Learning from L oops - Applying CLD to Identify Leverage
Points as Organizational Learning
Rod Sarah
Learning and Development Officer,
Monyx Services Pty Ltd
16 Miles St
Mulgrave 3170
Australia
Rod.Sarah@ monyx.com
Andrew O’Brien
CEO, Monyx Services Pty Ltd
Mulgrave, Australia
Dr Tim Haslett
Senior Lecturer,
Dept of Management, Monash University
Melbourne, Australia
Abstract
In this paper submitted to the special session on “System Dynamics and Organizational
Learning: purposes, practices and pitfalls”, we describe a system dynamics intervention
using the principles of Group Model Building to uncover the deep-seated issues in a
long-standing problem area of an organization. In describing this intervention, we will
outline how we used causal loop diagrams to represent the system in focus and identify
leverage points for improvement. This intervention is positioned as part of an
organizational capability-building program involving a university postgraduate degree.
This paper will analyse our learning in the context of the research question outlined in
the special session proposal.
Overview
In his call for papers to the special session on “System Dynamics and Organizational
Learning”, Max Visser raises several key issues regarding the goals of system dynamics
modeling and the achievement of learning. In this paper, we describe a system dynamics
intervention to uncover the deep-seated issues in a long-standing problem area of the
organization. In doing so, we contribute to the understanding of the level of learning
that can be achieved by a careful analysis of this intervention. The paper is presented as
a narrative case study to enable fellow system dynamicists to understand the context and
approach. It concludes with an analysis that is a comparison to the literature on
organizational learning.
Page 1 of 18
The Organizational Environment
The case study is positioned in Monyx Pty Ltd (hereafter referred to as Monyx), a
medium-sized student services organization within Monash University which is based
in Melbourne, Australia. The “service provider’ businesses that Monyx operates for the
university include careers and employment services, student affairs services, student
Tights, welfare assistance services, student co-curricular programs, sport and recreation
services and facilities, clubs and activities, bookshops, a range of retail services and
food and beverage services.
Monyx commenced trading in August 2003 following a start up phase that involved a
complex merger. The merger involved eight organizational entities, five different
accounting systems and multiple employers and employment instruments. Using the
language of system dynamics, the context was one of complex interactions and
interrelationships, with multiple feedback loops containing embedded lag effects.
Of note is the challenge associated with the merger was that none of the existing
accounting and finance systems had capacity to provide for the new organization. A
new accounting system was thus required whilst at the same time continuing to trade
using the pre-existing systems. It is this challenge of moving from five accounting
systems to one that is the focus of this paper.
Operational History of the Case Study
The merging and integration of eight organizational entities and five different
accounting systems during 2002-2004 was not without its challenges. Over this period,
several different managers had focused on improving the performance of the finance
reporting and processing function within Monyx, however despite some short-term
improvements and changes, various service provider business leaders continued to
express concer and frustration about financial reporting and processing support for
their work units. The locally driven improvement projects that were conducted each had
client representative involvement, yet despite some short-term improvement, the
dissatisfaction levels from clients would reappear. In appraising the situation, the CEO
expressed frustration with a perceived internal communication breakdown and an
increase in what Argyris (1991) calls defensive routines. Despite earlier work done on
building skills in dialogue and the intentional design of physical spaces for constructive
conversations, the real story remained hidden from view.
Frustrated with the lack of success with interventions for improvement to date, the CEO
reviewed these various improvement initiatives undertaken to date. Many of the
interventions followed a recurring pattern that began by using a “now - where - how”
representation of the problem or situation, with many of these problems fitting the
description of firefighting as outlined by Kim (2001) and Repenning et al. (2001). As an
expression of a gap analysis, and as a simple balancing loop, as system dynamicists, we
know that this simple dynamic will lead to goal seeking and improvement behaviour if
all elements associated with the system-in-focus had been considered.
The CEO’s assessment was that the improvement project teams had drawn a far too
narrow system boundary to define the problem domain in these earlier attempts to
improve service delivery and the increase in the associated defensive routines provided
and indicator that finance was operating as a closed system (Emery 1981).
Page 2 of 18
What became apparent was that a systems-informed inquiry into our finance problems
would treat the entire finance group as a system and thus we needed an intervention
methodology that acted with a principle of systems integrity. The principle used was
that systems perform the way they perform because of the way they are structured. The
elements of systems are linked together, where every element can affect, and be affected
by, every other element in the system by what are referred to as feedback loops. This
principle was fundamental in choosing our approach.
