The Two-Conference Solution

Anyone who follows the theoretical computer science blogs knows that two peoples—the Technicians and the Conceptualists—have been warring over the same tiny piece of land (the STOC/FOCS accepted papers list) for well over a generation.  The most fundamentalist of the Conceptualists believe that STOC and FOCS were promised to them in a divine covenant with Merlin, while moderates simply point out that the Conceptualists have maintained a continuous presence in these conferences since the time of Cook and Karp, always turning STOCward in prayer on the day of the submission deadline; and that, if not for STOC and FOCS, conceptual papers might get wiped entirely off the face of the earth (or worse, shunted to CCC).  For their part, the Technicians see the Conceptualists as unwelcome usurpers, infiltrating an ancient land of log factors with bizarre new models and definitions; and suggest that, if the Conceptualists feel so wronged by physicists, biologists, and economists who refuse to see the natural and social worlds in computational terms, then let the physicists, biologists, and economists give the Conceptualists sessions in their conferences.

To many of us, it’s become increasingly clear that the only long-term solution to this bitter conflict is partition: two sets of conferences for two peoples with irreconcilable intellectual aspirations.  (A few old-timers, such as Noam Chomsky, still advocate technical and conceptual papers side-by-side in the same conference, but others consider Chomsky’s proposal as quaint and outdated as his hierarchy.)

And thus I’m pleased to point my readers to two new conferences, one for each people, the first of which has the further merit of actually existing:

  • Innovations in Computer Science (ICS) (“encouraging new ideas, approaches, perspectives, conceptual frameworks and techniques”), to be held for the first time January 4-7, 2010 in Beijing.
  • SLOGN (“a new conference in theoretical computer science, narrowly construed, encouraging difficult arguments, analyses, and algorithms”), to be held April 1, 2010 atop Mount Everest.

Mihai Pătraşcu—arguably the most irăşcible of the Technicians—has announced his support for the new ICS conference, stating that ICS “seems like one of the best ideas in decades for improving the quality of STOC/FOCS.”  As one of the handwaviest of the Conceptualists, I wish to announce my wholehearted support of SLOGN, for precisely the same reason.

And if Mihai and I are in complete agreement about how the field should evolve, what could there possibly be to argue about?  Shalom, Salaam, and QED.



16 Responses to “The Two-Conference Solution”

  1. harrison Says:

    Scott, I’m not sure I should believe that the SLOGN website is actually real and isn’t a construction of spambots on the web.

    Also, I think SLOGN may need a catchier slogan…

  2. anonymous Says:

    Sounds like a great solution! I, particularly, am looking forward to the first proceedings of SLOGN. :)

  3. Anup Says:

    If only theory were so dramatic.

    I fear that the Technician/Conceptualist cut is not a balanced cut… those looking for a war will find no one on the Technician side to fight with…

    How to fight an enemy with no visible army and no representatives?? It’s the question of our times.

  4. Toby Says:

    Glad you’re blogging again, Scott. That conference looks like it will be fun. Beijing is great, though it’s a bit far away.

  5. juhan Says:

    “to be held April 1, 2010 atop Mount Everest”
    Wow! I come! :-D

  6. Bram Cohen Says:

    So, despite your snark about SLOGN, you DON’T think that ICS is a form of ghettoization, and will be happy to submit to it? Even after taking the time to parse the joke, I’m still not sure.

  7. John SIdles Says:

    The Intel Corporation hosts many web-sites and training-seminars that (essentially) are SLOGN-oriented. A search for “INTEL iterative sparse solvers” will find good examples.

    SLOGN-brand concrete maths is not widely regarded as “sexy”, but it is essential to the application of QIT/QIS conceptual maths (Berezin symbols, Lindblad forms) to practical engineering problems.

  8. Scott Says:

    Bram: I should clarify that I had nothing to do with setting up the SLOGN website; I believe Rocco Servedio and Ryan O’Donnell were the culprits. Boaz Barak pointed me to it, and it was also on Michael Mitzenmacher’s blog.

    Basically, I appreciate any good-natured snark, even if it mocks and ridicules everything I stand for. Honor among snarkers.

