100% found this document useful (1 vote)

695 views99 pages

Van de Ven - Engaged - Scholarship - A - Guide - For - Organizationa

Engaged Scholarship in a Professional School

Uploaded by

Trang Nguyễn

We take content rights seriously. If you suspect this is your content, claim it here.

Available Formats

Download as PDF, TXT or read online on Scribd

100% found this document useful (1 vote)

695 views99 pages

Van de Ven - Engaged - Scholarship - A - Guide - For - Organizationa

Engaged Scholarship in a Professional School

Uploaded by

Trang Nguyễn

We take content rights seriously. If you suspect this is your content, claim it here.

Available Formats

Download as PDF, TXT or read online on Scribd

You are on page 1/ 99

1 Engaged Scholarship in

a Professional School
Model

Th
sig

eo
De

ry
ch

Bu
ar

ild
se

in
Re

g
Solution Theory

n
io
at
Pr

ul
ob

rm
le

Fo
m

m
So

le
lv

ob
in

Pr
g

Reality

[Academics] appear to have entered a period of non-engagement,

cherishing their autonomy over engagement and retreating into the
ivory tower.
(Patrick Saveau quoted in Cushman 1999: 328)

Scholarship means something more than research, and engagement is

the means for scholarship to Xourish.
(Chapter 1, this volume, p. 9)
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

Understanding how research can advance scientiWc and practical knowledge is

an ongoing challenge for scholars who work in professional schools, such as
business, engineering, social work, medicine, agriculture, education, public
administration, journalism, and law. A central mission of scholars in profes-
sional schools is to conduct research that both advances a scientiWc discipline
and enlightens practice in a professional domain (Simon 1976). Professional
schools typically build their raison d’être on the mission of conducting
research knowledge that advances both science and practice (Simon 1976;
Kondrat 1992; TranWeld and Starkey 1998). But this mission remains an
elusive ideal.
Studies show that practitioners often fail to adopt the Wndings of research
in Welds such as medicine (Denis and Langley 2002; Dopson et al. 2002),
human resources (Anderson et al. 2001; Rynes et al. 2002), social work
(Kondrat 1992) and management (TranWeld et al. 2003; Rousseau 2006).

Van, D. V. A. H. (2007). Engaged scholarship : A guide for organizational and social research. Oxford University Press, Incorporated.
Created from uu on 2024-11-20 05:12:21.
2 ENGAGED SCHOLARSHIP

Many top journals1 have highlighted growing concerns that academic research
has become less useful for solving practical problems and that the gulf between
science and practice in a profession such as management is widening. There
are also growing criticisms that Wndings from academic as well as consulting
studies are not useful to practitioners and do not get implemented (Beer 2001;
Gibbons et al. 1994). Management scholars, for example, are being criticized
for not adequately putting their abstract knowledge into practice (Beyer and
Trice 1982; Lawler et al. 1985; Hodgkinson et al. 2001). Practicing managers, as
well, are criticized for not being aware of relevant research and not doing
enough to put their practice into theory (Weick 2001; Van de Ven 2002). As a
result, organizations are not learning fast enough to keep up with the changing
times.
Academic researchers sometimes respond to these criticisms by claiming
that the purpose of their research is not to make immediate contributions to
practice; instead it is to make fundamental advances to scientiWc knowledge
that may eventually enlighten practice. However, there is evidence that aca-
demic research is also not adequately advancing scientiWc knowledge. One
important indicator of the impact and use of published research by the
scientiWc community is the number of times this research is cited as inform-
ing subsequent scientiWc articles. Based on his citation analysis, Starbuck
(2005) reports that papers published in management journals were cited on
average only .82 times per article per year. Hence, much current academic
research is not contributing in intended ways to either science or practice.

Ways of Addressing the Theory–Practice Gap

This book focuses on the relationship between theory and practice primarily
in organization and management studies, which is my Weld of study. I do not
attempt a comprehensive review of the debate, either in general or with
respect to the management and organization literature. Rather, I review
three ways in which the gap between theory and practice has been framed
(as discussed by Van de Ven and Johnson 2006), and then focus on one
approach that motivates proposing a method of engaged scholarship. A
perusal of literature and discussions with scholars in other professional
domains suggest that the principles below for addressing the gap between
theory and practice apply equally well in many other professional Welds.

1 The relationship between management science and practice has received much attention in
special issues of the Academy of Management Journal (Rynes et al. 2001) and Executive (Bailey
2002), Administrative Science Quarterly (Hinings and Greenwood 2002), British Journal of Manage-
ment (Hodgkinson 2001), and several other more specialized management journals.

Van, D. V. A. H. (2007). Engaged scholarship : A guide for organizational and social research. Oxford University Press, Incorporated.
Created from uu on 2024-11-20 05:12:21.
IN A PROFESSIONAL SCHOOL 3

A KNOWLEDGE TRANSFER PROBLEM

The limited use of research knowledge for science and practice is typically
framed as a knowledge transfer problem. This approach assumes that
practical knowledge (knowledge of how to do things) in many professional
domains derives at least in part from scientiWc knowledge. Hence, the
problem is one of translating and diVusing research knowledge into
practice. I discuss this knowledge transfer problem in Chapter 8. Research
knowledge is not often communicated in a form that facilitates its transfer,
interpretation, and use by an audience as intended. I argue that a deeper
understanding of communicating knowledge across boundaries and a
more engaged relationship between the researcher and his/her audience are
needed if research Wndings are to have an impact in advancing science and
practice.

SCIENCE AND PRACTICE ARE DISTINCT FORMS OF KNOWLEDGE

A second approach to the theory–practice gap views scientiWc knowledge and
practical knowledge as distinct kinds of knowing. Recognition that science
and practice produce distinct forms of knowledge has been long-standing. It
dates back to Aristotle, who in The Nicomachean Ethics (1955), distinguished
between techne (applied technical knowledge of instrumental or mean–ends
rationality), episteme (basic knowledge in the pursuit of theoretical or analyt-
ical questions), and phronesis (practical knowledge of how to act prudently
and correctly in a given immediate and ambiguous social or political situ-
ation). More recently, Polanyi (1962), Habermas (1971), Latour (1986), and
Nonaka (1994) have made further distinctions between explicit epistemic
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

scientiWc knowledge and more tacit practical knowledge, which overlap Aris-
totle’s techne and phronesis distinctions. Each reXects a diVerent ontology
(truth claim) and epistemology (method) for addressing diVerent questions.
To say that the knowledge of science and practice are diVerent is not to say that
they stand in opposition or they substitute for each other; rather, they com-
plement one another.
In her review of the theory–practice gap in social work, Kondrat (1992)
points out that what has been missing from the discussion are empirical
studies of knowledge from practice. What knowledge does the practitioner
of an occupation or profession use, and how does he/she obtain it? So also,
Schon (1987) asks what does the competent practitioner know? and how does
he/she go about knowing ‘in’ practice? Rather than regard practical know-
ledge as a derivative of scientiWc knowledge, these kinds of questions address
the epistemological status of ‘practical knowledge’ as a distinct mode of
knowing in its own right. ‘When this status is granted, the practical takes its

Van, D. V. A. H. (2007). Engaged scholarship : A guide for organizational and social research. Oxford University Press, Incorporated.
Created from uu on 2024-11-20 05:12:21.
4 ENGAGED SCHOLARSHIP

place alongside the scientiWc as constitutive elements of professional know-

ledge’ (Kondrat 1992: 239).
Scholarly work and managerial work diVer in context, process, and purpose.
The context of the practitioner is situated in particular problems encountered
in everyday activities (Hutchins 1983; Lave and Wenger 1994). As such,
managers develop a deep understanding of the problems and tasks that arise
in particular situations and of means–ends activities that comprise their
solutions (Wallace, 1983). Knowledge of practice in a professional domain is
typically customized, connected to experience, and directed to the structure
and dynamics of particular situations (Aram and Salipante 2003: 190). In
contrast, science is committed to building generalizations and theories that
often take the form of formal logical principles or rules involving causal
relationships. ‘ScientiWc knowledge involves the quest for generality in the
form of ‘‘covering’’ laws and principles that describe the fundamental nature of
things. The more context free, the more general and stronger the theory’
(Aram and Salipante 2003: 1900). The purpose of practical knowledge is
knowing how to deal with the speciWc situations encountered in a particular
case. The purpose of scientiWc knowledge is knowing how to see speciWc
situations as instances of a more general case that can be used to explain
how what is done works or can be understood.
We may have misunderstood the relationship between knowledge of sci-
ence and practice, and this has contributed to our limited success in bridging
these two forms of knowledge. Exhortations for academics to put their
theories into practice and for managers to put their practices into theory
may be misdirected because they assume that the relationship between know-
ledge of theory and knowledge of practice entails a literal transfer or transla-
tion of one into the other. Instead, I suggest taking a pluralist view of science
and practice as representing distinct kinds of knowledge that can provide
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

complementary insights for understanding reality.

Each kind of knowledge is developed and sustained by its own professional
community, consisting of people who share a common body of specialized
knowledge or expertise (Van Maanen and Barley 1986). Each community
tends to be self-reinforcing and insular, and limited interactions occur between
them (Zald 1995; Cook et al. 1999). Each form of knowledge is partial—
‘A way of seeing is a way of not seeing’ (Poggi 1965). Strengths of one form
of knowledge tend to be weaknesses of another. Once diVerent perspectives
and kinds of knowledge are recognized as partial, incomplete, and involving
inherent bias with respect to any complex problem, then it is easy to see
the need for a pluralistic approach to knowledge co-production among
scholars and practitioners. This leads to a third view of the theory–practice
gap—namely, a knowledge production problem.

A KNOWLEDGE PRODUCTION PROBLEM

There is a growing recognition that the gap between theory and practice may
be a knowledge production problem. In part this recognition is stimulated by
critical assessments of the status and professional relevance of practice-based
social science (Simon 1976; Whitley 1984, 2000; Starkey and Madan, 2001;
Hinings and Greenwood 2002). Gibbons et al. (1994) and HuV (2000), among
others, question the status quo mode of research typically practiced in
business and professional schools.
This status quo approach to social research has many variations, but it
tends to reXect an unengaged process of inquiry. Researchers typically go it
alone to study a research question without communicating with or being
informed by other stakeholders (scholars from diVerent disciplines, practi-
tioners with diVerent functional experiences, and other potential users and
sponsors) who can make important contributions to understanding the
problem domain being investigated. This status quo form of unengaged
research is evident in the following characteristics of a research report: (1) a
research problem or question is posed but little or no evidence is presented
that grounds the nature and prevalence of the problem, its boundary con-
ditions, and why it merits investigation; (2) a single theoretical model is
proposed with little consideration given of plausible alternative models for
addressing the research problem or question; (3) the research design relies on
statistically analyzing questionnaire or secondary data Wles (such as PIMs,
patent data, Compustat, or census Wles) without the researcher talking to any
informants or respondents in the Weld; and (4) results are presented on the
statistical signiWcance of relationships with little or no discussion of their
practical signiWcance and implications. Because such research is not grounded
in ‘reality,’ does not entertain alternative models for representing reality, nor is
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

it informed by key stakeholders, it often results in making trivial advance-

ments to science, and contributes to widening the gap between theory and
practice. Anderson et al. (2001) characterize this kind of unengaged scholar-
ship as ‘puerile science’ that is often low in both relevance and rigor. As a
consequence, it joins the large proportion of research papers that are not used
to advance either science or practice.
Many suggestions have been made for revising and improving this status
quo approach to social science research. Many of these suggestions are
institutional in nature, such as modifying academic tenure and reward sys-
tems, funding criteria for competitive research grants, editorial policies and
review procedures of academic journals, and creating additional outlets for
transmitting academic Wndings to practitioners (Lawler et al. 1985; Dunnette
1990). Structural reforms such as these are important institutional arrange-
ments that enable and constrain research. But discussions of structural

Van, D. V. A. H. (2007). Engaged scholarship : A guide for organizational and social research. Oxford University Press, Incorporated.
Created from uu on 2024-11-20 05:12:21.
6 ENGAGED SCHOLARSHIP

reforms like this tend to overlook the choices and actions available to indi-
vidual scholars undertaking research in a professional domain. In this book I
focus on methods and strategies that have more immediate relevance to
individual scholars engaged in the knowledge production process.

Engaged Scholarship
At the level of the individual researcher, Pettigrew formulates the problem this
way:
If the duty of the intellectual in society is to make a diVerence, the management
research community has a long way to go to realize its potential. . . . The action steps to
resolve the old dichotomy of theory and practice were often portrayed with the
minimalist request for management researchers to engage with practitioners through
more accessible dissemination. But dissemination is too late if the wrong questions
have been asked. (Pettigrew 2001: S61, S67)
He goes on to say that a deeper form of research that engages both
academics and practitioners is needed to produce knowledge that meets the
dual hurdles of relevance and rigor for theory as well as practice in a given
domain (see also Hodgkinson et al. 2001).
Pettigrew sketches a vision that is not limited to business school research
but reXects a much larger movement of engaged scholarship for transforming
higher education (Zlotkowski 1997–2000). To Ernest Boyer (1990), a leading
proponent of this movement, engaged scholarship consists of a set of reforms
to break down the insular behaviors of academic departments and disciplines
that have emerged over the years. Engaged scholarship implies a fundamental
shift in how scholars deWne their relationships with the communities in which
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

they are located, including faculty and students from various disciplines in the
university and practitioners in relevant professional domains.
It’s about faculty members having a profound respect for those other than themselves,
whether they be practitioners or students. . . . There is a profound emphasis on the
concept of deep respect and, I might even say, humility vis-à-vis other kinds of
knowledge producers. Not because we don’t have an important and distinctive role
to play in knowledge production, but because we don’t have the exclusive right to
such production. As we begin to engage in partnerships with both our students and
outside communities of practice on the basis of such deep respect, we allow ourselves
to become real-world problem solvers in a way that is otherwise not possible. Indeed,
I would suggest that unless we learn to develop deeper respect for our nonfaculty
colleagues, we run the risk of becoming ‘academic ventriloquists’—speaking for our
students, speaking for the communities we allegedly serve—but not really listening to
them or making them our peers in addressing the vital issues that concern all of us.
(Edward Zlotkowski quoted in Kenworthy-U’ren 2005: 360)

Engagement is a relationship that involves negotiation and collaboration

between researchers and practitioners in a learning community; such a
community jointly produces knowledge that can both advance the scientiWc
enterprise and enlighten a community of practitioners. Instead of viewing
organizations and clients as data collection sites and funding sources, an
engaged scholar views them as a learning workplace (idea factory) where
practitioners and scholars co-produce knowledge on important questions and
issues by testing alternative ideas and diVerent views of a common problem.
‘Abundant evidence shows that both the civic and academic health of any
culture is vitally enriched as scholars and practitioners speak and listen
carefully to each other’ (Boyer 1996: 15).
Applying these notions of engaged scholarship to the full range of activities
of faculty in universities, Boyer (1990) discussed the scholarship of discovery,
teaching, application, and integration. These four dimensions interact to form
a rich and uniWed deWnition of scholarship. Subsequently, Boyer (1996)
further expanded his deWnition to include the scholarship of engagement,
which emphasizes how academics relate their teaching, discovery, integration,
and application activities with people and places outside the campus and
ultimately direct the work of the academy ‘toward larger, more humane ends’
(Boyer 1996: 20).
For many American public universities, engaged scholarship represents a
call to return to their charter mandate of a Land Grant University, as
established by the Morrill Land Grant Act of 1862 (Schuh 1984). Three
ideas of engagement were central to the founding ideals of a Land Grant
University. First, it would provide upper-level education for the masses—
a direct response at the time to the elitism and limited relevance of the private
universities in the country. Second, the Land Grant University would generate
new knowledge by addressing questions and problems of society. Although
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

agriculture was dominant at the time, every area of activity was to be a

legitimate subject of intellectual inquiry. Third, the Land Grant University
would have a strong outreach mission, which is to provide intellectual
leadership by applying the tools of science and technology to address the
problems of society. These three ideas gave rise to the familiar tripartite
mission of teaching, research, and service. As this brief history indicates,
engaged scholarship represents a re-enactment of the founding values and
roles of universities as institutions engaged in society and of individual
scholars engaging students and community practitioners in their teaching,
research, and service.
The engaged scholarship movement has proliferated into numerous uni-
versity-based initiatives of community outreach, service-learning, clinical
teaching, extension services, social emancipation causes, and community-
based participatory research. As evident in a Google.com listing of 36,000
entries on ‘engaged scholarship,’ in November 2006, these initiatives are

