January 13, 2007 (3): Decisions!

It all began when one of my colleagues and I parlayed the insights of knowledge management into an attempt to identify the role of communication and understanding in corporate decision-making. This began to take the form of a critique of decision support technology or, more properly, the ways in which that technology has now been appropriated by more recent enterprise software; and, with my tendency to always “push back to basics,” I found myself asking more general questions about the current state of decision theory and how consistent it is with decision practice.

There was a real irritant in asking these questions, however; and it seemed to lie in how “understanding,” as we were trying to study it, serves to “inform us to make informed decisions” (allowing for a circular turn of phrase). Well, for a start, what is it that we need to understand? The answer to that seems to be that we need to understand the situation (i.e. state of affairs, where I am choosing that word “state” because this involves taking a flow of events and relationships in which we are embedded and rendering it as something static) in which one or more actions need to be taken; and those actions need to be selected on the basis of one or more decisions. In other words, as any number of philosophers probably discovered long before I did, any theory of decision making must rest on a foundation of understanding the situation.

At this point I have to share some frustration with a book I am trying to read: Narrative Policy Analysis by Emery Roe. At the very least this book should wean me away from too much indulgence in postmodern deconstructionism; but, putting aside his stylistic preferences (obsessions), Roe is reluctant to come right out and say that he is trying to explore how narrative can serve explanation. This is not an unreasonable point to make; but for anyone of a philosophical bent it opens a real can of worms (if not several).

The most important can of worms is the very nature of explanation itself. I had A9 do an exact phrase search on “nature of explanation;” and the Amazon.com frame came up with 1355 books. My guess is that most of them have their roots in analytic philosophy, all sharing a positivist streak that can be traced back to Hempel (or someone who inspired Hempel). In my own intellectual development this is basically the pile of manure in the barn that convinces me that there must be a pony somewhere in the vicinity (but not in the pile of manure)! All that means, though, is that I am still looking for the pony.

However, finding the pony is only part of the story. Once you home in on the nature of explanation, you still have to worry about communicating an explanation to others. In other words you have to account for an explanation, and usually we do this by articulating it through text. This is another can of worms, since we now encounter all the subtleties associated with different text types and how we arrive at communication and understanding through those text types.

On the other hand, if we go back to Ancient Greece (well, if we go back to either Plato or Socrates as Plato tried to represent him), we discover that the two cans merge into one. The label on that can is the Greek noun λόγος, which means all sorts of things that, taken together, seem to orbit around the concept of an explanatory account rendered as text. (This is why “ology” is such a great suffix. The validity of the explanatory account is not necessarily part of the story. So “scientology” involves λόγος as much as “neurology” does!)

The good news is that, with all my experience in writing, I am never afraid of text. Furthermore, as my own blog header tries to proclaim, I do not try to hide from uncertainty or complexity behind fear, superstition, and/or pettiness. However, this may be an occasion to reflect on some wisdom of Richard Feynman:

In physics the truth is rarely perfectly clear, and that is certainly universally the case in human affairs. Hence, what is not surrounded by uncertainty cannot be the truth.

Thus, I am beginning to feel that just about any action we take (not to mention the decision-making preceding that action) cannot help but be “surrounded by uncertainty;” and all we can to do manage is to keep expressing ourselves through texts that we utter and/or write, knowing full well that they will never be clearer than our all-too-muddled thoughts! In other words the final say goes to one of Feynman's most honored predecessors, Niels Bohr:

Never express yourself more clearly than you think.

December 29, 2006: Organizational "Health Maintenance"

I would like to continue to explore the point I raised on December 26 to the effect that enterprises need "health maintenance," rather than "illness treatment." Obviously, if an organization is in a pathological state, then that pathology needs to be treated; but there are (at least) two aspects of health maintenance that are likely to facilitate treatment.

1. There is the need to cultivate an awareness of pathologies, since, as if often the case with human patients, it is often (frequently?) the case that the organization does not know when it is in a pathological state.
2. There is also the need to cultivate preventative measures: An organization that is aware of a potential pathology is better equipped to "nip that pathology in the bud," dealing with it before it has a serious impact.

So how do we implement such measures for an organization? In medicine we know about the physical examination and the frequency with which it should be performed (having now "graduated" to an age where I have to have mine annually). However, this time scale is to coarse for the world of business. An enterprise functions on the basis of ongoing business processes, and those processes need to be monitored in the course of their functioning. IT is now in a position the make sure that such processes are running reliably and accurately; so this is the first step towards making the health maintenance metaphor "work."

However, such monitoring cannot be conducted for its own sake. The data generated by monitoring drives the formulation of hypotheses that address detecting, diagnosing, and treating pathologies. Some rather interesting visualization technology has gone into "driving" dashboard displays that deliver monitoring data; but the hypothesis data generated by managers at all levels of the enterprise are equally important to health maintenance. When a hypothesis is posed, it will (almost?) always need to be vetted by several managers throughout the enterprise; so we need some kind of "dashboard visualization" to provide awareness of the current "state of the hypothesis space" and facilitate the vetting process.

How can all of this take place? In his book Images of Organization, Gareth Morgan addressed this question by bringing another metaphor into play:

The formal analysis and diagnosis of organizations, like the process of reading, always rests in applying some kind of theory to the situation being considered.

In other words the real purpose of visualization, of both monitoring data and hypothesis spaces, is to enable managers to read their organization, a concept that inspired the subtitle of James Taylor's fascinating book on organizational communication. I would further argue, to draw upon another previous discussion, that we are talking here about a literary approach to reading, rather than any kind of objective or analytic one. In the context of another favorite theme, managers should be able to think about the people in an organization as subjects rather than objects. The danger of visualization is that it serves up abstractions that entail excessive objectification, but this danger is more likely to add to any looming pathologies than to deal with them.

This then leads to my final point, which is the recurring question of whether or not tools such as Wikipedia and Google are eroding our skills for being critical readers, whether of texts or of organizations. We are more interested in tools that deliver answers, and reading is not necessarily about delivering answers. As I previously suggested, sometime one has to go "roundabout" the questions; and literary reading enables such "roundabout" thinking. This may take more time that finding an answer from a search result or a Wikipedia entry; but, where the health of an organization is concerned, getting to the most effective answer is more important than getting there "efficiently."

