Share First Profess Later

eek has a nice post about the concept of profession and collaboration. It's interesting: while collaboration is often very much about the creation of shared artifacts that help a group do something; a self-conscious profession, a professional society, often exists or is created to be the artifact that lets a group know they are attempting to do something.    (Q3X)

Especially when they aren't doing anything. The existence of the society defends against the existential dread of who are we and what do we do.    (Q3Y)

Professions and disciplines make me nervous, in a similar way to Eugene. What one does when collaborating is always more important than collaboration itself.    (Q3Z)

In the ideal situation, collaboration disappears into the background. If you find yourself enmeshed in the details of how your group should interact, you've missed a step. Take a pause, and work on your Shared Language.    (Q40)

Posted by cdent at 04:44 AM | Trackback This | Comments (4) | TrackBack (0) | | bl

Next Great P2P Application

I'm at work, where most of us telecommute and never see one another, and devised the following application to enhance productivity and shareholder value. Maybe you can use it in your distributed workplace too.    (Q3A)

Because, as myth has it, I hate everything and everyone, I tend to be quick on the anger draw. So in an effort to create a less hostile work environment I implore you to help me build the next great p2p application. Everyone in a team runs it on their workspace. You turn it on when you start work, it finds your peers.    (Q3B)

The system is based on a therapy that some Obsessive Compulsive Disorder people may use: They wear a rubber band on their wrist and snap it when they are doing "the bad thing".    (Q3C)

When a member of the team starts doing a "bad thing" someone else on the team, may, at their option, snap the offending teammate. The offender will see on their screen a big rubber band come in from the back of the screen and arrive at the front of the screen with an audible snap.    (Q3D)

With this we can snap each other when we aren't helping to increase shareholder value. When I'm getting angry too easily or being overly sensitive to some bizarre concept of justice, when Bob is talking too much, when Joe is yelling, when Ned storms out, when Barb doesn't read or write, when Norman is being "too mean", when Francine gets her math wrong, when Susan is too slow, when Barry is too indirect, everyone will be able to move on and get shit done.    (Q3E)

Or collapse under a hail of thunderous snapping.    (Q3F)

Clearly this thing needs a name.    (Q3G)

Posted by cdent at 11:30 PM | Trackback This | Comments (4) | TrackBack (0) | | bl

Feedback is Leadership?

Morning rambling. Take with drugs. Your milage may vary.    (Q2Z)

Feedback, in some fundamental way, is at the core of a successfully collaborating system. This was the fundamental insight of Wiener and his cybernetics. Whatever else we may think about what it has to say about life, the statement that information dispersal and feedback go hand in hand is one of the major intellectual contributions of the 20th century.    (Q30)

A collaborating system of any kind--ants, people, network services--is made of individual entities producing and consuming information. Individuals are the core contributors to the system. The sum of their amassed contributions may be greater than the parts, and often is, but at base it is the individual pieces and parts that are creating and contributing: having inspiration.    (Q31)

What those pieces and parts create and contribute is shaped, refined and eventually defined by feedback from the rest of the system.    (Q32)

No person is an island.    (Q33)

In an activity group, initial contributions are themselves feedback to some thing in the system: a catalyst. A meme or goal that becomes a leader, or is installed by a leader (which itself could be meme, goal or person). Activity (information) crystallizes around the catalyst, accretes around the growing mass of the signal which is the product (some fun, a paper, a tool, a house, a brand new idea) of the group.    (Q34)

If the growing product is not headed in the right direction how do we know? Feedback. How do we know where to go get headed in the right direction? Feedback. From the system itself.    (Q35)

If the system is insufficiently mature or the initial catalyst insufficiently clear (thus not really a catalyst) the meme, goal or person must be clarified, buttressed, tuned to be a true catalyst. That tuning is feedback, observation from outside the confines of the system, and is leadership.    (Q36)

Posted by cdent at 06:08 PM | Trackback This | Comments (2) | TrackBack (0) | | bl

Effective Reading

This is adapted from discussion at Socialtext where, because we are a distributed or virtual workforce, we move a huge amount of information around as text. It's been noted lately that the quantity of my blogging has slacked of late. This is mostly because I'm writing inside the walls of work, so here's something from the inside.    (PZ1)

Socialtext wikis are called workspaces. Individuals are members of some number of workspaces. These are usually divided up by topic, group or project. (Knowing this matters below.)    (PZ2)

I frequently find myself at the intersection of two events    (PZ3)

• some people wondering how I get so much read    (PZ4)
• some people complaining about not being able to keep up with the stuff to read; me becoming somewhat bewildered by that    (PZ5)

So mayhaps I should write down what I do and other folk can farm that for useful techniques or rationale to ignore me. This writing was somewhat inspired by discussion of how to maintain focus.    (PZ6)

In case you find this not to your liking, that's okay. This is what I do. It works for me. It may not work for you, but then again it may.    (PZ7)