Van, D. V. A. H. (2007). Engaged scholarship : A guide for organizational and social research. Oxford University Press, Incorporated.
Created from uu on 2024-11-20 05:12:21.
8 ENGAGED SCHOLARSHIP

highly diverse and diVused. Service learning is perhaps the most widely
diVused form of engaged teaching, due largely to eVorts by national organ-
izations and federal grants (such as Campus Compact, American Association
for Higher Education, the National Community Service Trust Act of 1993, and
others). Service learning is a credit-bearing educational experience in which
students become involved in an organized service activity that augments
understanding of topics covered in a university classroom with experiences
as volunteers in local sites serving community needs, such as philanthropic
agencies, primary and secondary schools, churches, old-age homes, half-way
houses, etc. (Bringle and Harcher 1996; DiPadova-Stocks 2005). Professional
schools tend to take less of a missionary and more of a training view of service
learning through a wide variety of university–industry internships, mentor-
ships, clinical research, and Weld study projects. An experiment conducted by
Markus et al. (1993) found that students in service learning courses had more
positive course evaluations, more positive beliefs and values toward service
and community, and higher academic achievement. Bringle and Harcher
(1996) review other research indicating that service learning has a positive
impact on personal, attitudinal, moral, social, and cognitive outcomes for
students.
Despite this diVusion and evidence, one of the major barriers to sustained
faculty involvement in engaged scholarship is the risk associated with trying
to achieve promotion and tenure. A number of national commissions and
professions have begun to address these institutional barriers. For example, the
2006 report of the Commission on Community-Engaged Scholarship in the
Health Professions focuses on recommendations for recruiting, retaining, and
promoting community-engaged faculty members in health professional
schools. In addition, the US Department of Education and the W. K. Kellogg
Foundation co-sponsored the development of a Community-Engaged Schol-
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

arship Toolkit that guides faculty in preparing their career statements and
records for faculty promotion and tenure in healthcare and other professional
schools (Calleson et al. 2004).
This book applies the principles of engaged scholarship to social research,
or what Boyer calls the scholarship of discovery.
No tenets in the academy are held in higher regard than the commitment to know-
ledge for its own sake, to freedom of inquiry and to following, in a disciplined fashion,
an investigation wherever it may lead. The scholarship of discovery, at its best, con-
tributes not only to the stock of human knowledge but also to the intellectual climate
of a college or university. Not just the outcomes, but the process, and especially the
passion, give meaning to the eVort. The advancement of knowledge can generate an
almost palpable excitement in the life of an educational institution. As William
Bowen, former president of Princeton University, said, scholarly research ‘reXects
our pressing, irrepressible need as human beings to confront the unknown and to seek
understanding for its own sake. It is tied inextricably to the freedom to think freshly,

to see propositions of every kind in ever-changing light. And it celebrates the special
exhilaration that comes from a new idea. (Boyer 1990: 17)
In addition to conveying this passion for knowledge discovery, the term
engaged scholarship reXects an important identity. Scholarship means some-
thing more than research, and engagement is the means for scholarship to
Xourish. Boyer resurrected the honorable term scholarship, gave it a broader
and more capacious meaning that conveyed legitimacy to the full scope of
academic work. ‘Surely, scholarship means engaging in original research. But
the work of the scholar also means stepping back from one’s investigation,
looking for connections, building bridges between theory and practice, and
communicating one’s knowledge eVectively’ (Boyer 1990: 16).
Pettigrew (2005: 973) asks the question, ‘How many of us see ourselves as
intellectuals, scholars, and/or researchers?’ He states:
An intellectual is a person having a well-developed intellect and a taste for advanced
knowledge, while a scholar is a person with great learning in a particular subject. And
a researcher is a person who engages in careful study and investigation in order to
discover new facts or information. Even from these rather limited deWnitions, the
narrowness of the researcher identity and role becomes very evident. . . . Scholarship to
me implies not just great breadth of learning and appreciation, but also the duty to
make these available in dedicated learning, teaching, and professing. An intellectual
would be capable of the appreciative system of a scholar but would be harnessing that
competence to engage way beyond the boundaries of academic and into the wider
reaches of society. I wonder how many of us have made explicit choices of engagement
with one or other of the three identities/roles? (Pettigrew 2005: 973)
This poses the important question of how an engaged scholar might
formulate a research study of a complex problem in the world that advances
both theory and practice? To do this a mode of inquiry is needed that converts
the information obtained by scholars in interaction with practitioners (and
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

other stakeholders) into actions that address problems of what to do in a

given professional domain. Many research questions and problems exceed the
capabilities of individual researchers to study them alone. A methodology is
needed that signiWcantly expands researchers’ capabilities to address such
complex problems and questions.
I propose a method of engaged scholarship for expanding the capabilities of
scholars to study complex problems and create the kind of knowledge that
advances both science and practice. Engaged scholarship is deWned as a
participative form of research for obtaining the diVerent perspectives of key
stakeholders (researchers, users, clients, sponsors, and practitioners) in study-
ing complex problems. By involving others and leveraging their diVerent
kinds of knowledge, engaged scholarship can produce knowledge that is
more penetrating and insightful than when scholars or practitioners work
on the problems alone.

Van, D. V. A. H. (2007). Engaged scholarship : A guide for organizational and social research. Oxford University Press, Incorporated.
Created from uu on 2024-11-20 05:12:21.
10 ENGAGED SCHOLARSHIP

ENGAGED SCHOLARSHIP RESEARCH MODEL

Past arguments for collaborative research have tended to be one-sided and
focus on the relevance of academic research for practice. I focus more atten-
tion in this book on the question of how scholarship that is engaged with
(rather than for) practice can advance basic scientiWc knowledge? Engaged
scholarship emphasizes that research is not a solitary exercise; instead it is a
collective achievement. Engagement means that scholars step outside of
themselves to obtain and be informed by the interpretations of others in
performing each step of the research process: problem formulation, theory
building, research design, and problem solving.
Using a diamond model as illustrated in Figure 1.1, I propose that scholars
can signiWcantly increase the likelihood of advancing fundamental knowledge
of a complex phenomenon by engaging others whose perspectives are relevant
in each of these study activities:
. Problem formulation—situate, ground, diagnose, and infer the research
problem by determining who, what, where, when, why, and how the
problem exists up close and from afar. As discussed in Chapter 3, answer-
ing these journalist’s questions requires meeting and talking with people
who experience and know the problem, as well as reviewing the literature
on the prevalence and boundary conditions of the problem.
. Theory building—create, elaborate, and justify a theory by abductive,
deductive, and inductive reasoning (as discussed in Chapter 4). Developing

Study Context: Research problem, purpose, perspective

Research Design Theory Building

Create, elaborate & justify a theory
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

Develop variance or process

model to study theory Model by abduction, deduction, & induction

Engage methods experts & people Engage knowledge experts in

providing access & information relevant disciplines & functions
Th
n
ig

eo
es

Criterion—Validity
ry
D

Criterion—Truth (Versimilitude)
Bu
ch

ild
ar
se

in
g
Re

Solution Theory
n
io

Iterate
Pr

at
ob

ul
rm

& Fit
le
m

Fo
So

m
lv

Problem Solving Problem Formulation

le
in

ob
g

Communicate, interpret & negotiate Situate, ground, diagnose, & infer

findings with intended audience. the problem up close and from afar

Engage intended audience Reality Engage those who experience

to interpret meanings & uses & know the problem

Criterion—Impact Criterion—Relevance

Figure 1.1. Engaged scholarship diamond model

this theory and its plausible alternatives requires conversations with

knowledge experts from the relevant disciplines and functions that have
addressed the problem, as well as a review of relevant literature.
. Research design—develop a variance or process model for empirically
examining the alternative theories. As noted in Chapters 5–7, doing this
well typically requires getting advice from technical experts in research
methodology and the people who can provide access to data, and of
course, the respondents or informants of information.
. Problem solving—communicate, interpret, and apply the empirical Wnd-
ings on which alternative models better answer the research question about
the problem. Chapter 8 argues that increases in the diVerence, dependence,
and novelty of knowledge between people at a boundary require more
engaged forms of communication, starting with written reports and pre-
sentations for knowledge transfer, then conversations to interpret diVerent
meanings of the report, and then pragmatic and political negotiations
to reconcile conXicting interests.
These activities can be performed in any sequence. I discuss these research
activities in a problem solving sequence beginning with problem formulation,
then searching for theories relevant to the problem, testing them, and applying
the Wndings. There are many other possible starting points and sequences. For
example, some scholars may start with a theory and then search for a problematic
situation that may be appropriate for applying and evaluating the theory. Other
scholars may be methodologically inclined, and interested in Wnding problems
and developing theories with their methodological tools (as was the case in early
developments of social network analysis). Still others may begin with a solution
or program that requires evaluation research in order to determine the particular
kinds of problems and context for which it may be appropriate.
These diVerent starting motivations and orientations quickly meld together
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

in the course of a study because the four activities are highly interdepend-
ent and are seldom completed in one pass. Multiple iterations and revisions
of these research activities are often needed throughout the duration of a
study. In the process, many sub-problems emerge in performing each research
activity, and all remain simultaneously active and need to be addressed as
an interdependent set. It is only when the process is complete that a fairly
coherent pattern emerges as reXected in Figure 1.1.
Maintaining balance in performing these tasks repetitively is important.
Given Wnite resources for conducting a study, I recommend that scholars
allocate their time and eVorts about equally to problem exploration, theory
building, research design and conduct, and problem solving activities. Spend-
ing too much time or eVort on only one or two research activities often results
in unbalanced or lop-sided results where some activities are ‘over-engineered’
while others are incomplete.

Van, D. V. A. H. (2007). Engaged scholarship : A guide for organizational and social research. Oxford University Press, Incorporated.
Created from uu on 2024-11-20 05:12:21.
12 ENGAGED SCHOLARSHIP

This suggestion of paying equal attention to all four research activities is

not evident in many research methodology texts in the social sciences. They
tend to focus on research design, and pay relatively little attention to the
processes of problem formulation, theory building, and problem solving. In
addition, while these texts provide good technical treatments of research
designs and data analysis, they largely ignore social processes of engaging
stakeholders in problem formulation, theory building, research design, and
problem solving (as illustrated in Figure 1.1). Social research is an intensely
social process. Throughout the book I emphasize that all four research
activities are equally important in conducting a study, and that each activity
entails a diVerent set of tasks that can be accomplished better by engaging
relevant stakeholders rather than going it alone.
The essential steps in performing the four activities of the diamond model
are illustrated in Figure 1.1. They can be evaluated in terms of Wve criteria:
relevance, validity, truth, impact, and coherence.2 The problem should be
grounded in a reality that is relevant to an intended research audience in
the scholarly and professional communities. The theoretical model should be
expressed clearly, it should consist of a logically valid argument. The design
and conduct of the research should apply the standards and methods that a
scientiWc community believes will produce a truthful solution. The Wndings of
the research should have an impact in advancing science and enlightening
practice in a profession. In addition to relevance, validity, truth, and impact, a
Wfth criterion—coherence—is equally important for evaluating the engaged
scholarship process.
In the PhD seminar I teach on engaged scholarship, the major assignment
is for students to develop a good research proposal.3 A good research
proposal is deWned as one that adequately describes each of the research
activities in terms of the criteria presented in Table 1.1. Students submit
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

diVerent parts of the proposal every few weeks of the semester. As an

instructor, I provide students feedback on their in-progress proposals, and
they revise their research proposals several times until it is judged to be
acceptable. Thus, through several iterations of revising-and-extending their
proposals, students develop a research proposal, which they submit for
funding and implement either as a research project or as an initial draft of
their dissertation proposal.

2 Scholars from diVerent philosophical persuasions often associate diVerent meanings with these
criteria. My interpretations of these criteria should become clear in subsequent chapters devoted to
each of the research activities in the engaged scholarship model.
3 The most recent version of this course can be accessed by following the link on MGMT 8101,
Theory Building and Research Design from my faculty web page at the Carlson School of Management,
University of Minnesota (available at: (http://umn.edu/avandeve). This course web page provides a
wealth of additional information, resources, and links that supplement the topics and issues discussed
in this book.

Table 1.1. Criteria for evaluating a research proposal

1. Statement of the research problem: _____

. is situated in terms of perspective, focus, level, and scope;
. problem symptoms or elements are clearly defined & grounded in reality;
. a diagnosis is made that analyzes patterns or relationships among elements;
. based on the diagnosis, an inference (a claim with reasons) is made for
the problem.
2. The research question: _____
. is stated in analytical and researchable terms;
. permits more than one plausible answer.
3. The research proposition (theory): _____
. clearly states an expected relationship among concepts or events;
. is supported with an argument (i.e., claim, reasons, evidence,
assumptions, & reservations);
. directly addresses the research question and problem;
. is compared with a plausible alternative theory or the status quo answer;
. travels across levels of abstraction.
4. The research design clearly spells out: _____
. theoretical unit of analysis and unit of observation;
. case/survey/experimental design for variance or process theory;
. sample or replication logic and sample selection;
. definitions and measurement procedures for variables or events;
. threats to internal, statistical, external, & construct validities.
5. Research implementation and problem solving for theory and practice: _____
. the contributions/implications of the research for science and practice
are clearly stated;
. methods for communicating and sharing findings with target audiences/users
are discussed;
. statement of how research findings will be used/applied is prudent;
. relevant stakeholders are engaged in each of the above steps.
Comments: Total Score: _____

Note: Please evaluate this report by using this five-point scale:

1 ¼ not addressed or evident in the report
2 ¼ attempt made but some errors occurred in the analysis/answer.
3 ¼ attempt made but the result needs more work, elaboration, or refinement.
4 ¼ attempt made with good result; issue accomplished; no further work needed.
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

5 ¼ attempt made with excellent result; issue accomplished with distinction.

Engaged scholarship can be practiced in many diVerent ways for addressing

a variety of basic and applied research questions. For example, researchers
might engage stakeholders in a study in order to: (1) obtain their perspectives
and advice on a basic research question; (2) collaborate and co-produce
knowledge; (3) design and evaluate a policy or program; or (4) intervene
and implement a change to solve a client’s problem. As these alternatives
suggest, principles of engaged scholarship apply to many forms of basic or
applied social research. They are discussed later in this chapter.
The four research activities in the engaged scholarship model illustrated
in Figure 1.1 serve as the organizing framework of this book. Following an
overview of the philosophy of science underlying this model of engaged
scholarship, I discuss each of the four research activities in the engaged

Van, D. V. A. H. (2007). Engaged scholarship : A guide for organizational and social research. Oxford University Press, Incorporated.
Created from uu on 2024-11-20 05:12:21.
14 ENGAGED SCHOLARSHIP

scholarship model. I also indicate how subsequent chapters treat detailed

steps and procedures entailed in each research activity.

PHILOSOPHY OF SCIENCE UNDERLYING ENGAGED SCHOLARSHIP

Underlying any form of research is a philosophy of science that informs a
scholar’s approach to the nature of the phenomenon examined (ontology)
and methods for understanding it (epistemology). Philosophers have
debated these issues endlessly, and constructed a variety of philosophies for
conducting research. Practitioners of science, in turn, are inXuencing how
these philosophies are developed and expressed in their research. Chapter 2
attempts to provide a synthesis of this reciprocal relationship between
the philosophy and practice of science with a historical review of four
philosophies of science—positivism, relativism, pragmatism, and realism. It
provides a discussion of how key ideas from each philosophy inform engaged
scholarship, and how the practice of engaged scholarship might inXuence
these philosophies of science.
Since the demise of the received view of positivism and logical empiricism
in the philosophy of science, it is now widely recognized that scientiWc
knowledge cannot be known to be true in an absolute sense (Suppe 1977:
649). Rather, from a critical realist perspective that I adopt, there is a real
world out there, but our attempts to understand it are severely limited and
can only be approximated. This perspective argues that all facts, observations,
and data are theory-laden and embedded in language. Moreover, most phe-
nomena in the social world are too rich to be understood adequately by any
single person or perspective. Consequently, any given theoretical model is a
partial representation of a complex phenomenon that reXects the perspective
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

of the model builder. No form of inquiry is value-free and impartial; instead

each model and perspective is value-full. This requires scholars to be far
more reXexive and transparent about their roles, interests, and perspectives
when conducting a study than they have in the past. For example, instead
of assuming an authoritative and objective ‘God’s Eye view’ of social phe-
nomena, I follow Henrickx (1999) in proposing that engaged scholars adopt a
participant frame of reference to learn about and understand a subject
through discourse with other stakeholders.
Critical realism views science as a process of constructing models that
represent or map intended aspects of the world, and comparing them with
rival plausible alternative models (Rescher 2000). For example, Giere (1999:
77) states,
Imagine the universe as having a deWnite structure, but exceedingly complex, so
complex that no models humans can devise could ever capture more than limited

aspects of the total complexity. Nevertheless, some ways of constructing models of the
world do provide resources for capturing some aspects of the world more or less well
than others.
Research knowledge of a complex phenomenon advances by comparing
the relative contributions of diVerent models. Azevedo (1997) discusses how
the coordination of multiple models and perspectives may reveal the robust
features of reality by identifying those features that appear invariant (or
convergent) across at least two (and preferably more) independent theories.
From her perspective, a pluralist approach of comparing multiple plausible
models of reality is essential for developing reliable scientiWc knowledge.
But the engagement of diVerent stakeholders in a study often produces
inconsistent and contradictory perspectives of a problem domain being
examined. Pluralistic perspectives should not be dismissed as noise, error,
or outliers—as they are typically treated in a triangulation research strategy.
Chapter 9 discusses how these diVerent outcomes require an expansion of
traditional explanations of triangulation that focus on convergent central
tendencies to include explanations based on inconsistent Wndings through
arbitrage and contradictory Wndings with methods for reasoning through
paradoxical Wndings.
It is often easier to construct meaningful explanations in cases where the
evidence is convergent. For example, Azevedo (1997) advocates the use of
multiple models for mapping a problem being investigated, and argues that
knowledge that is reliable is invariant (or converges) across these models.
Convergent explanations rely on similarities, consensus, and central tendencies
in explaining a problem or issue under investigation. Convergent explanations
tend to treat diVerences and inconsistencies as bias, errors, outliers, or noise.
More diYcult (but often more insightful) explanations emerge when
diVerent data sources yield inconsistent or contradictory information about
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

a phenomenon. Arbitrage provides a strategy for developing holistic, integra-

tive explanations based on diVerent accounts of the same phenomenon.
Friedman (2000: 24) points out that in academe and elsewhere, ‘there is a
deeply ingrained tendency to think in terms of highly segmented, narrow
areas of expertise, which ignores the fact that the real world is not divided up
into such neat little bits.’ He argues that the way to see, understand, and
explain complex problems in the world is to systematically connect the
diVerent dots, bits, and pieces of information through arbitrage—‘assigning
diVerent weights to diVerent perspectives at diVerent times in diVerent
situations, but always understanding that it is the interaction of all of them
together that is really the deWning feature of the [system]’ (Friedman 2000:
23–4). Arbitrage is a strategy of explaining diVerences by seeing the interde-
pendencies and webs of entanglements between diVerent and divergent dimen-
sions of a problem, its boundaries, and context.

