Вы находитесь на странице: 1из 17

Organization Science informs ®

Vol. 18, No. 4, July–August 2007, pp. 631–647 doi 10.1287/orsc.1070.0304


issn 1047-7039 ! eissn 1526-5455 ! 07 ! 1804 ! 0631 © 2007 INFORMS

Wakes of Innovation in Project Networks: The Case of


Digital 3-D Representations in Architecture,
Engineering, and Construction
Richard J. Boland, Jr., Kalle Lyytinen
Department of Information Systems, Case Western Reserve University, 10900 Euclid Avenue,
Cleveland, Ohio 44106 {boland@case.edu, kalle@case.edu}

Youngjin Yoo
Fox School of Business, Temple University, 1810 North 13th Street,
Philadelphia, Pennsylvania 19122, youngjin.yoo@temple.edu

C hanges in the technologies of representation in a heterogeneous, distributed sociotechnical system, such as a large
construction project, can instigate a complex pattern of innovations in technologies, practices, structures, and strategies.
We studied the adoption of digital three-dimensional (3-D) representations in the building projects of the architect Frank
O. Gehry, and observed that multiple, heterogeneous firms in those projects produced diverse innovations, each of which
created a wake of innovation. Together, these multiple wakes of innovation produce a complex landscape of innovations with
unpredictable peaks and valleys. Gehry’s adoption of digital 3-D representations disturbed the ecology of interactions and
stimulated innovations in his project networks by: providing path-creating innovation trajectories in separate communities of
practice, creating trading zones where communities could create knowledge about diverse innovations, and offering a means
for intercalating innovations across heterogeneous communities. Our study suggests that changes in digital representations
that are central to the functioning of a distributed system can engender multiple innovations in technologies, work practices,
and knowledge across multiple communities, each of which is following its own distinctive tempo and trajectory.
Key words: innovation; diffusion of innovation; innovation theory; IT-induced innovation; wakes of innovation; digital
3-D representation; intercalated innovations; trading zones; Frank Gehry; path creation; path dependency; distributed
systems; architecture; engineering and construction

Introduction Frank Gehry’s accomplishments as an architect are


We studied the adoption of digital 3-D representations well known, and he has received many distinguished,
in the design and construction projects of the architect, international awards in architecture, including the pro-
Frank O. Gehry, and observed that his adoption of digi- fession’s highest accolade, the Pritzker Prize. However,
tal 3-D models as a primary representation gave rise to his ability to spawn wakes of technology and process
multiple and diverse wakes of innovation throughout his innovations in the tradition-bound architecture, engi-
project networks. We use the image of a wake to depict neering, and construction (AEC) industries is another
an innovation as emerging in and traveling across an important part of his legacy. Consider, for example, the
innovation space, much as a wake travels through water. following sample of the innovations during the design
Research on technology innovation and diffusion has put and construction of the Peter B. Lewis Building in
considerable effort into studying the wake of a single Cleveland, Ohio (Figure 1).
technology innovation in a typically homogeneous envi-
In that one project:
ronment (Damanpour 1991, Swanson 1994, Zmud 1984,
(a) the structural engineering firm invented a new
Rogers 1995, von Hippel and von Krogh 2003, Bijker
1995, Latour 1997, Williams and Edge 1996). However, method for designing a steel roof with dramatically
in our study, firms involved in Frank Gehry’s project curved surfaces, which won an industry award for engi-
networks produced multiple and distinct innovations— neering innovation;
each of which became a different wake of innovation. (b) the drywall contractor invented multiple patent-
These multiple wakes overlapped on, intruded on, and able ways to frame undulating wall surfaces, and began
interacted with each other in ways that formed a tur- a new line of business, consulting on high-profile con-
bulent, self-propagating system of innovations—a com- struction projects;
plex, undulating surface of diverse innovations that fed (c) the Cleveland fire marshal developed new tech-
back into the project, stimulating further innovations in niques for modeling smoke evacuation, which were then
a staccato fashion. presented at their national training academy;
631
Boland, Lyytinen, and Yoo: Wakes of Innovation in Project Networks
632 Organization Science 18(4), pp. 631–647, © 2007 INFORMS

Figure 1 The Peter B. Lewis Building 5. the precise tolerances increased the risk of loca-
tion errors, leading the construction manager instead of
the contractors to measure the position of all pipes, win-
dows, ducts, etc.; and
6. the precisely curved interior surfaces made the
traditional techniques of acoustic control less desir-
able, stimulating a search for alternative construction
materials.
Note how each innovation responded to the evolving sur-
face of wakes created by prior innovations, and set the
stage for future ones by adding to the turbulence of the
innovation space.
In this paper we seek to develop an explanation of
these wakes of innovation and the recursive quality
of the innovation surface that characterized the Frank
Gehry projects we studied. The innovation turbulence
(d) the specialty metal contractor invented a water- began in Gehry’s studio, as it moved from being com-
proof shingling system that dramatically reduced the puter phobic to being a world-renowned center of digital
thickness and cost of the roofing; 3-D architecture and construction in less than a decade.
(e) the construction manager expanded its scope of In the early 1990s, while other architects were busy
work and begin providing location measurements to con- adopting 2-D computer-aided design (CAD) software,
tractors, increasing its own risk, but reducing construc- Frank Gehry refused to use 2-D CAD tools in his stu-
tion time and construction errors; and dio, in part because he disliked computers, and in part
(f) the drywall contractor became the first American because his design method relies heavily on working by
company to license a soundproof plaster system from hand with physical models. However, by 2000, Gehry
Swiss developers, which created a new line of specialty Partners LLC was regarded as the most advanced user of
business for them. 3-D digital tools in architecture (Lindsey 2001). Behind
that transformation is a fascinating network of innova-
For all these innovations, there exist alternative, tra-
tions in the distributed sociotechnical systems of AEC
ditional methods that the contractors, fire marshal, or
projects.
construction manager could have employed. However,
Gehry Partners’ adoption of digital 3-D technolo-
instead they invented a better way (less costly, more
gies disrupted the ecology of practices that had evolved
elegant, or more reliable) to accomplish their part of
among AEC project teams, allowing professional and
the overall project. Also, these innovations were car-
craft communities to mindfully deviate from their estab-
ried forward into subsequent, non-Gehry projects by lished practices of design and construction, and to create
the innovating firms, thereby strengthening each individ- new sets of self-reinforcing relationships between other
ual organization, as well as the AEC industries. Each actors (Garud and Karnoe 2001). Each firm attended to
innovation can be thought of as creating a wake in their disrupted ecology with reasoning grounded in their
their respective community of practice, but in the Lewis own facts, needs, and specifics (Swanson and Ramiller
Building project, they also served to influence each other 2004). They did not act with a global logic, but with the
as interacting wakes of innovation. Consider the follow- local logic of their own situation and concerns (Boland
ing examples: 2003). Gehry Partner’s adoption of digital 3-D represen-
1. the new roof structure enabled the exterior and inte- tations also altered traditional communication patterns
rior surfaces of the building to directly coincide with through which actors learned about each other’s plans
each other—an unusual condition in a complexly shaped and coordinated their work. Their individual mindful-
building; ness (Weick and Sutcliffe 2006) and new communication
2. the roofing shingle system enabled that parallelism practices allowed innovations that occurred in one com-
between the interior and exterior surfaces to be carried munity to occasionally be inserted into, or intercalated
out with extremely tight precision; with, the unique development process of another com-
3. the precision requirement challenged the interior munity (Galison 1997), which pushed each to expand its
surface contractors to invent new, closer tolerance, own innovation further, while at the same time strength-
curved surface construction techniques; ening interactions among them (Brown and Duguid
4. the curved interiors created a new challenge for the 2000).
fire smoke evacuation studies, stimulating the invention In the remainder of our paper, we first provide an
of new models of air flows for extremely complex inte- overview of the traditional practices in an AEC project
rior surfaces; network. Next, we synthesize a theoretical framework
Boland, Lyytinen, and Yoo: Wakes of Innovation in Project Networks
Organization Science 18(4), pp. 631–647, © 2007 INFORMS 633

