COMMONALITY VERSUS VARIABILITY
The Contradictory Nature of Enterprise Systems
Stig Nordheim
Agder University College, P.O Box 422, N-4604 Kristiansand, Norway
Keywords: Commonality, Variability, Enterprise Systems, Dialectics.
Abstract: This position paper argues that there is a major contradiction inherent in Enterprise Systems. The evidence
for this contradiction is seen in the meta level concepts of commonality and variability that characterize
Enterprise Systems. The inherent contradiction between commonality and variability is discussed in the
light of Enterprise Systems’ literature, and the contradiction is elaborated. Then an Enterprise Systems’
vendor perspective on commonality and variability is presented, based on interviews. A challenge of this
inherent contradiction is its synthesis, finding the right balance between commonality and variability.
1 INTRODUCTION
Enterprise Systems (ES) are commercial software
packages that enable the integration of transaction-
oriented data and business processes throughout an
organization (Markus and Tanis, 2000). Examples of
ES include: Enterprise Resource Planning,
Enterprise Content Management, Supply-Chain
Management, Customer Relations Management and
Data Warehousing.
This paper argues that there is a major
contradiction inherent in ES. A contradiction is a
relationship between two opposite aspects of a
phenomenon. The evidence for asserting that there is
an inherent contradiction in ES is presented by
means of the meta level concepts of commonality
and variability that characterize ES. In the following,
the meta level concepts of commonality and
variability are discussed in the light of the ES
literature, and then the concept of contradiction is
presented. A system level contradiction that is
asserted to be inherent in ES is then presented, and
the contradiction is illustrated by interview data
from an ES vendor.
2 COMMONALITY AND
VARIABILITY
Enterprise systems (ES) are configurable systems,
based on the concepts of “commonality” and
“variability” as discussed by Leishman (1999).
These concepts are related to fit and alterability. The
concepts are based on understanding the
commonality and variability (c/v) across industries,
geographies, customers, and systems.
The general meaning of commonality is
“possession of common (i.e. widespread or general)
features or attributes” (Merriam-Webster, 1996). In
software engineering a commonality is seen as “an
assumption held uniformly across a given set of
objects (S)” (Coplien et al., 1998), and commonality
is represented by the software code. Variability is
generally understood as “subject to variation or
changes” (Merriam-Webster, 1996). Variability in
the context of software engineering is seen as “an
assumption true of only some elements of S”
(Coplien et al., 1998.), and is implemented in
different ways. In this context, variabilities are
“bound” by placing specific limits on each of the
variabilities. One example of bounded variability
may be a range of legal values for a parameter
(Coplien et al., 1998).
By providing variability in the ES package the
vendor anticipates that the customer has a certain
need for flexibility. The scope of the variability is
typically predefined and with a set of options.
Parameter configuration is one example of
predefined variability, and black-box customization
of frameworks is another. This concept of a
predefined variability fits with the software
engineering perspective on variability (Coplien et
al., 1998), that variability is bounded.
Based on a review of different types of systems,
Leishman (1999) identifies a number of generic
mechanisms available for implementing variability.
572
Nordheim S. (2006).
COMMONALITY VERSUS VARIABILITY - The Contradictory Nature of Enterprise Systems.
In Proceedings of the Eighth International Conference on Enterprise Information Systems - ISAS, pages 572-575
DOI: 10.5220/0002469505720575
Copyright
c
SciTePress
Variability here applies both to extensions of the
system and integration with other applications. This
variability may be implemented by mechanisms
such as configuration, subclassing and inheritance,
specialization of patterns, multiple versions of
components, instantiation of abstract classes,
parameters, templates, DDLs, customer exits,
import/export mechanisms, adapters and connectors,
install scripts, registry, property sheets and
customizers (Leishman, 1999).
Another set of variability mechanisms is
presented by Brehm et al. (2001), who present a
typology of Enterprise Resource Planning (ERP)
adaptation in practice. These ERP variability
mechanisms include configuration, bolt-ons for
implementation of third-party packages, screen
masks, workflow programming, user exits for
programming additional software code in an open
interface, ERP programming using the programming
language provided by the vendor, interface
development, and in rare instances some package
code modification.
A special case of variability is found with Open
Source software, where the commonality itself, the
source code, also becomes subject to variability. But
even in this case some vendors have a kernel of
commonality, e.g. the enterprise content
management vendor eZ Systems (http://ez.no/).
3 CONTRADICTIONS
To explain the contradictory nature of commonality
and variability, we need to clarify the concepts
related to contradictions. A contradiction can be
seen as a relation between two opposite aspects of a
phenomenon. One aspect in a contradiction (also
called a contradistinction) cannot be fully
understood without considering the other aspect. The
two aspects of a contradiction are called thesis and
antithesis, where antithesis is the negation of the
thesis. So the two aspects (or contradistinctions) are
intrinsically related, yet opposite and distinct from
one another (Mathiassen and Nielsen, 1989; 1990).
The combination of thesis and antithesis may lead to
a synthesis, which may be different from both thesis
and antithesis.
