Remember the discussion a few months ago about Fowler’s commas-and-dashes-together idea – his insistence that a subordinate clause should always retain its closing comma, even if it ended with a parenthesis? Remember how it seemed unlikely that it would ever come up? Well, it’s just come up:
And, for the first time in an already incident-packed career, I want to put a comma straight after a dash:
”– though the IMF’s position has softened somewhat –, to protect themselves …”.
There are several places in the sentence where you could plug in the final clause. If you’re reading lazily and prepared to allow “the IMF” to be plural, you can almost read straight through the second dash without noticing it. An even likelier candidate is after the first dash – the Lure of Low Attachment again – until you realise, seven or eight words into the last clause, that it’s not really parsing. The correct place is high up in the sentence, after “controls” – and I think a Fowler comma after the dash would send you straight back there, as surely as a GOSUB/RETURN command would in Basic.
Of course, you could avoid any ambiguity by intervening – breaking it up into two sentences or repunctuating extensively, making a careful but denatured version of the original with the same ingredients. But I like this sentence: the rhythm, the length, the nested subordinates. One little extra punctuation mark would make it foolproof.
Except that the Fowler comma never caught on – never came remotely close to emulating the success of his which/that initiative for restrictive and non-restrictive clauses. No modern reader would understand why it was there if you put it in; you would cite Modern English Usage in vain to the puzzled audience, and even, probably, the writer.
But in a day when hands-off-the-wheel descriptivism is the intellectual consensus – in a time when “literally” is up on the ramp being safely drained of semantic coherence like a junkyard car being emptied of oil – it’s nice to know that, in another linguistics era, somebody cared. Cared enough to want to intervene; confident, or quixotic, enough to believe that things could actually be changed if you wrote with conviction and won hearts and minds. Jonathon Owen at Arrant Pedantry has been researching how professional editors can have an effect on the language, just as youth movements and technology do, if enough people are persuaded of the need for regularisation or clarity. And really, how else does language ever change except when a group of people are persuaded to adopt a usage that they find attractive?
Not this time, though. The only choice here is to trust the reader, or worry about the attachment ambiguity and pull it all to pieces for safety.
I’m going to leave it the way it is. I like it.