Chimp English

[dave pierce]

Members

Could someone direct me to research by Savage-Rumbaugh that describes the
method used to teach Kanzi to talk English? One report states that
"instead of using behvaiorism--rewarding the apes with or each time they
use a word correctly--she allows the amimals to pick up words in "normal"
conversation..." Do the chimps simply absorb the English language?
Exactly, how does she get this behavior and can it be replicated?

All the best
David

---------------------------------------------
To join the Behavior Analysis (Behav-An) forum, send the command
SUBSCRIBE BEHAV-AN YOURFIRSTNAME YOURLASTNAME;
to leave the forum, send the command
SIGNOFF BEHAV-AN
to LIST...@LISTSERV.NODAK.EDU or, if you experience difficulties, write to BEHAV-AN...@LISTSERV.NODAK.EDU.

[Joseph Cautilli]

It is not English, for chimps can not speak. It is American sign. Anyhow,
teh training difference between Sherman and Austin and her latest Kanzi is
that Kanzi is not discrete trial trained but more simlarly trained to free
operant situations. Alot of incidental teaching is used and a lot more
context design. If interested please readthe book KANZI

Joe

[Douglas Cunningham]

Actually, Joe, It was English that Kanzi was taught...for chimps can HEAR.
You see, most of the chimps at Yerkes (or whatever that institute is called
now that it has spun off from Yerkes) are taught a highly developed
artificial language (I believe it is called LANA, as is the first chimp it
was taught to). The language consists of a large array of graphemes --
pictures that exhibit functional relationships much like the relationships
exhibited by written English words (or, in cogno-babble...the pictures
represent concepts and ideas, and have meaning...*snicker*).
As for Kanzi, if I remember correctly, Kanzi's instruction was initially an
accident. The Yerkes researchers, like most chimp language folks, wait
untill a chimp has reached several years of age before begining language
instruction. Kanzi's mother's training progressed in much the same fashion
(She was, I think, one of the first Bono chimps to be taught the language).
Since Kanzi was being nursed at the time of the language instruction, Kanzi
was allowed to accompany the researchers (and Kanzi's mother) during their
walks and other training sessions, etc.
The "happy accident" with Kanzi occured this way: the human researchers
had developed the habit of talking out loud (in English) as they punched up
the correct graphic sequence (much in the same way that hearing adults will
speak or mouth the English equivalent words when using sign language). The
researchers noticed that Kanzi (who happened to be sitting in on the
mother's training) seemed to respond to (or "comprehend") the auditory
commands, etc. SOOO, they began to train Kanzi with the graphic,keyboard
language with the spoken accompanyment (Kanzi could hear the english, and
act on what was said, but could not talk back...until they taught Kanzi the
keyboard language).
While I have not read the book you mentioned, I suspect that it has been
reduced to mostly anecdotal ramblings, in an editorial attempt to capture
the non-technically trained audience. Perhaps a better reference for the
interested is a monograph that the Savage-Rumbaugh's came out with several
years ago (alas, the I have misplaced the reference, and my copy. It came
out some time after 1993 and before 1997) The monograph describes, among
other things, their instruction of Kanzi and a series of interesting
experiments where they directly compared the verablly directed behaviors of
Kanzi (e.g., a voice over a PA system comes on and says, "Kanzi, get the
red elephant and place it in front of the Green tiger"...and the
researchers measure Kanzi's compliance with the command) with the verbally
directed behavior of their (age-matched) daughter! (The 4 year old chimp
did slightly better than 4 year old human...but authors mitigated this blow
to human superiority with the insistence that eventually the daughter would
well outperfrom Kanzi, while Kanzi was not expeceted to get much more
proficient with the language...)
Anyway, fascinating research. The field certanly could use quite a bit of
input from radical behaviorism (for example most psycholinguist and animal
language folk have a fuzzy definition of verbal behavior (or "langauage"),
if they have a definition at all!).
-dc

------
Do you do Linux? :)
Get your FREE @linuxstart.com email address at: http://www.linuxstart.com

