1

PATHCREATIONININNOVATIONDESIGN:AMULTI‐LAYEREDPERSPECTIVE

Ola Henfridsson Jönköping International Business School & Viktoria Institute

Youngjin Yoo

Temple University

Fredrik Svahn Viktoria Institute

Abstract.

We extend the path creation perspective by outlining a dialectic process model of its multi-layered nature. To better understand path creation in innovation design, we draw on a six-year in-depth field study of designers at CarCorp and its owner, GlobalCarCorp. In particular, our empirically grounded theoretical model focuses on the dialectical process by which designers shift their attention across different layers of innovation in pursuit of a particular innovation path. Because of the reciprocal nature of path creation and path dependency, the design praxis of path creators are embroiled in a realm of tensions and contradictions within and across different layers of path creation, which then serves as a underlying generative force that propels technology innovation forward. Our study depicts the process of innovation that emerges out of a series of unpredictable turns and twists at different layers.

1. INTRODUCTION

On January 3, 2009, the New York Times featured an Op-Ed article covering “four ways for

Detroit to save itself”, all encompassing the adoption of new digital technology for making cars safer,

cheaper, cleaner, and more convenient (Thrun and Levandowski 2009). Calling Detroit to break away

from its path, the article echoes the widely held conviction that a new innovation path is essential for

turning around the troubled auto industry. However, the suggestion that such technology adoption would

be accomplished “without too much trouble” appears optimistic. Our 6-year study of path creation at

CarCorp and its parent company, GlobalCarCorp1, suggests that this is a long and arduous process that

cuts across multiple layers of path creating design activities. Outlining a new path creation perspective,

our research explicates digital innovation at CarCorp as a multi-layered process enacted by designers

within and across layers of dependencies in the microstructure of car design.

1 GlobalCarCorp is a pseudonym for one of the big three American automakers.

2

In the organization literature, the idea of path creation has emerged as a powerful theoretical

perspective to conceptualize innovation (Boland et al. 2007, Garud and Karnøe 2001). Viewing

innovation as ongoing agency embedded in sociotechnical structures (Garud and Karnøe 2003, Garud et

al. 2008), Garud and Karnøe’s (2001) original notion of path creation was developed in response to the

economic theory of path dependency (Arthur 1989; David 1985). While path dependency analysis focuses

on innovation as an outcome of complex interactions of seemingly random minor events, thus placing

human agency backstage, the path creation perspective highlights the role of designers and entrepreneurs

in mindfully deviating from the existing path (Garud and Karnøe 2001, Stack and Gartland 2003).

Embedded in the institutional order established over time, technology designers “shape emerging

institutions and transform existing ones despite the complexities and path dependences that are involved”

(Garud et al. 2008, p. ?).

The original idea of path creation focused on entrepreneurs and their path creating activities

(Garud and Karnøe 2001). Recently, scholars have recognized the role of multiple and distributed actors,

each contributing to the production of a new path (Boland et al. 2007, Garud and Karnøe 2003). For

instance, Boland et al. (2007) explored the cascading nature of multiple and intersecting path creation

among heterogeneous actors. Their work expands the path creation perspective by placing an

entrepreneur’s path creation activities in a larger sociotechnical context, thus noting that an entrepreneur’s

deviation often collides and interferes with other paths, resulting in unpredictable wakes of innovations.

In this paper, we further contribute to this stream of literature by noting the multi-layered nature

of innovation paths. If Boland et al (2007) take a step back from a single entrepreneur in order to ‘zoom

out’ to see multiple paths forming a complex form of wakes, in this paper, we take a step closer to ‘zoom

in’ to pay a closer attention to a single path. In so doing, we observe that an innovation path that CarCorp

takes in fact consists of multiple layers that are interwoven with one another. As a single thread is often

not a straight line, but is rather made up of multiple layers of thin fibers interwoven together, we find that

innovation path of a firm consists of multiple layers. Without paying a close attention to the layered

nature of path creation, one may mistakenly think that changes in the artifact layer (e.g., adopting new

3

digital technologies in the automotive industry) be sufficient for creating a new path. Indeed, earlier

studies of path creation examine paths as a one-dimensional shift from the established innovation

trajectory. In contrast, we outline a perspective that explicates path creation as a multi-layered process

where designers turn their attention to different layers of the technological innovation over time.

We draw on the notion of microstructure of design (Baldwin and Clark 2000) that includes task

structure, design structure, and artifact as three layers of design. We further note that the initiation of such

path creation inevitably involves a change in the identity of the designers occasioned by tensions in the

artifact layer. Enacting the envisioned innovation path, designers confront contradictions within and

across layers of innovation design. Informed by a dialectical view of organization (Benson 1977, Seo and

Creed 2002), we conceptualize such enactment as design praxis, involving active reconstruction of

sociotechnical reality. Essentially, the research question addressed in this paper is: what is the multi-

layered process by which designers create new paths in digital innovation?

To better understand layers of path creation in innovation, we conducted a 6-year longitudinal

study of car infotainment solutions at CarCorp and its owner GlobalCarCorp, and observed how a new

innovation path was created over time. Struggling with realizing the vision of the “connected car”,

CarCorp designers actively effectuated changes in the artifact layer of the car infotainment innovation,

later discovering that the envisioned innovation path needed to be established in the design and task layers

as well. Each such discovery involved contradictions between the established innovation path, colored by

CarCorp’s manufacturing origin, and the path created by infotainment designers dealing with new digital

technology.

In the remainder of the paper, we first outline a review of the path creation literature. We then

present our theoretical framework that draws on the microstructure of design and the dialectic view of

organization. Next, we present the research methodology, followed by a presentation of our longitudinal

case study of CarCorp’s path creation in the car infotainment area. Sensitizing the case study with the

theoretical framework outlined, we develop a process model of multiple layers of path creation. Finally,

4

we conclude our paper by discussing implications of our process model for innovation theory and

practice.

2. PATH CREATION

Path creation analysis focuses on the role of agency in creating new trajectories in innovation

(Garud and Karnøe 2001, 2003). Such analysis also recognizes that new paths are never created in a

vacuum isolated from already existing sociotechnical arrangements (Hanseth 2000). In this regard, path

creation is an idea that presupposes the existence of path dependency (Arthur 1989, David 1985). As

implied in Garud and Karnøe’s (2001) original definition, there exists a mutual interdependence between

path creation and path dependency, manifesting the classical duality of structure (Garud et al. 2008,

Giddens 1984). As Garud and Karnøe (2003) note, the term ‘path’ suggests “that the accumulation of

inputs at any point in the development of a technology is as much a position that actors have reached as it

is one that they may depart from” (p. 281).

The reciprocal nature of path creation and path dependency is reflected in actors’ ongoing

enactment of existing structures. Drawing on existing structures as resources in innovation, designers may

mindfully deviate from a subset of institutional arrangements and routines to accomplish change

(Feldman and Pentland 2003, Garud and Karnøe 2001). Designers can therefore be regarded as embedded

agents who create new paths as a step-by-step process over time (Garud and Karnøe 2001, 2003).

