5

I am handling the Collected Papers of a colleague who has died several years ago. We are preparing to publish (as a series of books) about 5000 pages of mathematics that he wrote. I have almost finished the LaTeX typesetting for the entire project. (The bibliography has almost 3000 entries, just FYI.) Now I want to spend some months working on the index to the entire project. I want it to be very detailed. This will be a balance between automation/scripts (as a starting point) and then lots of human experience too.

Question: I would like advice about "best methods" from people who successfully built an index for a very large collection of mathematics. What worked well? What techniques failed? Any advice and "lessons learned", that you wish you had known at the start? I would like to learn from the experiences of others, please.

For context here: I was one of 3 or 4 people who helped Don Knuth (at his personal invitation) by proofreading his Collected Papers years ago, before they were published by CSLI, but I did not help with the process of making the index to these volumes. Now that I am building a large index, I would like to rely on the expertise of the larger TeX/LaTeX community for some "words of wisdom" about how to efficiently and accurately build a comprehensive and detailed index. After all, the index to a book (or, in this case, series of books) is perhaps the most important part.

All advice about indexing is welcome here, please. Thank you in advance for the advice/suggestions/tips, including suggestions about where to hear the experiences of other editors in this regard.

5
  • P.S. Just as a point of interest/context: Approximately 80 percent of the pages are in English and 20 percent in French. (Rough estimate.) I switch routinely between the dictionaries in emacs while editing. Similarly, the bibliography respects the source language of all citations. (I used bibtex with babelbib for this purpose. Please don't suggest a switch to biblatex now; I made the choice many years ago, and with 3000 entries in the bibliography, I'm sticking with bibtex.) I mention English versus French distinction since I want to plan ahead with respect to the needs of the index.
    – mdw
    Nov 26, 2019 at 12:03
  • 1
    not suggesting that you switch to biblatex but it;s probably worth mentioning the size of the bibliography isn't an issue, as they both read .bib files. It is like switching bibliographystyles in bibtex in theory you can restyle an existing bibliography. There are always issues but same is true of switching bibtex styles some will assume matching tex macros to be defined in the document. so switching to biblatex might be an issue, but issues are not related to how big the bibliography is Nov 26, 2019 at 12:17
  • Thanks for this insight @DavidCarlisle! I only mention the size of the bibliography because it is complicated (really complicated). There are many different languages that appear in the bibliography, and I tried to respect the language and formatting of each one, as carefully as possible. You have been in the LaTeX community for a very long time. OK, question for you: Do you have colleagues who you suggest that I reach out to, with the goal of getting their sage advice about the topic that I posted? (P.S. It is very nice to hear from you again, after your recent help earlier this week!)
    – mdw
    Nov 26, 2019 at 12:27
  • 1
    Not really, I'd just think of some top level categories like "people" "math reviews subject classifications" or whatever and just read the entire thing adding categorised index entries everywhere then see what sorted index comes out, and iterate. only big index I have been involved in was the latex companion where we used a professional human indexer, quite an enlightening experience, the indexes were made by makeindex , but the index entries were specified by a human who had no prior knowledge of the subject area but inferred the main topics by er well reading the book, a skilled job:-) Nov 26, 2019 at 12:48
  • @DavidCarlisle I plan to be that human indexer (but definitely would never call myself a professional).
    – mdw
    Nov 26, 2019 at 12:55

1 Answer 1

1

I'm being presumptuous in trying to provide an answer as I am not a professional indexer, which I think is a skilled job. When I have developed indexes I have gone through the document noting people's names, topics, "significant" words, striking phrases, in fact anything that might be of interest. Thus ending up with a somewhat random collection of words, names, phrases. Then what are you to do with them? Are some subsidiary to others? Are some of no interest? Should some refer to others?

Do you want individual indexes, say of topics versus people, or a combined index, or ...? I do not envy your task. Any chance of getting a professional involved?

1
  • I don't think we will pay to get a professional indexer involved. We are publishing with an academic (university) publisher, and we aren't doing this for profit at all. It is just a labor of love. Yes, I'm thinking the same way that you are, namely, to pull out anything/everything that is interesting, by starting with some scripts and automated processed, and then filtering it to the things that actually seem useful. There will surely be lots of names (this person knew the literature in this area extensively) and the topics covered are relatively broad too. Thank you for chiming in!
    – mdw
    Nov 26, 2019 at 23:14

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .