p.3 Improvisation is described as an exceptional practice in organizations (Crossan, Cunha, Vera & Cunha, 2005; Weick, 1998). It is inefficient and risky and should thus be reserved only for the most difficult competitive challenges... improvisation is crucial for organizational routine... the experience of routine is the outcome of everyday micro-adaptations and unprescribed tactics.

p.4 where does improvisation come from? Studies... point to repetitive work itself as a favorable context for the development of mastery in improvisation. Repetitive work provides motivation, tailor-designed work and immediate and specific feedback - the core factors for the development of mastery in any skill (Ericsson, Krampe, & Tesch-Romer, 1993).
 Improvisation and repetitive work are thus mutually constituted and routine is the outcome of their dualistic relationship... We argue that repetitive work is an improvisational achievement and... routine is a dynamic achievement that depends on improvisation.

p.5 Improvisation is but one process of exploration, and a last resort for that matter... In all but the most challenging contexts it is lacking in efficiency and effectiveness... It centers on an emergent course by exploring multiple alternatives and by putting resources to multiple and often non-canonical uses (Orlikowski, 1996).

p.6-7 The literature on improvisation in organizations... circumscribes improvisation to competitive contexts where all other practices fail - contexts that couple uncertainty with a high rate of change (Miner, Bassoff, & Moorman, 2001)... Planning is difficult in such environments because information on competitive dynamics is ambiguous and has a short shelf-life... Research in fast-changing competitive environments has shown that improvisation is not only the main alternative but also the most robust practice for surviving and thriving (Brown & Eisenhardt, 1997). When improvising, agents address challenges effectively because they are able to learn and plan while acting. Improvisation is first and foremost about acting. When improvising, agents attempt to address challenges by acting on them, instead of choosing to reflect and plan the best way to address them... When improvising agents learn more about the challenge they are facing, making their attempts to address it increasingly knowledgeable and increasingly powerful (Weick, 1987). Improvisation is therefore effective in handling novel challenges because it allows agents to reflect and plan while acting... improvisation is faster than planning because it allows action to inform planning as it unfolds.

p.7 improvisation is more efficient than planning because it draws on available resources, instead of procuring new ones... When improvising to deal with novel and unexpected challenges, improvisers draw on available resources, using them creatively. The focus is on making do with resources that agent currently hold using them in new and creative ways (Machin & Carrithers, 1996; Peplowski, 1998).

p.8 Improvisers have a bias for action. They have developed a habitus (cf. Bourdieu, 1990) to address challenges, threats and opportunities by acting on them once they surface

p.9-10 The defining characteristics of repetitive work are that it is codifiable, that it is stable, and that it is relatively unambiguous. These three aspects of repetitive work allow the separation between its planning and its execution (Bendix, 1947).

p.10 repetitive work consists of observable, simple practices enacted to address recursive challenges. Organizations engage in repetitive work because part of the challenges they face are recursive. They occur consistently and recurrently across time.

p.10 If work processes are codifiable but need to be constantly readjusted to match competitive demands, standardization is not only impracticable but also unnecessary. Standardizable repetitive work needs an environment that absorbs a stable output (Liker, Collins, & Hull, 1999)... the stability necessary for repetitive work is an accomplishment of the organization (cf. Yan & Louis, 1999).

[JLJ - this suggests an interesting alternative definition of computer - a controlled environment where specific, codifiable work is performed according to human-written script, on demand, perhaps repetitively or not at all, in response to consistent and recurrent needs, recursively across time. This scripted work is performed mindlessly by the machine on pieces of information, such as arithmetic operations, loading and storing data from external locations such as databases, networks or memory, or logical branching operations, yet paradoxically, the work output can approach or even exceed the quality, consistency and accuracy of that performed/produced by a mind.]

p.11 Repetitive work can... be standardized in a set of prescribed procedures that can be applied mindlessly and effortlessly to carry out the organization's production process. According to this view employees do not need to deviate from prescribed roles, rules and procedures because these are designed to meet every challenge they face in their everyday work. [JLJ - perhaps at a workplace like a nuclear power plant you want to enforce strict following of procedures in order to reduce the chance of an accident]

p.13 we first explain how repetition is an improvisational accomplishment. We then argue that employees improvisational skills are themselves acquired in repetitive work. We end by discussing how routine emerges from the mutual constitution of repetitive work and improvisation.

[JLJ - I would argue that the concept of 'repetitive work', upon close inspection, is itself suspect. Any procedure is itself developed through improvisation and is maintained through constant improvisation, even though there might be long periods where no changes are made to the steps themselves. The world, the machines, the pieces to be assembled, the workplace environment, even the workers are slightly different as time passes. Most of the time this does not affect the repetitive work, but over time these small changes accumulate and require subtle improvisation to maintain the appearance of repetitiveness.]

p.15 There are some instances where prescribed procedures are adequate for the organization's goals. Their repeated enactment is sufficient to deliver their prescribed outcomes.

