When is Anselm's argument Anselm's argument?

An interesting paper from earlier this year is P. Schlenker, 'Anselm's argument and Berry's paradox' in Noûs, 43 (2009) 214-223. A version can be found here.

A problem with this paper, like so many others, is that it claims to have found where Anselm's Proslogion argument goes wrong. It is interesting to note how many different, competing ways of going wrong there supposedly are in Anselm's argument. I suspect that there needs to be more care taken with the hermeneutics of translation from natural to formal logical language.

As Geach and Strawson famously pointed out, 'there is no one logical form that is the form of a given concrete argument; one and the same argument may correspond to more than one abstract schema; and accordingly an argument is not shown to be invalid by sharing some logical form with an invalid argument, for it may simultaneously have some form that makes it into a valid argument.' (P. Geach, 'Why Logic Matters' in H.D. Lewis, Contemporary British Philosophy: Personal Statements, Fourth Series, London 1976, pp. 86-99, p. 95.

One lesson to be drawn from this is that coming up with something that sounds like it might bear some relation to what Anselm thought does not constitute the basis for an account of what he thought that is sufficiently nuanced to allow one to draw the conclusion that his argument was unsound and/or invalid. At best, using such an approach one could demonstrate the invalidity or unsoundness of one's own account. But from the perspective of Anselm studies, that would not be very interesting. It would be useful to produce a set of criteria that would allow one to come to a conclusion about the formal validity at least of a complex natural language argument. One of those criteria would have to be that provided by Geach and Strawson: 'there is no one logical form that is the form of a given concrete argument'.