Therefore, in October 2004 the CEO initiated, led and participated in a project to “map
the finance system”. A team of staff from across the organization was brought together
to conduct what emerged into a Group Model Building project. At the macro level, the
group was asked to map the finance challenge to create shared understanding of the
issues involved, identify short-term actions that would lead to long-term solutions,
involve the finance people and wider organization in identifying agreed leverage points,
and develop an implementation plan for long-term solutions.
Figure 1 below is an early representation as a “stakeholder diagram” of the key parties
in this project was developed by the project team as part of the process of understanding
the high-level relationships that existed in the problem and to assist in drawing the
“system boundary” of this problem (Churchman 1979). In this diagram, we show the
relationships between Monyx, the University and the Monyx Board. Monyx has a
number of Service Provider (SP) business units, who are in tum serviced by Finance
and related processing services, divided into a number of work cells, each staffed by
individuals.
Figure 1: High-Level Finance Stakeholder Diagram
=
&
S
As a systems thinking and system dynamics intervention, the systems mapping
approach using causal loop diagrams was chosen because of the organisation’s prior
exposure to systems thinking as well as through an executive development program that
was underway at that time.
Page 3 of 18
A primary interest in the design of that executive development program was the transfer
of classroom-based learning to practice within our organization. In designing the
program, a challenge was to identify what (infra)structures provides stability and
continuity for organizational learning to occur whereby students as practicing managers
could leverage the program to diffuse and consolidate their individual learnings and
practices? Running parallel to the formal three years of structured classroom learning,
we wanted to have organizationally structured opportunities to apply the learnings from
the program. This paper is the case study report of one such applied opportunity.
Reflection, sense-making and making sense using causal loop diagrams
As anew entity, Monyx was engaged in a significant cultural transformation associated
with becoming a values-based organization that valued learning and included the goal of
creating a learning culture with a commitment to becoming a Leaming Organisation.
Specifically, Monyx aims to have learning as an element central to all its activities so
that as it takes action, it learns what worked and what didn’t, reflecting on why but also
striving to improve through continuous improvement.
This philosophy of “learning (y)our way forward” is central to and consistent with the
organization’ s approach to learning and knowledge accumulation.
Learning by reflecting on our experiences and how we make sense of those experiences
has been another central idea adopted by Monyx - the idea that we retrospectively make
sense of a situation after the event, and how we may have contributed to that situation.
By integrating elements of our physical culture with social practices of reflection and
review with conversational capacity to support team leaming using dialogue, we have
had a base upon which to have constructive conversations that deal with the otherwise
undiscussable issues and defensive routines that block learning (Argyris 1999; Argyris
and Schon 1996). Consistent with the idea that sense-making occurs retrospectively
(Weick 1995), the ‘stock’ of organizational knowledge is seen as continually being
refreshed through processes of selection and retention as learning, as well as unlearning
that comes from conversation.
As a retrospectively imposed structure on our experience, Monyx has made extensive
use of what are variously referred to as influence diagrams (Coyle 2000) and causal
loop diagrams (Senge 1990; Senge et al. 1994). Preferring to use the latter term, Monyx
has made significant use of causal loop diagrams to make sense of organizational action.
A causal loop diagram is a form of knowledge and while expressed diagrammatically, is
often accompanied by a narrative as a set of events linked together sequentially. The
logic of a system dynamics narrative is not “if - then - else...” rather it is a sequential
“first - second - third...” logic where the sequence of events implies a plot of a story.
As part of this approach, the organization has developed an awareness of systems
archetypes (Kim 2000; 2000; Kim and Anderson 1998; Kim and Lannon 1997;
Wolstenholme 2004) and used them for understanding day to day challenges. For
example, in 2003 the central communication device in a Monyx Board paper revolved
around a causal loop diagram, locally referred to as the “Tragedy of the CAF” that
explained how competing parties in the organization were competing against each other
for increasing resources, with the unintended consequence of student fee increases
reaching a threshold limit beyond which further rises were not able to be automatically
Page 4 of 18
applied. This had the added advantage of legitimizing the use of system dynamics
method of causal loop diagrams at a strategic level in the organization.
Background to the Group Model Building Project Team and Methodology
The project team, akin to a Group Model Building team (Richardson and Andersen
1995; Vennix 1996) was drawn from inside the organization and consisted of four
members, including the CEO as sponsor. All members had previous experience in
system dynamics causal loop diagramming, with one member having almost ten years
experience that included formal simulation modeling expertise. None of the members
were members of the local Monyx finance team. This was done to minimise any
perception of pre-determined outcomes.
