INTERNATIONAL JOURNAL OF

PROJECT
MANAGEMENT
International Journal of Project Management 24 (2006) 687–698
www.elsevier.com/locate/ijproman

Fundamental uncertainties in projects and the scope

of project management
Roger Atkinson a, Lynn Crawford b,c,*
, Stephen Ward d

a
Business School, The Institute of Business and Law, Bournemouth University, Christchurch House, Talbot Campus, Fernbarrow, Poole, BH12 5BB, UK
b
University of Technology, Sydney, P.O. Box 123, Broadway NSW 2007, Australia
c
ESC Lille, Avenue Willy Brandt, 59777 Euralille, France
d
School of Management, University of Southampton, Highfield, Southampton, SO17 1BJ, UK

Abstract

This paper builds on discussions that took place over a series of meetings in the UK of the Rethinking Project Management Network.
The management of uncertainty is seen as a necessary condition for effective project management. Sources of uncertainty are wide rang-
ing and have a fundamental effect on projects and project management. These sources are not confined to potential events, and include
lack of information, ambiguity, characteristics of project parties, tradeoffs between trust and control mechanisms, and varying agendas in
different stages of the project life cycle. Common project management practice does not address many fundamental sources of uncer-
tainty, particularly in ‘soft’ projects where flexibility and tolerance of vagueness are necessary. More sophisticated efforts to recognise
and manage important sources of uncertainty are needed. Such efforts need to encompass organisational capabilities, including some
aspects of organisation culture and learning.
Ó 2006 Elsevier Ltd and IPMA. All rights reserved.

Keywords: Uncertainty management; Hard and soft projects; Trust; Organisational capabilities

1. Introduction the validity and usefulness of the authors’ work, in many

This paper is concerned with the kinds of uncertainty
present in projects, and what might be done to manage
them. The paper builds on discussions that took place dur-
ing meetings of the UK EPSRC funded Network on
Rethinking Project Management over the period 2004–
2006. Early on views about uncertainty in projects were
summarised in a series of learning propositions to orientate
further discussions. In developing this paper, previous
work of the authors was an initial base, but perspectives
and issues that emerged from the network presentations
including case studies and discussions, extended this base.
Follow up discussions after presentations largely supported
*
Corresponding author. Address: University of Technology, Sydney,
P.O. Box 123, Broadway NSW 2007, Australia. Tel.: +61 2 9514 8730; fax:
+61 2 9968 1274.
E-mail address: Lynn.Crawford@uts.edu.au (L. Crawford).
instances providing further related insights and questions.
Throughout network meetings the presence of uncer-
tainty in a variety of forms was constantly recognised as a
central issue. Initial deliberations about uncertainty focused
on appreciating the variety of sources of uncertainty requir-
ing management attention and that go well beyond a set of
possible events that might impair project performance. This
has implications for the development of formalised
approaches to project risk management. These approaches
need to recognise the full range of sources of significant
uncertainty associated with any given project. Subsequent
network discussion reflected on the reasons why project
management to date seemed to be lacking in attending to
all this uncertainty. One explanation is that conventional
project management is too focused on operational planning
and control. This prompted reflection on how projects with
particularly problematic sources of uncertainty might be
characterised, and what implications this has for modifying

0263-7863/$30.00 Ó 2006 Elsevier Ltd and IPMA. All rights reserved.

doi:10.1016/j.ijproman.2006.09.011
688 R. Atkinson et al. / International Journal of Project Management 24 (2006) 687–698
or extending a conventional planning and control perspec-
tive on project management. In particular, there is a need
to recognise that many project contexts are characterised
by very high, difficult to quantify, levels of uncertainty
where management flexibility and tolerance of vagueness
are necessary. Following this line of thought, network dis-
cussions concluded that more attention needs to be given
to understanding and developing less tangible, but more
generic management processes associated with building
trust, sense-making, organisation learning, and building
an appropriate organisational culture better suited to oper-
ating with high levels of uncertainty.
The structure of this paper follows the sequence out-
lined above. Section 2 considers some key areas of uncer-
tainty that warrant serious management attention in most
projects. This has implications for the desirable scope of
common project risk management processes that augment
project management. Section 3 suggests that one obstacle
to effective management of uncertainty is that conven-
tional project management does not pay enough attention
to conception and end stages of the project life cycle, or
to strategic aspects of projects. Section 4 describes a
‘hard’/’soft’ typology of projects that highlights major
sources of ambiguity and vagueness that may be problem-
atic using convention plan and control project manage-
ment. Section 5 briefly considers some implications of
‘softness’ for the scope of project management processes,
including the role of trust. Section 6 identifies the impor-
tance of organisation infrastructure and in particular the
importance of organisational culture and capacity for
organisational learning.
2. Uncertainty management as part of project management
Much good project management practice can be thought
of as effective uncertainty management, clarifying what can
be done, deciding what is to be done, and ensuring that it
gets done as noted earlier. For example, good practice in
planning, coordination, setting milestones, and change con-
trol procedures, seeks to manage uncertainty directly. How-
ever, common practice does not consider the range of
sources of uncertainty present in projects or what a coordi-
nated approach to proactive and reactive uncertainty man-
agement can achieve. Three key areas of uncertainty
featured in network discussions: uncertainty associated with
estimating, uncertainty associated with project parties, and
uncertainty associated with stages of the project life cycle.
2.1. Uncertainty in estimates
An obvious aspect of uncertainty in any project con-
cerns estimates of potential variability in relation to perfor-
mance measures like cost, duration, or quality related to
particular planned activities. For example, we may not
know how much time and effort will be required to com-
plete a particular activity. It was evident from network dis-
cussions, and all the case study presentations by project
managers, that estimating activity for planning and control
purposes is a critical project management process. The
causes of uncertainty about estimates include one or more
of the following:
 lack of a clear specification of what is required;
 novelty, or lack of experience of this particular activity;
 complexity in terms of the number of influencing factors
and associated inter-dependencies;
 limited analysis of the processes involved in the activity;
 possible occurrence of particular events or conditions
which might affect the activity;
 emerging factors unknowable at the start of the project;
 bias exhibited by estimators, typically optimism bias
[1,2].
Thus, uncertainty results from vagueness, ambiguity and
contradictions associated with lack of clarity because of
lack of data, incomplete and inaccurate detail, lack of
structure to consider issues, the working and framing
assumptions being used to consider the issues, known
and unknown sources of bias, limited control of relevant
project players, and ignorance about how much effort it
is worth expending to clarify the situation [3].
2.2. Uncertainty associated with project parties
In many projects, particularly large ones, key perfor-
mance issues are often less related to technology, but rather
are related to uncertainty introduced by the existence of
multiple parties and the associated project management
infrastructure.
While employees and other agents of a project owner are
essential to the achieving of project performance, they also
contribute to uncertainty about future performance. This
uncertainty arises from several factors associated with each
project party, including [4]:
 uncertainty about the level of performance that will be
achieved;
 the objectives and motivation of each party;
 the quality and reliability of work undertaken;
 the extent to which each party’s objectives are aligned
with the project owner’s objectives, and the scope for
moral hazard where one party is motivated to do things
which are not in the best interests of the project owner;
 the actual abilities of the party;
 availability of the party.
In any organisational context including all projects, the
introduction of agents is prone to the three problems of:
adverse selection, moral hazard and risk allocation. The
uncertainties presented by these problems can be substan-
tial [5]. When these agents are different organizations, these
problems can be particularly challenging. Where project
owner and agent(s) belong to the same organization it
might be expected that such problems would be less likely
 R. Atkinson et al. / International Journal of Project Management 24 (2006) 687–698 689