Van, D. V. A. H. (2007). Engaged scholarship : A guide for organizational and social research. Oxford University Press, Incorporated.
Created from uu on 2024-11-20 05:12:21.
16 ENGAGED SCHOLARSHIP

Finally, contradictory information from diVerent sources may represent

instances of conXicting values and interests among pluralistic stakeholders
about the problem or issue being examined. Explanations of a problem
domain should obviously reXect these contradictions when observed. In
Chapters 8 and 9, I discuss four general methods for reasoning through
paradoxes by either: balancing between opposites, shifting levels of analysis,
alternating positions over time, and introducing new concepts that dissolve
the paradox. Inconsistent and contradictory Wndings are important, for they
represent anomalies that trigger theory creation.
Campbell’s (1988: 389) evolutionary perspective of science provides a
possible avenue for addressing the simultaneous need to establish valid and
reliable representations of a problem domain being examined. He argues that
the models that better Wt the problems they were intended to solve are
selected, and the gradual winnowing down of plausible rival models or
hypotheses by the scholarly community produces an evolutionary conception
of the growth of scientiWc knowledge. This evolutionary perspective is based
on a pragmatic philosophy of science. Among the plausible alternative models
competing to explain a given phenomenon, the model that wins out at a
particular moment in time is the one that is judged to best represent the
phenomenon. Fortunately, only a Wnite set of three to Wve plausible models
tend to compete for selection at a given time, as indicated by Collins’s (1998)
historical review of competing models for explaining a phenomenon.
Explanations based on arbitrage and paradoxical reasoning represent dia-
lectical methods of inquiry where understanding and synthesis of a complex
problem evolve from the confrontation of divergent thesis and antithesis.
Dialectical reasoning is not a strategy for addressing narrow technical prob-
lems where one looks for expert judgments to converge on a correct answer.
Instead, it is a strategy for triangulating on complex real-world problems by
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

involving individuals whose perspectives are far from the average (MitroV and
Linstone 1993: 69). In a complex world, diVerent perspectives make diVerent
sorts of information accessible. By exploiting multiple perspectives, the robust
features of reality become salient and can be distinguished from those features
that are merely a function of one particular viewpoint or conceptual model.
Thus, engaged scholarship is essentially a pluralistic methodology. Azevedo
(2002) points out that communication across perspectives is a precondition
for establishing robust alternative models of a problem. She adds,
Individual theories are not considered true or false. Rather their validity is a function
not only of how well they model the aspect of the world in question but of how
connected they are, in terms of consistency and coherence, with the greater body of
scientiWc knowledge. These connections can be established a number of ways . . . but
communication across perspectives and willingness to work toward establishing
coherence is a precondition. (Azevedo 2002: 730)

Pluralism consists of not only multiple perspectives, but also a degree of

openness and equality among them for addressing complex social phenom-
ena. Participants often experience conXict and interpersonal tensions associ-
ated with juxtaposing people with diVerent views and approaches to a
problem. Managing conXict constructively is not only important but lies at
the heart of engaged scholarship. Attempting to avoid tensions between
scholars and practitioners, as we have in the past, is a mistake, for it blinds
us to very real opportunities that are possible from exploiting the diVerences
underlying these tensions in understanding complex phenomena.

PROBLEM FORMULATION
Problem formulation consists of situating, grounding, and diagnosing a
research problem or issue in reality. Of course, diVerent observers will see
diVerent ‘realities.’ In Chapter 3 I take a critical realist perspective and argue
that there is a real world out there, but our representation and understanding
of it is a social construction; reality does not exist independently of the
observer’s schemata or conceptual frame of reference (Weick 1989). As a
consequence, the formulation of a research problem involves a complex
sensemaking process of applying various conceptual templates or theories
to determine what to look for in the real world and how to unscramble
empirical materials into a recognizable and meaningful research problem.
Problem formulation plays a crucial role in conducting research and
potentially aVects succeeding phases, including theory building, research
design and conduct, and conclusions. Yet problem formulation is often
rushed or taken for granted. People tend to be solution-minded, rather
than problem-minded. When problem formulation is rushed or taken for
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

granted, in all likelihood important dimensions of the problem go undetected

and opportunities are missed (Volkema 1995).
Social science today suVers from elaborating theories that are often based
on insuYcient grounding in concrete particulars. It also suVers from a lack of
relevance as perceived by the intended audiences or users of the research (Beer
2001; Rynes et al. 2001). As a consequence, theories tend to be grounded in
myths and superstitions. Those who generalize from experience can answer
the questions, ‘For example? From whose point of view? What is the point of
view?’ Lacking answers to these questions often leads to unfounded general-
izations. In crime investigation, establishing the case is mandatory for pur-
suing it. Merton (1987: 21) cautioned that an important Wrst element in the
practice of science is ‘establishing the phenomenon.’ Evidence and argument
should clearly indicate that it is enough of a regularity to require and allow
explanation. In this way ‘pseudofacts that induce pseudoproblems are
avoided’ (Hernes 1989: 125).

Van, D. V. A. H. (2007). Engaged scholarship : A guide for organizational and social research. Oxford University Press, Incorporated.
Created from uu on 2024-11-20 05:12:21.
18 ENGAGED SCHOLARSHIP

Grounding the problem or phenomenon in reality is a crucial step in any

research study. You might ask, what kind of research problems require
engagement of others? I argue that the more complex the problem or the
bigger the research question, the greater the level of engagement is required of
researchers from diVerent disciplines and practitioners with diVerent func-
tional experiences. Engagement of others is necessary because most real-
world problems are too complex to be captured by any one investigator or
perspective. Caswill and Shove (2000b: 222) point out that there are many
signiWcant questions and problems whose formulation and theoretical devel-
opment depend on engagement and close interaction between scholars and
practitioners.
Big questions have no easy answers, and they seldom provide an immediate
pay-oV to practitioners or academics (Pettigrew 2001). By deWnition, big
questions often do not have clear solutions until after the research has been
conducted and policy questions have been addressed. Big questions also
require a process of arbitrage in which researchers and practitioners engage
each other to co-produce solutions whose demands exceed the capabilities of
either researchers or practitioners (Hodgkinson et al. 2001). Thus, at the time
of designing a research project prospective solutions to research questions are
secondary in comparison with the importance of the research question that is
being addressed. A good indicator of a big question is its self-evident cap-
ability to motivate the attention and enthusiasm of scholars and practitioners
alike. Indeed, as Caswill and Shove (2000b: 221) state, practitioners are ‘often
more attracted by new ideas and concepts than by empirical materials.’
Critics have argued that practitioner involvement in formulating research
questions may steer the questions in narrow, short-term, or particularistic
directions (Brief and Dukerich 1991; Grey 2001; KilduV and Kelemen 2001).
Ironically, this argument seems to assume that academics know better how to
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

formulate researchable questions than practitioners, but when interacting with

practitioners, researchers may behave as ‘servants of power’ (Brief 2000) by
cowering to the interests of powerful stakeholders. Like Anderson et al. (2001),
I view an engaged scholar as being more humble and also standing in a more
egalitarian relationship with practitioners and other stakeholders when trying
to understand an important question or phenomenon that requires research.
Big research questions tend to reside in a buzzing, blooming, confusing world.
Learning about the nature of the question or phenomenon in such ambiguous
settings is facilitated by obtaining the divergent perspectives of numerous
stakeholders. Heedful accommodation and integration of diverse viewpoints
yields a richer gestalt of the question being investigated than the sensemaking
of a single stakeholder (Morgan 1983; Weick 1995).
Caswill and Shove (2000a, 2002b) critique the assumption that the advance-
ment of theory requires academic detachment, and that collaborative research
merely implements and exploits, but does not advance, social theory.

The trouble is that arguments about independence and interaction, and about theory
and application are readily and sometimes deliberately confused. In everyday discus-
sion, it is sometimes asserted, and often implied, that interaction outside the academy
is so demanding of time and mental energy that it leaves no room for creative thought.
In addition, when distance is equated with purity, and when authority and expertise is
exclusively associated with analytic abstraction, it is easy (but wrong) to leap to the
conclusion that calls for interaction threaten academic inquiry. (Caswill and Shove
2000b: 221)
Indeed, the belief that interactions between people with diVerent views and
approaches advances academic (and practical) inquiry lies at the heart of
engaged scholarship.

THEORY BUILDING
Theory building involves the creation, elaboration, and justiWcation of a body
of knowledge that is relevant to the research problem. A theory is the mental
image or conceptual framework that is brought to bear on the research
problem. Theories exist at various levels of abstraction for representing
knowledge. A formal classiWcation of the structure of knowledge is the
Dewey indexing system found in all libraries. It classiWes all knowledge into
ten categories with ten subcategories, another ten sub-subcategories, and so
on. This classiWcation system packages knowledge by disciplines, paradigms,
schools of thought, and theories on various subjects. You may not like such a
formal hierarchical structure of knowledge, but you need to know it if you
hope to Wnd a book in the library.
This nested hierarchical structure not only indexes bodies of knowledge, it
also structures our views of reality by specifying what problems and what
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

aspects of problems are relevant and not relevant. Selecting and building a
theory is perhaps the most strategic choice that is made in conducting a study.
It signiWcantly inXuences the research questions to ask, what concepts and
events to look for, and what kind of propositions or predictions might be
considered in addressing these questions. Because a theory is so inXuential in
directing (or tunneling) a research study, Chapter 4 examines the activities
and patterns of reasoning involved in theory building, and the importance of
engaging others in the process of theorizing.
DiVerent and opposing views are often expressed about theory building.
They range from those who emphasize theory creation and argue that trivial
theories are often produced by hemmed-in methodological strictures that
favor validation rather than imagination (Weick 1989; Mintzberg 2005), to
those who focus on elaborating and justifying a theory by calling for clear
deWnitions, internal logical consistency, and veriWability (Bacharach 1989;
Peli and Masuch 1997; Wacker 2004). In part these writers are right in

Van, D. V. A. H. (2007). Engaged scholarship : A guide for organizational and social research. Oxford University Press, Incorporated.
Created from uu on 2024-11-20 05:12:21.
20 ENGAGED SCHOLARSHIP

describing one theory building activity, but wrong in ignoring other activities
involved in theory building. Many of these oppositions dissolve when theory
building is viewed not as a single activity, but as entailing at least three
activities—creating, constructing, and justifying a theory.
Chapter 4 discusses how these three activities entail diVerent patterns
of reasoning: (1) the creative germ of a promising (but often half-baked)
conjecture is typically created through a process of abductive reasoning to
resolve an anomaly observed in the world; (2) then a theory is constructed
to elaborate the conjecture by using basic principles of logical deductive reason-
ing to deWne terms, specify relationships, and conditions when they apply; and
(3) if the merits of the theory are to be convincing to others, the theory is
justiWed by crafting persuasive arguments and using inductive reasoning to
empirically evaluate a model of the theory in comparison with rival plausible
alternative models. In other words, theory creation involves an abductive
process of ‘disciplined imagination’ (Weick 1989), theory construction entails
logical deductive reasoning, and theory justiWcation requires inductive reason-
ing and argumentation. Hence, theorizing entails diVerent patterns of reasoning,
and much can be learned about the scientiWc enterprise by understanding
the complementary relations among these diVerent patterns of reasoning.
A key recommendation discussed in Chapter 4 is to develop alternative
theories and methods to study a problem. Multiple frames of reference are
needed to understand complex reality. As mentioned before, engaged schol-
arship is a pluralistic methodology. Any given theory is an incomplete abstrac-
tion that cannot describe all aspects of a phenomenon. Theories are fallible
human constructions that model a partial aspect of reality from a particular
point of view and with particular interests in mind. Comparing and contrast-
ing plausible alternative models that reXect diVerent perspectives are essential
for discriminating between error, noise, and diVerent dimensions of a com-
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

plex problem being investigated. Allison (1971) provides a good example of

triangulating on the Cuban Missile Crisis with three models—a rational actor,
organization behavior, and a political model. Each model is a conceptual lens
that ‘‘leads one to see, emphasize, and worry about diVerent aspects of an
event’’ (Allison 1971: 5). Combined, complementary models provide richer
insights and explanations of a phenomenon that would otherwise remain
neglected.
The choice of models and methods varies, of course, with the particular
problem and purpose of a study. The more complex the problem or question
the greater the need to map this complexity by employing multiple and
divergent models. Triangulation of methods and models increases reliability
and validity. It also maximizes learning among members of an engaged
scholarship team. Presumably diVerent models reXect the unique hunches
and interests of diVerent participants in the research project. Sharing ap-
proaches and Wndings enhance learning among co-investigators. Each strategy

represents a diVerent thought trial to frame and map the subject matter. As
Weick (1989) argues, undertaking multiple independent thought trials facili-
tates good theory building.
The typical strategy in social science research is to use a single theory to
examine a given phenomenon. I argue that you have much greater likelihood
of making important knowledge advances to theory and practice if the study is
designed so that it juxtaposes and compares competing plausible explanations of
the phenomenon being investigated (Kaplan 1964; Stinchcombe 1968a; Single-
ton and Straits 1999; Poole et al. 2000). Stinchcombe (1968a), for example,
advises researchers to develop ‘crucial’ propositions that ‘carve at the joints’ (as
Plato described) of positions by juxtaposing or comparing competing answers.
Examining plausible alternatives promotes a critical research attitude. It also
leverages knowledge diVerences by examining the extent to which evidence for
competing alternative models compares with status quo explanations. Know-
ledge of many topics has advanced beyond the customary practice of rejecting
a null hypothesis when a statistical relationship is diVerent from zero. Such a
Wnding is a cheap triumph when previous research has already shown this
to be the case. More signiWcant knowledge is produced when rival plausible
hypotheses are examined. Such studies are likely to add signiWcant value to
theory and practice. Testing rival plausible hypotheses also provides the insur-
ance of a win–win outcome for investigators—no matter what research results
are obtained, if properly executed it can make an important contribution.

RESEARCH DESIGN
Building plausible theories that address the research question and problem
typically sets the stage for designing operational models to empirically exam-
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

ine key aspects of the theories. Research design activities include developing
speciWc hypotheses and empirical observation procedures (based on the
theoretical model) that predict what data should be obtained if the model
provides a good Wt to the real world. A theory is typically not open to direct
inspection, while a model makes operational some speciWc predictions of a
theory, which can be subjected to empirical inspection. The theory and the
hypothesis are related by reasoning or calculation, while the real world and
the data are related by a physical interaction that involves observation or
experimentation. As Giere states,
it is understood that the model Wts only in some respects and then only to some
speciWed degree of accuracy. . . . If what is going on in the real world, including the
experimental setup, is similar in structure to the model of the world, then the data and
the prediction should agree. That is, the actual data should be described by the
prediction. On the other hand, if the real world and the model are not similar in
the relevant respects, then the data and the prediction may disagree. (Giere 1997: 30)

Van, D. V. A. H. (2007). Engaged scholarship : A guide for organizational and social research. Oxford University Press, Incorporated.
Created from uu on 2024-11-20 05:12:21.
22 ENGAGED SCHOLARSHIP

