Newsgroups: comp.ai.philosophy
Path: cantaloupe.srv.cs.cmu.edu!das-news2.harvard.edu!news2.near.net!satisfied.elf.com!news.mathworks.com!news.alpha.net!solaris.cc.vt.edu!insosf1.infonet.net!internet.spss.com!markrose
From: markrose@spss.com (Mark Rosenfelder)
Subject: Re: When is a simulation of a Y a Y? (Was Bag the Turing
Message-ID: <D3AB9n.Bs3@spss.com>
Sender: news@spss.com
Organization: SPSS Inc
References: <D2KLCv.C23@spss.com> <D2pzFL.L38@cogsci.ed.ac.uk> <D2xL96.AMB@spss.com> <D2z6t6.5r4@cogsci.ed.ac.uk>
Date: Tue, 31 Jan 1995 19:24:57 GMT
Lines: 90

In article <D2z6t6.5r4@cogsci.ed.ac.uk>,
Jeff Dalton <jeff@aiai.ed.ac.uk> wrote:
>In article <D2xL96.AMB@spss.com> markrose@spss.com (Mark Rosenfelder) writes:
>>Of course he hasn't *proven* that Hardin "can't bring himself
>>to say" plaid. 
>
>I'm not asking him to *prove* it; but "show" I didn't mean prove,
>and I'd accept much less than proof.  But why _does_ Dennett think
>this is what's going on?  What evidence does he give?

It may help to see a bit more of the passage from Hardin that Dennett cited.
The context is a discussion of differences between what we perceive
and the retinal image.  (The emphasis on "fills in" is Hardin's.)

  But if stabilized images disappear from view, what do we see in their
  place?  The remarkable answer is that the eye-brain _fills in_ the area
  with information gleaned from immediately adjoining retinal regions.
  Consider the receptorless optic disk, or "blind spot", formed where the
  bundle of optic fibers leaves the retina for the brain.  We recall that 
  this area is only 16 degrees removed from the center of vision.  It covers
  an area with a 6 degree visual diameter, enough to hold the images of
  ten full moons placed end to end, and yet there is no hole in the 
  corresponding region of the visual field.  This is because the eye-brain
  fills in with whatever is seen in the adjoining regions.  If that is
  blue, it fills in blue; if it is plaid, we are aware of no discontinuity
  in the expanse of plaid.  (_CFP_, p. 22)

Here and in other parts of this section (e.g. the only other section that
mentions blind spots, p. 8), Hardin uses entirely unobjectionable, neutral
language, saying only that we are not aware of the blind spot (or of our
own saccades, or of the shadows of retinal blood vessels, etc.).  

However, this passage seems to me to go much further than merely "saying
...that we don't notice the blind spots", as you claim; it actually declares 
that the blind spot is filled in "from adjoining retinal regions", and even
comments on how remarkable this process is.  Hardin does not say where
this notion comes from; one would have to examine his own sources to find
out.  I don't think that Dennett is misinterpreting anything Hardin
says about "filling in".  

About "filling in plaid", we can't be certain from this passage what Hardin
meant, but I'm dubious of your theory of "accidental syntactic variation".
The *direction* of the change seems significant; do you really think that
Hardin might equally well have written

  If that is plaid, it fills in plaid; if it is blue, we are aware of no
  discontinuity in the expanse of blue.

In other words, is Hardin really claiming that some neural process takes
complex patterns from the edge of the blind spot and reproduces them inside,
carefully lining up the pattern elements?  Perhaps, but he doesn't say
so directly, or explain how such a process would work, or why it is
hypothesized.

As for whether this idea is a relic of the Cartesian Theater-- what
would it take for you to accept that Dennett has "shown" this?  Citations
of Descartes, and discussions of the pineal gland?  All Dennett has to
show, IMHO, is that the authors he is discussing are assuming that the
brain has to construct a *visual image* out of what the eyes see, for
some part of the brain to look at.  Hardin's talk of filling in the blind
spot certainly sounds like such an assumption is at work, perhaps
unconsciously.

That the brain has to *interpret* the visual image registered on the
retina is not under dispute.  That it has to *edit* it, however, *as a
visual image*, is precisely what Dennett is questioning.  

His discussion of Shakey (p. 85+) raises the issue nicely.  Shakey displayed
a TV image of what its camera was recording, then displayed a corrected,
interpreted version of that image (a vector drawing of what it "saw").  It's
tempting to say that Shakey "looked at" that corrected image, but as Dennett
shows, quite wrong: the image was for the benefit of the human observer;
Shakey was not "looking at" anything at all.  

Similarly, it's tempting to imagine the brain editing a visual image
as it comes in from the eyes, as if we could tap into the nerves somehow
and reconstitute a nice visual image, seeing the improvements at each 
point, as if we were watching the output of a image enhancement program.

By proposing an alternative view, Dennett has not demonstrated that this
picture is *wrong*; he just forces us (if we want to keep it) to argue 
for it, rather than just assuming it.

>Before Dennett made a big issue of "filling-in",
>why would anyone have thought it was a particularly questionable
>thing to say?

This question, to my mind, quite misses the point.  It's like asking
why anyone would have thought that "simultaneous" was a particularly
tricky concept before Einstein made a big deal of it.
