Who are we learning about?

Among the many, many reminisces following the recent death of actor-comedian Robin Williams – some in honest tribute, some in shameless opportunism – we can find the video of his meeting with Koko, the female lowland gorilla who is famous for communicating by sign language; we also have the reports from the Gorilla Foundation, her caretakers, that when told about his death, Koko was distinctly sad over the news. From a scientific standpoint, however, we really can’t be sure of this. Courtesy of Not Exactly Rocket Science, we have Jane C. Hu’s article on Slate that discusses a lot of the difficulty with the conclusions and even the research methods of the Foundation.

This isn’t a new finding, either – the skepticism over the claims of Koko’s abilities have been around for a long time, and far too much of the purported findings look like the same kind of uncritical enthusiasm once expressed over facilitated communication, a method of using a human go-between to help functionally impaired children to communicate effectively. In facilitated communication, an adult, often a parent, would interact with a paraplegic child – by holding hands, for example – and interpret the child’s infinitesimal motions as the attempt to type onto a special symbolic keyboard, thereby producing a coherent message from someone who otherwise had no recognizable forms of communication. The problem is not knowing whether there really are distinct motions that the children make towards certain keys, or whether the adults themselves are introducing their own bias and, in effect, answering ‘for’ the child, much like an Ouija board uses the ideomotor effect to produce a message, not from spirits, but from subconscious movements of the participants (you’d think the spirits could move the planchette pointer without human intervention, and of course, we have myriad ways of detecting even nerve impulses that never produce motor functions much better than having to have an adult ‘feel’ the efforts from a disabled child.) And from numerous reports, it seems the case with Koko is more creative interpretation than her ability to communicate in a near-human manner, much less actually master a ‘language.’

A lot of this requires dismissing the human bias and trying to see it all from an uncluttered perspective. There is a radical difference between using any form of language and performing a function of cause-and-effect conditioning. A rat that presses a button to receive food is generally considered to have learned this through trial and error, and many species have this kind of cognitive learning ability. Putting a symbol on the button is an added step, as is then putting that symbol among many others in a group of buttons and only rewarding the rat when it presses the correct symbol; this does not mean the rat is now using that symbol as meaning, “food,” even though at a base level the correlation is likely present. For it to be differentiated, the rat (or any other species) would have to be able to use the symbols in an abstract manner far outside of simple associations, perhaps communicating something along the lines of “today’s food bad; yesterday’s food good.” This is what we consider language, and it applies as well to stringing together a selection of sign language symbols. The Slate article points this out in a distinctive way: does Washoe the chimpanzee signing water bird when she saw a swan indicate that she has created a new term, differentiating a ‘water bird’ from a ‘tree bird’ or ‘land bird,’ or is it simply a stream-of-consciousness type of response, giving the sign for ‘water’ when she saw it, followed by the sign for ‘bird’ when she saw that, both stemming from conditioning to associate a subject with a hand motion?

[We do this too, more than we tend to think: very often, we say, “Bless you,” when someone sneezes, never bothering to think about why someone with a nasal irritation deserves this special attention, nor whether we even have such power. But boy howdy, watch how some people get uptight when you fail to perform this pointless, conditioned response. And then there’s The Oatmeal’s take on it.]

There’s also the huge difference between language and communication, which can be performed with minor vocalizations, facial expression, body posture, or the erection of fur or feathers. Many species communicate in one way or another, from schools of fish reading the movements of their immediate neighbors to even plants releasing chemical responses to pests. So what studies are looking for is not that, but the distinction that separates language from communication, humans from all other species: the ability to express discrete, coherent abstracts. This would demonstrate that another species might even use such concepts, of which there is very little indication. And that naturally raises the question of why this might be – why us and no other species? What are the key differences?

One speculation is that the limiting factor is anatomy, specifically the larynx and tongues; this has some supporting evidence in that human children can master sign language much faster than speech. This indicates that infants’ delay in speaking is not entirely due to understanding abstracts or even assigning labels, but the difficulties of manipulating tongue and vocal cords. If apes possessed the mental ability to handle abstracts and used no language only because of the limits of anatomy, an alternate method of communication might demonstrate this readily. Yet, we’re well past the point where this could have been established firmly, without the results that we should expect if it were true.