First, there's no doubt that I read very quickly and skim at a very high level of presumed comprehension (that is, I'm able to convince myself that I got the bits that mattered; whether this is true or not is unknown). That's been true since I started reading. Second, I've become practiced with using tools that augment my ability to read (email filters, mail and rss reader settings, etc). Most of that practice happened when I was a sysadmin and my colleagues and I managed to get in the habit of feeling really nervous if our mail didn't get a response within a minute or so.    (PZ8)

I proceed from the assumption that as knowledge workers our primary job is to communicate. Communication is not overhead, it's the work. Things like writing code are reifications of previous communication. The quality of the code mirrors the quality of the communication and comprehension that precedes the generation of the code.    (PZ9)

So, starting from that assumption I have some general rules of thumb about what I should be doing:    (PZA)

• before writing or acting read everything in your circle of concern    (PZB)
  • all your email    (PZC)
  • all the new and updated content in workspaces you care about    (PZD)
• When it is time to write or act, do that until you are done, then read again.    (PZE)

Reading email and workspaces deserves more more explanation. First, turn off email notify in all workspace and get a good rss aggregator that can show diffs (this is perhaps the single most useful thing to do to enable "staying abreast"). Only subscribe to those workspaces that are germane, rely on others to let you know when something important is happening in some other workspace that is not in your immediate circle of concern. Second, use filters with your email: send email to different folders depending on topic or list. If possible get a mail program that lets you start it with a different set of folders available depending on the activity you are engaged in (for me this is work and not work). Use one that lets you hit the same key to just keep on reading.    (PZF)

For email:    (PZG)

1. as your read, delete everything that doesn't matter    (PZH)
2. if an incoming folder (other than annoying non-work mailing lists which have your non-continious always-partial attention) is above a certain size, say 20 messages or so, something is wrong. There either aren't enough incoming folders, or you aren't deleting and responding with alacrity.    (PZI)
3. read again    (PZJ)
4. respond to anything that needs a response and delete the responded-to message (if you can't trust your emailer to send your mail correctly, or keep copies of your sent mail, or otherwise make you feel confident in deleting messages, fix that!)    (PZK)
5. do whatever getting things done work-alike thing it takes for you to deal with messages you can't respond to but think are important signifiers in your life, but remember #2 above (I leave them in the folder and let #2 keep me in check)    (PZL)

For feeds:    (PZM)

1. when you are reading, anything that seems important, send to a tab in your browser    (PZN)
2. read everything    (PZO)
3. configure your aggregator so it updates itself differently for different kinds of feeds. I update Socialtext related feeds once an hour (or by hand whenever I'm in a break) but other feeds only update every six hours or so    (PZP)
4. when done reading go to your browser and review the tabs and deal with as appropriate (garden, comment, post to delicious, whatever), if it's been a while since the page was loaded in your browser, you may wish to reload    (PZQ)

Do these things without paying attention to IRC.    (PZR)

The trick here, for me, is that whether I'm in email or NetNewsWire I just sort of idly hit a key, soaking it all in. I generate a gestalt of the state of things I care about. Only after I've made it all the way through do I react because it is the whole state of the little work universe that matters when responding, not just the one atom of information that's currently under the cursor.    (PZS)

The snobby/arrogant part of me often feels like saying, "Have you read everything?" before having a conversation with people. I suspect this is annoying for those people. It is similarly annoying for me to have to come down to the limited MTU of speech...    (PZT)

I feel speech is best at two stages of the ShareLanguage? spectrum: when there is very little (and reading is not much help in developing understanding, either because there are no SharedGoals? to jumpstart understanding or there's just too much ignorance for a foundation to exist) and when there is a great deal (when so much SharedLanguage is present that MTU and bandwidth are high and speech has been transcended in favor of something like SharedBrain?). In the middle, reading can be a good tool because it situates wait states in the reader. Work colleagues are generally (or perhaps should be) in the middle, except for some occasions when pairing, face to face, or starting a new thing.    (PZU)

Clearly this is a self-serving attitude on my part, and I temper myself (at least a little) accordingly, at least when I remember to do so. But now you know. Feel free to comment, but please don't try to tell me I'm wrong for me or anyone else other than you. You're the only person you can be sure about. And even that's not clear.    (PZV)

There's an important corollary to the above rules:    (PZW)

• When possible, always send your email to a relevant group list that is archived    (PZX)

This helps insure the opportunities for people to know the gestalt of the environment. It also helps maintain artifacts.    (PZY)

Update: Eric Sinclair has some observations (of me and his own experiences).    (Q03)

Posted by cdent at 09:27 PM | Trackback This | Comments (2) | TrackBack (1) | | bl

