We can see details; How can we see wholes?

The Problem

However, while large and open wikis and comprehensive directories provide increasing opportunities to network and connect, it’s hard to see the Whole. (Points of greatest influence, Peter+Trudy Johnson-Lenz)
If we could start with universe, we would automatically avoid leaving out any strategically critical variables. – “Operating Manual for Spaceship Earth” by Buckminster Fuller

There are countless FreeSoftware projects, websites, subjects, questions, repositories, books, individuals, and so on – it is very hard to see them in context, as a whole.

FreshMeat is nice, but that’s part of what the Johnson-Lenz quote refers to: “comprehensive directories.”

Similarly, in our world of “institutes,” “collaboratories,” and individuals dedicated to the study of CollectiveIntelligence, we have the same problem: myriad myriad myriad. Again, can have the difficulty of seeing the whole.

Some things that point towards ways of “seeing the whole:”

Hard Drive Example

Here’s another example:

What is it to “look at a hard drive?”

Envision what it is to see the whole hard drive.

Every single file. Obviously, a small monitor can’t do it, but envision that an entire huge wall is made of monitor, perhaps part of the floors and ceiling as well, and that drawn onto the monitor, in a TreeMap or 3-D perhaps, is your entire hard drive: All of the folders, all of the contents. And more than that, that the contents are transparent. You see the heads of all text files beyond a certain size. Folders that are teeming with little itty bitty files are actually, visibly, … folders teeming with little itty bitty files.

And you can see everything, or at least everything not occluded by something else, in a gigantic, glorious, display.

This gives us something of a sense of “seeing the whole.”

Now apply this to, say, search results. You search for something, and the enormous universe of stuff appears before you. Countless web pages, books, people, etc., etc.,.

This gives us something of a sense of “seeing the whole.” Or more appropriately, “seeing the wholes” – because there are many wholes, many ways of seeing entire collections.

This is not to negate the important question, “What about things that aren’t in the collection?” But it is to point to something that people aren’t really thinking about, that we should be thinking about.

Solution Space

User Interfaces

Reguarding user interfaces: 3D is something to consider, because we are in a “real world arbitrariness” / chaos in display, and the arbitrary terrain / real-estate will have psychological and political effects, meaning that more degrees of freedom will be better.

There will be hoards of SemanticWebBrowser?s, for viewing all this data, all these wholes. Views of the semantic web will have independent URLs, and constitute a form of community digest in the form of a digest.

see also: SharedAwarenessSystem.

See Also


There are many aspects to this whole issue, quite a few of which are very much of interest to me this year. For example…

  • We are currently testing the limitations of a set of theories that presume the entire contents of the internet can be thought of as an EnDimensionalRaggedArray? that is entirely fractal.
    • Any cell of “TheArray?” can be Named (and addressed or accessed) by means of Tags.
    • Display (or “rendering” of TheArray? will generally be constrained by the hardware “window” (monitor size).
      • In this “relative” approach, it is possible to think of the display as a “lens” that can panned across the two dimensions of TheArray? that are currently being displayed (as a page). It is then possible to view more content by either scrolling down or across the page, or by moving the content which the ‘lens’ is displaying (i.e. changing the page being displayed by clicking on a link.)
    • Higher dimensions of TheArray? can be accessed by using a link to ‘jump’. This is especially powerful if the names of the cells of TheArray? are comprised of the elements of the vectors (dimensions) that can be mathematically manipulated since this makes it possible to ‘compute’ the name of the ‘next’ cell(s).
    • Obviously TheArray? is an extremely large Object, made up of an extremely large number of nested objects.
      • This is what necessitates the use of “Ragged” (or Associative) addressing models given that any single computer will never be able to allocate all of the memory needed even to create just the Name and Address tupels, much less to store the contents.
      • Since the (fully qualified) name of a cell in such an object is likely to itself be very large, it becomes essential to have a LocalNames capability that can infer, from the current context and the shortest possible acronym, what the fully qualified name is. This “inference”, based on the current contents of the data space, combined with the fact that TheArray? will never be completely defined, necessitates the use of Expert Systems methodologies as opposed to the more conventional Procedural programming methodologies.

At present, we are building prototypes that combine the use of TiddlyWiki for Client-Side computation with OddMuse for “repository” functions. While this hybrid approach is somewhat surprising (even to us), it seems it provides an effective compromise of the respective capabilities that is adequate to our current needs. That being said, we are already starting to incorporate other external processing results from other environments such as external tagging systems that generate RssFeeds.