In this facilitation, systems thinking gave the Group Model Building (GMB) team the
insight to inquire into how we as facilitators contributed to creating a space and climate
and behaviour set, and to improve our own behaviours to improve the quality of
conversation. We wanted a place where we could “diffuse tensions” and “air’ and
“share” in safety.
The methodology chosen was one of semi-structured group interviews using a list of
short prompting questions to act as conversation starters however, outside of appointed
times for stakeholder participants to join the GMB team; the interview structure was
highly fluid and contextual. In designing the process we felt that as facilitators and
modelers, if there had been a communication breakdown then if we begin our attempt to
understanding what appeared to be a difficult and complicated issue by asking leading
questions, then we actually could be contributing to a defensive reactions from others in
the group. Therefore, we framed our interviews by recognizing the value of questions
for leaming (Brown 2002; Marquardt 2005). The GMB team promised confidentiality
and that apart from asking a few prompting questions, our facilitation would focus on
listening and further questioning would be for clarification. In an advocacy/inquiry
sense, the primary mode of behaviour would be inquiry (Senge et al. 1994)
The Group Model Building approach has been well documented in the literature for
example, (Richardson and Andersen 1995; Vennix 1996) a special edition of the System
Dynamic Review (1997 - Vol 13 No 2) on Group Model Building. While it is
recommended that the five roles - a facilitator to monitor the group process and to elicit
group knowledge, a modeler/reflector to focus on the model building, a process coach to
focus on the dynamics of the GMB team, a recorder to write down the groups processes
as a basis for later reflection and learning and a gatekeeper who is close to the client
group and who takes responsibility for the project - are distinct and essential, with a
GMB team size of four there was an obvious potential for overlapping membership of
roles. In fact, what we had was akin to a Venn-Diagram where there were multiple
overlapping roles with different people moving in and out of roles based on the
situation.
This was a unique situation - the group had modeling expertise and all members were
all organizational insiders undertaking “insider research”. All had significant experience
in the conduct of group processes for eliciting information and drawing out individual
perception of elements of the problem. There was joint process facilitation
responsibility - no one had an appointed role, yet all performed this role both in the
larger group process and with the participants. What was further novel was that the
Page 5 of 18
gatekeeper role as sponsor was a member of the GMB team. This group did not use the
gatekeeper role. Thus, rather than managing across all five roles, the group was able to
focus on the remaining four roles to maximise the data gathering element of the project
As the interviews were conducted, each GMB team member acted in all roles
informally. The modeler/reflector activity was the least active of the four roles. During
the two days of data gathering interviews, it was more important to be facilitating the
group elicitation process than have a focus on getting the detail of the model correct.
The interview setting was a large room with circular table and chairs and whiteboards.
Some participants simply spoke and engaged in a dialogue with the GMB team, while
others chose to use whiteboard representations of their interpretation of the facts as they
saw it. A stand-out example used to describe the situation made use of a rich pictures
(Checkland 1999; Waring 1996) as shown at Appendix 1. The power of this “rich
picture” as a diagrammatic summary is that it uses images of grenades, brick-walls and
locked doors to represent some of the interpersonal issues and relationship dynamics
that one person feels occur in this system. It leaves all parties with no doubt as to what
the key issues and variables may be.
Each interview session was followed with a GMB team debrief of the group processes
and what key data items were gleaned. This was a deliberate process step by the GMB
team. The focus was on gathering data and sensing, and actively resisting moving to
making sense of the data and beginning to develop any concept models prematurely.
This was done explicitly to avoid what is referred to as the anchoring and adjustment
bias/heuristic (V ennix 1996).
Towards the end of the second day, the GMB team prepared an overview of issues to
the stakeholders, which included those interviewed from the finance team and wider
organization. Table 1 provides a summary of the issues raised during interviews.
Table 1 - Issues identified in interviews
High degree of complexity. Delays from outside finance derail finance.
Almost zero margin for error. Lack of understanding of the whole system.
Lack of shared vision and direction. Lack of confidence in accounting system.
Finance as a closed system. Pressure to contain cost of finance.
Delays and inaccuracies. Need to plan and find role clarity.
Different finance systems across the business.
Short-term appointments.
Damaged relationships including open hostility.
Us and them mentality.
Concems over leadership - outside group.
Concems over leadership - inside group.
People afraid to raise issues.
Concem over time taken.
Attack - constant criticism from outside.