that describes how wakes of innovation within AEC research on innovation. Recent research on innovation,
networks can interact with one another as one firm’s tra- however, is beginning to recognize the importance of
jectory of innovation becomes open to discovery, nego- networks and the limitations of studying innovations in
tiation, and knowledge creation with another’s. We then a single, vertically integrated firm. Next, we will exam-
use the proposed framework to analyze our case study ine the emerging body of innovation literature that forms
data, showing how innovations with digital 3-D repre- the basis of the theoretical framework of our study.
sentations were deployed in designing and constructing
Frank Gehry’s buildings, and how they generated multi-
ple other innovations in the AEC system of design and Theoretical Framework of the Study
construction. Our paper concludes by posing implica- Studies of IT innovation have often explored how certain
tions for developing theories of innovation in distributed, factors push (Damanpour 1991, Swanson 1994, Zmud
heterogeneous, sociotechnical networks. 1984) or pull (Rogers 1995) innovations through mar-
kets. More recent work focuses on the user’s active
Architecture, Engineering, and role in pulling innovations, as in continuous innovation
theories (von Hippel 1998, von Hippel and von Krogh
Construction as Context for the Study of 2003), or on the social processes shaping and stabilizing
Innovation an innovation, as in theories of the social construction
Complex building projects serve as good models for of technology (Bijker 1995, Latour 1987, Williams and
the network forms of organizing that are increasingly Edge 1996). They are primarily studies of a particular
found in today’s economy. AEC projects are distributed wake of innovation in relatively homogeneous adopter
(designed and constructed by multiple, autonomous groups.
actors), heterogeneous (composed of communities with Recently, a growing body of innovation research has
distinct skills, expertise, and interests), and sociotech- shifted its focus from a single innovator to a net-
nical (requiring trust, values, and norms, as well as IT work of heterogeneous actors (Tuomi 2002, Van de Ven
capabilities and complex fabrication processes). For ease 2005). When innovation is viewed as being a distributed
of expression, we will simply refer to complex building
phenomenon, it is characterized by network effects,
projects as “distributed systems” or “project networks.”
messiness, ambiguity, and combinability (Lyytinen and
The traditional practice of a standard, AEC design/bid
Damsgaard 2001, Tuomi 2002, Van de Ven 2005).
project begins with an architect working with a client to
Because the network approach assumes an innovation
create a set of drawings (blueprints) and specifications,
space to be dynamic and volatile, it includes political
which indicate the intention and form of what is to be
and institutional influences across the network’s multiple
built. Standard contracts assume that these drawings will
be paper based and two dimensional (Stein et al. 2001). communities. Studies of the communal and networked
These drawings do not show every detail required in features of innovation (Brown and Duguid 2000, von
construction, but instead indicate typical ways in which Hippel 2005) show that increased network diversity pro-
construction elements are to be handled. The architect’s motes new combinations, fosters learning, and enables
drawings leave it to the contractors and subcontractors to faster diffusion (Tuomi 2002), but at the same time,
identify a suitable method of constructing the intended it can build thicker boundaries that inhibit their spread
elements in a workmanlike way. Contractors take the (Ferlie 2005). These scholars stress that networked inno-
architect’s drawings and use them as a basis for cre- vations are associated with new types of interactions
ating their own documents for the architect’s approval. among innovators (Van de Ven 2005), which are enabled
In these “shop drawings,” they show how they intend and mediated by a wider variety of IT artifacts (Carlile
to fabricate and install their part of the building. In 2002). Thus, IT is not just a target of innovation, but is
preparing their shop drawings, contractors ask archi- also an engine of innovation because it increases knowl-
tects questions by submitting RFIs (requests for infor- edge permeability across boundaries, thereby enabling
mation) or RFCs (requests for clarification). Even in new knowledge to be created, and distributing it in
projects of modest complexity, RFIs and RFCs are to be new ways.
expected, and contracts specify the format and timing Traditionally, the pace of innovation in networks is
that is required for making the requests and providing explained by network topologies (Ahuja 2000, Obst-
responses. Within this framework of 2-D representa- feld 2005), which themselves can be transformed by IT
tions and communications, responsibilities for errors are capabilities (Butler 2001). The new emerging patterns
traced to either the architect or the engineer for having of interaction enable further knowledge creation among
specified an inadequate design, or to the contractor for the network’s communities (Tuomi 2002). Indeed, we
having failed to apply the correct workmanlike expertise. observed that new interactions among heterogeneous
The dynamics of these project networks are distinct actors was the modus operandi in shaping the wakes
from the operations of a single firm, even though sin- of innovations that followed Gehry Partners’ introduc-
gle firms have frequently been the site for much prior tion of 3-D representations in their construction projects.
Boland, Lyytinen, and Yoo: Wakes of Innovation in Project Networks
634 Organization Science 18(4), pp. 631–647, © 2007 INFORMS

An emphasis on the interactions and relationships among with respect to the alternatives available in their environ-
heterogeneous actors is particularly important in study- ment, and that they follow reinforced patterns of feed-
ing IT innovations, because IT is more than just the tools back relations in choosing their courses of action and
that are deployed in a project. IT is an environment—a targets of learning (Arthur 1989, March 1991). Path cre-
“web” of equipment, techniques, applications, and peo- ation, in contrast, emphasizes human agency and the
ple that creates a social context, including the history way that actors can and will mindfully deviate from
of commitments that formed that web, the infrastructure what appears to be the common expectation (a tradition-
that supports its development and use, and the social ally reinforced pattern of action) in order to sample new
relations and processes of its use (Kling and Scacchi experience, explore new forms of practice, and create
1982). new resources (Garud and Karnoe 2001).
Recent studies on network-based innovation have ex-
amined interactions and knowledge sharing within at Trading Zones and Intercalation of Innovations
most two communities or one firm (Bechky 2003). Although path creation highlights an actor’s entrepre-
Moreover, these studies primarily concentrated on the neurial deviation from established patterns of practice as
structural features of social networks (Ahuja 2000, a key source of innovation, it neither explains how such
Obstfeld 2005), paying little attention to the role of tech- deviations are sustained and expanded, nor how they
nology in facilitating new forms of interactions across relate to other actors’ path-creating activities. In Gehry
networks that span multiple professional communities Partners’ building projects, there was no lone actor devi-
(Boland and Tenkasi 1995) or entrepreneurial activities ating mindfully. Quite the contrary, multiple actors were
of individual actors who are embedded in such struc- deviating mindfully in response to challenges posed
tural arrangements. Thus, in order to fully account for by Gehry Partners’ digital 3-D representations—often
the complex patterns of innovations that we observed in unexpected and unrelated communities. These com-
in our study of Gehry’s digital 3-D projects, we need munities invented novel uses of 3-D representations,
a theoretical language that allows us to explain diverse crafted new knowledge and skills with them, enacted
actors’ individual innovations, as well as how those new construction practices, and generated new organi-
actors and their activities are interrelated and produce zational forms (Yoo et al. 2006). What we observed
wakes of innovation spreading beyond the boundaries of was a series of abrupt and unexpected breakdowns in
their communities. the established systemic relationships that had evolved
In summary, in order to understand and analyze the over many centuries within and across AEC communi-
wakes of innovation induced by the adoption of digital ties. The AEC industry typically follows long-standing
3-D technologies, we need to understand: (1) how sep- traditions and well-established structural arrangements
arate communities integrate and adopt digital 3-D and in their project governance, in which each element in a
other innovations into their work practices, and (2) how project’s system of relations is interrelated and interde-
and under what conditions such innovations can traverse pendent with others. A significant change in one element
heterogeneous communities. We will address the first in the AEC system can disturb the established pattern
question with the concept of path creation (Garud and of relations among other elements (Churchman 1968).
Karnoe 2001), and the second question with the notions This is especially true when the change affects the prin-
of trading zones and intercalation (Galison 1997), which ciple forms of representation among project actors, dis-
will be explained in the next sections. turbing their ability to communicate and coordinate. The
wakes of innovations that we saw were an expanding
Innovation as Path Creation trial-and-error search for new configurations of relation-
The concept of path creation (Garud and Karnoe 2001) ships among the actors, ideas, and technologies in AEC
is exemplified in Gehry’s willful intention to push the projects. We saw these new relationships take shape in
boundaries of architecture and construction. The theoret- sporadic patterns as actors’ paths crisscrossed at criti-
ical construct of path creation is a reaction against the cal junctures of a project, propelled by the expanded
limitations of path dependence as used in evolutionary deployment of 3-D representations for a specific design
economics (David 1985, Arthur 1989). Although path challenge, or a local construction task.
dependence brings a historical systems view to economic The image of wakes captures the spreading patterns of
explanations of technological innovation, and recognizes mutually dependent and reinforcing deviations that fol-
that temporality and dynamic adaptation to unanticipated lowed engagements with 3-D representations in specific
events are important in understanding how technolog- contexts. However, this image of a sociotechnical system
ical innovations emerge and are adopted over time, it resetting its dynamic equilibrium—no matter how vivid
does not normally include agency as a creative force that it was for us—did not explain why it was unfolding in
makes history by establishing new cycles of positively front of our eyes. We felt a pressing theoretical need
reinforced feedback relations. Path dependence assumes to delve more deeply into this phenomenon. In mak-
that human actors are generally passive or conservative ing sense of the staccato-like spurts of innovation that
Boland, Lyytinen, and Yoo: Wakes of Innovation in Project Networks
Organization Science 18(4), pp. 631–647, © 2007 INFORMS 635