    And yes, I’ll probably submit to ICS, among other reasons because I’m long overdue for a visit to Andy Yao’s group in Beijing. I just need a new conceptual insight between now and September…

  9. matt Says:

    As a more radical suggestion, TCS could follow the practice of just about every other scientific field, and move away from conferences and use journal articles instead and maybe even electronic, open access publication. When some TCS papers are basically math, does a conference referee really provide a careful enough reading? And often the full details of the result are hidden behind publisher portals which require a payment to see the article even for most university libraries. Plus journals have much more convenient submission deadlines.

  10. John Sidles Says:

    Anup says: I fear that the Technician/Conceptualist cut is not a balanced cut…

    Perhaps it is not even a particularly useful cut?

    Two terrific articles in the Dec 2008 IEEE Information Theory Society Newsletter (the first by David Donoho, the second by Candes and Tao) discuss compressed sampling vis-a-vis not from the viewpoint of a Technician/Conceptualist cut, but rather (what amounts to) a FrameBuilder/TheoremProver cut.

    Donoho’s article argues eloquently that “Frameworks do matter and we should honor the framework builders” while Candes and Tao offer case studies in which theorem-proving provides the necessary foundations for framework-building.

    Interestingly, the Candes and Tao article and the Donoho article both embrace the work “crystallize” to describe the interplay of ideas ⇔ theorems ⇔ frameworks.

    These two articles are (IMHO) in good accord with one another. Aren’t the most exciting conferences—in any discipline—those at which crystallization is in the air? Conversely, isn’t privileging any one style of research a short-sighted practice that leads to boring conferences?

    By the way, in that same IEEE newsletter, the words “idea”, “theorem”, and “framework” appear in the proportions 36/38/12 … to me this seems like a pretty good ratio.

  11. Razvan Musaloiu-E. Says:

    Off-topic: irăşcible is irascibil in Romanian. :P

  12. anonym Says:

    Scott,

    i think you should read the comment of “xiaowei” in mihai’s blog in detail

  13. Scott Says:

    anonym: I do have serious misgivings about cooperating with an authoritarian regime, as the ITCS in Tsinghua clearly does. On the other hand, curiosity-driven science has a well-known corrosive effect, and a priori seems as likely to threaten repressive regimes as prop them up. What do others think? It certainly merits some discussion in our community.

    In any case, my understanding was that the ICS conference will move around, and not ultimately be tied to any one country or university (but maybe I’m mistaken).

  14. xiaowei Says:

    hi scott, I have been following your website and have seen a reference to a comment i made on a previous blog.

    please dont be mistaken by thinking that ITCS in Tsinghua DOES NOT work with an authoritarian regime as an expat and someone who has seen the system works its wonders you should know that the communist system does its best to make sure everyone is connected to it. they most probably have a person responsible for decision in that group that reports and partially belongs to the communist party. i say partially because if not wholly owned then this person is sharing its mentalily with the regime.

    if such a person exists in that group it will not be tremendously difficult to point that one person out over a duration of time.

    so, as i said before everything that shines is not gold.

  15. Anonymous Says:

    In any case, my understanding was that the ICS conference will move around, and not ultimately be tied to any one country or university (but maybe I’m mistaken).

    What is your understanding based on? I have no authoritative information, but my impression is that ICS is intended to be based primarily in China (or, at the very least, Asia). I wouldn’t be surprised if it occasionally moves, the same way FOCS and STOC have been known to happen in Europe, but I think the current organizers would be very unhappy if the 2012-2018 locations were mostly in North America or Europe. In fact, I predict that there will be few serious proposals to host ICS outside of Asia, since nobody wants to upset many Chinese theorists by seeming to hijack their conference.

  16. anon Says:

    xiaowei wrote: they [the communist party] most probably have a person responsible for decision in that group [ITCS] that reports and partially belongs to the communist party. [...] if such a person exists in that group it will not be tremendously difficult to point that one person out over a duration of time.

    From a quick look around the ITCS website, that would seem to be Professor Amy Yuexuan WANG.