Defense and blame - your fault not mine.
We can’t make a difference.
Giving up, someone else will fix it.
Lack of senior manager and CEO input.
Too much senior manager and CEO input.
Wrong company structure.
Problems with some software.
Fatigue.
Project creep.
Lack of feedback to finance.
Feedback to finance not heard or accepted.
Acceptance of poor performance.
Lack of understanding of interconnectedness.
Low and falling morale.
Band-aid solutions.
Multiple and conflicting stakeholder priorities.
Lack of resources.
Loss of confidence.
Constant firefighting.
Lack of recognition for achievement.
Poor business literacy outside finance.
Focus on short-term deadlines only.
Lack of time for training and planning.
Source: Authors
Page 6 of 18
A number of systems archetypes were evident within the issues outlined. The
breakdown in relationships and increasing hostility were recognized by the GMB team
as representing the escalation archetype (Kim and Lannon 1997) represented by a
spiraling cycle of blame and defensive reactions amongst a number of the parties. Also
noticeable was a giving up or reliance on others to solve problems and the subsequent
reinforcement of this behaviour as others solved problems. The GMB team recognized
the shifting the burden archetype (Kim and Anderson 1998) as a useful explanation of
this dynamic. Drifting goals and the tragedy of the commons (Kim 2000, 2000) were
amongst the other archetypes identified and these insights assisted understanding the
“wicked” problem facing the organization.
As part of this review the GMB team presented a “first cut” or “in the moment” causal
loop diagram with a view to telling the story they had discovered it from the interviews.
Figure 2 outlines this loop diagram.
Business
Literacy
Time for Training /
Planning
Complexity
Volume
Abiity b Use
pe
Lack of Shared
Understanding Actual
Systems
Capacity
Wim Lack ofrecogniton of
Achievernent
in System
Us/ Them
Unclear Expectations
Lack ofRole Clariy
Responsibitiy/
Accountabies
Focus on Abily
Relationship eee
Decline
Time ness:
Leadership
Role Clariy
Me Pressure/Stress
Defensive eet
Leaders Negatve
We
Feedback
Figure 2: Initial Loop diagram
Whilst generally well received and validated by most participants, one of the finance
leaders reacted with hostility and suggested that the real issue was complexity and that
the loop diagram as presented totally overlooked this issue. Whilst recognizing this as a
valid perspective, the GMB team took the view that legal structures, government policy,
changing stakeholder expectations and a diverse business with complex accounting
requirements were all reasons for the complex environment, and the real leverage was in
finding ways to address complexity.
Page 7 of 18
Following an overview of the issues the stakeholder group participated in a world café
(Brown 2001, 2002; Brown and Isaacs 1996, 2001; Brown and Isaacs 2005) to explore
options for solutions. The stakeholder group was asked to consider short term actions to
address major concerns whilst the group model builders refined the causal loop
diagram, provided opportunities for more input and complete and final loop diagram
and identify major leverage points. The first cut causal loop diagram provided initial
clarity of the issues and facilitated an agreed set of actions designed to build momentum
as outlined in Table 2.
Table 2 - Initial steps for a long term solution
Finish the causal loop diagram of the “whole” | Establish and communicate role clarity.
system.
Conduct training program to develop capacity to | Develop strategies for finance to be seen as and
act. feel part of the wider organization.
Upgrade cashbook module. Integrate stock and general ledger systems.
Loop diagram finance processes. Develop business literacy so as to reduce delays
outside of finance.
Source: Authors
Following the initial interviews and stakeholder group report back as part of the world
café, each of the GMB team members began to develop their own representation of the
data as causal loop diagrams. This was done for two reasons: the pragmatics of four
mid-to-senior managers being off-line for an intensive period of time, and not wanting
not come to a premature conclusion of what an ‘integrated’ causal loop diagram would
look like. Each GMB team member developed individual causal loop diagrams.
There was however a strong requirement for “evidence-based modeling” and in the
presentations, each model and representation was subjected to critical review, challenge,
correction and iteration as a final diagram began to emerge. The need for evidence or
data driven modeling whereby the causal relationships could be traced back to
individual statements recorded in the interviews was felt essential for the eventual
acceptance of the model by the participants.
Thus, while the models were built individually at an initial stage, they were created in a
group process within the GMB team. Potential policy insights began to emerge because
of group dynamics within the GMB team gatherings, not because of one expert modeler
presenting a finalised model back to the GMB team. The skill of dialogue proved
invaluable here for GMB team learning to occur which led to a more robust final
version that could then be tested and validated back to the stakeholder group.