to arise, to the extent that the parties can share informa- Table 1
tion, responsibilities and objectives more readily. Unfortu- Typical uncertainty management issues in each stage of the project life
cycle (Chapman and Ward, [3]. Copyright John Wiley & Sons Ltd., used
nately, this is not always the case. In any of these ostensibly with permission)
cooperative situations, the different parties involved are
Stages of the PLC Uncertainty management issues
likely to have different performance objectives, or at least
different priorities and perceptions of objectives. One con- Conceive the product Level of definition
Definition of appropriate performance
sequence of this is that different parties will have different objectives
perceptions of risks associated with these objectives, and Managing stakeholder expectations
therefore may wish to adopt different strategies for manag-
Design the product Novelty of design and technology
ing related project uncertainty. This divergence will be strategically Determining ‘fixed’ points in the design
aggravated if different parties also have different knowledge Control of changes
and perceptions of the nature of sources of uncertainty and
Plan the execution Identifying and allowing for regulatory
different capabilities for managing them [3,6]. strategically constraints
Concurrency of activities required
2.3. Uncertainty associated with stages in the project life Capturing dependency relationships
cycle Errors and omissions
Allocate resources Adequate accuracy of resource estimates
Many significant sources of uncertainty that need to be tactically Estimating resources required
managed in projects are associated with the fundamental Defining responsibilities (number and scope
of contracts)
generic management processes that make up the project life Defining contractual terms and conditions
cycle. A fair number of sources are implicitly acknowl- Selection of capable participants (tendering
edged in lists of project management ‘key success factors’ procedures and bid selection)
such as those offered by Gallagher [7]. Potential sources Execute production Exercising adequate coordination and control
typically identified in this way are listed in Table 1 against Determining the level and scope of control
various stages of the project life cycle. (By presenting Table systems
1 we do not mean to imply that projects can be character- Ensuring effective communication between
participants
ised as a simple linear process, or simply in terms of the
Provision of appropriate organizational
stages listed, but the scope of stages and level of detail is arrangements
sufficient for present purposes. Ward and Chapman [8] Ensuring effective leadership
provide a more detailed description of project life cycle Ensuring continuity in personnel and
stages and possible complications in practice.) responsibilities
Responding effectively to sources which are
All of the uncertainty management issues in Table 1 are
realized
best addressed very early in a project and throughout the
project life cycle and should be informed by a broad appre- Deliver the product Adequate testing
Adequate training
ciation of the underlying ‘root’ uncertainties. Chapman Managing stakeholder expectations
and Ward [3] offer a six Ws framework for this purpose Obtaining licences to operate
based on the following six questions about a project:
Review the process Capturing corporate knowledge
Learning key lessons
1. who are the parties ultimately involved? Understanding what success means
2. what do the parties want to achieve? Support the product Provision of appropriate organization
3. what is it that each party is interested in? arrangements
4. which way (how) is each party’s work to be done? Identifying extent of liabilities
5. what resources are required? Managing stakeholder expectations
6. when does it have to be done?
ciently well-defined specifications for production. During
Understanding the uncertainty associated with each of execution this gives rise to difficulties necessitating addi-
these basic questions, and the implications of interactions tional design development and production planning, and
between them, is fundamental to effective identification consequently adverse effects on the performance criteria of
and management of project risk. Use of the six Ws frame- cost, time and quality. This problem of ‘premature defini-
work from the earliest stages of the project life cycle can tion’ can be most acute in novel, one-off projects involving
usefully inform development of project design and logistics new technology, particularly when key stakeholders attempt
by clarifying key sources of uncertainty. In particular, fail- to impose unrealistic completion dates or cost targets.
ure to clarify stakeholder expectations and priorities at an Sometimes this can be caused by politically motivated per-
early stage can cause major difficulties later in the project. formance criteria, targets, and operating constraints: e.g.
A common source of difficulty in projects is a failure to budgets too small to allow adequate resources to complete
carry out the design and plan stages thoroughly enough. the project by a stated time, or to a given level of function-
Thus a project proceeds through to execution with insuffi- ality [9]. In addition, some uncertainty about operating
690 R. Atkinson et al. / International Journal of Project Management 24 (2006) 687–698
conditions and related factors outside the control of the pro-
ject manager will always remain. Inevitably, judgments have
to be made about the degree of detail and accuracy that is
practicable in the design and plan stages.
The allocate stage of the project life cycle is a significant
task involving decisions about project organization, identi-
fication of appropriate agents by the project owner, and
allocation of tasks between them. The sources of uncer-
tainty and risk associated with this stage of any project
are considerable, primarily due to the three agency prob-
lems of: adverse selection, moral hazard and risk allocation
mentioned earlier. Risk allocation is particularly important
because it can strongly influence the motivation of princi-
pal and agent, and the extent to which uncertainty is
assessed and managed. Insofar as principal and agent per-
ceive risks differently, and have different abilities and moti-
vations to manage sources of uncertainty, then their
approach to risk management will be different as noted ear-
lier. In particular, either party is likely to try to manage
uncertainty primarily for their own benefit, perhaps to
the disadvantage of the other party. Even in the same orga-
nisation, there can be significant issues in allocating risk in
a hierarchical structure or between different units in the
same organisation. Chapman and Ward [6, Chapter 6]
explore ‘internal contracts’ to address such issues.
A common source of uncertainty in the execution stage is
the introduction of design changes. Such design changes can
lead to disruption of schedules and resourcing, and affect
cost, time and quality measures of performance directly to
an extent that is difficult to predict. A potentially serious
concern is that changes are introduced without a full appre-
ciation of the knock-on consequences. Apart from direct
consequences, indirect consequences can occur. For exam-
ple, changes may induce an extension of schedules, allowing
contractors to escape the adverse consequences of delays in
works unaffected by the change. Changes may have wider
technical implications than first thought, leading to subse-
quent disputes between client and contractor about liability
for costs and consequential delays [10–12]. Standard project
management practice should establish product change con-
trol procedures that set up criteria for allowable changes
and provide for adequate coordination, communication
and documentation of changes.
In the plan stage, looking forward to the deliver and
support stages, and developing appropriate responses for
key sources of uncertainty, can reduce or eliminate poten-
tial later problems at relatively low cost. The key here is
identifying which issues need this attention in the plan
stage, and which do not.
2.4. Extending project risk management
Effective uncertainty management needs to address
uncertainty in a broad sense, with the early consideration
of all sources of significant uncertainty and associated
responses. Even when integrated with project management,
formal project risk management processes that adopt a
focus on threats will not adequately address many sources
of variability and ambiguity. Risk management processes
concerned with threats and opportunities will do better,
but will still tend to be focussed on uncertain events or cir-
cumstances. This does not facilitate consideration of
aspects of variability that are driven by underlying ambigu-