There is a paucity in the literature of research that closely examines the underlying nature of path

creation. Outlining technology entrepreneurship as a collective enterprise involving distributed

competences (Tsoukas 1996), Garud and Karnøe (2003) highlight path creation as a process that involves

multiple and distributed actors. Using evidence from the wind power market, they explain how multiple

actors deploying modest resources can co-shape a successful innovation path. Boland et al. (2007) further

extend this analysis of path creation by showing how wakes of innovation are occasioned when

heterogeneous actors’ paths collide. To date, however, the literature offers little explanation of the

internal dynamics of path creation by which a group of designers or entrepreneurs break away from its

5

past. We suggest that such path collision may involve a shift of attention across different layers of

institutionalized structures over time. Next section outlines a theoretical perspective that offers a basis for

understanding such shifts of attention in the path creation process of specific groups of designers.

3. TOWARDS A MULTI-LAYERED PERSPECTIVE

In view of this literature review, this section presents the basis of a path creation perspective that

combines Baldwin and Clark’s (2000) model of layers of structure (LoS) and a dialectical view of

organization (Benson 1977, Seo and Creed 2002). While the LoS model provides a conceptual tool to

explore different layers of path creation, the dialectical view facilitates analysis of the tensions between

path creation and path dependence evoked in the praxis of designers.

3.1 Layers of Innovation Design

Summarizing a research agenda on design and industry evolution developed over more than a

decade (Baldwin and Clark 1993, 1997a, 1997b; Henderson and Clark 1990), Baldwin and Clark (2000)

advance an evolutionary account of innovation design. The theory encompasses several levels of analysis

and ranges from the “microstructure of designs” to higher levels of economic organization including

social institutions and governments. For our purposes, we concentrate on the microstructure of designs.

The fundamental element of the theory at this level is a model describing layers of structure in the design

of artifacts. The model is depicted in Figure 1.

The LoS model outlines three layers of structure in the design of an artifact: task structure, design

structure, and artifact structure. The artifact structure refers to the tangible instantiation of a particular

design. The artifact can be seen, heard, touched, and used; it performs a set of specific functions that

create value for its user. Second, in order to produce the artifact, designers need to configure design

elements in a specific way. The design structure is a description of the artifact and encompasses both its

structure and functions. For instance, it specifies the hierarchical relationship and interdependences

among design elements. The final layer of structure, the task structure, specifies the list of activities and

6

their interrelationships. It describes the design process and links design activities with particular units of

organizations.

Figure 1: Layers of structure in the design of an artifact (Adapted from Baldwin and Clark, 2000, p. 48)

The innovation literature portrays alignment between organization and product design as

imperative to successful management of knowledge resources and sensemaking processes in developing

complex products (Henderson and Clark 1990; Sosa et al. 2004; Yoo et al. 2006). As suggested in Figure

1, the LoS model refers to this alignment as the fundamental isomorphism of design structure and task

structure. Such isomorphism is reflected in the task structure through centralized communication

channels, information filters, and standardized operating procedures that mirror the product design

(Henderson and Clark 1990), and it suggests that different layers of path creation are interdependent.

3.2 Path Creation Dialectics

The LoS model provides a language to understand layers of path creation in innovation design. In

order to explore path creation as a process, however, we adopt a dialectical view of institutional change

(Benson 1977, Seo and Creed 2002). This view implies understanding an organization as a social entity

that always is in a state of becoming (Benson 1977, Weick 2004, Yoo et al. 2006). The organization is a

dynamic entity embedded in a larger sociotechnical system that constrains its actions. As outlined in the

7

path dependency literature (Arthur 1989, Bassanini and Dosi 2001, David 1985), firms often have

economic, organizational, and technological interdependences both within and across the boundary that

limits its design choices. Over time, these repeated patterns of interactions and interdependence between

firms and units within them create powerful networks of institutional forces (Hargadon and Douglas

2001; Powell 1991). Firms are thus locked-in and path-dependent (Arthur 1989), sometimes falling into

competency traps (Levitt and March 1988). Thus, any organization engaged in innovation needs to

overcome such path dependence in order to transcend its present socio-technical configurations.

Dialectics offers a perspective for understanding the process by which firms break away from the

powerful and systemic force of path dependencies (Boland et al. 2007; Garud et al. 2008; March 1991).

Using the dialectical view, we conceptualize designers in product developing firms as embedded

agents situated in contradictory and multiple layers of artifact, design, and task structures. While

designers are constrained by the dominant design and other sociotechnical configurations that limit design

options, they actively and artfully exploit contradictions in these different layers of structures in order to

seek an alternative future. The dialectical view suggests that “the future is not necessarily a project of the

present order; rather, the future is full of possibilities and one of them has to be made” (Benson 1977, p.

18). Therefore, at the core of a dialectical view is the tension between the familiar path dependencies and

the unfamiliar and uncertain projected innovation paths. This tension constitutes a force that propels

willful and competent human entrepreneurship toward new designs.

A key aspect of dialectics is captured by the notion of contradiction (Poole and Van de Ven 1989;

Seo and Creed 2002). Seemingly stable and coherent relationships between different layers in

sociotechnical structures are only temporary and arbitrary patterns (Benson 1977, Weick 1979). While a

detailed analysis of contradictions is beyond the scope of this paper (Seo and Creed 2002), we suggest the

multi-layered nature of innovation design as an important source of contradictions in product innovation.

As layers of artifact, design, and task are embedded in wider sociotechnical contexts that follow

idiosyncratic evolutionary patterns, seemingly coherent and stable relationships across these structural

layers can be ruptured by inconsistencies and incompatibilities due to the changes in any one of these

8

layers over time. Contradictions are therefore the source of a mindset shift, from an unreflective and

passive mode to a reflective and active one, among designers (Seo and Creed 2002).

Furthermore, the transformation of contradiction is achieved through the praxis of designers

(Benson 1977; Giddens 1984; Seo and Creed 2002). Praxis involves “the free and creative reconstruction

of [sociotechnical] arrangements on the basis of a reasoned analysis of both the limits and the potentials

of present [sociotechnical] forms” (Benson 1977, p. 5)3. Reconstruction of sociotechnical reality

necessarily involves critical self-examination and re-shaping of identity of the organizational actors.

Furthermore, praxis involves not only the reflective moment, characterized by critical awareness of the

current sociotechnical situation, but also the active moment. The active moment draws on the alternative

sociotechnical resources and logic found in surrounding environments.

4. RESEARCH SETTING AND METHODS

CarCorp is a manufacturing firm that produces, markets, and sells around 125,000 cars per year in

primarily Europe and the USA. CarCorp is a fully owned subsidiary of GlobalCarCorp, which is a major

global vehicle manufacturer. The number of employees at the main production plant of CarCorp was

4,500 in 2007. Concurring with GlobalCarCorp’s attempts to streamline their global business including

many other brands than that of CarCorp, business functions are now tightly integrated with

GlobalCarCorp’s global organization. While many areas of R&D have been re-located within the global

firm to avoid redundancy, car infotainment – which is the empirical focus of this paper – is one R&D area

of which CarCorp has been attributed significant global responsibility.

The research approach taken is process-oriented and focuses on the interplay between actors,

context, and technology (Langley 1999, Markus and Robey 1988). The primary source of our research is

a longitudinal study (Pettigrew 1990) at CarCorp over six years (2002-2008). During these years, we have

participated as academic researchers in a number of R&D projects in the car infotainment area. In this

3 In his original definition, Benson (1977) only mentions social arrangements and social forms in defining praxis. In order to emphasize the sociotechnical nature of digital innovations, however, we replaced the word social with sociotechnical.

9

regard, we have chosen to plunge ourselves “deeply into the processes themselves, collecting fine-grained

qualitative data – often, but not always in real time – and attempting to extract theory from the ground up”

(Langley 1999, p. 691). A recurring theme in these projects has been digital convergence and the need of

a new path for car infotainment and telematics. Because this theme was incepted at CarCorp before we

started our field studies, we also made an effort to collect data from the preceding period (1996-2002)

throughout the study.

According to Yin (2009, p. 48-49), a single case study can be useful when the purposes of the

study are revelatory and longitudinal in nature. In our case study of CarCorp, we sought: (a) to examine

dependencies in different layers of innovation design; and (b) to analyze the pattern of path creation

within and across these layers over time. Using a single case, we sought to firmly trace events, activities,

and choices ordered over time in order to build process theory (Langley 1999). This manifests an

understanding of “process as developmental event sequence” (Van de Ven 1992). In view of the four

families of process theories outlined by Van de Ven (1992), we develop a dialectic process theory

because of its appreciation of colliding events, forces, and contradictory values in organizational life (see

also Poole and Van de Ven 1989). In particular, this dialectic orientation resonated well with our ambition

to extend insights documented in the path creation literature about different actors’ colliding paths

(Boland et al. 2007). Such collision is likely to manifest itself in contradictory ways for designers forming

their own innovation path over time.

Our field study included multiple data collection methods including interviewing, participant

observation, and project document analysis. Using Walsham’s (2006) distinction between styles of

researcher involvement, we have been “involved researchers”. Over the different R&D projects, the

involvement has varied between participant observation and action research, suggesting a relatively high

degree of engagement (Nandhakumar and Jones 1997). The main advantage of involved research is that

extended engagement with actors enables researchers to acquire an in-depth understanding of practices

and problems of the world (Van de Ven 2007). The other side of the coin is that researchers risk

becoming “socialized to the views of the people in the field and loses the benefit of a fresh outlook on the

10

situation” (Walsham 2006, p. 322). As it was only the first and the third authors who were directly

involved in the data collection, the second author provided an outside view of the material and its

interpretation.

One important source of data is interviews. All in all, 73 formal interviews have been conducted

over a five-year period. All interviews were tape-recorded and almost all were transcribed, equivalent to

more than 1,000 pages of transcribed interview material. Interviews have primarily been conducted with

CarCorp and GlobalCarCorp engineers and managers working with car infotainment including

applications, platforms, and software architecture. Among the 55 interviews with

CarCorp/GlobalCarCorp personnel, some respondents were interviewed several times over the years. This

especially goes for a number of engineers, referred to as designers in this paper, who were deeply

engaged in R&D projects in the car infotainment area. Because this core group of designers has

participated throughout the entire process, typically in different organizational roles over time, they play a

major role in this case story. In addition, 18 people outside CarCorp and GlobalCarCorp are included in

the interview study. These respondents, working at automotive suppliers, competitors, consultancy

organizations, mobile device manufacturers, and mobile network operators, were all engaged in projects

together with CarCorp.

Participant observation is another important source of data. Over the years, we have participated

in over 50 meetings related to four main R&D projects in the infotainment area. In addition to these

formal meetings, we have conducted frequent visits at various sites of CarCorp and GlobalCarCorp, both

in Europe and North America. Lastly, the study includes significant amounts of archival data including

reports, strategies, and sales forecasts. While many of these documents could not be used directly in this

research because of their confidentiality, it has corroborated interpretations made throughout the analysis

of the data.

Following the suggestions by Charmaz (2006) and Miles and Huberman (1994), we repeatedly

read and coded the data to identify the key themes from major events, activities, and technology choices

over time. Our strategy for theorizing from the process data can be described as a temporal bracketing

11

strategy, i.e., a type of temporal decomposition intended for structuring the process analysis (Langley

1999). As Langley notes, such a strategy is especially useful for incorporating multidirectional causality

into the theorization. It allowed us to organize the data chronologically and across different layers of

structures. We then analyzed the relationship among key concepts, which led to a theoretical

understanding of the interrelatedness among these concepts and their evolution over time. Throughout the

study, we organized many workshops to validate intermediate understandings and to contribute initial

analysis back to CarCorp managers and engineers. The latter aspect is an important element in pursuing

engaged scholarship (Van de Ven 2007).

5. PATH CREATION WITH DIGITAL TECHNOLOGY AT CARCORP: 1996-2008

Some twelve years ago, a design group at CarCorp began exploring the opportunity to connect

the car to external networks and devices in the infotainment4 domain. Recognizing the ongoing

digitization of the car, designers envisioned that novel information and entertainment services could be

offered to customers.

In what follows, we present events, activities, and decisions of this design group’s efforts to

implement the vision of “Car Connectivity” at CarCorp. Along with it, we also present the design group’s

own reflection on the changes in its identity over time. We begin with the background to these efforts,

followed by a detailed description of four distinctive phases of the designers’ path creation.

5.1 Car Connectivity at CarCorp

In the late 90s, the embedded phone was one of the most important infotainment offers at

CarCorp, sold as options in high-end car models. Consistent with most infotainment products, the

architecture of the embedded phone was modular. Different components of the system were integrated by

4 Car infotainment refers to information and entertainment features for drivers and/or passengers. This product segment ranges from basic products such as CD-players and FM/AM radio tuners to more high-end ones such as embedded telephone systems, navigation, and rear-seat entertainment systems. Car infotainment has become increasingly important since infotainment has had a relatively high profit margin compared to other product segments in the automotive industry, especially when sold as options or accessories. Given that GlobalCarCorp made an overall loss the last couple of years, the significant profit made on infotainment is indicative of this fact.

12

CarCorp in a modular fashion. The typical embedded phone system included amplifier, antenna, dial pad,

displays, microphone, telecommunication unit (e.g., a GSM-module), loudspeaker system, and steering

wheel controls. While the responsibility for integrating the main modules (e.g., amplifier and

telecommunication module) was sourced to first tier suppliers, CarCorp was in control of the product

design and offer. This control was manifested in the specification of requirements on the system’s

functionality as a whole, as well as on the interfaces between subsystems.

At the time, designers at CarCorp initiated several R&D projects that would enhance and

complement the embedded design. CarCorp’s first priority was so-called telematics solutions, as they

would extend the embedded phone through additional services such as airbag deployment notifications,

remote diagnostics, and stolen vehicle tracking. Telematics solutions represented a bold initiative of

expanding CarCorp’s business model into services, going beyond the current business model based on

hardware. In addition, telematics solutions followed a conventional view of the design hierarchy.

Over time, however, the embedded phone became a commercial disappointment for CarCorp. It

simply failed to keep up with the explosive growth and rapid developments in mobile

telecommunications. Because components of a embedded phone face tough environmental conditions

(e.g., cold, electromagnetic compatibility, heat, vibrations) and must be reliable over time, all hardware

and software components went through comprehensive and rigorous testing procedures to ensure

durability and to prevent interference with other components in the car. These procedures were costly and

time-consuming. In addition to a costly design process, the embedded phone suffered from low sales

volumes compared to the conventional cell phone.