Making Sense of the Data
During the concluding phases of developing a causal loop diagram, two insights
emerged. The first came from those GMB team members attending the 22"! ISDC at
Oxford in July 2004 when they drew upon the Forrester Award winning presentation
(Wolstenholme 2004). In his acceptance speech, Wolstenholme presented his research
findings on the importance of organizational boundaries as they apply to system
archetypes to highlight the difference between what are intended and unintended
consequences of individual actions. The boundary management problem (Wolstenholme
2004, p. 344) where flows across boundaries need to be managed appeared part of the
analysis of this project.
Page 8 of 18
The second insight came as another GMB team member was attending the Systems
Thinking in Action (STIA) Conference in Cambridge, Massachusetts in December 2004
where John Sterman and Bill Isaacs presented work on change efforts and the
challenges associated with fire-fighting and tipping points. As part of this presentation a
model on process improvement was presented drawing on the work of Repenning and
Sterman (2001) in which the work harder, work smarter and shortcut balancing loops
represented similar if not identical dynamics to those being experienced by the Monyx
finance group.
Taken together, these offered archetypical reference modes, not in the traditional sense
of reference mode as used in quantitative system dynamics simulation modeling, but as
case study narratives that described very similar dynamics to what the participants in the
finance, causal loop diagramming project had told to the GMB team.
Drawing the Loop diagrams for Organization-wide Learning
In developing our finance loop diagrams as causal loop diagrams, we were confronted
with what Espejo et al. (1996) termed “the Regulatory Models Dilemma”, which is that
the more complex the models, the more difficult it is to share them and the less likely
organizational learning will occur (Espejo et al. 1996, p. 187). In the language of
cybemetics that Espejo et al. writes from, for managers who “control” an organizational
function, the detailed complexity of the model necessary for “regulation” of a system
means that those who develop the model are more than likely the only organizational
members able to comprehend and understand the model. However, these complex
models are difficult to use for communication and learning and thus the requirements
for models used for “regulation” are different to those if the model is to be used for
organizational learning (Espejo et al. 1996, p. 185).
The “final” causal loop diagram is shown as Appendix 2. This was an example of what
the GMB team would categorize as a “regulatory” level causal loop diagram. As a GMB
team, we were cognizant of the aforementioned “regulation-leaming” dilemma. This is
one of the great advantages of system archetypes (Wolstenholme 2004) namely to create
simple models that are explicit. Without this simplicity of what is the far side of the
organizational complexity, there can only be individual learning.
Organizational leaming can occur when the causal loop diagram or model can be
communicated in clear and unambiguous terms (Espejo et al. 1996). As representations,
archetypical diagrams that are simple enough to be shared are not the detailed
regulatory models themselves, but are “statements” about the interfaces between the
relationships between stakeholders, sub-systems and local sections or sectors that enable
learning (Espejo et al. 1996, p. 188). This appeared to be an example of the
organizational (sub-system) boundaries management problem referred to by
Wolstenholme (2004, p. 344).
In response to this finding, the GMB team produced a final model for learning which is
reproduced at Appendix 3. While it is not a “classic” system dynamics causal loop
diagram, complete with appropriate signage of the causality on causal arrows or
identifying loop polarity, our purpose was for communication and learning.
Page 9 of 18
Additionally, while some in the target audience had prior exposure to causal loop
diagrams at this time, many had not, so it was decided to “keep it simple and short”.
Using PowerPoint build features, this loop diagram was built up in a series of steps as
the GMB team “unpacked” the loop diagram incrementally and told a story based on
interview and café responses. The simpler model was well received, specifically by our
“naming” of certain problematic dynamics in loops.
One initiative the GMB team chose to follow was that of creating “quick wins”. This is
not the same as quick fixes, with their associated unintended consequences and side
effects. A quick win in this situation is akin to the importance in Kotter’s “eight stage
process” of change (Kotter 1996) which he outlines the importance of generating short-
term wins. Working as external consultants to the finance group, the GMB team was
very cognizant of the importance of producing demonstrable results. This meant that
people needed to see quick wins to build the confidence in the process that would
potentially create long-term benefit. The ideas generated in the large group world café
as outlined in table 2 provided the basis for actions that met the criteria for ‘quick wins’.
As these actions were generated by the group it was felt that to implement as many of
these at possible would develop confidence in the process and create a sense of
momentum.
