Call me crazy but I liked the pun of the nun’s order and indexing. I guess enjoying such things is just a bad habit of mine.
Yes, a concordance uses all words, an index, only significant ones.
So far, only humans can mark the significant words in a sensible manner. One of the problems with computer created help files is the program makes a cross-reference to all words in a master text file. What could be more logical? However, if some new term is used that is different from usage elsewhere, no one will find it and the help index is useless. There still needs to be a human involved to make it practical.
Wow, Ed, dBase III? That takes me back. And it’s still a good program, but all those 5" floppies drive me crazy.
I think by the time I did this I had acquired a 30-meg hard drive, and having this vast storage capacity at my disposal was able to load everything onto the HD and dispense with the floppies. Boy, those days were sweet.
Just out of curiosity, what makes one index noticeably better than another? Besides omitting the location of an important concept in the book, what else can affect the quality of an index?
Absolutely. But what a computer can do is keep page references straight, across revisions. As Musicat said in the first place, you put in invisible index tags, saying, in effect, “Put this word or passage into the index under lookup word ‘xyz’.” You can also insert “xyz, see abc” entries, and, for larger books, make the index hierarchal, like[ul]
[li]Elves[LIST][/li][li]Anderson[/li][li]Rowling[/li][li]Tolkien[/li][/ul]
[/LIST]IBM had this stuff ready to use out of the box (on mainframes) by the mid 70s. They had to. In the 60s and 70s, IBM, reckoned merely as a publisher, was the largest publisher in the world, except for the US government.
And they had indices in those manuals? Where was it, in the 15th 4" looseleaf binder when there were only 14 binders on the shelf?
I’m familiar with embedded indexing. I think that Scarlett67 and I are not talking about the difference between embedded and standard indexing so much as the difference between doing an actual index – i.e., indexing important concepts – and just making a list of terms and phrases, which to my mind is more like a concordance or maybe a glossary. You can do a proper index with embedded tags, such as in Microsoft Word or FrameMaker or whatever. But simply highlighting important phrases and clicking “add tag” (or, God forbid, “add every instance”) is not creating an index.
As for Lakai’s question about index quality, there are some basic indexing rules, such as not having more than 10 page numbers per main entry; not having subentries that all have the exact same page number; no circular references (e.g., “Dogs. See canines,” and then “Canines. See dogs”), etc. Beyond that basic technical stuff, though, what makes an index for me is flipping through and seeing that the indexer has made the information easy to access. If I look something up using terminology different from the author’s, there’s a cross-reference waiting there to direct me to the correct location. The subentries are informative without being overly wordy. The “See also” references give me a sense of how the information in the book is connected.
I may not be doing a great job of describing this. Anyway, what struck me about the Straight Dope indexes was that they were really funny while still being informative. When I read the first one, I initially thought it was an in-joke of sorts, like funny credits at the end of a movie comedy. Then I realized, no, this is an actual index, it’s just humorously written. A couple of examples from “More of the Straight Dope”:
Great Britain, what’s so great about, 153
guillotine, whether one can stay briefly conscious after beheading by, 262-264
intestines, whether suction toilet sucked out someone’s, 260-262.
And so forth. 
I can’t remember a time when IBM manuals didn’t have indices, and my memory goes back to the mid-60s. (And I’ve handled some that went back to the 50s.)
Actually, for some subjects, like operating systems, IBM /did/ issue master-index volumes. They stopped doing it when operating systems were broken up into multiple products with staggered release schedules, but with the current arrangement where all the things that used to be separate products are now features of a single operating-system product, as it was in the 60s, with a single release schedule, maybe they have master-index volumes again.
I seem to recall Dave Barry’s books, or at least one of them, having similar indices. Although it’s possible I’m remembering one of the Straight Dope books.
(Off-topic: Firefox’s spellchecker doesn’t recognize “indices” as a valid word? Nor even provide “indexes” as a suggestion?)
Powers &8^]
I once prepared an index for a very lengthy deposition transcript. This was before I even went to law school. I actually thought it was kind of fun - a little like a literary treasure hunt, and I got paid for it, too. It didn’t offer the opportunity for quips and puns, though, more’s the pity.
MsWhatsit said:
It may not be intentionally funny, just that is the actual topics of the columns. Those are, in fact, paraphrasings of the column titles, slightly rephrased to make the noun of interest the first word.