This process can be generalized when comparing alternative predictions

or hypotheses from plausible alternative models. Empirical evidence can
be obtained on alternative predictions and compared to determine which
empirically-based prediction oVers the better or stronger explanation. When
data evaluating the hypotheses from one model oVer worse explanations
than hypotheses from other models, then presumably the former model is
abandoned in favor of the latter models.
A wide variety of research designs can be employed to gather empirical
evidence for evaluating the predictions or hypotheses from diVerent models.
Research methodology texts typically divide and discuss these research
designs in terms of experiments (e.g., Kirk 1995), quasi-experiments (Shadish
et al. 2002), comparative case studies (Yin 2003), and various qualitative
research methods (Denzin and Lincoln 1994; Miles and Huberman 1994).
Before delving into the operational details of these research designs in Chap-
ters 6 and 7, Chapter 5 provides an overview of two basic approaches that are
often undertaken to examine process versus variance models. These two
models capture basic distinctions between research studies undertaken to
investigate either: (1) variance or causal questions of ‘what causes what’; or
(2) process questions of ‘how things develop and change over time.’
Mohr (1982) Wrst distinguished variance and process models in an explan-
ation of organizational behavior. In developing a formalism for the represen-
tation of social action, Abell (1987) contrasted variance and narrative
approaches, while Abbott (1984, 1990) compared stochastic and narrative
explanations in sociology. The common thread running through these works
is the diVerence between scientiWc explanations cast in terms of statistical
associations between independent and dependent variables versus explan-
ations that tell a narrative or story about how a sequence of events unfolds
over time to produce a given outcome. Chapter 5 discusses these divergent
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

explanations between variance and narrative explanations. They constitute

fundamentally diVerent research approaches for examining variance theories
that make causal predictions among variables, as distinct from process theories
that examine progressions in the temporal development of how events unfold
in a social entity, be it an individual, group, organization, or larger community.
An example from the study of organizational change may be useful to
clarify these distinctions between variance theories and process theories. Van
de Ven and Huber (1990) point out that studies of organizational change tend
to focus on two kinds of questions:
. What are the antecedents or consequences of the change?
. How does a change process emerge, develop, grow or terminate over time?
The ‘What’ question usually entails a variance theory explanation of the
input factors (independent variables) that statistically explain variations in
some outcome criteria (dependent variables). The ‘How’ question requires a

process theory explanation of the temporal order and sequence in which a

discrete set of events occurred based on a story or historical narrative (Abbott
1988). In terms of causality, the ‘What’ question requires evidence of co-vari-
ation, temporal precedence, and absence of spurious associations between
the independent and dependent variables. The ‘How’ question explains an
observed sequence of events in terms of some underlying generative mechan-
isms that have the power to cause events to happen in the real world and the
particular circumstances or contingencies when these mechanisms operate
(Tsoukas 1989).
A researcher adopting a variance model is inclined to decompose organ-
izational processes into a series of input–output analyses by viewing each
event as a change in a variable (e.g., the number of product innovations), and
then examining if changes in this variable are statistically associated with
some other independent variable (e.g., R&D investment). From a variance
theory perspective, events represent changes in the states of a variable, and
these changes are the building blocks of variations among variables in an
input–process–output model. But since the process question is not whether,
but how, a change occurred, one needs to narrate a story of the sequence of
events that unfolded as the product innovation emerged. Once the sequence
or pattern of events in a developmental process is found, then one can turn to
questions about the causes or consequences of the event sequence.
Having distinguished the two questions, it is important to appreciate
their complementary relationship. An answer to the ‘What’ question typically
assumes or hypothesizes an answer to the ‘How’ question. Whether implicit or
explicit, the logic underlying an answer to a variance theory is a process
story about how a sequence of events unfold to cause an independent
(input) variable to exert its inXuence on a dependent (outcome) variable.
For example, to say that R&D investment causes organizational innovativeness
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

is to make important assumptions about the order and sequence in which

R&D investment and innovation events unfold in an organization. Thus, one
way to signiWcantly improve the robustness of answers to the Wrst (variance
theory) question is to explicitly examine the process that is assumed to explain
why an independent variable causes a dependent variable.
By the same token, answers to ‘How’ questions tend to be meaningless
without an answer to the corresponding variance theory questions. As Petti-
grew (1990) argues, theoretically sound and practically useful research on
change should explore the contexts, content, and process of change through
time. Just as change is only perceptible relative to a state of constancy, an
appreciation of a temporal sequence of events requires understanding the
starting (input) conditions and ending (outcome) results.
Given the diVerent but complementary epistemologies of variance and
process theories discussed in Chapter 5, I delve into detailed considerations
for designing variance and process studies in Chapters 6 and 7. Chapter 6

Van, D. V. A. H. (2007). Engaged scholarship : A guide for organizational and social research. Oxford University Press, Incorporated.
Created from uu on 2024-11-20 05:12:21.
24 ENGAGED SCHOLARSHIP

focuses on experimental, quasi-experimental, and survey designs for empir-

ically evaluating causal models in variance research. Chapter 7 discusses
methods for designing and conducting longitudinal cases, historical, and
Weld studies to examine processes of how phenomena develop and change
over time.
You might question if this ‘theory testing’ approach admits to a more
exploratory ‘grounded theory building approach’ to research? My response
is that the diVerence between these two modes of inquiry is a matter of timing
and sequence in performing the theory building and research design activities
of the diamond model. In exploratory studies, propositions typically develop
after data are collected and analyzed. Thus, I recommend that the methods
discussed in Chapter 4 for developing theories be applied after or while the
data are being collected and analyzed. However, as discussed in Chapter 2, all
data, facts, and observations are laden with theories that are tacit or explicit in
the minds of the investigators. Any observations presuppose a selective frame
of reference of a chosen object and concepts. Before collecting data, the focus
of an exploratory study can be signiWcantly clariWed by meeting with key
study stakeholders to discuss and explain what concepts might be used to
observe the phenomenon.
Most studies, of course, include elements of both theory building and
theory testing. Numerous iterations in running the paths of the diamond
model are typically required in conducting any research project. Seldom, if
ever, can a researcher complete a study by running the paths in one linear
sequence; much back-tracking and jumping from one base to another is the
typical process sequence.

USING RESEARCH FOR PROBLEM SOLVING

The problem solving activity of the engaged scholarship process focuses on

linking the research Wndings back to the problem observed in the practitioner
and the scientiWc communities. Generally, this involves executing the research
design to produce empirical evidence for a solution to the research problem
and question that initially motivated the research. At a minimum, a research
solution entails a report of research Wndings and a discussion of their impli-
cations for theory and practice. Many researchers consider their communi-
cation task completed when they publish their report in a scientiWc journal
and make verbal presentations of it at professional conferences as well as to
host organizations and practitioners who sponsored the research.
This practice assumes that communicating research Wndings entails a one-
way transfer of knowledge and information from the researcher to an audi-
ence. The underlying assumption of this view is that if an idea is good enough,
it will be used. But research knowledge based on sound empirical evidence

is often not used or adopted as intended by either scientists or practitioners.

I argue that a deeper understanding of communicating knowledge across
boundaries and a more engaged relationship between the researcher and his/
her audience are needed if research Wndings are to have an impact in advancing
science and practice.
It is one thing to write a research paper, and quite another to transfer,
interpret, and implement study Wndings at the communication boundaries of
both scientiWc and practitioner communities. Estabrooks (1999: 15) points
out that ‘Many factors get in the way of using research, and empirically, we
know very little about what makes research use happen or not happen.’
Recently, scholars have begun to reconceptualize knowledge transfer as a
learning process in which new knowledge is shaped by the learner’s pre-
existing knowledge and experience. Individuals are not simply sponges,
soaking up new information without Wltering or processing. ‘Knowledge use
is a complex change process in which ‘‘getting the research out there’’ is only
the Wrst step’ (Nutley et al. 2003: 132). Neither scientists nor practitioners
simply apply scientiWc research, but collaborate in discussions and engage in
practices that actively interpret its value to accomplish their tasks.
I anchor Chapter 8 in Carlile’s (2004) framework of knowledge transfer,
translation, and transformation. It provides useful insights into how resear-
chers might communicate their study Wndings at the knowledge boundaries
with diVerent audiences. The framework emphasizes that communication
across boundaries requires common knowledge among people to understand
each other’s domain-speciWc knowledge. When the diVerence, dependence,
and novelty of domain-speciWc knowledge between people at a boundary
increase, then progressively more complex processes of knowledge transfer,
translation, and transformation are needed to communicate the meanings
and potential uses of that knowledge.
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

When the people at a knowledge boundary share the same common lexicon
and syntax for understanding their diVerent and interdependent domain-
speciWc knowledge, then it can be communicated using a conventional infor-
mation processing view of knowledge transfer from a speaker to listeners
through written and verbal reports. The major challenge of knowledge trans-
fer is to craft a suYciently rich message and medium to convey the novelty of
the information from the speaker to the audience. For example, written
reports, verbal presentations, and face-to-face interactions between the
speaker and listeners represent three increasingly rich media for knowledge
transfer. In addition, logos, pathos, and ethos represent three increasingly rich
dimensions of a message.
Knowledge transfer, however, even when communicated in the richness of
a rhetorical triangle, typically remains a one-way transmission of information
from a sender to a receiver. The listener in knowledge transfer remains
relatively silent, but is never inactive. Authors of research reports will not

Van, D. V. A. H. (2007). Engaged scholarship : A guide for organizational and social research. Oxford University Press, Incorporated.
Created from uu on 2024-11-20 05:12:21.
26 ENGAGED SCHOLARSHIP

know this unless they engage in conversations with readers or listeners of a

report. Then it becomes clear that listeners often have diVerent interpret-
ations and meanings of the novel information than the speaker intended. A
research report is not treated as a social fact or as having a ‘Wxed’ meaning.
Rather, it is open to multiple and unlimited meanings, interpretations, and
actions among participants (speakers and listeners) engaged in the text.
Hence, when communicating research Wndings, a research report should be
viewed as a Wrst—not the last—step for researchers to engage in conversations
with potential users, and thereby gain a broader and deeper appreciation of
the meanings of research Wndings.
When interpretive diVerences exist in the meanings of research Wndings,
then a more complex communication boundary of ‘knowledge translation’
must be crossed. At this boundary, speakers and listeners engage in con-
versations and discourse to mutually share, interpret, and construct their
meanings of research Wndings. Speakers and listeners become co-authors in
mutually constructing and making sense of their interactions. At the know-
ledge translation boundary, conversation is the essence and the product of
research. Engaging in conversation and discourse with an audience requires
researchers to adopt a hermeneutic ‘participant view’ rather than a ‘God’s Eye
view’ of research Wndings.
Communicating across knowledge transfer and translation boundaries may
surface conXicting interests among parties that entails an even more complex
political boundary where participants negotiate and pragmatically transform
their knowledge and interests from their own to a collective domain. As
Carlile (2004) states, ‘When diVerent interests arise, developing an adequate
common knowledge is a political process of negotiating and deWning com-
mon interests.’
Finally, seldom can knowledge transfer, translation, and transformation be
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

accomplished with only one communication among people across boundar-

ies. Numerous interactions are required to share and interpret knowledge,
create new meanings, and negotiate divergent interests. The engaged schol-
arship process provides a strategy to approximate this by repeated engage-
ments of stakeholders in each activity of the research process: problem
formulation, theory building, research design, and problem solving.

FORMS OF ENGAGED SCHOLARSHIP

Engaged scholarship can be practiced in many diVerent ways and for many
diVerent purposes. Figure 1.2 illustrates four diVerent forms of engaged
scholarship. As discussed in Chapter 9, these diVerent forms of engaged
scholarship depend on: (1) whether the purpose of a research study is to
examine basic questions of description, explanation, and prediction or on

Research Question/Purpose
To Describe/Explain To Design/Control

Extension Basic Science Policy/Design Science

Detached with Evaluation Research
Outside Stakeholder Advice for
Professional Practice
Research 1 3
Perspective 2 4
Intension Co-Produce Action/Intervention
Attached Knowledge Research
Inside with Collaborators for a Client

Figure 1.2. Alternative forms of engaged scholarship

applied questions of design, evaluation, or action intervention, and (2) the
degree to which a researcher examines the problem domain as an external
observer or an internal participant.
1. Informed basic research is undertaken to describe, explain, or predict a
social phenomenon. It resembles a traditional form of basic social science
where the researcher is a detached outsider of the social system being
examined, but solicits advice and feedback from key stakeholders and
inside informants on each of the research activities as listed in Figure 1.1.
These inside informants and stakeholders play an advisory role, and the
researcher directs and controls all research activities.
2. Collaborative basic research entails a greater sharing of power and activ-
ities among researchers and stakeholders than informed research. Col-
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

laborative research teams are often composed of insiders and outsiders

who jointly share the activities listed in Figure 1.1 in order to co-produce
basic knowledge about a complex problem or phenomenon. The division
of labor is typically negotiated to take advantage of the complementary
skills of diVerent research team members, and the balance of power or
responsibility shifts back and forth as the tasks demand. Because this
collaborative form of research tends to focus on basic questions of mutual
interest to the partners, it has much less of an applied orientation than
the next two forms of engaged scholarship.
3. Design and evaluation research is undertaken to examine normative ques-
tions dealing with the design and evaluation of policies, programs, or
models for solving practical problems of a profession in question. Variously
called ‘design or policy science’ or ‘evaluation research,’ this form of re-
search goes beyond describing or explaining a social problem, but also seeks
to obtain evidence-based knowledge of the eYcacy or relative success of

Van, D. V. A. H. (2007). Engaged scholarship : A guide for organizational and social research. Oxford University Press, Incorporated.
Created from uu on 2024-11-20 05:12:21.
28 ENGAGED SCHOLARSHIP

alternative solutions to applied problems. Evaluation researchers typically

take a distanced and outside perspective of the designs or policies being
evaluated. Inquiry from the outside is necessary because evidence-based
evaluations require comparisons of numerous cases, and because distance
from any one case is required for evaluation Wndings to be viewed as
impartial and legitimate. But engagement of stakeholders is important so
they have opportunities to inXuence and consent to those evaluation
study decisions that may aVect them. In terms of the engaged scholarship
model, these decisions include the purposes of the evaluation study
(problem formulation), the criteria and models used to evaluate the
program in question (research design), and how study Wndings will be
analyzed, interpreted, and used (problem solving).
4. Action/intervention research takes a clinical intervention approach to
diagnose and treat a problem of a speciWc client. Kurt Lewin, a pioneer of
action research, suggested a learning strategy of both engaging with and
intervening in the client’s social setting. The foundation of this learning
process was client participation in problem solving using systematic
methods of data collection, feedback, reXection, and action. Since Lewin’s
time, action research has evolved into a diverse family of clinical research
strategies in many professional Welds. Action research projects tend to
begin by diagnosing the particular problem or needs of an individual client.
To the extent possible, a researcher utilizes whatever knowledge is available
from basic or design science to understand the client’s problem. However,
this knowledge may not apply or may require substantial adaptation to Wt
the ill-structured or context-speciWc nature of the client’s problem. Action
research projects often consist of N-of-1 studies, where systematic com-
parative evidence can only be gained through trial-and-error experiments
over time. In this situation action researchers have argued that the only way
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

to understand a social system is to change it through deliberate interven-

tion and diagnosis of responses to the intervention. This interventionist
approach typically requires intensive interaction, training, and consulting
by the researcher with people in the client’s setting.
Sometimes advocates of a particular form of research make disparaging
remarks about other forms. This is unfortunate because all four forms of
engaged scholarship are legitimate, important, and necessary for addressing
diVerent research questions (description, explanation, design, or control of a
problematic situation). Which is most appropriate depends on the research
question and the perspective taken to examine the question. Pragmatically,
the eVectiveness of a research approach should be judged in terms of how well
it addresses the research question for which it was intended (Dewey 1938).
Although the four forms of engaged scholarship entail diVerent kinds of
relationships between the researcher and stakeholders in a study, engagement

is the common denominator. The more ambiguous and complex the prob-
lem, the greater the need for engaging others who can provide diVerent
perspectives for revealing critical dimensions of the nature, context, and
implications of the problem domain.

CAVEATS OF ENGAGED SCHOLARSHIP

Several caveats of engagement should be recognized. As discussed in Chapter
9, the practice of engaged scholarship raises a number of issues that are often
not salient in traditional approaches to social research. They include: (1) the
challenges of engagement; (2) being reXexive about the researcher’s role in a
study; (3) establishing and building relationships with stakeholders; and (4)
spending time in Weld research sites. Engagement does not necessarily imply
that a researcher loses control of his/her study, but it does entail greater
accountability to the stakeholders involved in a study. Engagement often raises
false expectations that concerns expressed will be addressed. Engagement does
not require consensus among stakeholders; much learning occurs through
arbitrage by leveraging diVerences among stakeholders. Negotiating diVerent
and sometimes conXicting interests imply that creative conXict management
skills are critical for engaged scholars. Without these skills, engagement may
produce the ancient Tower of Babel, where intentions to build a tower to reach
heaven were thwarted by the noisy and confusing language of the people.
Engaging stakeholders (other researchers, practitioners, sponsors, users, or
clients) in problem formulation, theory building, research design, and problem
solving represents a more challenging way to conduct social research than the
traditional approach of researchers going it alone. But throughout this book I
argue that the beneWts exceed the costs. By involving stakeholders in key steps of
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

the research process, engaged scholarship provides a deeper understanding of

the problem investigated than is obtained by traditional detached research. My
argument assumes, of course, that the primary motivation of engaged scholars
for undertaking research is to understand this complex world, rather than to get
published and promoted. The latter is a by-product of the former.

Discussion
This chapter introducedaresearch process model ofengaged scholarshipthatserves
as the organizing framework for this book. This model incorporates a contempor-
ary philosophy of science and a set of methods for undertaking research with the
aim of advancing knowledge in both a scientiWc discipline and in the practice of
a profession. I argued that a research project involves four activities:

Van, D. V. A. H. (2007). Engaged scholarship : A guide for organizational and social research. Oxford University Press, Incorporated.
Created from uu on 2024-11-20 05:12:21.
30 ENGAGED SCHOLARSHIP

1. Problem formulation—ground the research problem and question in the

real world;
2. Theory building—develop or select a conceptual model that addresses
the problem as it exists in its particular context;
3. Research design and conduct—gather empirical evidence to compare
plausible alternative models that address the research question; and
4. Problem solving—communicate and apply the research Wndings to solve
the research question about the problem existing in reality.
Subsequent chapters discuss ways to perform each of these activities in this
process model of engaged scholarship. Scholars can cover the four bases of the
diamond model in any order they like. But all the bases must be covered to
complete a research project.
This engaged scholarship diamond model incorporates to a wide variety
of research methods including: basic or applied; theory building or theory
testing; variance or process theory; cross-sectional or longitudinal; quantita-
tive or qualitative; and laboratory, simulation, survey, archival, or other
observation methods. Depending on the problem or question being investi-
gated, engaged scholarship may involve any of these diVerent categories of
research. While engaged scholarship entertains a wide variety of research
methods, it directs the research process by specifying the core set of activities
of a research project that need to be performed from start to Wnish. Because
the core activities of problem formulation, theory building, research design,
and problem solving are highly interdependent, so also are the methods that
are selected for doing these activities. Thus, the critical task is to adopt and
execute the research models and methods that Wt the chosen research problem
or question being addressed.
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

ALTERNATIVE RESEARCH MODELS

A basic proposition of the ES model is to compare and contrast a proposed model
with plausible alternative models. To ‘walk this talk,’ I compare the engaged
scholarship model with two other plausible alternative models for conducting
social research: a general systems model of problem solving by David Deutsch
(1997) and a model of the scientiWc episode by Ronald Giere (1999).