ity and lack of information.
To address all sources of significant uncertainty requires
a more explicit focus on uncertainty management as part of
managing projects. Particular attention needs to be paid to
the parties involved in a project and their respective objec-
tives in three ways.
2.4.1. Treat the definition of objectives as a key part of
managing projects
Selecting relevant performance criteria, formulating
objectives for these criteria, and modifying objectives,
should be regarded as important, even fundamental aspects
of project management.
2.4.2. Project management should clarify and manage
desired trade-offs between multiple performance objectives
Typically, in any given project context, there is more
than one performance criterion and associated objective,
often even a hierarchy of objectives. The implication is that
variations in performance on each criterion are possible
and measurable, and hence that uncertainty exists in
respect of each of these performance criteria. A simple
example is the common presentation of project perfor-
mance in terms of cost, time and quality related objectives.
The cost criterion might be addressed in terms of capital
cost or ‘whole life’ cost, and the quality attribute might
be divided into technical specification, functionality, reli-
ability, and appearance, each of which may be ‘at risk’ to
different degrees. Objectives may be set for each of these
performance attributes, and the project will be ‘at risk’ to
different degrees with respect to each objective.
Active management for performance usually involves
making trade-offs between objectives because different
courses of action involve different combinations of uncer-
tainty in respect of the various performance criteria. Fail-
ure to recognise these trade-offs and articulate preferred
trade-offs can result in ineffective and often inappropriate
management of risk. For example, to ensure that a project
is completed on time is it really sensible to adopt any
course of action whatever the cost?
In the extreme, performance criteria that are not readily
quantified may be treated as inviolate constraints for man-
agement purposes. This may lead to neglect of uncertainty
in these performance criteria and failure to manage associ-
ated risk, even though these criteria represent important
aspects of performance.
2.4.3. Ownership of uncertainty needs explicit consideration
Decisions should be made about how uncertainty and
associated issues should be allocated to various project par-
ties, recognising as noted earlier, that different parties have
 R. Atkinson et al. / International Journal of Project Management 24 (2006) 687–698
different objectives, different perceptions of project risk,
and different capabilities for managing associated sources
of uncertainty.
3. The limited scope of conventional project management
An emergent conclusion from network discussions was
that common perceptions of projects and project manage-
ment practice do not encompass all the stages of the project
life cycle shown in Table 1. This is another reason why
common risk management practice often fails to address
basic sources of uncertainty that drive problems in the pro-
ject life cycle. In particular, professional guidelines mini-
mise the role of conception at the front end of the life
cycle and support at the tail end. An associated boundary
problem is clearly delineating the conception stage and
understanding explicitly how this stage is linked to subse-
quent project life cycle stages (especially design and plan-
ning) in relation to determining appropriate objectives,
performance requirements, and constraints.
Even in other life cycle stages, professional guidelines
often fail to distinguish between strategic, whole project
considerations, and lower level, operational procedures.
Ward and Chapman’s [8] depiction of the design, plan
and allocate stages as highly iterative processes comprising
a number of development steps, goes some way towards
drawing this distinction. However, it seems as if the more
procedural (teachable?) elements of these stages receive
more attention from trainers of project managers and in
project management text books, than the more strategic
issues. Unfortunately, it is the latter that includes the most
important, fundamental sources of uncertainty. The result
is that project management is commonly regarded as con-
cerned with ensuring things get done right, assuming that
there is a well defined remit of what needs to get done. With
this view, project management is not concerned with think-
ing about whether the right things are being done, why the
project should proceed, or what performance criteria
would be appropriate. This may explain why it has been
possible for the subjects such as ‘whole life costing’ and
‘value management’ to emerge somewhat separately from
traditional project management concepts.
Perhaps the conventional common view of project man-
agement is essentially to see the project task as a set of pro-
cesses to ensure a project meets its (predetermined)
objectives. Then the whole raison d’etre of project manage-
ment is to remove (or substantially reduce) uncertainty
about meeting specified objectives. However, project man-
agement in this sense is a castle built on shifting sands if in
practice objectives are unclear, contradictory, or impossi-
ble. Many endeavours recognised and ‘managed’ as pro-
jects experience problems for this reason.
A common interpretation of a project plan is that it is an
attempt to define an intended future. Project design and
planning activity creates a set of plausible working assump-
tions as a basis to move the project forward. The tools and
techniques of conventional project management are very
691
useful in the right place. However, they too reinforce a
focus on operational planning and control with consequent
lack of attention to strategic issues and associated funda-
mental uncertainty management issues.
A view expressed in network discussions was that con-
ventional (common practice) project management pro-
cesses are concerned with legitimising the project plan,
and uncertainty (particularly from fundamental sources)
is played down. The project manager is regarded as a con-
venient recipient of project risk, providing psychological
relief to the project owner (if not actual physical relief)
from the burden of uncertainty and risk bearing, and some-
one who subsequently unwillingly serves as scape-goat if
things fail to turn out as desired by the project owner.
4. Hard and soft projects
In considering the appropriate scope for project man-
agement and associated uncertainty management, it is use-
ful to characterise the range of project types and contexts
in terms of the scope of uncertainty involved. Conventional
project management approaches may be more effective for
some kinds of project than others.
In practice, the concept of ‘project’ has been broadened
from an initial focus on management of largely unitary,
standalone projects with well defined and agreed goals
and end products, to include multiple projects and pro-
grammes that are multidisciplinary, and which are ‘‘not
pre-defined’’ but ‘‘permeable, contested and open to rene-
gotiation throughout’’ [13, p. 4]. These two ends of a spec-
trum are often referred to as ‘hard’ and ‘soft’ but in reality
there are a number of dimensions of hardness and softness
of projects [14,15]. Projects and programmes may simulta-
neously exhibit both hard and soft characteristics on these
dimensions and these characteristics may change through-
out a project life cycle.
Crawford and Pollack [14] identify seven dimensions of
hardness and softness of projects based on: previous
research [16,17], use of the terms ‘hard’ and ‘soft’ in project
management practice and literature, and differences in the
philosophical basis of the hard/soft dichotomy [18]. These
seven dimensions of hardness and softness are illustrated in
Fig. 1.
For some projects, goals are generally clear and well
defined and goals may be ‘given’ at the start. At the other
end of the spectrum, projects may have multiple purposes
and there may be differing views and expectations of the
outcome of a project. Consequently, goals may be initially
ill-defined, emerging as a result of negotiation and consen-
sus building throughout the project. There is also the
uncertainty that the ‘real’ purpose for a project is hidden
for political, economic, social or technical reasons, while
only the secondary or declared reason is being managed.
Projects that have tangible end products such as build-
ings, equipment and machinery, can be represented as phys-
ical models and prototypes, providing a basis for clarity in
terms of what will be produced. Clarity and agreement
692 R. Atkinson et al. / International Journal of Project Management 24 (2006) 687–698