A thesis (A) may be challenged by an antithesis
(Not-A), and the resolution of the conflict becomes a
synthesis (which is Not Not-A). By its very nature,
the synthesis is a novel construction that departs
from both the thesis and the antithesis. Over time,
this synthesis may become a new thesis as the
dialectical process continues (Van de Ven and
Poole, 1995).
Analysing contradictions is also referred to as
dialectics, where the dialectical tension is the
opposition between two interacting forces or
elements. The main point in dialectical reflection is
to find the principal contradiction of a process in
order to understand a situation (Bjerknes, 1992).
According to Dahlbom and Mathiassen (1993),
contradictions can in some cases be seen as trade-
offs: “When working with computers … we are
typically faced with what is traditionally called
trade-offs. From a dialectical perspective, these
trade-offs are manifestations of contradictions
inherently related to the use and development of
computer systems” (p63). According to Israel (1979)
dialectics contributes to the production of
knowledge, by an increased understanding of a
phenomenon. For the purpose of this paper, we note
one dialectic to be of particular interest here, namely
the age-old contradiction between stability and
change (Lewis, 2000).
4 THE INHERENT
CONTRADICTION OF
ENTERPRISE SYSTEMS
The mechanisms for commonality and variability
built into Enterprise Systems (ES), may also be seen
as representing a contradiction (cf. Figure 1).
Viewed as a thesis, commonality assumes that the
system properties involved should not and cannot be
changed. A rationale for this is that the system is
based on best practices, and therefore it should not
be changed. This argument is used for ERP systems.
An ES following this thesis to the extreme, would
simply be installed with no configuration
whatsoever, as it would only consist of
commonality. A pure thesis is not a feasible
solution. Viewed as an antithesis, variability
assumes that the system properties involved can and
probably should be changed. A rationale for this is
that the organizational contexts where the system is
installed, may be different, and therefore the system
needs to have its properties set or changed. An ES
following this antithesis to the extreme, would only
consist of variability and have no commonality at
all. So a pure thesis is not a feasible solution either.
COMMONALITY VERSUS VARIABILITY - The Contradictory Nature of Enterprise Systems
573
The thesis of
Commonality:
“None of the
system properties
can be changed”
The antithesis of
Variability:
”All of the system
properties can be
changed”
The contradiction:
Commonality
vs variability
Synthesis:
“Some of the system
properties can be
changed”
The thesis of
Commonality:
“None of the
system properties
can be changed”
The antithesis of
Variability:
”All of the system
properties can be
changed”
The contradiction:
Commonality
vs variability
Synthesis:
“Some of the system
properties can be
changed”
Figure 1: The inherent contradiction of Enterprise
Systems.
The contradiction itself is about the fixed versus
the variable, or stability versus change (Lewis,
2000). Commonality versus variability may
therefore be seen as an inherent dialectic within ES,
expressing the contradictory nature of ES. These
design trade-offs are indeed manifestations of
contradictions inherently related to the development
of the systems (cf. Dahlbom and Mathiassen, 1993).
Because of its inherent commonality/variability
architecture, an ES may therefore be seen as the
embodiment of a contradiction that is fundamental to
the nature of ES. This contradiction is viewed at a
system level and from a vendor perspective.
The synthesis for most ES vendors implies
finding an optimal balance between commonality
and variability, i.e. between stability and change.
This balance means that a subset of the properties of
the system may be changed. Different vendors will
reach very different decisions about the right balance
between commonality and variability (Davenport,
1998).
A puzzling phenomenon is that some ES vendors
provide powerful customization tools, while
emphasizing the importance of limited
configuration. So the vendor’s attitude to stability
and change may even be contradictory.
5 A VENDOR PERSPECTIVE ON
THE CONTRADICTION OF
ENTERPRISE SYSTEMS
To illustrate the inherent contradiction presented
above, the results of interviews with two
representatives from an ES vendor are presented in
the following.
The case studied is eZ Systems (http://ez.no/), an
Open Source enterprise content management (ECM)
vendor with 30-40 thousand downloads per month.
One of the informants interviewed has a split role
between management of customer projects and
programming, and the other informant is responsible
for all customer projects in the company.
The idea behind eZ Systems is flexible solutions
rather than “off-the-shelf” software. So they
emphasize the antithesis of Fig 1. Their customers
are extremely heterogeneous. As this is an Open
Source software, most customers download and
implement it without any contact with the vendor.
An interesting point is that even with this Open
Source enterprise system, there is commonality. The
commonality is in this case a kernel that is
developed and controlled exclusively by eZ
Systems.
There are two types of variability provided by eZ
Systems, configuration and programming.
Configuration is referred to as a supported
variability, and programming is referred to as an
unsupported variability. In addition to configuration,
supported variability also consists of modifying
templates, e.g. HTML templates. Unsupported
variability consists of interface programming,
usually based on existing libraries. So a considerable
variability is achieved by plug-ins into the kernel,
and as the system evolves, plug-ins are extended
without affecting the kernel. Consultancy on
variability ranges from adapting templates for small
businesses, to complex integration with legacy
systems in large enterprises.