[Genae Hall]

For a behavioral perspective of the contingencies involved in teaching
verbal behavior to Kanzi, it would seem useful to look at Mark Sundberg's
(1996) JEAB review of Savage-Rumbaugh et al.'s (1993) monograph: Language
Comprehension in Ape and Child. Very briefly, Savage-Rumbaugh et al.
claimed that Kanzi learned to "associate words of interest to him with
symbols" without explicit training. In their view, no training or
reinforcement occurred because food reinforcers were not used. In this "no
training" procedure, the caretakers interacted with Kanzi during daily
activities as if he was a young human child. Indoors, "Kanzi's symbol board
was attached to a speech synthesizer so that he could 'talk' aloud by
pressing the symbols." Although Savage-Rumbaugh et al. did not specify the
environmental contingencies involved in Kanzi's acquisition of verbal
behavior, Sundberg illustrates how they likely included verbal prompts,
conditioned reinforcers (praise), conditioned punishers (reprimands), and
extinction. This informal approach was also used with a typically
developing 2-year-old young human child, and Kanzi's language acquisition
was comparable to that of the child.

One advantage of these "naturalistic" conditions is that they would seem to
promote automatic reinforcement of verbal behavior (like the conditions
which seem to promote high rates of verbal behavior described by Hart &
Risley in their book Meaningful Differences). It appears that Kanzi's
caretakers were frequently pairing verbal stimuli (lexigrams that were
selected and the auditory response-products of talking) with highly
reinforcing items and activites. As a result, similar verbal stimuli
produced by Kanzi may have acquired the capacity to automically reinforce
his verbal behavior.

It is unclear whether the exact approach used with Kanzi and the 2-year-old
child was also used with the animals described in the UK TImes article. If
it was, automatic reinforcement might conceivably account for why "some
animals were successfully copying human words and phrases."

Finally, I'm not sure if the apes are "speaking English." They are, however,
functioning within an English verbal community. The lexigrams are not
English--yet they are programmed by the experimenters to produce English
words via the speech synthesizer. The apes are producing English verbal
stimuli by selecting particular lexigrams, and trained English listeners are
responding.

Genae Hall

>> Members
>>
>> Could someone direct me to research by Savage-Rumbaugh that describes the
>> method used to teach Kanzi to talk English? One report states that
>> "instead of using behvaiorism--rewarding the apes with or each time they
>> use a word correctly--she allows the amimals to pick up words in "normal"
>> conversation..." Do the chimps simply absorb the English language?
>> Exactly, how does she get this behavior and can it be replicated?
>>
>> All the best
>> David

[McGinnis]

> RE: "they didn't use behaviorism"
>
> I have a dream. That one day man will see that his behavior is the
product
> of the workings of the principles of reinforcement. That researchers of
all
> disciplines will exhibit evidence of their knowledge, understanding, and
> reverence for behaviorism and its basic principles. That one day man will
> see the dismissal of behaviorism and its principles by his fellow man as
an
> admission of ignorance and arrogance.
>
> My two cents,
>
> chris
>
> J. Christopher McGinnis, Ph.D., NCSP, CBA/e/fl
> Psychology Resident, Family Psychology Clinic
> 4461 Camino Real Way, Fort Myers, FL 33912
> Office Phone: (941) 936-1336
> Office Fax: (941) 936-4927
> E-mail: mcgi...@gate.net

[John Davies]

if you look in a book by Richard Gross called Psychology of the mind and
Behaviour you will find a whole section on chimps learning ASL it tells
about alot of the main researchers in the field and gives lists of commands
learnt hope you find it helpfull
dave pierce wrote in message ...

>Members
>
>Could someone direct me to research by Savage-Rumbaugh that describes the
>method used to teach Kanzi to talk English? One report states that
>"instead of using behvaiorism--rewarding the apes with or each time they
>use a word correctly--she allows the amimals to pick up words in "normal"
>conversation..." Do the chimps simply absorb the English language?
>Exactly, how does she get this behavior and can it be replicated?
>
>All the best
>David
>