Today was the fourth day of SFN 2006, and I find myself seriously frustrated with the setup of the meeting this year – there seems to have been little to no thought put into where to schedule the various concurrent sessions. Usually there are multiple slide sessions each morning or afternoon that are of interest. The natural thing to do is to skip back and forth between sessions to catch the most interesting talks, but here in Atlanta this is all but impossible given the distance separating even closely related sessions. The Atlanta convention center consists of three connected buildings, each of which has multiple levels connected with escalators. It takes a good 10 minutes to walk from Building B to Building C, and yet sessions that are fairly similar in focus are routinely allocated to different buildings. On Sunday the session on objects and faces (mostly fMRI of the ventral pathway) was in B and one on extrastriate visual cortex was in C, as was a session on attention. No doubt most of the people interested in one of these sessions would have been interested in the others, and yet it was all but impossible to go back and forth. This has happened to me every day thusfar. Many of the scientists here have expressed a similar frustration. The problem would be easy to remedy – someone just needs to put related sessions in nearby rooms. The room assignments, as far as I can tell, are totally random, which has detracted from the meeting.

This afternoon featured a well attended talk on the current funding woes, given by the heads of the various agencies of the NIH (the NIH director was supposed to give it, but his flight was delayed). The talk focused on explaining why we are currently experiencing a funding crunch in the US, and attempted to highlight what the NIH folks say are widespread misconceptions about NIH funding. The increased competition for funds has apparently been mostly driven by an increase in grant applications (rather than a funneling of resources to more clinically oriented research), fuled in part by a boom investments in the biomedical sciences by universities. New facilities have brought with them new jobs, and there are now many more people applying for grants. NIH funding doubled between 1998 and 2002, but has leveled off since then, so supply has not kept pace with demand. Tom Insel stressed that they are still committed to basic research; apparently the proportion of directly translational research they fund has not increased. He did not address the issue that most of the scientists I know are interested in – whether basic research grants need to explain how the research might ultimately benefit human health in some way. There was a Q&A session at the end that I had to skip out on, so perhaps someone queried him about this. They did offer some encouragement, citing previous periods when funding dried up and funding rates were as low as 10%, and explaining how grants that are ending now will free up uncommitted funds for new grants. They also concluded by stating the need to better explain the importance of the NIH to the public (apparently only 10% even know what the NIH is).

The most striking news for me at the®evolution in Scientific Publishing session yesterday was that fewer than 100 people showed up, even though prominent posters advertising the discussion greeted all attendees on their way into the meeting. I guess neuroscience journal editors are going to have to figure out how to navigate the brave new world of scientific publishing without much help from readers.

For those of you who haven’t been following the debate, there are a couple of linked hot-button issues that publishers have been dealing with over the last few years. One is whether journal content should be available to anyone (open access) or only to subscribers or site license users. Another is how journals should make the transition from print to web publication, which includes the question of who’s going to pay for our work once journals move away from print subscriptions as a major revenue source. (If you doubt that journal editors add value by publishing your work, ask yourself why you don’t just post a description of the experiment on your own website and skip all that pesky peer review hassle.) Like the rest of us, the editors of the Journal of Neuroscience are trying to figure out how to keep the community happy and the journal sustainable at the same time, and this session was their attempt to bring society members into the conversation.

The panel included people from the American Physical Society, the librarian organization SPARC, Science magazine, Highwire Press, Elsevier and PNAS. We heard some cautionary tales – when PNAS tried making their content freely available after one month in 2000, they had an 11% drop in subscriptions, and when the APS tried asking high-energy physicists to pay for publication in the early 1990s, many of them stopped submitting their work to the journal. But there were also some success stories, including several journals at which authors or third parties can pay to make articles free online at publication. At PNAS, about 20% of authors take this option. Donald Kennedy of Science emphasized that his journal is available to virtually all scientists through site licenses, though Andrew Watson, who edits the open-access Journal of Vision, challenged that view in the question period, calling the discussion “too nice”. He had a point: I’m not sure that someone unfamiliar with the debate would have been able to determine which panelists disagreed with each other on what issues.

