Tuesday, March 20, 2007

Should I publish research ideas on my blog?

(Note: The ideas in this article aren't strictly my own. I'm borrowing most of this from discussions I've had with various people including: Dave Bacon, Jennifer Dodd, Mike Nielsen, Aram Harrow, Caterina Mora, Joe Renes, Andrew Hines, Gavin Brennen, Frank Verstraete, and Scott Aaronson.)

I've been thinking a lot lately about the ways in which us quantum information geeks discuss and disperse ideas. Traditionally, academic discourse has been via published journals and conferences (that often have conference proceedings where people have the opportunity to formalize whatever it was that they talked about in their talk or poster). There are a lot of downsides to both of these processes. Leaving aside the very large issues of the sometimes prohibitive cost of journals and conferences and the general worth of the refereeing process, there is the very physical problem of time.

Writing, refereeing, and ultimately publishing papers is a time intensive process - as is the process of reading, understanding and categorizing the relevance of research results. The birth of the arXiv saw scientists realize that many of the tasks of discussion and dispersion of new research results could be done in parallel. People began to put unpublished preprints on the arXiv which sped up the process of scientific discussion. Scientists who are confident in their abilities can "sort the wheat from the chaff" and can proceed to use non-refereed results to further their own research. In essence, for many papers that hit the arXiv there is little necessity to wait for the published version to appear in a journal before using the results.

Obviously, this whole arXiv experiment has its downsides. Once someone has permission to post papers on the arXiv the only moderation that exists is to prevent potentially offensive material from being published. There is little to no control over the validity of the content of any papers. If you want to use a result that you see on the arXiv that hasn't been refereed you are doing so at your own risk. Fortunately this has not proven to be a huge problem for the majority of arXiv users however as science is driven by reputation.

If you've ever spent much time involved in the "climate change wars" on the interwebs you will have seen endlessly the argument that science is not a democracy as opinions are always weighed against authority, experience and reputation. This reality about the nature of science provides a measure of control over the content of the arXiv - if you throw a lot of junk up then you will eventually end up with a reputation as a junk scientist. Not unlike the tale of "the boy who cried wolf" no-one will notice if you publish quality material as they will simply cease to read your work and your scientific credibility will gradually dissolve.

So far, the arXiv has worked pretty well as a way of speeding up scientific discourse and making the job of searching for relevant results simpler. The big question is whether the arXiv will continue to be useful with the passage of time and in the face of a growing number of contributed papers?

My experience is with the quant-ph subsection of the arXiv. The quant-ph arXiv has in recent years seen a gradual increase in the average number of papers posted daily. This has been due to the explosion of interest in the fields of quantum information science and quantum and atom optics. As the number of daily articles increases it becomes harder to notice important new results especially from relatively unknown authors. The problem of increasing field popularity has always existed in science. Traditionally it is dealt with in a number of ways. Physics journals have normally managed this problem by dividing popular fields up into sub-specialties and by providing new journals and sections for emergent fields. However, for a field like quantum information science, which draws it strength from being at the intersection of many fields, this approach could be seriously flawed. In addition to this we have always used conferences as a way of dispersing new ideas. The thesis has always been that the really good ideas don't get missed, they might take a while to surface but eventually they will be noticed by the field.

My question is, can we use communications technology to improve the nature of our scientific discourse? A very relevant, and very useful example of how "web 2.0" ideas can be adapted directly to science is Dave Bacon's recently launched website (http://scirate.com) which uses the idea of "digging" and applies it to the daily quant-ph feed. This website works by letting you "scite" a paper that you like, then it is hoped that with time popular articles will become noticed as they will be "scited" often. In addition to this, Dave has launched through his blog "The Quantum Pontiff" and the "SciRate blog" a monthly article discussing the months most "scited" papers. It is important to note that such services are designed purely to simplify the task of sorting through papers, not to replace it.

SciRate has been developed on the back of a trend for people in the field of quantum information science (I'm picking this as an example, it is more than apparent in other fields as well) in the last couple of years which has seen people increasingly use the internet to discuss current research. While physicists have been using email since it began to discuss work, this process is essentially one that is done behind closed doors. What we have seen emerge in recent years, especially with the advent of blogs and wikis, is the public discussion of current research.

Anyone who has hung out with physicists knows that "shop talk" can almost never be avoided. So it isn't surprising in the slightest that when people like Dave, Scott, Mike, Aram, and to a much lesser extent people like me (and the many others that I've left out) began to blog that work discussions would become unavoidable. More and more we are seeing Scott and Dave post articles that are relevant to very recent papers and we are seeing real-time discussions between various leaders in our field taking place in a very public forum. Such discussions aren't normally available to people who aren't working at the premier institutions in quantum information theory and as such I think that these blogs are becoming an invaluable asset.

I have often mused, as have many others, as to whether we'll see blogs or some other form of technology take over from the old process of writing and refereeing papers. While my mood on this seems to change with the week, I'm currently of the opinion that this won't happen. Papers give people the opportunity to carefully think out and refine their ideas over a relatively long period of time. A good paper can be the compilation of years of work, carefully condensed to produce a particular result. While there's nothing stopping people from doing this in a blog article, it seems that blogs are, for the most part, about the rapid discussion of ideas. In my own head I see a blog as being more like a chalkboard in the middle of a discussion with twenty people than a well-reviewed PRL. Blogs, in the quantum information context, seem to me to be developing into a form of media for commentary on our own field. While blogs since their inception have been a way of commenting on events in the news media. We are now seeing the development of a alternative media culture within the quantum information community. I guess my experience within the political blogosphere has biased me to think that this has to be a good thing for the community, but I guess time will tell.

Finally, I guess I should get around to the topic the title of this post. In my discussions with other physics bloggers one question frequently arises, "should I post new research ideas on a blog?". For the most part, this question is asked out of a fear of losing credit for your idea. As I've had it put to me, "I don't have enough good ideas to give them away for free". Personally, I'd like to see a culture develop where people are credited for their ideas no matter the forum where they are presented. However, being a postdoc who is finding it hard enough to get employment as it is, does it make more sense for me to hoard my knowledge and use what I can of it to publish papers to further my career? Or should I put all my ideas out here for the world to see in the hope that if they are useful to someone they will give me the credit (and that there is some way of showing a prospective employer that I got the credit for that idea)?

I'd really love to hear people's ideas on any or all of the things that I've mentioned in this post. So if you have some thoughts, please post them.