Garrett Dummett
—-
What is it?
A summary of Brian Garrett’s chapter on Dummett’s anti-realism.
Where can it be found?
Garrett, What is this thing called metaphysics?, Ch. 9: Realism & Anti-Realism.
—-
Summary
The Dispute - Dummett as trying to find middle ground between idelaist and realist — but still questioning independence dimension - crux of the debate: belief in statements having truth-value independently of us having the means to find it - dispute over what is the appropriate notion of truth (and thus over meaning) - relation of disputed class of statements to a reductive class of statements (e.g. other minds and behaviour) — anti-realist: claims in the disputed class actually amount to claims in reductive class - note that reductionism is not related to the realist debate in any simple way (see Metaphysics Realism for more on this) — in particular not necessary or sufficient for anti-realism
Case Study: Character - Under investigation: “Jones was brave” — with no possible evidence as to Jones’ bravery - Want to assess truth of counterfactuals: “if jones had been exposed to danger, he would have acted bravely” & “if jones had been exposed to danger, he would not have acted bravely.” — Dummett: counterfactual can only be true in virtue of some categorical statement — anti-realist: neither counterfactual T => “Jones was brave” is neither true nor false — realist: one or other of the counterfactuals T - Dummett then argues that one would have to be naive to be a realist about character — {[green but no details of argument given, will ahve to look into this ]} — BG: ~this on basis that it is intuitive to think about psychological states giving rise to behaviour — Hence Dummett characterised as a naive behaviourist.
Case Study: The Past - looking at statements about the past that we will most likely never have evidence for - Dummett: realist needs to appeal to truth-value links between present and past tensed statements — need grasp of this link to gain understanding of what it is for statement re: past to be T — problem: this seems to assume that the realist admits that there is difficulty re: past statements — looking to ground statements re: past in statements re: present
Maths - does appear to be a reasonable categorical difference in case of maths: when don’t have proof or counter-example — {[green But does this mean that Dummett would accept that one can be a realist re: mathematical properties that have been proved? ]}
Dummett’s Motiviation - D: Can’t understand what it is for a statement to be T without what we’ve learned to treat as establishing its truth. — {[green This seems quite pragmatic, presuming that we learn truth-establishing in a social way. ]} - prob re: Goldbach Conjecture: seem to know what it means without having ability to show it is true — {[green I think this highlights the large amount of contextual consideration at play in the background of Dummett’s position: consider for e.g. how Goldbach’s Conjecture woulf be treated in other areas of science (where inductive inference is accepted as valid). ]} - anti-realism as an attempt at improving logical positivism
Fitch’s Paradox - anti-realist committed to “for all p, if p T, it is possible to know p” - willing to accept that some truths are unknown - Paradox (K is knowing): - 1. q and ~Kq - Assumed - 2. Possibly K(q and ~Kq) - From anti-realist - 3. Possibly (Kq and K~Kq) - Dist. of knowl. over conj. and poss. - 4. Possibly (Kq and ~Kq) - Knowl. entails truth and possibility - 5. 4 is contradiction, therefore ~ Anti - Realist principle - Reductio coupled with plausability of other steps. - 6. Therefore ~anti-realism. - problems: — implies that if some truths are unknown, then some are unknowable — contrapositive: if all truths are knowable, then all truths are known
—-
What do I think? - Need to weed out just what role the social context of a statement plays in Dummett’s position. - Not sure that I agree with BG’s statement: “The sentence is true just if reality is that way; otherwise it is false.” P 147 — Does Dummett need to take on some form of anti-representationalism in order for his view to work though? — i.e. If one wants evidential considerations to be relevant to understanding, does one need to hold an anti-representationalist view of language? — otherwise we seem to be able to gain understanding just in the way BG indicates above.
—-