December 27, 2006: Identity Meets the Cluetrain Manifesto

What is Hecuba to you or you to Hecuba? This injunction (shameless appropriated from Shakespeare's Hamlet) applies not only to actors but to all of us in the “roles” we play (our “presentation of self,” as Goffman called it) in everyday life. However, there is also a grander scheme of things that I introduced some time ago. This involves what I feel is the most important lesson from Plato’s “Theaetetus” dialogue. This is that, while none of the four definitions of knowledge considered can survive Socrates' critical examination, Socrates does demonstrate how the concept of knowledge is tightly coupled to three other key concepts: memory, being, and description (λόγος).

I would now like to pick up where my last discussion left off and examine the concept of “Being.” I would argue that this concept needs to be sorted into the being of objects and the being of subjects, i.e., the agents who engage with objects; and this latter class of being has to do with identity. In other words any inquiry into the nature of identity involves pulling at a thread that is tightly woven to many other critical threads, including the thread of knowledge itself!

These couplings have been particularly well appreciated by George Herbert Mead, particularly in his exploration of the concept of symbolic interactionism. I like to say that the motto of symbolic interactionism is: “No perception without personal interaction.” For my money this is the underlying premise without which the assertion that “markets are conversations” cannot make any sense. Indeed, it is also the premise behind Habermas’ theory of communicative action, which argues that “communicative action” (as Habermas defines it) is the fundamental prerequisite for understanding. Since it seems valid to assume that markets can only operate effectively within a context of understanding between buyers and sellers, Habermas’ theory ultimately explains why markets are conversations; but I am afraid that this kind of foundational thinking has gotten lost amid the 95 theses of the Cluetrain manifesto!

September 19, 2006: Did Kleist Anticipate Sokal?

To call Heinrich von Kleist a strange writer is to mince around with understatement. There are too many examples of his fiction and drama that leave you thinking, "I can't believe I'm reading/seeing this!" So one has to wonder whether or not one should react the same way to some of his essays. He could have perpetrated the sort of hoax that, in our time, Alan Sokal pulled off in publishing "Transgressing the Boundaries: Towards a Transformative Hermeneutics of Quantum Gravity" in the journal Social Text. As we now know, Sokal tried to "publish an article liberally salted with nonsense if (a) it sounded good and (b) it flattered the editors' ideological preconceptions" and succeeded. Two hundred years earlier, Kleist published some "learned essays" on topics sufficiently odd for his time to leave the reader wondering whether or not to take him seriously.

My favorite is "On the Gradual Fabrication of Thoughts While Speaking." The basic thesis is that the thoughts we have "in our head" are, for the most part, vague and ill-formed; and it is only when we try to express them in words that they become concrete enough to be called ideas. (At this point I should not that the fact that Kleist had a thesis at all makes him more credible than Sokal. The editors of Social Text apparently did not require him to have a thesis statement, so he did not provide them with one: He wrote no abstract and avoided anything that looked like a thesis sentence in both his opening and closing paragraphs. Kleist may have decided to explore a weird thesis, but at least he told you where he was going!) After stating this thesis, he backs it up with a series of examples and concludes by reaffirming that thesis. In that respect it is a model essay.

As a matter of fact, I think everyone should read it, not only because it is a model essay but also because I think his examples do support this thesis! Why not? In his "Theaetetus" Plato has Socrates lead Theaetetus to explore the definition of "knowledge." Theaetetus proposes four definitions, each of which is deftly refuted by Socrates, leaving us at the end of the dialog with no definition for knowledge! Perhaps we have such muddled ideas about "thoughts in the head" that Kleist may be right: Those thoughts really are not there and only take shape when we try to express them. In Kleist's own more poetic language, "it is not we who 'know'; it is rather a certain condition, in which we happen to be, that 'knows.'"

So, the next time you encounter an article that does have an authentic thesis, rather than a mere concatenation of meaningless jargon, no matter how absurd it may seem, accord it the respect of a serious claim. Then see if the author can back up the claim with an argument, making sure that the argument is not pulling off any rhetorical sleight-of-hand. If you can drag the claim through that test and it emerges intact on the other side, it may be less absurd than you thought!

September 05, 2006 (2): My Life and Hard Times in Requirements Analysis

From 1978 to 1981 I worked at General Research Corporation on problems of requirements analysis for the Army. I had taught programming methodology at the University of Pennsylvania at a time when people were beginning to use the term “software engineering” instead of “programming.” The focus on methodology grew out of the recognition that the design and construction of computer programs should be as much an engineering discipline as the design and construction of bridges or electronic devices. Engineering was about building something for someone else’s benefit; and that third party needed some confidence that the resulting artifact would do what it was supposed to do, reliably and safely. How would a software engineer determine what that artifact was supposed to do? It was assumed that the party for whom the program was being written would initiate the engineering process with some set of requirements, and software engineering methodologies were concerned with how one could effectively and productively proceed from those requirements to a reliable program that would satisfy the beneficiary.

Unfortunately, the real world has never been particularly sympathetic to abstract engineering methodologies; and, where software is concerned, the problem tends to start right with that beneficiary. How does anyone, the beneficiary or any programmer, know whether or not the requirements provided by the beneficiary are an accurate representation of what that beneficiary really wants? The history of software engineering has been plagued with case studies that begin with the sad truth that requirements are actually a very poor representation, particularly if the beneficiary does not have a clear idea of what the program is supposed to do; and the lack of a clear idea is often a product of the unholy alliance of a beneficiary who does not appreciate what it is reasonable to expect and a software engineer eager to promise the world.

Thus, it is not enough for the Department of Defense to want to protect the United States from missiles with nuclear weapons on their warheads. While that is an understandable request, it is not a requirement that can be translated into the design of a computer program. Knowledge engineering operated under a similar illusion: One cannot simply identify an expert, say in the area of medical diagnosis, and “engineer” a “representation” of the “knowledge” behind that expertise. When I worked at Schlumberger, I had the good fortune to work alongside some of the best minds who knew how to interpret the complex and obscure measurements taken in the boreholes of potential oil wells; but I quickly learned that they did not live in a world of requirements and specifications, which could then be converted into computer software.

