Job Market Writing Samples

It’s the season for students entering the job market to put the final touches on their writing sample. Hopefully by now, most of the arguments are in place. In fact, for people who write a dissertation that is a single, sustained piece of argument, rather than an n-paper option as I did, there’s a surfeit of arguments, and the problem is to figure out what to put into the writing sample. So I thought I’d put together a few items that might be helpful to think about explicitly as you’re shaping your piece.

Writing Sample vs. Publication

The writing sample should, of course, be really polished. That means, for example, that you have a main argumentative thread, that everything in the sample is there for a very good reason, that your language is exactly as you’d like it to be, no argument is there just because it’s cool, etc. These are the virtues of writing for publication, as well, and so many people think of a writing sample as a paper that is ready to sent off for publication—or at any rate, as close to ready to send off to publication as you can make it.

However, a writing sample and a published piece do not have the same aim, and it’s important to realize that. A good publication stands on its own. If you’re lucky and people read it and find it interesting, people will discuss it. A writing sample, by contrast, is part of the whole dossier, and it also has to serve as an entry point into that dossier.

Then there’s the issue of accessibility. Obviously, both a published paper and a writing sample need to make their main points accessible to reasonably patient and well-intentioned readers who don’t already know the contents of the author’s mind. But a published paper can place higher demands on the background knowledge of the reader. I work in the philosophy of language, and if I write a paper that has a formal semantic component, I don’t explain everything I need in the paper. For example, I won’t explain what model-theoretic semantics are, or even how simple formulas of the lambda-calculus work. These are just part of the lingua franca among researchers in the field. In every field, there are views and tools that are part of the lingua franca in the same way, and these aren’t necessarily formal tools. They can be forms of arguments or what have you. A writing sample, by contrast, shouldn’t take that kind of work for granted.

The Basic Problem

How could you possibly achieve all of these goals in a single paper? Let alone while keeping that paper to a reasonable length?

The obvious answer is to make parts of your paper do several jobs at once. Let’s look at a particular example of how that might go: how to make the writing sample an entry into your project. The most obvious way is clearly to say something in your conclusion about some questions that you’ve had to leave open or addressed only in passing that receive a fuller treatment in your dissertation. The problem with that strategy is that many people won’t read all the way to your conclusion. It also ends the paper on a down-note—rather than the final impression of your paper being a strongly worded and well-argued conclusion, you talk about things you didn’t do. That’s suboptimal, perhaps even deeply so.

You could mention such open questions in the course of the paper, mention that you can’t really discuss them, but that you do it in your other work. That strategy, too, has some problems. You can put that kind of reference into the text proper. If you do that more than once, you run the risk of making your prose choppy by distracting the reader from the main argumentative thread. And having a main argumentative thread to keep the reader on board is very important. It really shouldn’t be compromised. Alternatively, you might put it in a footnote. Those often don’t get read. But of the options we’ve considered so far, they’re by far the best.

Here’s what looks to me like the most elegant solution: your other work is foreshadowed in the way you write your introduction. The introduction needs to introduce your topic, the problem(s) you’ll address, and say something about the basic outline of your solution or intervention. One way of writing such an introduction is to cite a bunch of people who’ve already addressed the issue, perhaps spending a paragraph each on the main positions that you’ll situate yourself in relation to. That way of doing it has the advantage of turning your introduction into the proof that you know the literature.

However, using the introduction as a literature review has the drawback of making it seem as if you’re writing about a problem because other people have written about them. This is probably not the best light to cast yourself in. So it’s better to explain why the problem is a problem on grounds that are independent of the literature.

That in turn naturally recommends a funnel-structure to your introduction. You might begin with a relatively general problem, and then introduce the particular topic of the paper in relation to that general topic. For example, imagine someone who’s writing her dissertation on the relationship between context and semantic interpretation. Perhaps she’s doing some case-studies, so that each of her papers is about a kind of locution; perhaps she’s supplementing case-studies with a more abstract, foundational chapter/paper. Her introduction might look like this:

blah … blah … blah … context in communication … blah … blah. It’s exhibited by many locutions, including [insert topic of paper 1], [insert topic of paper 2], and [insert topic of paper 3]. In this paper, I focus on [insert topic of paper 2] … blah … blah.

Or, depending on the thesis, it might look like this:

blah … blah … blah … context in communication … blah … blah. The problem has both foundational and more directly empirical aspects. On the one hand, we want to know exactly what form context-dependence has, which we can investigate by considering particular locutions, such as [insert topic of case-study 1], [insert topic of case-study 2]. On the other hand, we also want to know about [insert topic of foundational paper]…. blah … blah

So what happens to that demonstration that you know the literature? Well, chances are that if you’ve thought about some topic long enough to write a dissertation about it, you’re not completely happy with how some/many/all of the philosophers in the debate formulate the issues. That means that, in order for your paper not to look like it just misses the point, you need to say something about why your way of setting up the issue is the right way. That’s where you do your literature survey. One nice thing is that the literature survey is not just a list, but another place where you can illustrate some argumentative skills.

More to follow.