about the end product are far harder to achieve where the
end products of the project are intangible, as in information
systems development and organisational and cultural
change initiatives. Stakeholders are likely to have different
interpretations and expectations of what will be produced.
Judgement concerning project success is far more difficult
to assess in projects with intangible products. Further, if
success is to be measured, it will be more likely to rely on
qualitative rather than quantitative measures that once
again will be susceptible to differing interpretations and per-
ceptions. Initiatives to extend the ‘Iron Triangle’ of cost,
time and quality to include benefits and value criteria have
been addressed by such as Atkinson [19].
Where the boundaries of the project are defined by con-
tracts, relationships and project boundaries are defined,
and exchange between the project and its environment is
amenable to control. At the other end of the spectrum
are projects where the boundaries are unclear and perme-
able. An example is an organisational change project where
it is difficult to identify where the project starts and ends. In
such cases, scope may be difficult to define. Project bound-
aries may be moved or redefined, or subject to differing
interpretations by different stakeholders. Such projects
are often undertaken by staff internal to an organisation
who are not isolated from their environment by contractual
structures. Staff are often shared with other projects and
with functional or line positions, further blurring the
boundaries of the projects. Such projects can be highly sus-
ceptible to external influence.
Projects that have clear boundaries providing a degree
of isolation from environmental influences allow for the
use of tools and techniques of project management such
as network planning, risk analysis, and computerised
resource allocation and task coordination. As the bound-
aries of the project become more permeable and projects
assume more of the characteristics at the soft end of the
spectrum in Fig. 1, the use of many of the tools and tech-
niques of traditional project management for dealing with
uncertainty become more difficult, less reliable and further
removed from reality.
For some projects it is assumed that there is one best
solution generally developed by technical experts. At the
soft end of the spectrum, multiple world views and perspec-
tives are recognised and solutions are developed through
negotiations and debate between multiple stakeholders.
Where the project involves refinement of a single solution,
the path to the known and agreed best solutions is man-
aged. Projects at the soft end may require facilitation of a
process for identifying and agreeing on possible solutions.
The degree of stakeholder involvement and the expecta-
tions of stakeholders in terms of interaction will also affect
the ability to reduce uncertainty in the interests of achiev-
ing clarity and control.
A general observation that arises from the conceptuali-
sation of projects in terms of Fig. 1 dimensions is that pro-
jects closer to the hard end of the spectrum are more
amenable to uncertainty reduction than those at the soft
end of the spectrum where the primary emphasis may be
on reduction of ambiguity. This suggests different
approaches for dealing with uncertainty and stakeholder
expectations depending on the nature of the project or pro-
gramme in terms of hard and soft dimensions. The softer
the project against these dimensions, the greater the degree
of flexibility, tolerance of vagueness but less so ambiguity,
and acceptance of residual uncertainty that will be required
of stakeholders in assessing the progress and success of the
project. Where uncertainty about future events is high, tol-
erance of uncertainty may be particularly necessary. Flexi-
bility is necessary in projects where the goals are unclear or
open to negotiation, strategy is emergent and the project is

Goals/objectives 1. Goal Clarity Goals/objectives highly

0 100
clearly defined ambiguously defined

2. Goal Tangibility
0 100
Physical artefact Abstract concept

Only quantitative 3. Success Measures Only qualitative

0 100
measures measures

Not subject to 4. Project Permeability Highly subject to

0 100
external influences external influences

Refinement of 5. Number of Solution Options Exploration of many

0 100
single solution alternative solutions
Expert practitioner, Facilitative practitioner,
0 6. Participation and Practitioner Role 100 high stakeholder
no stakeholder
participation involvement

Values technical 7. Stakeholder Expectations Values relationships,

0 100
performance and culture and meaning,
efficiency, manages by manages by negotiation
monitoring and control and discussion

Fig. 1. Depiction of the hard and soft dimensions framework [14].

 R. Atkinson et al. / International Journal of Project Management 24 (2006) 687–698
highly subject to external influences. Keeping options open,
and adopting a flexible, robust approach may be much
more effective than prematurely crystallising plans and
relying on conventional control mechanisms to deliver
performance.
Analysis of projects against this hard/soft dimensions
framework provides a basis for questioning assumptions
about the nature of projects. When examined in this way,
even projects that may initially be considered hard (e.g.
construction projects) and therefore amenable to tradi-
tional control approaches for dealing with uncertainty
and managing expectations, may be found to have some
characteristics towards the soft end of the spectrum and
visa versa.
5. Implications of ‘softness’ for the scope of project
management
Widely available and promoted project and risk man-
agement methodologies, tools and techniques have been
developed to deal with uncertainty in projects with charac-
teristics primarily at the hard end of the spectrum described
above. Associated with these approaches to project man-
agement are expectations of clarity and certainty that
may not be either feasible or desirable for projects with
characteristics that are at the soft end of the spectrum.
Such projects may require both different approaches and
levels of performance expectation. For example, Thiry
[20] proposes (see Fig. 2) the use of sense-making and value
analysis for projects that are at the soft end of the spectrum
where both uncertainty and ambiguity are high. As Fig. 2
suggests, Thiry [20] makes a clear distinction between
uncertainty and ambiguity/equivocality. According to
Thiry, uncertainty is defined by the difference between the
data required and the data already possessed; it is a ‘lack
of information’. Ambiguity, on the other hand, means
the existence of multiple and conflicting interpretations; it
is linked to confusion and lack of understanding. Whereas
uncertainty warrants the acquisition of objective informa-
tion and the answering of specific questions, ambiguity
warrants sense-making, the exchange of views and the def-
inition of situations/problems.
Such sense-making is particularly important in the con-
cept stage of the project life cycle, and during preliminary
design and planning activities where uncertainty is high
and a strategic perspective of context possibilities is
important. Unfortunately, if the need for ongoing sense-
making is not acknowledged, later pressures to crystallise
plans and subsequent reification of the project plan,
together with escalating commitment, may increasingly
preclude further sense-making as the project progresses.
Additionally, lack of time and limited resources may dis-
courage sense-making efforts. Further, such sense-making
that takes place may be limited by being inherently retro-
spective rather than forward looking, and political consid-
erations may produce more concern for plausibility than
accuracy.
693
ft
So
High
Ongoing
Ambiguity
Sensemaking
Conversations & Value Analysis
Ongoing Risk Analysis &
Operations Problem Solving
Low
d
ar
H
Low High
Uncertainty
Fig. 2. Uncertainty – ambiguity relationship in change situations –
adapted from Thiry [20].
Ability to formulate qualitative success measures for
projects is another tool that should be added to the project
management armoury to assist in managing softer projects.
Projects with low uncertainty and ambiguity can generally
be assessed using quantitative success measures such as
time and cost performance, and measures related to their
tangible end products. Projects at the soft end of the spec-
trum require different forms of performance evaluation
that recognise the validity of different perspectives and
worldviews. This calls for ability to develop sensitive per-
formance evaluation frameworks that match the complex-
ity of the project.
5.1. The role of trust
As projects take on more ‘soft’ characteristics, the
importance of project parties as contributors to project
uncertainty increases. Then much depends on the motives
of different parties, the incentives operating, and the level
and type of trust between parties.
A root problem is that a project context is more tempo-
rary than ongoing operations where routinisation, learning
from past experiences, memory of past experiences is easier
(more available), the parties are relatively constant, and
experimenting and the evolution of optimum practice is
possible. The subject matter of the project may involve
the creation of artefacts outside the project owner’s capa-
bility (buildings, physical infrastructure, software etc) and
this necessitates the involvement of outside parties, who
may be unfamiliar to the project owner. Such new and
temporary relationships increase the importance of trust,
since project parties may have little or no prior knowledge
of the other parties’ technical or fiduciary standards, and
there is a lack of time for familiarity to develop from
shared experiences or demonstrations of non exploitation
of vulnerability.
A further problem is that total control over the activ-
ities of project parties is neither possible nor desirable.
694 R. Atkinson et al. / International Journal of Project Management 24 (2006) 687–698
An understandable reliance on controls can lead project
staff to feel that they are not trusted (and vice versa),
and this can have adverse consequences of a moral hazard
nature. Moreover, any lack of trust can produce a
dynamic that serves to ratchet up control mechanisms’,
thus preventing success Coopey [21]. The problem with
trust is that it can spiral, both positively and negatively
[22], consequently there is always a balance to be struck
between the use of controls and trust. An increase in
one implies a compensating reduction in the other. Handy
[23] argues that where you cannot trust, you have to
check all the systems of control involved, to ensure that
any gaps in trust are replaced with controls. However,
while acknowledging that a mixture of trust and controls
are needed, O’Neill [24] argues that ‘all guarantees are
incomplete’ and so ‘. . . elaborate measures to ensure that
people keep agreements and do not betray trust must, in
the end, be backed by trust’.
There are a number of benefits related to uncertainty
reduction from using trust in place of controls. These
include [25]:
 more accurate risk calculations (through more open
communications between project parties),
 reduced control costs (through for example faster and
more effective contracting),
 teams more effective (through improved confidence),
 improved planning (through more honest specifications
and estimates).
Zaghloul and Hartman [26] discuss the problems of mis-
trust within construction projects, in terms of the cost sav-
ings that can be obtained through reducing the use of
exculpatory clauses in contracts and building trust via:
 understanding project risks and who owns them,
 allowing sufficient time to manage or mitigate the risks,
 building trust through negotiations prior to a written
contract,
 risk-sharing and the design of appropriate risk-reward
systems.
If a relationship is to start in a state of mutual trust, it
requires prior confidence in the reliability of the other
party, otherwise those involved are being asked to trust
without evidence that it is safe to do so. One requirement
for trust to exist is to have it built-up over time. This can
be observed to be taking place when individuals do what
they say they will do, thus creating a basis of trust build-
ing. However, with project work, there is often limited
time to build trust between parties and there is often no
history of their behaviour, but the parties still have to
work together to achieve the project objectives. Hence
formal team-building exercises are often used to build
trust as an alternative to something which would have
naturally occurred over time. But there can be a problem
with these, in that the type of trust operating in team
building workshops is often fiduciary, while decision mak-
ing in projects can often be influenced by rational eco-
nomic reasons. Meyerson et al. [27] have termed the
trust involved in such circumstances ‘swift trust’. When
team members have little knowledge of other team mem-
bers’ likely actions, ‘swift trust’ is entered into as team
members know that they have to trust others in order
to move a project forward, but have little evidence of
the level of their vulnerability as they do not know other
team members. Meyerson et al. [27] studied temporary
groups (such as found in weak matrix project structures)
and concluded that swift trust is found in high risk, high
stake projects, where there is a lack of normative struc-
tures and institutional safeguards to minimise things
going wrong.
To help create trust, Cook [28] advises that uncertainty
can be reduced by knowing the competence of other staff
while vulnerability can be decreased by the use of enforce-
able contracts and insurance schemes. Heimer [29] explains
the method required to promote a trust strategy ‘. . .
worked by reducing uncertainty’ while a distrust strategy
is ‘. . . worked by reducing vulnerability’. Overall, Heimer
suggests that the trust problem can be solved by manipulat-
ing vulnerability and uncertainty about intentions and
competence. Similarly, Munns [22] proposed a ‘Graduated
and Reciprocated Initiatives in Tension reduction’ (GRIT)
model to be followed in a process of continual building of
mutual trust during a project.
In an attempt to improve the success rate of projects,
trust (which takes several forms) should be included for-
mally within the discipline of project management. Trust
does not figure in the current APM Bok version 4 [30]
nor the BS 6079 [31]. One way in which trust could be used
by practitioners is to include an audit of trust as an element
of uncertainty management.
A framework for such a trust audit proposed by Atkin-
son [32] contains the following factors:
 vulnerability (to the action of others),
 credibility (related to both ‘can do’ and ‘will do’),
 culture (of the organization),
 visibility (and openness of information).
Each factor relates to different types of trust, and ques-
tions related to those topics enable a comprehensive view
of the trust within a project team to be obtained. These fac-
tors could be considered by different project stakeholders
such as: the Project Manager, Project Board, Quality, Sup-
port, Development and the Users to develop as complete
an audit as possible.
By including a trust audit as part of an uncertainty man-
agement process, practical measures of governance can be
included to ensure that the balance of trust and control is
known and that appropriate measures are put into place.
If there is an over reliance on controls in place of trust,
organisations must expect and prepare for the dysfunc-
tional behaviour that will follow.
 R. Atkinson et al. / International Journal of Project Management 24 (2006) 687–698
6. Supporting organisational capabilities
All project management activity takes place in a wider
organisation context, and how the organisation operates
will have a major impact on what can be achieved by its
members in terms of project management. Organisation
structure, co-ordination and control systems, environmen-
tal scanning capability, communications and information
systems, knowledge management, and support for organi-
sation learning, all affect the quality and scope of project
management undertaken [33]. Such factors define the basic
resources that project management must work with, and
they set the tone for how project management will be able
(or allowed) to operate. Such factors can enable project
management to flourish, or can present barriers to its
development no matter how capable and determined the
champions of project management effort. Consequently,
efforts to improve and broaden the scope of project man-
agement need to consider the organisational infrastructure
that facilitates project management as much as the partic-
ular processes for specific projects.
Organisations which have efficient and effective systems
for co-ordination and control, environmental scanning,
and organisation learning will be comparatively well placed
to foster efficient and effective uncertainty management.
Unfortunately, many organisations exhibit deficiencies in
their approach to uncertainty and learning which can act
as barriers to the development of uncertainty management.
Sometimes shortcomings in organisational capabilities are
not evident until systematic attempts to identify and man-
age uncertainty are made. The following extract from the
executive summary of the Space Shuttle Columbia Acci-
dent Investigation Board [34] is instructive of the issues
involved:
The organisational causes of this accident are rooted in
the Space Shuttle Program’s history and culture, includ-
ing the original compromises that were required to gain
approval for the Shuttle, subsequent years of resources
constraints, fluctuating priorities, schedule pressures,
mischaracterization of the Shuttle as operational rather
than developmental, and lack of an agreed national
vision for human space flight. Cultural traits and organ-
isational practices detrimental to safety were allowed to
develop, including: reliance on past success as a substi-
tute for sound engineering practices (such as testing to
understand why systems were not performing in accor-
dance with requirements); organisational barriers that
prevented effective communication of critical safety
information and stifled professional differences of opin-
ion; lack of integrated management across program ele-
ments; and the evolution of an informal chain of
command and decision making processes that operated
outside the organisation’s rules.
The Board concludes that NASA’s current organisation
. . . has not demonstrated the characteristics of a learn-
ing organisation.
695
Accepting that culture can operate as an organisational
control mechanism, the quality of uncertainty management
undertaken as part of project management can be driven or
at least influenced by the organisation culture prevailing in
associated business units. This culture can be manifest in a
variety of ways, such as attitude to: planning, formal pro-
cedures, regulations, criticism, mistakes, uncertainty, and
risk. These cultural characteristics can either facilitate or
hinder the development of uncertainty management.
Chapman and Ward [6, Chapter 12] argue that some of
the most significant barriers to effective uncertainty man-
agement are based on unfavourable features of organisa-
tional culture. Sometimes these barriers are induced
from outside the organisation in the form of convictions,
prejudices, biases, and routines of professional groups
that can blinker thinking. These may arise from a wish
to make decisions efficiently in a particular context
according to recognised scientific or professional stan-
dards. Such professionally based convictions can strongly
influence different groups within the same organisation,
making it very difficult to adopt effective uncertainty man-
