Daily Archives: March 2, 2012

The Scientific Method (“Talking” chimps: Controlling the variables)

Charles Rulon

Where humans, in general, fail in their ability to think critically is in the area of con­trols.  Endless cause and ef­fect errors and wrong conclusions have resulted because of our failure to con­sider the many variables in a situ­a­tion. 

Recently articles have appeared of orangutans using iPads at zoos (Google “Apps for Apes” and “orangutan outreach”).  Apparently, zookeepers are planning to “set up play-dates when the apes can use iPads to video chat with friends in other zoos”.

So, consider the question: Can the great apes actually learn to talk with us—to have a two-way communication—via an iPad or, perhaps, American Sign Language (ASL)?  If they can, that would be absolutely extra­or­dinary.  Of course, extraordinary claims require extraordinarily careful research.[i]  And, of course, “talk” has many mean­ings: dogs “talk” when they bark, growl, whimper and leave a message on a fire hydrant.

So let’s go back to the 1970s when research­ers at the University of Nevada reported success in teaching ASL to an infant chim­panzee named Washoe.  For the first time in history it was pro­claimed that a non-human pri­mate, a chimp (our closest evolutionary rela­tive), had mas­tered a lan­guage in which it could actu­ally communicate with humans.  Washoe was reported to not only understand over a hundred different ASL sign gestures, but also to be able to also com­bine them in ways that suggested elementary grammar.  For exam­ple, when a swan flew by, Washoe is reported to have signed the words “water” and “bird.”

Other research­ers soon began teaching ASL and other sign lan­guages to young chimps and even a gorilla with seeming success.  Books, articles and even a film documentary soon appeared.[ii]   Writer Michael Crichton even had a fictional gorilla named Amy extensively communicate with her keeper using ASL in his 1980 novel Congo, a novel I just finished reading.

However from the begin­ning a num­ber of experts on language and an­imal behavior had remained skeptical of these extra­ordinary claims.  But their criticisms regarding the many uncontrolled variables appeared only in tech­ni­cal journals.  Then in 1979-1980 two books (Nim, 1779; Speaking of Apes, 1980) were published.  Both au­thors presented a strong scientific case for the view that, although chimps have a re­mark­able memory that en­abled them to master over a hun­dred different visual signs (dogs and horses can also master several dozen signs), they do not com­pre­hend sign sequences in any way essentially different from a dog’s under­standing of such com­mands as “Go get the newspaper.” These chimps have simply learned to do “clever tricks” for a reward.  The authors documented, by extensively studying unedited video tapes, that:

a. Much of the signing by the trained chimps imitated parts of what the trainer had just signed.  In many cases trainers were astonished to see how often they had unconsciously started a sign that the chimp had noticed and copied.  For example, an un­cut version of a Nova documentary called, “The First Signs of Washoe,” showed that almost all of Was­hoe’s multi­-sign statements came after similar signs by trainers.

b. Most of the chimp’s signing were random combi­na­tions of signs plus the sign for “me” and for the chimp’s name—signs that fit al­most all other signs and which they had learned were likely to be rewarded.

c. The trained chimps never learned the two-way nature of conversation as young children do. They con­tinuously interrupted. The research­ers had ex­plained this away by attri­buting such inter­ruptions merely to the chimps’ “eagerness to talk.”

d. Many times the chimps’ signs were wrong, vague, or only partially complete, resulting in the train­er either “reading in the rest,” or claiming that the chimp was either “making a joke”, “teasing”, or “being bratty.”

e. In the course of several years, these chimps put together signs in thousands of random ways.  No re­searchers bothered to record all of the nonsense com­binations produced by these chimps, such as “Banana eat Nim.”  But every lucky hit such as “Nim eat banana,” was reinforced by cues of approval and went into the re­searcher’s records.   So, claim the skep­tics, these chimps just ran on with their hands until they got what they wanted.

f. Most damaging, deaf native users of ASL not only reported a failure in two-way communication with the trained apes, but also that these apes were not signing ASL at all, but were just making many gestures and partial signs.  In retrospect, it seems obvious that a precondition for any experimental attempt to teach a true sign language to primates would beto ensure that the main contact people are all native speakers of that sign language.  Otherwise it’s somewhat like a non-Italian-speaking trainer with an Italian dictionary trying to raise a human child who hasn’t yet learned a language to speak Italian.