we observed, we drew upon Galison’s (1997) seminal These three concepts—path creation by entrepreneurial
study of the history of microphysics in the 20th century. actors, intercalated innovations across communities, and
Galison reveals the dynamics of concurrent and inde- the intensity of interactions in trading zones at perme-
pendent paths of innovation associated with at least three able boundaries—will be used to explain how and why
communities necessary to do microphysics. These com- wakes of innovations across AEC communities followed
munities are concerned with instruments (measurement the adoption of digital 3-D representations in Gehry
and observation), experiments (design and execution), Partners’ projects.
and rival theories. He shows that each of these commu-
nities innovates at different rates and generates its own
(path-dependent) innovation trajectory. Innovation shifts Research Setting and Method
within and between communities are not simultaneous— According to Yin (1994, p. 40), a single case study can
a shift in theory is not necessarily accompanied by a be useful when the purpose of the study is revelatory
synchronic shift in instrumentation or experimentation. in nature. In our case studies of Gehry Partners’ build-
He refers to this staccato pattern as an intercalated pro- ing projects, we sought: (a) to examine how patterns
cess of change, in which multiple innovations that occur of diverse innovations unfolded within and across dis-
at different rates and times in the different communities tributed systems; and (b) to analyze the mechanisms by
are inserted into the project network. which such innovations take place and interact in their
Galison’s idea of intercalated innovations across mul- project networks. With a single set of cases, we sought
tiple communities offered us a fruitful way to make to firmly ground our theoretical accounts with empir-
sense of the heterogeneous and nonlinear innovation pro- ical observations (Eisenhardt 1989, Martin and Turner
cesses we observed. Like microphysics, the AEC indus- 1986). Overall, we expected these theoretical accounts
tries consist of many distinct communities, each with to offer faithful explanations of the wakes of innovation
a unique professional identity and its associated exper- in the distributed systems of the large AEC projects we
tise, skills, tools, and technologies. These communities studied.
chart their own innovation paths, which follow their During our study, we conducted interviews with the
own logic at their own tempos, and produce intercalated architects, engineers, construction managers, and sub-
changes as their paths cross into each other’s innovation contractors who worked on a range of Gehry Partners’
trajectories. The organization and structure of commu- projects. We visited the home offices of participants
nity boundaries, which either enable or inhibit cross- to review their internal processes, to examine construc-
boundary learning and knowledge sharing, are critical in tion documents with them, and to observe their design
shaping innovation trajectories of each community and and construction practices. The objectives of these inter-
the set of communities as a whole. In particular, we views were to identify: (a) the ways in which the various
need to explore what makes boundaries permeable so actors experienced the digital 3-D representations used
that dynamic learning can take place. by Gehry Partners to be different from their usual doc-
Galison (1997) observes that structural features such umentation and information-sharing practices (i.e., was
as formal organization charts, responsibilities, and con- there a change in routines and knowledge related to their
trol structures, as well as social networks and informal use of digital 3-D representations indicating changes in
communications defined by friendships, mentor-student their trading zones?), (b) the ways in which the actors
relationships, memberships in committees, or traffic adapted to those differences (i.e., how these new repre-
patterns of buildings, all affected the permeability of sentations and boundary exchanges related to their own
boundaries and interaction of knowledge needed to invention of new routines in construction and design
do microphysics. He observed that when community practices), (c) the difference it made to their practice
boundaries cross or overlap during a project, trad- (i.e., did these deviations lead to new and reinforced
ing zones can emerge (Kellogg et al. 2006). In these modes of experiential learning and innovation?), and (d)
trading zones, rudimentary cross-specialty languages can the ways in which they communicated or shared their
develop, allowing formerly distinct knowledge to flow in innovation experiences with other communities (i.e., did
both directions. Galison sees this as similar to the pidgin learning take place in trading zones?).
and creolized languages that emerge when different cul- In collecting the data, we were particularly interested
tures come into sustained close proximity. The stronger in how the deployment of digital 3-D leads to the gen-
and richer the trading zone, the easier it is for learning eration of new knowledge, acquisition of new technolo-
to travel from one community into another, and for com- gies, retraining of employees, hiring of employees with
munities to coproduce new knowledge. Trading zones new skills, restructuring of organizations, changes in
are thus both physical and cognitive arenas for commu- routines, changes in how actors perceived and enacted
nities with separate innovation trajectories to negotiate, their environment, changes in how actors related to
collaborate, and learn through mutual perspective mak- other actors, and ultimately, changes in an actor’s iden-
ing and perspective taking (Boland and Tenkasi 1995). tity. We first focused on how Gehry Partners changed
Boland, Lyytinen, and Yoo: Wakes of Innovation in Project Networks
636 Organization Science 18(4), pp. 631–647, © 2007 INFORMS

their work practices following their adoption of digital Gehry Partners LLP played a central role in innovating
3-D representations. Thereafter, we examined how their with 3-D technologies, and the complexity of Gehry’s
clients, contractors, engineers, and regulators appropri- designs, along with the use of 3-D representations, cre-
ated digital 3-D tools. To accomplish our research goals, ated the trading zones in which the wakes of innovations
we asked Gehry Partners associates (partner architects, occurred. We will next describe the main uses of 3-D in
designers, project architects, and CATIA™ operators) Gehry’s office and the nexus of contexts in which other
how designing and constructing recent buildings (and the actors applied and appropriated 3-D technologies.
Peter B. Lewis building in particular) differed from their
earlier projects, and what mindful deviations took place. Gehry Partner’s Use of Digital 3-D Technologies
Second, we asked contractors and regulators involved Over the last decade the use of digital 3-D representa-
in Gehry projects to compare and contrast that expe- tions such as CATIA™ and Rhino™ software has become
rience with other projects they have had in the past more commonplace in AEC due to the lowered cost and
years. We also asked them about projects subsequent to improved functionality of 3-D tools. In many ways, dig-
their involvement in the Gehry project in order to assess ital 3-D is a potentially disruptive innovation because
ways in which innovations from the Gehry project were it uses enhanced computer capabilities to fundamentally
carried forward into later projects. We videotaped and change the representational systems (or “genres”) by
photographed models and drawings of buildings, and which knowledge about a building is represented, com-
followed demonstrations of how they used 3-D and other municated, manipulated, and shared. It meets the defini-
digital technologies and what changes in their work tion of radical innovation (Zaltman et al. 1973) in that it
practices occurred to help understand how knowledge is a “significant departure from existing practices” cog-
was generated and shared. We started with a set of ini- nitively, materially, and relationally. In what follows we
tial interviewees for each project obtained from sponsors highlight the main functionalities of the 3-D software
or Gehry Partners. We updated and expanded the list tools that have been adopted in Gehry’s office between
through snowballing as we traced the network of rela- 1992–2004, and the benefits they offered (see Lindsey
tions associated with each construction project. 2001 for detailed description of their use of 3-D CAD
A total of 83 interviews—48 interviewees from 16 in their design practices).
organizations—were conducted.1 Most interviews were Digital 3-D representations are more interactive and
transcribed and analyzed verbatim. We also contacted intelligent than 2-D representations (Shih 1996), and
some interviewees afterwards to clarify issues raised in allow more daringly shaped structures to be built cost
the interviews and to collect additional documents from effectively. Three-dimensional tools allow for a full
them. Our data collection branched out from an initial visualization of designs in actual scale, and support
focus on three Gehry Partners’ projects (the Peter B. simulation as well as integration and coordination of
Lewis Building, the MIT Stata Center, and the Princeton detailed design information for digital manufacturing.
Science Library) as we traced the involvement of key When fully developed, a digital 3-D representation is a
actors in earlier and later Gehry Projects. Following complete digital prototype of a building that acts like
these actors forward and backward in time led us to the actual building. A digital 3-D prototype can reflect
study aspects of the following additional Gehry projects: all elements of the design and construction process and
the “Fish Sculpture” at the Barcelona Olympics, the permits different tasks and elements of the building to
Guggenheim Museum in Bilbao, Spain, the Experience
be dynamically interconnected, so that multiple actors
Music Project in Seattle, the Millennium Band Shell in
can coordinate and share their knowledge more openly
Chicago, and the Bard College Center for the Perform-
(Lacourse 2001). Digital 3-D also enhances designers’
ing Arts. Following the recommendations of Eisenhardt
cognition by allowing multiple aspects of a building
(1989), Boyatzis (1998), and Yin (1994), we then repeat-
to be displayed in a single representation and explored
edly read the transcripts and interview notes and viewed
interactively (Baba and Nobeoka 1998). It also enables
the photos and video streams to identify key themes and
designers to visually move back and forth between views
topics in each interviewee’s report of their use of 3-D
of detail and views of the whole in a hermeneutic pro-
representations and their participation in the construction
cess of learning (Boland et al. 1994). In addition, digital
project.
3-D affects how design information is used by project
stakeholders, making previously tacit design knowledge
Innovating with Digital 3-D Representations explicit (Yap et al. 2003) and enabling richer, more
by Gehry Partners detailed knowledge transfer and interactions. This fea-
Due to the distributed nature of the AEC system, path- ture was critical for Frank Gehry because he strives to
creating innovations did not always begin with Gehry fully engage his clients, consultants, and contractors dur-
Partners. Sometimes it was the software developers, ing the design process.
sometimes the contractors, and sometimes the fabrica- Another particularly valuable aspect of 3-D represen-
tors who initiated a new wake. However, we note that tations is that they integrate structural elements of the
Boland, Lyytinen, and Yoo: Wakes of Innovation in Project Networks
Organization Science 18(4), pp. 631–647, © 2007 INFORMS 637