Leveraging from the finance loop diagrams as causal loop diagrams revealed several
actions with short implementation times that could achieve some improvements, with
none of the unintended consequences of a quick fix. This is an application of the
concept of acting locally within a sub-system while thinking globally across the
containing system.
Learning
We were able to map in causal loop diagrams the feedback loop driven behaviours
exhibited in part as a result of multiple locations across the finance function which were
affecting the group dynamics within the finance sub-teams and between the finance
function and their clients, the Monyx businesses. While the geographic location was
raised as an issue, there were other dynamics revealed in our interviews.
There was another dominant pattern of explanation embedded in the stories which
related to the complexity of the legal structures, the taxation implications of the formal
structure and the compliance and govemance responsibilities on the finance team and
management. The solution suggested in these stories was to simplify the formal
structure of the organization - and that this story had been being told for three years and
if only the senior executives, leadership and Board of Management taken this advice,
then the issues behind the review would not exist.
In parallel with the explanation of the complex legal structure, were the problems with
the computerized accounting systems. The organization used multiple systems, some of
which were quite old and, as legacy systems were difficult to maintain in part due to a
lack of skill and expertise. The lack of a simpler and single accounting system further
complicated the skilling and capacity of the finance people to perform their functions.
Similar to the legal structures, the solution in these stories was to simplify the
accounting systems by consolidating them into one new system and that had this advice
been taken over the last three years, then this review would not have been required.
Page 10 of 18
The GMB team was not swayed by any one of these single categorizations. What most
of the stories talked about was a physical, legal or technical solution to what appeared to
be an out of control social problem around interpersonal relationships caused in part by
an incremental deterioration of communication. Despite establishing a set of guiding
principles and values that include the aspiration towards a learning organization and the
importance of learning, a structured training program conducted during 2002 on
dialogue, the purposeful design of our physical spaces to support conversation and
interaction, and a cultural practice of valuing conversation, the communication
breakdown emerged as the enduring theme. In complex human social systems such as
modem organizations, the soft issues are the hard issues.
The GMB team reflected on the research and writings on undiscussables and
organizational defensive practices (Argyris 2000) which deal with defensive routines
established by people to protect themselves from the threat of embarrassment. What the
GMB team noted was that in their one-on-one dealings with participants in our project,
this was demonstrably lacking in their interactions. However, in their descriptions of the
workplace dynamics both within the finance functions and between finance and its
clients, genuine conversations were not occurring. We heard many examples of stories
that appeared to the GMB team to be examples of the ladder of inference (Argyris 1993;
2000; Dick and Dalmau 1999) in that the application of attributions that were untested.
What we heard were stories of behaviour that the GMB team was able to rephrase in
their making sense phase as examples of escalating feedback dynamics resulting in the
breakdowns of personal relationships and poor team dynamics.
‘General Laws of Systems Thinking’
In making sense of the learnings of the processes, we were also looking for generalised
learnings that could be applied in other situations in Monyx. There was a recurring
“plame-no blame” tension during the interviews that we wanted to capture. As well, the
system's history in terms of cause and effect are distant in time and place was a repeated
theme in describing the finance system as complex. Both these elements are revealed in
Laws of Systems Thinking in Chapter 4 of The Fifth Discipline (Senge 1990).
These set of eleven Laws of Systems Thinking are of course interrelated and it is hard to
talk about one law without referring to another. However, two reoccurred most
frequently and appeared to the GMB team as tightly coupled:
Law 11 - There is no blame.
Systems thinking tells us that there is no outside or “them” in looking for
the cause of a problem, rather that the individual or “you” are a part of
the cause of the problem and that you and the problem are
interconnected as part of a system that we can define as the “problem
situation”.
In this intervention, our systems thinking premise that we are all part of
the system was that we should look to our own actions rather than others
as the drivers of the problematic behaviour. In the final communication
to all the finance team, this was a core message from the GMB team.
Page 11 of 18
Law 7: Cause and effect are not closely related in time and space.
We typically think of problems from the perspective of an isolated event.
However, in organizations structured as complex networks of
relationships that have evolved over time and the genesis of a problem
occurs as issues deteriorate over time as each part builds actions based
on higher and higher levels of untested assumptions and information.
This was again a focus of the debriefing and this law reinforced the no-
blame stance of Law 11.
