps because it sounds better than saying that these apes spent only a total of 37 seconds in water -- mostly on one day -- and that they were bipedal on dry land over 4 times as long as they were in the water, unless you go with percentages and try not to mention the actual times. Since in his study the bonobos spent less than 1% of their time in the water (less total time in the water than they spent bipedal on land, by the way) this does tend to skew things. Yet Kuliukas wants this study to essentially replace much longer observations which give different results, as well as pretending that this is applicable to all bonobos when we know that they vary in behavior from place to place, much as chimps do in many of their habits.

A further note: in thinking about the relatively low amount of bipedalism Kuliukas saw in his study, I found a publication by another person who'd studied the bonobos in that same park a few years before, and the differences between that study and Kuliukas's were fascinating. Duchêne (1997) took a different tack, and I'd say a far more sensible tack, in looking at bipedalism. Whereas Kuliukas compared the time spent bipedal to total time, Duchêne compared the time spent bipedal to the time spent quadrupedal. This seems to me likely more apropos (and in Kuliukas's case it would still give him the big number he so obviously wants for bipedalism in water, although it would still leave all those unanswered questions he should've thought of and answered). It would, however, give a far more representative number for the bipedalism on land for those bonobos, because most waking time -- the vast majority of time -- on land for these bonobos was sitting around rather than any sort of locomotion ("65% of the time" in Kuliukas's case), while apparently no non-locomotor time was spent in water. This wouldn't give Kuliukas the extremely low number he also so obviously wishes to have for bipedalism on land, but it would still be low in his study and would be far more honest.

By the way, in light of what I mentioned above about apes so often using bipedalism in regard to food, Duchêne found that the percentage of bipedalism (compared to quadrupedalism) jumped from "3.9% for spontaneous bouts to 18.9% when abundant food is supplied" -- 4.8 times more because of food. This brings to mind Hunt's finding that "80% of chimpanzee bipedalism occurred most often during feeding, with no other context making up more than 4%" and that "Bipedalism was observed most commonly among chimpanzees when they fed on the small fruits of diminutive, open-forest trees". This also fits with the results of Videan and McGrew's experiments which showed carrying food as a big incentive for bipedalism in both chimpanzees and bonobos. Remember that all of Kuliukas's observations of bipedalism in water were connected with getting food.

We saw at the start of this page an example of how Kuliukas started a true statement and turned it into a false "fact" by leaving out an some apropos but "inconvenient" bits of information. Similarly, he's taken his observations of a true but apparently highly unusual pattern of behavior on the part of 10 captive bonobos and -- by leaving out many inconvenient but apropos pieces of data, turned it into another false "fact". He's then used this false "fact" as if it were true:

"Since 1994 a great deal of film footage and observational data from the field has shown that the one place apes are pretty much *always* bipedal is in water. This is blindingly obvious to anyone with an ounce of common sense."
Algis Kuliukas May 2 2003 sci.anthropology.paleo

"Extant apes are pretty much 100% bipedal in water..."
Algis Kuliukas May 13 2003 sci.anthropology.paleo

And despite any later waffling and qualification, this false "fact" will out there to trap unwary and uninformed people whose only "crime" is wanting to learn about human evolution, who will learn and repeat such misinformation entirely innocently. What's "blindingly obvious to anyone with an ounce of common sense" is, sadly, that the biggest legacy of the AAT/H and its proponents, as far as I've seen: a long litany of false "facts", like Kuliukas's statements above, that will ensnare the innocent and no doubt, as Darwin warned, "endure long".

References

Duchêne, V
1997 Inleidende studie van de locomotie van Pan paniscus: de bonobo als model voor de evolutie van menselijke bipedalie. License thesis, University of Antwerp.

Hunt, Kevin D
1994 "The Evolution of Human Bipedality: Ecology and Functional Morphology" in Journal of Human Evolution 26: pp 183-202.

Kuliukas, Algis
2001 Bipedal Wading in Hominoidae past and present. M. Sc. dissertation, University College London.

Videan, Elaine N. and W.C. McGrew
2002 "Bipedality in Chimpanzee (Pan troglodytes) and Bonobo (Pan paniscus): Testing Hypotheses on the Evolution of Bipedalism" in American Journal of Physical Anthropology 118:184–190.

Home

Feedback: E-mail me

Next