building design (such as size, material properties, and (1997), they were sporadic and opportunistic—forming
cost) with its 3-D geometry (or solid modeling). The and moving between firms as each firm followed its own
combined geometric and structural information can be local logic and innovation trajectory.
stored, modified, and later reused through a repository
(Toupin 2001). Designers can easily manipulate complex In the Beginning: Intercalation of 3-D from
surface geometries, more freely experiment with non- Aerospace into Frank Gehry’s Design Practice
traditional shapes, and quickly explore their costs and Frank Gehry’s design method is based on building phys-
structural implications, which aligns with Gehry’s exper- ical, three-dimensional models by hand, which, ironi-
imental design style (Lindsey 2001). Changes to sections cally, facilitated his appropriation of 3-D technologies.
of an integrated 3-D model can be automatically prop- His method is to always begin with a free-form, impres-
agated (Argyres 1999), showing consequent changes to sionistic sketch intended to convey the spatial and emo-
adjacent sectors of the building, and helping to manage tional sense he is trying to achieve with the building,
complex design processes (Greco 2001). For all these which he calls a “dream image.” (Frank Gehry, Decem-
reasons, 3-D representations have the potential to dra- ber 6, 2004). After that, he and his colleagues build
matically reduce the cost, effort, and error rates of design models with wood, paper, metal, cloth, plastic, or what-
and construction projects (Koutamanis 2000). ever is at hand in an effort to explore the design ideas
Global positioning systems can be combined with 3-D latent in his original sketch. From the earliest massing
tools to change the traditional process of measurement models, to rough study and design models, to the more
at a construction site by creating accurate location data refined final models, there is no drafting on paper and
and managing it centrally as part of a complete digi- no 2-D CAD involved. As one of the staff expressed it:
tal prototype (Gragg 1999). Three-dimensional represen- We persisted much longer in Frank’s office where if you
tations of the building show every point on a line or had a computer, then you weren’t a real architect. You
surface as an "x! y! z# Euclidian coordinate relative to know, you were some kind of drafting support team. (Reg
an established absolute X = 0, Y = 0, Z = 0 point at Prentice, November 13, 2003)
the building site. Instead of locating building elements During a single design project, hundreds of physical
in relation to each other with a tape measure, special- models are constructed to explore design options and
ized surveyors locate sets of absolute points within a alternative forms. They include models of the site and
3-D space,2 thus reducing the possibility of propagating its surroundings, models of interior spaces, models of
measurement errors. Three-dimensional representations the entire building, and models of especially important
can also be used to generate accurate data for guiding elements of the building. The design team works at mul-
downstream fabrication. As a result, Gehry partners have tiple scales simultaneously (usually three) because each
intensively collaborated with a new breed of contractors, scale reveals different problems and possibilities with
like A. Zahner & Co. and Permasteelisa, who are lead- the evolving design. Other architects use physical mod-
ers in the use of advanced digital fabrication techniques els, but not as their primary design and thinking tools
(Koutamanis 2000). as Frank Gehry does. They use physical models primar-
In the next section, we summarize our main findings ily as presentation devices with their clients, not as a
under four headings that emerged from our analysis of tool for thinking during design. In an important sense,
the interviews and documentary data. First, we present Gehry’s emphasis on physical models and lack of expe-
how Gehry Partners adopted 3-D technologies from the rience with 2-D CAD tools meant lower learning barri-
aerospace industry. Second, we examine a series of path- ers in adapting to digital 3-D. When they began using
creating events within Gehry Partners that surrounded 3-D tools, Gehry’s staff did not have to unlearn practices
the initial adoption of 3-D. Third, we explore the uses of that centered on using digital 2-D. In contrast, they were
digital 3-D technologies within the heterogeneous net- able to build upon their physical 3-D design practice by
work of actors in construction projects. These findings digitizing their most precise physical models.
show how trading zones emerged among participating A fortuitous event that enabled Frank Gehry’s adop-
communities and how intercalation of innovations took tion of 3-D came in the late 1980s, when he won a
place. Finally, we discuss how the multiple path-creating competition for the design of the Disney Concert Hall
activities within Gehry Partners’ projects resulted in a in Los Angeles, which he worked on in several phases
stream of innovations among participating firms at dif- during the 90s until the hall finally opened in the fall
ferent times and generated the wakes of innovation we of 2003. Disney Concert Hall was much larger than any
observed. These wakes result from concurrent path cre- of the projects Gehry had undertaken to that date. To
ation events in technologies, work arrangements, con- successfully carry out its design and construction, Frank
struction methods, and forms of knowledge among a Gehry recruited Jim Glymph to join his office as a senior
project’s architects, engineers, construction managers, partner. Jim had extensive experience and was known for
contractors, and trades. These multiple innovations were his design ability, his deep technical knowledge of con-
never synchronized, but similar to Galison’s analysis struction, and his project management skills. Jim firmly
Boland, Lyytinen, and Yoo: Wakes of Innovation in Project Networks
638 Organization Science 18(4), pp. 631–647, © 2007 INFORMS

believed that computer-based tools would be necessary improve the quality of outcomes. The software capabili-
to support the design and construction of the increas- ties of CATIA were just one element in the many inno-
ingly complex surface geometries imagined in Gehry’s vations that emerged across the sociotechnical network
studio. He began to explore computer tools that would of the Fish Sculpture project. The unique work practices
satisfy some or most of their needs, but concluded that that they enacted, with architects, fabricators, and con-
none of the available tools for architectural design (e.g., structors together in one trailer, and the new knowledge
AutoCAD) were suitable. base provided by the 3-D building model, along with
The architectural commission that triggered adoption intense communication protocols, fewer management
of digital 3-D technology at Gehry Partners was the oversight restrictions, and a thoroughly collaborative
design of a pavilion for the 1992 Barcelona Olympics. problem-solving approach, were necessary additional
After completing the Pavilion, Frank Gehry proposed innovations. The Fish project effectively created new
construction of a separate Fish Sculpture on the Olympic trading zones among architects, engineers, and fabrica-
grounds. He had long been fascinated by fish shapes, tors. They learned that the use of 3-D representations
and had developed a physical model of his sculp- allowed project members to participate in frequent trad-
ture, but no detailed plans, when the proposal was ing of expertise and coproduction of knowledge during
accepted. The Fish Sculpture had an extremely tight, all phases of design and construction. This intense level
six-month timetable and a limited budget. At this highly of continuous collaboration and invention has never hap-
constrained moment, Glymph renewed his search for pened again, although it has continued to serve as an
suitable software tools, and found that the CATIA™ sys- ideal “organizing vision” for present and future projects.
tem, which had originally been developed to design the
French Mirage fighter jet, filled many of their needs and Path Creation: Integration and Expansion of
could offer a possible solution. In addition to adopting 3-D into Frank Gehry’s Design Practice
the CATIA™ software, Glymph proposed that they work The experience of the Fish Project taught Gehry Part-
closely with a small team of contractors who would also ners that 3-D tools could be effectively integrated into
use the tool, and they could attempt to do the Fish Sculp- their firm in ways that complemented and empowered
ture project “paperless.” their existing design practice. The 3-D representations
These unique events—the Disney Concert Hall com- offered higher levels of precision and flexibility, gener-
petition, the recruitment of Jim Glymph, and the accep- ated 2-D drawings as a by-product, and also allowed the
tance of the Fish Sculpture project—were, in retrospect, continued use of physical models as the primary design
turning points that set the stage for a significant shift in medium. Digital 3-D tools also enabled them to scale
the trajectory of Frank Gehry’s architectural practice in small physical models up to larger prototypes without
losing the subtleties of a design. As a result, 3-D rep-
ways that were not evident at the time. The Fish project
resentations became a central element of the emerging
enabled them to gain positive experience in using 3-D
digital design practice at Gehry Partners. Two significant
representations through an entire construction project,
steps in developing their digital practice were the design
and to modify their design practice to benefit from 3-D
and construction of the Guggenheim Museum in Bilbao,
tools.
Spain (1991–1997), and the Experience Music Project
During the Fish Sculpture project, Gehry and the con-
(EMP) in Seattle, Washington (1995–2000). However,
tractors deviated from several traditional building prac-
Gehry Partners have remained committed to the use of
tices in order to realize the benefits of using digital 3-D
physical models as the primary medium for exploring
models. As Jim Glymph recounts it:
design ideas:
The first project (using 3-D) was the most successful we When we’ve got a very free-form gestural model, which
may have ever had. " " " And what convinced us was we has all this energy in it, one of the things is to capture it.
got the job done a month early. Everybody made money, And if you go to substandard Euclidian geometry and you
there were no extras – that never happens. That convinced cut sections, and you build this thing up, and you get nice
us that “wait a minute, what we just did because we arcs that are close and everything just seems so close. But
had to, maybe we should be doing this all the time, it when you get the model put back together it’s dead. And
worked.” But the other thing that happened was a lot the measurable differences are tiny and " " " even I was sur-
of (management) rules were suspended in order to allow prised. " " " he (Frank Gehry) tried to create an illusion in
that to happen " " " . (Jim Glymph, November 9, 2002) this building (Bilbao) and it’s a very subtle curvature that
The Fish construction turned out to be a striking you would think would make no difference. The building
without it is completely dead. So there are these gestu-
success—a daring sculpture completed ahead of sched-
ral things, which I think, are why people have emotional
ule and on budget. The project significantly increased reactions to Frank’s buildings. There’s something about
Gehry Partners confidence in the use of digital 3-D tech- building something with your hands that leaves some-
nology, as well as their belief that if they could effec- thing we can all recognize, maybe not consciously, but
tively break from traditional practices within the industry, subconsciously. It humanizes it.
they could create better collaboration on projects and (Jim Glymph, November 9, 2002, p. 4)
Boland, Lyytinen, and Yoo: Wakes of Innovation in Project Networks
Organization Science 18(4), pp. 631–647, © 2007 INFORMS 639

