Knowledge/Information centricity vs document centricity

Knowledge/Information centricity vs document centricity

Summary: Mentioning "knowledge centricity" is like using the word "paradigm."  People really roll their eyes when the abstract terms slip off tongues as evidenced by the way my colleague Dan Farber characterized my usage of the phrase as a dropped bomb at the end of yesterday's Dan & David Show.

TOPICS: Software

Mentioning "knowledge centricity" is like using the word "paradigm."  People really roll their eyes when the abstract terms slip off tongues as evidenced by the way my colleague Dan Farber characterized my usage of the phrase as a dropped bomb at the end of yesterday's Dan & David Show.  The subject came up during a conversation about the document format battle between the OpenDocument and Open XML camps.  Dan, who was about to end the show pinned me down and asked me to explain what the heck I was talking about.  it's a fair question and rather than transcribe the entire dialogue, here's what I said in the punchline:

Imagine if the way the Wikipedia was built was through all of the contributors to it passing around documents to each other via email... It would just grind to a halt and the Wikipedia would never be what it is today.  Take the same way the Wikipedia has been built and figure out how to organize the way your company runs so that knowledge is easily collaborated on an distributed in a very efficient manner and you don't rely on proprietary systems [to manage your content].

In fairness to Dan, I poo-pooed on the idea that document formats are relevant in this wiki/blog/rss world which isn't totally true.  For example, HTML may be the standard format for accessing and searching knowledge and information.  But HTML is not a good native format for housing the raw material (not to mention that it can't house some raw material like images).  So, we do need authoring standards for the raw material.  PNG and JPG for images for example.  And when creating text with WYSIWYG, the final resting place for that text should be a place that all systems can equally access (much the same way so much software can access JPG and PNG images).  

Imagine for example, you edit some text with SocialText, MediaWiki, or JotSpot and press the save button.  Should that text be natively saved in the non-standard markup language used by whichever of those wikis are using? Or should it be stored in a format that can be easily accessed or authored by not just wikis, but other software -- perhaps word processors -- as well?  I could see myself using a my favorite word processor to author some knowledge (not a document per se) while on an airplane and clicking "save to wiki" when I'm done. Then, when I land and connect to the Internet, that knowledge is automatically poured into a wiki where it instantly becomes available as highly accessible (via a browser) and highly searchable (because of HTML) knowledge.  A common format for encoding that knowledge -- a format understood by both the word processor and the wiki --  would be the enabler. 

Here's another example.  As I write this, Dan Bricklin is huddled in his study somewhere in Massachusetts, toiling over his next brain-child: WikiCalc.  What wikis are to text and images, WikiCalc is to spreadsheets (the etymology dating back to the first electronic spreadsheet, invented by Dan Bricklin and Bob Frankston -- known as VisiCalc).  Back to the airplane, should I need to be connected to the Web to author a WikiCalc formatted body of information or knowledge? Or, should I be able to encapsulate knoweldge that's well-suited row-and-column with any spreadsheet software and click Save To Wiki when I'm done?

So, in my podcast with Dan, where I did at one point say "who cares about formats?," I erred.  Standard formats are critical to the sort of frictionless authoring, flow, accessibility to, and collaboration over knowledge.  But the minute we pidgeon hole ourselves into thinking about knowledge and information as documents, we also doom ourselves to requiring bloated, expensive, and largely non-interoperable layers of software that add friction to the creation and distribution of important knowledge and the generation of business value.  The sort of value that the Wikipedia generated in its world.  Either that, or we should start referring to knowledge workers as document workers.  Which brings up a final point.  In some ways, the OpenDocument format has the wrong name. Perhaps it should be called the OpenKnowledge format.

Topic: Software

Kick off your day with ZDNet's daily email newsletter. It's the freshest tech news and opinion, served hot. Get it.


Log in or register to join the discussion
  • It's ok, David...Dan misuses the term "deconstruct"...

    as well.

    There seems to be a penchant for media types to expropriate terms philsophopy: paradigm (Thomas Kuhn), deconstruct (Derrida, Heidegger) amonng others. Y'all think it makes ya sound real cool! But we (who know better...ex-philosophy majors) forgive, you mean well!
    • Terms of art

      After all, a word means whatever the people using it agree on.

      You philosophers have a bad habit of getting math wrong, too [1] but it's cool as long as you're not trying to communicate with mathematicians who use the same (overloaded) terms of art.

      [1] Yes, I know who used the words of symbolic math first. Thus my point.
      Yagotta B. Kidding
      • well, those of us understanding both, try to be consistent ...

        with our terms in whichever Art we are expressing. I just think that too often terms are mis-appropriated and therefore lose their original force and meaning. But I like both Dave & for the of IT Art, I'll suffer the misappropriation! btw, YGBK...I generally love your comments on the board...tx!
        • <blush!>

          Yagotta B. Kidding
  • It's not really about information OR knowledge!

    In the immortal words of Wittgenstein, "Wovon man nich sprechen kann, darueber muss man schweigen." (Anyone unfamiliar with German and/or Wittgenstein should have not trouble finding a translation on the Web!) The point is that it really does not matter whether we are speaking about information or knowledge if all the supporting examples we give are ultimately about MEMORY (or recollection . . . I am still not sure why Aristotle distinguishes those two terms).

    Of course, to the extent that we CAN speak of memory, we encounter some awkward pieces of biological evidence. While we all seem quite happy to talk about memory as if it were some kind of repository and then argue about indexing structures (or ontologies, if we have been seduced by philosophy) and search algorithms, there are a lot of "wet brain" biologists who feel very strongly that it only makes sense to talk about memory as an ONGOING PROCESS. (I got this from reading Gerald Edelman.) This is painful. We are all very comfortable talking about static objects (like repositories) that we can build, probe, and even grow into other static objects (change of state). We are far less comfortable when it comes to "thinking in time" (and we have had this discomfort since the pre-Socratics). So we can follow Wittgenstein's advice and just shut up about the whole thing, leaving well enough alone; or we can grab the bull by the horns and try to start analyzing the human dynamics of recollection in the course of work and play. If our analysis ever progresses far enough, we might be able to model it with technology!
    • wow...great response!

      couldn't agree more. Since its convenient for us to believe that things per-sist, we deceive ourselves into thinking they in fact do so...but as you point out, such seems the case.

      I just think it's sad that terms into which a great deal of work/thought has been put, end up being commoditized (Marx on Language???)
    • Utility of memory

      Both memory and recollection dredge up the past, but one has already been sorted out.

      We have, in the next place, to treat of Memory and Remembering, considering its nature, its cause, and the part of the soul to which this experience, as well as that of Recollecting, belongs. For the persons who possess a retentive memory are not identical with those who excel in power of recollection; indeed, as a rule, slow people have a good memory, whereas those who are quick-witted and clever are better at recollecting.

      I prefer memory when I have the time; more savor, less purpose, and the possibility of surprise. I also read as slowly as I can, trying to withhold response.
      Anton Philidor
    • Perhaps a distinction is necessary

      We can talk on the one hand about methods that might model how human memory works and on the other about tools that support us by representing information in a useful way.

      The first task may of be great significance from a psychological point of view but the scope is, I think, very ambitious and probably not realisable with current methods (and may not turn out to be very useful from a practical perspective in building information systems).

      I think we can build systems that can represent and manipulate information in useful ways but these may not and need not reflect how the human brain processes information. This is a central theme of Winograd and Flores's book "Understanding Computers and Cognition". We can build useful tools, but they don't have to emulate human understanding or be "intelligent".

      Now it is fairly clear that we can build a system that we can guarantee is logically consistent with itself and this should prevent us from making assertions that are logically contradictory. However we cannot systematise how people interpret the symbols we represent in the system, this is a matter of the set of users having some common (and necessarily informal) understanding of what the system "means".

      So we could assert that two words are synonyms and use this information to infer further information from our system. However getting the system itself to reliably infer that two words are synonyms is an altogether more challenging task (and it might even be difficult to decide what "reliably" actually means in this context). This is because meaning is a fundamentally social construct and the relationship between word and meaning is fundamentally arbitrary.

      However we can construct the system to prevent us asserting that two words are both synonyms and antonyms of each other.
  • A sort of Wikipedia for our copany?

    Hi there,

    After googling for a while, I stumbled upon your story..
    We are a relatively small consulting firm with two offices in the U.S. and one in Europe..
    In the business of consulting, each client's needs are different, but in the beginning stages, we often refer back to old proposals and templates as new business begins...
    As the process continues, we rely on our library of key terms and ideas used in several of our transformations with clients..
    So..we have a drive with all of these documents, etc..but no one really knows where things are, who said what in that e mail months ago, etc.
    We want to make a "wikipedia" for our company. These important templates, information, keywords, everything..we want it all together, easy to access by our employees.

    What kind of platform is needed to do this?
    What kind of professional help might we need? We do not have a full-time IT expert with the company.

    Any information/resources/advice would be SO greatly appreciated. Thank you in advance.