Fortunately, I discovered, essentially by accident, an alternative approach. Rather than working among my colleagues in the software systems group, I asked to have an office in the wing in which all these experts worked. Using a Xerox Lisp Machine, I would prototype programs to interpret test cases of these measurements, yielding displays of both the measurements and the interpretations. This became an excellent conversation-starter. An expert would wander past my office, see the display and come in for a closer look. Examining both the data and the results, the reaction would almost always be the same: “Why did your computer do a damn fool thing like that?” My reply was also almost always the same: “Why is that a damn fool thing? I just did what the training manuals said!” This would start an extended lecture on why what you did in the real world had nothing to do with what the training manuals taught you; and, enlightened by that lecture, I could go back to work on the program. Essentially, I had discovered a user-centered, evolutionary approach to identifying and satisfying requirements that reflected what the beneficiary really wanted. Since those days I have discovered that I am far from alone in appreciating this methodology, but my opinions are still very much in the minority.

In retrospect, however, I can appreciate why this methodology has not caught on in the software engineering community; and the reason can be traced back to that distinction, first raised in Plato’s “Republic,” between lexis (word) and praxis (act), which has become one of my favorite topics. The idea that requirements can be “represented” at all presumes the construction of an artifact that is basically a lexis structure, even if the “words” are elements of a formal language, rather than a natural one. At Schlumberger, however, I came to understand the nature of requirements by becoming familiar with the praxis of the experts who could come into my office and make fun of what my prototype programs were doing. Unfortunately, documentation is the bread and butter of engineering methods, whether they are based on natural language or some combination of formal representational systems (blueprints, flow charts, algebraic specifications, etc.). There is no place for praxis in such documentation. Indeed, as I have already observed, John Seely Brown and Paul Duguid have gone so far as to call those documents “Abstractions detached from practice [that] distort or obscure intricacies of that practice.”

The problem, which has been a recurring theme in this blog, is that lexis structures are static, which makes them conducive to representations, which, in turn, are conducive to analysis. Praxis, on the other hand, is, by its very nature, dynamic, making it particularly elusive to most methods of analysis. However, rather than ignoring praxis because it eludes our methods, we should be seeking out alternative methods that, to invoke the language of Richard Neustadt and Ernest May, facilitate our ability to “think in time.” This is a lesson that still continue to grow on me.

September 05, 2006 (1): The Bøyg

A variety of different paths have been leading me to think about the Bøyg. This is an episode of a fight in the Seventh Scene of the Second Act of Ibsen's Peer Gynt play (which does not seem to get performed very much, probably become of the overwhelming demands it makes for production). (Grieg wrote music for the Bøyg scene, but it was not included in either of the suites he later compiled.) According to the script, the Bøyg should be realized as a disembodied voice. In my translation, Peer describes it as follows:

Not dead, not living; all slimy, misty. Not so much as a shape! It's as bad as to battle in a cluster of snarling, half-awakened bears!

As a matter of fact, when Peer tries to lunge at the Bøyg, he succeeds only in cutting himself.

One of the paths that led me to the Bøyg is the current Confused of Calcutta discussion about Wikipedia and "Brickipedia" (i.e. the traditional multivolume print-version of an encyclopedia). I did a free association of Wikipedia with the shapelessness of the Bøyg. (By the way, Wikipedia does not currently have a "Bøyg" entry!) I realize that some may argue with Peer's description, at least to the extent that Wikipedia is living, in contrast to any "dead" printed artifact; but, to my own critical eye, Wikipedia seems to have certain zombie-like qualities (although I would not accuse it of being slimy)!

The important lesson from this scene, though, is the Bøyg's advice to Peer: "Go roundabout!" Peer wounds himself when he tries to confront the Bøyg head-on, ignoring this advice. After that, he is more inclined to listen; and this advice turns out to impact how the rest of the story evolves. My point here is that, where information is concerned, this can also be an important piece of advice. A direct path taken through Wikipedia or Google may take us quickly to the answer to a question; but the context-free nature of that answer may end up cutting us (a bit like the joke about the Swiss Army Knife that has a tool for anything you want to do: the problem is that, when you try to get the tool, you cut yourself on one of the blades)! If there are still people out there trying to distinguish the difference between "information" and "knowledge," I would offer up the claim that, because knowledge is always intimately tied to context, we can only arrive at it by going "roundabout!"

August 14, 2006: "Du Descriptif"

I have chosen to appropriate today's title from Philippe Hamon's treatise on the text type of description. (To the best of my knowledge, this book has not been translated into English.) I see that the current Wikipedia entry for text types is still in stub form. Perhaps I shall do something about that, but not right now. What Wikipedia does say is that description is one of four text types, the other three being narrative, exposition, and argumentation. The path to my current concern with description is a bit circuitous, so bear with me.

Two of my recent entries (August 9 and August 11) were inspired by my reading the survey paper, "Emergence of Communication Networks," by Peter R. Monge and Noshir S. Contractor and written for the second edition of the Handbook of Organizational Communication. (Yes, I know I said this on August 11!) While it should be apparent that this paper triggered a variety of thoughts while I was reading it, I still have to confess that the act of reading, itself, was quite a slog, to say the least. By the time I reached the "Conclusions" section and was informed that "additional work is required to reduce or eliminate the extensive redundancy that exists among the different theoretical perspectives," I felt like screaming at the authors, "Isn't that what a survey paper is supposed to do?" In ignoring the very possibility of organizing their content to reduce, if not eliminate, all that redundancy in the existing literature, the author's reduced their paper to yet another instance of Winston Churchill's definition of the writing of history: "One damned thing after another."

