While answering a comment in another recent post, I was struck by a thought I have had other times, but which I tend to remove. This is about the fact that it is surprisingly hard to produce a
paper in a large experimental collaboration in high-energy physics. The amount of work required to put together a sound analysis of collider data is quite sizable, and the pains of going through the internal review process may last months, when not a year or even longer.
Of course it is nice to have the benefit of signing each and every paper that our collaborators wrote: my publication list counts over 440 entries by now. But what that number means, it is not so clear -I edited about twenty of those papers, contributed significantly to another 20 or so, and barely read some other 200, while I have not even browsed through the remaining half.
A simple computation will clarify matters. Let us compute the rate of publications per author by CDF, an experiment which has now produced articles for over 20 years:
Number of publications: O(600) (I do not care about the exact number since this is an example).
Number of authors per paper: O(600)
This gives a rather startling picture of improductivity of any given collaborator, doesn't it ? It basically means that every CDF collaborator produced the equivalent of one paper in 20 years of work!
... On a different note, the number of colleagues who actually worked at the analysis which will soon become our first CMS paper -the one I announced here- is 9. This means I will soon be able to claim the merit of one ninth of a publication. Mumble mumble... This basically justifies about two years of work! I can go back to blogging for quite a while now, apparently. Or can't I ?
Regardless of what you think of the previous question, please consider: the typical theorist works in a group of three to five, and produces one paper a year. Okay, you may question these numbers, but it is the right order of magnitude. And it means that a theorist is, on average, five times more productive than a CDF collaborator, since his paper-per-head-per-year rate is 0.25, as opposed 0.05. Food for thought!
Of course it is nice to have the benefit of signing each and every paper that our collaborators wrote: my publication list counts over 440 entries by now. But what that number means, it is not so clear -I edited about twenty of those papers, contributed significantly to another 20 or so, and barely read some other 200, while I have not even browsed through the remaining half.
A simple computation will clarify matters. Let us compute the rate of publications per author by CDF, an experiment which has now produced articles for over 20 years:
Number of publications: O(600) (I do not care about the exact number since this is an example).
Number of authors per paper: O(600)
This gives a rather startling picture of improductivity of any given collaborator, doesn't it ? It basically means that every CDF collaborator produced the equivalent of one paper in 20 years of work!
... On a different note, the number of colleagues who actually worked at the analysis which will soon become our first CMS paper -the one I announced here- is 9. This means I will soon be able to claim the merit of one ninth of a publication. Mumble mumble... This basically justifies about two years of work! I can go back to blogging for quite a while now, apparently. Or can't I ?
Regardless of what you think of the previous question, please consider: the typical theorist works in a group of three to five, and produces one paper a year. Okay, you may question these numbers, but it is the right order of magnitude. And it means that a theorist is, on average, five times more productive than a CDF collaborator, since his paper-per-head-per-year rate is 0.25, as opposed 0.05. Food for thought!
Such numbers mean very little, the majority of theorist papers are worthless junk. It's probably safe to say that many if not most living theoretical physicists will have zero impact on the future of physics (all the susy and string ones for example). While the impact of many experimental physicists may also be small it is much less likely to be zero.