I don’t believe that a computer model will ever be able projects, enabled design and construction knowledge to
to tell me what a physical model can. flow more freely in both directions, and provided oppor-
(Craig Webb, Design Partner, Gehry Partners, Decem- tunities for active trading among the different commu-
ber 22, 2003) nities. Within these newly invigorated trading zones,
The Guggenheim Museum in Bilbao, Spain was their alternative beliefs and knowledge could be negotiated
first use of CATIA™ to support documentation and com- without homogenizing the architect’s intention for the
munications among contractors during a large-scale con- building, or the unique practices within the different
struction project, although 3-D technologies were used communities. (Boland and Tenkasi 1995).
only for the most difficult and critical parts of the The traditional 2-D based process in AEC resembles
exteriors and main gallery designs. In the EMP build- Manning’s (1979) concept of a semiotic loose coupling
ing, Gehry Partners expanded the use of CATIA™ and that characterized interactions between a police com-
employed 3-D representations as the primary contract munication system and people making emergency calls.
documents for the whole building and all its parts. Dur- Like Manning’s police communication system, the stan-
ing EMP, they began to use layered CATIA™ models to dardized 2-D blueprints enabled actors in diverse com-
communicate separately with different trades and con- munities to “conventionalize” their interactions, creating
tractors. During the Peter B. Lewis Building project an adequate basis for coordination (Stinchcombe 1959,
(1997–2002), they used CATIA™ ’s capabilities during Bechky 2006). Actors in AEC projects know how to
design to identify double curved surfaces in complex encode, classify, decode, and transform ideas with the
exterior and interior geometries in order to reduce con- 2-D code shared by all participants. When such 2-D
struction cost. By the time Gehry Partners were com- information crosses boundaries among communities dur-
pleting work on the Lewis Building, their architects had ing a project, it: (a) maintains the integrity, consis-
begun using CATIA™ to create fast prototypes during the tency, and autonomy of each participating actor; and
initial design phase of their projects in order to develop (b) loosens couplings between actors by hiding details of
early and accurate cost estimates. their construction practices that cannot be readily com-
Gehry Partners use of 3-D representations to commu- municated through 2-D drawings.
nicate with contractors and clients throughout design and Such loose coupling is well entrenched in most con-
construction has not decreased their reliance on physi- struction practices, as the project architect for the Peter
cal models—instead, it has added new routines to their B. Lewis building explained to us:
design practice, whereby they analyze constructability, With typical projects, many architects have a standoff
cost, and even maintenance- or energy-related aspects of position from contractors. They just basically enforce
their architectural designs earlier than previously possi- their documents and their specifications and criticize.
ble. During their appropriation of 3-D technologies by And they expect that the contractor knows how to do
Gehry Partners, the horizon of what they feel confi- everything. They don’t talk about process - they just talk
dent to design and build has expanded, and the agency about results. Architects run around and measure. Mea-
they can jointly mobilize by using digital 3-D represen- sure if the deviation is in acceptable limits and write up
reports " " " " (Gerhard Mayer, September 20, 2002, p. 11)
tations has become more powerful and pervasive. This
has resulted in a positive feedback cycle (Levitt and However, when 2-D blueprints were replaced with dig-
March 1988, March 1991) in which learning how to ital 3-D representations at the AEC community bound-
use 3-D technologies throughout the construction pro- aries, the traditional contract language and associated
cess has enabled more daring structural forms to be suc- principles of loose coupling did not provide actors with
cessfully built. The growing sense of confidence is well sufficient ability to understand and negotiate their new
echoed by comments one of the lead architects made roles, or with a sufficient understanding of their risk.
about designing the Peter B. Lewis Building: This is the case even though digital 3-D representa-
" " " CATIA™ makes the whole process and the whole tions carry much more information about the building
method to build it somewhat predictable. And that’s what than the 2-D images do, because the standard contrac-
all our processes are looking for. " " " Without CATIA™ , it tual relations and traditional communication patterns in
would have not been possible to do it (the Lewis Build- the construction industry are inseparable from the use
ing). (Gerhard Mayer, September 20, 2002, p. 8) of 2-D representations at the community borders (Stein
et al. 2001). They may have more information with 3-D
Building Trading Zones with 3-D Tools representations, but they also have less ability to trace
By adopting 3-D representations in their design practice, the cause of an error, or even to agree on what infor-
Gehry Partners disrupted many traditional patterns of mation the representation contains, because the number
interaction that are usually taken for granted within the of perspectives it makes available is indefinitely large.
AEC system, because they shared more information with Thus, the use of 3-D representations at community bor-
more project members earlier. This increased boundary- ders on an AEC project challenged the key organiz-
spanning and knowledge-trading activities during the ing practices and encouraged new interactions in trading
Boland, Lyytinen, and Yoo: Wakes of Innovation in Project Networks
640 Organization Science 18(4), pp. 631–647, © 2007 INFORMS

zones, where knowledge creation across communities conditions before fabricating the next element to be
could take place. installed. In fitting the windows, this would mean mea-
In these trading zones, the architects, engineers, con- suring the actual sizes of the openings in the concrete
struction managers, contractors, and subcontractors could walls before preparing the metal frames and glazing
continually redefine and mutually adjust their relation- for installation. At the Lewis Building project, however,
ships through the use and interpretation of digital 3-D because of the extremely complex geometries of the
representations as “boundary objects” (Star and Griese- walls and the windows, manufacturing of metal frames
mer 1989). Such changes made project management and glazing required more lead time than the traditional
an interdependent and mediated form of organization, approach of field measurement would allow. The con-
displaying what Thompson (1967) calls coordination struction manager recounted this incident as follows:
through “mutual adjustment.” These new project net- So, [the window and glazing subcontractor] came to me
works acted more like tightly coupled systems, in which and says, I can field measure but you won’t get the win-
actors are less independent, with identities that become dow for 16 weeks, and so therefore you can’t put the
blurred as their tasks affect each other continuously, sig- brick in. I can’t do that. I said, you go make it per
nificantly, and directly. CATIA, I’m waiving your requirement to field measure,
A dramatic example of how these new trading zones I’m guaranteeing that concrete opening will be where it’s
operated comes from GQ Contractors, the plaster and supposed to be, and that that window will fit provided
drywall subcontractor on the Lewis Building project. you build it the way the architect designed it " " " " (Dan
Seib, March 4, 2003, p. 42)
Their chief operating officer reported that in his 20 plus
years of working in the industry, he had spent fewer In the future, 3-D representations are expected to en-
than eight hours in architects’ offices. Normally, they gender even more open and intense trading zones as
would receive drawings, analyze them to make a bid, the models become more fully parametric and con-
rework them as shop drawings if they were awarded tain cost, construction time, and other information for
the contract, and do the work. On the Lewis Building each element of the building. Theoretically, such 3-D
project, however, he spent the equivalent of 17 weeks representations will form one central model with all
in the architect’s office, working with 3-D modelers on the information needed by all the contractors and
detailed CATIA™ models to plan how the framing and consultants—information that contractors previously had
drywalling could be done. to create by themselves from the architect’s 2-D
blueprints. A subcontractor commented explicitly about
I have never, ever spent more than an hour in an archi-
tect’s office prior to this job. And I spent 22 trips, 4 and this increasingly open and intense trading within his own
5 days at a time in their office. And I spent some days firm:
where I was in there at 8:00 in the morning and I didn’t It used to be this problem that Zahner had: this con-
get out of there until 10 or 11 at night, working on this crete wall separating the shop from the office, " " " Well
frame. (Ed Seller, September 18, 2002, p. 8) then they built the engineering into the shop. And they’re
going on and moving into the shop now. Then all of a
Similarly, in these newly shaped trading zones, the sudden all the people in the shop are coming in here
temporal organization of work deviated significantly #the office$ mixing in with engineering people and mix-
from that of traditional construction work. The construc- ing people from the trade and so it’s kind of like, tear
tion manager and key contractors were brought into the that wall down. (Bill Zahner, May 17, 2002, p. 22)
process at the very beginning stages of the design, with a
“design assist” contract. During this phase, key contrac- Generating Wakes of Innovation
tors were involved in developing the 3-D models in order The use of 3-D during the design and construction of
to help invent better ways to build parts of the building the projects that we studied expanded and strengthened
and to reduce cost. The design assist contract creates Gehry Partners’ architectural practice. However, con-
a special trading zone for mutual learning, negotiation, structing Frank Gehry’s designs could not have been
and invention between architect and contractor during accomplished without myriad other innovations emerg-
the design process, without a commitment to award the ing from multiple actors in trading zones between the
actual construction contract. different communities that collaborated with them.
Finally, the knowledge created in trading zones can These involved intersecting wakes of innovations in
change the temporal structure of a project. Some activi- a community’s tools, technologies, work practices, and
ties that had been traditionally carried out in a sequen- organization structures and strategies. The multiple inno-
tial order could now be done concurrently, tightening vations and the trading zones that enabled them to
their interdependencies. The window fitting at the Lewis flow between communities emerged from each firm’s
Building provides a vivid example of this. The long- response to the potential of 3-D technologies in light
standing method in construction is a loose coupling of their own unique history and strategy in the project
technique of making “field measurement” of “as built” setting. Not all firms involved in the projects made
Boland, Lyytinen, and Yoo: Wakes of Innovation in Project Networks
Organization Science 18(4), pp. 631–647, © 2007 INFORMS 641

significant adjustments or innovations in response to the not submitting to an altruistic common good. Their
disruption posed by 3-D. Some, like a concrete con- push to innovate was shaped by their professional
tractor and a steel fabricator that we studied, had little identity, unique vocabularies, and craft-specific knowl-
interest in changing their work practices or technologies, edge, rather than by shared identity, common vocabu-
because they believed their traditional practices were laries, and mutual understanding. Innovations in these
sufficient. Others innovated deeply, like A. Zahner and separate communities followed their own temporal logic
Co. (metal fabricators), DeSimone Structural Engineer- and structure rather than aligning with a coherent and
ing, and Columbia Wire and Iron, who were already far synchronized plan. Still, underneath this seemingly ran-
along in their use of 3-D, and pushed even further. Still dom phenomenon, we found systemic and structural pat-
others, agnostics like the construction manager (Hunt terns of interrelations among the innovations generated
Construction) were more cautious in embracing innova- by distinct actors.
tions, accommodating 3-D technology at the work site to Path-creating agency by one actor sets an innovation
the extent that they could effectively work and trade with trajectory for that firm. However, its trajectory can cross
Gehry and other contractors. As a result, they pragmat- those of other actors in an AEC project if specific con-
ically incorporated some lasting innovations into their ditions for trading knowledge are present. Such encoun-
firms, but they did not make any large-scale practice ters can become occasions for change in both actors’
changes. paths, setting up multiple wakes of innovation, while
Across this spectrum of engagement in path cre- each actor continues traversing their own, but now newly
ation, firms also moved at different temporal paces. At shaped, path. The primary example of path-creating
GQ Contractors, for instance, there was an immediate, activity traversing communities in our study is the adop-
early embrace of 3-D technology, some changes in their tion of digital 3-D representations by Gehry Partners.
work practices during construction (including the inven- This change opened a series of new trading zones for
tion of new, patentable framing techniques), and later innovations in tools, techniques, management, and gov-
changes in their structures and strategies. At Hoffman ernance within Gehry Partners’ construction projects,
Construction, we saw a strong commitment to 3-D tech- and reshaped the innovation trajectories of many project
nology at the very beginning of the Experience Music actors. Furthermore, subsequent innovations by those
Project, when they put CATIA™ stations in all subcon- actors brought about changes in Gehry Partners’ own
tractor offices and insisted that they learn and use 3-D. use of 3-D tools, as well as their design and managerial
At the end of the project, they were pleased to say: practices—making the innovation process reflexive and
“even the ditch diggers were using 3-D.” They subse- chaotic. Just as multiple wakes create unpredictable pat-
quently created a new division dedicated to 3-D based terns of interactions when they collide with one another
projects. Examples of the wakes of innovation that we on an otherwise placid surface, we found innovations
saw appearing at different paces in trading zones across from different communities producing nonlinear patterns
communities are summarized below in Table 1. Sim- of interactions.
ilar to the example from the Lewis Building that we
presented in the introduction, many of the innovations Dynamics of Wakes of Innovation
were conditioned by the use of 3-D technologies and Our research suggests that studying IT-induced innova-
were related to other innovations during the project. For tions in distributed systems is not only a question of why
instance, the new form of professional liability insur- and how a particular piece of technology is developed
ance developed for the MIT Stata Center project was and adopted by some users or populations. It is also a
adopted to encourage the free flow of information in question of identifying and accounting for a multiplicity
trading zones, by reducing the threat of litigation from of innovations and their associated cognitive and mate-
using 3-D representations; Hunt’s use of workstations rial processes as they unfold within and across diverse
in their construction site trailer engaged all parties in a communities. This characterization differs from the tra-
more intensive sharing of knowledge; and Hunt’s cen- ditional attempts to explain the diffusion of innovations,
tralized control of measurements offered transparency which have focused on understanding the adoption of
and better understanding of each contractor’s work sta- a single technology application like enterprise resource
tus and interdependencies, reducing location errors and planning systems, or computer-aided software engineer-
rework. ing tools within an organization (Cooper and Zmud
1990, Fichman and Kemerer 1997, Orlikowski 1993),
or among homogeneous firms (Kauffman et al. 2000).
Discussion It also differs from the argument that garnering com-
Throughout our field study we were struck by the stac- plementary assets is necessary to innovate effectively
cato spurts of innovation that followed Gehry Part- (Teece 1987). Although we agree that radical innova-
ners’ use of 3-D technologies. Actors on his projects tions, like the use of digital 3-D representations in AEC,
were pursuing their own economic interests and were will be accompanied by ancillary innovations across
Boland, Lyytinen, and Yoo: Wakes of Innovation in Project Networks
642 Organization Science 18(4), pp. 631–647, © 2007 INFORMS