This could be an excellent excuse for me to get back on my hobby-horse about the deleterious effects of excessive specialization in education, the corrupting force of techno-centrism on today's writing, or just a rant on why writing is more important to an academic career than whether or not what one writes ever gets read. Instead, however, I would prefer to address the question of why survey papers are such formidable intellectual challenges, so much so that it is rare that we run across a really good one. In a way the survey paper is the magnum opus of the descriptive text type, just as the novel is the magnum opus of the narrative text type, the topic-based essay is the magnum opus of the expository text type, and, I suppose, the hard-clad forensic text is the magnum opus of argumentation. The challenge is so great that it is a wonder that surveys get written at all, except that the demand for them has been so heavy in the past. The result is that almost none of them count for exemplars of good descriptive writing.

This is more than a little sad, particularly when we take into account to ancestry of the descriptive text. After all, if we want good examples, we can go as far back as the Homeric epics to find passages such as the description of the shield of Achilles; so we know that good descriptions have been around at least as long as good narratives. So is this just another case of our losing our touch as writers, of those authors who once knew better words now using only four-letter words? (Anything goes, indeed!)

I would argue that this problem is a bit more complicated; and now I have to take a jab a techno-centrism! See, in a world that is now driven by all the databases that store our "knowledge" (those are scare quotes, by the way), we tend to be deceived into flipping the coin around and thinking that knowledge is what we store in databases. So the object of descriptive writing is just that: an object. As database designers will tell you, an object is described in terms of its attributes and its relations to other objects. This is a nice, neat abstraction, so nice and neat that it often can be most conveniently illustrated by a directed graph; and, invoking language that Donald Knuth introduced into our vocabulary, one can then describe the object by traversing the graph that represents it. In other words, if you can represent your description in the abstract form of one of these directed graph, then you should be able to formulate a text description by stringing together smaller textual units for the nodes and edges of that graph. Of course when you do that you get exactly what aggravated Churchill: one damned thing after another. It usually does the reader a greater service to just illustrate the graph and keep the text to a bare minimum.

Needless to say, such a description is not a literary composition; but the prevailing counter-argument is that people who read survey papers are not reading them for their literary value. Granting that, we should still honor the fact that those people are still readers. Even if you are not expecting the quality of Lattimore's translation of the description of the shield of Achilles, shouldn't you expect to take some satisfaction in your act of reading? I suspect that the answer is, "Not any more;" and this, too, has a reason.

One way to view a survey paper is as a reference resource; and we no longer seem to expect literary quality when reading reference resources (even if this used to be a sign of the best of them). This gets me back to the problems I have with invoking the World Wide Web as the ultimate reference resource, be it for Wikipedia, Google, or even that wonderful composite, A9. The thing is that we do not really read these resources. We consult them, often in a highly fragmented manner. (How many times to you get your answer from Google in the excerpt from the hit, rather than actually linking to the hit itself?) We have come to believe that this is all we need; and, if a survey paper is a dry traversal of "one damned thing after another" that happens to be on-line, then Google will find the paper and then the find command will let us find the fragment of text we need. Why should the old-fashioned approach to reading be of any use any more?

One good reason is that the above scenario works best for experts. The people who still have a need for enough literary quality to make the act of reading worth while are the ones trying to come up to speed on a new topic. They often have a need for expository and descriptive texts that really are worth reading (and may be so through the skilled incorporation of argumentation and narrative). This is why "Dummies" guides are actually some of the best examples of writing currently on the market. The people who edit those series realize that their readers need to be provided with reading matter that is as good as the content being covered by the text. Nevertheless, I cannot help but fear that even these resources will eventually get sucked into (ensnared by?) the World Wide Web and lose that quality of writing style that made them so successful in their past. Then, unfortunately, we will have little more than Wikipedia and Google; and I cannot help but believe that our knowledge will suffer for it.

August 11, 2006: Normal (?) Science

One of the interesting ideas to come out of the "knowledge movement" was the attempt to view an business organization as a "knowledge structure." One of the champions of this position was Bruce Kogut, who, together with his colleagues Shan and Walker, published a paper in 1993 with the intriguing title, "Knowledge in the network and the network as knowledge: Structuring of new industries." In a recent survey paper on communication networks, Peter Monge and Noshir Contractor summarized one of Kogut's key insights:

Once organizations choose partners, however, they tend to spend less time seeking other partners. As Kogut, Shan, and Walker (1993) say, "because information is determined by previous relations and in turn influences the subsequent propensity to do more relations, the structure of the network tends to replicate itself over time. The early history of cooperation tends to lock in subsequent cooperation" (p. 70). Further, they observe that "The replication of the network is a statement of the tendency of learning to decline with time. The structure of the network is a limiting constraint on how much new learning can be achieved. ...But when viewed from the perspective of the evolution of networks, there is a tendency for old lessons to be retaught" (p. 71).

This got me to thinking about the concept of "normal science" and the role it played in Thomas Kuhn's study, The Structure of Scientific Revolutions. Here is how Kuhn characterized the concept:

In this essay, ‘normal science’ means research firmly based upon one or more past scientific achievements, achievements that some particular scientific community acknowledges for a time as supplying the foundation for its further practice. Today such achievements are recounted, though seldom in their original form, by science textbooks, elementary and advanced. These textbooks expound the body of accepted theory, illustrate many or all of its successful applications, and compare these applications with exemplary observations and experiments. Before such books became popular early in the nineteenth century and until even more recently in the newly matured sciences, many of the famous classics of science, by such writers as Aristotle, Newton and Ben Franklin, served for a time to define the legitimate problems and methods of a research field for succeeding generations of practitioners.

It strikes me that Kogut's characterization of an organizational network as knowledge bears a strong family resemblance to Kuhn's characterization of normal science. The kind of knowledge that Kogut seems to have in mind is knowledge of "how things are supposed to be done" in an organization; and one strategy for maintaining that organization is to replicate that knowledge. In a similar way scientific communities may be viewed as organizations; and, to the extent that any community is based on the need to distinguish a collective "self" from an "other," normal science defines the community, not by its scientific results, but by the implicit knowledge of how things are supposed to be done. Straying from the path entails leaving (or being banished from) the community. Abandoning normative standards entails abandoning the recognition of one's activities as legitimate.