Over time, nomadic device solutions (NDS) emerged as an increasingly important alternative for

CarCorp’s innovation efforts in the infotainment area. The main difference between the embedded phone

and NDS is that the latter relies on a distributed telecommunication module. That is, instead of

embedding a telecommunication module in the in-car computer platform, NDS include a gateway that

interconnects an external mobile device with the in-car system. In short, the system uses the driver’s (or a

passenger’s) mobile handset as telecommunication module. Even though the initial technological

13

difference between the embedded phone and NDS may appear to be trivial, over time, the designers’

struggle with NDS brought significant changes in CarCorp’s overall infotainment innovation strategy.

5.2 Phase 1, 1996 – 2001: Artifact Layer Tensions

In 1996, CarCorp designers initiated a product development project with Nokia. The idea was to

produce, market, and sell an NDS solution that was compatible with a specific Nokia cell phone. While

the firm’s primary strategic infotainment focus was the embedded phone, the project was sanctioned by

management as a possible alternative for low-end cars.

For the designers, however, NDS was something more. They viewed NDS as a new technical

solution that would increase customer convenience compared to the embedded phone. As noted by an

infotainment designer, who later would become the most dedicated proponent of the new infotainment

path:

A CarCorp customer has a life outside the car too. It is important to think about this [infotainment] from the customer’s point of view: “these are my tools, my personal choices that I make to increase convenience in all situations. They should work together.” If you have chosen a CarCorp car, chosen a personal handset, chosen a PC at work, then you must be able to get these devices to work together.

In order to enable such interoperability, designers collaborated with Nokia engineers to create a

cradle-based solution for docking the cell phone in the car. The cradle interconnected the Nokia device to

in-car resources including audio system, control buttons, power supply, and external antenna. The

application relied on AT commands5 for controlling the communication between the gateway in the car

and the Nokia handset. Figure 2 depicts a simplified version of the NDS architecture with an emphasis on

the interfaces between the vehicle and the Nokia device.

In line with the designers’ intention to increase customer convenience, the artifact structure of the

NDS solution was different from the traditional embedded phone. Rather than defining interface

5 AT is a telephone network modem standard based on a solution developed by Hayes Communications. It is part of the GSM standard.

14

requirements that would be sourced to an automotive supplier, requirements were adapted to interconnect

the off-the-shelf Nokia device.

Figure 2: NDS Architecture 1996-2001

However, the project turned out to be a failure in that no commercial application was launched.

Nokia changed the specifications of the compatible phone model before CarCorp even had the chance to

include the solution in any vehicle roll-out plan. Reflecting on the early attempts, the CarCorp project

manager noted:

The problem was that they [Nokia] changed their interface specifications on several occasions, which made it impossible to put a product on the market. They did not do this in order to create problems for us. Their market is just so much bigger and they wanted to keep their competitive edge in relation to Ericsson and Motorola.

The Nokia project gave the design group at CarCorp early experiences of working with consumer

electronics firms. Unlike traditional automotive vendors, Nokia was not dependent on the cooperation

with CarCorp. Following its own innovation path, the cell phone was a stand-alone product with a much

higher sales volume, making its role as a component in CarCorp’s solution marginal to Nokia. In this

regard, CarCorp designers learned that the instability of the proprietary hardware-based interfaces (e.g.,

cradles) made the traditional product control strategy virtually impossible. An infotainment designer

recalled:

All the time, we were running for developing the right cradles timely, and because devices come and go rapidly and that tools had to be developed for producing the cradles that fitted the specific device, we couldn’t keep up. Because it wasn’t core competence, it was really difficult…

15

CarCorp designers noted that consumer electronics had a much shorter product life cycle and

development time horizon. While CarCorp’s development time horizon spanned at least 3-4 years,

anything beyond 18 months was considered an eternity for Nokia. Such temporal differences, manifested

in the artifact layer of the design, made it difficult for designers at CarCorp and Nokia to collaborate with

the objective of designing stable interfaces between cell phones and cars.

5.3 Phase 2, 2002 – 2006: Forming a New Identity

After the early failure, it was not until 2002 that the design group picked up the NDS idea again.

The NDS designers decided to adopt a new emerging wireless technology standard, the Bluetooth

protocol, to overcome previous interface problems of proprietary cradle-based interfaces. As an

infotainment designer recalled: “it [Bluetooth] gave new life and hope in light of our mechanical

concerns”. The infotainment product manager at the time commented:

Now [2002], the technology is in place. We have overcome the barriers associated with proprietary standards in mechanics, electronics, buses, and so on. General standards such as the Bluetooth protocol now exist, making us believe that this will actually work, also beyond a particular phone model’s life-cycle.

The newborn optimism however coincided with frequent tensions on the designers’ role at

CarCorp. They used to belong to the traditional electrical engineering department, who overall favored

embedded solutions. However, working at the intersection between car infotainment and consumer

electronics, NDS designers began to find themselves caught between two different realms of design.

Their articulation of identity during the phase, thus, reflects this tension. As an infotainment project

manager noted:

“We are [CarCorp is] still a sheet-metal company. We are beginning to learn electronics,… software and stuff… Many people don’t think like the way we, software people, do. The possibilities that we see are not options that they prioritize.”

Therefore, this was the period when a unique identity of the designers who worked on NDS and car

connectivity began to emerge within the electrical engineering department. Reflecting back on this phase,

one of the champions behind the NDS concept commented:

16

”We were rebels. We have always worked on ideas and solutions that have been difficult to appreciate from an automotive perspective. We have always seen ourselves as outsiders in view of mainstream automotive designer.”

Pointing to the promise of the Bluetooth standard, CarCorp designers managed to initiate a number

of new R&D projects related to Bluetooth-enabled NDS. Rather than developing a proprietary solution,

the rationale was now to concentrate on a standard used in consumer electronics. Contrary to established

practice, CarCorp relinquished its control over an essential component (the cell phone), as well as the

communication interface, of the solution by implanting a device-independent solution (see Figure 3).

Stimulated by their regained confidence, designers generated and explored more than 20 different

use cases based on the Bluetooth standard. As a result of the efforts, a comprehensive prototype for

supporting hands-free use of cell phones was developed and later evaluated and verified in actual car use

(cf. Henfridsson and Lindgren 2005). The solution was implemented on the basis of the service discovery

protocol and the handsfree profile of Bluetooth. Demonstrating the solution to top management at both

CarCorp and GlobalCarCorp, designers managed to get the specifications included in vehicle rollout

plans.

Figure 3: NDS Architecture 2002-2006

A significant problem with the design, however, was that it only supported a limited range of

mobile devices. Because the Bluetooth protocols were, and still are, interpreted differently by mobile

device manufacturers, unanticipated interoperability problems emerged for CarCorp when

commercializing the solution. While Bluetooth core protocols were stable and generally consistently

17

implemented, the application profiles such as the hands-free profile and address book were difficult to

handle over multitudes of devices. An infotainment manager ironically reflected upon the unanticipated

problems:

We thought so [that Bluetooth could handle interoperability problems]. Bluetooth got a huge backlash though. If you were an early adopter you run into troubles. CarCorp was a really early adopter [of Bluetooth] in automotive.[…] Standard proved not to be standard. There was a very complex relationship between devices over brands and models, which made the process rather hazardous.

