(Co-Authored with Steven A. Cavaleri)
I recently alerted Steven Spear, author of Chasing the Rabbit, to Parts Two and Three of this series which discuss his very important book. I guess my posts prompted him to post a blog on how high velocity organizations out learn and out race the competition. Since Steve’s blog post is directly related to the problem solving pattern in both high velocity organizations and open enterprises, my next two posts will provide a close but appreciative analysis of his views. He begins with a statement about how knowledge is created.
“Knowledge of what to do and how to do it is created by building theory, testing theory, and building new theory when the old ideas fail. High velocity organizations build the creation of useful knowledge directly into process design and process improvement. As a result these organizations out improve, out innovate, and out invent their rivals. They out learn them, so they out race them. How they do so is a vital lesson now, when old answers on how to compete no longer apply.”
I like this statement very much, especially its last four sentences, but find the first sentence a bit too optimistic as a general characterization of how knowledge is made. It would fit better if it started with: “In high velocity organizations, knowledge of what to do and how to do it is created by seeing gaps between expectations and reality, building theory . . . ” and continuing on as is, and then the second sentence was slightly modified to refer to “they” or “these organizations” at the very beginning.
I offer this suggestion for two reasons. First, while I think all organizations arrive at theories and related expectations which are then sooner or later met with experience that cannot be denied, many of them, unlike high velocity organizations, neither deliberately engage in “theory building,” nor self-consciously “test theories,” nor try to “see gaps” by looking for them. And if they do come up with new theories when old ideas fail, the processes by which they do so, will rely much more on intuitive and relatively uncritical leaps to another alternative, than on attempts to generate, using various methods, alternative theories that are subjected to exacting tests and evaluations. And second, in a general statement like this about how knowledge is generated by high velocity organizations, it’s just as important, I suspect Steve will agree in light of his book, to reference gaps between expectations and reality, or if you like “problems,” as it is to talk about theory building and theory testing.
Steve continues his discussion by pointing out, with reference to the views of Clayton Christiansen and Thomas Kuhn, that the theory building and theory testing cycle involves both inductive and deductive reasoning, with theory building using induction and theory testing deduction. I tend to agree very much with this, provided the terms “induction” and “inductive reasoning” are understood as referring to processes of creating or discovering new theories. In my view there is no such thing as an “inductive logic” which can warrant inductive conclusions based on the premises of inductive arguments. But if by “induction” one means the kind of activity discussed by John Holland and his collaborators, and by Paul Thagard in his work, then I’m happy to agree that theory building involves inductive “reasoning,” always including leaps to conjectural, often creative, “inference” to the best explanation.
Steve next goes on to say:
“With a theory in hand–a statement of what attributes and behaviors cause what outcomes in what circumstances, you go about testing it deductively. This means that:
— You make predictions of what you expect to happen.
— Observe what actually happens.
— Identify anomalies–the events that actually occurred contrary to what you expected.
These anomalies are critically important because they tell you where your existing theories have failed and where you need to look again (inductively) for deeper insights.”
I generally agree with this statement though I’d like to note two things. First, there might well be alternative theories that need to be tested rather than just one. The pattern of testing still involves deduction, and we would have to compare how well expectations from contrasting theories compared to events. Also, selection among theories might involve more than just comparison with events. We’d get into that kind of situation if our observations and measurements couldn’t distinguish among competing theories. In that case alternative evaluation perspectives or criteria might have to be used to decide which theory was best, until we could develop new measurement techniques, instruments, or models to empirically distinguish between the alternatives.
Second, should the term ‘anomaly’ be used here while also referencing Kuhn as Steve does? I’d prefer words like ‘deviations’, or ‘expectation gaps’ instead. My reason is that Kuhn didn’t always use the word ‘anomaly’ consistently, and that he sometimes contrasted ‘puzzles’ (deviations which could be accounted for by a new theory sharing the same paradigm as an older theory), with ‘anomalies’ (deviations which resisted resolution in the context of a single paradigm even when theories formulated within the paradigm were changed or revised). If we follow this use of ‘anomaly’ and ‘puzzle’, (which, however, Kuhn didn’t always do) then in testing, very few of the deviations that are detected would be ‘anomalies’, as opposed to just less troublesome deviations or ‘puzzles’ that refuted our theories, but didn’t motivate us to change our paradigm.
Next, Steve asserts the important conclusion:
“High velocity organizations institutionalize these inductive/deductive cycles to they are consistently building and applying new knowledge faster and with more breadth and consistency than their rivals.”
I very much agree with this and would like to call attention to its similarity to the view I’ve been writing and teaching about for some time, namely that the most adaptive organizations are those that are best at initiating and executing Knowledge Life Cycles (KLCs), where KLCs include: seeking, recognizing and formulating problems, solving those problems by making new knowledge, and integrating (broadcasting, searching and retrieving, sharing, and teaching) the new knowledge into one’s organization. The step of problem solving at the organizational level of analysis includes: information acquisition (the theory building step), individual and group learning (the theory building step), knowledge claim formulation (the theory building step), and knowledge claim evaluation (the theory testing step).
The best short account of the KLC and its relation to KM is here. The framework used in earlier posts in this series on “the problem solving pattern” reflects the KLC framework as well, but the vocabulary has been changed, for a variety of reasons. In my next post, I’ll continue my commentary on Steve’s Blog by discussing his four capabilities of rabbit organizations once again.
To Be Continued