Lowland gorilla
Not Koko – just a gorilla pic I had

In the middle of all this sits the urge, all too often, to consider that using language makes humans ‘higher’ than other species, more evolved or more successful, which is unwarranted and mostly ego talking. Any species that survives is successful enough, and we ourselves remain in a constant battle with mere bacteria and viruses. Gorillas might have never developed the traits to use language because those traits, appearing spontaneously within individuals through mutation and genetic drift, never produced a significant advantage and thus never spread throughout the species. Language is clearly a social benefit, requiring a highly-interactive species to produce a significant advantage, and gorillas fall lower on that scale than, for instance, sardines. They are not predators which could gain an advantage from coordinating in packs to obtain their food, and the niche that they inhabit is largely free from serious hazards – except, ironically, for humans. In short, they don’t need language, and thus are highly unlikely to have either the necessary thinking structure, the desire to engage in it, nor any function to put it to. And the results have primarily supported this: despite decades of using the language, Koko (and others like her) have demonstrated no marked increase in abilities, understanding, or function, no exploitable advantages, and in fact, might arguably be said to be functioning less optimally, relying far too much on the environment of the research centers to indulge themselves while not even expressing their thoughts and desires to notable advantage. Even though we are told that Koko wants a baby and has even selected a mate through a form of video dating service, this event has not come to pass and, due to her advanced age, is now unlikely to.

Further indication of how untrustworthy the various claims of communication are is how loosely the research appears to be run. Instead of dispassionate observation, the caregivers, first among them Penny (Francine) Patterson, are deeply interactive and openly interpretive, accepting, rejecting, and translating the various sign language missives from their subjects. Unedited, the transcripts of signs from Koko bear little relation to what it is claimed she is saying, and in an especially questionable exchange, Patterson has claimed that Koko’s signing of nipple is actually intended to mean people (and this is not the first time I’ve heard this account.) Aside from the obvious problems with this, we have to remember that Koko is signing, which negates anything resembling rhyming entirely, and presumably the signs she’s been taught for these two concepts have no similarity in structure. Not to mention that if she knew this language and actually meant people, this would be a far more often used sign and one that she should easily be capable of using appropriately. Even if we allow for the possibility that Patterson intentionally mistranslated to hide Koko’s indelicate interest (at least to humans) in seeing nipples, as intimated by the article, this still indicates that Koko was not answering the inquiries at all. The resemblance to a solid research project is tenuous at best.

Slate‘s article goes deeper into the Gorilla Foundation’s activity and history, getting away from the initial subject and raising some serious questions about how well the program is being run. Admittedly, most of this is hearsay due to few former employees going on the record, denied permission through non-disclosure agreements, so drawing conclusions is unwarranted – though the numbers of disgruntled workers isn’t a very good sign at all. But what is apparent is how much effort is put into promotion and media attention, which raises some interesting questions on its own – ones that go much farther than the Gorilla Foundation or even the concept of teaching apes how to use language.

Funding for scientific endeavors is a tricky thing. It usually comes in the form of grants, which can be (and usually are) quite capricious in nature – most areas of potential research sit unexamined because of a lack of funding or patronage. So another avenue to obtain funding is through the public, and there are few topics that can spark enough interest to generate a significant, sustainable level of funding; animals are probably the most prominent (followed by, in no order, childhood diseases and cancer research.) So there is a significant incentive to make animal-related research studies very prominent, or to create a public appeal offshoot that brings in funding on its own; most zoos use this model, where the income from captive display animals is partially directed towards endangered species breeding programs and wildlife research. Others, like the Gorilla Foundation, expend a lot of effort into making their research as public-oriented as possible.

The ethics of this can be debated ad nauseum (and are,) partially because ethics isn’t defined well enough to get everyone onto the same page – I’m not going to get into that here, since that’s worth about 5,000 words in itself. More to the point, the danger of a public-funding model is that it requires constant interest, and thus a lot of effort put into wooing the media and providing fresh content. Now, most scientific studies don’t make for interesting articles, much less TV spots, and even less so when there’s little progress being made. The nature of science is that not every avenue of research is going to produce positive results, and many that do are in the future expansion, speculate-on-potential-impact variety – this isn’t ‘news.’ So the incentive to over-promote, hype, and even generate not-entirely-accurate reports of progress is also present, and where the dividing line sits in such situations is never particularly clear.

This appears to be where most of the ape language field falls: a tiny handful of test subjects yielding sporadic results wide open to interpretation, without any serious progress being made. While, “Koko wants a baby,” and, “She’s sad over the passing of Robin Williams,” generate a lot of public interest, from a scientific standpoint they have virtually no value, even if the efforts had been made to verify that these communiqués were accurate, which is far from apparent. Any study funded by the typical grant process would not have gone on for the decades that these have, partially because the controls are not in place to determine if the claimed results are solid, and mostly because the results, even as reported, are fairly thin. There remains a lot to be learned about the cognitive processes of apes, and indeed any species, but to do this accurately, the process must be rigorous, objective, and above all, independently confirmed.

*     *     *     *     *

I have just a minor annoyance to express here. Despite the ridiculous number of articles on Koko and the Gorilla Foundation that can be located, and the repetition that Koko is an endangered species, not one source that I found confirmed which species Koko actually is, among only four likely candidates – I was therefore unable to provide this myself. Such a simple detail, but apparently far beyond mass media’s capabilities.