[ Pobierz całość w formacie PDF ]
can think about this experience, then and later, as the experience: --- , filling the gap
with a state of perceptual classification or re-creation. And she can use the concept so
formed to think thoughts with determinate truth conditions, as when she hazards I'll have
that experience again before the day is out , or wonders whether or not That experience
is the one normally produced by ripe tomatoes .
Yet Mary's concept looks like a paradigm of the kind of thing Wittgenstein's private
language argument is designed to discredit. For a start, Mary's use of the concept will not
conform to any public criteria. Since there are no a priori links between phenomenal
concepts and psychological ones, Mary's mere possession of the phenomenal concept will
give her no idea of the characteristic external causes or behavioural effects of her new
experience. Nor will she be able to communicate the thoughts that the concept enables
her to form: if she coins a word ( qual , say) to express the concept, she will not be able
to convey to her hearers what it means. Even so, I say, Mary's concept is a concept in
good standing, in that it enables her to form thoughts with definite truth conditions. If this
is so, then neither conformity to public criteria nor communicability can be essential to
determinate thought.
There are two issues here: conformity to public criteria and communicability. Let me deal
with these in turn. Public criteria first. If you think that representational content is
somehow constituted by normative rules governing the deployment of
end p.128
concepts, then you may be inclined to resist the suggestion that Mary has a good concept
even in the absence of public criteria. How could Mary's concept possibly have a
determinate content, you will ask, if Mary is not sensitive to any normative principles
tying its use to public criteria? However, I take this line of thought to cast doubt on the
premiss that concepts require such normative rules. Since Mary clearly can think good
thoughts with her new concept, say I, it follows that normative rules are inessential to
representational content, at least the kind of rules that Mary lacks.
There are some large issues here, but my own view is that content does not derive from
normative rules, but rather from the kind of non-normative natural facts invoked by
causal or teleosemantic theories of representation. In so far as there are norms in the area
of judgement, these follow from the prior naturalistic constitution of content, and are not
a precondition thereof (cf. Papineau 1999). So, on my view, it is no deficiency in Mary's
concept that she is not sensitive to any normative principles tying its use to public
criteria. It is enough that her concept has appropriate causal or teleosemantic credentials,
since this in itself will ensure that her concept refers determinately, and that judgements
made by using it have definite truth conditions. (Of course, if we assume that it is
correct to make true judgements, and incorrect to make false ones, then Mary will be
subject to the norm that she should judge truly; but this norm doesn't require that
Mary be sensitive to public criteria, only that her judgements have truth conditions,
which requirement I take to be satisfied, for the reasons given.)
What about the incommunicability of Mary's concept? ( Well, let's assume the child is a
genius and itself invents a name for the sensation! But then, of course, he couldn't make
himself understood when he used the word : Wittgenstein 1953: § 257.) Again, since I
take Mary to have a concept in good standing, I do not take communicability to be
essential to determinate referential content. The thoughts Mary forms with the concept
the experience: --- have quite definite truth conditions, even if she can't communicate
them to anybody else.
end p.129
It would be worrying, however, if phenomenal concepts were necessarily
incommunicable, if no one else could ever understand words used to express phenomenal
concepts, even outside the special circumstances of our Mary thought-experiment.
Certainly much of the argument of this book presupposes that such communication is
possible. However, there is no great difficulty here. Our Mary may not immediately be
able to make herself understood to normal English speakers with her term qual . But
nothing stops other better-placed speakers from communicating their phenomenal
concepts, or indeed our Mary herself doing so, given more propitious circumstances.
What exactly is required to understand someone else's expression of a phenomenal
concept? A weak requirement would be that you understand that the speaker is
expressing a phenomenal concept, and that you know which experiential property it
refers to. A stronger requirement would be that you be able to identify this experiential
property via the same phenomenal concept, and not just via some material concept.
To see that there is no principled barrier to understanding expressions of phenomenal
concepts, in either the weak or the strong sense, consider our Mary example again. To get
a case of someone who satisfies the weak requirement, but not the strong one, let us
suppose that Mary has a companion, Jennifer, who similarly has never seen colours but
knows all about colour vision in material terms. Jennifer isn't shown the piece of paper [ Pobierz całość w formacie PDF ]
zanotowane.pl doc.pisz.pl pdf.pisz.pl rafalstec.xlx.pl
can think about this experience, then and later, as the experience: --- , filling the gap
with a state of perceptual classification or re-creation. And she can use the concept so
formed to think thoughts with determinate truth conditions, as when she hazards I'll have
that experience again before the day is out , or wonders whether or not That experience
is the one normally produced by ripe tomatoes .
