Ordinary language is all right.
One could divide humanity into two classes:
those who master a metaphor, and those who hold by a formula.
Those with a bent for both are too few, they do not comprise a class.
You can imagine Socrates asking an interlocutor how an argument goes—but as a prelude to the old routine, extracting something like a full performance from the interlocutor. Professional philosophers, in contrast, can seem to be suspended in rehearsal. They defer the performances: 'Let me just get it right first, how does it go again?'.
If you ask a musician 'how does it go?', he might sing a bit of it to you. Even if he's not a singer; even if he could, say, play it on the piano there.
He probably won't go look for a recording of it, either.
—Which shows something about what we do with arguments. It's hard to imagine someone being convinced by the gist of an argument. Or, in some settings, being asked, 'just gimme the gist of it'. By a judge in open court, for example. (On the other hand, in closed chambers, with just the lawyers…)
We say not only 'what's the argument?', but also 'how does the argument go?'. —We think there is such a thing as the gist of an argument.
My favorite guilty verdicts are the ones where the whole gallery gasps.
When I think of Kant's Groundwork, the absolute first thing I think isn't anything about the categorical imperative or a good will or anything like that—it's that it has four parts. (Yes, the preface counts.)
Maybe once you've read it carefully, the way you think about the Groundwork is by thinking about one of those other things. But I am not confident that people have read it carefully.
When I meet someone who starts from the other side (of the experience or activity of reading—keeping in mind what we'll see before we've read, thus our initial means of access to whatever else is to be had by reading, or to come of reading), I'm a little more confident.
Certain scholars like to write on Wittgenstein or Nietzsche by, essentially, constructing tissues of citations to their entire corpuses of published and unpublished works or notes. There are assumptions about Wittgenstein or Nietzsche, or about intellectual work or philosophical work in general, which can seem to validate this form of scholarly practice. Personality driving thought; continuity of preoccupations or style of thought; inherent stability of problems or their shaping effect on accurate or approximate solutions to them; organic conceptions of genius. That kind of thing. There are also all sorts of little scholarly norms whose satisfaction gives cover to work which would otherwise be better conceived and judged relative to the one who did it, but which because of its appearance of scholarly competence, 'has to be taken seriously' on impersonal or community terms. And, I suppose, less-little scholarly norms, or rather, maxims that characterize our practice. Like: if you can't see how to make something of it on its own terms, at least try to make something of it on 'our' terms, which I suppose is the sort of operating principle that gives rise to papers that pay lip service to the idea of 'Nietzsche's style' before quickly moving on to to state his 'views' or reconstruct his 'arguments' in ways that appear to deliberately neglect anything having to do with style. Small or large, these parts of our practice obviously play a big role in the attraction of scholarly writers and readers to work that mines Nietzsche or Wittgenstein for quotations. And something similar could probably be said of work with an aversion to this kind of citation-weaving in preference to cleanly stated views and arguments, attributed to Nietzsche or Wittgenstein with a kind of escape clause (think of Kripke's 'Wittgenstein's argument as it occurred to Kripke') that basically disclaims any interest in accuracy or fidelity of response.
When I think of the sorts of complaints or apologies made around these issues (they have to do with 'bad writing' or 'bad thinking' or 'bad philosophy' and their opposites), it occurs to me that the memorability of the original writing is rarely part of them. But from a reader's perspective (and thus that of a scholar whose work is supposed to result in writing), isn't the difficulty with remembering work by Wittgenstein or Nietzsche a leading source of our sense that they're so hard? By comparison, consider how hard it can be to keep in mind even one of Plato's shorter dialogues (and thus how astonishing it should be when they present themselves as recollections and recountings of entire conversations that would have lasted several hours).
Thomas' 'radical' idea about research practices and the writing process seems not so much radical as old to me (which is a fair way in which something might be radical)—remembering the stages of composition in rhetoric—but that may just be a reminder that advances in sophistication and complexity aren't necessarily correlated with progress. And a reminder of what the radicality of the writings of a Nietzsche or a Wittgenstein might really involve. What sort of an effort, what sorts of patience and commitment to dwelling on what one is really capable of holding in memory and thinking about, would be called for to work on and respond productively in writing to, Human, All Too Human or Philosophical Investigations in something like the manner Thomas describes? How much would it take in, considering the ways we usually have of dealing with texts?