DOI stands for Digital Object Identifier, and it’s a relatively new (ca. 2000) standard number that has been introduced to identify digital objects. For ancient studies, its most relevant use is providing a unique standard number that identifies a specific journal article. Books have long had standard numbers like ISBNs, as well as Library of Congress Classification Numbers (LCCN) and OCLC numbers that can be used to identify them. Until the DOI, journals could be identified by the ISSN, but individual articles were out of luck. DOIs in print citations for journal articles look like this, and are usually found at the end of the citation, after page numbers (if any): doi: 10.1080/1357627021000025469
In addition to providing citation precision, DOIs can be used to find digital copies of the articles in question. The International DOI Foundation (http://www.doi.org/) provides a DOI resolver at http://dx.doi.org/. (I love the simplicity of this page, by the way: a box, a sentence or two, and boom, there’s your article.) You can also resolve a DOI on the fly by just sticking the DOI numbers onto the end of that URL prefix: http://dx.doi.org/10.1080/1357627021000025469 will take you to the article in question. (Note that if the article is by subscription only, and you or your institution does not subscribe, you may not actually be able to access the full article.)
The question by @DrKillgrove that got the conversation started was (paraphrased) ‘I have an article in hand – how do I find out what its DOI is if it doesn’t say so anywhere?’ @YaleClassicsLib pointed out the free DOI lookup page at Crossref.org: http://www.crossref.org/guestquery/ This lets you type in a citation and see if a DOI exists for a given article. It really needs an ISSN to work, which you may have to track down, and it did not find the DOI when I entered basic citation information for the article linked above, so it’s not as lovely as the DOI resolver linked above. A Google Scholar search for your article title and “doi” might be a decent backup.
I should note that many articles do not have DOIs, so it was possible that none existed for the article Dr. Killgrove was looking for. DOIs are provided by publishers, who join the DOI system and pay a fee (which helps guarantee that the DOI system will continue to work in perpetuity). Journals that almost certainly have DOIs for their articles are US/Western scientific journals put out by major publishers. Journals no longer being published and journals not affiliated with major publishers may not have DOIs. Most journals that provide DOIs for current issues have also applied them retrospectively.
In sum: DOIs are good, because they allow
- precision citation in a numeric string and
- the quick finding of a digital copy of a given article using http://dx.doi.org/.
Please consider including them in your print citations, and using them in online citations to link directly to articles.