[hist-analytic] Question for Bruce on
Baynesr at comcast.net
Baynesr at comcast.net
Sun Nov 8 09:49:03 EST 2009
One reason for dwelling on this is that in connection with the paper
Bruce cites Putnam provides a very detailed account of what amounts
to being a definitions of definite color. Unlike Bruce, one of his
central concerns arises from the realization that color is to be undertood
as a continuum. Because of this there are no discrete colors, since
in a continuum there is always a color between any two others. So
all this business about fuzzy sets we have in Bruce's account is
unnecessary. Instead we have a bunch of postulates, but a very good
accounting of what a determinate shade is in terms of the relation
'exactly the same color as'. The procedure is familiar, going back to
the definition of number in Russell, where we define a number in terms
of sameness of correlation etc.
But for Bruce there ARE determinate colors and, so, there is presumably
no color continuum in Putnam's sense. Can Bruce "get away" with not
viewing color as a continuum? Only if he can give a definition of what it
is to BE a determinate color! This, despite his protestations regarding my
failure to understand, has not been done. We don't say a "determinate" dog
can be defined as whatever dog we happen to see. There is no continuum
running from beaglehood to snouzerhood, etc. But set this aside. If color
is a continuum then in the sense that there is always a color between
any two others there are no discrete colors, which is the view I take.
You see, once you reject the idea of the continuum then an explicit
definition of what such a discrete color is is something quite significant,
as Putnam fully understood. I will respond, briefly, to each of Bruce's points.
1. "I did talk about how three different
people classified the color (the specific, determinate color) that all three of them
saw. But I had noting to say about a mere appearance."
You repeatedly describe What 'Tom describes' what 'Mary describes' what 'Harry...describes'
as "the color." Until you tell me what it is that makes what they describe a
distinctive color, then there will be some question of what we are talking about. But
couching matters in terms of how people might, rightly or wrongly describe as
"the" color obscures more than clarifies what you mean by a "determinate" color.
What is a determinate color, so described? Again, reporting on how people report a
color doesn't get us down to an answer to the question what IS a determinate color?
Do you believe colors exist; can a color be individuated; what is a color; are
colors secondary properties? These are the sorts of things people are wondering
about when they wonder what a "determinate" color is. So when you say "the specific
determinate color" you use a term you have yet to explain. The confusion Tom or Harry
may experience in arguing over greenish-blue or bluish-green serves only to illustrate
the need to resolve the matter. Their confusion doesn't substitute for an answer or
even the beginning of an answer.
2. "Because the shades in question are
not sharply delimited. There are a great many clear cases of red shades
(vermillion, scarlet, red madder would be instances) but there are also many
Well, then, it appears there are no determinate colors! And if there were
we still would know what they are. You cannot have a determinate color when
that color is understood as made up of "fuzzy sets." You are trying to have
your "fuzz" and "determinate" colors at once. This strikes me as counterintuitive.
What is not counter intuitive is the view I would take: there are no
determinate colors! Let's put it this way: how do we distinguish there being
no determinate colors from there being "fuzzy" sets of colors, only? This
brings us to your next point.
3. "When you conclude, "so it's not a crazy question to ask how you get
determinateness out of a fuzzy set," you get things exactly backwards:
We have no need to get determinateness out of a fuzzy set; we have to
form a set from the determinate shades that we can encounter in experience."
Now if we form a set consisting of determinate shades, then wherein lies the
fuzziness? Can we say that a "determinate shade" is that particular shade
we encounter whenever we experience an ordinary object? Whether there are fuzzy
sets is, on your view it seems, irrelevant to what constitutes a "determinate"
shade, since the shades are determinate but it is merely *the sets in which
they are included that are fuzzy*? But what we are interested in is what makes
the shade a determinate one, regardless of what set it belongs to, fuzzy or not.
3. "The ball you see (assuming it to be homogeneous in color)
is, if red, a determinate shade of red."
All shades are determinate it would seem. But what do you mean when you say this?
What account can you give that makes it true?
4. "When I speak of a determinate color, I mean the specific color shades that
people can see."
Now we are stuck with "specific color."
5. "Steve's misunderstanding of my discussion may have resulted from an ambiguity
that is attached to a world like "red." A red object--that is, an object belonging
to the class of red things--may change in color but nevertheless remain red. How
is this possible? Because it may change from being scarlet to being vermillion.
Things that are scarlet are rightly classified as red and so are things that are
vermillion. As I see it, it is entirely possible (conceptually) for an object
with a single shade of color to be rightly classifiable (according to some
accepted classificatory scheme) as an instance of two different generic colors--for
instance, green and yellow. But no thing, I argue, can have more than one specific
color-shade (if it is not spotted, for instance)."
I agree with most of this. More later, perhaps.
Bruce concludes saying (among other things):
"...(countable and uncountable) Any of a range of colours having the longest wavelengths,
670nm, of the visible spectrum; a primary additive colour for transmitted light: the
colour obtained by subtracting green and blue from white light using magenta and
Once you start talking about color as wavelengths then when you argue that
nothing can be two colors "all over" you have to give an account of what "all over"
means in this context. Here the issue of surfaces emerges; there are no sufaces to
a field, and an object having these spectral properties does not have a surface
in any definable topological sense.
Color has always been a hot topic in philosophy. Whether they are quale, universals,
pariculars, tropes, etc. All these questions figure in the issue we are discussing.
We cannot dismiss them with a wave of the hand.
I'm cutting off the message I'm replying to; it screws up the archives, I'm told.
I'm formulating a "policy" of sorts in deferrence to the archives, which are becoming
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the hist-analytic