The Unclassification of Knowledge
Does knowledge have a structure?
Embedded Files
Even if knowledge quantity cannot be directly measured, brave attempts have been made to construct meta-structures of it. It is important to note the meta qualification. Knowledge meta-structuring is not the same as the structuring of specific knowledge at a domain or application level (eg., the taxonomy of: Domain, Kingdom, Phylum, Class, etc., which is the structuring of the knowledge of biology). Meta-structuring of knowledge is the structuring of knowledge itself, no matter what the knowledge pertains to or represents.
A popular meta-structure for knowledge was proposed by Russell Ackoff, though it likely has a much earlier history (including from the poet T.S. Eliot [1]). This is often referred to as the DIKW pyramid.
I (and others) have added the “lowest” level of Noise to define a (N)DIKW [2] pyramid.
1. W: Wisdom
2. K: Knowledge
3. I: Information
4. D: Data
5. N: Noise
Noise is just stuff: unstructured, perhaps random, unpredictable, context-free stuff. It might be electrical static, random digits, molecular vibrations, whatever. It is present and has some agency, but is neither defined nor delineated much beyond the application of the labels ("electrical", "random", etc.)[3]. Further definition and identification of characteristics would require some structuring process which, by the very definition (sic) of Noise, it does not have.
Data was considered by Ackoff to be the application of some symbology (since he didn’t originally mention “noise” or any other source, one might well ask: to what is the symbology being applied and how would one apply it?).
Information, then, is defined as “useful” Data manipulated (without necessarily explaining how the data is manipulated, by whom, and how it is qualified as being "sufficiently" manipulated to actually qualify for the label "Information"?) to help answer open-ended “who?”, “what?”, “where?” and “when” questions (about, um, what?).
Knowledge, according to this model is derived from Information then answers “why?” questions (again, useful and sued for what purpose?). The unanswered considerations listed for Data-to-Information also pertain here.
Wisdom is finally derived from Knowledge at the apex of the pyramid. It was considered by Ackoff and others as some rather amorphous spiritual aspect required by and restricted to humans; it wasn't better defined.
This is all good and as a model it has proved useful [4]. In addressing what I think are its limitations I am not simply being nit-picking; these models were thought up and have been used extensively by some very smart people. It is just that, in wrestling with some of these definitions I think there are some loose ends.
FOOTNOTES
[1] In The Rock, the poet T.S.Eliot wrote: “Where is the life we have lost in living? Where is the wisdom we have lost in knowledge? Where is the knowledge we have lost in information?” (my emphasis)
[2] In the following sections, I have identified the Noise, Data, Information etc., as it refers to these labelled things in the (N)DIKW model by capitalizing and underscoring the words. I identify general noise, data, information un-capitalized without underlining. I hope it’s not too confusing; the challenge is the slightly different meanings and contexts of the same words.
[3] Of course, the labels themselves do imply both the existence of the item (static, digits,...) and some degree of their characteristic and function--that being the purpose of labels. Without labels applied to something, we cannot process that something at all. The transition to Information is the beginning of the acquisition (identification, discovery, creation,...) of the knowledge of the something and requires the application (identification, discovery, creation,...) of further labels (characteristics, behavior,...).
[4] Statistician George E. P. Box is usually credited with the aphorism “All models are wrong; some models are useful.”
The (N)DIKW model is useful.
Page updated
Report abuse