The line of demarcation between commonality
and variability is by eZ Systems perceived as an
interesting design trade-off. So from this vendor’s
perspective, the contradictory nature of the ES is
recognized, and an optimal synthesis is seen as a
design challenge. As pointed out by one of the
informants, if there is too much commonality
compared to variability, the customer is being
locked up due to lack of configuration options. But
if there is too much variability compared to
commonality, the customer will be confused due to
the lack of standard functionality. According to the
informants at eZ Systems, an implementation based
on as much commonality as possible means better
maintainability, higher quality and a cheaper
solution.
The vendor’s goal is in this case to have as much
configuration as possible, and as little programming
as possible. The ideal is to empower non-
programmers to establish complex solutions by
simple configuration. A goal is to cover 95% of a
customer’s needs by configuration. This can
therefore be seen as eZ Systems’ view of an ideal
synthesis (c.f. Fig 1). The vendor also tries to
influence the customer’s requirements on to what is
configurable. Configuration constitutes the typical
effort, together with modifying HTML templates.
Functional adaptation and user interfaces are usually
adapted through configuration. In a typical project,
ICEIS 2006 - INFORMATION SYSTEMS ANALYSIS AND SPECIFICATION
574
programming may constitute 10% of the work,
whereas 90% of the effort is configuration and
template adaptation. When programming occurs,
typical programming tasks include: integration with
back-end systems, including the integration of user
administration and access rights. Special processes
for document approval within the implementing
organization may also require some programming.
There is very little programming in a typical project,
but the extent of programming varies enormously.
Larger projects are characterized by some
programming, typically modifying scripts to handle
import/export. In some larger projects there may
also be some programming to create new
functionality, based on existing libraries.
6 CONCLUSIONS
Most of the practice literature referred in this paper
is from an ERP context, since ERP systems is a
prime example of Enterprise Systems. ERP systems
constituted only one of the six example systems that
the vendor literature (Leishman, 1999) was based
on. By examining another type of ES in the case of
eZ Systems, it was found that the categorizations of
commonality and variability were similar to those of
ERP. The ECM vendor eZ Systems followed
categorizations similar to ERP systems, and both are
ES.
We argue that Enterprise systems by nature are
contradictory. They embody an inherent
contradiction at the system level. This is evident
from the concepts of commonality and variability.
The vendor challenge as seen in the case presented,
consists of finding an optimal synthesis. The
synthesis is a good balance between commonality
and variability. As part of this, an apparent vendor
goal is to provide sufficient variability by means of
configuration.
For ES customers the inherent contradiction of
these systems also may give rise to another question:
What are the possible implications of this
contradiction for organizations that are
implementing ES?
REFERENCES
Brehm, L., Heinzl, A., Markus, M.L., Tailoring ERP
Systems: A Spectrum of Choices and their
Implications. Proceedings of the 34th Hawaii
International Conference on Systems Sciences,
Hawaii, 2001.
Coplien, J., Hoffman, D., and Weiss, D., Commonality
and Variability in Software Engineering. IEEE
Software, 1998. 15(6): p. 37-45.
Dahlbom, B. and L. Mathiassen (1993). Computers in
Context : The philosophy and practice of systems
design. Cambridge, Mass., NCC Blackwell.
Davenport, T. H. (1998). "Putting the Enterprise into the
Enterprise System." Harvard Business Review 76(4):
122-131.
http://ez.no/
Israel, J. (1979). The Language of Dialectics and the
Dialectics of Language. Copenhagen, Munksgaard.
Lewis, M. W. (2000). "Exploring Paradox: Toward A
More Comprehensive Guide." Academy of
Management Review 25(4): 760-776.
Leishman, D.A., Solution Customization. IBM Systems
Journal, 38,1 (1999) 76-97.
Markus, M. L. and C. Tanis (2000). The Enterprise
System Experience - From Adoption to Success. In
Framing the domains of IT management : projecting
the future through the past. R. W. Zmud (ed.)
Cincinnati, Ohio, Pinnaflex Education Resources: 173-
207.
Mathiassen, L. and P. A. Nielsen (1989). "Soft Systems
and Hard Contradictions - Approaching the Reality of
Informations Systems in Organizations." Journal of
Applied Systems Analysis 16: 75-88.
Mathiassen, L. and P. A. Nielsen (1990). Surfacing
Organizational Competence. Soft Systems and Hard
Contradictions. In Organizational Competence in
System Development: a Scandinavian Contribution. G.
Bjerknes (ed.) Lund, Studentlitteratur.
Merriam-Webster's collegiate dictionary. 10th ed. 1996:
Merriam-Webster, Springfield.
Van de Ven, A. H. and M. S. Poole (1995). "Explaining
development and change in organizations." Academy
of Management Review 20(3): 510-540.
COMMONALITY VERSUS VARIABILITY - The Contradictory Nature of Enterprise Systems
575
