Spender

Nice, if slightly chummy piece on Stephen Spender's New Selected Journals in the most recent LRB by Karl Miller, who was there for much of this and isn't shy about telling us so.  This bit struck me in particular:

He had failed. He had failed even to fail.

I blame myself not so much for failure -- but for not having pressed ideas of work original work to the point of proof where they either failed or succeeded. What I blame myself for in a sense is that I didn't have enough failures -- but that I so often put aside the things I most deeply wanted to do -- the things that were my own thing from inside myself -- and did things that were proposed from the outside. [LRB 30 August 2012, 12]

Brave of him to admit it, as Miller notes. And striking. It made me think that, in my own work, I have (as it were) only failed, not failed even to fail. Though I must say, that's not quite as reassuring a circumstance as you, or Spender, might think.