Affinity for Categories

I went to IdeaDay last night to listen to MattMay and AlexWilliams talk about podcasting. I've not become a podcasting fan, perhaps because I don't use my ipod (headphone issues) or perhaps because I'm too nervous to go around with environmental noises blocked out, but none of that matters because I was primarily there to meet the BlueFlavor guys, see who else goes to these things and participate in Sabrina's forays into the Seattle social network of folks what do things like what she do.    (PX9)

The presentation was interesting but a little disjointed. Matt and Alex clearly know their stuff and could have talked for ages but I think they needed an agenda. An initial query to the audience of what they wanted to hear about threw things in the direction of how to make money podcasting. I think they should have started with a gentle introduction. It may have been a review for many of the people there, but would have established some baseline understanding and SharedLanguage.    (PXA)

The most relevant discussion was a tangent in response to a "how do I find stuff, it's hard to index audio" question from the audience. Matt mentioned discovering community and making use of affinity groups to find stuff.    (PXB)

This made some things related to tagging click. I've harped for a long time about the importance of SharedLanguage in collaboration and used naming as an ax to grind against the insidious evil of FreeLinks. Tags, WikiWords, nicknames are all ways of establishing affinity: If you have developed the SharedLanguage to understand the name, it's like having an invitation to a group. If you have affinity for the name, you may have affinity for the group using it; it's time to seek that invitation.    (PXC)

All these things--tags, nicknames, WikiWords--are markers for conceptual categories. They are _not_ labels for classes.    (PXD)

(An aside into my defintions of class and category may be necessary:    (PXE)

• a class is a defined grouping of entities in which the members fulfill the definition of the class and can be listed.  T    (PXF)
• a category is a cognitive label applied to a non-enumerable grouping of entities wherein membership is determined by typicality amongst the members and not some overarching definition.  T    (PXG)

)    (PXH)

Markers assume a measure of doubt, treasure it and get value out of connotation and suggestion. In other words, affinity.    (PXI)

Affinity based systems assume a measure of tolerance for the happy accident of discovery: "This isn't exactly what I was looking for, but, damn, it's cool!" It's not about information retrieval, but discovery and enhancement.    (PXJ)

Posted by cdent at 07:54 PM | Trackback This | Comments (2) | TrackBack (0) | | bl

Wiener, Grammar, Action and Me

I've been reading a biography of Norbert Wiener. The title is the ominous Dark Hero of the Information Age. I'm not that far in but it's been quite interesting thus far, despite the sometimes stilted writing.    (PV9)

When he was ten (c. 1904) he wrote his first philosophical paper, described in the book as "a treatise on the incompleteness of all knowledge". His conclusion: "In fact all human knowledge is based on an approximation."    (PVA)

Wiener went on to lay most of the foundations for information theory, stochastic processes, dealing with feedback in electronic systems; all based on probabilistic statistics. He rejected Bertrand Russell--one of his tutors and one of the authors of The Principia Mathematica--within a few months of meeting. Russell said there could be internal completeness, Wiener said there could not, and Gödel eventually proved Wiener correct.    (PVB)

Wiener also happened to be a wildly depressed guy. When there was noise in his own personal information system or he was not getting good feedback, he was unable to produce his usually good work or generally interact well with his life. He experienced self doubt and confusion within that mirrored his belief that nothing was certain, anywhere or anywhen.    (PVC)

To what extent did one pattern lead the other? Where was the balance in the symbiosis between the perception of self and perception of the outer world. Wiener's later work informs views of "it's all one big system".    (PVD)

I resonate with some of what's going on here. My drive to create information resources that are accessible, referenceable and reusable is driven by a personal need to externalize noise in my own system so I can do (occasionally) good work and interact well with my life. I want to make stigmergic structures out in the environment that I and others can use as what amount to optimized external decision makers and information chunks.    (PVE)

Simply tossing information out into the ether doesn't cut it. For the chunks to be useful they need to be transparent, transportable, composable and authentic (in the Heidegger sense). Transparent means you can see what the information is for or about. Transportable means that the information survives being moved out of its initial context (carries, creates or refers to its own context). Composable means that the information can be effectively reused inline with other information.    (PVF)

To put it another way, the information is like Lego: it is a building block in a system whose grammar can be perceived. This block can go on this block in these numerous but constrained ways.    (PVG)

Information, here, is one word for many things: exchanges of knowledge, perceptions, tools and processes.    (PVH)

In many cases the perception of a grammar relies on expertise. In order for more than one actor to collaborate in an information system, they must share some level of experience in the domain and have shared language. If they do not already have or first build shared language they will waste their time arguing the grammar of the available blocks with little composition. Until there is shared language very little will get done.    (PVI)

(See also Collaboration Requires Goals for related discussion.)    (PVJ)

So if found in a situation where shared language is too incomplete (as it always must be somewhat incomplete, see above) what are the tools to make more?    (PVK)

Posted by cdent at 11:36 PM | Trackback This | Comments (0) | TrackBack (0) | | bl