Deutsch’s Problem Solving Model

Several scholars have observed that science can be seen as a problem solving
activity (Campbell 1988; Azevedo 1997, 2002; Deutsch 1997). For example,
David Deutsch, a quantum physicist at the University of Oxford, describes
science in terms of Wve problem solving stages: (1) the problem; (2) a proposed

model or conjectured solution; (3) criticism with experimentation; (4) a

solution of replacing erroneous theories; and (5) a new problem that recycles
the process. Deutsch (1997: 62) explains the problem solving stages as begin-
ning when a problem surfaces. A problem starts when a theory is not adequate
and a new theory is needed. It is deWned not only as an emergency or the root
of anxiety, but is more when ideas are not adequate and there may be a better
explanation (Deutsch 1997). In other words and as discussed in Chapter 3,
research often begins with an anomaly requiring abductive reasoning because
the current explanation or theory may be too narrow or not broad enough to
explain the anomaly.
Following the discovery of a problem (stage 1), the next stage is conjecture.
This is ‘where new or modiWed theories are proposed in the hope of solving
the problem (stage 2). The conjectures are then criticized . . . using scientiWc
methods of experimental testing. This entails examining and comparing them
to see what oVers the best explanation, according to the criteria inherent in
the problem (stage 3)’ (Deutsch 1997: 64). A conjectured theory is not
adopted when it seems to provide explanations worse than other theories.
But, if one of the principle theories is abandoned for a new one (stage 4), then
the problem solving exercise is deemed a ‘tentative’ success. Deutsch says the
success is tentative since later problem solving may involve replacing or
changing these new theories and in some cases even going back to and
revising the ideas that were deemed unsatisfactory. Deutsch states, ‘the solu-
tion, however good, is not the end of the story: it is a starting point for the
next problem solving process (stage 5)’ (Deutsch 1997: 64).
Deutsch points out that the objective of science is not to Wnd a theory that
is deemed true forever; it is to Wnd the best theory currently available. A
scientiWc argument is intended to persuade us that a given explanation is the
best one available. It cannot say anything about how that solution will fare in
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

the future when it is subjected to a new type of criticism and compared with
explanations that have yet to be invented. Deutsch says (1997: 64–5), ‘A good
explanation may make good predictions about the future, but the one thing
that no explanation can even begin to predict is the content or quality of its
own future rivals.’
As with the engaged scholarship model, Deutsch points out that the stages
of speciWc problem solving are seldom completed in sequence at the Wrst
attempt. There is usually repeated backtracking before each stage is completed.
Only when the process is finished does a coherent pattern emerge that reflects
the Wve linear stages of problem solving.
While a problem is still in the process of being solved we are dealing with a large
heterogeneous set of ideas, theories, and criteria, with many variants of each, all
competing for survival. There is a continual turnover of theories as they are altered or
replaced by new ones. So all the theories are being subjected to variation and selection.
(Deustch 1997: 68)

Van, D. V. A. H. (2007). Engaged scholarship : A guide for organizational and social research. Oxford University Press, Incorporated.
Created from uu on 2024-11-20 05:12:21.
32 ENGAGED SCHOLARSHIP

Deutsch cites Popper for this evolutionary epistemology. However, he

cautions not to overstate the similarities between scientiWc discovery and
biological evolution, for there are important diVerences. One diVerence is
that biological variations (mutations) are random, blind, and purposeless. In
human problem solving, the creation of models or theories is itself a complex,
knowledge-laden process driven by human intentions. Perhaps an even more
important diVerence is that there is no biological equivalent of logical reason-
ing and argument. The stronger the arguments for problems and theories, the
more inXuential or persuasive they are. Science, like problem solving, justiWes
an explanation as being better than another currently available explanation.

Giere’s Model of a ScientiWc Episode

Ronald Giere, a philosopher of science at the University of Minnesota, has
been inXuential in introducing a pragmatic realist epistemology of science.
This view downplays the idea that there might be universal natural laws
encoded in true general statements. Rather, scientists are seen as engaged in
constructing models that represent or Wt the world in relatively better or
worse ways. It is a kind of realism regarding the application of models to the
real world, but it is a realism that is perspectival rather than objective or
metaphysical (Giere 1999: 60–1). Giere states,
My account of scientiWc epistemology pits one model, or family of models, against
rival models, with no presumption that the whole set of models considered exhausts
the logical possibilities. This means that what models are taken best to represent the
world at any given time depends on what rival models were considered along the way.
And this seems, historically, a contingent matter. So the models of the world held at
any given time might have been diVerent if historical contingencies had been diVerent.
(Giere 1999: 77)
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

Based on this perspectival realist epistemology, Giere (1997) proposed a

model of the scientiWc episode (or research project) consisting of four com-
ponents: (1) a real-world object or problem under investigation; (2) a theor-
etical model of the real-world object or process; (3) some operational
hypotheses or predictions derived from the model including a research design
of what the data should be like if the model really does match with the real
world; and (4) some data (or solutions) that are obtained by observation or
experimentation with the real world (Giere 1997). Giere arranges these
components as shown in Figure 1.3, which correspond closely to the four
research activities in the engaged scholarship diamond model. The Wgure
illustrates four important relations.
1. The relationship between the real world and the model is expressed by a
conceptual proposition or analogy asserting that the model Wts the real-
world problem or phenomenon being examined. It is understood that the

Model Fits/Doesn’t Fit MODEL

REAL WORLD
(Problem/Phenomenon) Hypothesis True/False (Theory)

Observation/
Experimentation Reasoning/
Calculation

DATA PREDICTION
(Solution) (Model)
Agree/Disagree

Figure 1.3. Ronald Giere’s model of a scientific episode

Note: The activities of ES diamond model inserted into Giere’s (1997: 30) figure of the four elements of
a scientific episode.

model Wts only in some respects and then only to some speciWed degree of
accuracy. If the model does not Wt accurately in the intended respects,
then the theoretical model is false.
2. The model and the prediction (what we call theory) are related by reason-
ing or argumentation. The real world and the data are related by a physical
interaction that involves observation or experimentation. ‘If what is going
on in the real world, including the experimental setup (our research
design), is similar in structure to the model of the world then the data
(solution) and the prediction or theoretical hypothesis should agree. That
is, the actual data should be described by the prediction. On the other
hand, if the real world and the model are not similar in the relevant
respects, then the data and the prediction may disagree’ (Giere 1997: 30).
3. The top half of the Figure 1.3 pictures the relationship between the real
world and the model in question. Are the model and the real world
similar in the respects under study and to an appropriate degree of
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

accuracy? This relationship is typically not open to direction inspection.

The bottom part of the Wgure, by contrast, pictures a relationship that can
be evaluated by relatively direct inspection. Scientists can examine the
data and see whether they agree with the predictions derived from the
operational theory or model.
4. The left side of the Wgure illustrates relationships existing between the
problem or phenomenon and data obtained from the real-world obser-
vations. The data are generated through physical interactions with bits of
the real world. The right side of the Wgure between model and theory, by
contrast, consists of conceptual relationships that are mainly symbolic.
The model exists mainly as a description of a possible type of object.
Like the four bases of the engaged scholarship model and Deutsch’s stages of
problem solving, Giere’s Wgure illustrates a fully developed scientiWc episode
containing all four components of a research project arranged to make possible

Van, D. V. A. H. (2007). Engaged scholarship : A guide for organizational and social research. Oxford University Press, Incorporated.
Created from uu on 2024-11-20 05:12:21.
34 ENGAGED SCHOLARSHIP

an evaluation of how well a model Wts the real world. Giere points out that
many scientiWc reports do not include all four components, and many do not
unfold in the deductive, model-testing manner as outlined here.
It is common for example, to Wnd reports that describe only the part of the real world
under investigation together with some new data. There may be no mention of
models or predictions. Similarly, we often Wnd discussions of new models of real-
world entities or processes with no mention of data or predictions. Occasionally we
Wnd accounts of models of real-world things that include predictions but no discus-
sion of data. We can learn a lot from such reports. Unless all four components are
present, however, there may be nothing we can subject to an independent evaluation.
(Giere 1997: 31)

Conclusion
You may wonder if engaged scholars in professional schools should conduct
more applied and less basic research? The answer depends on the research
question and perspective taken to study a problem domain. As Figure 1.2
illustrates, engaged scholarship can be practiced to study a variety of basic and
applied questions. Engaged scholarship represents a strategy for surpassing
the dual hurdles of relevance and rigor in the conduct of fundamental
research on complex problems in the world. By exploiting diVerences in the
kinds of knowledge that scholars and practitioners from diverse backgrounds
can bring forth on a problem, engaged scholarship produces knowledge
that is more penetrating and insightful than when scholars or practitioners
work on the problem alone. More speciWcally, the quality as well as the impact
of research can improve substantially when researchers do four things: (1)
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

confront questions and anomalies arising in practice; (2) organize the research
project as a collaborative learning community of scholars and practitioners
with diverse perspectives; (3) conduct research that systematically examines
alternative models pertaining to the question of interest; and (4) frame the
research and its Wndings to contribute knowledge to academic disciplines, as
well as one or more domains of practice.
Simon (1976) argues that signiWcant invention in the aVairs of the world
calls on two kinds of knowledge: practical knowledge about issues and needs
from the perspective of a profession and scientiWc knowledge about new ideas
and processes that are potential means for addressing these issues and needs.
Historically invention is easier and likely to produce incremental contribu-
tions when it operates among like-minded individuals. Thus we Wnd applied
researchers who tend to immerse themselves in the problems of the end-users
and then apply available knowledge and technology to provide solutions for
their clients. We also Wnd pure disciplinary scholars immersed in their

disciplines to discover what questions have not been answered and then apply
research techniques to address these questions. In either case if researchers
cannot answer their initial questions, they modify and simplify them until
they can be answered. As this process repeats itself, the research questions and
answers become increasingly speciWc contributions to narrow domains of
problems and inquiry. TranWeld and Starkey (1998) point out that researchers
may locate themselves in diVerent communities of practice and scholarship at
diVerent times,
but they cannot stay Wxed in either the world of practice (without risking epistemic
drift driven by politics and funding) or in the world of theory (without retreating to
academic fundamentalism). The problems addressed by management research should
grow out of the interaction between the world of practice and the world of theory,
rather than out of either one alone. (1998: 353)
In the conduct of engaged scholarship, researchers are equally exposed to the
social systems of practice and science, and are likely to be confronted with real-
life questions that are at the forefront of the kind of knowledge and policies that
are used to address problems in the world. This setting increases the chance of
signiWcant innovation. As Louis Pasteur stated, ‘Chance favors the prepared
mind.’ Research in this context is also more demanding because scholars do not
have the option of substituting simpler questions if they cannot solve real-life
problems. Engaged scholarship is diYcult because it entails a host of interper-
sonal tensions and cognitive strains that are associated with juxtaposing inves-
tigators with diVerent views and approaches to a single problem. But focusing
on the tensions between scholars and practitioners, as has often been the case in
the past, may blind us to the very real opportunities that can be gained from
exploiting their diVerences in the co-production of knowledge. As Simon
(1976) observed, if research becomes more challenging when it is undertaken
to answer questions posed from outside an academic discipline, it also acquires
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

the potential to become more signiWcant and fruitful.

The history of science and technology demonstrates that many extraordin-
ary advancements have often been initiated by problems and questions posed
from outside the scientiWc enterprise (Rattan 2001). Necessity is indeed the
mother of important invention. Scholarship that engages both researchers
and practitioners can provide an exceedingly productive and challenging
environment; it not only fosters the creation of knowledge for science and
practice, but it may dissolve the theory–practice gap.

Van, D. V. A. H. (2007). Engaged scholarship : A guide for organizational and social research. Oxford University Press, Incorporated.
Created from uu on 2024-11-20 05:12:21.
2 Philosophy of Science
Underlying Engaged
Scholarship*
John P. Bechara and Andrew H. Van de Ven**
Model
n

Th
sig

eo
De

ry
ch

Bu
ar

ild
se

in
Re

Solution Theory
n
io
at
Pr

ul
ob

rm
le

Fo
m

m
So

le
lv

ob
in

Pr
g

Reality

It is better to choose a philosophy of science than to inherit one by

default.
(Chapter 2, this volume, p. 37)
Many of us are practitioners—not philosophers—of science. We don’t think
much about ontology and epistemology so that we can get on with the craft of
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

doing research instead of talking about it. But underlying any form of research
is a philosophy of science that informs us of the nature of the phenomenon
examined (ontology) and methods for understanding it (epistemology).
Whether explicit or implicit, we rely on a philosophy of science to interpret
the meanings, logical relations, and consequences of our observational and
theoretical statements. Many of us inherit the philosophy of science that
underlies the research practices of our teachers and mentors. Inheriting a
philosophy of science is understandable if an orthodox view of the scientiWc
method exists and is simply taken for granted by the scientiWc community.
While such consensus may have existed among social scientists in the 1960s
and early 1970s, the past 30 years have witnessed a major deconstruction and
revision of traditional views of social science.
* There is a glossary to the philosophy of science terms on page 298.
** John Bechara is a doctoral student and Andrew Van de Ven is a professor at the Carlson School
of Management at the University of Minnesota.

Van, D. V. A. H. (2007). Engaged scholarship : A guide for organizational and social research. Oxford University Press, Incorporated.
Created from uu on 2024-11-20 05:12:33.
UNDERLYING PHILOSOPHY OF SCIENCE 37

As discussed in this chapter, critics argue that social science cannot be

objective, rational, and cumulative because language, culture, social norms,
political ideologies, mental biases, and selective perception constitute the
inputs and processes of science. Science is an intensely human social process,
and hence subject to all of these factors that limit the capabilities for social
science to be objective, rational, and cumulative. Since the demise of the
‘Received View’ of positivism1 in the 1960s and 1970s other philosophies of
science—such as relativism, pragmatism, and realism—have been developing
and are competing for adoption by social scientists. They provide a repertoire
of alternative ways to interpret the nature of things we study and the methods
for doing so. Practitioners of science, in turn, are inXuencing how these
philosophies are developed and expressed in their research.
This chapter attempts to provide a synthesis of the reciprocal relationship
between the philosophy and practice of science by undertaking a brief
historical review of four philosophies of science—positivism, relativism, prag-
matism, and realism. It provides a discussion of how key ideas from each
philosophy inform engaged scholarship, and how the practice of engaged
scholarship might inXuence these philosophies of science. Engaged scholar-
ship requires a comparative understanding of diVerent philosophies of science.
An understanding of a complex problem or phenomenon being investigated
can be enhanced by engaging the perspectives of diverse scholars and stake-
holders. Appreciating these diverse perspectives often requires communicat-
ing across diVerent philosophical perspectives. It also requires maintaining the
diverse intellectual diVerences that not only create an opportunity for arbi-
trage, but also for a productive interplay of perspectives, models, and world
views (Alvesson and Sköldberg 2000).
This chapter also emphasizes that the philosophy underlying our scientiWc
practice is a choice, and should not simply be a default inherited without
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

question from our teachers and mentors. Understanding the implications of

this choice is important not only for engaged scholarship, but for any reXective
and responsible scientiWc inquiry. We turn to philosophy of science to provide
us with the conceptual tools and frameworks to reXect on our practice, and to
understand alternative ways to do social science.
Before reviewing four alternative philosophies of positivism, relativism,
pragmatism, and realism, it is important to clarify in a reXexive spirit that our
own version of engaged scholarship adopts a critical realist perspective. This
view takes an objective ontology (i.e., reality exists independent of our
cognition) and a subjective epistemology. More speciWcally, this perspective
is based on the following principles.
. There is a real world out there (consisting of material, mental, and
emergent products), but our individual understanding of it is limited.

1 Putnam (1962) referred to logical positivism and logical empiricism as the received view.

Van, D. V. A. H. (2007). Engaged scholarship : A guide for organizational and social research. Oxford University Press, Incorporated.
Created from uu on 2024-11-20 05:12:33.
38 ENGAGED SCHOLARSHIP

In general, physical material things are easier to understand than reXexive

and emergent social processes.
. All facts, observations and data are theory-laden implicitly or explicitly.
Social sciences have no absolute, universal, error-free truths, or laws as any
scientiWc knowledge.
. No form of inquiry can be value-free and impartial; each is value-full.
Some methods are better warranted than others depending on the phe-
nomenon.
. Knowing a complex reality demands use of multiple perspectives.
. Robust knowledge is a product of theoretical and methodological triangu-
lation where evidence is not necessarily convergent but might also be
inconsistent or even contradictory.
. Models that better Wt the problems they are intended to solve are selected
allowing an evolutionary growth of knowledge.

