Deep Citations



The point of citations is to enter your work into the academic discourse, connecting previous work with your own - putting it in context: therefore you could say that citations are the synapses of the academic discourse - it is through citations that explicit connections are shown so you could say that citations are a form of high level academic thought. It therefore follows that the richer we can interact through citations, the richer academic dialogue we can take part in.


The same thought can be applied to all industries, not just academia, but this is where the notion of citations have the deepest roots.


Please note that bibliography and references are something else - this is about reading a text and being able to summon additional resources to verify the veracity of claims and to get further context.




“c. 1300, "summons, written notice to appear," from Old French citation or directly from Latin citationem (nominative citatio) "a command," noun of action from past participle stem of citare "to summon, urge, call; put in sudden motion, call forward; rouse, excite" (see cite). Meaning "passage cited, quotation" is from 1540s. From 1918 as "a mention in an official dispatch."”





core citation attributes


When you come across a citation you should immediately be able to ascertain:


  • Who and what is being cited. This should be immediately apparent
  • Basic Academic Citation Information: Document Title, Author, Date, Web Address, Location, Publisher, as well as identifiers such as ISBN, DOI, ISSN, PubMed or arXiv IDs




'deep' citation attributes


To become deeper, quick interactions should also provide:


  • The value/veracity of the cited resource
  • What further information the cited resource offers
  • What the context for the citation is
  • Any tags for this citation by the author, editor or fellow readers
  • Any 'foot' notes the author has chosen to include
  • Indication of how the citation relates to the text (what is the link type)


The reader should also be able to easily interrogate the properties of the citation to answer the following questions:


  • Is it accurately cited? Does the source really do state what the presentation of it asserts?
  • Is it authoritative? Is it itself based on correctly carried out and analysed primary work or does it build correctly on correct previous work?
  • Is it primary? Or does it relate to the original work which can easily be accessed?
  • Is it relevant?
  • What more can this source offer of relevance?
  • What is the context in which this was stated?
  • What do others have to say about this?
  • How can I find other relevant information through this citation?


Importantly, it is crucial that the interaction is smooth and fast, otherwise the citations will not be interacted with and will not be followed in the normal course of reading, which is where the most value can be added.




specific proposals for an improved citation system


With a basis in the goals of what citations should show, whether the cited resources is accurately cites, authoritative and relevant, and with general concerns of optimizing connections in general, these are the current proposals as to how to improve citations:


citation implementations


  • Citations should be added to specific text in the document (such as an assertion or a direct quote) as a richer version of a URL, not added in brackets after text as just plain text. This is to help make it clear what the citation refers to, both to the writer, the reader and the computer systems for analysis in the future (see the Citation Analysis section below).


address everything + everyone


  • Web addresses for more resources
    • It should be possible to have web addresses for are required for as many resources as possible, including raw data, chat logs and so on. This is not so much a technical issue as it is a hosting, organisational, commercial and privacy issue.
  • High resolution addressability
    • Within digital documents it will be important to deliver on one of Doug Engelbart's most crucial points: Addressable Paragraphs (or higher resolution). It will also be increasingly important to reference a wider range of information such as transcripts and other media. Furthermore, why is it we only cite documents and not people? We should work towards a way to reference people directly, to tag names as people, so that we can follow flows of what people are saying or doing, independently of their large and rare magnum opus tomes.
  • Rich Multimedia Support
    • Make it possible to cite sections of audio/video at specific moments in time, with instant playback and context.
  • Person Addressability
    • There is a need to be able to address people through some sturdy means, to make sure that what is refereed to really is associated with a specific person. (I have a feeling this can be facilitated by building up basic databases of any information associated with any person involved in the dialogue and letting powerful computers (any modern computer) sort out any discrepancies).
  • Space Addressability
    • Universal GPS coordinates can be useful when citing spoken words at specific locations.
  • Time Addressability
    • The notion of addressing a specific time can be powerful for citations and for movement around time vectors. Please see the related Time Browser Project for more on this.


implementation issues


  • We are accustomed to dealing with large image files and video files, so why not import the entire cited work into the document being worked on? The only issues here are commercial, not technical.




You can refer to a page on the Current implementations & issues.