Dimensions in Group Communication Media

Anyone who has ever used a combination of email, phone, meetings and other media to communicate in a small group has experienced at least one moment where they've thought, "this would be going much better if we were having this conversation using X".    (PTD)

While the choice of X is often driven by personal preference and comfort levels (as someone who spent a few years being a sysadmin, my X is email), I believe there are several scales along which the various media can be measured. The resulting values (which are very much dependent on the context of the medium's use) can be used to choose a medium depending on the goals of the group or the individual initiating the conversation.    (PTE)

The three scales that are usually most relevant for me (I'm primarily concerned with enabling action and reflection) are:    (PTF)

Degree of Synchrony    (PTG)

Does the medium provide for synchronous or asynchronous responses. This depends not so much on the quickness with which a participant responds but rather whether the communication to which a response will eventually be made is stored for review. On the phone you can wait for a long time to make a response, but eventually you may forget what the other person said. This isn't the case with email.    (PTH)

Desired result of communication    (PTI)

Action versus Reflection. Should the participants be inspired to perform some action or to continue thinking at the end of the conversation? How does the medium encourage one or the other?    (PTJ)

Degree of Ambiguity    (PTK)

How clearly does the medium transmit already clear concepts? How effectively does it indicate there is doubt? What degree of doubt or noise does the medium inject into the conversational space?    (PTL)

These scales can be used to make three axes defining a three dimensional space. Each medium fits somewhere in the space. Each scale impacts the others.    (PTM)

• synchronous ... asynchronous    (PTN)
• action ... reflection    (PTO)
• ambiguity ... clarity    (PTP)

Neither end of any scale is bad. Each is good in some ways: effectively expressing doubt is a way of showing that there is room for reflection or a need for more information. However, using a medium that introduces or enhances ambiguity in a setting where quick action is required is a bad idea.    (PTQ)

Wikis, Weblogs and Email are all asynchronous media (the content is stored for later use) but they work more effectively in different areas. Email, because it is pushed to the reader, is often more effective at causing action. RSS feeds from wikis and weblogs could inspire action, but thus far there is not much of a tradition of use along these lines. They are more effective at reflection, especially wikis, which allow in place refactoring of content.    (PTR)

IRC, phone calls and face to face meetings are synchronous. IRC can introduce a great deal of noise where processes of use have not emerged. It can be a place for causing action, but leadership is required. It is often good for dipping into a shared pool of understanding for a bit of reflection. An unexpected phone call can inspire action while a pre-arranged conference call or face to face meeting is only effective with the help of processes and structures that facilitate the use and construction of information artifacts that will live on after the call.    (PTS)

Where would you map the various media? What dimension matters most to you? Someone could make a graph.    (PTT)

Posted by cdent at 02:39 AM | Trackback This | Comments (1) | TrackBack (0) | | bl

Old Stuff: Classification v Categorization

When I was in school I took an Information Architecture class that required a readings journal. Some of those entries deserve revision.    (PPB)

I was, at the time, especially in the difference, similarities and interrelationships between classification and categorization. What follows began life as Studer: Classification v. Categorization. The first version was written November, 2001.    (PPC)

The Studer article suggests there is a lack of consistency in the literature in the use of the terms classification and categorization. Studer uses the terms carelessly, especially when quoting: while he uses the term classification in his own text the quoted text uses category.    (PPE)

Studer makes it sound like the process of creating classifications is a step following the creation or identification of categories. This conflicts with my interpretations.    (PPF)

In my view classification is an artificial (synthetic, non-fundamental) process by which we organize things for presentation or later access. It involves the arbitrary creation of a group of classes which have explicit definitions and may be arranged in a hierarchy. In other words a class is strictly defined and once inhabited the inhabitants can be enumerated.    (PPG)

Categorization, on the other hand is a natural process in the sense that humans do it as part of their cognitive fundament. It is, like Studer reports, an act of simplification to make apprehension and comprehension of the environment more efficient. Categories spring up out of necessity and because they are designed to replace the details of definition are themselves resistant to definition. When provided with a list of stuff we are able to categorize the stuff, but when asked to list the full contents of a category we cannot.    (PPH)

So to put it more succinctly:    (PPI)

• a class is a defined grouping of entities in which the members fulfill the definition of the class and can be listed.    (PPJ)
• a category is a cognitive label applied to a non-enumerable grouping of entities wherein membership is determined by typicality amongst the members and not some overarching definition.    (PPK)

This is important to me, in part, because I'm playing around with trying to determine if computers can ever be actually intelligent or must always fake it. I vote for the latter because computers, thus far, cannot categorize.    (PPL)

The ability to categorize may be the basis for intelligence (On Intelligence, by Jeff Hawkins, presents some data to support this, as well as some assertions that may blow my "thus far" out of the water, given time). On the fly categorization allows us to place data in an informational context. Once in that matrix we can do what amounts to an endless recursive dialectic wherein each new synthesis becomes thesis.    (PPM)

