Re: Communication based on private states

From: Harnad, Stevan (
Date: Mon Oct 23 1995 - 21:21:23 BST

ws> From: "Smith, Wendy" <>
ws> Date: Mon, 23 Oct 1995 10:42:55 GMT

ws> By "private states" are we meaning experiences or internal physical states
ws> (i.e. the experience of hunger rather than the physiological experience
ws> of a low blood sugar). In the seminar, we discussed the correlation
ws> of experiences to certain physical states. Although, as in the
ws> example of hunger, the two may correlate very well, one is physical,
ws> and the other is non-physical. As an experience is non-physical, we
ws> have no access to it, and so it can't be measured or observed.
ws> Humans can't share experiences; although they can share descriptions
ws> of experiences. If this is the case, sharing an experience with a
ws> pigeon would appear very ambitious.

Whether the source of the stimulation is external or internal is a bit
of a red herring. A pigeon responding to an external touch or an
internal itch does not cover a very useful distinction (except for the
behaviourist, who cannot see the internal itch). Both events have a
private, experiential side. The source of the stimulation is irrelevant.

And I hope what you mean above is that I have no access to YOUR
experience, rather than that YOU don't!

Let's not prejudge whether/how an experience is physical. What is sure
is that it is not measurable directly by anyone except the experiencer.
That's what makes it private; not that it comes from an invisible source
inside the body (which is what L & T seem to want to equate it with).

ws> To a certain extent, everything can be explained at a biological
ws> level, because we (and pigeons) are biological creatures, and
ws> therefore all behaviours (and experiences?) will be mediated through
ws> biological apparatus. However, although the experience may arise out
ws> of the activation of the biological apparatus, it is not the same
ws> thing as the activation (ie cause is not the same as effect, however
ws> closely related they are). Therefore, even if the same biological
ws> apparatus is activated in two beings, it does not
ws> mean that the experiences will be the same.

True, but it's a step in the direction if making it more likely rather
than not, wouldn't you say?

ws> Are the pigeons communicating an experience, or are they doing
ws> something else entirely? We see them pecking at keys, etc, but by
ws> assuming that this is due to an experience and that they are
ws> communicating this we could be anthropomorphising. If they are
ws> communicating, is this about a non-physical experience, or a physical
ws> internal state?

That is the question: Are they trying to tell signal to one another how
they feel, or they doing an elaborate sensorimotor, interactive, but
completely instrumental dance, choreographed by the experimenter, who
then projects a big linguistic overinterpretation on it?

ws> Are pheromonal communication and language best described by two ends of
ws> a continuum? Even if they are due to some form of quantitative
ws> difference, the differences may be sufficient to cause qualitative
ws> differences (like colour - it is a continuum of wavelengths, but these
ws> lead to qualitative differences). I think I'm talking about
ws> categorisation here.......

I highly doubt that the involuntary communicative effects of releasing
pheromones is on a continuum with language (you can neither do it
voluntarily, nor do you know when you're doing it; and the receiver
cannot actually consciously perceive them either; they react
involuntarily) -- or if they are, then it's on the same continuum as the
knee-jerk reflex...

> l&t> Linguists and philosophers single out the intent of the communicative
> l&t> exchange as an essential element in human communication.

> l&t> This does not mean that chimpanzees (or other organisms) cannot learn
> l&t> such discriminative repertoires, only that there is typically little
> l&t> adaptive utility for them to do so.

ws> Animals do not appear to share communication in this way spontaneously,
ws> and perhaps this is important. Chimps should, perhaps, be regarded as
ws> first class chimps rather than second class humans. "Chimpness" or
ws> "pigeonness" may be essential to an experience - as in what is it like
ws> to be a bat, the only way to find out is to be a bat. Perhaps something
ws> within a human experience leads to the sort of communication that chimp
ws> and pigeon experiences don't. Perhaps "why" is as important as "how".

But chimps are in many respects surprisingly like us, so it's all the
more suprising that, being smart too, they don't communicate more than
they do, especially symbolically. Social communication they show in
abundance, though perhaps more sluggishly and obtusely than us (and that
may indeed be a sign of difference rather than similarity).

> l&t> The pigeon experimental situation shares some features of the
> l&t> interactions between parents and children with autism or other
> l&t> developmental disabilities (Keogh, Whitman, Beeman & Halligan, 1987;
> l&t> Reichle, Lindamood, & Sigafoos, 1986). Autistic children are often
> l&t> minimally verbal and unable to report their private experiences to
> l&t> others. Indeed, a common diagnostic feature of autism is the failure to
> l&t> use gestures or other symbolic communicative devices to indicate needs
> l&t> or wants. Thus, to create the desired behavior or outcome in their
> l&t> parents, autistic children may pull or push their parents to the
> l&t> refrigerator and then begin to scream or to hit them until they open
> l&t> the door and retrieve the orange juice. Yet, just as in pigeons,
> l&t> extensive training can often be used to teach autistic children to
> l&t> discriminate familiar objects correctly (e.g., food items, articles of
> l&t> clothing, household objects, places).

ws> This may add to knowledge about how communication first develops in
ws> children, but again, "why" may be necessary to understand the
ws> differences between autistic and normal children. How well do pigeon
ws> studies address this point?

The fact that you can shape autistic children's behaviour by reward and
punishment is only evidence that autistic children, like us, are
trainable vertebrates. I think the resemblance between pigeons and
autistic children ends there...


This archive was generated by hypermail 2b30 : Tue Feb 13 2001 - 16:23:55 GMT