“Both these worries aggravated my already long-standing misanthropy.” (J. L. Borges, “The Book of Sand”)

Jon Winokur runs a blog called Advice to Writers and an associated Twitter account to remind us of the “writerly wisdom of the ages”. The other day he tweeted Shannon Hale’s approach to writing a first draft, which she describes as “shoveling sand into a box so that later I can build castles.” I retweeted it with a caveat. “Please remember,” I said, “that what is good advice for writers of young adult fantasy is not necessarily good advice for early career researchers.” Novelists and literary types are not always good models for researchers and scholars. You may admire the result, but you don’t want to write like Henry “Vas-y” Miller.

Jo Van Every, an academic career guide, came to the defense of Hale’s metaphor,  however, suggesting that early career researchers, perhaps, “already have sand in a box. They are now using that sand to create different published outputs.” This was a good prod for me to clarify the issue I have with the image of writing as filling boxes with sand. I prefer to think of scholarly writing as building something out of relatively well-defined and sturdy parts, I said, not as shoveling and shaping a mass of undifferentiated particles. I could have added that I don’t think it is helpful to think of producing scholarly output on the model of weeding a garden or pruning a tree. That metaphor has other, perfectly legitimate, uses in academic writing.

The correct metaphor, if you ask me, is that of a construction. It’s not as limited as, say, Lego, but more like a constructor set that also lets you include bits and pieces of everyday reality, ordinary household objects and other toys. There are general structural elements and more specialized parts. Your paper will present a theoretical “framework”, for example, built out of concepts that your reader recognizes and it will then then put it to work by subjecting it to a “load”, i.e., by introducing data that has been gathered according to a methodology that, again, is recognizable to the reader. A paper can certainly “fall apart” on you (or in the hands of your reader) but it cannot, meaningfully, be “smashed to atoms”. Its meaning does not erode like a castle in the sand.

Maybe some novelists have a more rugged conception of their materials but, like I say, I’m not going to tell novelists how to write a first draft or how to think of their writing process. I’m just trying to help scholars avoid a less than apt metaphor with which to understand their own writing.

In any case, Jo rightly reminded me that researchers aren’t usually “starting from scratch” when they’re writing journal articles. They’ve “already got a conference paper, a working paper, pages and pages of analysis” or some basis like that to proceed from. This happens to be something I have an opinion about too., and I answered that conference papers and working papers are best seen as unfinished journal articles. They should be written in the same way. You still need to decide what to say (i.e., what you know) before you begin one. As for “pages and pages of analysis”: I would encourage researchers to think of them merely as a warmup. After they have helped you decide what to say, throw them out. Now write what you know for the purpose of discussing it with your peers.

Understandably enough, this suggestion puzzled Jo. “Isn’t that ‘warm up’ the first draft?” she asked. “I’m not sure what is gained by calling it something that isn’t writing.” And this is indeed exactly my point. The metaphor we are evaluating is one of shoveling sand into boxes and then later shaping that same sand into castles. I’m suggesting that you should not try to shape your drafts, and therefore that you should not produce them as though they are made of a “malleable” substance. Instead, write it around claims you identified through the “free-writing” process.

I realize that it is counter-intuitive to say that that process isn’t actually “writing”. But I really do believe that it stands in the same relation to your final text as drawing a mind-map, talking to a colleague about your results,  or just going for a walk and thinking things through. It’s as far from scholarly writing as that. Or as close to it, if you will. And here Hale’s image of a box of sand may have some carry after all.

Some researchers (especially ethnographers, I have found) approach the writing of their analysis as shoveling particles of experience into their paper where they will gradually be given meaning. That is, they are simply importing their data set into their word processor, which they think of as a tool to help them with their analysis. (There are much better tools for this purpose, I’m told.) In the first instance, it’s just a box to distinguish their “sample” from the “population”. If they had been working with more quantitative data, there would be no confusion here; it obviously wouldn’t be writing. But because qualitative analysis is, indeed, very much like drafting a novel–they are drawing , not just on their interview transcripts and field notes, but  also on their memory of their research experiences–it feels like they are actually in the first stages of their writing. This is the feeling I’m trying to get writers to understand better.

Just because you are putting words together, even in sentences, doesn’t mean that you are writing. You might, for example, be speaking. Even if you are typing, you might be transcribing or, to come closest to drafting a novel, thinking “out loud”, i.e., transcribing what is on your mind about something. But to be really engaged in scholarly writing is to be composing a paragraph–at least six sentences and at most two-hundred words that say one thing and support, elaborate or defend it. If you’re not doing that you may as well be talking or drawing a picture … or, of course, thinking. That’s also something you should do, of course. But it isn’t writing.