Computers can presumably replicate this process but if they do, it is imitation. Their distinctions must be made by definition, by classification, not categorization. They can be made to appear to do categorization but the alternate representations they provide are rules (definition) based. Until recently the most promising research in creating seemingly intelligent machines has used what can be called a brute force approach: supply the computer with as much information as possible, related in as many ways as possible. This is the method that IBM used to get Deep Blue to become a chess champion and is one of the keys to the Semantic Web.    (PPN)

If we want to create truly intelligent machines we must determine how categorization works. I wonder, though, why we want intelligent machines. What do we gain from that? Don't we instead want machines that are tools to augment our own intelligence? If that's the case, then we are already have the understandings to make progress: we simply need to improve on what we have.    (PPO)

Posted by cdent at 05:08 AM | Trackback This | Comments (1) | TrackBack (0) | | bl

Queues and Stacks and Creepy Crawlies

In separate conversations with an old coworker and a new coworker I've stumbled upon a new word that I rather like.    (POV)

It seems to me that the most effective way to manage a collection of tasks is to model them as a queue: first in, first out. Real life requires that we be able to reorder the queue in response to changes in the world, but the general principle of FIFO applies.    (POW)

A queue can be lengthened or shortened as needed, but provides a fairly robust cognitive aid in the face of additional requests that happen when the queue is already full: "There's no room in here, what would you like me to take out to make room?" or "You can get in here, but you have to go on the end."    (POX)

Very often organizations or people that are too busy or unfocused lose the discipline to maintain a queue of tasks and switch to using a stack: first in, last out. I know I do this. A lot of forces influence this: It's important not to say no to someone or something; A mess of dependencies makes it difficult to choose which tasks to sacrifice in the queue ; When there is a perception that stack or queue overrun will cause an organizational crash of some kind, a stack absorbs tasks into its dark confines more easily.    (POY)

It's in these dark confines where the new word lurks. Those tasks which are in the depths of the stack, pushed in a long time ago, warped and sickly from lack of attention, are filobytes.    (POZ)

Posted by cdent at 10:45 PM | Trackback This | Comments (0) | TrackBack (0) | | bl

Genius Social Networks

Back in my junior year of high school, my physics teacher, who had a penchant for pausing class to say something pithy, etched the following phrase into my brain: Genius is the ability to find the connections between seemingly disparate things.    (POG)

That idea has driven much of my thinking since then.    (POH)

Phil Jones (with whom I'd been having a lively chat about purple numbers) recently found my Why Wiki? posting. He responds with some comments on his own aspirations:    (POI)

I've convinced myself recently that if there were a universe where it were possible for statements like "there are two kinds of people" to be true then it would be true that there are two kinds of people interested in developing collaboration tools. Neither better nor worse, just different. Both are necessary and useful.    (POK)

One type is interested in enabling or augmenting the subtle interplay of people found in synchronous encounters, in synchronous settings as well as extended into asynchronous settings. These extroverts are the true and hopeful believers in collaborative action.    (POL)

The other type is more interested in augmenting the individual to allow them to manipulate information so it can be found, created and then distributed in a way that it can be manipulated by others. Introverts in an augmented dialectic.    (POM)

My own predilections, fears and interests place me in the second camp. For a long time I thought the main reason was because I didn't much like people and couldn't stand the dreadful noise, small packet size and high overhead of synchronous interaction.    (PON)

There's truth to that, but my experience with Phil suggests more: The internet as a whole and personal information tools that operate with it allow us to leap great spaces between disparate places and topics to draw and discover inferences that are like sparks of genius in a giant shared mind.    (POO)

Only a diversity of tools and a diversity of people can create the complexity in the technological and social network to both enable and ensure the distant leaps between hubs and echo chambers that signal the big ideas. Tools that are focused on the individual and their tasks and interests and publish to the universe simultaneously break down hegemony and synthesize the new groups and ideas that will be built from and broken down soon thereafter.    (POP)

Thanks Mr. Riehle.    (POQ)

Update: Phil has some additional comments.    (POR)

Posted by cdent at 08:18 AM | Trackback This | Comments (1) | TrackBack (3) | | bl

Blogs: Moving the Bait

SB received her copy the SLIS Network alumni magazine today. I suspect they've lost me somehow (off the hook for now, but see below).    (PIF)

It opens with a note from the Dean, Blaise Cronin, lovingly entitled: BLOG: see also Bathetically Ludicrous Online Gossip.    (PIG)

Blaise has a strong reputation as someone who loves to stir controversy. It helps him support his persistent belief that citation is a greater indicator of academic importance than content. Blaise is a brilliant strategist on a fishing trip so I find myself reluctant to write in response, but his last paragraph compels:    (PIH)

Under some circumstances, I might be willing to agree with Blaise. Ross reports back from Les Blogs:    (PIJ)

thoughts I've had myself on a rainy day.    (PIL)

