Complex Project Management as Complex Problem Solving: A Distributed Knowledge Management Perspective

See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/306344720
Complex project management as complex problem solving: A distributed
knowledge management perspective
Article in International Journal of Project Management · December 2014
DOI: 10.1016/j.ijproman.2013.06.007
CITATIONS
38
READS
1,016
3 authors, including:
Some of the authors of this publication are also working on these related projects:
CACTOS - Context Aware Cloud Topology Optimisation and Simulation View project
SIMCT - Simulation-based Contract Costing for Outsourcing Enterprises View project
Terence Ahern
Dublin City University
5 PUBLICATIONS 71 CITATIONS
SEE PROFILE
P. J. Byrne
Dublin City University
82 PUBLICATIONS 865 CITATIONS
SEE PROFILE
All content following this page was uploaded by Terence Ahern on 13 January 2018.
The user has requested enhancement of the downloaded file.

Complex project management as complex problem solving:
A distributed knowledge management perspective
Terence Ahern ⁎, Brian Leavy, P.J. Byrne
Dublin City University Business School, Dublin City University, Ireland
Available online 2 July 2013
Abstract
Traditional project management (PM) privileges planning and downplays the role of learning even in more complex projects. In contrast, this
paper draws inspiration from two organisations that were found to have developed complex PM expertise as a form of complex problem solving
(CPS), a practice with implicit learning because complex projects are unable to be completely specified in advance (Hayek, 1945). Central to this
view of complex project management as a form of complex problem solving is the governance challenge of knowledge management under
uncertainty. This paper proposes that the distributed coordination mechanism which both organisations evolved for this contingency can best
be characterised as a ‘common will of mutual interest’, a self-organising process that was fostered around project goals and paced by the
project life cycle (Kogut and Zander, 1992). The implications for theory, research, and practice in complex PM knowledge management are
examined.
© 2013 Elsevier Ltd. APM and IPMA. All rights reserved.
Keywords: Complex project management; Distributed knowledge management; Bounded planning; Problem solving; Common will
1. Introduction
In the project management (PM) literature, the management of
complex projects as an important focus for more intensive
research is an emerging tradition, along with the need to
understand the particular governance challenges associated with
it (Baccarini, 1996; Miller and Hobbs, 2005; Morris and Hough,
1987; Müller, 2009). This research paper highlights and
examines knowledge management as a key aspect of governance
in the case of complex projects, based on an empirical study of
complex project management featuring two Irish state-owned
organisations, referred to here as GovCo-1 and GovCo-2. In the
late 1990s and early 2000s, each of these complex organisations
(Pollitt and Bouckaert, 2000; Thompson, 1967) was challenged
to take on major infrastructural development projects of a scale
and complexity well beyond what had been the norm for either
organisation up to then. In GovCo-1, the stimulus was provided
by the government's National Development Plans for infrastruc-
ture investment (NDP, 2000, 2007) and the stimulus for GovCo-2
was provided through EU deregulation in the energy sector. In
this context, GovCo 1&2 provided a valuable opportunity to
explore more closely in what ways the management of ‘complex’
projects differs most from that of other kinds of projects reflected
in the mainstream PM literature (APM, 2011, 2012; PMI, 2013).
The main empirical finding was that complex project
management (PM), as manifested in the two organisations
under study, could best be understood as a form of complex
problem solving (CPS) that does not lend itself to being
completely specifiable in advance. In the mainstream PM
literature, such projects undertaken by GovCo 1&2 tend to be
viewed as just more ‘complicated’ projects that can still be
planned and managed in the traditional way as “the application
of knowledge, skills, tools, and techniques to project activities
to meet the project requirements” (PMI, 2013, p. 5, italics added).
In this approach, there is little learning anticipated beyond
the application of prior knowledge. In contrast, the empirical
finding that complex PM is a form of complex problem solving
⁎ Corresponding author at: Dublin City University Business School, Dublin
City University, Ireland. Tel.: [phone].
