Various systems have been put in place to facilitate communication among employees and liberate organizations by removing or de-layering managerial hierarchies that were perceived to be barriers to open communication and collaboration. By the numbers - which is to say, by the systems own internal metrics - they appear to be effective ... but are they really?

A closer look at organizational; behavior reveals a "Ubiquitous" process of tinkering and improvisation around these systems that, coupled with other factors, points to the conclusion that "the technology does not seem to work completely according to plan."

The term the author uses is "drift" - comparing the agendas of what the organization currently does, what it wishes to do in the future, and the technology solutions implemented to achieve a desired course - are all like currents, and the actual behavior seen is affected by each of these, but not strictly controlled by any.

Drifting is the outcome of two separate processes: first, where a technology has openness and plasticity (it can be adapted), users will gradually learn to discover and exploit the affordances and potentialities of the systems; and second, where users are compelled to achieve results, without strict control of the means, they will seek the path of least resistance, which often means working around the rigid constraints of a system.

True Stories, Again

The author suggests that "drift" is not necessarily a bad thing. He provides examples of how a groupware application (Lotus Notes) was implemented by a number of companies for various purposes. In each case, while the software failed to achieve its intended outcome, the mis-use of the product had positive consequences:

He also cites instances in which the intended users "bypassed" many of the features and functionalizes of the application because it was unsuitable to their needs. They did not use it as intended, but used it nonetheless to suit needs that the designers did not intend. Where the application was adaptable, it was adapted. If it could not be adapted, it was abandoned.

Drifting and Systems Development

Examining the actual use of information systems reveals that the "pre-defined goals" of an IT solution are often out-of-line with what the users actually want or need to do, and the "designed tasks" often miss a more obvious course of action that will lead to success. And while IT architects are "horrified" by improvisation by users, they rather should consider such behavior to be opportunities for innovation, which would lead to the discovery of systems improvement.

A General Model

The author asserts that software design takes place in a laboratory, a clean place where design is based on theory, coupled with a general notion of the high-level goals of an organization. That solution is then delivered into "the swamp," where human users must attempt to utilize the clean-room solutions in a dirty environment, and adapt them as needed to make them work under real-world conditions.

The real world is a messy place where things don't go according to plan, and where the needs of the moment often take precedence. In such an environment, people must rely on affordance (what an object appears to be capable of doing) rather than instructions, on improvisation rather than training, and on tactics rather than theory.

(EN: I disagree with this dichotomy, and the suggestion that design and use are entirely separated rather than interdependent entities. One cannot design without considering real-world needs and uses, nor is it efficient to use something for without considering its intended use.)

Another aspect of drifting is that, over time, actions drift away from their original goals. Other goals intervene; the original goals are invalid; alternate agendas emerge.. This is, to some extent, the reason system designers regard improvisation, hacking, and drift as negative influences. Rigid adherence to intended use, it is assumed, ensures the original goals are accomplished. In truth, rigid adherence creates ritual behaviors, which are no more aligned with the goals that purposeful deviation. And as a system becomes designed to control behavior, not to achieve goals, it can lead to counterproductive ritual behaviors.

Swampy Time and Space

The concepts of "plan" and "method" are from the laboratory environment - the orderly progression of steps in a designed process flow. - are based upon a linear perception of space and time, with minimal alternatives. In the swamp, behavior is less structured and options are unlimited. However, improvisation occurs in a "moment of vision" - where an individual in the swamp envisions a pattern or sequence that will lead to the accomplishment of a goal.

Generally speaking, the laboratory is a predictive environment, whereas the swamp is a reactive environment: people follow the laboratory plan, but when a contingency arises, they improvise to overcome it. The contingency may be an unusual occurrence, or it may be a shortcoming of the plan itself.

The two modes are, and have ever been, coexisting and complementary. Constant improvisation is highly inefficient, but rigid obedience to plan, regardless of circumstances that arise, is ineffective.

Global Consequences

Drifting phenomena are pronounced in the infrastructure of global companies. This is because a single solution is not universally applicable - the peculiarities of a given location require deviation from the master plan, nor is it feasible to control behavior on such a large scale without significantly damaging the outcome. In practice, success in achieving the goal always outweighs success in conforming to the system's requirements.

The author strays a bit, to discuss the concept of standards. A standard may be developed in a laboratory or emerge from common practices. The latter seems more practicable (and practicable), but even a standard developed from experience in the swamp can be limiting, as it is based on past practice of a majority. Any attempt to enforce standards will sacrifice effectiveness for conformity - or more often, it will create additional work to mollify the demands for conformity to standards, while meanwhile doing whatever is necessary to accomplish the goals.

Back to globalization: when globalization is done by means of information technology, the rigidity of that technology becomes exposed. The larger an organization, the greater the heterogeneity of its units, and the more difficult it becomes to implement a homogeneous solution. For this reason, many companies that reach across borders have a policy of open toleration for adaptation and improvisation in the "foreign" offices, and bricolage is expected to do business in spite of the restrictions of the standard solution.

There is an ongoing effort of global companies to adapt their systems for greater flexibility, that foreign offices may use the corporate solution, but achieving this control is difficult because of the unpredictability of the foreign elements. More to the point, if the "standard" solution is flexible, the greater the opportunity for bricolage in each office, and the greater the resulting drift.

The author reiterates that bricolage is not isolated to global companies, merely that it is more pronounced. In any sizable operation, there will be elements for whom the standard solution, by design, is ineffective and improvisation becomes necessary.


(EN: the chapter doesn't have a "conclusion" section, but the topic changes back at this point to a more general view)

The prevailing mindset in contemporary management is for a geometric model: a pyramid structure, a linear process, etc. Btu as organizations grow and become increasingly global, the conditions of the swamp become more pronounced, and the attempt to demand conformance to laboratory standards becomes increasingly counterproductive and out-of-touch with reality. AS the pace of change accelerates, an organization must be able to adapt quickly to the present, rather than being chained to the past by standards and procedures.