This returns us to the theme of order and chaos. Organizations, both industrial and scientific, depend heavily on a sense of identity in order to "get things done." However, we are now confronted with the question of how one establishes identity (which implies some sense of stability) in a context that is becoming increasingly chaotic. Normal science, in a sense, denies that there is chaos out there (or, to invoke Henry Miller's language, denies that there is no confusion that cannot ultimately be understood). Romanticism challenges this premise and, in so doing, challenges a normative approach to science (which is why Kuhn introduces the concept of scientific revolution in the first place) and may also challenge a normative approach to business organization.

Meeting that challenge may involve returning to the theme of the opposition of faith and interpretation. At the end of the day, the decision to accept that the conduct of science or business management rests on normative standards is an act of faith. One cannot invoke scientific method to justify the conduct of normal science. Indeed, much of the reaction against Kuhn had to do with the fact that the concept of a scientific revolution could be seen as a refutation of the validity of the normative conduct of normal science, thus challenging the prevailing faith of scientists at the time Kuhn's book appeared. Similarly, our commitment to normative standards of business management is also an act of faith; and all it takes are intimations of refutation to remind us how strong that faith is in the established community of managers.

In posing interpretation as an opposition to faith, I argue that anything we confront in the world always needs to be subject to interpretation; and, because the context is always changing, we cannot assume that any interpretation is stable or enduring. All we can do is hone our interpretive skills and continue to exercise them as a safeguard against atrophy. There is no guarantee that this will solve any major problems like unemployment or war, but it should at least keep us from sleepwalking through the world that others want to make for us!

August 09, 2006: The Sense of Identity on the Other Side of the Chaos Coin

I prepared the above diagram after reading Plato's "Theaetetus." This dialogue is often invoked by the knowledge management crowd for its definition of knowledge as justified true belief. However, if you go to the source itself, you discover that, in the course of the dialogue, definitions of knowledge are considered (justified true belief being the last of these); and all of them are found wanting. The dialogue concludes by accepting (almost in the Kubler-Ross sense) the fact that Socrates and Theaetetus have failed in their attempt to define knowledge. What Socrates manages to demonstrate, though (although he never says this explicitly), is that the difficulty of defining knowledge may have to do with how tightly coupled it is to at least three other equally fundamental concepts: memory, being, and description (which may be a poor translation of the Greek λόγος).

I raise this summary because I think that the concept of being (along with the related concept of identity) must not be neglected in considering the strategy of embracing chaos that I discussed yesterday. This occurred to me while reading some of the business literature invoking terms such as "network organizations," "boundary-less organizations," and "virtual organizations." I realized that, from the point of view of actually getting the work done, this kind of language can raise some challenging questions of identity, for both individual workers and the "entities" (who knows what we want to call them any more) that, at least nominally, serve as their employers. As I suggested yesterday, this issue of identity is critical, because it is only through our sense of self that we can engage our own subjective skills of perceptual categorization, which are all we have in order to manage ourselves in that chaotic setting that we may be obliged to embrace. To some extent Social Psychology has recognized this with its sub-discipline of Identity Theory and the roots of that sub-discipline in George Herbert Mead's symbolic interactionism. In that respect it should be no surprise that Habermas' work on his theory of communicative action draws heavily on Mead's symbolic interactionism. However, what matters at the end of the day is how the issue of identity surfaces when each of us faces the prospect of going to work every day or confronts the fate of having lost one's job. Pop psychology has trivialize the phrase "identity crisis;" but it may be that one of the most critical consequences of the Internet has been the cultivation of an identity crisis of global proportions. As I have already suggested, these consequences have economic, social, and political implications, all of which have received little more that gratuitous lip service as a price for the worship of innovation. However, an emerging identity crisis may be an instance of what Malcolm X had called the chickens coming home to roost!

July 23, 2006: "The Irrational Foundation of Rational Knowledge"

I had forgotten that much of my interest in irrationality may have originated in my reading of Karl Mannheim's Ideology and Utopia: An Introduction to the Sociology of Knowledge; so I decided to take today's commonplace book entry from that source:

As a rule, human thought is not motivated by a contemplative impulse since it requires a volitional and emotional-unconscious undercurrent to assure the continuous orientation for knowledge in group life. Precisely because knowing is fundamentally collective knowing (the thought of the lone individual is only a special instance and a recent development), it presupposes a community of knowing which grows primarily out of a community of experiencing prepared for in the subconscious. However, once the fact has been perceived that the largest part of thought is erected upon a basis of collective actions, one is impelled to recognize the force of the collective unconscious. The full emergence of the sociological point of view regarding knowledge inevitably carries with it the gradual uncovering of the irrational foundation of rational knowledge.

June 25, 2006: Service is a Science, or is it?

Thinking back on yesterday's comments about nostalgia, I have decided that not all of the agonies of Hell involve happy memories. One can also be tortured by unpleasant memories, particularly when you feel you could have taken some action (or just said something) to change them. (You might call these the agonies of "coulda-shoulda-woulda.") I had one of those hauntings this morning, so now I have to get it off my chest.

This is a memory from March 24, 2004 and the Almaden Institute I attended that day. The topic was "Work in the era of the global, extensible enterprise;" but the focus was on work in the service sector. The tone was set by the welcoming remarks, which, as I recall, were delivered by Paul Horn, Senior Vice President of IBM Research. The main thing I remember from those remarks can be roughly paraphrased (if not directly quoted) as follows:

Fifty years ago IBM partnered with Columbia University to invent the academic discipline of computer science; today I am announcing that we are partnering with the University of California at Berkeley to invent the discipline of service science.

This was my first unpleasant experience of the day. I had always felt (courtesy of one of the many citations dug up by Knuth) that the foundations of computer science were laid in a series of lectures in the summer of 1946 at the University of Pennsylvania, after their pioneering work in digital computing (the Eniac) for the military during the Second World War. I have to confess to a strong personal bias behind this belief, since the first lecture was delivered on July 8, 1946, which happens to be the day I was born! However, I also remember Edsger Dijkstra's Turing Award speech, in which he talked about those people who can only understand concepts when delivered in their own terminology. The examples he gave were Catholics and Communists; but, when history is involved, I have discovered there are plenty of other examples!