agement processes. However, most organisations also
exhibit a number of more generic culture based behav-
iours or conditions inimical to effective uncertainty man-
agement such as ‘conspiracies of optimism’, ‘macho
management’, blame culture, and ‘management by misdi-
rection’ [6, Chapter 12]. Essentially such behaviours seem
to evidence the difficulty management has in coping with
complexity and uncertainty. In particular, these behav-
iours can reflect an inability or unwillingness on the part
of managers or groups to recognise the difference between
(a) bad management and poor performance due to factors
that are not under a manager’s control; and (b) good
managers who apply proactive uncertainty management
to reduce problems and enhance performance, and man-
agers who are just lucky. Addressing such conditions
can be one of the most significant benefits of formal
uncertainty management processes.
6.1. Learning by experience
Knowledge management and learning, both organisa-
tional and individual, are major contributors to uncer-
tainty management in a variety of ways. At a very
practical level, readily accessible repositories of data from
past projects either specific to the organisation or available
from industry sources, are fundamental to the quality of
estimates. Availability of reliable data for the estimating
and planning of projects in itself contributes to the reduc-
tion of uncertainty. However, many organisations fail to
collect such data on projects, and even where it is collected
it is often not made available for those embarking on new
projects or is not analysed and presented in a form that is
useful for such purposes. The problem is compounded
where new project types are undertaken by organisations.
The level of uncertainty may be higher for management
of projects in some organisations than in others due to lack
696 R. Atkinson et al. / International Journal of Project Management 24 (2006) 687–698
of data from past projects. Such data may or may not be
available from other organisations or industry sources.
Even where data on past performance is available, a
major challenge is failure to access such inputs during the
planning of new projects and therefore failing to learn from
experience. There are a number of factors that contribute
to this including the organisational culture as discussed
above, time pressures and the attitudes and behaviours of
project management personnel. Certainly the non-repeti-
tive nature of project work in the context of temporary
organisations presents a particular challenge for knowledge
management, and transfer of learning.
‘‘Lessons learned’’ is a popular term in the project man-
agement literature and amongst practitioners, yet it often
masks payment of lip service only to the idea of learning
from experience. The capture and re-use of learning from
one project to another is generally accepted as something
that should be done but it often goes no further than cap-
ture. It is often associated with post project reviews where
learning has significant potential to reduce uncertainty.
However, the value of these reviews may be undermined
by factors outlined earlier in this paper, including failure
to report important lessons for the future due to cultural
issues, especially those surrounding perceptions of failure.
When a project is finished, the lessons learned are linked
to whether the project was delivered on time within cost
and to the agreed quality. The benefits of value manage-
ment may not be realised for some time after the project
is finished. The question is who writes a post implementa-
tion review when the team have been disbanded, who pays,
and who would the circulation list of readers be? The
excitement of a new project contains energy to get it started
while at the end of a project that energy is reduced. How-
ever, while Senge [35] presents the need for reflection and
reviews, unless those lessons change working practices,
no organizational learning has taken place. Now compare
that with the notion that projects can be described as a
temporary organisation: the problems involved in learning
within project work start to become evident. So a new
method of learning from projects needs to be explored
since this is different from typical organisational learning.
Potential benefits from learning from experience may be
affected by the nature of different types of projects and
their contexts. As outlined in Section 5 above, the nature
and degree of uncertainty reduction and therefore control
that is possible in projects that satisfy mainstream and a
primarily ‘hard’ conception of projects, may not be either
feasible or desirable for ‘soft’ projects involving higher lev-
els of abstraction, stakeholder involvement, and residual
ambiguity. The same levels of performance expectation,
and therefore measures of success, may not be appropriate
when applied to different types of projects, and this may
lead to adverse effects in terms of learning. For instance,
projects may be categorised as failures due to unrealistic
expectations. This may lead to transfer of lessons that are
misleading or discourage honest transfer of learning due
to the stigma of failure. If uncertainty is acknowledged,
then it is more likely that an environment conducive to
openness and learning can be created. Further, the learning
extracted is likely to be more reliable and useful in reducing
uncertainty in future projects.
7. Conclusions
Discussions of the UK EPSRC funded Network on
Rethinking Project Management raised many issues with
implications for the fundamental uncertainties of projects
and the scope of project management. The network pro-
posed rethinking of the traditional view of projects as hav-
ing a linear life cycle involving a set of apolitical processes
undertaken to achieve an objective or goal ‘given’ at the
start, to reflect the actuality of projects as social processes
requiring ongoing construction of the appearance of cer-
tainty and clarity in the midst of complex uncertainty
and ambiguity [36]. Some but not all aspects of uncertainty
can be categorised and treated as risks, and risk, generally
considered as a threat to achievement of project objectives,
receives far more overt attention than the broader concept
of uncertainty in the traditional view of projects and their
management. An inescapable conclusion from the network
meetings was that management of uncertainty is a neces-
sary condition for effective project management, but that
management of uncertainty needs to be given more atten-
tion and be rather more sophisticated than current com-
mon practice. This paper has outlined directions for
development of project uncertainty management.
Sources of uncertainty are wide ranging and have a fun-
damental effect on projects and project management. These
sources are not confined to potential events, and include
lack of information, ambiguity, characteristics of project
parties, tradeoffs between trust and control mechanisms,
and varying agendas in different stages of the project life
cycle. Risk management processes that focus on identifying
potential events as threats (or opportunities) will not
address many important sources of uncertainty. Further,
common practice project management tends not to address
many fundamental sources of uncertainty, particularly in
the conception and post delivery stages of the project life
cycle, or in ‘soft’ projects where flexibility and tolerance
of vagueness are necessary. More sophisticated efforts to
recognise and manage important sources of residual uncer-
tainty are needed.
Such efforts within a given organisation require some
attention to organisational culture, capabilities, and the
development of appropriate infrastructure [4,33]. At a
basic level, there is a need to understand stakeholder toler-
ance or intolerance of uncertainty in processes and out-
comes. Why is uncertainty tolerated? By whom?
Inexperienced project owners may be inappropriately intol-