Alternative Philosophies of Science

We turn now to a brief historical review of positivism, relativism, pragma-
tism, and realism. Table 2.1 summarizes the discriminating characteristics of
these four philosophical schools, and provides an outline of the review
discussed below. In addition, the Appendix to this chapter contains a glossary
of key philosophical terms that may be a useful reference while reading the
chapter. (The Appendix is located at the end of this volume, before the
Bibliography.) The four philosophies of science are featured in this chapter
not only because they inXuence our view of engaged scholarship, but also
because they reXect many of the current practices and debates among social
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

scientists.
Johnson and Duberley (2000, 2003) distinguish positivism, relativism,
pragmatism, and realism in terms of their ontological and epistemological
perspectives.2 Logical positivism is ontologically objective (implicitly assumes
an objective world independent of cognition) due to its construal of an
empirical reality devoid of metaphysical entities and epistemologically object-
ive due to its emphasis on correspondence between statements and reality
using inductive veriWcation. In contrast, relativism is ontologically subjective

2 Ontology focuses on the nature of things, while epistemology deals with how we gain knowledge
about these things. Campbell (1988) points to a circularity problem in these deWnitions because any
ontological description presupposes an epistemological one, and vice versa. ‘Ontology has to do with
what exists, independently of whether or not we know it. But to describe what exists I have to use a
language of knowledge claims, and hence contaminate the deWnition with epistemology’ (Campbell
1988: 440).

Van, D. V. A. H. (2007). Engaged scholarship : A guide for organizational and social research. Oxford University Press, Incorporated.
Created from uu on 2024-11-20 05:12:33.
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.
Created from uu on 2024-11-20 05:12:33.
Van, D. V. A. H. (2007). Engaged scholarship : A guide for organizational and social research. Oxford University Press, Incorporated.

Table 2.1. Comparison of the characteristics of Logical Positivism, Relativism, Pragmatism and Realism

Dimensions Logical Positivism Relativism Pragmatism Realism

Definition Philosophical movement Contemporary intellectual Philosophical movement Philosophical movement

inspired by empiricism, movement characterized by its characterized by the relation of characterized by the existence of
instrumentalism, and positivism skepticism about the theory and praxis and specifically a mind-independent reality and
(Vienna Circle, Berlin School). foundations of Western in the predetermined outcomes the ability of a theory to capture
philosophy (historical relativism, of an inquiry (relativism—Dewey partial aspects of reality
social constructivism, and Rorty; realism—Peirce, (conjecture realism, structural
postmodernism, critical theory, James, and Rescher). realism, realistic pragmatism,
hermeneutics). critical realism . . . ).
Ontology Objective: Reality is the empirical Subjective: Reality is socially Subjective: Similar to Objective: Reality exists
world (the world of the senses constructed. postmodernism. Objective: independent of our cognition.
i.e., the rejection of the Reality places limitations and Thus, there is no basis to reject
metaphysical). constraints on our actions. the metaphysical (epistemic
fallacy).
Epistemology Objective: The correspondence Subjective: There is no privileged Subjective and dependent on Subjectivist: There is no
between our statements and epistemology due to the practical consequences. predefined or predetermined
reality through inductive incommensurability of methodology or criteria to judge
verification or deductive discourses. the veracity our knowledge.
falsification.
Knower Positivist: The knower is Constructivists: The knower is in Consequentialist: The knower Perspectivalist: The knower has
independent of the empirical the world and cannot stand has a priori cognitive a priori cognitive frameworks
world (passive observer). outside of his/her socio-linguistic frameworks which affect his/her which affect his/her perception
Furthermore, the mind can constructs to view it objectively. perception of the world. of the world.
mirror the empirical reality.
Language Language is value free and Language is self-referential, i.e., Language is not self-referential Language is not self-referential
provides a means to mirror and it does not refer to any but actionable and functions to or theory neutral but describes
correspond to the empirical transcendental entity beyond meet the agent’s goals and albeit partially, the underlying
world (analytic/synthetic itself. This presupposes its value purposes. mechanism and structure of a
distinction). and interest-laden nature. phenomenon.
40 ENGAGED SCHOLARSHIP

due to its construal of a socially constructed reality and adopts a subjective

epistemology due to its denial of an objective and impartial representation of
social reality. Pragmatism includes philosophers, who take either objective or
subjective views of ontology, but all adopt a subjective epistemology that
emphasizes the relation between knowledge and action—knowledge is ‘truth-
ful’ to the extent that it is successful in guiding action. Finally, realism adopts
an objective ontology (there is a reality out there independent of cognition)
and an objective or subjective epistemology. In short, positivism and relativ-
ism represent the outer limits of philosophical thought with their contrasting
ontology and epistemology, and hence, bracket the discussion of pragmatism
and realism, which lie in-between positivism and relativism.
The choice of these four philosophical schools and their labels was based
in part on the historical development of the philosophy of science as an
academic discipline. According to Boyd (1991) and Suppe (1977) it began
with the logical positivists. Subsequent literature largely developed in
reaction to it, such as social constructivism, which is placed under the
relativist label and scientiWc realism, which is placed under the realist label.
Understanding how these diVerent perspectives developed in reaction to each
other provides a useful Wrst step in appreciating, selecting, and possibly
synthesizing a philosophy of science that overcomes some of the concerns
and criticisms of contemporary skeptics and hopefully initiates a process of
reXexivity.
Before beginning, we admit that you may interpret the philosophies of
science reviewed here very diVerently, and may disapprove of the choice
of perspectives and the labels used to frame our discussion.3 We say this in
an open and reXexive spirit of choosing and synthesizing the philosophy
of science that Wts your scholarly practice. We do not presume that this
chapter captures the breadth and depth of perspectives needed to make a
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

well-grounded choice or synthesis. However, we do hope to raise awareness

that the particular philosophy of science that is practiced is, and ought to be, a
critical choice, rather than a default. Further study in the philosophy of
science may be needed to make this choice.

3 In previous drafts we received critical feedback on a variety of ways to classify and label the many
philosophies of science. We confess to not having found a solution that adequately reXects and is
sensitive of the philosophical identities of various scholars. In particular, we appreciate and are
sympathetic of the critical feedback from interpretive, postmodern, and hermeneutic scholars who
objected to our grouping of their perspectives under the label ‘relativism.’ One reviewer said she ‘felt
oVended’ by the label because of a pejorative ‘anything goes’ ethical connotation associated with the
term ‘relativism.’ This is not our intent. Following Suppe (1977), Laudan (1984), and McKelvey
(2002a), we use relativism as an ‘existence concept’ of the philosophical perspectives that view reality
as socially constructed and that ‘deny the existence of any standard or criterion higher than the
individual by which claims to truth can be adjudicated’ (McKelvey 2002b: 896).

BACKGROUND
Human beings have a unique capacity to represent reality and reXect upon it.
This capacity coupled with the desire to control nature prompts scientiWc
inquiry, or what Reichenbach (1963) called ‘the art of correct generalization.’
The goal of this art is to create knowledge—not mere opinion—that can
be generalized across space and time. Philosophy of science examines the
conceptual foundations and methods of this process of scientiWc inquiry.
Like most forms of Western intellectual thought, the history of philosoph-
ical thought of science can be traced back to the ancient Greeks. Reichenbach
(1948, 1963), for example, begins his historical lineage with the debate
between the rationalists and the empiricists. The rationalists believed that
reason was the sole source of reliable knowledge. Reason was able to control
empirical observations and order them into a logical system that made the
prediction of future observations possible.
One of the Wrst rationalists—also known as an idealist—is Plato (427–347
bc). He believed that an ‘idea’ exhibits the properties of objects in a perfect
way, and thus we learn about these objects through their respective ideas not
through the objects themselves. The laws of ‘ideas’ govern and provide reliable
knowledge of the physical world. Plato’s student, Aristotle (384–322 bc),
similarly believed that the mind was the source of those laws. The leading
rationalist of the Enlightenment period and the founder of modern philosophy,
Rene Descartes (1596–1650), argued that the distinctive feature of rationalism
was a belief that the laws that control the physical world can only be discovered
through the reasoning of the mind (Russell 1972). Implicit in this assertion is
the distinction between observer and the world, also referred to as Cartesian
dualism. However, Descartes never denied completely the contributions of
empirical observations to our knowledge of the physical world. He relied on
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

logical deductive rules, beginning with a set of axioms or premises considered

to be true to infer valid conclusions about the world. For example, if all swans
are white, and the particular bird that we observe is a swan, then we deduce
that the bird is white.
In contrast, empiricists believed that sensory experience was the sole source
of reliable knowledge. The most devout empiricist was Francis Bacon (1561–
1626), recognized as the pioneer in the logical systematization of the scientiWc
inquiry (Russell 1972). He, as well as the ancient Greek philosopher Democ-
ritus (460–370 bc) and eighteenth century philosophers John Locke (1632–
1704) and David Hume (1711–76), attempted to replace the rationalist
method of deduction with induction as the proper method for attaining
reliable knowledge of the physical world. Induction starts with the enumer-
ation of past and present empirical observations to draw inferences about the
physical world. For example, if all the swans we observe are white, then we
conclude that all swans are white.

Van, D. V. A. H. (2007). Engaged scholarship : A guide for organizational and social research. Oxford University Press, Incorporated.
Created from uu on 2024-11-20 05:12:33.
42 ENGAGED SCHOLARSHIP

Although rational deduction and empirical induction remain with us

today, both forms of inquiry suVer from numerous shortcomings. In particu-
lar, Reichenbach (1963) discussed three major shortcomings: (1) the assump-
tion of rationalism that the premises in a deductive argument are true without
resorting to sensory observation; (2) the assumption of empiricism that all
knowledge is based on sensory experience while it is clear that the method of
induction is not a product of sensory experience; and (3) the assumption of
empiricism that the limited past and present observations provide a basis to
predict future observations. This last shortcoming was one of the major
contributions of Hume in specifying the problem of induction.
The rationalist cannot solve the problem of empirical knowledge because he construes
such knowledge after the pattern of mathematics, and thus makes reason the legislator
of the physical world. The empiricist cannot solve the problem either; his attempt
to establish empirical knowledge in its own right as derived from sense perception
alone breaks down because empirical knowledge presupposes a non-analytic method,
the method of induction, which cannot be regarded as a product or experience.
(Reichenbach 1963: 90–1)
Immanuel Kant (1724–1804) attempted to reconcile the rationalist and
empiricist views by synthesizing their respective contributions. He contended
that there exist synthetic a priori principles of the physical world that preclude
any sensory experience. This synthetic a priori consists of axiomatic principles
such as the premises of geometry or causality that are assumed to be ‘given’ to
the human mind. Furthermore, he contended that axiomatic theorizing can
be used to derive other synthetic statements about the physical world that act
as regulative mechanisms to organize sensory experience and subsequently
create knowledge. Thus, with the synthetic a priori, Kant believed he showed
that knowledge was a combination of a priori and a posteriori principles, or a
combination of rationalism and empiricism.
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

Kant’s contributions occurred during a period of cultural development in

the eighteenth century known as the Age of Enlightenment. The Enlighten-
ment was characterized by liberation from the theo-centric view and replaced
by an anthropocentric view that emphasized human reason as the sole means
to understanding the world (Russell 1972; Popkin 1999). It was in this cultural
context that Auguste Comte (1798–1857) coined the word positivism as a
philosophy aimed at showing that human knowledge had reached a stage of
development that transcends religious dogma and places hope in the progres-
sive accumulation of knowledge using the empirical sciences and speciWcally
physics as the model of all sciences. Years later logical positivism emerged in
Germany as an extension of the Enlightenment, a synthesis of the scientiWc
development of the nineteenth and early twentieth centuries, and a reaction
to G. W. Hegel’s (1770–1831) metaphysics, which sought to explain reality in
terms of abstract metaphysical entities that did not have any empirical

manifestation. According to Johnson and Duberley (2000), the three major

rationalist tenets used by the positivists were: subject–object dualism (Carte-
sian dualism—the world is independent of the subject), truth as an agreement
of the mind with reality, and truthful knowledge is to correctly represent
reality in the mind. In addition, the three empiricist tenets used by the
positivists were: induction as the mode of reasoning to gain knowledge of
reality, reduction of reality to empirical data, and reduction of causality to
Humean4 constant conjunction.

LOGICAL POSITIVISM5
Logical positivism emerged from the Vienna Circle—Moritz Schlick
(1882–1936), Rudolf Carnap (1891–1970), and Herbert Feigl (1902–88);
and the Berlin School—Hans Reichenbach (1891–1953) and Carl Hempel
(1905–97). These pioneering scholars were mainly scientists and mathemat-
icians who became philosophers. Logical positivism rejected Kant’s a priori
elements in science due to their analytic nature/self-referential character,
adopted a blend of positivism, empiricism, instrumentalism, and beneWted
from the contributions of Frege, Russell, and Whitehead in mathematics
and Wittgenstein in language. It construed the role of philosophy as the
analysis of science from a logical perspective using what was known as a
language of veriWable propositions (Blumberg and Feigl 1948). According to
Suppe (1977), the goal of logical positivism was to eliminate all metaphysical
entities from philosophy and science that implied ontological neutrality (i.e.,
emphasis on epistemology) (Niiniluoto 1999). From August Comte, it
adopted the privileging of science (and speciWcally physics) as a model for
all other sciences. From Ernest Mach (1838–1916), logical positivism adopted
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

a radical empiricist attitude whereby the only source of knowledge of the

physical world was sense observation. From Henri Poincare (1854–1912)
logical positivism adopted instrumentalism, which denied theoretical terms
any referential value.6 From Gottlob Frege (1848–1925) and Bertrand Russell

4 Hume deWned causality as a product of habitual experience. The four conditions to ensure
causality are: constant conjunction (two events are constantly associated with each other), antecedence
(events occur sequentially in time), contiguity (both events are spatially in the same location), and
necessity (no alternative observation). The last condition, necessity, was problematic since it was
impossible to observe all instances of the phenomenon under investigation and thus any universal law
from a Wnite number of observations can never be certain.
5 Another form of logical positivism was logical empiricism, which substituted the ontological
neutrality of the former with a realist ontology (i.e., that there exist a partial mind-independent reality).
One of the most ardent logical empiricists was Herbert Feigl (a member of the Vienna circle) at the
University of Minnesota who founded the Center for the Philosophy of Science, and was instrumental
in diVusing logical positivism in the United States during the Wrst half of the twentieth century.
6 Referential value refers to the existence of unobservable entities in the physical world, which are
represented using theoretical terms in science.

Van, D. V. A. H. (2007). Engaged scholarship : A guide for organizational and social research. Oxford University Press, Incorporated.
Created from uu on 2024-11-20 05:12:33.
44 ENGAGED SCHOLARSHIP

(1872–1970) logical positivism adopted logic as a means to analyze science

and accepted the analytic nature of mathematics. It is this development in
mathematics that led to the demise of neo-Kantianism and the synthetic
a priori (Ayer 1982) since the latter relied on mathematics as a source
for knowledge of the physical world. Finally, from Ludwig Wittgenstein
(1889–1951), logical positivism adopted the veriWability theory of meaning,
which states that understanding the meaning of a proposition consists of
understanding the circumstances in which it could be veriWed or falsiWed. In
the remainder of this section, we will focus on some of the main positivistic
tenets followed by some of its earliest critics.
According to Boyd (1991), the veriWability theory of meaning, or veriWca-
tionism, was a doctrine used by logical positivism to address the demarcation
problem between science and non-science (metaphysics). One implication of
veriWcationism was the distinction between science and mathematics/logic.
Science was considered to be the only source of synthetic knowledge of the
world based on empirically observable terms. Mathematics and logic were
considered to be sources of analytic knowledge. Giere (1988: 26) illustrates this
relationship between theoretical and observational terms in Figure 2.1. He
points out that logical positivism distinguished scientiWc theories from pure
axiomatic logic or mathematics by the empirical interpretation of non-logical
terms and the use of correspondence rules to explicitly link theoretical terms
with observational terms.
The standard doctrine was that the meaning of theoretical terms is totally a function
of the meaning of the observational terms together with purely formal relations

POSTULATES

PRIMITIVE
CONCEPTS
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

DEFINED
CONCEPTS

EMPIRICAL
CONCEPTS

'SOIL' of
OBSERVATION
(EXPERIENCE)

Figure 2.1. A logical empiricist picture of a scientific theory

Source: Reproduced from Feigl (1970) in Giere, R. N. (1988). Explaining Science: A Cognitive Approach.
Chicago: University of Chicago Press, p. 25.