The real memory-agony, however, dealt with the whole idea of their being a science of service, particularly since one of my activities that day involved a break-out session to discuss the curriculum for that new academic discipline. To this day I believe that our session went absolutely nowhere, because we could not think of any area of study that could be excluded from the curriculum; and so it is that I am haunted. Fortunately, I think Hegel has been helping me see some light. My entry about putting philosophy in its place can actually be boiled down to deciding whether the objective of the study of philosophy is descriptive or prescriptive. Hegel came down strongly on the former option, arguing that those descriptions should be as clear as possible, since they depicted "reality as it is." As a result I would extend Hegel and argue that, just as philosophy is a descriptive discipline, science is a prescriptive one, concerned primarily with understanding enough about the natural world to exercise control over it. So, in the tradition of answering one question with another question, before asking if there is a science of service, we might consider the extent to which the study of service can lead to prescriptive results.

Here I would like to cite some of the work of Orion White, a sociologist who was studying service organizations before their operations had been influenced by digital technology. He drew the distinction that a service organization is dialectical, rather than bureaucratic (Public Administration Review, 1969). In more contemporary language we could say that the operations of a service organization cannot be reduced to business processes, which may then be "engineered" for optimal efficiency. Instead, operations are all about recognizing underlying oppositions and resolving them synthetically, rather than coming down on one side or the other as "the way it has to be." I would then argue that, while it may be possible for a bureaucratic organization, one cannot take a prescriptive approach to the operations of a dialectical organization. One can only build up a repertoire of keen descriptive skills, through which one can grasp the reality of the entire situation, and then, with a Habermas-like ambition, apply that grasp to achieving an understanding that is social and subjective, rather than strictly objective. Since science only "plays" in the objective world, the very concept of a service science entails an impoverished view of the world.

What, then, is the primary descriptive goal in the study of service? I would argue that it involves the nature of value and how value is perceived by individuals in their social settings. In this respect one could do worse than begin with the study of the production of value that Marx began to develop at great length in his Grundrisse (even if much of that work went through a major overhaul by the time he was writing Capital). Marx was well aware that the production of value was not strictly a matter of the commodities that enter a production process in one form and leave in another. Much of Marx' attention focused on the consequences of viewing labor, itself, as a commodity; but Habermas expanded his own scope beyond the basic operation and operators of the production line. He presented evidence "that we consider whether the work incorporated in rationalization must not be understood and evaluated as second-order productive labor—as an independent one, because it depends on productive labor of the first order." (In many ways this is Habermas' first recognition of the role that understanding plays in work practices.) Thus, any understanding of the nature of value has to account for the ways in which knowledge contributes to value (an issue that was given lip service during the heady days of the "knowledge movement" but never subjected to the standards of descriptive discipline that Hegel had set for philosophy).

So am I just playing out a pun that the study of service should be grounded in the Grundrisse? Am I trying to send a be-careful-what-you-wish-for message to IBM, because the path to that wish may lead through Marx? No, I am just content to play Hegel's game and do my part towards the depiction of "reality as it is," however, complex that reality may be!

June 23, 2006: Hegel Puts Philosophy in its Place

While reading Habermas' Theory and Practice I came upon his summary of an observation that Hegel made in the preface to the Philosophy of Right: "Philosophy cannot instruct the world about what it ought to be; it is solely reality which is reflected in its concepts, reality as it is."

June 15, 2006: On Intelligence?

I have to begin by saying that I am always skeptical when anyone, even a close friend, tells me I absolutely must read a book. This is the way I was introduced to On Intelligence, by Jeff Hawkins (with the assistance of Sandra Blakeslee); and I am writing this to give some sense of the way I felt when I checked the book out of the San Francisco Public Library. While I do not always judge a book on the basis of the most immediate first impressions, the "Prologue" had such an impact that I figured I had better set down my thoughts while they are fresh. My most important observation is that I think I have already read the punch line (the memory-prediction model, or at least something with a strong family resemblance to it). I had to do a bit of digging; but I am pretty sure I came across this story in Karl Pribram's Languages of the Brain, which I read when I was teaching at the University of Pennsylvania. This book came out in 1971. On the one hand I am surprised that Hawkins did not find it in his local library, since the timing would have been about right; and I remember Pribram being the first author I encountered who tried to write with clarity about how the brain worked. (A major theme in Hawkins' "Prologue" is that he decided to write this book because he has been looking for a book about how they brain works ever since his youth.) On the other hand I discovered that the book was neither in the San Francisco Public Library or in any of the on-line collections I managed to find with the help of My Web 2.0. I have not kept up with my reading of Pribram to determine whether his more recent stuff still holds to the same model; but the appearance of "holonomy" in the subtitle of Brain and Perception leads me to believe that his basic theme is still there.

So I am not going to fault Hawkins for being unaware of Pribram, even if I find it a bit strange that he did not encounter the material over the last 25 years. (Pribram seems to have been on Howard Gardner's radar.) Where I am more skeptical is what I would call the extreme reductionism of Hawkins' position that essentially abstracts out anything that has to do with behavior. The result is that "it's all about the brain;" and that brain may as well be in a vat, rather than integrated into a "whole body" system. (Having made that extreme reduction, I can then appreciate that he takes it one more step to "it's all about the neocortex.") So he writes off the body as relevant context for the brain, which means that he also writes off the social context of the body.

None of this means that I shall dismiss the book. I intend to take advantage of the fact that I have it for at least three weeks. However, both my enthusiasm and my expectations have diminished. I am still willing to believe that there is value in what Hawkins has achieved, but the framework may be narrower than the title leads one to believe. This, I fear, is a standard symptom of techno-centric thinking: Carve off the slice of the problem that you think you can work and then posit it to be the whole ball of wax!

June 11, 2006: Hermeneutics 101

Having provided introductory outlines to phenomenology and Habermas' theory of communicative action, I figure I ought to do the same for hermeneutics, since I feel that this is another non-technical discipline that offers some critical "survival skills" for the Internet age. I must confess that I was led into this one by a conversation I had with a friend at PARC who made the pronouncement that it was time for us to stop wasting time on concepts that really had nothing to contribute, like hermeneutics. So if this particular introduction reads something like an apology, then that can be attributed to my personal past history!