E-mail addresses: [e-mail] (T. Ahern),
[e-mail] (B. Leavy), [e-mail] (P.J. Byrne).
www.elsevier.com/locate/ijproman
0263-7863/$36.00 © 2013 Elsevier Ltd. APM and IPMA. All rights reserved.
http://dx.doi.org/10.1016/j.ijproman.2013.06.007
Available online at www.sciencedirect.com
ScienceDirect
International Journal of Project Management 32 (2014) 1371 – 1381

(CPS) means that managing project knowledge becomes more
problematic.
In terms of governance, this alternative perspective means
that a central aspect of knowledge management in complex PM
settings involves managing intrinsic knowledge uncertainty.
This is manifest as incomplete pre-given knowledge in complex
projects that necessarily limits complex PM to ‘bounded
planning’, which implies the need in complex PM to
continuously create knowledge over the project life cycle that
is not specifiable at the outset (Engwall, 2002). This, in turn,
requires the development of an effective mechanism for
coordinating this emergent knowledge. In the cases of GovCo
1&2, both were found to have evolved a distributed governance
approach to knowledge management that revolved around
problem solving as a mode of learning and organising. In effect,
in order to create project knowledge that was unspecifiable at
the outset in project designs, plans, etc., the project team
became a community of learners that was learning the project
though organisational CPS. In order to coordinate this emergent
knowledge, GovCo 1&2 harnessed the agency of what this
paper terms a ‘common will of mutual interest’ that was
fostered around project goals and paced by the project life
cycle. This can be thought of as a high-level organising
principle that is irreducible to individual project actors (Kogut
and Zander, 1992), by which the project team can know more
than it can tell (Polanyi, 1967) and can know more than its
individual members can know separately.
The full empirical inquiry that led to these findings is
reported elsewhere (Ahern, 2013), which is an exploratory case
study investigation of two complex organisations in the public
sector using a Contextualist research perspective that includes
51 semi-structured interviews (Pepper, 1942). This longitudinal
process approach facilitates the study of the development of
organisational processes that are ‘in flight’ during periods of
important change in organisations (Pettigrew, 1990, 1997,
2012). The primary purpose of this paper is to examine some of
the main conceptual and practical implications for the
traditional PM literature associated with the above two
important empirical insights in complex PM, namely, incom-
plete pre-given knowledge and coordinating emergent knowl-
edge. This will be done by reviewing the literature on related
themes and drawing on further findings from the data
(Siggelkow, 2007).
The remainder of the paper is organised as follows:
Section 2 reviews the literature on complex PM with particular
attention to the contrast in knowledge management assump-
tions between traditional PM and those implied by viewing
complex PM as complex problem solving (CPS). In addition,
learning modes are reviewed for generating knowledge in
complex PM, which can be coordinated through a distributed
organising approach. Section 3 discusses the implications for
governance in complex PM of knowledge management as a
process of learning and organising under ‘bounded planning’
rather than ‘total planning’ assumptions. This includes the
scaffolding of distributed learning and organising using
documented procedures as well as the fostering and pacing of
a common will of mutual interest for coordinating emergent
project knowledge. In Section 4, the concluding section, the
implications of inherent knowledge uncertainty in complex
PM as a form of organisational CPS are discussed in relation
to the following areas of research and practice: (i) planning,
knowledge creation, and knowledge coordination; (ii) leader-
ship; (iii) knowledge transfer; and (iv) PM complexity.