Table 1 Intercalated Innovations in Three Gehry Partners’ Projects

1. Hunt Construction, Indianapolis, Indiana. (Lewis Building)


(a) Hunt brought CATIA™ workstations into their construction trailer.
(b) Hunt later began using them as a representational tool for morning meetings of subcontractors.
(c) They expanded their scope of work to include providing "x! y ! z# location services for most contractors, who are usually
responsible for their own location measurements.
(d) They adopted the "x! y ! z# measurement system on other projects, bringing the survey work for all location measurements into their
construction management responsibilities.
2. A. Zahner and Co., Kansas City, Missouri. (Lewis Building; MIT Stata Center; Experience Music Project)
(a) For the EMP project, they transformed their fabrication facility for mass customization of metal panels, using 3-D software
to drive machinery.
(b) They invented a new self-draining steel shingle for the Lewis Building roof.
(c) They invented new fasteners to precisely assemble exterior wall panels before welding that were used in the MIT Stata Center.
(d) They invented a way to create double-curved surfaces at comparable cost to single-curved ones, thus freeing a major
design restriction for architects.
3. GQ Contractors of Cleveland, Ohio. (Lewis Building)
(a) They used 3-D workstations at Gehry Partners studio to prepare their shop drawings, but did not bring the technology into
their shop floor.
(b) They invented new, patentable framing techniques for interior walls with curving surfaces.
(c) They developed capabilities with new sound-deadening plaster materials that lead to a new line of specialty business.
(d) They redirected their business strategy toward high-end high-quality projects and expanded into consulting
services with other drywall firms.
4. The Cleveland Fire Inspection Office. (Lewis Building)
(a) They employed new techniques for modeling and evaluating smoke behavior in complex interior spaces.
(b) They presented their Lewis Building experience to their national training academy for teaching new fire inspection techniques.
5. DeSimone Structural Engineers, New York. (Lewis Building)
(a) They invented two new methods for framing structural steel.
(b) They won a national award from their profession for the most innovative structural engineering solution of the year.
6. Columbia Wire and Iron of Portland, Oregon. (Experience Music Project)
(a) They invented a new machine for making double bends in large steel beams, enabling construction of the Experience
Music Project.
(b) They have subsequently become the major source of exotically shaped steel components.
7. Hoffman Construction, Seattle. (Experience Music Project)
(a) They placed CATIA™ workstations in all subcontractors offices and expected their use on the project.
(b) In a subsequent project for the Seattle Central Library by the leading Dutch architect, Rem Koolhaas, they adopted
complimentary 3-D imaging technologies to digitally compare the design in the CATIA models with the “as built”
surfaces of the building, for fine-tuning of precise surface locations.
(c) They have now created a new construction management company to work solely on high-profile, 3-D based
projects with demanding structures and geometries.
8. MIT, Cambridge, Massachusetts. (MIT Stata Center)
(a) The project team obtained a unique insurance policy for professional liability that put all actors on the project team under one
policy, thereby reducing incentives to search for a party to “blame” for errors.
(b) The university adopted a unique payment method that guaranteed contractors a funds disbursement within 24 hours of submitting
an approved invoice (as opposed to 45 days), thereby reducing the incentive to argue over small “change orders,” which could
delay payment by months.

communities, our analysis does not support the claim subsequently shaped and reshaped by multiple, rever-
that such assets can be reliably identified, garnered, and berating innovations in which new knowledge, technolo-
mobilized, as economic analyses would suggest. IT inno- gies, and practices were constantly generated in newly
vations involve a complex ensemble of tools, practices, built and expanding trading zones. The resulting wakes
beliefs, commitments, and history (Kling and Scacchi of innovation go beyond coercive, mimetic, and norma-
1982, Orlikowski and Iacono 2001). Our study shows tive explanations of innovation found in institutional the-
that innovations involving an IT ensemble are sporadic, ory (DiMaggio and Powell 1983). The firms in these
messy, rife with ambiguity, and situated in the local log- networks were not coerced by institutional legitimacy,
ics and trajectories of each project participant. and they did not imitate one another by following others’
We also found that the adoption of 3-D representa- “best practices.”
tions in Gehry Partners’ projects was a complex pro- Our study also portrays how uncertain the diffu-
cess that connected diverse innovations from multiple sion of a complex and radical IT innovation can be.
firms of each project network. The initial IT innovation Whereas the traditional model of diffusion of innova-
of adopting 3-D representations in Gehry’s practice was tion adopts an image of linear progression, with social
Boland, Lyytinen, and Yoo: Wakes of Innovation in Project Networks
Organization Science 18(4), pp. 631–647, © 2007 INFORMS 643

forces pulling and pushing innovations through popula- for new ways of working together on a different set of
tions of adopters (Zmud 1984), an image of wakes and tasks. The more porous boundaries and active trading
intersecting paths paints a different landscape. In this among communities spawned multiple wakes of inno-
landscape it is difficult to pinpoint who is pulling and vations along each community’s innovation trajectory,
who is pushing. In one sense, we can argue for a cen- spreading out from their projects. Whether this more
tralized push, because nothing in this network moved tightly coupled and communicatively intense arrange-
before the challenging designs of Frank O. Gehry were ment will persist, or will become newly conventionalized
embedded in 3-D representations. However, in another around 3-D representations, remains to be seen.
sense, Frank Gehry is not the only highly creative archi-
tect designing buildings with challenging geometries. Driving Forces of Wakes of Innovations
Other architects could have played a similar innovator Our study suggests that the forces behind such processes
role. In addition, we can see a decentralized and dis- are a combination of design vision and mindfulness,
tributed pull as well, because Frank Gehry (or any other together with the dynamics generated in new trading
architect) would not have been able to innovate with zones.
such complex geometries without drawing on relevant
resources that became available outside their practice Design Vision and Mindfulness. Our interviewees
(e.g., CATIA™ 3-D modeling from aerospace, Com- often referred to the criticality of a mindful creation
puter Numerical Control–based fabrication technologies of the new that was shared at some level by all actors
from manufacturing, etc.), and enrolling networks of involved in the network. They were mindful of others
actors willing to experiment across communities. This and of the opportunities for invention that were open to
landscape of innovation can be thought of as a set of them in their interactions (Weick and Sutcliffe 2006).
autonomous, vibrating nodes that create complex sets For Gehry Partners, the source of mindfulness was the
of connections from relatively simple interactions taking architectural vision of Frank Gehry himself, which con-
place in their trading zones (Van de Ven et al. 1999). tinually expands and creates new forms of architectural
Instead of the project networks of AEC becoming more language, exploring new materials, techniques, and tech-
homogenized from the “diffusion” of an innovation, we nologies to fulfill his vision. There is an excitement in
see the landscape of innovations becoming more color- being mindful in this way, which keeps the possibilities
ful and rugged as new actors join the network, with each for innovation open during the construction process:
actor animated by its own unique character and voice at " " " we always involve, we always break the problem
their intersections. down, make sure the building works, what that means,
Studies of single innovations within unique adopter spend a massive amount of time on making it work before
groups are useful in understanding implementation prob- there’s any form giving. People always think we have
lems associated with a specific innovation. However, some idea of what it’s gonna look at the end, and we
in order to understand how IT innovations substan- never do, they make sure we don’t. So, you know, it
tially transform the practices and industrial organization is a process of not knowing where you’re going, you
of distributed systems, they need to be amended with know, you have a few basic things you do about the pro-
cess itself. Probably easier to show you at the office, but
“ecological” studies that trace intercalated innovations
they’re very minimal and beyond that, you’re reacting to
across heterogeneous networks, such as the AEC. Such
the people that you’re, the new client. So every project’s
an ecological approach to IT innovations is necessary new. (Jim Glymph, November 9, 2002)
for studying the increasingly complex technologies that
rapidly digitize wider ranges of tools, activities, pro- The experience of the Fish project provided a glimpse
cesses, and places (Zuboff 1988). In this regard, we of how the managerial and work practices of an AEC
need to examine the means by which innovations emerge project could be accomplished in a more integrated way
and are traded, the likelihood of transferring innovations by using 3-D technologies. It opened a vista of imagin-
across boundaries, the structural properties of these com- ing and experimenting with radical, digital architecture,
munities, and the way that their interrelationships affect and construction practices. As Gehry Partners has tra-
the pace, direction, and scope of innovation. versed its own innovation path over the past decade, this
Our analysis shows a systemic process of innovation initial vision has become more concrete through experi-
in which an initial (path-dependent) condition of loosely ments in multiple construction projects that have helped
coupled, dynamic stability in the distributed system of them garner an expansive suite of competencies.
AEC project networks was disturbed. The convention- We observed the contractors to be mindful of the
alized use of 2-D boundary representations to support strategic question of how their firm could become a
a loose coupling with sufficient communication for the leader or change agent in their craft. Many project par-
distributed system to complete design and construction ticipants, from architects to engineers to craft work-
projects became characterized by tighter coupling in ers, share a path creation impetus and exercise mindful
trading zones, as a subset of the communities searched control over their activities. They are, or can become,
Boland, Lyytinen, and Yoo: Wakes of Innovation in Project Networks
644 Organization Science 18(4), pp. 631–647, © 2007 INFORMS

