I read the story of Kanzi of which i had only vaguely heard,
Sue Savage-Rumbaugh - Wikipedia, the free encyclopedia, and I must say I found quite a few similarities between their approach and mine, but also differences. A few quotations:
We recently warned about the danger in ape language research of confusing language with what might perhaps just be tempting ways
of conceiving language intellectually
When we observe how similar the dimensions are in Kanzi?s way of acquiring language and in human children?s first-language acquisition,
it seems less far-fetched that an ape can develop language.
why have they not developed it in the wild?? An important part of the answer is: just as evolution is not driven by an aim of developing Homo sapiens, so cultural changes are not driven by an aim of developing our human form of language, as if this form of language were some kind of ultimate state towards which all conscious creatures strive.
.
However, in a young bonobo finding himself in a cultural environment where human companions speak and gesture to him while doing exciting things together, there arises a profound drive to develop more humanlike language.
.
Especially the last quote shows the importance of feelings (motivation) in the use of cognitive abilities. Still, the problem remains, aptly worded by Fitch:
Kanzi cannot communicate all the concepts he can entertain. For example, Kanzi can successfully carry out quite complex motor actions, such as starting fires and making and using simple stone tools (Toth et al., 1993), but his ?linguistic? productions never even come close to the complexity required to describe these abilities.
This is also in compliance with my analysis: adding to the actions array does not change the limits of cognition.
So, comparing Kanzi to Amy, we can state that:
- Kanzi's sensation/emotion array seems to be large enough to accommodate much wider cognitive abilities. Apes are capable of love/caring, jealousy, deception, curiosity, playfulness, fear, joy... Whereas we have endowed Amy with no more than hunger until now.
- Both action arrays seem to be more or less similar. There is no reason for any differences except those based on body shapes.
All that rests are:
1) the connections between sensations/feelings and actions
It would seem that Kanzi's repertoire of actions is limited to very few patterns:
- beg
- play,
- fight,
- flee,
- imitate,
- ...
- innate or learned sequences, like grooming, hunting, or making a simple tool.
The question Fitch asks is very legitimate. Kanzi's vocabulary should be at least as wide as his actions repertoire.
2) Kanzi's sensations/feelings seem also to be limited by these very actions:
desire for
- food, (hunger)
- comfort, (fear)
- distraction (joy, curiosity)
These basic emotions are accompanied by more sophisticated ones like jealousy, deception, but they are all determined by the status quo: their world as it is.
This really should not surprise us. After all, humans are also capable of living the same routines for thousands of year without any significant change. But if you pluck someone from deep in the Amazonian jungle and drop him in a modern metropolis -the equivalent of a travel in time of 10000 years!-, he will eventually learn how to cope, if not thrive.
In my model, that means that humans have flexible connections between the different arrays their brain is made of, whereas animals are limited by direct connections.
We discovered, with the Amys, that they had only three fundamental arrays:
- sensations/emotions and their copy,
- external stimuli and their copy,
- actions and their copy.
Every direct link between one element of an array and the others, will limit the number of possibilities. This is different from the distinction in statistics between permutations and combinations. Only combinations (1 of each array in any order) are meaningful in this context.
Kanzi's behavior does not seem to answer this description. After all, he is capable of using fridges, can openers, gates etc. All modern objects that have no equivalents in his natural habitat. But they are all objects that can be manipulated with the same action repertoire he already possesses.
Language means applying the action array to any other array. Also, it can only originate, in my model, in the sensation array.
Instead of, to keep it simple
stimulus---> action
we have
virtual stimulus--->virtual action--->real action (gesture or speech).
There is no reason for the real action not to be possible, unless all Kanzi is experiencing is:
real stimulus--->real action.
That would mean that Kanzi is able to imitate making fire, but unable to imagine himself making fire on his own. Making fire does not belong to the standard, virtual, repertoire, even if it is a concrete possibility.
The making of stone tools is a little more complex, at first glance. But the fact that the tools are discarded after use, would also indicate that it is something that can be learned, but only as a concrete action, and not as an abstract pattern, which would necessitate a virtual copy.
edit: "Me Jane, you Tarzan" is grammatically correct in many languages. Native speakers of those languages who learn a second language like English can get very confused by, in their eyes, or ears, redundant "am" and "are", whereas their children have no difficulty using both grammars appropriately. In the case of humans, it is more a matter of lack of flexibility in the face of new patterns, For apes, it looks like a congenital impediment.
