Faster resolution of ambiguous citations?

A place for users to ask each other questions, make suggestions, and discuss Bookends.
Post Reply
myeary
Posts: 6
Joined: Wed Jul 05, 2006 9:54 am

Faster resolution of ambiguous citations?

Post by myeary »

Hi Jon,

I've been doing a lot of citation scanning as I wrap up my diss, and I had a thought that might make a useful feature: allow the user to limit the fields that are searched when "try to resolve ambiguous citations" is selected. (Disclaimer: I don't use unique identifiers in my citations. I type them in as I go.)

As an example, I cite {Bregman 1990} quite a bit; he only wrote one thing in 1990, so this should be easy, right? Problem: two later articles refer to this work in their abstracts, so wherever it occurs in my document I'm asked to choose from three sources: the book I want, and the two articles I don't.

As a workaround, I've *erased* all abstract and keyword fields -- essentially keeping two databases, one with full information and one that's stripped down for citation formatting. This is obviously not sustainable, but it's getting the job done for now. What *would* be sustainable is an option to restrict the fields that are used in the lookup -- I would suggest limiting the search to Author, Title, and Date.

Thanks for your time!

- Mark
Jon
Site Admin
Posts: 10104
Joined: Tue Jul 13, 2004 6:27 pm
Location: Bethesda, MD
Contact:

Re: Faster resolution of ambiguous citations?

Post by Jon »

Hi,

That's a thought, although it does limit the usefulness of being able to create temp citations yourself based on a word or two you remember from the paper (including the abstract). An alternative would be to do a Proofreading scan occasionally, and let Bookends resolve the citations for you, replacing the ones you entered with "professional" (so to speak) citations that are unambiguous (I recommend citing by Author, Date, Unique ID for this purpose).

Jon
Sonny Software
myeary
Posts: 6
Joined: Wed Jul 05, 2006 9:54 am

Re: Faster resolution of ambiguous citations?

Post by myeary »

Hi Jon,

Thanks for the tip! Proofreading does the trick.

(I'd still like to make the case for limiting as stated above, though; there's still a lot of clicking to do when proofreading a 50-page document. I guess it comes down to me regarding the "temporary" citation markers as rather non-temporary; it's what I've used when writing them "by hand" and replacing them later, and using {author date first-word-of-title-or-journal} seems to be somewhat of a standard among citation software practices.)

- Mark
ozean
Posts: 461
Joined: Fri Mar 04, 2005 11:53 am
Location: Norway
Contact:

Re: Faster resolution of ambiguous citations?

Post by ozean »

Hi Mark, have you tried to insert citations directly from Bookends? Usually you can either hit a keyboard shortcut (e.g. command+Y) to switch from your word processor to Bookends, where you would just enter the first letter of the author’s last name to move to the approximate location of the reference in your database, then you select the appropriate reference with the mouse or by using the arrow keys and then you hit command-Y once more to be taken back to your word processor with the reference already inserted in Bookends’ preferred manner.

I know that this breaks your document internal workflow, because you have to switch to another application. But it might actually be faster than typing in the contents of the { } brackets, since you just need to hit keys/click four times:
(1) keyboard shortcut, (2) first letter of author’s name, (3) klick to select, (4) keyboard shortcut
before you can continue writing your text. :)
myeary
Posts: 6
Joined: Wed Jul 05, 2006 9:54 am

Re: Faster resolution of ambiguous citations?

Post by myeary »

Thank you for your suggestion. Yes, I have tried the command-Y switch-and-paste method, but I prefer not to, for two reasons:

One: it's not any faster. the time it takes to switch applications, look up an author (from a database of about 1000 entries), and switch back is about the time it takes me to type {Smith 1990 Pantalones}.

Two: as you mention, it breaks my workflow. I feel this is a non-trivial point, since people shouldn't have to change workflows to suit applications -- they find applications to suit their workflows. (I imagine I'm not alone in this, given the cornucopia of "focused" writing applications presently flooding the market.)

With the suggestions above, I'll be able to use Bookends to finish my current project; I'd like to keep using it for future projects as well, and more "intelligent" scanning of non-unique citations would be a big help.

- Mark
ozean
Posts: 461
Joined: Fri Mar 04, 2005 11:53 am
Location: Norway
Contact:

Re: Faster resolution of ambiguous citations?

Post by ozean »

I can really understand the workflow argument and agree it is quite important too. I guess I might even follow the same routine as you, if only my memory for names and titles wouldn’t be so bad… ;)
Jon
Site Admin
Posts: 10104
Joined: Tue Jul 13, 2004 6:27 pm
Location: Bethesda, MD
Contact:

Re: Faster resolution of ambiguous citations?

Post by Jon »

Bookends has had the ability to handle citations entered by the user (not Bookends) from the very beginning for just the reasons you outline. Proofreading scans were invented to deal with the problem of incorrect/ambiguous citations you may have entered (and don't forget you're only talking about ambiguity, but there is the problem of incorrectly cited works, e.g. misspelling of a name or incorrect year, which proofreading scans take care of, too). They work pretty well to "clean up" your document, and only need to be run occasionally (once if you are ready to submit). Yes, Bookends could omit certain fields during a scan. But be aware that such scans would carry some not insignificant overhead. First, they'd be slower (Bookends has an index that keeps track of all fields -- searching 27 or so individual fields will take more time). Second, there would need to be a UI for this in Preferences (as if we need more preference options), perhaps a rather complex one if the user gets to select which fields to scan and which to ignore. For these reasons, and given that relatively few people cite this way and there is a proofreading scan in place for people who do, I'm inclined to leave things as they are. Do others feel strongly about this?

Jon
Sonny Software
Post Reply