Yet Mary's concept looks like a paradigm of the kind of thing Wittgenstein's private
language argument is designed to discredit. For a start, Mary's use of the concept will not
conform to any public criteria. Since there are no a priori links between phenomenal
concepts and psychological ones, Mary's mere possession of the phenomenal concept will
give her no idea of the characteristic external causes or behavioural effects of her new
experience. Nor will she be able to communicate the thoughts that the concept enables
her to form: if she coins a word ( qual , say) to express the concept, she will not be able
to convey to her hearers what it means. Even so, I say, Mary's concept is a concept in
good standing, in that it enables her to form thoughts with definite truth conditions. If this
is so, then neither conformity to public criteria nor communicability can be essential to
determinate thought.
There are two issues here: conformity to public criteria and communicability. Let me deal
with these in turn. Public criteria first. If you think that representational content is
somehow constituted by normative rules governing the deployment of
end p.128
concepts, then you may be inclined to resist the suggestion that Mary has a good concept
even in the absence of public criteria. How could Mary's concept possibly have a
determinate content, you will ask, if Mary is not sensitive to any normative principles
tying its use to public criteria? However, I take this line of thought to cast doubt on the
premiss that concepts require such normative rules. Since Mary clearly can think good
thoughts with her new concept, say I, it follows that normative rules are inessential to
representational content, at least the kind of rules that Mary lacks.
There are some large issues here, but my own view is that content does not derive from
normative rules, but rather from the kind of non-normative natural facts invoked by
causal or teleosemantic theories of representation. In so far as there are norms in the area
of judgement, these follow from the prior naturalistic constitution of content, and are not
a precondition thereof (cf. Papineau 1999). So, on my view, it is no deficiency in Mary's
concept that she is not sensitive to any normative principles tying its use to public
criteria. It is enough that her concept has appropriate causal or teleosemantic credentials,
since this in itself will ensure that her concept refers determinately, and that judgements
made by using it have definite truth conditions. (Of course, if we assume that it is
correct to make true judgements, and incorrect to make false ones, then Mary will be
subject to the norm that she should judge truly; but this norm doesn't require that
Mary be sensitive to public criteria, only that her judgements have truth conditions,
which requirement I take to be satisfied, for the reasons given.)
What about the incommunicability of Mary's concept? ( Well, let's assume the child is a
genius and itself invents a name for the sensation! But then, of course, he couldn't make
himself understood when he used the word : Wittgenstein 1953: § 257.) Again, since I
take Mary to have a concept in good standing, I do not take communicability to be
essential to determinate referential content. The thoughts Mary forms with the concept
the experience: --- have quite definite truth conditions, even if she can't communicate
them to anybody else.
end p.129
It would be worrying, however, if phenomenal concepts were necessarily
incommunicable, if no one else could ever understand words used to express phenomenal
concepts, even outside the special circumstances of our Mary thought-experiment.
Certainly much of the argument of this book presupposes that such communication is
possible. However, there is no great difficulty here. Our Mary may not immediately be
able to make herself understood to normal English speakers with her term qual . But
nothing stops other better-placed speakers from communicating their phenomenal
concepts, or indeed our Mary herself doing so, given more propitious circumstances.
What exactly is required to understand someone else's expression of a phenomenal
concept? A weak requirement would be that you understand that the speaker is
expressing a phenomenal concept, and that you know which experiential property it
refers to. A stronger requirement would be that you be able to identify this experiential
property via the same phenomenal concept, and not just via some material concept.
To see that there is no principled barrier to understanding expressions of phenomenal
concepts, in either the weak or the strong sense, consider our Mary example again. To get
a case of someone who satisfies the weak requirement, but not the strong one, let us
suppose that Mary has a companion, Jennifer, who similarly has never seen colours but
knows all about colour vision in material terms. Jennifer isn't shown the piece of paper [ Pobierz całość w formacie PDF ]