Actually, one of the difficulties in introducing the concept of hermeneutics is that it seems to have meant many different things to different people. Therefore, I should being by stating up front that my own primary influence comes from the writings of Paul Ricœur, whose book on metaphor I discovered on a business trip to Basel and has had me hooked ever since. What has most interested me about Ricœur is that he took the traditional view of hermeneutics, as the study of methods for interpreting texts, and generalized it to the interpretation of meaningful actions. This puts him on the same field of interest as Habermas, who has had his own prolonged debate with Gadamer over hermeneutics. John Thompson's Critical Hermeneutics focuses explicitly in the intellectual relationship between Ricœur's phenomenological stance and Habermas' foundation in social theory (anchoring them both in Wittgenstein's ordinary language philosophy). This is what I mean when I say that any attempt to get a sense of the nature of hermeneutics quickly leads you into a vast Web of opposing opinions!

Ricœur's approach is basically a dialectical one. He begins with the exploration of two opposing concepts: explanation and understanding. From his point of view, "explanation" means "structural explanation," a syntactic perspective that can be generalized beyond the sentence to the broader scale of a paragraph or an entire work (if not an entire corpus). "Understanding," on the other hand, is concerned with interpretation and thus involves a semantic perspective. The opposition comes from the fact that, often, you cannot determine the syntactic structure of the text unless you know what it is trying to say; and you cannot interpret the text without knowing how it has been structured. (This was discovered "in the trenches" when Terry Winograd was doing his pioneering doctoral research on a computer system that understood natural language.) Ricœur proposed that hermenutic interpretation could only succeed through a Hegelian (Fichtean?) style synthesis of these two concepts, dealing with explanation and understanding as interrelated processes, each of which informs the other (which is one of the most important features of Winograd's SHRDLU system). This is why I clipped in the above diagram (which I discovered through an A9 image search, back when A9 was still using Google for image searches).

Why is this important? The primary reason is that computer science (or information technology) still tends to assume that intepretation is concerned strictly with the denotation of symbol structures, which is fine if you are writing a compiler or a system that interprets database queries but (as Winograd discovered) starts to break down when you more from formal languages to natural ones. So this takes us back to the debate over what we can realistically expect from the Semantic Web and whether or not we can call it "semantic" in any way that makes sense the the real world of how we all use the Internet (i.e. the world in which the Internet is used by ordinary people trying to get things done, for business or pleasure, as opposed to computer science students or professionals).

It is nice to see that Ricœur seems to have had a healthy appreciation of the pioneering work of C. S. Peirce on the interpretation of symbol structures. It is nice to reflect that, in one of his early essays ("Some Consequences of the Four Incapacities"), Peirce came up with the wonderful observation that "men and words reciprocally educate each other." I think this gets to the heart of the synthesis upon which Ricœur's approach to hermeneutics is based, and I think we can do worse than live by that insight.

June 10, 2006: No Ontology without Phenomenology!

Since no one picked up the gauntlet I threw down yesterday on "Between the Lines," I would like to use this space to clarify why some might view me as an ontology-basher. This is actually material that I had prepared for my talk in Ireland that I mentioned in my June 5 entry; but it probably deserves more attention as the world (or some fraction thereof) lines up to ride the Semantic Web hobby-horse. This is also a good place for me to give credit to the late Gian-Carlo Rota for tweaking my curiosity about phenomenology. Like Hubert Dreyfus, Rota seemed to want to use Husserl as a stick for bashing artificial intelligence; but I always felt that Rota's discipline as a mathematician gave him a far sharper mind than Dreyfus had. (Rota reminds me of a remark that Bertrand Russell once made about his own intellectual pursuits: He devoted his youth to mathematics, turned to philosophy in middle age, and devoted his twilight years to writing fiction! Rota never made it to fiction, but he read Husserl with the same intensity he applied to reading mathematics.)

As I tried to do with Habermas, let me just try to being a few key points into the foreground. Most important is the nature of the subject matter: Phenomenology is the study of appearances. Once you fix on that basic premise, everything else is icing on the cake (although it would probably be a good idea to sort out who made the icing, be it Hegel, Husserl, Heidegger (philosophy's answer to music's three B's?), or any of the more recent sources). A key element of Husserl's icing is the study of noesis, which is his term for the "knowing act" that mediates between appearances and what the mind takes to be reality.

Note the "scare stress" in that last sentence: At the end of the day, I probably have to confess to being an unabashed idealist. I may not necessarily embrace Berkeley's extreme solopsism; but I still tend to live by Ringo Starr's immortal line in Help!: "It's all in yer mind!" This is why I have a general problem with the very concept of an Ontology of Everything: It presumes some representation of reality that can provide a foundation for all person-person (and person-machine) interactions (what Habermas (June 8 entry) called the "foundation of shared concepts"). However, the mind reshapes what it takes to be reality with every experience it has; and it seems naive and/or vain to assume that there is going to be some digital resource out there that can be compatible with every mind that tries to engage it.

This is why I have tried to live by the motto in today's title: No ontology without phenomenology! Unless the Semantic Web figures out how to get a grip on this concept of noesis, it is likely to open itself up for any number of bloopers once we try to use it for "real world" applications, whether they involve business or leisure. As Habermas tried to emphasize, at the end of the day, it is all about understanding; and understanding does not reside in "objective facts" but in the negotiation between individual "knowing acts" that is part of our everyday discourse.

June 08, 2006: Habermas 101

As I promised yesterday, I shall try to provide a relatively concise outline of the framework Jurgen Habermas developed in his Theory of Communicative Action. There are basically two parts to the framework. The first is his model of the process of reaching understanding. Then, because reaching understanding is an action, the framework also deals with the "four sociological concepts of action.

The model of the process of reaching understanding is based on four interrelated perspectives:

1. Personal interactions that take place with regard to three "worlds;" the objective, the social, and the subjective.
2. Validity claims that may be applied to the "transactions" of those interactions; these validity claims may be based on propositional truth (objective), acceptability determined by social norms (social), and the sincerity or authenticity of the individual (subjective).
3. The agreement to negotiate by recognizing and criticizing validity claims.
4. The grounding of negotiation on "common definitions of the situation," i.e. the recognition that negotiation can only be based on a foundation of shared concepts.

