The Great Divide : A Rant

May 28, 2009

There’s a good posting on the Open Science Project called Scientific Software Wants To Be Free. I generally agree with their manifesto, we’ve kicked similar ideas around the pubs of Cambridge for a while, without such concrete suggestions.

In the comments, Geoff Hutchison makes an excellent point : –

I think these are also key points:

  • Software releases also rarely garner wide recognition, and without an accompanying paper to cite, the work cannot be paid back in citation currency.
  • Additionally, emphasizing one’s work on software or publishing methods papers carries a risk of being perceived as a programmer first and scientist second, falling on the wrong side of the technician/scientist divide.

I have been told to be sure to publish “real chemistry” more, even though I want to get “citation currency” for open projects. After all, without publications or citations, I can’t get grant support for these activities either.

So is the trick to get grants for “real research” that have “broader impacts” in software development?

I sympathize with Geoff. I have friends and colleagues whose academic careers have been made difficult, jeopardized or even ended because they gained technology expertise in order to make their research more effective, and ended up being perceived as too techie to be “proper scientists”.

The problem, as I see it, is the very existence of a “scientist/technician” divide. I’m beginning to suspect that the problems software engineers have with existing metric systems based on journal articles are more the symptom than the cause of a lack of respect and recognition afforded to technical staff in academia, and that fixing it through software journals and the like will be a temporary palliative cure.

Perhaps I’m just getting more bolshy the more streaks of grey appear, but perhaps academia needs to change to embrace software engineering, rather than software engineering twisting itself out of shape to fit in with the anachronisms of scientific research.

Admittedly, I’m pretty short on “how”. The publication / citation model clearly isn’t working so well for researchers any more – perhaps fundamental changes there will help. Funding of the sort JISC, and the e-Science programme certainly help provide support and some credibility, but you still need a Principal Investigator, and if they’re academic they risk being dragged toward the divide for their pains. Post-grad training could also help to change the culture from the ground up. The JISC-CRIG Dev8D initiative is a band-aid for this very problem; an extra-mural support network to give developers some well-deserved recognition they don’t necessarily receive in their own institutions. Beyond vague ideas and short term fixes, I’ve got nothing.

It’s an important issue though, because so much of modern research depends on software development, and when (not if) the software / technical talent walks out of the door to work in industrial research (where it is respected), how will academia compete?


2 Responses to “The Great Divide : A Rant”

  1. In the Netherlands there is a division between Technical and General Universities. The first focus on application of scientific knowledge, while the latter focuses on theory. But your analysis the academia looses the scientist to convert his theory and hypotheses into a valid and working tests, might work one a small scale, but not at large.

    I think the fact that qualified scientific programmers actually get bullied away ensures that they will never found solid ground anyway, because they get no chance explaining what Joe Average Scientist is doing wrong when dealing with data.

    I know that this is quite blunt, but just look at the scaring amount of unreproducible literature in current cheminformatics, and you know this is true. “Oh joy, yet another SVM model for a specific target.”

  2. Diana Drennan Says:

    I agree – it’s an unfair divide. I think the “trick” is to develop the software, and then use it to do “real” science. Thus, you cite your own work.

    Often, the experience of being a user of your own software leads you to figure out the bits of code that need to be added /refined /reworked /improved /fixed as you go. And being a developer /programmer leads you to think of new questions you can answer with your programs as you go. This is absolutely true for me. Unfortunately, my current job is mostly as a user, and I don’t have time to develop the software I want to. Nor am I encouraged to. My personal opinion of this is that it makes me less of a “real scientist”.

    As a personal example – My PhD research was to be 4 papers. 1 to introduce the software, and 3 to show how it could be used. (unfortunately, they never got published — One of my co-authors (the more published one) said the papers were ready to publish, but my advisor never got around to putting in the bits he wanted to include and so never agreed to publish. (sigh))

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: