The egregious error
Here is a typical instance of the grammatical error. Edwards wrote:
and his error-rich editor (ERE for short) changed this to
This is now sanctioned by the Oxford University Press, one of whose editors introduced it throughout a book of mine a few years ago, forcing me to go through the whole typescript cleansing it. But it is an error. What it produces is a mistake, wrong, ungrammatical, indefensible.
Here are two reasons for saying this.
First reason: The error turns something that can be grammatically explained or analysed into something that cannot. (1) says that Arminian liberty consists in an act of self-determination. Whose act? The will’s.
Nothing corresponding to this can be done for (2). The word-sequence ‘the will determining its own acts . . .’ isn’t a recognisable grammatical entity.Another example of the error:- Edwards’s correct
was changed by ERE to
There is no way of cleanly construing (4). Trying to move towards an understanding of it, one naturally thinks that it includes something of the form ‘whether the will [blah blah] is not the same thing as the soul [blah blah]’. But of course what is meant by (3) isn’t about equating the will with the soul, but rather equating a certain will-event with a certain soul-event. (Perhaps that equation entails that the will is the soul; but that isn’t what it says.)
Second reason: In many instances, everyone shrinks from the error, feeling that the possessive is required in the context of ‘-ing’. To avoid it in those cases while implementing it in others is to strike a further blow against the rule-guidedness (and thus the ultimate intelligibility) of our language. That wouldn’t be so if there were a clean line around the cases where the error feels wrong even to error-makers, but there isn’t. Very many of them involve possessive pronouns, including the following from the opening three pages of the error-mauled version of Edwards’s book:
Had the ERE remained true to his error, he would have written ‘is not him being removed’, which would have sounded grotesque, even to him.
Fidelity to the error would require ‘him willing’, ‘him choosing’ and ‘him liking’.
But it isn’t only with pronouns that the ERE shrinks from his error. This is also from the abused version of Edwards:
Perhaps he left the possessive “man’s” untouched here because if you change that to “man”, you get “. . . a man choosing, liking best, or being pleased with a thing, are not the same as . . .”, and the singular “man” clashes with the plural “are”. Just so; what is being talked is not a man blah blah, but a choosing, a liking best, a being pleased with - three items. Whose choosing etc.? A man’s.
But in the very next sentence, though the ERE still suspends the error for the pronoun, he returns to it for the noun:
The very same noun, and another case of singular/plural trouble; perhaps the difference the ERE felt between these comes from there being three items in (7) but only two in (8), which is less plural!
This is chaotic: the ERE was steering by vague, unexamined intuitions that had nothing principled behind them. The obviously best way out of the chaos is to restore the possessive in every instance.
There is an excellent further discussion of this matter on the internet.