erant of uncertainty, particularly if they hope to transfer
risk and responsibility for managing uncertainty to agents,
and projects exhibit a significant degree of ‘softness’. This
intolerance of uncertainty may induce project management
behaviours such as cautious/safe ways of working and
 R. Atkinson et al. / International Journal of Project Management 24 (2006) 687–698
missed opportunities, the mindless/uncritical/mechanical
application of project management principles and tech-
niques, and actions designed to avoid apportionment of
blame when things do not turn out as hoped. Conversely
contractors may be inappropriately tolerant of uncertainty
because of optimism, the felt need to accept risk and asso-
ciated uncertainty in order to win work, or because of igno-
rance about the scope of uncertainty present.
Replacing ambiguity with vagueness is one possible
method of reducing uncertainty. Managing stakeholder
expectations is a further method of bringing uncertainty
into project discussions. While this might mean that some
stakeholders are mildly disappointed at the end of a pro-
ject, this is preferable to having stakeholders being sur-
prised by the final outcome of a project. Managing
expectations transfers the uncertainty of surprise into pos-
sible disappointment, thus not eliminating the problem, but
transferring the nature of the problem to a different more
manageable form.
Uncertainty is created in part by the quality and com-
pleteness of information, diversity of interests and suscep-
tibility to external influences in a project; all of which
makes us vulnerable to the action of others. The most eco-
nomic method of compensating for gaps in information is
through trust, of which there are many types and levels,
each requiring different coping strategies. However, com-
plex projects require the controls of governance. The out-
come is for a balance of trust and control, with an
acceptance that trust will be ultimately overarching due
to the lack of a guarantee of controls. For the practitioner,
the link and dynamics between uncertainty, control and
trust could be improved if the factors of trust were included
in uncertainty management processes.
References
[1] Buehler R, Griffin D, Ross M. Inside the planning fallacy: the causes
and consequences of optimistic time predictions. In: Gilovich T,
Griffin D, Kahneman D, editors. Heuristics and Biases – the
Psychology of Intuitive Judgment. Cambridge University Press; 2002.
[2] Armor DA, Taylor SE. When predictions fail: the dilemma of
unrealistic optimism. In: Gilovich T, Griffin D, Kahneman D, editors.
Heuristics and Biases – the Psychology of Intuitive Judgment.
Cambridge, UK: Cambridge University Press; 2002.
[3] Chapman CB, Ward SC. Project Risk Management: Processes,
Techniques and Insights. second ed. John Wiley & Sons; 2003.
[4] Ward SC. Requirements for an effective risk management process.
Project Management Journal 1999(September):37–42.
[5] Eisenhardt KM. Agency Theory: An Assessment And Review. The
Academy of Management Review 1989;14(1):57–74.
[6] Chapman CB, Ward SC. Managing Project Risk and Uncertainty: A
Constructively Simple Approach to Decision Making. Chichester,
UK: John Wiley & Sons Ltd.; 2002.
[7] Gallagher K. Chaos: project success factors, In: Proceedings of
Implementing Technology and Project Management Conference,
September, 1995, pp. 21–36.
[8] Ward SC, Chapman CB. Risk management and the project life cycle.
International Journal of Project Management 1995;13(3):145–9.
[9] Flyvbjerg B, Bruzelius N, Rothengatter W. Megaprojects and Risk:
An Anatomy of Ambition. Cambridge: Cambridge University Press;
2003.
697
[10] Williams TM, Eden C, Ackerman F, Tait A. The effects of design
changes and delays on project costs. Journal of the Operational
Research Society 1995;46:809–18.
[11] Williams TM, Eden C, Ackerman F, Tait A. Vicious circles of
parallelism. International Journal of Project Management 1995;13:
151–5.
[12] Cooper KG. Naval ship production: a claim settled and a framework
built. Interfaces 1980;10(6):20–36.
[13] Winter, Mark, Smith, Charles. Making sense so far: emerging
directions and future research: Interim Report May 2005.
<www.rethinkingpm.org.uk>. 2005; (accessed on 17 June 2005).
[14] Crawford L, Pollack J. Hard and soft projects: a framework for
analysis. International Journal of Project Management 2004;22(8):
645–53.
[15] Crawford LH, Hobbs JB, Turner JR. Project Categorization
Systems: Aligning Capability with Strategy for Better Results,
Newtown Square. Pennsylvania: Project Management Institute;
2005.
[16] Crawford L, Costello K. Towards a transferable methodology for
managing strategic change by projects. In: Crawford L, Clarke CF,
editors. IRNOP IV Conference - Paradoxes of Project Collaboration
in the Global Economy: Interdependence, Complexity and Ambigu-
ity. Sydney, Australia: University of Technology; 2000.
[17] Costello K, Crawford L, Bentley L, et al. Connecting soft systems
thinking with project management practice: an organizational change
case study. In: Ragsdell G, editor. Systems Theory and Practice in the
Knowledge Age. New York: Kluwer Academic/Plenum Publishers;
2002. p. 002.
[18] Midgley G. Systemic Intervention: Philosophy, Methodology, and
Practice. New York: Kluwer Academic / Plenum Publishers; 2000.
[19] Atkinson RW. Project management: cost, time and quality, two
best guesses and a phenomenon, its time to accept other success
criteria. International Journal of Project Management 1999;17(6):
337–42.
[20] Thiry M. Combining value and project management into an effective
programme management model. International Journal of Project
Management 2002;20(3):221–7.
[21] Coopey J. Learning to trust and trusting to learn. Management
Learning 1998;29(3):365–82.
[22] Munns AK. Potential influence of trust on the successful completion
of a project. International Journal of Project Management
1995;13(1):19–24.
[23] Handy C. Trust and the virtual organization. Harvard Business
Review 1995(May-June):40–50.
[24] O’Neill O, A question of trust [Web Page]. <http://www.bbc.co.uk/
radio4/reith2002/1.shtml.2002; (Accessed on 15 April 2002).
[25] F. Hartman, Smart trust: a foundation for more effective project
management, in: Proceedings of 15th IPMA World Congress on
Project Management, London, 22–25 May 2000, London, IPMA,
2000.
[26] Zaghloul R, Hartman F. Construction contracts: the cost of mistrust.
International Journal of Project Management 2003;21:419–24.
[27] Meyerson D, Weick KE, Kramer RM. Swift trust and temporary
groups. In: Kramer RM, Tyler TR, editors. Trust in Organisations.
London: Sage Publications; 1996. p. 166–95.
[28] Cook KS. Trust in society. In: Cook KS, editor. Trust in Society.
New York: Russell Sage Foundation; 2001. p. xi–xxviii.
[29] Heimer CA. Solving the problem of trust. In: Cook KS, editor. Trust
in Society. New York: Russell Sage Foundation; 2001. p. 65–73.
[30] Dixon M. APM Project Management Body of Knowledge. fourth
ed. Peterborough, England: Association for Project Management;
2000.
[31] British Standards Board Guide to project management: BS6079.
London, British Standards Board, 1996.
[32] Atkinson RW. Information System’s Project Management and the
Phenomenon of Trust. PhD Thesis. Bournemouth University.
[33] Ward SC. Risk Management Organisation and Context. London:
Witherbys; 2005.
698 R. Atkinson et al. / International Journal of Project Management 24 (2006) 687–698

[34] CAIB Columbia Accident Investigation Board Report, Washington,
DC, CAIB, NASA and Government Printing Office, 2003.
[35] Senge PM. The Fifth Discipline: The Art and Practice of the Learning
Organisation. London: Bantam Doubleday Dell Publishing Group,
Inc.; 1992.
[36] Smith, Charles, Winter, Mark C. EPSRC Network 2004–2006:
Rethinking Project Management: Meeting 5: Actuality and Uncer-
tainty: Sensemaking Paper 5 [Web Page]. <http://www.mace.man-
chester.ac.uk/project/research/management/rethinkpm/pdf/papers/
actuality_uncertainty.pdf>. 2005; (Accessed on 21 May 2006).