The Philosophers' Cocoon

I just received the following query from one of our readers (a grad student):

I had a question for you
and the fellow cocooners that I thought might be an interesting topic. I've got
a paper that I've been working on for a couple of years and I can't figure out
how to make it into a journal type article. I have been thinking recently about
submitting it as a discussion piece. I've seen plenty of stuff on the web about
publishing articles, but I haven't seen much about publishing discussion
pieces. Is there a rule of thumb for the content? How recent does the work
being discussed need to be? Any help would be appreciated. 

What say you, my fellow pupae? Any advice?

Here are a few thoughts of my own. The first piece I ever published was actually just the kind of piece the reader refers to: a longer paper that I had trouble placing, and which I then transformed into a much shorter piece. So I guess my first thought is that this kind of thing can work. A second thought is that Thom Brooks' guide to publishing for grad students is a good place to look, since he discusses replies there a bit. His advice, if I recall, is to make sure that the piece makes one very good point very clearly, and stays on point at every instant. In other words, the piece should probably begin with a very short summary of the point — and no more than the point — you mean to object to, and then provide the objection as quickly and powerfully as possible.  As for how recently the piece being discussed needs to be, I don't know if there's a good rule to follow here besides a "sliding-scale": namely, the newer the piece, the better.  An exception would perhaps be a slightly older piece that is very influential.  My final thought (echoed by Brooks) is that it's probably not a good idea to spend a great deal of time on any given reply.  Since very few journals accept replies to pieces in *other* journals (JESP is a rare exception), you pretty much only have one shot to get the piece in print. I have several replies myself that went nowhere, and I'm glad I didn't waste any more time on them than I did. In other words, I suggest that if you want to write a reply, just write it up quickly, maybe get a couple people to read it, edit it for gross errors, and send it off. If the point you're arguing for really is a good one, you'll probably get an R&R and be able to polish it later (this happened to me twice).