Quick reaction on spnetwork part 4

Triggered by the spnetwork 4 post by Christopher Lee, hosted at John Baez’ Azimuth.

Only a very brief reaction, written from a beach, will come back later to it.

I am intrigued by this:

Think about it: that’s what search engines do all the time—a search engine pulls material out of all the worlds’ walled gardens, and gives it a new life by unifying it based on what it’s about. All selectedpapers.net does is act as a search engine that indexes content by what paper and what topics it’s about, and who wrote it.

There seems to be a huge potential here.

OK, I am thinking about it and I’m having the usual conversation with the regular naysayer, who tells me that in order to switch to the spnetwork, ot to ANY alternative of the actual publishing system, you need to have an incentive for that. What’s wrong with the actual system, besides the double monopoly (monopoly and monopsony, hence a banana republic situation) of greedy publishers hand in hand with managers in academia? Not much (with the condition that you pay the publisher with OTHER PEOPLE MONEY). The researcher writes an article, which is peer-reviewed, everything is verified and working nicely, why change that?

The regular naysayer tells me that the real problem is the huge number of articles. Which one to read and which not? Which one to check to the bones, even if already peer-reviewed? The answer is this: is statistically better to read the articles appeared in good journals.

Any system aiming to improve  the old one should solve this problem of picking articles from the huge pile which is produced every day.

Apparently, the name, “Selected Papers Network”, suggests that Lee’s project tries to solve this. But now here he comes with a really interesting and different idea!

Forget the incentive, let’s think about articles. The truth is that even if there are many articles, too many to read, there are very few readers of an article  chosen at random. As authors, we all want our articles to be read. As readers, we long for fewer, more interconnected articles.

There are too many articles either because the article is written for ISI points, or because there are too many articles writers, or even because the article is not touching the readers who might do something with it, because they read other articles or, rarely, because they are not yet born (sorry, but I can’t stop to mention again the comparison of what is happening now in research publishing with the impressionists revolution, so why not accept that there are already articles which don’t have yet readers, like, say, Van Gogh paintings during his lifetime).

Or, an article, as it is written today, with the manifold stupid conventions which are reppelant now, but have reasons in the past, is a very bad vector of information. There are a lot of articles, each trying to get a bit og brain time, on it’s own, without trying to collaborate with others. In this respect, I believe this is a far consequence of  the cartesian method, which I think is obsolete in some respects, because it is  “designed as a technique for understanding performed by one mind in isolation, severely handicapped by the bad capacity of communication with other isolated minds. It was a very efficient technique, which is now challenged by two effects of its material outcomes:

  • better communication channels provided by the www,
  • mechanical, or should I say digital, applications of the method which largely surpass the capacity of understanding of one human mind, as witnessed for example by the first computer aided mathematical proofs, or for another example by the fact that we can numerically model physical phenomena, without understanding rigorously why the method works.”

As far as I understand the new idea of Christopher Lee, the tagging system proposed by spnetwork could be a part of a solution for the problem of having too may articles not communicating one with another (by grouping them).

Another part of the solution could be using other vehicles than articles for communicating science, I am thinking  about open notebooks. There are not too many open notebooks, but they have the following advantages over articles:

  • more honesty, be it about negative results, apparent dead ends, more clear background data and real motivations for research,
  • more lively, welcoming discussions, than the dry and often hidden peer-review
  • naturally interactive.

Articles are like movies, open notebooks are more like games.

Therefore, to conclude, it seems to me that Christopher Lee’s federated ecosystem could have more chances if it allows open notebooks (besides articles, which are still necessary) to join the party.