The Tyranny of the Citation

The recent and rightfully praised comparison of Zotero and EndNote raises many excellent points. A considerable amount of the article and the discussion it generated addresses citation prowess—the ability for the software to get properly formatted citations into your document (complicated by choice of word processor, of course). Such a focus makes perfect sense because citations function as a crucial professional apparatus (if not discourse) that must match the (identifying?) standards of one’s field and publication outlets and expectations.

But while citation formatting is one major reason to use bibliographic software, it isn’t necessarily the only or even primary reason. We should also think more broadly about the curation and sustainability of our reference libraries and how we might shake up our assumptions about what we value in our tools. The inevitable advice that one should employ the tool that best fits your needs or workflow is correct, provided we’re able to step back and evaluate whether aspects of our workflow couldn’t be improved. I worry that feature by feature comparisons don’t really facilitate such reflection or highlight the values that aren’t so easily compared head to head.

As wholly dependent as I am on Zotero for managing my reference library, I don’t use it for creating citations. Of course one reason I first used reference management software was because I wanted all references tied to the same tool that could reformat all citations with a few clicks. What if I needed to reformat all my citations to submit to a particular journal?!? It turns out that after many years I’ve never wanted to do that, nor even known anyone who has wanted to do that (of course people have find this very valuable in theory).

So why don’t I use the citation functionality? As robust and powerful as the CSL formatting language is, it doesn’t work all that well with the rather messy variety of sources I cite and how I like to splice commentary into my notes. This is not to say it’s impossible, but that it takes me more time to battle Zotero than to manage the notes independently. This is what I find remarkable about the ubiquitous debate about citation accuracy amongst various software options and citation mechanisms. It doesn’t take that long to manage citations by hand (while you’re walking uphill to/from school in the snow, etc.). Yes, I may have to manually correct ibid and short and long titles, but i just do this as a last revision so that I’m not constantly redoing it. It takes a little extra time, but seriously less time than trying to get Zotero (or especially anything else) to output the exact format I want.

To create the reference in the first place, I use the Shift-Cmd-A shortcut to grab a nicely formatted version from the Zotero item, then paste it into the document. I rarely need to edit it again, and i don’t have to worry about any particular formatting problems or losing manual edits when Zotero reformats citations. If you’re citing mostly more recent and standardized sources, and you don’t pollute your notes with prose as I do, these concerns are less relevant. But anyone who regularly deals with non-standardized sources might refocus from citation formatting (regardless of software preference) to other virtues that facilitate long term library maintenance and organization. As for creating bibliographies, I maintain (with surprisingly little effort) a collection of works cited for each writing project, readings course, etc. When I need a bibliography, I can simply generate one from the collection and edit it as need be. Usually the need is rather little, as bibliography formatting is usually more straightforward than that of notes.

As for speed, I think it’s the ultimate red herring in bibliographic software. Granted, efficiency is a real concern and i get a blood pressure spike whenever i see the spinning color wheel for any reason. Other similarly impatient folks have legitimately complained that Zotero is too slow under certain (sometimes bizarre) conditions. Some of this is due to its dependency on Firefox, a dependency that is being dissolved as i write. But the amount of time i spend waiting on Zotero in comparison to the time i spend researching and writing (not to mention other distractions) is so absurdly small, it’s laughable—even if it’s an easy an check in the +/- columns of a tool comparison. But missing the big picture in favor of the obvious metrics is not unlike the beer-bellied cyclist spending an extra $1000 for a lighter bike or a high-performance derailleur. Actually, just skip lunch.

OK. i lied. I do care about speed in one respect, namely the way that Zotero is magically convenient for quickly getting references from library catalogs or databases into a library (which i usually do by hand rather than via massive searches; i prefer a lean library), thus saving tons of time otherwise lost to manually populating Zotero items. More important for my sanity and future utility of my library, it’s a snap (in the process of saving an item) to get references into the right collections and subcollections. I use collections to denote stuff i need to read on a certain topic, stuff i’ve read but need to work into a project, stuff i have consulted but won’t use or cite, stuff i need to consult on my next visit to national library of medicine, etc. Of course Zotero is not the only tool that can help keep your library organized (nor are collections), but i’ve found this method the easiest, even though i resisted collections for some time simply because they were a new way of organizing for me.

Incidentally, i might mention another way that Zotero is flexible that i came to appreciate later on–namely in note-taking with the detached tinyMCE window next to the PDF i’m reading, or just if i want a small window because i’m doing other things at the same time. And those notes are of course immediately available from wherever i’m accessing my library. Once upon a time i thought i should be annotating PDFs, but found that i preferred to have notes in a simple text box that i can see at a glance. The annotations were, like citations, more interesting in theory than in practice for me. Zotero is not flexible in all ways, of course. I have remained frustrated by its lack of custom item types for many years. I won’t enumerate the many more valuable features that have been implemented in the meantime. When I stopped worrying about citing everything with Zotero, I stopped caring as much about custom item types.

Flexibility can built into workflows as well. People have also complained about wanting to use other browsers than Firefox; some have even abandoned Zotero because they prefer another browser, as if they must preserve one particular workflow at all costs. I have found it perfectly easy to use Firefox for finding and saving references and to manage my library, while using Chrome for everything else. Having Firefox open just for Zotero is really no different than running a standalone application like EndNote. In any case, as Zotero decouples from Firefox, browser issues are a diminishing concern for the end user.

Citations, speed, and other automation features aside, one simply cannot neglect the value of openness and unimagined uses. One commenter mentioned that openness should have been a part of the original comparison, and Croxall defended its omission by saying that it’s not a concern of the general user, even if it should be. But it seems that part of the value of such comparative pieces like Croxall’s is how it can take users beyond their own limitations.

Openness IS a crucial concern for long term use, which is where any bibliographic manager pays real dividends. It’s about neither the philosophy nor the supposed moral superiority of open source. With EndNote (and others), you’re buying into a system that the producer has every incentive to keep you locked into. If you don’t buy upgrades continuously, can you still get your data (and modifications to it) out? I completely agree that it’s not unreasonable to pay even $100 for software you that you like and use. But it’s not just a one time payment, but that amount every few years. Or your library won’t sync. Or you can’t import new records. Or it won’t work with your OS upgrade. Or whatever. But there will be a reason why you’ll need to pay again. Ad infinitum.

Yes, virtually all reference managers can all export in various standard formats. But I’m not the only one who has moved libraries more times than I care to admit, and who has subsequently concluded that it’s never easy as instructions say, never lossless, and only slightly preferable to, say, contracting the plague. This is especially true if you have non-standard or modified items in your library. Openness does have real implications for the average user, and not just because closed source and proprietary formats are the kryptonite of the digital humanities. Even if typical users do not value openness, comparative articles ought to explain why they should.

It is indeed convenient and useful to compare products and establish pros and cons across a matrix of functionality. The enthusiastic response to Croxall’s informative and well-written article shows how useful this work is. But in looking for the “best” software for managing citations, we must not lose sight of the long term sustainability of our personal reference libraries in favor of marginal differences between routine tasks. And we need to shed our workflow blinders to recognize other possibilities and features of the tools besides those that fit well with old habit or desired new ones, like a utopian need for seamless citation integration.

I certainly don’t mean to suggest that my priorities or methods should be anyone else’s. But I do enjoy reading how people use various tools, as I always learn something. But if we do want to put products head to head, i think it’s worth considering how Zotero is about far more than citing references, especially in terms of openness and flexibility that results from it not having a financial agenda tied to how we use it or its social network.