Wednesday, December 18, 2013

Some big news

Here’s some big news: scientists are human! They cherish their own views and find it bracing if the views of their intellectual opponents fail to gain traction. Moreover, given the evident hierarchical organization of scientific investigation (e.g. people do not enter the game on an equal footing given that prestige is not evenly allocated across institutions or practitioners, journals are managed by people with skin in the game, friends like to help friends, careers depend on the relative success of one’s ideas, big and small etc.) it should come as no surprise that some try (and succeed) to prevent good ideas that gore sufficiently prominent oxen from seeing the light of day.  How? By denying funding by trashing grants asking for such, by making it hard to get into print, by limiting conference invitations to the non-fashionable etc. The official line is that though scientists as individuals may lack the requisite impartiality to give a wide range of views a fair hearing, Science (btw, always capitalized) as an institution allows good ideas to bubble up to the surface given enough time. In other words, Science encourages what scientists don’t.

Let’s assume that this optimistic tale is correct (but who knows really). It is still intriguing to see just how strong the forces of intellectual self-interest can be and just how long “given enough time” can be. It can be quite a long time. A Nobel Prize eventually came to Barry Marshall and Robin Warren for their work on ulcers being bacterial effects, but only after a long fight with the medical establishment that was sure that it was due to stress (here). Recently, a little window into the ways that thought leaders try to protect their favorites seeped in the NYTs science section (see here). Nathan Mhyrvold recently disputed some findings in dinosaurology in PloS One. The paper he criticizes was published in Nature in 2001, the lead author being Gregory Erickson.  Mhyrvold claimed to find a flaw in how growth rates were computed. The details, though interesting, are not important here. What is intriguing is how Erickson responded. He tried to stop publication of Mhyrvold’s paper claiming that the paper, if published, would “hurt our field by producing inherently flawed growth curves, misrepresenting the work of others, and stands to drive a wedge between labs that are currently cordial with one another.” This did not prevent publication in PloS One, an open source journal, but it also surely helped that Mhyrvold is himself a really big deal (here). One can imagine how things might have gone had this issue been raised by a somewhat less prominent critic, say a graduate student submitting her first publication and looking for a “prestigious” venue, like Nature, in which to publish it.

I don’t wish to suggest that there is anything nefarious in all of this. It’s another “crooked timber of humanity” sort of thing. However, I suspect that this sort of political maneuvering is more efficacious in smaller fields like linguistics. This is especially true given the paucity of publication outlets. Publication rates in our major journals are very low and so it is easier to prevent unfortunate ideas from seeing the light of day.

On a more personal note, I have found that reviewers often mistake agreement with the position expostulated to be a criteria for favorable review. I don’t mean that I have personally felt this to be true about my own work (though of course, whenever any paper of mine is rejected I am sure that it is because the reviewers have it in for me!) but because I have been asked on more than one occasion concerning a review I have submitted how I can recommend publication despite my professed views that I believed the position being explored to be wrong. It seems that (some) editors find it odd that one might think a paper incorrect and yet worthy of publication.  I don’t have to agree with a paper to find it interesting and provocative. Indeed its interest may lie precisely in advancing a view that I disagree with.

This problem is further exacerbated by the peer review process, I believe.  Peer review pursues the lowest common denominator. It’s hard enough to convince one person that  an off-beat idea is worth investigating. It’s harder still to convince five.[1] I once asked a friend of mine if he believed that Einstein’s 1905 papers could have been published today? They were very quirky given the standards of the time; quite informal, from a patent clerk, sent to the most prestigious physics journal in the field at the time. He thought not likely.  Luckily, the journal editor at the time was Max Planck and editors were more like impresarios than scholarly bureaucrats. Today, one threatens eviction if one is too interesting. I personally believe that this is what happened to Jacques Mehler at Cognition. His journal was too different (and IMO, by far the most interesting cog sci journal ever), a problem remedied by appointing Gerry Altmann to replace him.[2]

One of the hopes for the web is that it will allow debate over ideas to flourish without having to be squeezed through the narrow review portals of the major journals, something that Paul Krugman notes (here) has begun to happen in economics. Maybe something similar will occur in linguistics and debates that are currently largely confined to the review process (where submitters and reviewers hash it out endlessly), will be held in public and given a good airing (but see here: Mark Liberman notes that the most linguistics institutions lag far behind other scientific fields in disseminating new ideas; a holdover perhaps from our illustrious humanistic philological tradition?  The timeless humanities don’t pursue “cutting edge research.”).

Let me end: I am curious to know how singular my impressions are. The above may just be the maunderings of an embittered isolate.  Do you find that institutional Linguistics does a good job at finessing the all too natural self-serving aspects of linguists?  Do you find that the journals and granting agencies try to promote worthwhile debate and investigation?  If not, can you think of how things might be improved? I’d love to know.

[1] The inherent empiricistic data of linguistics, its basic lack of interest in theory as opposed to data (no doubt another holdover from our philological past) is another serious impediment against encouraging new ways of looking at things. But this is a topic for another day.
[2] For how dramatic a change this entailed, take a look at Mehler’s farewell address in Cognition (here) coupled with Altamnn’s initial editorial foray (here). The indicated concerns could not be more different.

No comments:

Post a Comment