In view of new devices emerging, the Bluetooth-based solution proved inflexible and soon was

outdated for customers. Since it only supported a handful of devices, the system needed to be updated for

each new model year. At its core, the challenge was that there were two parallel design hierarchies (one in

automotive and one in mobile phone manufacturers) with a very weak form of interface between the two,

forming a unstable dyarchy at the task structure later. Designers at CarCorp had to find a better design

and task structures in order to deal with this challenge.

5.4 Phase 3, 2006 – 2007: Recognizing the Need of a New Design Structure

CarCorp now faced increased competition from the consumer electronics industry. The rapid

consumer uptake of a wide range of mobile devices in combination with significant improvement of

functionality, reasonable prices, and portability created significant competitive pressure on the car

infotainment market. Still, the NDS design group felt that the threat posed to core infotainment

applications was not really acknowledged broadly at GlobalCarCorp. The project manager of the early

Nokia project commented:

We are a couple of people who think that this [selling embedded navigation and cd-changers] will not be possible in the future. Except from a particular customer group, top-end customers, who don’t care but tick all available options when buying a car, no customer will select embedded navigation. […] When you have navigation in your pocket, why have an embedded navigation system in the car? You will not have a cd-changer in the car AND a mp3-player in your pocket. We believe that this type of car equipment won’t be there in the future, that the market will disappear for us.

Reflecting upon this conviction, the manager further noted:

18

Now, I should not presume that this is GlobalCarCorp’s official stance. I get a lot of shit for saying this, especially from our marketing people […] they don’t believe in this, they don’t think it is reasonable to think like this. They still believe that it is going to be possible to sell embedded navigation in large volumes, and that it still will be possible to sell CD changers.

As illustrated in the excerpt above, there were tensions between the designers’ view and the

mainstream CarCorp and GlobalCarCorp employee. Yet, NDS design group was slowly gaining attention.

In proposing a new NDS project in 2006, they managed to convince the global infotainment manager pf

GlobalCarCorp, based in the Detroit headquarters, to serve as the chairman of the steering committee.

This was important to sanction the project more broadly. The global infotainment manager underlined his

support of the project:

With regard to a number of functional areas, this project is incredibly important. It is the only current activity in infotainment that is forward-looking. [...] The commercial potential is huge.

However, knowing the NDS design group’s point of view, he still highlighted that NDS was only a

complement to the embedded agenda:

The infotainment market is not threatened. There are some people pretending it is. We are selling more navigation systems than ever before and we are going sell even more next year.

Inspired by the attention received from Detroit, the designers were carefully reviewing their

previous efforts in thinking about the direction of the new project. The Nokia project had shown that

when innovation paths collide, the mobile device manufacturer will prioritize whatever sells more cell

phones. The Bluetooth experience had shown that CarCorp could not keep the pace with the steady

stream of new mobile devices coming out on the market. In view of these disappointments, designers

were turning their attention to designs that could give the possibility to invite outside collaborators.

The idea of devising a flexible NDS design became the mantra of the new project. This broader

scope, beyond specific functionality such as handsfree or navigation, was facilitated by GlobalCarCorp’s

decision to make CarCorp one of two main centers for infotainment R&D. The former manager of the

19

Nokia project was appointed project manager of the project. Reflective of the management support

earned, he reflected upon current systems and the road ahead:

“They [our systems] are very inflexible. If you want to put something into production, it takes three years, almost regardless of what it is. So, if we could get away from hardware solutions, we might address the problem of long lead times for introducing new functionality in the car. […] The hardware used should remain the same over time, while the software modules should enable the adaptation needed.[…] We are envisioning a design that boosts the car’s capacity to handle the digital world. The solution must enable us to follow the technical development in telecommunications during both the construction and production time of the car, which taken together is around 7 years.”

Figure 4: The Open Architecture, 2006-2007

These words of the project manager reveal a radical re-orientation of CarCorp’s attitude in terms

of product design. Rather than holding on to the automotive design tradition, he envisioned a car

architecture that would be malleable to changes in the consumer electronics and telecommunications

world. Much effort was therefore invested in lowering the differences of the in-car platform and outside

technological paradigms. The technical solution to be devised was a device-independent, flexible solution

based on multiple communication channels. Unlike the previous generation that only supported limited

number of cell phone models through a single Bluetooth standard, the new solution was designed to

support new types of devices such as portable navigation systems, portable music players, and portable

DVD players.

20

In order to respond to the pressure from the rapidly changing consumer electronics market,

CarCorp formed a closed project network with few selected partners in order to explore new design

options that would be easily updated over time. Rather than holding onto traditional operating systems

such the UNIX-variant QNX, the project team -- now including online navigation vendor, a system

integrator and a large mobile manufacturer -- built a new platform based on Linux embedded and a Java

Virtual Machine (JVM). The new design was anticipated to make changes to the applications running on

the platform, or addition of new functionality adopted from the telecommunications side, more easily

executed. On the basis of the design, GlobalCarCorp would be able to engage in dedicated projects in

which outside application developers could port their mobile device applications for in-car use.

The project team developed and tested the new flexible platform by porting different applications

directly into the flexible platform. As the project was still in the research and development stage, the team

relied on ad-hoc closed coordination among the participation firms. On the basis of successful results, a

demonstration car was developed where the newly ported applications could be demonstrated. At the turn

of 2007, the results of the project were demonstrated for multiple GlobalCarCorp managers visiting

CarCorp. It was soon decided that the new platform would be included in a series of vehicle roll-out

plans.

5.5 Phase 4, 2008 – : Building External Application Development Community

In view of the success of the flexible platform, the CarCorp designer who managed the

implementation project soon got invited to take on a new infotainment project based on open source

business models. Originating from the European headquarters of GlobalCarCorp, the vision of relying on

open source-like application development and revenue-generation was something that she appreciated as

a natural extension of the flexible platform-project. After all, having 12 years of experience with NDS

design, she, and her design team, viewed the traditional stage-gated and closed project model of CarCorp

as the major hurdle in making the flexible platform a success.

21

The new project was explicitly inspired by Chesbrough et al’s (2006) writings about open

innovation. The project manager commented:

“I very much believe in not trying to solve everything for ourselves, to believe we are the champions, but rather trust in the capacity of others to generate creative and useful ideas. Then we have to manage that creativity so that everyone gains something.”

Because there were no automotive counterparts, the benchmark of the strategy was done in view of

the recent successes of so-called developer programs in the consumer electronics and telecom worlds.

Apple, SonyEricsson, Nokia, Navteq, T-Mobile, and the Android project all exemplify attempts to create

application development communities for generating an ecology of different actors around the promoted

platform. Initially inspired by SonyEricsson’s and Nokia’s developer programs, the idea was to develop

an Application Programing Interface (API)6 and Software Development Kit (SDK)7 that would be

available to application developers becoming members of GlobalCarCorp developer’s community.

In May 2008, the project manager went to the headquarters in Detroit to brief and get feedback on

initial ideas of the project from relevant top executives. Meeting infotainment managers, software

designers, and R&D engineers, the project was not only sanctioned but also attributed additional budget.

