Too Busy For Words - the PaulWay Blog

Working at the ANU has its perks. You get a pipe to the internet so big that you can download an ISO image of a Linux distribution faster than you can burn it to a CD. You get to work in a place where the gardens are large and well cared for, and often you have a great view over them as you work. Especially in the research labs, you get great flexibility with when you work during the day. And (for me) it's close enough to ride to, so I can get a little more exercise.

There's a downside, though - you work in an academic environment. Now, scientists hold themselves up as a model of rationality and reasonability. It's all about peer review, open publication, and building on the work on others. No-one's idea has any more weight than anyone else's unless it can be proven. Scientists say they abhor professions like marketing, where the only important things is spin and image and branding. Scientists are always happy to be proved wrong, often by newcomers to the field, because if the new theory holds up then the great cause of Science is advanced.

This, I'm here to tell you, is absolute crap. Scientists are the cliqueiest^[1] people in the campus, and they've built a heirarchy of students, lecturers, professors, deans and chancellors to prove it. The indoctrination starts early, with undergrads being given the cold shoulder by many senior academics who only want to work with more prestigious PhD students. You don't write a paper if your professor doesn't already think it's a good idea. Who gets to sit on what boards, and what weight there opinions are given, is a closely guarded privelege - most of these are invitation only. No-one asks too many questions when the Dean of Science, who gets to approve research grants to do with bioinformatics, sits on the board of a Biotech company. It's "perks for the heads" all the way.

If you need any further proof, ask yourself why some journals are given more credence than others. Yes, they have higher standards of publication, but those are only applied to newcomers - reviewers know better than to question too closely the work of the established names in the field. In order to publish, you have to sign an agreement that you're not getting anyone else to publish the work as well - so if you get rejected and decide later to get your paper published in a 'lesser' journal, you've just wasted six months of review process from the first journal. This makes no sense if the aim of Science is to publish the truth. Who reviews what for which journal is another closely-held, fiercely competitive status credit: a reviewer for "Nature" holds great power and prestige and knows it. The idea of a journal that just takes any paper regardless of who else is publishing it is actively laughed at.

Reviewers themselves often work anonymously - which, if you think about it, makes no sense in a field which says it doesn't have any ego to puncture. Often, your paper is being reviewed by people who not only have an opposite theory to support but are getting a lot of money in funding and lecture talks and corporate sponsorship from it. You can bet that they don't hold back the criticism. The amount of planning we've gone through to present some of our ideas, which are pretty radical breaks from the established conventional areas of research in molecular biology, is phenomenal - simply because a reviewer that doesn't think we're on the "right track" (i.e. their own track) and will just dismiss it out of hand as being "technically flawed". Good luck publishing when that happens.

(Maybe I should state here that this is stuff that I've picked up from working with the people here. I haven't experienced any of it directly, thank Buddha. But you can tell the signs are there.)

The more I look at the Open Source software development model, the more I wonder how much we've gone down that path. What cults of personality are their already? Whose opinions do we listen to merely because they're that person rather than because their technical analysis is correct? This is especially damaging in areas where the person has no technical experience - me giving opinions on correct kernel module design, for instance. Why, to pick a favourite bugbear of mine, do so many people say "Oh, but you've got to have two GUI systems to get competition" but they don't support the same principle for having two Linux kernels, or two networking stacks, or two slightly-incompatible versions of the OpenGL standard? How can we make sure we value ideas and principles and tests rather than opinions and preferences and whims?

[1]: hey, I just discovered another four-consecutive-vowel word!