I was a reading some proofs of an article yesterday, one that you will soon be reading too if you would like since it will on line. It was one of the first in which I deliberately suppressed obvious "signposting" while still signaling the direction of the argument clearly in a more implicit way. I was very pleased with the writing in this article: enough time had passed so that I had not remembered the specific thought-processes that went into the prose composition, so I could judge it almost as another person. I did remember that I had purposefully ended a few sentences with a preposition and written others with a slightly more colloquial tone than I usually employ.
I did notice some metadiscursive makers that were not signposting per se, but served that function. In other words, I didn't use markers like "In the next section I will consider...." but I did use phrases like "in other words." I could have gone a little further even here, but I am still pleased with my progress. When you read it I would like you to tell me whether you would miss the signposting, or whether I could take the next step and eliminate even more meta-discourse.
There was also one very hideous-sounding sentence with two syntactically ambiguous adverbs sticking out like bony elbows. It's fine if there is only one, though of course other readers might find more they don't like.
What I really want to see is the editor's introduction to this issue of HIOL and the afterword, to see how other people construe my argument. That will be a more significant test.