2. Complex project management as complex problem solving
Informed by the two empirical findings highlighted earlier, this
section will review the literature on complex projects in relation to
the management of knowledge under the traditional PM
paradigm, which assumes full pre-given knowledge, and under
more recent pragmatist perspectives of PM, which accept the idea
of incomplete pre-given knowledge in projects and the need for
learning. In this, a distinction will be made between ‘complicated’
projects that can be completely specified in advance and
‘complex’ projects that are unable to be completely specified in
advance. Finally, different modes of problem solving learning are
discussed, including complex PM as a form of organisational
CPS, which facilitates the creation of emergent knowledge that is
un-specifiable at the outset; and the coordination of this emergent
knowledge through what this paper terms a ‘common will of
mutual interest’ as a distributed tacit dimension. This term is new
to the literature and is inspired by an interaction between the case
study data and the literature to represent the synergy that is
achieved in projects when a team spirit is successfully fostered to
the extent that it becomes self-reproducing as a common will
around an interest that is mutually desired and experienced. In this
way, it becomes a self-organising process for coordinating the
behaviour and, hence, the collective learning of project teams in
complex PM settings.
2.1. Complex PM as applied science — planned knowledge
In early work on the complexity of project settings, Shenhar et
al. (1995) distinguish two dimensions of project complexity—
‘technological uncertainty’ and ‘system scope’. This typology is
used in advocating a contingency approach to PM (Lawrence
and Lorsch, 1967; Shenhar, 1998, 2001; Shenhar and Dvir,
1996), rather than the “one size fits all” approach of traditional
PM (Shenhar, 2001, p. 394). In subsequent research, Shenhar
et al. (2002) extend the framework to encompass three di-
mensions of project complexity, namely, ‘uncertainty’, ‘pace’,
and ‘complexity/scope’ (UPC Model), where ‘pace’ is added to
reflect the “speed and criticality of time goals” (ibid., p. 101).
Implicit in this research is the assumption that knowledge relating
to project complexity can be analysed and integrated as ‘tech-
nical’ complexity under the norms of technical rationality
(Ashby, 1956; Cleland and King, 1968; von Bertalanffy, 1950),
rather than as ‘social’ complexity that requires a socio-technical
approach (Davies and Hobday, 2005; Nightingale and Brady,
2011; Sapolsky, 1972; Williams, 1999, 2005). Under the former
approach, knowledge is detached from the knowing subject like a
commodity and is pre-given at the outset, while, under the latter,
knowledge is integrated with the knower as a process of knowing
over time, because it is not completely pre-given at the outset.
1372 T. Ahern et al. / International Journal of Project Management 32 (2014) 1371–1381

In recent literature on complex PM, scholars have sought to
incorporate insights from research in complexity, chaos, self-
organising, and evolution with traditional PM (Cooke-Davies et
al., 2007; Geraldi et al., 2011). This emerging area in PM is termed
‘complex project management’ (Whitty and Maylor, 2009). For
example, Saynisch (2010a,b) analyses complexity using the two
dimensions of ‘project complexity’ and ‘environmental complex-
ity’ and calls for a governance approach that integrates the two
cybernetic cycles of traditional PM and the management of
complexity (evolution, self-organisation, edge of chaos). Interest-
ingly, he maintains that getting the balance right between these two
“will be the future management art” (Saynisch, 2010b, p. 8, italics
added), which suggests that, in order to deal with situations of
project complexity, PM may have to reposition its self-image to
one of a craft with implied learning rather than an applied science
under the traditional PM paradigm with little learning anticipated
(Mintzberg, 1979, 1987).
2.2. Complex PM as organisational practice — emergent
knowledge
In a paper that recognises the limitations of the planned
approaches of traditional PM in complex project settings,
Berggren et al. (2008) advocate a practice-oriented approach,
termed ‘neo-realistic’, which involves three key managerial
practices. These include “reducing complexity by transforming
expectations”, understanding of interdependencies for better
“systems integration”, and, importantly, “public arenas for
handling the unknown amount of errors” in complex PM
settings (ibid., p. S112, italics added). This analysis implicitly
acknowledges Hayek’s (1945) ‘specification problem’, which
here applies to complex projects that are unable to be fully
specified in advance, by recommending organic integration
for coordinating distributed contextual knowledge. In a recent
publication on knowledge integration in a complex PM
setting, Enberg et al. (2010) also encounter Hayek's ‘specifi-
cation problem’ in terms of “unforeseeable and unimaginable
multiplying effects of small changes” (p. 762). Informed by
Weick's (1995) sense-making and Polanyi’s (1967) tacit di-
mension of knowledge, they adopt a ‘segregated team’ ap-
proach to knowledge integration that relies in part on the “gut
feelings” of senior project team members, which this paper
views as a distributed tacit dimension (Polanyi, 1967, 1974).
