Monday, February 26, 2007

Dating a result..

The PCP results appeared in conferences in 1992, Sanjeev Arora's thesis appeared sometime after, and the "journal version" appeared in 1998. Similarly, the PRIMES in P result was announced in 2002, but was published in the Annals of Math in 2004.

If I'm trying to write something that talks about the history of a set of results (let's say I'm talking about PCP), it is generally recommended that I cite the definitive version (i.e the journal paper). So I'd talk about a PCP result "published in 1998", which seems silly given that everyone knows it appears much earlier. Given the lag time of CS journals, this is more of a problem than in other areas.

Is there any clean way out of this dilemma ? Should I cite the journal, but attempt to avoid any mention of dates in the text ? Should I use the "first published date" in the text, and cite both the earlier, conference version, AND the journal version ?

Maybe it doesn't matter, because the only results worth announcing well before publication are so famous that such questions are moot, and "everyone knows when it appeared". I don't have answers here: I'm just confused.


  1. Just cite both versions...

    But it can be a problem - consider the Clarkson-Shor technique (journal version), which is in fact Clarkson technique (conference version).

  2. In the text, cite the announcement date and then in the bibliography give the details of the journal version with a note saying "First appeared in ..." or "First announced at ..."

  3. For situations like this, I typically write in the bibliography under the same item (journal version) that earlier or preliminary version appeared in conf X (specially if such a note exists in the journal version).

  4. Suffix array was defined in the early '90s \cite{1993journal}. (where conf version was in 1990).

    "starting with \cite{conf92} there was a lot of activity in PCP \cite{definitive version}the 90s ..."-type of usage.

    "Primality... \cite{2004} was announced in 2002."

    It should be clear from the text (and the fact that ideas are in the air for sometime) that if you give a first anouncement year and cite a journal version of a different year there is no error.

    What if you do not know about the earlier anouncement of some result?

    BTW, What does Knuth do?

  5. The second anonymous comment sounds perfect actually.
    From what appears in the text, one is interested in the info that helps for the easy reading, and a pointer to the details. It is sufficient to give all the details like journal version etc etc in the reference text, but it is always better to have the first date of official announcement of the result (may be the refereed conference date) in the main text. This helps in getting the picture about the timing of the result right there without having to search through the reference.

  6. I use the same solution as anon 2 - the key will be [ALMSS92] but the actual citation is to the journal version, where I add a note about when the preliminary version appeared.

    In fact, I even made a small modification to the bibtex style file that adds a field "PrelimYear" for this purpose. You can find it here

    Boaz Barak

  7. That's neat, Boaz. Thanks for the pointer.

  8. What if you do not know about the earlier anouncement of some result?


Disqus for The Geomblog