Page 205 -

194 MANAGING KNOWLEDGE WORK AND INNOVATION

Critics of the linear model argue, in fact, that linearity is little more than a
feature of retrospective and rationalized descriptions of innovation and knowl-
edge creation processes by those involved. For example, the Medico story, when
retold, looks very much like a success resulting from the rational decisions of
the Medico team. However, in practice, it was a very messy, highly ambiguous
and context-dependent process, the outcomes of which could not have been
foreseen. The country weekend meeting, for example, happened to result in a
group of professionals getting together to set standards that then aided diffusion
of the brachytherapy treatment. Although the Medico team made the event
possible, this outcome was not planned by them. These kinds of ‘accidental
encounters’ are actually typical of innovation processes, especially in the early
stages (Kreiner and Schultz, 1993).
Linear descriptions, then, gloss over the discontinuities, iterations and politi-
cal uncertainties that characterize all but the most simple innovation processes
(Clark, 2003). For example, in the Medico case, political uncertainties amongst
radiologists and urologists played a key role in driving, and also interrupting,
the innovation process. That is not to say, however, that there are no ‘stage
gates’ in the innovation process. In drug development, for example, regulatory
approval mechanisms (e.g. for phased clinical trials) do play a role in shaping
and concentrating innovation activities at particular points in time. For example,
‘first-in-man’ safety studies must be done prior to wider patient population tri-
als. Yet, even here, there is considerable movement back and forth in terms of
knowledge flows and processes, with the so-called ‘preclinical’ work often having
to be revisited or redone after clinical results are known (Swan et al., 2007a). In
other industries, where the regulations may be less disciplining, the stage gates
are typically self-imposed in order to try and control the innovation process, as
for example with the use of project management methodologies which include
extensive planning protocols requiring scheduled milestone identification.

>> PROCESS VIEWS – INNOVATION AS AN INTERACTIVE
PROCESS
In contrast to the linear model, process views depict innovation as a cumulative
and iterative set of episodes, activities and fortunate (and unfortunate) coinci-
dences, where multiple actors, multiple forms of knowledge and organizational
tasks interact and serendipity has a major role to play (Clark, 2003). Key features
of this view are summarized in Box 9.3. ‘Managing’ innovation, then, is more
about creating a context that allows such new combinations of knowledge and
practices to happen than about ‘management’ in the traditional control sense.
Key aspects of this enabling context were described in Chapters 2 and 6. The
use of the term ‘episodes’ (as opposed to ‘stages’) reflects the sporadic, iterative,
recursive and sometimes discontinuous nature of innovation processes (Van de
Ven, 1986). This view, in contrast with more broad-brushed prescriptive models,
highlights the effective deployment and management of knowledge as critically
contingent upon the different episodes of innovation as well as upon the social and
organizational/political features of the context in which those episodes unfold.

6/5/09 7:20:35 AM
9780230_522015_10_cha09.indd 194
9780230_522015_10_cha09.indd 194 6/5/09 7:20:35 AM