This model grew out of work Habermas initially published in his essay "What is universal pragmatics?," which is the first chapter in the collection Communication and the Evolution of Society. As many social theorists have observed, this model presumes an “ideal speech situation,” meaning that the parties involved want to reach understanding. We are all familiar with (too many?) cases where this is not the case; but Habermas’ model is still a useful way to identify four concepts that provide a grounding for the nature of understanding.

The three-worlds concept surfaces again in the classification of actions into “four sociological groups:”

1. The teleological concept is based on the formalist principle than an action effects a state transition. We thus understand actions in terms of the states they are intended to achieve, either directly or as part of a more extensive plan; and actions are described through rules that map a state and an action to a resulting state. For Habermas this concept has its origins in the pioneering work of John von Neumann and Oskar Morgenstern on a theory of strategic games, although von Neumann had already been exploring theories of states and transitions in his early work on computer design. Note, however, that, because the focus of this concept is on state, it resides more in the domain of the static objects of lexis, rather than praxis. Put another way, it is more concerned with the results of action than with the dynamic nature of action, itself.
2. The normative concept assumes that actions are regulated by socially determined norms. In other words the “goal state” of an action is less important than whether or not that action is acceptable in the social setting in which it is taken. Because that setting is rarely (if ever) static, this concept is more firmly grounded in the dynamics of praxis than is the teleological concept.
3. The dramaturgical concept views the acting subject as “playing a role” in a social setting that provides an “audience.” It is perhaps best captured by the title of Erving Goffman’s pioneering analysis, The Presentation of Self in Everyday Life. This concept, again, is grounded in the dynamics of praxis.
4. The communicative concept refers to the interpersonal relations (based on both lexis and praxis) of at least two subject who seek a shared understanding of a given situation “in order to coordinate their actions by way of agreement.” These are precisely the personal interactions that take place over the coordinated work practices of office work. It should therefore be no surprise that Habermas identifies origins of this concept in the ethnomethodological techniques of Harold Garfinkel, as well as George Herbert Mead’s earlier pioneering research in communicative action.

The connection to the three worlds has to do with the fact that teleological action takes place in the objective world, normative action in the social world, dramaturgical action in the subjective world, and communicative action in the integration of all three worlds.

June 07, 2006: The Road to Lexis and Praxis

I figure I probably ought to offer some justification as to why I invoked Greek terminology from Plato and the general plan of Kant's Critique of Pure Reason in yesterday's entry. Accounting for Kant is definitely the taller order, particularly since I feel I am just getting beyond the Cliff Notes stage in my ability to read him. So I may end up postponing that question until I have given the Critique a second reading (which will definitely be more thorough than my first reading, meaning that this postponement is likely to take some time). On the other hand I feel I can address the justification of my invoking the lexis-praxis terminology at the present.

I suppose I can begin that justification by repeating yesterday's citation to the paper by John Seely Brown and Paul Duguid ("Organizational learning and communities-of-practice"). I first read this paper when John was running Xerox PARC and I was involved with knowledge research at the Palo Alto laboratory that Fuji Xerox had established (which was on the same campus as PARC). The most influential passage in this paper for me was in its second paragraph:

Formal descriptions of work (e.g., "office procedures") and of learning (e.g., "subject matter") are abstracted from actual practice. They inevitably and intentionally omit the details. In a society that attaches particular value to "abstract knowledge," the details of practice have come to be seen as nonessential, unimportant, and easily developed once the relevant abstractions have been grasped. Thus education, training, and technology design generally focus on abstract representations to the detriment, if not exclusion of actual practice. We, by contrast, suggest that practice is central to understanding work. Abstractions detached from practice distort or obscure intricacies of that practice. Without a clear understanding of those intricacies and the role they play, the practice itself cannot be well understood, engendered (through training), or enhanced (through innovation).

This was a pretty radical position for a representative of "The Document Company," since the primary medium for those formal descriptions was the document! So I saw in this paper the invocation of a fundamental opposition between the document-based and the practice-based. In my own research (which, at that time, was focused primarily on the frustrating problem of organizational memory) I tried to generalize this to an opposition between the noun-based and the verb-based, because, while documents certainly used verbs, the document itself was an artifact; so both the documents themselves and the formal descriptions that were often the content of documents resided firmly in the world of nouns. They were, indeed, "abstractions detached from practice." (For a while I was even using the phrase "communication artifact" instead of "document.")

I was quite happy with pursuing the opposition between nouns and verbs until I got involved with reading Kenneth Burke's Grammar of Motives. There I discovered that Spinoza had also been interested in such oppositions. Burke cited three variations, all of which I felt could be relevant: passive-active, object-subject, and product-producing (which resonated nicely with my interest in the term "communication artifact"). With this as background, along with a natural desire to seek out generalizing categories, I was primed for the opposition of lexis and praxis that Jurgen Habermas invoked in the first chapter of his Theory and Practice. This led me to explore the specific definitions of both of these words in the Perseus online dictionary of classic Greek. The nice thing about the entries is that they invoke citations from classical sources; and that is how I discovered the citation of lexis and practice in Plato's "Republic" (along with a context-based translation of those words)!

Now Plato's context is concerned with giving faithful accounts. More specifically, he is addressing relating "some word or act of a good man" to others; and this is where the words lexis and praxis appear in the original Greek. However, this takes us back to the original theme of the Brown-Duguid paper. At the end of the day, work is all about both words and acts. Our training in information technology has served us well with formal descriptions in the domain of words; so, like the drunk under the lamp-post who looks for his keys because the light is better there, we try to assume that only the words that matter, because they are what we know how to describe. In so doing we invoke abstractions that distort our understanding of acts.

What is to be done? Clearly, we need a better handle for describing and talking about acts that can stand up to the formal descriptions we invoke for words. The first volume of Habermas' Theory of Communicative Action proposes such a framework for talking about acts. However, an account of this framework is very much another story; and I shall postpone it to another entry (which I can probably write before I get around to Kant)!