But there are two things:    (PIM)

Blaise is the dean of a major site for information science education and research. He's showing a, um, bias here that's inappropriate for a place that should be pushing the boundaries of communication and knowledge enhancement forward, not holding them back. He's also playing political games with his own faculty that are too ridiculous to mention further.    (PIN)

And that bit about Wikipedia certainly brushes up against my chosen lines of work. Blaise speaks there as if he is the voice for all librarians. The librarians I know and cherish, of course, know better than Blaise.    (PIO)

Where I am in life today has a great deal to do with what I was able to milk out of my SLIS education. It came from my own assiduous exploration and the support of some very special people. It had very little to do with the policies and programs overseen by Blaise, and if he continues with these blinders the school will be unable to produce the sort of graduates the world needs.    (PIP)

Update: Blaise has posted a response to all the static he's received.    (PMH)

Posted by cdent at 12:56 AM | Trackback This | Comments (3) | TrackBack (0) | | bl

Why Wiki?

A regular unanswered question (for myself and others) of "what do you do?", some conversations with Lee LeFever about social design and community, and recent updates at work have given me a chance to think a bit about the question and stir some ideas.    (PG6)

At work I'm primarily a developer, but I tend to think of my vocation as a builder of augmenting, computer-based tools for personal and collaborative work. I go to the trouble of making that mouthful of a statement to distinguish between types of activities that computers do and types of activities that people do.    (PG7)

Computers have two types of applications, those that automate and those that augment (21Q, 22J). An augmenting application assists a person in performing some activity which cannot or is not automated. Only activities which can be formally described in theory can be automated. As such there are many tasks, especially those related to human discourse, which cannot be automated; they can, however, be augmented.    (PG8)

At a fundamental level computers are tools for creating representations (22L). An augmenting application supporting discourse is engaged in representing and transmitting information. The application is used as tool to evaluate, craft and remodel information (22K).    (PG9)

   (PGA)

Most of my development effort in the last few years has been with wikis (first PurpleWiki, and now Kwiki and Socialtext) and purple numbers. Purple numbers may eventually change the universe of information handling, but that's a discussion for another day.    (PGB)

Wikis are a type of augmenting discourse tool optimized for a particular set of behaviors. Under ideal(tm) conditions they provide an easy path to participation in evolving communication. They do this by being straightforward to learn, quick to respond, and accessible in a distributed fashion. They support changing content and provide an easy way to create and explore connections between things. How something fits in to the larger picture is a large part of how we infer meaning.    (PGC)

I think there are three primary audiences for wikis: the individual who hopes to use the wiki as an outboard brain or memory; the nascent group that hopes to discover and solidify the community that lies as potential in their loose connections; and the existing community that hopes to support a shared goal or perform some action.    (PGD)

Those three categories could be used to describe any set of people, but a wiki is not the perfect tool for every task. There are multiple types of discourse and multiple tools to support them. Some are better at certain aspects than others, none are really good enough (we have a long long way to go, but each day and in every way we are becoming better and better).    (PGE)

Blogs have become a central tool in the distribution of narrative discourse. With a blog there is usually a single author or small group expressing outwardly in a gesture that leads, over time, to the distribution of language and understanding outside the immediate clan. Very often the initial discourse is not fully refined but is rather some author's speculation: a seed that may lead to more knowledge later, as a separate piece of content. As has been said many times, the connections in the network of blogs is often loose and distributed.    (PGF)

Email continues to be a primary tool for discussion within a clan. The members of an email group have already discovered some bit of shared language or understanding that has brought them together. Email discussion can reinforce and solidify language, providing stability from which action can be performed.    (PGG)

With both blogs and email, content tends to be relatively static. Typos may be corrected in a blog entry and email threads may carry on forever but there is little in the way of refinement of the content. This is where wikis step in: they are good tools for summarizing, annotating and connecting information. These are the actions of a knowledge enhancement system.    (PGH)

Wikis do not match all the requirements for a knowledge enhancement system, but experience has demonstrated that this is good. Wikis are here now, today, helping people to do good work generating and supporting communities, developing and creating shared language, and refining information into new knowledge. Their simplicity makes them available.    (PGI)

   (PGJ)

When I chose to join up with Socialtext back in September, it was an attractive choice because the people there believe in two things: people matter more than tools; and tools should help people do what they want to do, not get in the way.    (PGK)

Socialtext, in its various incarnations, is based on wiki but integrates concepts from email and blogs to allow the action and narration those systems support. The latest release is a fine improvement: it enhances email integration, adds support for backlinks (placing information in context, leading to deeper understanding) and for PC Forum 2005 we've created a special prototype of Eventspace, running under mod perl for improved response time.    (PGL)

Architecting these sorts of tools may not solve poverty and hunger, or alleviate suffering in the aftermath of a ? disaster, but the tools can augment people actively doing that work. I happen to be good at making the tools go, so that's where I look to fit myself into the puzzle.    (PGM)