intrigued by the possibilities of the new, and consistently as a medium for communication and coordination with
push to expand their expertise. This reflects a con- contractors, engineers, builders, and fabricators. Even
scious desire to create something different and better though many leading architects now use advanced 3-D
than what has gone before. Their innovation trajectories tools to generate detailed structural and surface models,
were stimulated by the demands and opportunities of they rarely share these models with contractors and fab-
high-profile projects and the state of development within ricators “due to the concerns related to litigation, risk
each firm. Sometimes they were more advanced in the sharing, or intellectual property rights” (Dale Stenning,
use of technologies, sometimes more collaborative in its Hoffman Construction, December 16, 2003). As a result,
project organization, or sometimes more experimental their use of 3-D tools, no matter how innovative it may
with materials and work practices. be internally, does not have second-order effects in gen-
Mindfulness, then, became a kind of propelling force erating spurts of innovation in the same way we saw in
that kept innovation rippling out from each part of the Gehry Partners’ case.
network—especially for those who progressed further Our study supports the position that IT innovation
in the adoption of 3-D technologies. With a state of does not take place through a single heroic innovator
mindfulness, everyone on the project had the potential or company (Swanson and Ramiller 1997). The fact
to become an innovator. One constructor commented on that Frank Gehry does not use computers and dislikes
this spirit as follows: their screen images shows that 3-D technology adoption
Because the format in which these documents was put at Gehry Partners cannot be explained by his personal
together on this job [it] did not complete the design, and needs or goals. However, Frank Gehry’s role in shaping
we found ourselves completing the design. So being that this innovation is critical because it is his architectural
this was a magnified case of that, now we’re paying more vision and collaborative practices that challenge tradi-
attention to the design of the other jobs to see where tional ways of organizing project work and open trading
the architect left off so that we can address design issues zones where actors can exchange ideas and invent meth-
early on, because a big problem here was the incom- ods for building. He has triggered a socially and tempo-
plete design. The shape was there, we always knew the
shape, but we didn’t know how we were going to create it
rally distributed transformation of artifacts (Tuomi 2002)
because the documents didn’t fulfill all, they didn’t cover over a 10-year period and has reconfigured structural
all the problems. And that’s what they ran into in the relations across a diverse network of communities in the
field, was that they had this set of documents that they AEC industries. Our study also shows how seemingly
typically would use and it would tell them everything separate innovations actually connect in time and space.
where they didn’t have it. And so even the carpenters We did not observe an immediate and direct “balancing”
in the field became part of the design, because they had of such innovations within or across firms in the form of
to figure out ways to do things " " " " Actually, they had instant “gestalt shift.” Instead, we saw the emergence
to help complete some of the design as far as structural of pockets of productive communication within trading
elements. (Ed Sellers, November 11, 2002, p. 13)
zones, where communities who operated within different
Trading Zones and Intercalated Innovations. Al- innovation trajectories learned to negotiate, argue, invent
though design vision and mindfulness provide psycho- and thereby innovate.
logical and cognitive explanations for the path-creating
innovations we observed, they do not provide a structural
explanation as to how such individual innovations came Limitations
together to form wakes of innovations. We found that There are several limitations related to this study. Our
answer in the notions of trading zones and intercalation. findings cannot be generalized across all innovation pro-
Different actors’ innovation paths were interwoven with cesses in the construction industry, because we focused
each other, and innovations from one community were on a leading and early innovator with radically new IT
inserted into another’s innovation path, thus opening a capabilities. First, such early innovators operate in high-
trading zone and changing the trajectory of both paths end markets that can attract unique resource pools (in
(Galison 1997). In addition, certain innovations became terms of architectural, engineering, and IT talent), and
mutually dependent, providing a propelling force for can invest in technology trials and design assist con-
pushing the innovation forward. Without richer trading tracts. Second, our study reports on innovation in a lux-
zones generated by 3-D technologies, a path-creating ury market by a “super star.” It is not clear how fast
innovation would most likely have remained within its and under what conditions such innovations could be
originating community. Thus, it is not only the individ- adopted by other parts of the high-end building mar-
ual actors’ path-creating activities, but also the way they ket, or in other market segments in the AEC industry.
interrelate through trading zones that creates wakes of This would require a better understanding of both cost
innovation. and demand factors related to different types of building
It is also important to note the critical role played projects, and of how the intensive use of 3-D representa-
by Gehry Partners’ explicit intention to use 3-D tools tions can reduce cost across the value chain in different
Boland, Lyytinen, and Yoo: Wakes of Innovation in Project Networks
Organization Science 18(4), pp. 631–647, © 2007 INFORMS 645

pockets of the industry. Finally, our study focuses on of a single wake of innovation, with a network-centric
AEC, which is a distributed industry made up of many perspective that emphasizes multiple, diverse wakes of
independent subtrades, each with their own unique com- innovation emerging in a sporadic and nonlinear way
puting and information practices. The pattern of innova- from trading zones in organizational networks. The out-
tions we observed in our study may be quite different come of such wakes is not a uniform world full of
from the patterns of innovation diffusion in more con- like-minded actors and averages. Instead, it is a vibrant
centrated and centralized industries with a few dominant and interesting world, filled with distinctive colors and
firms. voices, producing everchanging combinations of patterns
of technologies, routines, and cognition.
Directions for Future Research
Our findings suggest several avenues for future research. Acknowledgments
A first research question is when and how do 3-D tech- The authors are indebted to the senior editor and the reviewers
nologies or other novel representations constitute effec- for their very helpful suggestions and patient guidance of this
tive boundary objects in trading zones (Henderson 1991, manuscript.
Star and Griesemer 1989). In particular, we wonder how
Gehry Partners’ representational practices differ from Endnotes
1
those of other highly accomplished architects who do not We interviewed members of the following organizations,
use digital 3-D representations, or use them differently among others: Gehry Partners (Los Angeles), Hunt Construc-
to influence innovation and learning. tion Company (Indianapolis), Hoffman Construction Com-
pany (Seattle), A. Zahner Company (Kansas City), Mariani
Second, 3-D modeling tools are just one of many
Steel Fabricators (Toronto), Cleveland Fire Prevention Bureau,
information technologies that mediate the coordination Cleveland Building Department, Desimone Engineering (New
and communication in distributed systems. It would York), Donnelly Concrete (Cleveland), Spark Steel (Toronto),
be interesting to investigate the generative capabilities Columbia Wire and Iron (Portland), GQ Contractors (Cleve-
of a broad range of digital information technologies, land), Skanska Construction, USA (Boston), Coop Him-
including RFID, product-lifecycle tools, or scientific col- melblau Architects (Vienna), Etalliers Jean Nouvel (Paris),
laborations for enabling wakes of innovation in other Westbrook Architects (Cleveland), and Magnusson Klemencic
distributed networks. Associates (Seattle).
2
Third, our study suggests the opportunity for using With complex buildings like Frank Gehry’s, this increases
other methods in examining wakes of innovations in dis- exponentially the volume and complexity of measurements.
tributed systems. In particular, we anticipate that future For example, in the Peter B. Lewis building the amount of
measures was close to 500,000 points, which were all managed
analyses of wakes of innovation could use formal and
centrally in a separate (Excel) data base.
dynamic approaches such as multimodal network analy-
sis or simulation, which would allow the examination of
References
different patterns of innovation between heterogeneous Ahuja, G. 2000. Collaboration networks, structural holes, and inno-
communities (Monge and Contractor 2003). vation: A longitudinal study. Admin. Sci. Quart. 45 425–455.
Finally, the processes of path creation and interca- Argyres, N. 1999. The impact of information technology on coordi-
lated innovations should be generalized and validated nation: Evidence from the B-2 “Stealth” bomber. Organ. Sci.
in other industries where IT innovation forms a central 10(2) 162–180.
locus for triggering business transformation. Increasing Arthur, W. B. 1989. Competing technologies, increasing returns, and
digitization of products and business processes allows lock-in by historical events. Econom. J. 89 116–131.
for innovations at boundaries by linking heterogeneous Baba, Y., K. Nobeoka. 1998. Towards knowledge-based product
communities in networks. These innovations can cross development: The 3-D CAD model of knowledge creation. Res.
the boundaries of multiple communities, and therefore Policy 26(6) 643–659.
the network perspective followed in our study can be Bechky, B. A. 2003. Sharing meaning across occupational communi-
helpful. Such studies could help identify organizational ties: The transformation of understanding on a production floor.
conditions and processes for generating wakes of inno- Organ. Sci. 14(3) 312–330.
vation. For example, mobile multimedia, biotechnology Bechky, B. A. 2006. Gaffers, gofers, and grips: Role-based coordina-
or advanced logistic systems can offer interesting fields tion in temporary organizations. Organ. Sci. 17(1) 3–21.
to examine. Bijker, W. E. 1995. Of Bicycle, Bakelites, and Bulbs: Toward a Theory
of Sociotechical Change. The MIT Press, Cambridge, MA.
Boland, R. J. 2003. An ecology of distributed mediated cognition.
Conclusion C. Ciborra, C. Angelou, eds. Information Systems and Social
We hope to stimulate a shift in thinking about diffu- Theory. Cambridge University Press, Cambridge, MA, 119–128.
sion of IT-induced innovation in networked contexts. Boland, R. J., R. V. Tenkasi. 1995. Perspective making and per-
We want to augment the individual and firm-centric spective taking in communities of knowing. Organ. Sci. 6(4)
perspectives on innovation that focus on the diffusion 350–372.
Boland, Lyytinen, and Yoo: Wakes of Innovation in Project Networks
646 Organization Science 18(4), pp. 631–647, © 2007 INFORMS