As a result of the successful Detroit trip, the project manager decided to propose an additional project that

would not only take the application development community idea to implementation but also investigate

its consequences for work processes at GlobalCarCorp. Looking closer at the internal task structures

including the closed innovation model were deemed necessary but challenging, since it broke

substantially with the proposed new mode of innovation in infotainment.

In September 2008, the strategy was formally sanctioned at the project’s final stage-gate review

meeting including managers from GlobalCarCorp’s worldwide regions. Despite the immediate crisis of

the automotive industry primarily triggered by the world’s financial problems, the idea of going further in

6 API is set of software routines, data structures, object classes and protocols of computer operating systems that are published in order to assist the development of applications. 7 SDK is a set of software development tools including an integrated development environment and debugging tools.

22

this previously untested direction was supported and funded. After 12 years’ struggle, a new path

spanning over artifact, identity, design, and task structures of designing car connectivity was established.

6. DISCUSSION

In this paper, we seek to extend the path creation perspective (Boland et al. 2007, Garud and

Karnøe 2001, 2003) by examining the internal dynamics of path creation process by noting its multi-

layered nature. Specifically, we combine Baldwin and Clark’s (2000) LoS model and a dialectical view of

organization (Benson 1977; Seo and Creed 2002) to explore how embedded designers shift their attention

within and across different design layers as they deal with the reciprocal nature of path creation and path

dependency in their design praxis. In what follows, we first discuss the path creation process at CarCorp.

On the basis of the discussion, we then outline implications for theory and practice.

6.2. Contradictions in Multi-Layer Path Creation

So,what is the layered process by which designers create new paths in digital innovation? Figure

5 depicts the multi-layered dialectic process that we observed at CarCorp and its owner GlobalCarCorp as

designers created new paths. Our findings clearly demonstrate that the digital innovation at CarCorp is not

a singular path creation. When examined closely, it shows that four layers of paths are involved: artifact,

design, task and identity. And, we further note that each layer evolved idiosyncratically, although they

also maintained a certain degree of isomorphism. Our analysis further suggests that the dialectic

contradiction in single layer of path leads to a contradiction across layers. The interlacing of intra-layer

and inter-layer contradictions then acts as propelling force for innovation that moves the digital

innovation forward. Below, we will examine these two contradictions in turn.

23

Figure 5: Path Creation at CarCorp and GlobalCarCorp

6.2.1 Intra-layer Contradictions Consistent with path creation literature (Boland et al. 2007, Garud and Karnøe 2001, 2003),

designers at CarCorp faced contradictions between the ongoing production of new sociotechnical

arrangements and the established sociotechnical order as they attempt to break away from an existing path

of infotainment innovation. Yet, this path was not a single line: instead, it was more like a thread that

consists of interlaced layers of innovation design. Furthermore, designers experienced contradictions at

each layer separately. These contradictions are the tension between what is familiar and what is alien

within the realm of each layer, or the tension between the present innovation order and the projected

innovation trajectory (see Figure 5). What is important to note here is that our data suggests an additional

layer of designer identity as a separate layer in path creation of innovation design. As designers deal with

different artifacts and struggle to find the most appropriate design and task structures, the transformation

of designer identity -- slowly but decisively -- played a critical role.

24

Intra-layer contradictions appear in a staccato-like manner across different layers in different

points in time. For example, in phase 1 (1996-2001), the contradiction between the established logic of

embedded telematics solutions and the unfamiliar logic of NDS was most pronounced at the artifact

structure layer. In phase 2 (2002-2006), a contradiction emerged at the identity layer between the familiar

identity of the automotive-centric engineering group and the newly formed rebel identity by NDS

designers as they tried to break away from the established dominant design of infotainment solutions. In

phase 3 (2006-2008), the contradiction appeared at the design layer became most pronounced. In the

design structure layer, the contradiction between the familiar logic of vertical integration with horizontal

loose coupling and the emergent logic of horizontal integration with vertical loose coupling implemented

emerged as a core issue for the design team. Finally, in the most current phase since year 2008, we noted

a contradiction in the task structure layer between the familiar arrangement of hierarchical control and the

emerging configuration of open polyarchical coordination

While the contradictions within each layer led to a synthesis within the layer which then led to a

path creation at that layer, we further found that such intra-layer contradictions and subsequent changes

fundamentally disturbed the established isomorphism across layers. Such disruption across layers then

instigated contradictions at other layers as we discuss next.

6.2.2 Inter-layer Contradictions and Multi-layered Path Creation A closer examination of the intra-layer contradictions and changes shows that the path creation at

these layers follow idiosyncratic patterns over time. Our findings indicate that contradictions appeared at

a particular layer at a time (see Figure 5). That is, designers do not radically break away from the existing

path at all layers all at the same time. Furthermore, contradictions and subsequent path creation in one

layer introduced new contradictions in other layers that resulted in further changes. For example, in the

first phase, it was the changes in the artifact structure that was most evident, no matter how minor that

change might have seemed at that time. Designers at CarCorp resolved the contradiction at the artifact

structure layer by embracing an NDS solution with Bluetooth interface. However, the embracing of the

NDS solution caused a deep chasm in the identity of the designers in the subsequent phase. With the

25

realization of new possibilities of having an external mobile device, yet frustrated with the unsuccessful

attempts to create a new sociotechnical reality as envisioned jointly with Nokia, the core design team took

on the identity of rebels, breaking away from many of the established norms in the infotainment area.

Here, the contradiction in artifact layer effectuated the path creation at the identity layer. This new

identity then led them to pursue a new path at the artifact structure based on Bluetooth and USB. But,

more importantly, these changes in the identity and artifact structure layers caused effectuated

contradictions in the design structure and task structure layers in the third and fourth phases. The inter-

layer contradictions between the task and design layers led the designers at CarCorp to break away from

the path of traditional horizontal loose coupling with vertical integration to choose a new path of vertical

loose coupling among multiple design hierarchies supported by a task structure with open polyarchical

coordination. CarCorp no longer played the role of product-lead firm that dictated the design hierarchy

with a small circle of handpicked vendors. The future of infotainment was no longer controlled by the

design rule specified by CarCorp. Instead, it became a member of a wider and expanding community that

forms a design polyarchy.

Our analysis of the CarCorp case study shows the path creation by of a firm as it embrace digital

technology. The picture that emerges from our study of CarCorp is that such a path in digital innovation

is not an unproblematic singular straight line, as the name might imply. Instead, it is a problematic thick

thread of multiple layers entangled together (Orlikowski 2007). We see institutional dialectics of path

creation and path dependency at each layer. Contradictions arise and are recede at different pace at

different layers, forming staccato-like pattern of interactions across layers. Therefore, the path creation at

the firm level is not a sudden and swift, almost teleological, shift as the firm embraced digital

technologies. The process is not a clear-cut crisp turn of the path from the “as-is” status to the “to-be”

status. There is no global vision that guides this rather dramatic transformation from a tightly controlled

centralized structure to a distributed and open polyarchical structure. Instead, it is a series of incremental

and localized experiments, in response to the observed contradictions between the familiar and the new

(Orlikowksi 1996). It is a journey that was characterized by frustrating setbacks, surprising discoveries of

26

new tools and technologies, and the emergence of a new identity and different forms of collaboration. As

designers at CarCorp wrestled with the challenges of new technology in infotainment systems, a

contradiction in one layer gives its way to another contradiction in another layer. In a sense, path creation

spilled over from one layer to another. It is as if designers at CarCorp stand upon the pedestal of one layer

where a new path has just been created and discover new future possibilities that have been hidden to

them until they step on that pedestal. As if there is a pool of never-ending riddles, designers at CarCorp

continue to move from one contradiction to another, as they pursue new paths. It is such inter-layer

contradiction that propels ongoing innovation dialectics, as long as organizational actors are willing to

continue to play with them.

