Do Digital Docs Need Page Numbers?

Perspectives on HistoryFrom time to time, we receive requests from authors for bibliographic information about articles—their own or by others—they have seen on the Perspectives Online web site. As often as not, they are looking for the specific page numbers from the print version of the newsmagazine, for reference in an essay or a book being set up for publication—typically because a sharp-eyed editor is pressing them to give the full citation, page numbers and all. As regular readers already know, Perspectives Online articles are almost always drawn from the pages of the printed newsmagazine, and while the digital version carries information about the issue (month, year, and column title) in which the article first appeared in print, it does not carry the page numbers. I appreciate the scholarly diligence of the publisher or editor who is not satisfied with a mere URL of the web page where the digital article appeared—especially these days, when hard-pressed publishers and their copyeditors seem to pay less and less attention to such details. But are page numbers of the printed version really necessary, especially when citing an article read only in its digital form?

On this question, the magisterial Chicago Manual of Style declares in its 16th edition (at 1.116): “One of the primary advantages of a traditional printed-and-bound book or journal is the presence of page numbers. Page numbers allow students and researchers to make precise citations to the works they consult, allowing readers to retrace their steps.” The manual encourages publishers, therefore, to “incorporate stable page numbers … or to otherwise number the elements in their web-based publications (e.g., through paragraph or section numbers) whenever possible.” But even magisterial tomes must be flexible and adaptive. So the CMS 16 suggests (at 14.4 and passim) using URLs and DOIs for citing digital material. We are all constantly using a URL, or uniform resource locator, (such as http://www.historians.org/Perspectives/issues/2011/1105/1105pre1.cfm, for instance) to locate and read material on the Internet. Unless the web page changes, that URL is an excellent pointer to the digital text, more precise and reliable, in fact, than the insistently dulcet tones of the friendly GPS voice in the car, which can, and sometimes does, lead you up the strange mazes of unknown roads. But the DOI is less familiar. The acronym stands for Digital Object Identifier, but is not used, appearances to the contrary, only for identifying digital objects. Instead, it is a digital identifier for all texts—in any format, whether in print or in an electronic avatar—that are assigned a unique DOI by a registration agency and by the concerned publisher.

A scrupulous scholar might ask: It is all very well to have a URL or a DOI to locate an article or book online, but without page numbers, how can I help a reader find a particular quotation or passage? The simple answer is that you don’t have to. The curious reader can use the site’s search engine or search box, or even more simply, the “Find” function available in all web browsers that can locate any text string on a web page.

Implementing DOIs can be complex, cumbersome, and expensive for a publication on a shoestring budget, especially since they have not become ubiquitous yet. But URLs are easy. Every page on a website or weblog has one, even if it is sometimes disguised as a TinyURL or some other variant. Yes, “link rot,” changing domain names, and the other afflictions that web pages and hyperlinks are prone to, may hinder the retracing of the scholarly path. But even a bound volume of a printed journal can have missing pages or issues. And, of course, born-digital texts don’t (and can’t) sport page numbers—at least not the traditional kind. So let us not worry too much about not having page numbers for digital docs. They can, should, and will have URLs for now. But they can certainly dream on about their DOIs, and leave page numbers behind in the BC (before-the-cloud) era.

Share on FacebookTweet about this on TwitterShare on LinkedInShare on Google+Digg thisShare on StumbleUponEmail this to someonePrint this pagePin on PinterestShare on RedditShare on Tumblr

Back to Top

Leave a Reply

Comment

* Required field

  1. Michael P. Procter, Sr.

    Great article! The downside to relying on urls is that they tend to change over the years. No system is perfect as you implied, but page numbers would make it easier to find an item. I just wrote a paper where one of my sources was a book in the Kindle format. It would be nice if there was a standard agreement on how to cite the Kindle; I’m sure it’s coming soon.

    Reply
  2. Alexander Duryee

    String-match searches are wonderful for direct quotations, but for cited statistics or paraphrased data, it becomes impossible to find without some level of granularity. For citation-chasers, this means a research dead end: imagine trying to find one non-quoted statement in a major work!

    Pagination may die, but removing that level of bibliographic granularity is a disservice to anyone seeking further sources.

    Reply
  3. Diana Williams

    What if you’re trying to track down someone’s source and they don’t give you a precise quote to search for? Or what if you need to search for a phrase in a text that contains many instances of that phrase, like, say, the writings of Gertrude Stein? Not having page numbers—a way to break down a very long text into discrete smaller pieces—means that what ought to be a quick lookup turns into its own research project.

    Reply
  4. michael les benedict

    Digital citations will sooner or later become the norm. But they aren’t the norm yet. In the interim, it is hard on a researcher to omit the page number that she needs in order to provide the citations still required by most publishers today.

    Reply