speciWed by the axioms of the theory. The result is instrumentalism with regard to the
theory part of theories. Theoretical terms do not refer to real entities; they are mere
instruments for organizing claims about the things referred to by observational terms.
The view of scientiWc theories was typically pictured as in Figure 2.1. (Giere 1988: 26)
Another implication of veriWcationism was the separation between the gen-
esis of a theory and its validity. The genesis of a theory was viewed as the context
of discovery, which was believed to be the concern of psychology and history.
The validity of a scientiWc theory provided the context of justiWcation, which
was believed to be the concern of logic and philosophy. Implicit in this separ-
ation is the independence of the social, psychological, and economic factors
inXuencing the scientist and his/her scientiWc theories. Reichenbach states:
What we wish to point out with our theory of induction is the logical relation of
the new theory to the known facts. We do not insist that discovery of the new theory is
preformed by a reXection of a kind similar to our expositions; we do not maintain
anything about the question of how it is performed—what we maintain is nothing
but a relation of a theory to facts, independent of the man who found the theory.
(1938: 382)
This led logical positivism to focus attention on theories as Wnished prod-
ucts waiting to be justiWed and to ignore factors in the genesis of theories
(Suppe 1977: 125). In Weick’s (1999) terms, logical positivism was concerned
with theory, rather than the process of theorizing as discussed in Chapter 4.
Furthermore, logical positivism emphasized induction as a means of develop-
ing empirically veriWable generalized propositions from empirically veriWable
particular propositions.
Logical positivism formulates research questions and hypotheses in prop-
ositional form and necessitates the use of empirical tests to verify these
propositions through careful control (manipulation) to avoid confounding
conditions and outcomes. The researcher is assumed to be independent of the
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

objects observed in the world, and capable of studying the objects without
inXuencing, or being inXuenced by, them. When inXuence in either direction
(threats to validity) is recognized or suspected, various strategies are followed
to reduce or eliminate it. Inquiry takes place ‘through a one-way mirror’
(Guba and Lincoln 1994: 110) in a sort of correspondence between our
thoughts/signs and reality. By following rigorous experimental procedures,
values, and biases are prevented from inXuencing outcomes and empirical
truth is established through replicable Wndings. This view of a researcher is
discussed later as a ‘God’s Eye’ frame of reference.
Suppe (1977) provides an extensive discussion of the criticisms of logical
positivism. Only a brief summary of selected criticisms can be mentioned
here. One of the earliest criticisms of logical positivism was by one of its
pioneers, Hans Reichenbach (1948) of the Berlin School. He argued that
logical positivism could not adequately solve the problem of induction and

Van, D. V. A. H. (2007). Engaged scholarship : A guide for organizational and social research. Oxford University Press, Incorporated.
Created from uu on 2024-11-20 05:12:33.
46 ENGAGED SCHOLARSHIP

subsequently the predictive nature of science. The problem of induction is

that in using it, it is never possible to arrive at a single theory. The positivistic
belief that science, by induction, can ultimately converge on the ‘real’ truth is
thus brought sharply into question.
Quine (1951) asserted the impossibility of having a clear distinction between
analytic and synthetic statements and furthermore, the reduction of complex
propositions into clear observational terms. He examined a diYculty Wrst
raised by Duhem (1962) that a theory cannot be conclusively falsiWed, because
the possibility cannot be ruled out that some part of the complex test situation,
other than the theory under test, is responsible for an erroneous prediction.
This diYculty is called the Duhem/Quine thesis (Chalmers 1999: 89).
Popper (1959) showed that logical positivism fails to provide an adequate
answer for the demarcation problem. Popper replaced positivism’s induction
and veriWcation with abduction and falsiWcation. Following the pragmatist,
C. S. Peirce, Popper argued that the process of developing a theory does not
begin with an inductive enumeration of observational data, but rather with
creative intuition (as will be discussed in Chapter 4). Furthermore, he avoided
Hume’s radical skepticism by showing that any process of veriWcation (proving)
is illusory and should be replaced with a process of falsiWcation (dis-proving).
The process of falsiWcation leads to an epistemological Darwinism where
the Wttest of theories survive empirical refutation. Like Popper, Donald
T. Campbell argued for an evolutionary growth of knowledge in which the
scientiWc community selects those models that better Wt the problems they are
intended to solve (see the section on realism).
Along with Popper, Norman Hanson held that a major defect of logical
positivism was that it conWnes attention only to the Wnished product of
scientiWc theorizing and gives no attention to the process of reasoning
whereby laws, hypothesis, and theories receive their tentative Wrst proposal
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

(Hanson 1958: 71). In ‘Patterns of Discovery’ Hanson (1958) extended the

work of Peirce (see pragmatism) to develop the logic of scientiWc discovery.
He emphasized that theories are not discovered by inductively generalizing
from data but rather are retroductively inferred hypotheses from conceptually
organized data. Hanson viewed observations and facts as theory-laden, and
notions of causality as reXecting a certain form of conceptual organization.
Causation is not a property of the physical world; it is a way people make
sense of the world. He developed a logic of discovery (retroductive reasoning)
that reXects the process in which scientiWc conjectures and laws are developed.

RELATIVISM
We use relativism as a general term denoting a set of alternative philosophies
of science that emerged in reaction to, or in denial of, positivism. The set

includes many variations: historical relativism, social constructivism, post-

modernism, critical theory, and hermeneutics. This grouping of philosophies
into a single category is based on our judgment, since we cannot cover all
their individual viewpoints. However, the perspectives included in our broad
category of relativism all break away from the positivist assumption that
scientiWc knowledge is a cumulative, unmediated, and complete representa-
tion of reality. They deny a solution to the demarcation problem; emphasize
the intertwinement between the genesis and validity of a theory, view reality
as socially constructed, and the goal of social science as that of understanding
what meanings people give to reality, not only to determine how reality
works. Furthermore, they reject the positivistic belief that scientiWc methods
provide a way to develop an objective ‘Truth’ of the concrete reality in the
world. Instead, they believe that scientists construct an image of reality based
on their interests, values, and viewpoints in interaction with others. From this
standpoint, observations and data give us ‘nothing more than facts. The
truth (small t) is what we make consensually of these facts. And in the social
world, truth is therefore collectively constructed. . . . Truth referents are not
in the facts but in the collective interpretation of the facts’ (Gioia 2003: 288).
We now examine some of the main tenets of the varieties of relativism:
historical relativism, social constructivism, postmodernism, critical theory,
and hermeneutics.
Some of the Wrst critics of positivism were the historical relativists (Toulmin
1953; Feyerabend 1962, 1975; Kuhn 1962, 1970) and social constructivists
(Bloor 1976; Latour and Woolgar 1986). Kuhn (1962, 1970), a historical rela-
tivist, called into question the belief that scientiWc knowledge is cumulative. He
argued, instead, that the development of science is dependent on the sociological
paradigm agreed upon within the scientiWc community. He replaced the a priori
cognitive structures of Kant with a paradigmatic view that reXected the set of
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

beliefs, values, assumptions, and techniques that guided the scientist. Kuhn
viewed scientiWc knowledge as progressing in a cycle of three phases. The Wrst
phase is normal science, where one particular paradigm has control over a
scientiWc community. The second phase is a crisis where abnormal or inexplic-
able observations arise. Finally, the last phase is revolutionary science where a
new paradigm replaces the old paradigm. The replacement of the old paradigm
is through a consensus of the scientiWc community and not through corres-
pondence with reality as had occurred with positivism. Furthermore, the
acceptance of a new paradigm presupposes incommensurability with other
paradigms because of the absence of an agreed-upon, objective criterion for
comparing the truth claims of alternative paradigms.
In a similar way, Feyerabend (1962, 1975), Toulmin (1953), and other
historical relativists argue for the idea of a socially constructed nature of
scientiWc knowledge. However, they diVer in some historical interpretations
and conclusions. For example, Toulmin construed scientiWc theories as neither

Van, D. V. A. H. (2007). Engaged scholarship : A guide for organizational and social research. Oxford University Press, Incorporated.
Created from uu on 2024-11-20 05:12:33.
48 ENGAGED SCHOLARSHIP

true nor false but more or less adequate answers for observed irregularities.
These irregularities occur when the current theories break down or are unable
to provide an adequate answer. Moreover, the development of new scientiWc
theories is based on a Weltanschauung, which is an evolving socio-conceptual
framework similar to Kuhn’s paradigm. However, unlike Kuhn’s view that
paradigms change all at once, Toulmin construed the change in Weltanschau-
ung as more or less gradual.
Perhaps the strongest negation of positivism is postmodernism, which is
skeptical of modern science, technology, and social transformations produced
by the Enlightenment. The demarcation between modern and postmodern eras
is unclear, but it is claimed to have started with Nietzsche in the late nineteenth
century (Dallmayr 1987; Sim 2001) and reached its epitome in the second half
of the twentieth century during a period of unprecedented socio-economic
and technological transformation. Postmodernism is an eclectic school of
thought that encompasses post-industrialism, post-capitalism, and post-
structural forms of thought. One common theme is skepticism about the
major foundations of Western thought, and in particular positivism being a
product of the scientiWc and mathematical developments of the nineteenth and
early twentieth century (Alvesson and Deetz 1996; Sim 2001). This skepticism
undermines the attainment of truth, the criteria determining truth, and even its
very existence. It is based on anti-essentialist and anti-foundationalist notions
about the nature of reality and the ways of knowing reality respectively. Overall,
postmodernism denies positivism’s logico-linguistic turn to the analysis
of science through the use of propositions, blurs the distinction between
observable and theoretical terms, and more importantly denies the distinction
between the genesis and validity of a theory.
Anti-essentialism refers to a rejection of the essence of phenomena and the
causal mechanisms underlying them. The essentialist notion is the cornerstone
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

of scientiWc inquiry. As Sim (2001) suggests, essentialism regards the attain-

ment of truth, meaning, and origin as its goal. This parallels the Hobbesian and
Humean critiques of Aristotelian essences, which consisted of the distinction
between ‘what’ an object is and ‘how’ it is. Hence, postmodernism rejects the
possibility of capturing the essence of the phenomenon and subsequently
rejects its existence.
Anti-foundationalism refers to the rejection of foundational or self-evident
beliefs required in the pursuit and acquisition of knowledge. Hence, postmod-
ernism rejects the basis of epistemology that asserts the existence of self-justifying
or self-evident Wrst principles that guide scientiWc inquiry.
Postmodernism’s anti-essentialist and anti-foundationalist rejections have
implications for our conception of the word ‘truth,’ which lacks a common
intrinsic feature that permits us to judge our theories. According to Rorty
(1979), the conception of truth has become nothing but a sign of approval or
agreement given to promote one theory over others (Engel 2002). An extension

of this particular theme lies at the heart of Jean Baudrillard’s denial of the
possibility of distinguishing between reality and simulation in our postmodern
world (Sim 2001). According to Baudrillard, the world is a ‘simulacra,’ where
reality and simulation are intertwined and undistinguishable (as often experi-
enced when playing virtual reality games).
According to Cahoone (1996), postmodernism denies the distinction be-
tween the presence of an entity and its representation. One aspect of this
criticism stems from the complexity, diYculty, and to a certain extent im-
possibility of representing the world in an unmediated and holistic fashion.
Another aspect stems from the denial of linguistic essentialism whereby
language is viewed as a mirror to objectively represent the presence of an
object (Hassard 1994; Alvesson and Deetz 1996). An example is Jacques
Derrida’s deconstruction, which denies the possibility of using language to
represent reality. He supports his denial by demonstrating that we think only
in signs and that the process of signiWcation is an endless shifting from sign to
sign which can never be terminated by reducing the signifying process to some
transcendental starting-point or end-point. This leads Derrida to state that
there is nothing outside the text that is represented. Furthermore, Rorty
claims that diVerent languages constructed within diVerent socio-cultural
contexts are incommensurable and thus knowledge is incommensurable.
This inevitability of incommensurability means that a consensus as a general-
izable epistemic standard is rejected.
Next, Cahoone (1996) argues that postmodernism attempts to show unity
as plurality whereby entities are shown to be a product or function of their
respective relationship with other entities. Hence, any apparent unities are
implicitly repressing their dependency on and relations to others. A proponent
of this view is Jean-Francois Lyotard, who rejects the grand theoretical enter-
prises or ‘grand narratives’ that serve only to justify our actions. He believes in
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

a multiplicity of alternatives to explanation or mini-narratives, where any

claim to a unitary or linear progression is a suppression of other possible
alternatives. ScientiWc knowledge becomes a plurality incapable of legitimizing
itself and based on multiple language games (Gasche 1988). As discussed in the
next section, this notion of selves is rooted in the William James’s construal of
pragmatic philosophy.
Postmodern theorist and feminist scholars (such as Martin 1990) have
deconstructed and surfaced a number of voices or interests that are typically
‘marginalized’ in positivistic accounts of social organization or behavior.
In the guise of technocracy, instrumental rationality has pretenses to neutrality
and freedom from the value-laden realms of self-interest and politics. It celebrates and
‘hides’ behind techniques and the false appearance of objectivity and impartiality
of institutionalized sets of knowledge, bureaucracy, and formal mandates. Not surpris-
ingly, technocracy is promoted by each of the management ‘specialisms’ as they claim a

Van, D. V. A. H. (2007). Engaged scholarship : A guide for organizational and social research. Oxford University Press, Incorporated.
Created from uu on 2024-11-20 05:12:33.
50 ENGAGED SCHOLARSHIP

monopoly of expertise in their respective domains. Human resource specialists, for

example, advance and defend their position by elaborating a battery of ‘objective’
techniques for managing the selection and promotion of employees (Hollway 1984;
SteVy and Grimes 1992). Strategic management institutionalizes a particular way
of exercising domination through legitimizing and privileging the ‘management’ of
the organization-environment interface, producing some actors as ‘strategists’ and
reducing others to troops whose role is to subordinate themselves to implement cor-
porate strategies (Alvesson and Willmott 1995; Shrivastava 1986). The concept of
technocracy draws attention to some of the darker and more disturbing aspects of so-
called ‘professional management.’ It points to a restricted understanding of human and
organizational goals; those that are identiWed and validated by experts. By associating
management with technocracy and its instrumentalization of reason, the domination of
a narrow conception of reason is at once exposed and questioned. (Alvesson and Deetz
1996: 203–4)
Hence, what positivists’ thought were impartial, objective, and value-free
accounts of science, relativists have shown to serve the interest and values of
people in power. Moreover, Zald (1995) claimed that ‘Most of the ‘‘brute
facts’’ that are subject to enumeration in positivistic social science gain their
force because of the cultural/social meanings in which the subjects participate.
Explanation in the causal sense must give way to, or be embedded in,
hermeneutic unveiling and interpretation’ (Zald 1995: 456).
Critical theory emerged from the Frankfurt school. Its founders included
among others Max Horkheimer (1895–1973), Theodor Adorno (1903–69), and
Herbert Marcuse (1898–1979). They adopted a Marxist framework to oppose
the destructive eVects of capitalism. The aim of critical theory was to diagnose
the problems of modern society and identify the nature of the social change
needed to produce just and democratic societies. Critical theory shares some
commonalities with postmodernism although a key diVerence is that it main-
tains hope that knowledge can lead to emancipation and progress. Habermas
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

(1979, 1984, 1987, 1990), a second generation critical theorist, adopts a conven-
tionalist position that deploys a consensus theory of truth, as a regulative
standard to assess the extent of systematically distorted communication. He
also avoids extreme relativism because he assumes that through ideal speech
communication, we might attain a consensual view of truth.
The last major relativistic perspective reviewed here is hermeneutics. Ori-
ginally focused on interpreting the meaning of the Scriptures, hermeneutical
philosophers expanded their scope into philology (the science of linguistic
understanding and the study of interpretative processes and beings) and
many additional areas beyond biblical interpretation, including the social
sciences. Although positivists focused almost exclusively on epistemology,
hermeneutical philosophers (such as Heidegger 1927/1962; Gadamer 1960/
1975; and Bernstein 1983) emphasized that epistemological issues are strongly
related to the ontological positions we might take.

As against the positivists who assumed it is possible to gather knowledge

about entities that are observed independent of the observer, hermeneutics
questioned this shallow approach to epistemology:
It argues that how we view the existence of objects in the extra-mental world is a
function of how we frame our own existence and relationship with the environment.
(By ‘extra-mental reality,’ I mean the world as it exists independently from how an
individual perceives it.) Our perception of the outside world is a function of how
we perceive our own position in and relationship with the phenomena—in the
past, present, and future. We therefore, need a better understanding of the reference
frames we use to make sense of this relationship and the extent to which these frames
confuse us or help us clarify and focus our thought processes—especially if we want
to develop a proper understanding of the processes taking place in the world.
(Hendrickx 2002: 341)
Margaretha Hendrickx applies this hermeneutical perspective by contrast-
ing a positivistic ‘God’s Eye frame’ with a critical realist ‘participant frame’
that Mr. Jones, a hypothetical management researcher, might take to conduct
research. Hendrickx (1999: 344) uses Figure 2.2 to illustrate ways that
Mr. Jones might make connections between Popper’s (1979) ‘Three World
View’: (1) a material world illustrated by the globe; (2) a world of mental
states and processes as illustrated by the human Wgure, and (3) an emergent

Knowledge
The management embodied in texts
researcher
Evolutionary
(2) describe Dynamics of
Organizations
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

(1) observe
(3) verify

extra-mental reality

Figure 2.2. God’s Eye frame of reference: triangular reasoning

Source: Hendrickx (1999: 344).