Both these empirical papers highlight the need to generate
emergent knowledge in complex PM settings that is un-
specifiable at the outset and the need to coordinate this
knowledge in a distributed approach through higher-order
principles that are self-organising (Kogut and Zander, 1992).
2.3. Complicated PM versus complex PM — planned versus
emergent
Once Hayek's (1945) ‘specification problem’ is acknowl-
edged in complex PM settings, it is no longer tenable to
proceed under the assumption of ‘total planning’ of traditional
PM. In his classic paper on the workings of markets as complex
phenomena, Hayek (1945) highlighted a practical difficulty
with a centralised governance approach to knowledge. This is
because the complete data are never given “to a single mind
which could work out the implications, and can never be so
given” (ibid., p. 519), which he describes as “a problem of the
utilization of knowledge not given to anyone in its totality”
(ibid., p. 520). The knowledge Hayek (1945) had in mind was
knowledge that was specific to the “man on the spot” (p. 524),
which can be viewed as contextual ‘knowing’ knowledge. He
recommended that any solution to this practical problem
needed to harness contextual knowledge “that is dispersed
among many people” (ibid., p. 530).
This insight draws attention to an important difference between
the terms ‘complicated’ and ‘complex’. An aircraft is a
complicated machine that relies on a large number of servo-
mechanisms (single-loop) and crew members (double-loop, etc.)
to operate the machine system within normal parameters. In
aviation history, aircraft design progressed from being a complex
project, when the technology was poorly understood, to being a
complicated project, when detailed designs could be documented
for production assembly and, therefore, comprehensible to a single
mind. However, like an emerging prototype that is only partially
understood, a one-off complex project may not transition from
complex to complicated until after it is delivered and retrospec-
tively comprehended in its entirety (Snowden, 2002). Even with a
team of planners on a complex project, if no single individual can
comprehend the project interconnectivity in its entirety, then, no
one can preclude the possibility of knowledge gaps between
constituent parts of the plan (Lenfle and Loch, 2010). While
adjacent interfaces can be specified between parts of a linear plan
like links in a chain, this approach may reduce but not eliminate
the potential for gaps in a complex network plan that no single
individual comprehends in its entirety, e.g., PERT diagrams. 1
These potential gaps are like untapped knowledge, or
‘unknown knowns’ (Cleden, 2009), that may exist at the outset
of the project or emerge over time. Using metaphors, viewing a
complex project as complex problem solving (CPS) is more
like painting a landscape than the mechanical assembly of an
elaborate jigsaw. In a jigsaw, the pieces and their connectivity
are known in advance but, in a landscape painting, while the
major features may be known in outline in advance, the final
connectivity has yet to emerge due to shifting light, clouds,
shadows, etc. This emphasises the contextual specificity of
complex projects (Engwall, 2003), which operates on the
contextual knowledge of the community of learners that is
delivering the project by learning the project (Wenger, 2001).
2.4. Organisational learning under knowledge uncertainty
If complex projects are distinguished from complicated
projects by unspecifiable pre-given knowledge that must be
continuously generated over the project life cycle, then, the
creation of new knowledge and its coordination become central
aspects for governance in complex PM. Because of this
incompleteness of project plans, delivering a project is partly
about discovering its hidden reality through “tacit foreknowledge”
1 Project Evaluation and Review Technique (PERT).
1373T. Ahern et al. / International Journal of Project Management 32 (2014) 1371–1381