Today I spent part of the afternoon at the Nature booth that the company has set up here at the meeting. Lots of people dropped by, some to chat about their interest in getting involved in editing, others to talk about their papers, and still others with general questions about the way things work behind the scenes. One question I get asked a lot is whether there is any way to tell if a paper that you’ve just written is something that Nature Neuroscience would be interested in publishing. I can certainly rattle off a list of traits that tend to characterize papers published in high profile journals, but I don’t know how helpful that is. One thing that has become clear to me in my own experiences writing papers, and in talking to authors, is that it is extremely difficult to objectively evaluate one’s own work. When you’ve spent a year or two conducting experiments and then writing them up, it is quite hard to step back and get a sense of how important or interesting the work is to a broad audience. Partially because of this, we do offer authors the option of submitting a Presubmission Inquiry to gauge the journal’s interest in a potential submission. This is easy to do, but the effectiveness of this depends critically on writing a useful description of the work. Quite often authors will describe the work in very abstract terms, or mainly describe the conclusions that they have drawn from the work, rather than what experiments were conducted. The most useful thing for an editor is to get a concise description of the science that was done, as well as it’s motivation and a brief statement of what is learned from it. For us to give useful feedback about whether the work is likely to be appropriate for the journal, the abstract should not be too abstract.

The evening brought the annual MIT Picower party. This year it was held at a large club in the Midtown area of Atlanta, and was totally packed. I would say there were well upwards of 1000 people in attendance, talking, drinking and dancing. Attendance was tilted towards the younger end of the SFN attendee spectrum, but there were quite a few faculty in attendance as well. When I left at around 1 things showed no signs of slowing down…

One important thing I get out of this meeting is the opportunity to talk with scientists about how the journal is doing and what we might improve. People seem generally happy with the papers that we’re publishing, but I’ve gotten a lot of interesting suggestions too.

The most controversial one was that we should publish the reviews of accepted papers (anonymously) online. Some people loved this idea, saying that reviews contain a wealth of information, from the reasons that the referees found the work exciting to the potential limitations of the authors’ approach, which would be especially helpful to scientists reading papers outside their field. Another potential advantage would be to make the editorial decision process more transparent to readers. Knowing that the reviews would be posted might also influence referees to write them more carefully and do a better job of citing references to document their concerns about novelty or conceptual advance.

The naysayers are equally passionate, telling me that posting reviews would place an additional burden on the referees, making them more reluctant to review and probably leading them to express their opinions less candidly. People worry that allowing a large number of readers to see the reviews might increase the odds of someone identifying a referee. One scientist said that editors would be dodging their responsibilities by posting the reviews online, making it look like we were trying to shift the blame for the decision onto the referees. Others note that many flaws identified during the review process will have been fixed by the time the paper is published, so the criticisms raised by referees are no longer relevant.

I’m sympathetic to these concerns. Of course we would never post reviews without the permission of the affected authors and referees, and I don’t advocate making it a regular habit for all accepted papers. On the other hand, I can imagine that it would be helpful to readers in certain cases. It will be a while before we decide whether to try it, but in the meantime, I’d be interested in your opinions on the subject.

I flew in Thursday night for the Friday auditory physiology satellite meeting organized by Xiaoqin Wang and Yale Cohen. It was an interesting meeting – in addition to posters and morning talks on new research findings, an afternoon session was focused exclusively on methods. A wide range of auditory neurophysiologists gave 10 minute slide presentations of training techniques and the details of their preps. It is widely acknowledged that training animals on auditory tasks tends to be harder than training them on visual tasks, which I think made everyone inclined to share tips and techniques.

Today brought the start of the main meeting. I also missed Gehry’s talk, but made it to the visual coding session, which featured some nice talks. The poster session was, as always, more about running into people than about absorbing many new results, but for me that is one of the best things about this meeting. You run into everyone…