Related writings:    (PGN)

• The Computer As Tool    (PGO)
• Hypertext and Knowledge Enhancement    (PGP)
• Collaboration Requires Goals    (PGQ)
• Blue Oxen Vision    (PGR)
• Anarchic Emergent Collaboration    (PGS)
• A Web for Augmentation, not Automation    (PGT)
• Community and Collaboratory Tools    (PGU)

Posted by cdent at 05:12 PM | Trackback This | | bl

outboard info design goals

From a memorable conversation with Sunir, so I'm memorializing it here:    (PFJ)

Norman "feels" (rubs me funny) as if he thinks of learning as a process that is finite. That filling the cache should be quick and complete, and that the cache should be internal.    (PFK)

A well groomed wiki is an external cache, personally and organizationally.    (PFL)

One way to optimize understanding of the universe is not to cache the understanding of particular bits, but instead to cache references to the (internal or external) compressed or summarized learning event. Or, in other words, to communicate effectively it makes sense to make signs, symbols, words, language that point to concepts at increasingly larger levels of granularity (while never forgetting the importance of small pieces of lego).    (PFM)

By shoving some of the cached learning out to externalities there's more room for compare and contrast and new sign discovery.    (PFN)

That's very similar to how you describe Norman: speed up access. And that's nice, but it's not how Norman writes. Norman writes with too weak a dissatisfaction with the state of the world and too mundane a criticism of how people think and create. I guess that's why I don't like him: he presents little hope for large change. His most memorable discussions are about the burners on a stove. There's just so much more fun that can be had with discussion of interaction.    (PFO)

I don't believe in outboard brains, and I have no expectation that the AI dream will ever succeed, but I do know that if I have memories of handles to larger pieces that exist elsewhere, I can keep the allegories in motion in my head and know a _lot_ of stuff with a fairly high degree of confidence without needing to fully know it.    (PFP)

To bring this around to somewhere that might relate to what you're after: I've found that in order for outboard processing to work there's several design and process guidelines that have to be reached. Here are some: interaction must be highly responsive, noise in the interface must be minimized, structural mechanics and metaphors in content need to be consisent, names must have value, it must be there when you want it, when there is a shared brain its context is shared as well (e.g when some members of the company have a discussion about design it it is done in an archivable fashion).    (PFQ)

Posted by cdent at 09:57 AM | Trackback This | | bl

At Least Two Pieces of Good News about Views: Mountains and Purple

Two nice things have happened today:    (PEG)

On the walk back from the grocery I found a good place from which to get my eyes on Mt. Rainier. This has been a long running problem for me. First the weather is usually cloudy, so none of the mountains (Cascades or Olympics) are in view. Second, in the neighborhood, views to the east are fairly plentiful, but narrow. Rainier is to the Southeast, usually outside the angle of view while tromping around the streets.    (PEH)

Today things are different. Very clear day. And I found a good spot for a view. Somewhere I walk past often, but usually with my back facing the view. At the northwest corner of a nearby playfield if I site across the center line of the soccer field and look a bit right there's the mountain. And when I say there, I mean THERE. On a day like today it is stunningly huge. Nails me to the spot in some kind of religious ecstasy.    (PEI)

Meanwhile, back inside, where my view is the brick wall of next door, I've finally cut a usable version of Kwiki::Purple with support for good linking and internal to the wiki TransClusion. See it: http://www.burningchrome.com/pwiki/    (PEJ)

This is the culmination of a huge amount of work and experimentation, and I'm sure there will be much more to come. Thanks again to Brian Ingerson, Matthew O'Connor and Eugene Eric Kim.    (PEK)

I hope to get it to CPAN asap.    (PEL)

Posted by cdent at 10:27 PM | Trackback This | | bl

A Web for Augmentation, not Automation

Earl Mardle comments on Russell Beale's How the Web is Changing:    (8QZ)

I was going to write this as a comment on his blog and then decided to move it here.    (8R1)

In his posting and some others from around the same time, Earl seems to be bouncing around the difference between tools that augment and tools that automate.    (8R2)

Tools that augment extend a human's ability while still leaving the human in control. They are often small things (like purple number stuff) that provide a way to grab or manipulate stuff of all sorts.    (8R3)

Tools that automate "do it for you", often operate in large swaths, and are based on performing tasks that can be formally described.    (8R4)

There's an ethical or world view difference between the two. Some people prefer the latter, some people the former. I'm deep in the augmentation style, I hope. The Semantic Web, as imagined by the W3C, strikes me as in alignment with the latter.    (8R5)

Being in one camp or the other doesn't make someone wrong, but it does make conversation across the boundaries of the camps a little confusing and disorienting.    (8R6)

I first came upon the augmentation/automation split while writing my Computer as Tool paper:    (8R7)

Landauer's book:    (8R9)

is a good read.    (8RB)