6.2.3 Praxis as Design Action Our analysis of CarCorp shows that contradictions within and across layers of structure that are

introduced as a result of path creation at each level are a propelling force of innovations. However, it is

important to note that the presence of contradictions does not deterministically lead to innovations (Seo

and Creed 2002). Our study suggests that design praxis transforms contradictions within and across

layers of structure into new negotiated sociotechnical orders that lead to innovations. In order to

reconfigure established sociotechnical relationships, designers, as the inhabitants of such sociotechnical

relationship, need to reconstruct the current sociotechnical reality. Praxis constitutes the foundation of

such reconstruction processes. Praxis involves a critical reflection of sociotechnical reality as presented in

the form of established orders in the layers of structure, and active reshaping of those established orders in

order to create alternative structures. Designers artfully and actively exploit contradictions within and

across layers of structure to construct new possibilities. In the context of innovations and

entrepreneurship, then, active moments of design praxis follow the logic of effectuation processes

(Sarasvathy 2001; Sarasvathy et al. 2008). Effectuation processes involve an active construction of

possible futures based on the current set of means and identity.

The actions and choices made by the designers at CarCorp show how contradictions across layers

of structure cause designers to be reflective, and thus transform them from passive participants in the

27

reproduction of the existing sociotechnical order into active change agents (Seo and Creed 2002). For

example, the failure of the joint project with Nokia made the designers become aware of the limit of the

established innovation order in which CarCorp occupied a powerful position. Equally important to note

here is that not only did these designers become self-aware of the limit of the current sociotechnical order,

but they also actively pursued alternative sociotechnical orders. Our study shows that the designers at

CarCorp continued to explore new design and task structures, as they became aware of new

contradictions, typically in a different layer, that arose as a result of the changes introduced in the

previous cycle of actions. Therefore, design praxis is an essential micro-level mechanism that underpins

path creation with new technology.

Our study also shows that design praxis is often bound by what they already knew, reflecting the

dialectic relationship between path creation and path dependency. Designers are also constrained by the

material reality of the artifacts that they have built and used until now. CarCorp designers did not know

and could not anticipate new technological options that would become possible in few years. Similarly,

they could not foresee the emergence of open innovation (Chesbrough et al 2006) as a viable innovation

logic just a few years back. Design praxis is localized and distributed in time and space. To designers, the

future is unknowable and open. The design goal is never fully given and disclosed, but need to be

discovered as they continue to explore new structures for artifact, design, and task. To echo Cooren et al.

(2006), in design praxis designers discover the capacity to act “when studying how worlds become

constructed in a certain way.” Our study also shows that designers at CarCorp could not control the

design space as digital technology continues to bring new possibilities, which continue to expand the

design space of car infotainment.

Our finding of the interrelationship between contradictions within and across the layers of

structure and design praxis complements previous studies in digital innovation. Boland et al. (2007) show

that designer’s mindfulness and design visions act as a powerful force that instigates and keeps the wakes

of innovation rippling through the design network. Our finding provides a condition under which

designers become mindful in their design actions. As shown in CarCorp’s case, designer’s mindfulness

28

becomes heightened as the contractions within and across layers of structure develop, deepen, and

penetrate designer’s social experiences (Seo and Creed 2006).

Finally, our study shows that design praxis is inseparable from changes in identity. As designers

at CarCorp search for new structure for artifact, design, and task over time as they strive to transform the

contradictions that they experience, exploring alternative sociotechnical orders, their identities also

evolved as well. In part, the changes in the identity were spurred by the infusion of new personnel

including telecommunication and software engineers in order to cope with the changing knowledge needs.

But at the same time, it was also caused by the keen consciousness of the designers on the changing

nature of the sociotechnical reality in which they are situated. This suggests that the introduction of a new

digital technology into an existing product cannot be accomplished as a mere movement on the same

trajectory of the existing innovation path. It involves discovery of new identity of the designers.

Therefore, the innovation dialectic cycle of contradiction-praxis-transformation-contradiction exacerbate

the continuous state of “becoming” of the organizations that embraces digital technology.

6.3 Limitations

There are several limitations related to our study. First, our findings cannot be generalized across

different types of innovations, as our study focused on the impact digital technology on product-lead

firms relying on modular designs. Product-lead firms have greater responsibility in designing and

integrating products compared to other firms participating in the design process. Therefore, it is likely that

the product-lead firms go through more dramatic changes in its task structure and identity, compared to

their contractors. Second, our empirical study ends in the midst of the fourth phase as CarCorp is

implementing its new direction. Therefore, it is not possible to tell if their open innovation strategy based

on polyarchy of design community will be successful. However, we expect that designers at CarCorp will

become aware of unexpected contradictions as a result of their current efforts. Consequently, they will

exercise their design praxis, continuing to transform the layers of artifact, design, task, and identity.

Lastly, our study is conducted in the context of the automotive industry. The industry is unique in that it is

29

highly concentrated. The pattern of innovation dialectics might be different in a more distributed and

heterogeneous industry.

6.4 Directions for Future Research Our study suggests several implications for future research. First, future research should replicate

our findings in other industries. By studying other industries and different types of products, we can

generalize and validate the innovation dialectics and the emergence of design polyarchy as a result of

digitization. Second, we can adopt other methodologies to examine the general impact of digitization of

physical products on design and task structure. For example, a cross sectional survey across different

industry using econometric methods or agent-based simulation can deepen our understanding on this

phenomenon and generate further theoretical insights. Third, our study focused on the changes in the

identity of a single group of designers over 12 years. Future research can explore the evolution of the

identity of the broader group of actors and the organization itself. Fourth, in this study, we treated the

entire designers as a homogeneous set of actors. However, past research clearly shows that members of

contemporary organizations carry multiple, and often conflicting, identities that are embedded in multiple

institutions. Finally, in this study, we explored the digitization of car infotainment as if it is independent

of the development of other technologies in cars. In reality, infotainment systems are connected to car

communication networks, which themselves are going through rapid changes. This suggests that future

research on the consequences of digitization through an innovation dialectics lens need to carefully

consider the multiplicity of social actors and heterogeneous materiality of different technologies and their

interactions.

7. CONCLUSION The digital technology is often seen as a holy grail for innovation. Our study shows that digital

technology changes not only the innovation path of the artifact and its design layers, but also brings

inevitable changes in the task structure and identity layers. Path creation through digital technology then

30

is multi-layered and much more internally dynamic than it was previously perceived. Invoking the image

of Homo Faber, which means man as his own makers, Sennett (2008) notes the identity-shaping role of

artifacts. Similarly, Orlikowski (2008) notes social and material are constitutively entangled in everyday

life, with materiality of technology is inextricably entangled and interwoven into human agency and

identity. Taking a multi-layered perspective on path creation, our work sheds light on internal dynamics

of such constitutive entanglement between material and social reality of artifacts and the identity of

designers in the innovation process.