The final conclusion was that when all the above variables were tightly con­trolled, the ability of chimps to have a two-way con­ver­sation with a human dropped almost to chance.

How could re­search­­ers have over­looked all of these seemingly ob­vious vari­ables?

A.  The “successful” chimp trainers had min­imal, if any, training in con­trolling their uncon­scious fa­cial move­ments, breathing rhythms, bodily ten­sions and so on that could cue the apes.  The litera­ture is full of “learn­ed” dogs, horses, pigs, even ducks, that respond to the smallest unconscious cueing.  “Talking” apes don’t perform well at all for skeptical strangers.

B.  Psychologists refer to “confirmation bias” and “ex­peri­­menter ef­fect” for all of the insidious ways that re­search­ers’ con­vict­ions can unwit­tingly deceive them and dis­tort the data.  The past few decades of research in cognitive, social and clinical psychology suggest that such biases may be far more common than most of us realize. Even the best and brightest scientists can be swayed by them, especially when they are deeply invested in their own hypotheses and the data are ambiguous.[iii]   Consider:

a. Eminent scientists tend to be more arrogant and confident than other scientists. As a consequence, they may be especially vulnerable to confirmation bias and to wrong-headed conclusions, unless they are perpetually vigilant.

b. Researchers have a tendency to look for and perceive evidence consistent with their hypotheses and to deny, dismiss or distort evidence that is not.

c. Researchers who get positive re­sults often have their careers advance faster and their work more likely funded. The pressure on scholars to disregard or selectively reinterpret negative results that could doom their careers is considerable.

d. Assistants are strongly moti­vated to produce results that will please an employer who pays their salaries.

e. If the work is controversial, there is a ten­dency for research teams to close off from the outside world and to form a cluster of insid­ers deeply suspi­cious of outsiders.

This brief coverage of “talking” chimps:

a. Was presented to emphasize how cri­tically important (and often how difficult) it is to control all the variables in scientific experiments.  One major variable is human fallibility.  Thus the necessity of having independent impartial investi­ga­tors reproduce the work.

b. Was presented to il­lustrate the power of the human mind to deceive itself.  Such self-deception is particularly wide-spread in areas deal­ing with the paranormal, the supernatu­ral, UFO’s and so-called alternative medicine “cures”.

c. Was not presented to demean the brains of apes and other mammals.  The great apes are probably much smarter than we give them credit for.  Each mammal species has a unique set of evolved mental capabili­ties that we are just on the frontiers of un­der­standing.  We “civilized” humans, for the most part, have seen our­selves as superior to other animals, an atti­tude that has resul­ted in wholesale indifference, care­less­ness and wide­spread species extinction.

d. Was not presented to indicate the last word.  Much research will con­tinue.  Scientific knowledge grows through an openness to correct past er­rors.

Update

Since the 1970s, much research has continued into great ape language, involving chimps, bonobos, gorillas and orangutans.  In recent years, computer keyboards and iPads have been added to ASL. A quick Google search reveals that many researchers remain convinced that two-way communication has been achieved.  Their conclusions, however, continue to be disputed.[iv]  So far, at least as reported by the linguistics department at UCLA, no breakthroughs have been confirmed; no unequivocal evidence exists that apes can learn and use a sign language, which incorporates most of the significant features of human language.[v]

Finally, to quote eSkeptic: “Next time you see [a talking chimp] on a television documentary, turn down the sound so you can just watch what he is doing without interpretation from the ape’s trainers.  See if that really appears to be language. Somewhere in the history of our kind there must have been the first beings who could rearrange tokens to create new meanings, to distinguish Me Banana from Banana Me. But the evidence from many years of training apes to press buttons or sign in ASL is that this must have happened sometime after we split off from chimps, bonobos and gorillas.  Since then we have been talking to ourselves.”[vi]

Charles Rulon is an emeritus in the biology department at Long Beach City College.



[i]This article draws heavily from Martin Gardner’s ex­cellent book, Science-Good, Bad and Bogus (1981).

Also see  http://www.ling.ohio-state.edu/~swinters/371/nimchimpsky.html; http://en.wikipedia.org/wiki/NimChimpsky

[ii] Omni, Jan., 1980; Nat. Geogra­phic, Oct. 1978; Koko, A Talking Gorilla (film-1979).

[iii] Scott O. Lilienfeld -Scientific American, Nov. 2010, p. 18.

[vi] http://www.skeptic.com/eskeptic/07-10-31/#feature