Additionally, in the final presentation, the GMB team drew on Law 6 - Faster is Slower
- to loop diagram out a suggested improvement schedule. In addition to some quick
wins to make some short-term improvements, several medium to longer-term
improvements were simultaneously initiated. These longer-term improvements involved
changes that would be implemented slowly over the coming period. It was this message
given to both members of the finance function and the business units by the GMB team,
with CEO endorsement as gatekeeper, sponsor and GMB team member, which allowed
the finance team to move beyond the issues that lead to the intervention. From this
perspective the intervention was seen as a success.
Organisational Responses
As already outlined, several short term “quick win” initiatives were implemented. The
longer-term “slower is faster” responses centred on restructuring the internal dynamics
within the various finance functions, and between the finance functions and their clients.
The purpose is not to outline in detail what changes Monyx implemented because of the
GMB team study.
However, the premise of this paper remained ‘structure drives behaviour’ in the terms of
organizational performance, thus the focus was on the interpersonal structural
relationships. The levels of perspective model (Kim 2001) provided a useful framework
for working with mental models and developing collective understanding of the
relationship between vision, mental models, systemic structures and behaviours and
outcomes. This model provided the background from which we identified actions
including a re-focus on finance team communication skills, including team dynamics,
team roles, personality types and preferences, as well as a revisiting and refreshing of
communication and listening capabilities including dialogue training.
Insights into what is Organizational Learning
Organizations exist as socially constructed realities in which the reality we know is
interpreted, constructed, enacted and maintained through discourse (Ford 1999, p. 480).
This view is supported by Berger and Luckman (1966) when suggesting that
conversation creates and maintains our reality whilst Fonseca (2002) explores the link
between conversation and social reality when discussing the view of Shotter (1993) that
communication and language provide the vehicle of influencing thinking. The group
model building process as outlined in this paper has drawn on this view of organizations
and the challenges they face as being socially constructed by the participants.
Page 12 of 18
The concept of collective intentionality, which involves people forming shared beliefs,
desires and intentions, is introduced by Searle (1995) when discussing the construction
of social reality and it is this suggestion that organizations can reach shared objectives
that underpins the approach taken in this group process that is the focus of this paper.
By forming a shared view of the system, the problems involved and the actions to be
taken the people involved in this process have been able to form a degree of collective
intentionality and move in the desired direction. Constructivist validation is important
for interpretive models as the idea is to provide new keys that might unlock new
insights (Pidd 2003, p. 310) and the outcomes achieved have supported the social
construction of the model that was developed.
Implicit in our organizations approach to organizational learning is organizations work
the way they do because people work the way they do. And people work they way they
do because of the way they think, and further, they work the way they do together
because of how they think together. Therefore, if you want to change the way people
work together, then you need to enable people to think differently - together. In the
language of Argyris and Schon (1978; 1996), and Argyris (1999), we are enabling
people to engage in “double-loop learning”, and in the language of Checkland (1985) to
reflect on their framework of ideas; that is to change their mental models.
This study has shown how Group Model Building, when based within a conversational
learning-based culture, has achieved organizational learning. The use of causal loop
diagrams for learning has led to changes in behaviour and improvements in
performance. We have demonstrated that a causal loop diagram is both a source of
knowledge and a process for learning.
Like many similar studies, the real learning is in the process, rather than in the content
or outcome. In addition, the deep learning exists in the premises that sit behind the
process of the intervention.
Page 13 of 18
References
Argyris, C. 1991. Teaching smart people how to learn. Harvard Business Review 69
(3):99-109.
———.1993. Knowledge For Action. San Francisco: Jossey-Bass Inc.
———.1999. On Organizational Learning. Oxford: Blackwell.
———. 2000. Flawed Advice and the Management Trap. New York: Oxford
University Press Inc.
Argyris, C., and Schon, D. 1978. Organizational Learning: A Theory of Action
Perspective. Massachusetts: Addison-Wesley Publishing.
——-—. 1996. Organizational Learning II: Theory, Method and Practice.
Massachusetts: A ddison-W esley Publishing.
Berger, P. L., and Luckman, T. 1966. The Social Construction of Reality. First ed. New
Y ork: Anchor Books.
Brown, J. 2001. The world cafe - living knowledge through conversations that matter.
Ph.D. dissertation, The Fielding Institute.
— ——. 2002. The World Cafe: A Resource Guide for Hosting Conversations That
Matter. Mill Valley, California: Whole Systems Associates.
Brown, J., and Isaacs, D. 1996. Conversation as a core business process. The Systems
Thinker 7 (10):2-6.
———. 2001. The world cafe: living knowledge through conversations that matter. The
Systems Thinker 12 (5):1-5.