REFERENCES Arthur, W. B. 1989. Competing Technologies, Increasing Returns, and Lock-in by Historical Events. The

Economic Journal, 89, 116-131. Baldwin, C.Y., K.B. Clark. 1993. Modularity and Real Options. Harvard Business School Working Paper

93-026. Baldwin, C.Y., K.B. Clark. 1997a. Managing in an Age of Modularity. Harvard Business Review 75(5)

84-93. Baldwin, C.Y., K.B. Clark. 1997. Sun Wars - Competition within a Modular Cluster. D.B. Yoffie, ed.

Competing in the Age of Digital Convergence. Harvard Business School Press, Boston. Baldwin, C.Y., K.B. Clark. 2000. Design Rules - The Power of Modularity. MIT Press, Cambridge, MA. Bassanini, A.P., G. Dosi. 2001. When and How Chance Can Twist the Arms of Clio: An Essay on Path

Dependence in a World of Irreversibilities. R. Garud, P. Karnoe, eds. Path Dependence and Creation. Lawrence Erlbaum Associates, Mahwah, New Jersey, 41-68.

Benson, J.K. 1977. Organizations: A Dialectical View. Administrative Science Quarterly 22(1) 1-21. Boland, R.J., K. Lyytinen, Y. Yoo. 2007. Wakes of Innovation in Project Networks: The Case of Digital

3-D Representations in Architecture, Engineering, and Construction. Organization Science 18(4) 631-647.

Charmaz, K. 2006. Constructing Grounded Theory: A Practical Guide Through Qualitative Analysis.

SAGE Publications, Thousands Oaks, CA. Chesbrough, H. 2006. Open Innovation: A New Paradigm for Understanding Industrial Innovation. H.

Chesbrough, W. Vanhaverbeke, J. West, eds. Open Innovation: Researching a New Paradigm. Oxford University Press, New York, 1-12.

Cooren, F., Taylor, J., & Van Every, E. (Eds.). 2006. Communicating as Organizing: Empirical and Theoretical

Exploration in the Dynamic of Text and Conversation. Mahway, NJ: Lawrence Erlbaum.

31

David, P. 1985. Clio and the Economics of QWERTY. Economic History 75 227-332. Feldman, M., B.T. Pentland. 2003. Reconceptualizing Organizational Routines as a Source of Flexibility

and Change. Administrative Science Quarterly 48 94-118. Garud, R., C. Hardy, S. Maguire. 2008. Institutional Entrepreneurship as Embedded Agency: An

Introduction to the Special Issue. Organization Studies 28(7) 957-969. Garud, R., P. Karnøe. 2001. Path Creation as a Process of Mindful Deviation. R. Garud, P. Karnøe, eds.

Path Dependence and Creation. Lawrence Erlbaum Associates, Mahwah, New Jersey, 1-38. Garud, R., P. Karnøe. 2003. Bricolage versus Breakthrough: Distributed and Embedded Agency in

Technology Entrepreneurship. Research Policy 32(2) 277-300. Giddens, A. 1984. The Constitution of Society - Outline of the Theory of Structuration. Polity Press,

Cambridge. Hanseth, O. 2000. The Economics of Standards. C. Ciborra, K. Braa, A. Cordella, B. Dahlbom, A. Failla,

O. Hanseth, V. Hespø, J. Ljungberg, E. Monteiro, K.A. Simon, eds. From Control to Drift - The Dynamics of Corporate Information Infrastructures. Oxford University Press, Oxford, 56-70.

Henderson, R.M., K.B. Clark. 1990. Architectural Innovation: The Reconfiguration of Existing Product

Technologies and the Failure of Established Firms. Administrative Science Quarterly 35 9-30. Henfridsson, O., R. Lindgren. 2005. Multi-Contextuality in Ubiquitous Computing: Investigating the Car

Case through Action Research. Information and Organization 15(2) 95-124. Langley, A. 1999. Strategies for Theorizing from Process Data. Academy of Management Review 24(4)

691-710. Markus, M.L., D. Robey. 1988. Information technology and organizational change: Causal structure in

theory and research. Management Science 34(5) 583-598. Miles, M.B., A.M. Huberman. 1994. Qualitative Data Analysis. Sage, California. Nandhakumar, J., M. Jones. 1997. Too close for comfort? Distance and engagement in interpretive

information systems research. Information Systems Journal 7 109-131. Orlikowski, W. J. 1996. Improvising organizational transformation over time: A situated change

perspective. Information Systems Research 7(1) 63-92. Orlikowski, W. 2007. Sociomaterial Practices: Exploring Technology at Work. Organization Studies, 28(09):

1435-1448. Pettigrew, A.M. 1990. Longitudinal Field Research on Change: Theory and Practice. Organization

Science 1(3) 267-292. Poole, M.S., A.H. Van de Ven. 1989. Using Paradox to Build Management and Organization Theories.

Academy of Management Review 14(4) 562-578.

32

Sarasvathy, S.D. 2001. Causation and Effectuation: Toward a Theoretical Shift from Economic Inevitability to Entrepreneurial Contingency. Academy of Management Review 26(2) 243-263.

Sennett, R. 2008. The Craftsman. New Haven, CT: Yale University Press. Seo, M.-G., W.E.D. Creed. 2002. Institutional Contradictions, Praxis, and Institutional Change: A

Dialectical Perspective. Academy of Management Review 27(2) 222-247. Sosa, M.E., S.D. Eppinger, C.M. Rowles. 2004. The Misalignment of Product Architecture and

Organizational Structure in Complex Product Development. Management Science 50(12) 1674-1689. Stack, M., M.P. Gartland. 2003. Path Creation, Path Dependency, and Alternative Theories of the Firm.

Journal of Economic Issues 37(2) 487. Thrun, S., A. Levandowski. 2009. Four Ways for Detroit to Save Itself The New York Times,

(www.nytimes.com/interactive/2009/01/03/opinion/20090103_opedthrun.ready.html, accessed: January 16, 2009)

Tsoukas, H. 1996. The Firm as a Distributed Knowledge System: A Constructionist Approach. Strategic

Management Journal 17(Winter Special issue) 11-25. Van de Ven, A.H. 1992. Suggestions for Studying Strategy Process: A Research Note. Strategic

Management Journal 13(Summer Special Issue) 169-188. Van de Ven, A.H. 2007. Engaged Scholarship: A Guide for Organizational and Social Research. Oxford

University Press, Oxford. Walsham, G. 2006. Doing Interpretive Research. European Journal of Information Systems 15(3) 320-

330. Weick, K.E. 1979. The Social Psychology of Organizing, 2 ed. McGraw-Hill, New York. Weick 2004 [Yongjin: please insert correct reference here] Yin, R.K. 2009. Case Study Research: Design and Methods, 4 ed. SAGE Publications, Thousand Oaks. Yoo, Y., R.J. Boland, K. Lyytinen. 2006. From Organization Design to Organization Designing.

Organization Science 17(2) 215-229.