[JLJ - imagine a worker at a healthcare test center, drawing blood samples. The work itself is repetitive, but the worker must take steps to clean the sample site, seal and label the samples, and arrange prompt shipment to the lab that does the actual analysis work. Contaminated samples, untimely delivery, samples too small in size, can all affect the end results. While repetitive in nature, the work also requires a quality-control attention to detail so that unanticipated events do not end up influencing the results, which perhaps might misguide a doctor taking actions based on test results which he assumed would be performed correctly. A worker on an assembly line might install a nut and washer that don't quite go on all the way. Now what? Is it "good enough" to hold the part, or is the part defective and have to be replaced? Fast action is necessary, as the assembly line is moving. This is hardly "repetitive" mindless work. There is an implied, skilled quality inspection that goes along with repetitive work, necessary and difficult to script.]

p.16 Independently of how difficult it is, bricolage allows employees to be able to fall back in prescribed procedures by creating on the spot tools and materials these require to be enacted repetitively.

p.16 there are instances when a breakdown of conditions for actions is of such an extent that employees need to enact a whole interpretive and action structure where prescribed procedures make sense again (Lanzara, 1983). In such crises, people struggle to make their conditions for action sensible as they decide on an appropriate course of action. The challenge here is not [JLJ - only?] to improvise work practices but also to improvise a set of conditions for action to help build a frame under which they can act. Research has shown that, when pressed to do it, agents are able to draw on their ongoing action to reconstruct a sensible frame of interpretation that is enough to support repetitive work procedures (Hutchins, 1991). This triggers a sensemaking process

p.17 The extent of improvisation required to sustain repetitive work suggests that standardized work in the office and [on] the factory floor is all but repetitive.

p.19-20 Research on loafing (e.g. Lim, 2002) has shown that employees are able to improvise [backstage] even in contexts of extensive surveillance, where they can engage in non-work activities... cyberloafing [JLJ - if you are doing research on loafing, what do you call it when you are not being productive in your research task? Can you be "loafing" from loafing research? Or, are you instead collecting detailed data for your research work?]

p.21 It is thus in employees everyday work that their improvisational skills are developed... or, more specifically, in employees' early experience at work (Bensen, 1986; Van Maanen, 1976).

p.26-27 Routine is the outcome of three reinforcing processes of improvisation. The first process includes those improvisations that seek to create the conditions for repetitive work... The second process includes those practices that seek to create the conditions for employees to develop their improvisational skills... Finally, and perhaps more importantly, employees need to enact a third set of improvisations to keep improvisation and its development hidden from managers' scrutiny... Routine thus emerges from mutual constitution of improvisation and repetitive work.

p.27 This approach highlights the adaptive nature of routine. The literature on strategy and organization in fast-changing competitive contexts has underscored the need to create dynamic capabilities that allow the organization to adapt to ever-changing competitive challenges... The assumption here is that innovation and novelty are the major, if not the only processes that are able to keep the organization moving fast enough to keep up with and, hopefully, outpace its competitors... If routine is the outcome of multiple improvisational processes, then it is as adaptive as novelty and innovation.

p.28 improvisation is about exploration - finding and taking advantage of new opportunities as they present themselves in the market.

p.29 The core contribution of this paper is to suggest an alternative view of routine. Drawing on evidence scattered through thick descriptions of everyday life in organizations, we argued that routine is the outcome of situated deviations from prescribed roles and procedures and is, in itself, a source of emergent strategic action. The dominant conception of routine see it as a mindless and effortless process of compliance. We explained that routine can also be a mindful and effortful process of improvisation. We further suggested that this process is adaptive. As employees enact local improvisation to consistently enact repetitive procedures and achieve prescribed goals, they are contributing to the sedimentation of unprescribed practices into an emergent strategy for the organization as a whole.

Previous research has shown that routines, meaning habitual procedures to carry out a specific task, are sources of adaptation. We used the descriptions of repetitive work dispersed across available organizational ethnographies and other micro-accounts of everyday life in organizations to push this literature further by showing that routine, meaning the work needed to consistently and repeatedly enact prescribed procedures and goals is in itself a powerful process of adaptability.

p.29-30 Routine is a powerful process of adaptation... The resilience that many large bureaucracies and complex organizations have shown in highly competitive marketplaces... is due to their ability to adapt to environmental challenges through their employees' improvisations as they struggle to produce and reproduce routine.