Paul Ginsparg: Next Generation Implications of... - science21

Ginsparg: founder of the physics preprint archive, http://arxiv.org - Michael Nielsen

Yes, indeed questions usually at least as interesting as talk. - carolh

Apology of somewhat lack of org due to this mornings various distractions. (Reduce expectations for talk?) - carolh

Entering college students don't remember Reagan... - Chad Orzel

(Don't remind me...) - Chad Orzel

recommendation for a talk by Jon Eisen - basically the story of his first PloS Biology editorial - can't immediately find the audio - anyone got it? - Cameron Neylon

Is it this one? "A genomic encyclopedia of bacteria and archaea" http://www.scivee.tv/node/7258 - Jen Dodd

'taxpayer access to the taxpayer funded literature' - Cameron Neylon

Maybe this one: "Responsible Authorship and the Ownership of Scientific Knowledge: Thoughts on Open Access Publishing" http://depts.washington.edu/uwbri... (scroll down to lecture 2). - Jen Dodd

Stakeholders not paying attention. What to do about that ....? - carolh

Why was it not necessary for Congress to pass a law requiring users to post to YouTube or Flickr? - Michael Nielsen

arxiv: about 500,000 articles, growing at a rate of 60000 articles per year. 20 references per article. About 50 million downloads in 2007. About 600 per article (mean, not median). - Michael Nielsen

"know if had a counter up there, we'd have 20,000 people aiming for that number 500,000 article." - Chad Orzel

Mirror sites no longer necessary, but politically impossible to close down. - Michael Nielsen

Average 600 dowloads/article from 1996-2007. That sounds awfully high-- I wonder what the median number is? - Chad Orzel

Got arxiv.org because Brewster Kahle had already got archive.org! (Benefit - Google unique.) - Michael Nielsen

Hasn't changed much in ten years... - Michael Nielsen

Linear growth in submission rate to a pretty good approximation. - Michael Nielsen

Surprisingly linear-looking growth curve - Chad Orzel

http://backreaction.blogspot.com/2007... - Sabine Hossenfelder

Ramped up somewhat more quickly than Nature Precedings. Thoroughly enthusiastic about the latter. - Michael Nielsen

Talks about difficulties facing Nature: people may be suspicious, might think they have a hidden agenda. - Michael Nielsen

Interesting repeating pattern of peaks and dips, especially more recently - wonder what causes them all? - Jen Dodd

Breaks down the growth: different communities gradually turn on interest, and then saturate, and then new communities come in. Very interesting breakdown by community. - Michael Nielsen

High-energy saturates around 2000. cond-mat starts to saturate more recently, not quite complete. - Michael Nielsen

really interesting that the different communities start at very different rates and then increase at very different rates - maths looks exponential - big community - Cameron Neylon

Saturation in cond-mat is mostly theory-- not complete participation of experimental community. - Chad Orzel

I was thinking that the curves looked pretty similar, but math is just going to a higher asymptote... - Chad Orzel

Enormous spike in submissions at about 5pm. - Chad Orzel

Submisions spike is for astro-ph, just after the deadline for the next day. It's about a factor of 5 or so above every other time period! - Michael Nielsen

They do it so they appear first the next day... - Michael Nielsen

I'm curious that there's also a (much smaller) peak around 5am! - Jen Dodd

I guess right the increase in citations for being first on the list! Do I get a prize? - Cameron Neylon

Factor of two difference in number of citations, two years later, if paper near top of next day's listing. - Chad Orzel

Massive attempt to be more visible. - Michael Nielsen

Fascinating: attempt to disentangle visibility bias versus self-promotion bias. I.e., compare whether it's because the papers are more visible, or whether it's because these same people are also likely to promote elsewhere. - Michael Nielsen

I would guess the other peak is a timezone effect of some sort - but it seems early for American timezones - Cameron Neylon

Self-promotion effect is a significantly larger effect, although the visibility does help. - Michael Nielsen

closer look at the graph - looks like a North American and European midday hump to me - Cameron Neylon

Of course - that would make sense. - Jen Dodd

Astrophysicists have a really short attention span. - Chad Orzel

Why are there periodic ripples in the fraction of articles with 10 reads? - Chad Orzel

7 days, 14 days... hard to tell, eyeballing. - Michael Nielsen

They are just mondays? - Cameron Neylon

I think that's right - you can follow the little peaks out to around 80, and there's the right number of them to fit in (11 peaks before 80). - Jen Dodd

The horizontal axis is days since publication, though, so there shouldn't be a day-of-the-week effect. Unless articles are more likely to be posted on Mondays? - Chad Orzel

Good point - I'd certainly believe there would be a Monday bias. - Jen Dodd

rushing through a variety of things ebcause we're 10 minutes over - Cameron Neylon

Network benefits to both readers and authors. - Michael Nielsen

oh -and I'm supposedly chairing :-) Presuming it doesn't matter if we're late to lunch because its scheduled for two hours - Cameron Neylon

Neo-Minsky quote: "Can you imagine they used to have an internet in which authors, databases, articles, and readers didn't talk to each other?" - Michael Nielsen

"new research and cognitive methodologies..." - Michael Nielsen

Questions... - Michael Nielsen

Garrett Lisi: How do you decide which sites to accept trackbacks from? - Michael Nielsen

Some context for the last question: http://latticeqcd.blogspot.com/2006... - Michael Nielsen

2.1k TrackBacks to 1.5k articles this year, about 4% of total - Chad Orzel

expanded TrackBacks somewhat to include news articles and so forth - Chad Orzel

"a lot of our methodology is stuck in the 1990s, because that's where I'm stuck." (with a laugh) - Michael Nielsen

Planning to change the way they order articles in feeds, etc. to reduce need for post-deadline spike. - Chad Orzel

Re: the self-promotion effect, I wonder if Ginsparg was referring to this article: http://arxiv.org/abs... - Hilary

Recording (video and slides) is at http://pirsa.org/07090080/ - Jen Dodd