Tuesday, February 5, 2013

Ditch the Job Talk...

. Tuesday, February 5, 2013

Dan Nexon sparked a conversation about job talks. The whole exchange is curiously ahistorical. I searched the web for something that might document the historical development of this tradition. I found nothing. So, let me advance an hypothesis. (If someone knows a source that discusses this historical development, please give me the cite).

People invented job talks because the cost of reproducing non-published written work was too high to allow distribution of large volumes of printed material in support of job applications. It seems that in the pre-photocopier era, the cost of reproducing written work, such as an academic paper, was about $.15 to $.25 per page. This was, according to wiki, at a time when the minimum wage was $1.65. This per page cost did not include the labor of typing the manuscript onto a stencil which could then be used to generate the mimeograph. In this environment, it was cost prohibitive to circulate hard copies of non-published research for three candidates to an entire faculty(somewhere in the range of $540 to $900 in 1968 dollars). How then, by golly, would a department faculty ever come to learn about the current research being conducted by the persons they are considering for a position? And thus was invented the job talk. A low-cost method of disseminating knowledge across a large community of interested listeners.

This hypothesized explanation for job talks hews close to what I take to be a pretty reasonable explanation for the origins of lecturing in the University. In the pre-Guttenberg era, books were costly to reproduce. Brad DeLong suggests that a single student's textbook costs for an eight course per year four year college degree based on today's "all students buy books" model would have been around $1.6 million. There were not many students who could afford that. Hence, teachers stood at the podium and read to their students from a single copy of an illuminated manuscript.

Back to job talks. I think the marginal cost of reproducing and distributing unpublished work today is about zero (Yes, neither the transmission nor the storage of data is free. But, still). Arguably, therefore, the function for which job talks were created no longer needs to be filled by job talks. And yet, we show up at the appointed hour to listen to nervous candidates present less sophisticated versions of the very paper they uploaded to the personnel website when they applied for the job. The very same paper that the committee sent, or could have sent, to the rest of the faculty as an email attachment when they notified us about the forthcoming job talk. In short, the tradition persists long after the technology that made it necessary has been supplanted by technology that renders it obsolete (arguably, the identical logic applies to conferences, but I'll save that for another post).

And thus robbed of its original quite useful purpose, we invent new purposes that the job talk supposedly serves. It's an initiation. It's how I can determine what kind of teacher you will be. Faculty are lazy, so it's the only way they will ever become familiar with candidates' work. We couldn't run a pro-seminar (why would we want to?) A much simpler explanation seems more compelling: We have job talks because that's how we do things and nobody stops to consider whether they actually provide any useful information that isn't better and more cheaply attained elsewhere. (Indeed, as Uncle Wuffle notes, there are two main purposes for giving a job talk. (1) to get a job and (2) to practice the job talk so you can get a job). We give job talks because we ask people to give job talks because that's what we do.

In fact, I might even suggest that the job talk creates the suboptimal outcomes that Dan's initial post deplored. Our belief that we can learn enough about a candidate's work by attending the job talk removes the obligation to take the time to read the research. As a result, we make decisions based on dumbed-down and overly general presentations of the candidate's research. If we dumped the talk, faculty might feel a greater obligation to read the work in order to have an informed opinion. That might generate more informed discussions and better hiring decisions.

I am sure our practice eventually will catch up with technology. I mean, how many of us teach big lecture classes anymore?


Dan Nexon said...

"I am sure our practice eventually will catch up with technology. I mean, how many of us teach big lecture classes anymore?"

Thanks a lot. Now I'm envisioning a hellish future of massive open online job talks (MOOJTs).

Joey said...

This post is the best I've read on the subject. The two points that should be at the forefront of any discussion of this sort of institutional stickiness are "robbed of its original quite useful purpose, we invent new purposes" and "nobody stops to consider whether they actually provide any useful information". These processes are rife in academia, despite the fact that some of us are supposed to be experts at studying social behavior.

Anonymous said...

I assume you have no objection to a flyout to meetthe an interview, right? I would never hire someone based solely on the paper. Also, if you believe there is n value toa job talks, then why does theis same logic not apply to to conferences and seminars? Why don't we all just sit in our caves and email papers around? Because there is value to live exchange, that's why.

Jimbo said...

A committee does an initial screening. The decision then goes to the whole department. What if someone thinks there's a flaw in the candidate's paper that big enough to be an argument against hiring him or her?

The JMP presentation gives the candidate an opportunity to address such concerns.

It also gives the department a chance to estimate how much depth of thought is behind the JMP.

DeLong can be right about the origins of the lecture without being right about the reasons for the persistence of in-person discussion.

Ditch the Job Talk...




Add to Technorati Favorites