———. 2005. The World Cafe. San Francisco: Berrett-K oehler Publishers, Inc.
Checkland, P. 1985. From Optimizing to Learning: A Development of Systems
Thinking for the 1990's. Journal of the Operations Research Society 36 (9):757-
767.
— ——. 1999. Systems, Thinking, Systems Practice. Chichester: John Wiley & Sons
Ltd.
Churchman, C. W. 1979. The Systems Approach and Its Enemies. New Y ork: Basic
Books Inc.
Coyle, R. G. 2000. Qualitative and quantitative modelling in system dynamics: some
research questions. System Dynamics Review 16 (3):225-244.
Dick, B., and Dalmau, T. 1999. Values in action - applying the ideas of Argyris and
Schon. Second ed. Chapel Hill: Interchange.
Emery, F. E., ed. 1981. Systems Thinking (2 Volumes). Harmondsworth, England:
Penguin Books.
Espejo, R., Schuhmann, W., Schwaninger, M., and Bilello, U. 1996. Organizational
Transformation and Learning: A Cybernetic Approach to Management.
Chichester: John Wiley & Sons Ltd.
Fonseca, J. 2002. Complexity and Innovation in Organizations. London: Routledge.
Ford, J. D. 1999. Organizational change as shifting conversations. Journal of
Organizational Change Management 12 (6):480-500.
Kim, D. H. 2000. Systems Archetypes I, Toolbox Reprint Series. Waltham: Pegasus
Communications, Inc. Original edition, August 1992.
———. 2000. Systems Archetypes II - Using Systems Archetypes to Take Effective
Action, Toolbox Reprint. Waltham: Pegasus Communications, Inc.
———. 2001. Organizing For Learning. Waltham: Pegasus Communications Inc.
Kim, D. H., and Anderson, V. 1998. Systems Archetype Basics. Waltham: Pegasus
Communications Inc.
Page 14 of 18
Kim, D. H., and Lannon, C. 1997. Applying Systems Archetypes. Waltham: Pegasus
Communications Inc.
Kotter, J. P. 1996. Leading Change. Boston: Harvard Business School Press.
Marquardt, M. J. 2005. Leading with Questions. San Francisco: Jossey-Bass.
Pidd, M. 2003. Tools For Thinking - Modelling in Management Science. Second ed.
Chichester: John Wiley & Sons Ltd.
Repenning, N. P., Goncalves, P., and Black, L. J. 2001. Past the Tipping Point: The
Persistence of Firefighting in Product Development. California Management
Review 43 (4):44-63.
Repenning, N. P., and Sterman, J. D. 2001. Nobody Ever Gets Credit for Fixing
Problems that Never Happened. California Management Review 43 (4):64-88.
Richardson, G. P., and Andersen, D. F. 1995. Teamwork in group model building.
System Dynamics Review 11 (2):113-137.
Searle, J. R. 1995. The Construction Of Social Reality. New Y ork: The Free Press.
Senge, P. 1990. The Fifth Discipline: The Art and Practice of the Learning
Organization. New Y ork: Doubleday Currency.
Senge, P. M., Kleiner, A., Roberts, C., Ross, R. B., and Smith, B. J. 1994. The Fifth
Discipline Fieldbook. London: Nicholas Brealey Publishing.
Shotter, J. 1993. Conversational Realities - Constructing Life Through Language.
London: Sage Publications.
Vennix, J. M. 1996. Group Model Building: Facilitating Team Learning Using System
Dynamics. Chichester: John Wiley and Son.
Waring, A. 1996. Practical Systems Thinking. London: Thomson Leaming.
Weick, K. 1995. Sensemaking in Organizations. Thousand Oaks, CA: Sage.
Wolstenholme, E. F. 2004. Using generic system archetypes to support thinking and
modelling. System Dynamics Review 20 (4):341-356.
Page 15 of 18
Appendix 1: Rich Picture Diagram representation of some of the ‘systems’ dynamics as told by one respondent.
Note: Names have been altered for anonymity
fosT—-UW ree / Neil
eT
/
. Berra”
(ile ss a OEY iin: ay
Al
RON oes in lad ; ou
on ~ wr
ag LE
ire, otis
Page 16 of 18
Appendix 2: Detailed Loop diagram of the Finance System
Appendix 3: Final Loop diagram of the Finance System used for Organizational Leaming
Same People in
Different Rok
Levels
“Ss ‘Multiple Competing .
Demands & Pade i
Desire to work:
collaboratively
Page 18 of 18
