Xref: utzoo rec.birds:477 sci.bio:995
Path: utzoo!mnetor!uunet!lll-winken!lll-lcc!lll-tis!ames!pacbell!att-ih!cuuxb!ltuxa!ll1a!cej
From: cej@ll1a.UUCP (Jones)
Newsgroups: rec.birds,sci.bio
Subject: Re: Intelligent Parrots, or Self-deception and Gullibility.
Message-ID: <2037@ll1a.UUCP>
Date: 10 Mar 88 02:48:00 GMT
References: <1988Mar4.162334.18184@utzoo.uucp> <4687@sigi.Colorado.EDU>
Organization: AT&T, Lisle, Il.
Lines: 61
Summary: Clever Hans

In article <4687@sigi.Colorado.EDU>, (Anthony Pelletier) writes:
> In a typical demonstration, she, or one of her assistants will
> hold up a few different objects that are all the same color and
> ask "what same?" to which alex will respond "color" and, if asked,
> tell what the color is. Or, several geometric shapes of the same
> color might be held up and Alex asked "what different?" to which
> he would respond "shape." 

	From this description it seems that the animal may just be
expressing a classical conditioned response to external stimuli.

	All this really indicates is that a parrot (or, more
properly, a macaw) can distinguish between shapes.  (I don't believe
that macaws have color vision.  It may be that a certain "answer" is
associated with a certain combination of items.)

	As far as the questions the trainer asks go, macaws can
differentiate between different "human sounds", without
understanding their meaning.  A friend's macaw can be prompted to
"speak" any number of phrases that it knows by hearing that phrase.

> I think the "clever hans" model, mentioned by R. Snell, has not
> yet been ruled out.

	No "Clever Hans like cuing" is required in this case.  The
objects and the question are cuing enough.

> In the "Hans" case, it was discovered when the researchers
> realised that the horse would not perform when his master was not
> in the room. Presumably, Hans watched his master and stamped his
> foot until the trainer reacted in some consistant way which hans
> had learned meant "stop stamping." He would do this no matter who
> asked the questions. But, with the master out of the room, there
> were no cues given.

	Not exactly accurate.  Hans didn't perform as well with his
master out of the room, and his performance varied depending on who
was present.  It seems Hans could read the "unconscious cuing" that
many people seem to exhibit.  If you were watching Hans, might your
eyebrows go up when he got to the right number?  Might you cross
your arms?  Shift your weight?  Blink?

> Does anyone know if Alex the parrot will respond the same way for
> other people even if Pepperburg (or other trainers) is not
> present? 
>
> -tony

	With classical conditioning, it wouldn't matter.  The real
trick with Hans was that anyone could ask any questions that could
be answered a number of stamps, not just a fixed set of questions
dealing with a fixed set of objects.

	But it would be interesting to see just what shapes the bird
can tell apart.  What kind of depth perception does it have?  Does
shadowing or contrast play the biggest part in its perception of
shape?

...ll1a!cej		[Just me, not AT&T]		Llewellyn Jones

	"Can't act, can't sing, can dance a little"
			- results of Fred Astaire's screen test