Organic reviews

Long ago, on the nascent form of this blog, I wrote a little diatribe on the shortcomings of the peer-reviewed journal system. My basic gripe is that the system slows the dissemination of information for marginal benefit. For example, people claim peer review makes science more reliable, but it has been found that, "at least 50% of published studies from academic laboratories cannot be repeated in an industrial setting." And that's for the most reproducible natural science, chemistry.

Review papers are frustrating for different reasons. Certainly, most review papers are not delayed by peer reviewers. Instead, reviews are hobbled by their very format: review articles come out every few months, and cover a field of research as a whole. Their infrequency means they become outdated as soon as another important paper comes out. And their scope means they are forced to rehash the same basic information (there's only so many ways to say AMPA receptors are important for LTP). I often find reading reviews tedious, trying to segregate what's new from what I already know.

I wish we had a form of organic review. A format wherein you could write a complete overview of a field, and then update it piecemeal as new findings emerge; wherein you didn't have to rewrite the entire review; wherein you could stay current to within a month, or even a week.  In effect, I wish we had review wikis.

There are a few non-traditional review sources out there, but they are all lacking in different ways. Most obviously there's Wikipedia. Like a lazy undergrad, I often turn to Wikipedia first when I read an unfamiliar term (what's a diencephalon, again?). The articles on popular things, like AMPA receptors, are fairly thorough, while the articles on more obscure things like T1R3 contain enough information for me to look elsewhere. Yet Wikipedia is true to its nature as an encyclopedia, and is rarely up-to-date, or technical enough to be useful to scientists.

Some people have tried to improve Wikipedia. A couple years ago the Society for Neuroscience tried to ameliorate the situation by launching the "Neuroscience Wikipedia initiative." Unfortunately, it appears to have netted less than 100 edits.

I myself have dabbled in editing Wikipedia. When I train students, I try to get them to read papers, and synthesize them into a whole. Instead of getting them to write a staid essay, I have them edit the Wikipedia page on whatever they're studying. For example, I worked with one student to develop a PI3K FRET sensor, so he added to the section on PI3K in long term memory (his username is Wc18, mine is Amphipathic).

Besides Wikipedia, there are a few other web resources that almost act like organic review. There is wikigenes, which has useful lists of citations, but lacks any bird's-eye perspective on research. Some labs have wikis, but they are often quite focused (the Hayashi lab's is quite good). And some adventurous souls have set up regular ol' web pages dedicated to their field of interest, but static webpages by their nature cannot organically evolve. In general, I'd say these alternative forms of review fail because they are too superficial, lack Weltanschauung, or are too focused.

The review wiki is such an obvious idea that it must come to fruition. The biggest obstacles are probably authorship (people want credit), reliability (no one trusts a random web page), and quality control. The easiest solution to these problems would be for a known organization to sponsor a wiki. For example, I bet a neuroscience department could gain reputation by starting an awesome, up-to-date wiki. Over time, as the wiki grew, it could serve as an alternative sort of textbook (in fact, if you search for science wikis, you'll see many hits from teachers looking for textbook alternatives). They could brand the wiki with the department or university. Then when undergrads inevitably discover the wiki, they'll assume the authors are important. The first mover advantage here would be huge.

(Why don't I take action and start a wiki? I don't have the stature to get people to use it, nor get buy-in from others to expand it.)

Until then, I will continue to read traditional reviews, and supplement them as best I can. The precious few neuroscience bloggers out there do a decent job reviewing recent papers, and in doing so comment on the state of the field. I hope some of my blog posts can do the same.