18 March 2011

Arsenic life, four months (and a bit) later: Reviewers with shovels

Recently, I wrote about a new review of the arsenic life story. I criticised the authors for mischaracterising online commentary, and alerted a few folks about this on Twitter, including Rosie Redfield. Rosie went over and commented on the journal website, and Rosen and colleagues, the authors of that review, have now replied to her.

They grab their shovels and dig themselves in.

And they’re not particularly careful about where they throw the mud.

Fundamentally, the Internet discussions we encountered contained comments that did not require citation for at least two reasons: i) they reflected general knowledge that in our opinion was not necessary to cite, not even as a personal communication because they were not specifically directed to any of the authors via telephone, email or in person; and more importantly ii) at present, Internet communications/contributions of the sort being discussed here are not components of the peer-reviewed literature and thus are not placed on record as part of the official scientific discourse.

Wow. Just... wow. A blanket statement of “Internet doesn’t count” is astonishing in an age when many journals have recommended ways to format references to web sites. Not online journals, mind you, but web sites. I have seen many papers that reference Internet materials. I’ve published some, and never heard “boo” about it from reviewers.

Heck, I have even seen Wikipedia - yes, Wikipedia! - cited as a reference in a original, peer-reviewed, technical paper published in a long-established journal. I have reservations about that, but it points out that what peers consider to be acceptable references changes all the time.

Worryingly, a logical extrapolation of this view would be that because you don’t have to credit anything on the Internet, you could plagiarise anything you wanted.

We are not sure what a “Twitter hashtag” actually is, but since they presumably are not actual references, we will continue to ignore this type of commentary for purposes of formal citations until it becomes clear that the scientific journals and community have adopted new criteria, standards, and procedures. For example, BioEssays currently does not allow authors formally to cite Internet commentary of the sort referred to by Dr. Redfield.

Then BioEssays needs to get with the program. The editor needs to expand the reference format to allow citations of more kinds of materials.

Journals have generally had an inclusive policy concerning “gray literature”, and Internet commentary isn’t fundamentally different from, say, conference abstracts. But people often cite conference abstracts. It’s generally not a problem.

And while admitting that some Internet commentary about the arsenic life paper was not anonymous, they go on to say that it’s still bad.

Signed comments should be applauded and indeed offer a measure of authenticity, but nevertheless these “chat room” environments are not constrained or screened and at times become ad hominem attacks, which have no place in the scientific literature.

“Oh my goodness! People are talking about non-peer-reviewed data in environments unscreened by selected moderators! However will science survive this chaos!”

This might well be Rosen and company’s response to the Internet, but I was actually thinking that might be something that could be said almost every scientific conference ever.

We are all aware of the massive misinformation that pervades the Internet and that is why so many simply choose to ignore it rather than track it as part of their efforts to stay informed of the scientific literature pertinent to their research. Because of such accuracy issues, at present and in its current structure, organization and process, the term “authoritative blogs” far too often represents an oxymoron.

Yes, that’s why we never see blogs by people who are the presidents of major scientific societies... Oh, wait. We do.

Well, that’s why we never see scientific societies asking people to blog at their conferences... Oh, wait. We do.

Well, that’s why major scientific journals of note would never write articles about online science blogging, Twittering... Oh, wait. They do.

Rosen and company forget that nobody ever gets to declare their own authority. Authoritative is something that other people say about you. And guess what? Other people see bloggers - even those using pseudonyms - as authoritative. Nature quoted Biochem Belle. The New York Times quoted Dr. Isis about arsenic life.

Rosen and colleagues have, perhaps unwittingly, provided a wonderfully unfiltered and honest view of how conservative some researchers are, and how tall and white is their ivory tower.

The past needs to die for the future to be born.


Rosen BP, Ajees AA, McDermott TR. 2011. Life and death with arsenic. BioEssays: in press. DOI: 10.1002/bies.201100012 (DOI link may not be working yet; try http://onlinelibrary.wiley.com/doi/10.1002/bies.201100012/abstract if it isn’t.)

Wolfe-Simon F, Blum J, Kulp T, Gordon G, Hoeft S, Pett-Ridge J, Stolz J, Webb S, Weber P, Davies P, Anbar A, & Oremland R. 2010. A bacterium that can grow by using arsenic instead of phosphorus. Science. DOI: 10.1126/science.1197258

Photo by Jeff_Werner on Flickr; used under a Creative Commons license.


Anonymous said...

"Journals have generally had an inclusive policy concerning “gray literature”, and Internet commentary isn’t fundamentally different from, say, conference abstracts. But people often cite conference abstracts. It’s generally not a problem."

In my experience, none of the journals I've published in explicitly do not allow citations of conference abstracts. And it would be foolish to do so - I've seen results presented in several meetings that never make it to the published literature.

Andrew Moore said...

Diversity is an important part of cultural progress in scientific communication… (Part 1)

As the editor of BioEssays, I must first clarify something and correct a misperception: BioEssays most certainly does allow websites, including blogs or blog posts (no special rules there), to be specifically quoted, and the URL given, in the body text, information boxes, tables or figure captions of articles - i.e. “informally” cited - in fact, anywhere but in the reference list of a scientific article. I am also in favour of relevant information from non-peer reviewed websites being included in order to contextualise a topic - as long as the material has a degree of credibility and value, of course. For those necessary credibility/value ratings - and I am not referring to any particular article of topic here - I rely on authors themselves, and comments from peer reviewers, as those scientists are bound to be more expert in their field than I am in their field. Such “informal” citations are, therefore, at the discretion of the author(s) of the article concerned, possibly informed by peer review comments; and yes, peer review is not perfect, so it is not expected unfailingly to catch the absence of informal citations.

