18 May 2015

CSCW Reflections

It’s been a while since my last post (I keep saying that, reducing the value of my apologies, but I’m going to keep apologizing anyway). My bad, sorry. I was working toward a CSCW deadline until about a week ago I realized that the paper I was writing was never going to go to CSCW (or if it would, it would get so soundly rejected that it wouldn’t have been clear whether I would have learned anything from the first round of feedback). Before I made this realization, I made some observations that I wanted to jot down and get back to later, and it’s sort of twofold:

  • Anthropological writing doesn’t really “get to the point” the way CS papers do (this is a gross oversimplification, but bear with me).
  • There’s a lot of really interesting CSCW and general HCI research going on out there that a lot of social scientists seem unaware of, and it might be worth getting vaguely familiar with this burgeoning body of work since, you know, a lot of researchers in CS readily benefit from social scientists’ work and that really ought to be a two-way street.

Get to the point

I was commenting out sections of text from my Quantified Self thesis as I was looking at parts that would be relevant to a potential CSCW paper (the contents of which might show up in a blog post later, but the gist is that people kind of sort of communicate their collections of activity data and visualizations similarly to a Cabinet of Wonders), and I realized something: When I comment out a sentence in my QS thesis, I comment out dozens of words.

This might not seem that significant, but the average sentence has about 9-14 words. In essence, my sentences are these long, rambling soliloquies with no end in sight, communicating some point that - by the end - the reader might have fully forgotten. This troubles me.

It’s not just the design of a long sentence that bothers me (in fact, what?). Apparently, reading comprehension correlates negatively with sentence length, suggesting that these long sentences are making it difficult for people to understand my writing.

But it’s not just my writing. I mean, yes, I’m certainly (egregiously) guilty of this practice, but it’s not just me; social scientists have engaged in this kind of writing for decades, maybe centuries, before I rolled in. I might even argue that my elaborate language comes from training rather than some innate characteristic. Not that I’m laying blame anywhere.

Case in point, let’s look at a random book I happen to be carrying around: The Savage Mind, by Claude Levi-Strauss. Here’s a sentence from the first chapter - indeed, the first page of the first chapter - which typifies the writing style of the field:

But to begin with, while these cases are cited as evidence of the supposed ineptitude of ‘primitive people’ for abstract thought, other cases are at the same time ignored which make it plain that the richness of abstract words is not a monopoly of civilized languages.

46 words. If you immediately comprehended that, then, I dunno, congrats I guess. Me? I had to read it a second time before I understood what was being said. If you’re still reading along, hoping I’ll double back to explain what the better part of 50 words was trying to say, you’re in luck. In essence, it said:

Yes, there are cultures which lack abstract language and make it convenient to call them “uncivilized”, but there are other “uncivilized” cultures that have abstract language.

26 words. I bet I could get it to 20 if I really tried, but then it’d be kind of choppy and weird-sounding.

As far as I can tell, the problem isn’t that social scientists are trying to fill more pages (maybe a few are, but for some reason I doubt that Claude Levi-Strauss was ever worried about filling pages). The problem, instead, seems to be this overwhelming introspection and reflexivity that forces social scientists to write meandering narratives avoiding all sorts of language to make the same point. Trying to get a single idea into the same sentence becomes this obstacle course of rhetorical devices and obscure diction.

The result of this meandering is a beautifully ornate sentence in every connotative way; it’s gorgeous to look at, and deeply rich with meaning, but damn if any novice in the field will ever understand it fully on the first read.

Does academic research need to be this opaque? Does it need to be such a trial by fire for all but the most careful and seasoned readers? I don’t think it does, but the prevalence makes me worry a bit. Still, I feel obligated to soldier on.

If you’re writing a very gnarled paragraph, I urge you to take a walk around your table, sit back down, and read the paragraph with fresh eyes. If any sentences take a rambling path toward the end, where you can barely remember the state the world was in when you started, please consider revising that sentence. Contemplate the value in breaking it up.

I’m not saying you need to do this with every sentence you write, or even that you should try to do this in all of your writing. It’s just that, in academic writing, your goal is to get your point across as clearly as possible. When you write long, rambling sentences, you risk one of two things happening. First, you run the risk that someone will navigate your murky circumlocution poorly and come out the other side with the wrong takeaway. The second danger, however, is no less troubling: you run the serious risk that people won’t bother to see the insight you’ve buried in your murky writing.

Then what was all of it for?


This rant got away from me (in part because occasionally the threat of hypocrisy got the best of me and I went to break down some run-on sentences). I’ll write about cool stuff computer scientists have been doing that could inform social sciences research later.

If you have something to say about this post, email me or tweet at me.