Boland, R. J., R. V. Tenkasi, D. Te’eni. 1994. Designing information Levitt, B., J. G. March. 1988. Organizational learning. Annual Rev.
technology to support distributed cognition. Organ. Sci. 5(3) Sociol. 14 319–340.
456–475.
Lindsey, B. 2001. Digital Gehry: Material Resistance, Digital Con-
Boyatzis, R. 1998. Transforming Qualitative Information: Thematic struction. Birkhauser, Basel, Switzerland.
Analysis and Code Development. Sage, Thousand Oaks, CA.
Lyytinen, K., J. Damsgaard. 2001. What’s wrong with the diffusion
Brown, J. S., P. Duguid. 2000. The Social Life of Information. Harvard of innovation theory: The case of a complex and networked
Business School Press, Boston, MA. technology. M. A. Ardis, B. L. Marcolin, eds. Diffusing Software
Butler, B. S. 2001. Membership size, communication activity, and Product and Process Innovations. Kluwer Academic Publishers,
sustainability: The internal dynamics of networked social struc- Norwell, MA, 173–190.
tures. Inform. Systems Res. 12 346–362. Manning, P. 1979. Producing drama: Symbolic communication and
Carlile, P. R. 2002. A pragmatic view of knowledge and bound- police. Symbolic Interaction 5 223–241.
aries: Boundary objects in new product development. Organ. March, J. G. 1991. Exploration and exploitation in organizational
Sci. 13(4) 442–455. learning. Organ. Sci. 2(1) 71–87.
Churchman, C. W. 1968. The Systems Approach. Dell, New York.
Martin, P. Y., B. A. Turner. 1986. Grounded theory and organizational
Cooper, R. B., R. W. Zmud. 1990. Information technology imple- research. J. Appl. Behavioral Sci. 22(2) 141–157.
mentation research: A technological diffusion approach. Man-
Monge, P. R., N. S. Contractor. 2003. Theories of Communication
agement Sci. 36(2) 123–139.
Networks. Oxford University Press, New York.
Damanpour, F. 1991. Organizational innovation: A meta-analysis of
effects of determinants and moderators. Acad. Management J. Obstfeld, D. 2005. Social networks, the tertius iungens orienta-
23(3) 555–590. tion, and involvement in innovation. Admin. Sci. Quart. 50(1)
100–130.
DiMaggio, P., W. Powell. 1983. The iron cage revisted: Institutional
isomorphism and collective rationality in organizational fields. Orlikowski, W. J. 1993. CASE tools as organizational change: Investi-
Amer. Sociol. Rev. 48(2) 147–160. gating incremental radical changes in systems development. MIS
Quart. 17(3) 309–340.
Eisenhardt, K. M. 1989. Building theories from case study research.
Acad. Management Rev. 14(4) 532–550. Orlikowski, W. J., C. S. Iacono. 2001. Desperately seeking the “IT” in
IT research: A call to theorizing the IT artifact. Inform. Systems
Ferlie, E. 2005. Conclusion: From evidence to actionable knowledge?
Res. 12(2) 121–134.
S. Dopson, L. Fitzgerald, eds. Knowledge to Action? Evidence-
Based Health Care in Context. Oxford University Press, New Rogers, E. 1995. The Diffusion of Innovation, 4th ed. The Free Press,
York, 182–197. New York.
Fichman, R. G., C. F. Kemerer. 1997. The assimilation of soft- Shih, N.-J. 1996. A study of 2D and 3D oriented architectural drawing
ware process innovations: An organizational learning perspec- production methods. Automation Construction 5 273–283.
tive. Management Sci. 43(10) 1345–1363.
Star, S. L., J. R. Griesemer. 1989. Institutional ecology, “translations”
Galison, P. 1997. Image and Logic: A Material Culture of Micro- and boundary objects: Amateurs and professionals in Berkeley’s
physics. The University of Chicago Press, Chicago, IL. Museum of Vertebrate Zoology, 1907–1939. Soc. Stud. Sci. 19
Garud, R., P. Karnoe. 2001. Path creation as a process of mindful 387–420.
deviation. R. Garud, P. Karnoe, eds. Path Dependence and Cre- Stein, S. G., R. E. Alexander, C. L. Noble. 2001. AIA general condi-
ation. Lawrence Earlbaum Associates, New York, 1–38. tions in the digital age. Construction Contracts Law Rep. 25(25)
Gragg, R. 1999. Museum design tests Hoffman’s learning curve. The 3–20.
Oregonian. Portland, OR. Stinchcombe, A. 1959. Bureaucratic and craft administration of pro-
Greco, J. 2001. CATIA Version 5 Release 7. Cadence 16(12) 26–31. duction: A comparative study. Admin. Sci. Quart. 4 168–187.
Henderson, K. 1991. Flexibile sketches and inflexible data bases: Swanson, E. B. 1994. Information systems innovation among organi-
Visual communication, conscription devices, and boundary zations. Management Sci. 40(9) 1069–1092.
objects in design engineering. Sci., Tech. Human Values 16
Swanson, E. B., N. C. Ramiller. 1997. The organizing vision in infor-
448–473.
mation systems innovation. Organ. Sci. 8(5) 458–474.
Kauffman, R. J., M. James, Y.-M. Wang. 2000. Opening the “black
Swanson, E. B., N. C. Ramiller. 2004. Innovating mindfully with
box” of network externalities in network adoption. Inform. Sys-
information technology. MIS Quart. 28(4) 553–583.
tems Res. 11(1) 61–82.
Kellogg, K. C., W. J. Orlikowski, J. Yates. 2006. Life in the trading Teece, D. 1987. Profiting from technological innovation: Impli-
zone: Structuring coordination across boundaries in postbureau- cations for integration, collaboration, licensing, and public
cratic organizations. Organ. Sci. 17(1) 22–44. policy. D. J. Teece, ed. The Competitive Challenge. Ballinger,
Cambridge, MA, 285–305.
Kling, R., W. Scacchi. 1982. The web of computing: Computer tech-
nology as social organization. Adv. Comput. 21 1–90. Thompson, J. D. 1967. Organizations in Action. McGraw-Hill,
New York.
Koutamanis, A. 2000. Digital architectural construction. Automation
Construction 9 347–360. Toupin, L. A. 2001. Knowledgeware makes reusing existing designs
a snap. Design News 56(20) 42–53.
Lacourse, D. 2001. CATIA V5 r6 serves up robust tools for mechan-
ical engineers. CADalyst 18(12) 32–36. Tuomi, I. 2002. Networks of Innovation: Change and Meaning in the
Age of the Internet. Oxford University Press, New York.
Latour, B. 1987. Science in Action: How to Follow Scientists
and Engineers Through Society. Harvard University Press, Van de Ven, A. H. 2005. Running in packs to develop knowledge-
Cambridge, MA. intensive technologies. MIS Quart. 29(2) 368–378.
Boland, Lyytinen, and Yoo: Wakes of Innovation in Project Networks
Organization Science 18(4), pp. 631–647, © 2007 INFORMS 647

Van de Ven, A. H., D. E. Polley, R. Garud, S. Venkatraman, eds. 1999. Yap, A. Y., O. Ngwenyama, K. M. Osei-Bryson. 2003. Leverag-
The Innovation Journey. Oxford University Press, New York. ing knowledge representation, usage, and interpretation to help
von Hippel, E. 1998. Economics of product development by user: reegineering the product development cycle: Visual computing
Impact of sticky local information. Management Sci. 44(5) and the tacit dimension of product development. Comput. Indust.
629–644. 51 89–110.

von Hippel, E. 2005. Open source software projects as user innovation Yin, R. K. 1994. Case Study Research: Design and Methods. Sage,
networks—no manufacturer required. J. Feller, B. Fitzegeral, Thousand Oaks, CA.
S. Hissam, K. Lakhani, eds. Perspectives on Free and Open Yoo, Y., R. J. Boland, K. Lyytinen. 2006. From organization design
Source Software. MIT Press, Cambridge, MA, 267–278. to organization designing. Organ. Sci. 17(2) 215–229.
von Hippel, E., G. von Krogh. 2003. Open source software and the Zaltman, G., R. Duncan, J. Holbeck. 1973. Innovations and Organi-
“private-collective” innovation model: Issues for organiation sci- zations. Wiley, New York.
ence. Organ. Sci. 14(2) 208–223. Zmud, R. W. 1984. An examination of “Push-Pull” theory applied to
Weick, K. E., K. M. Sutcliffe. 2006. Mindfulness and the quality of process innovation in knowledge work. Management Sci. 30(6)
organizational attention. Organ. Sci. 17(4) 514–524. 727–738.
Williams, R., D. Edge. 1996. The social shaping of technology. Res. Zuboff, S. 1988. In the Age of the Smart Machine: The Future of
Policy 25 865–899. Work and Power. Basic Books, New York.

Вам также может понравиться