Such citations are “informal”, and not mixed with peer reviewed citations in the reference list, because, as most scientists will surely agree, there are some significant differences between blog posts and the peer reviewed scientific literature; there are important questions too. Here are a few observations and questions - which I don’t claim to be the complete list:

1. Blog posts are not peer “reviewed” by scientists who have been specifically appointed to comment on an article on the basis of their authority (via publication record) in a given field - thought, by their very nature, blog posts may - or may not - be commented on by suitably qualified peers. There is, hence, some degree of peer appraisal in blogs, but not in such a structured, consistent or procedurally reliable way as in formal peer review.

2. Scientific blog posts are not necessarily subject to editorial moderation that seeks to ensure that only material dealing with the discussion of the topic/paper at hand in an “un-personal” tone, is made public; though I appreciate that people who stake their reputation on Internet communication will largely contribute to blogs that they respect for their quality of content.

3. How, given the multitude of un-indexed material on the Internet, can scientists find all, or even much of, blog material that might be relevant for them to mention in an article, or sort through it all in terms of priority? It is hard to see how that is possible if blog entries do not have the same kind of metadata as scientific articles and are not indexed similarly. A reference list is rather different, as it draws on the indexed literature that can be found via keyword searches of literature indexes and a variety of other means to give the best possible chance that all relevant material can be cited.

4. Where, and how reliably/safely, are blog posts archived for retrieval years later when a reader wants to access that “informal citation”? Scientific literature is, on the whole, very reliably archived, and relevant updates linked in later.

5. Are there ethics, guidelines and procedures for the correction or retraction of an influential blog post that is later found to be misguided? In the case of a faulty scientific paper, a corrigendum, or even retraction, can be published, and can then be found together with the original reference, hence correcting the scientific record in a very accessible way.

Andrew Moore

Editor-in-Chief, BioEssays

Disclaimer: The opinions expressed in this post are my own, and do not necessarily represent those of my employer. This account is not sponsored or endorsed by my employer.

Andrew Moore said...

Diversity is an important part of cultural progress in scientific communication… (Part 2)

There is another thing that, at present at least, makes it hard to reconcile blog posts with the peer reviewed literature: speed. A blog post generally goes live within a day or two; scientific articles can take anywhere from a day to weeks from point of decision to go live online, the intervening time being necessary for author corrections, proof reading, typesetting, preparation of electronic publishing data etc. (though many primary literature articles do, these days, go live almost immediately after acceptance, without correction and typesetting, as “accepted” versions). Journals on the whole, and particularly review journals, do not YET, have such fast production times to keep up with rapid electronic communications on the Internet.

Besides that, it would be impractical and excessively time-consuming to search for all Internet commentary between the start of peer review and point of publication in order to update the article. It would not be feasible for authors to check on all possibly relevant and important blog material between submission and publication of an article, and send according revisions to the editorial department to update the record, so to speak. Important updates will, regularly, not be possible to include. There is no solution to the problem of speed, because the “blogosphere” will always be much faster than the peer reviewed literature. That might mean that it is better not to mention individual, named, blog posts at all, but rather just give a link to a blog’s main page with a note to the effect that interested readers can get an update on the developing discussions in the community there. At BioEssays we recognise some of the limitations of the conventional literature, and that is one reason why we introduced an article commenting tool.

Andrew Moore

Editor-in-Chief, BioEssays

Disclaimer: The opinions expressed in this post are my own, and do not necessarily represent those of my employer. This account is not sponsored or endorsed by my employer.

Andrew Moore said...

Diversity is an important part of cultural progress in scientific communication… (Part 3)

In a general sense - and I wish to make it clear here that I am not particularly referring to the article by Barry Rosen et al. recently published in BioEssays – I think that a diversity of communication cultures should be encouraged: culture advances via diversity and the new ideas that it brings. That diversity applies not just at the level of the individual - i.e. an individual might communicate via very many different media - but also at the level of differences BETWEEN individuals. Hence the opinions of a large diversity of people actively working in science are important to incorporate into our design for the future of scientific communication. Sharp criticisms and emphatically worded negative opinions about a certain medium are important to listen to, because the scientific community will, I hope, remain diverse, and scientific communication is there to serve that community. If that diversity in the community results in one or other scientist criticising the blog medium, we could choose to use that to recognise areas for improvement in that medium, e.g. greater attention to developing quality stamping of blog posts or whole blogs, and greater efforts to make relevant Internet-based communications findable, e.g. by keywords, other metadata and formal indexing.

I would like to envisage developments in the scientific and publishing spheres that take account of high quality rapid communications happening on the Internet. We clearly all have a challenge there. At present, we are all - editors, bloggers and scientists - some way off addressing it, in my opinion. I do not personally think that any particular medium should be “dismissed” - particularly if it is relatively new (on the time scale of publishing history), and open to interesting developments; at the same time, I respect others’ opinions and their freedom to voice them.

There is clearly a transition in progress from the strict, conventional, literature communication habits to a more flexible and eclectic one. It is likely that the formal scientific literature of the future will include much more in the way of citations to moving Internet commentary. If blogs had indexed metadata and other identifiers that were easy to track, publishers could offer authors the option of inserting a continually updating Internet citation within an article, linking to new blogs as they arise.

Technology will help us solve some of the above problems, but not without concerted efforts and thoughts from all involved, proponents and opponents. And I think it would be good for us all to write a bit more on what could help us in that respect.

And, just for the record, I don’t mind if this three-part blog post is not cited (formally, informally or by name) by anyone else ;-)

Andrew Moore

Editor-in-Chief, BioEssays

Disclaimer: The opinions expressed in this post are my own, and do not necessarily represent those of my employer. This account is not sponsored or endorsed by my employer.

Brian Krueger said...

"honest view of how conservative MOST researchers are"

Fixed that for you.