On Agrippa's Trilemma

Chapter 8 of Luck is on skepticism. Duncan first presents the metaepistemological skeptical challenge, then classical Pyrrhonian methodological skepticism.

Here's the trilemma:

If challenged on a claim to know, one has open to them the following options:

1. Refuse to respond.
2. Repeat a claim made earlier.
3. Say something new.

Duncan glosses these options this way:

1'. Make an undefended assumption.
2'. Reason in a circle.
3'. Embark on an infinite regress.

I'm not sure 1'. is the best way to interpret 1, but my focus is on 2'.

Consider this statement from Carnap quoted in Chisholm's "The Myth of the Given":

"When an experiential sentence is accepted 'on the basis of direct experiential evidence,' it is indeed not asserted arbitrarily; but to describe the evidence in question would simply mean to repeat the experiential statement itself."

This is not exactly "coherentism" as the term is usually understood and which 2'. would suggest, but it does satisfy 2. of the Pyrrhonian Problematic above represented in Agrippa's Trilemma. (though see my comments on the potential fuzziness of that line here)

Chisholm's defense of the doctrine of the given is masterfully summed up by Matt McGrath:

"Having made the statement "There lies a key," if one is asked 'What is your justification for thinking that?,'' one must provide an answer, but eventually, in the chain of questions, a claim about one's present experience will be challenged: 'What is your justification for thinking you have such-and-such experience?' To this, Chisholm thinks one can do no better than to answer: 'My justification is that I have such-and-such experience.'" (Sosa and Kim, Epistemology, p. 105)

The main objection to this is the Sellers Dilemma and I think the best responses have come from van Cleve, Alston, and Sosa, but that's another matter.

So I don't think that 2'. is quite the right take on 2.