Posted by cdent at 12:59 AM | Trackback This | | bl

Less Rules and Less People

Two things I wanted to remember from Acts of Volition:    (84M)

Following rules makes you stupid:    (84N)

There's something there about the difference between structures that constrain and those that support.    (84P)

The rules posting points to Garrity's Law of Inverse Congregational Intelligence:    (84Q)

I suppose that runs a bit contrary to the collaboration ethic I've claimed to buy at various times, but there is a great deal of evidence in both social and personal history to support Garrity's Law. I suspect the strength of ties is another variable as well as the strength of goals.    (84S)

Posted by cdent at 03:43 PM | Trackback This | | bl

PurpleWiki 0.9.2 Released

PurpleWiki version 0.9.2 has been released. See the ChangeLog for details on the changes. You can get the release either from CPAN or from the distribution page.    (4M8)

The parts that I like about this release are that it has:    (4M9)

• Much improved view drivers (the parts that take a parsed wiki page and turn into HTML or some other format) that make some forms of transclusion faster and more reliable.    (4MA)
• Support for remote NID and TransClusion handling. That is, if you configure things correctly, you can create content on one machine, and with some ease, transclude it on another.    (4MB)

Now I need to find some people who want play with remote TransClusion. Anyone?    (4MC)

Posted by cdent at 03:53 AM | Trackback This | | bl

My Patterns of Collaboration

EugeneEricKim, the other cofounder and remaining director of BlueOxen, has spent the last few days hosting a workshop on patterns of collaboration. I was unable to attend, but, along with several other people, watched from the sides through a wiki and email list.    (41H)

On the first day of the workshop, the participants developed a definition of collaboration (at Collab:Collaboration). Members of the email list were asked to introduce themselves and comment on the proceedings. I combined an introduction and a response to the definition in one mail message that I include here for the record. This version is edited to add some WikiWords. The original message is archived.    (41I)

This will be review for some of you, but it was a nice review for me.    (41J)

Hello, I'm ChrisDent. Sorry for this late message and sorry for missing the gathering. I had some timing conflicts and more importantly a distinct lack of funds.    (41K)

I'll begin with a bit of intro and then move on to some comments on the discovered definition of Collaboration (which I like). I've discovered this messasge is quite long. I seem to be using it for a bit of mental cleanout. Thanks for the opportunity and sorry for the length.    (41L)

I've been interacting (sometimes collaborating) with Eugene off and on in various capacities for about 3 years. We encountered one another in mailing list communities associated with Doug Engelbart's Bootstrap organization (sometimes known as Alliance, sometimes known as Institute).    (41M)

Around then, I had started a masters program in Information Science at IndianaUniversity. I had left a technical leadership role at a mid-sized ISP to fill up some of the holes in my brain. In my first class I was introduced to Engelbart, attracted to his ideas of augmentation, co-evolution and the necessity of collaborative effort to solve wicked problems. I started scrounging around for ways to know more.    (41N)

My first project of note in the IS program[1] laid the groundwork for a continuing sense that the foundation of a good collaborative toolset is the ability to access and reuse existing information. By access I don't mean find; I mean having a graspable handle and being aware of how things are being grasped.    (41O)

My second project[2] was based on some of Engelbart's ideas but explored them through the writings of other authors. I was trying to describe a productive "using" rather than "partnering" or "communicating" interaction with computers. I didn't quite hit it, and I've since discovered much more fodder in the notions of embodied and situated cognition combined with a bit of phenomenology[3] that await a book or PhD? thesis if I can find the steam.    (41P)

I came to the Bootstrap mailing lists with these things in my head. Eugene and I noticed each other as people who thought interesting things and often backed up our noodling with experiments or tools. Around the time Eugene was thinking about creating Blue Oxen, I was looking for a way to get some credits to finish up my degree without taking yet another boring class. We concocted an internship that turned me, with time, into a cofounder of the organization. Eugene and I had met in person only once.    (41Q)

Eugene and I spent several months fleshing out what's since become TheBlueOxenWay. The same semester I did the internship I was taking an extremely hard core practical class in software design patterns[4]. I suspect that class had some impact on the direction we chose. The class, as described in the referenced document, is one of the most significant collaborative events of my life. Another is the interactions I had with the team I worked with before going back to school.    (41R)

Both cases strongly support the idea that a shared goal is a very important part of successful collaboration.    (41S)

I never gained the traction with BlueOxen that I needed to feel successful. In part this was because I was spread too thin, with too many other obligations. I part this was because Eugene and I demonstrated another important aspect of collaboration: while we had a fairly robust shared language and to some degree a quite robust shared understanding, the details of our shared goals were not as well understood as they could have been. I, at least, found it difficult to get the necessary food out of the shared system.    (41T)

So I've since moved on to other things. I think this has been positive for us both. I hope Eugene agrees. We are exploring more diverse areas now th