One of the interesting thngs about this type of Research and Development is that it is a source of delightful surprises regarding what is and isn’t really possible and practical!

Continue the churn… There is this theory of “wholeness” of Christopher Alexander. And there is the idealistic / marxist misconception of “the whole” I was often confronted with (the tale of the wise men and the elefant). Actually these conceptions are quite different (1) the whole as an organism like a city (2) the whole as “everything”. There is also the Doug Hofstaedter concept of intelligence / perception that builds very much (most intensely in the COPYCAT project, done my Melanie Mitchell) on the concept of slippage, a kind of stochastic pattern matching that allows to see the similarity in differing patterns. This allows for different solutions, depending on the perspective.

For example, “harddisk as a whole” for me (not necessarily for anyone else) means (1) how much is used, how much is free (2) how much of the used space is assigned to the various topic (might be 30% for programming, 20% for office related things, 50% multimedia data) (3) how much space could I make available on removing outdated or overly large objects.

Two quick notes:

  • There are always multiple wholes.
  • All wholes have holes. :)

The second is a quote from PeggyHolman.

I sense that we are all agreeing about some of the various aspects of this relatively complex subject. Some of the specific aspects that are consuming my time just now result from trying to push the envelops of…

  • Pattern recognition - by trying to define “objects” that can be recognized by my digital Agents and then me enhanced or refined automatically. (My over-arching term for this is Hwo:PdaEs - a personal Digital Advisors (not Assistants, I don’t trust it them enough yet to allow that Autonomy yet!)
    • NameConventions? for such Objects and the Tags that are associated with them, such that a central Wikipedia:Inference_engine can sense the NameSpace(s) surrounding it, and Recommend actions and options for my consideration.
  • Discrimination of Contexts, so that a block of MicroContent may be evaluated and processed differently in each of the contents within which it appears. Obviously, this includes the challenges of dynamic translation of LocalNames between communicating environments (or contexts).

A better way of stating that “My sense that we are agreeing…” might of course be better explained by stating that a number of these ideas are currently resonating with me, if only because of the contexts within which I am working.

The unique ability that all life is being built around is that of pattern recognition. Although computers are billion of times faster in crunching numbers, humans are unmeasurable better at recognizing patterns. There is this PatternTheory of ChristopherAlexander (I wrote this German Book MusterTheorie about) who localizes 15 fundamental properties of living systems. Two of these properties are tightly related to recognition / perception. One is named ECHOES by CA (I name it SIMILARITY). The other one is named ROUGHNESS by CA (I name it INDIVIDUALITY). Anyway, these are about “the identical in the different” and about “the different in the identical”. The constructivists (Piaget et al.) name this functions “assimilation” and “accomodation” IIRC.

To make long things short, there is an field of “pattern theory” outside of academia, that is slowly developing and that deals with such problems.


  • It’s a genuine pleasure to see a post from you!
  • “MicroPatterns?” is what I should define as my subset of your much broader interests in Patters, that I consider to be a component of my MicroContent interests.
  • With any luck, I may actually have a bit of time to start reading your book this summer, assuming…
  1. I can buy a copy (advice would be appreciated)
  2. I can brush up my German reading skills enough that my efforts to read it are sufficient to let me acquire and retain its and its insights between the inevitable interruptions that generally causes me to have to start over too frequently to make any real progress.
  • With even more luck, I might be able to finish reading it before the fall AND find the time to come back to Europe and perhaps even chat about this at the conference in Portugal OR find the time to go to Portugal and leave from Berlin (allowing me to buy you “ein Bier oder Kaffee, as I pass through Graz)

Hans, the pleasure is on my side. Beer or coffee is always welcome. I think you have my email address. Send me your postal address and I send you the book.

Das Ganze - the whole. “Die Ganzen” - the wholes, it simply doesn’t exist, in German a least. ;)

Agreed, though I don’t think I understood it when I wrote this.

“Seeing the Wholes.”

Furthermore, “All wholes have holes,” as PeggyHolman told me.

Define external redirect: EnDimensionalRaggedArray TheArray SemanticWebBrowser NameConventions MicroPatterns

EditNearLinks: UserInterface RssFeeds IncidentalCollaboration ChristopherAlexander FreeSoftware TreeMap