Van, D. V. A. H. (2007). Engaged scholarship : A guide for organizational and social research. Oxford University Press, Incorporated.
Created from uu on 2024-11-20 05:12:33.
52 ENGAGED SCHOLARSHIP

world of products of individual and collective human minds, represented by

the book.
In the God’s Eye frame, ‘the world consists of some Wxed totality of extra-
mental objects. There is exactly one true and complete description of ‘‘the way
the world is.’’ Truth involves some sort of correspondence relation between
words or thought-signs and external things’ (Putnam 1981: 49). Here, the
researcher views him/herself as a value-neutral observer of real things existing
in the world (illustrated by arrow 1). Arrow 2 stands for the relation between
the researcher’s mental interpretation of the observations and his/her mental
model as represented by linguistic signs in a text. Arrow 3 represents the relation
between the linguistic signs in the text and the real-world phenomena.
The God’s Eye point of view inspires Mr. Jones to think of his relationship with
phenomena under investigation as one of the sides of a triangle (Figure 2.1). He
functions in this triangle as a (very complicated) mirror. His research activities may
be broken down in primarily three tasks, ‘observing,’ ‘describing,’ and ‘verifying.’ First
he inductively observes what is happening in the world. The reXected photons fall on
his eyes’ retinas and induce an electron cascade that leads to the creation of photo-
graph-like images of the phenomena in his brain. Via a very complex set of biochemical
and neurological reactions, these images are translated in patterns of dots on paper or
digital signals stored on a computer disk. Mr. Jones describes what he observes in the
world. He then generalizes his empirical Wndings in hypotheses and deductively tests
(and observes) whether the postulated relations hold true.
This triangular reasoning squeezes Mr. Jones out of the world, so to speak. He
believes that with proper training, he is capable of transcending his own subjectivity, as
if he was able to turn his values and preferences oV as easily as he turns his computer on
to write up his research Wndings. It is the job, the duty, of Mr. Jones to publish articles
with ‘true’ descriptions of what happens in corporations. He believes that it is possible
to obtain the value-neutral state of mind of an outsider. . . . He perceives his relation-
ship with these as independent of time, space, and mind, as if he is like God. . . . He
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

believes that what he sees is the way the phenomena in the extra-mental world are.
(Hendrickx 1999: 342–3)
A participant frame of reference views the researcher as an active participant
in the domain he/she attempts to study and understand. Although this frame
has also been called the ‘internal’ view (Evered and Louis 1981) and ‘pragma-
tists’ worldview’ (Putnam 1981; Rorty 1982), Hendrickx (1999: 375) prefers
the term ‘participant’ for being neutral on us-vs.-them and inside-vs.-outside
dichotomies that are implicit in the God’s Eye frame. Figure 2.3 presents
Hendrickx’s depiction of the participant frame.
Mr. Jones now thinks of himself as participant in a discourse about ways to help
companies succeed in the long run [the research problem being investigated]. He
perceives himself as a voice in a universal conversation, in which the various points of
view of actual persons reXect their various interests and purposes (Putnam 1981: 49–50).
One of these purposes is to Wnd the most clarifying lens with which to discover [and
represent the problem and its resolution].

Text

Evolutionary
Dynamics of
Organizations

(2) writing (1) reading

(3) listening

Author Reader

(4) talking

(5) experiencing (5) experiencing

Extra-mental reality

Figure 2.3. Participant frame of reference: quadrangular reasoning

Source: Adapted from Hendrickx (1999: 345).

In the participant worldview, a management researcher explicitly acknowledges

that he is the product of a certain history and culture. Thus, Mr. Jones realizes that he
knows as much as he learned from the books that he read, the experiences he
underwent, and the conversations in which he participated. He has come to terms
with the subjective nature of what he knows and understands the futility of attempting
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

to reason in a value-neutral way. Instead, Mr. Jones talks openly about his research
values and investigates whether they make sense after all. Mr. Jones attempts to
understand whether or not his espoused values are the values he actually uses in his
research. He also wants to know the extent to which his values in use are consistent
with values beneWting the human species as a whole (Campbell 1979: 39; 1982:
333–4). His values motivate him (Campbell 1993: 36). He looks upon his research
questions as issues with practical consequences for him, his neighbors, and the top
management teams he studies. (Hendrickx 1999: 346)
Hendrickx illustrates Mr. Jones’ participant frame of reference in a quad-
rangle, as depicted in Figure 2.3. As a participant, the researcher performs the
roles of an investigator and author with other co-investigators, co-authors, or
readers engaged in the discourse. The extra-mental real world and the text or
model represents the material and socially-constructed worlds, respectively,
that the participating researcher(s) and others construct by experiencing,
talking, listening, reading, and writing.

Van, D. V. A. H. (2007). Engaged scholarship : A guide for organizational and social research. Oxford University Press, Incorporated.
Created from uu on 2024-11-20 05:12:33.
54 ENGAGED SCHOLARSHIP

So, compared to the God’s Eye view, where the Other is either an onlooker like Mr.
Jones, or alternatively, someone down there to be observed, the participant frame of
references does not classify readers and writers as a function of whether they know less
or more; rather it implies that they know something diVerent. (Hendrickx 1999: 346–7)
Hendrickx (1999: 341) concludes by advocating the participant frame of
reference and rejecting the God’s Eye frame. She states that a God’s Eye frame
tends to encourage an authoritative and dogmatic attitude on the part of the
author, which promotes close-mindedness and intolerance of alternative
perspectives. Such an attitude is not conducive to listening to and learning
of the viewpoints of others about the real-world phenomenon or alternative
ways to represent it. A participant frame of reference requires an open-
minded attitude that encourages engagement and learning with others.

PRAGMATISM
Pragmatism is an American philosophical school of thought that emerged in
the late nineteenth century with Charles Sanders Peirce (1839–1914). In an
article, ‘How to Make our Ideas Clear,’ Peirce (1878/1997) introduced the term
pragmatism, a term derived from Kant and traced back to the Greek word
action. Pragmatism sought to reconcile rationalism and empiricism by show-
ing that knowing and doing are indivisibly part of the same process. In
philosophy of science, pragmatism was viewed as an alternative to logical
positivism and was aligned with instrumentalism, which is the view that
scientiWc theories are not true or false but are better or worse instruments for
prediction (Misak 2001). Some philosophers went further to assert that Peirce’s
thought not only provides an alternative to logical positivism, but actually
repudiates, in advance, some of its major developments (Rorty 1961). Accord-
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

ing to Meyers (1999) pragmatism espouses three theories: (1) a theory of the
mind, where beliefs are hypotheses and ideas are plans of action; (2) a theory of
meaning, where ideas can be clariWed by revealing their relationship with
action; and (3) a theory of truth, where beliefs are true when they succeed in
guiding action and prediction. Pragmatism is multifaceted and seems to vary
according to each pragmatist (Lovejoy 1908). To minimize confusion we focus
on the main arguments and criticisms of three pioneering pragmatists—Peirce,
James, and Dewey—and two contemporary scholars—Rorty and Rescher.
Peirce introduced the pragmatic maxim of ascertaining the meaning of an
idea in terms of the practical consequences that might conceivably result from
the truth of that conception. The sum of these consequences constitutes the
meaning of the conception (Rescher 2000: 9). Peirce viewed meaning as an
inference for repeatable actions, both as habitual behavior in a reoccurring
situation over time and as generalizations of actions to larger contexts or
diVerent situations (Dewey 1916). In his article ‘Fixation of Belief,’ Peirce

states that the function of beliefs is to commit us to action. This criterion for
truth bears not only on the success of the application, but also on the extent to
which it sustains a long term commitment from the scientiWc community.
Peirce acknowledged the fallibility of inductive scientiWc inference. Instead, he
proposed a method of scientiWc discovery through systematic observation
and creative inference.
Peirce introduced abduction or retroduction as a creative mode of discov-
ery that follows neither inductive nor deductive modes of inference. ‘Induc-
tion was widely believed to be the basic process in science. Peirce denied this,
arguing that induction serves not to initiate theory but rather to test it’
(Mounce 1997: 17). As discussed in Chapter 4, abduction is a hypothetical
inference, framed to solve a problem. The new conception is not Wnal. Further
inquiry will reveal problems that can be solved only by framing a fresh
conception (Mounce 1997: 17).
Peirce’s belief in science and the oneness of truth was rooted in an onto-
logically realist stance. He defends his realist stance by arguing that there is no
reason to believe that a mind-independent reality does not exist. He suggests
this belief is harmonious with our practice of science. He states:
There are real things, whose characters are entirely independent of our opinions about
them; those realities aVect our senses according to regular laws, and, though our
sensations are as diVerent as our relations to the objects, yet, by taking advantage of
the laws of perception, we can ascertain by reasoning how things really are and any
man, if he have suYcient experience and reason enough about it, will be led to the one
true conclusion. (Peirce 1878/1997: 21)
Peirce intended pragmatism to be a rational and empirical substantiation
of knowledge claims (Rescher 2000). He construed the meaning of a term or
proposition as constituted by its practical consequences and its truth by its
success to satisfy the intended aims. These aims, which included successful
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

prediction and control, were the aims of science.

William James (1842–1910), a contemporary of Peirce, elaborates upon
and alters Peirce’s philosophical approach in his seminal lectures in 1907 on
Pragmatism (James 2003). In these lectures, James describes the current
dilemma of philosophy as a diVerence in temperament between rationalism
and empiricism. He contends that individuals inevitably exhibit character-
istics of both sides of the debate. Thus, in order to continue the abstractness of
rationalism and the particularism of empiricism he proposes pragmatism:
You want a system that will combine both things, the scientiWc loyalty to facts and
willingness to take account of them, the spirit of adaptation and accommodation, in
short, but also the old conWdence in human values and the resultant spontaneity,
whether of the religious or of the romantic type. (James 2003: 9)
For James, pragmatism provides a method to settle metaphysical disputes
because through it one compares the practical consequences of adopting

Van, D. V. A. H. (2007). Engaged scholarship : A guide for organizational and social research. Oxford University Press, Incorporated.
Created from uu on 2024-11-20 05:12:33.
56 ENGAGED SCHOLARSHIP

alternative views. Two alternative or rival views are identical if their practical
consequences are identical. Thus, the goal of philosophy is to discover the
diVerence between rival views based on their consequences to diVerent
individuals. James described his version of pragmatism as being a comprom-
ise between empiricism (which claimed that an objective world commands
thought) and idealism (which claimed that subjective thoughts construct the
world). He construes pragmatism as a less objectionable but more radical
version of empiricism. Pragmatism replaces abstraction, a priori reasons, and
Wxed principles with concrete facts and action. It is this emphasis on experi-
ence that James uses to depict action or practice, which lies at the heart of
pragmatism, as a means for solving metaphysical problems and developing a
theory of truth. He states:
Theories thus become instruments, not answers to enigmas in which we can rest. It
agrees with nominalism for instance, in always appealing to particulars; with utilitar-
ianism in emphasizing practical consequences; with positivism in its disdain for
verbal solutions. (James 2003: 24)
James adopted a realist ontology and asserted the existence of a reality
independent of our cognition. He states, ‘The notion of a reality independent
of either of us, taken from ordinary experience, lies at the base of the pragma-
tist notion of truth’ (James 1908: 455). According to James, the truthfulness of
a theory is evident through its success as an instrument that is loyal to past
experience but also is able to transcend it to generate new facts and to hold so
long as it is believed to be ‘proWtable for our lives’ (James 2003: 34). In contrast
to Peirce (who was inXuenced by Kant), James was inXuenced by the British
empiricists such as Locke, Hume, and Mill. James, therefore, viewed pragma-
tism quite diVerently from Peirce. While Peirce viewed pragmatism as a
methodology for converging on a Wxed standard, James invited pluralism.
He entertained a diversity of views about a phenomenon that allowed not
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

only for diVerences among individuals, but even diVerent inclinations and
viewpoints within individuals (Rescher 2000: 19).
James interprets pragmatism as: (1) a method to solve metaphysical dis-
putes by which one compares the practical consequences of adopting alter-
native views; and (2) a theory of truth where truth is veriWcation, which is
consistent with our beliefs and experience. Truth is made and can change over
time. James refuses the rationalistic relationship between mind and world,
which presupposes a passive reality. He contends that this relationship is and
should be viewed as pragmatic, future looking, and dynamic, which presup-
poses an active reality in line with Darwinian evolution.
John Dewey (1859–1952), a student of Peirce, viewed pragmatism as a
means to attain societal goals. While Peirce’s pragmatism was theoretical and
oriented to natural science, and James’s was personalistic and psychological,
Dewey’s pragmatism was communalistic and society-oriented. Dewey’s pos-
ition was intermediate between Peirce and James and emphasized its social

aspect by viewing truth as a ‘matter of communally authorized assertability’

(Rescher 2000: 27). Like Peirce, Dewey refers to his version of pragmatism as
‘instrumentalism,’ which is grounded in scientiWc realism as a means to
remove doubt through social consensus.
The presupposition and tendencies of pragmatism are distinctly realistic; not idealis-
tic . . . Instrumentalism is thus thoroughly realistic as to the objective or fulWlling
conditions of knowledge. (Dewey 1905: 324–5)
According to Rescher (2000), Dewey diVers from Peirce in his conception
of social consensus. Dewey’s view is that social consensus is not based on
epistemic factors (empirical evidence) but on socio-political factors. Dewey
contends that the success of theories is based on their ability to realize the
goals of societal improvement and development.
Two contemporary pragmatists, Richard Rorty (1931–) and Nicholas Rescher
(1928–), take clearly diVerent views of pragmatism. Rorty (1980), inXuenced by
James and Dewey, adopts a postmodern view of pragmatism. To Rorty, prag-
matism is subjectivistic, anti-foundationalistic, and anti-essentialistic, where
truth and validity lose any type of decisive weight and lack any generalizable
epistemic standards. Rorty illustrates his anti-essentialism by arguing that
truth does not have an essence or any type of isomorphic correspondence
with reality. Therefore, any attempt at a progressive accumulation of truth is
fruitless. For Rorty, the consequences of knowledge in practice provides a way to
state something useful about truth. He believed that all the vocabulary of
isomorphic pictures, models, and representations of reality should be replaced
with one of practical consequences and implied actions. Rorty states:
The whole vocabulary of isomorphism, picturing, and mapping is out of place here as
indeed is the notion of being true of objects. If we ask what objects these sentences
claim to be true of, we get only unhelpful repetitions of the subject terms—‘the
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

universe,’ ‘the law,’ ‘history.’ (1980: 723)

In making these remarks Rorty was probably reacting to the formal syn-
tactical view of theories. However, other philosophers—such as Suppe (1989),
Giere (1999), Morgan and Morrison (1999)—had a less relativistic response
by replacing this syntactical view with a semantic view of theories. In this view
models (rather than correspondence rules) provide the interpretation for the
theory. As discussed in Chapter 6, models mediate between theories and data
(Morrison and Morgan 1999: 5).
Rescher (2000), following Peirce, adopts a more realistic view of pragmatism
and repudiates Rorty’s relativistic pragmatism. Rescher proposes three steps to
realign pragmatism with its Peircian roots: (1) pragmatism should be construed
as a philosophical system that holds ‘success’ as epistemic for eVective predic-
tion, control, and explanation; (2) success is objective and independent of
personal preferences; and (3) pragmatism is a method and not a doctrine.

Van, D. V. A. H. (2007). Engaged scholarship : A guide for organizational and social research. Oxford University Press, Incorporated.
Created from uu on 2024-11-20 05:12:33.
58 ENGAGED SCHOLARSHIP

Rescher emphasizes pragmatic success as inextricably intertwined with the

scientiWc enterprise. Principles of eYcacy in prediction and eVective interven-
tion in nature are essential to pragmatic success and are the foundations
of scientiWc inquiry. Rescher provides a pragmatic justiWcation for realism,
which maintains that there is a real world—a realm of mind-independent,
objective physical reality—out there, even though our abilities to understand it
are severely limited. He emphasized that the stable aim of science is to provide
useful models of reality. To Rescher (2003), realism is only justiWed by the fact
that our knowledge of reality is itself fallible and we can never fully comprehend
its complexity. The existence of a mind-independent reality is not the result of
scientiWc inquiry, but a presupposition of inquiry. As discussed in the next
section, Rescher argues that it is pragmatically useful for scientiWc inquiry to
presuppose realism.

REALISM
Realism contends that there is a real world existing independently of our
attempts to know it; that we humans can have knowledge of that world; and
that the validity of our knowledge is, at least in part, determined by the
way the world is. Realism is a philosophical theory that is partly metaphysical
and partly empirical. It transcends experience but is testable by experience
(Leplin 1984). This section discusses some of the historical underpinnings
of realism followed by some of its variations including: scientiWc realism,
conjectural realism, realistic pragmatism, and critical realism (evolutionary
critical realism).
Historically, realism was concerned with the existence of unobservable
entities that lie beyond human perception. Rescher (1987) traces the debate
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

regarding unobservables to three schools of thought: instrumentalism (his-

torically known as nominalism), realism, and approximationism (historically
known as conceptualism). As discussed before, instrumentalism rejects the
existence of unobservables and regards any reference to such entities in
scientiWc theories as a means or tool to help explain the observable phenom-
ena. In contrast, realism accepts the existence of unobservables and contends
that scientiWc theories reference and capture such entities as they exist in the
real world. Finally, approximationism asserts the existence of unobservables;
however, it contends that scientiWc theories roughly capture these unobser-
vables as they exist in the real world. In other words, a weak form of realism
adopts a loose isomorphic representation of reality, whereas the strong form
of realism contends a direct isomorphic relationship, and instrumentalism
repudiates any type of isomorphism.
Chalmers (1999) simpliWes the debate to be between anti-realism versus
realism. Anti-realism (like instrumentalism) restricts scientiWc theorizing

to the observable and avoids any metaphysical/speculative claims. For

anti-realism the criterion of success for a scientiWc theory is its ability to
predict observable phenomena. Here theories simply serve the function of
‘scaVolding to help erect the structure of observational and experimental
knowledge, and they can be rejected once they have done their job’ (Chalmers
1999: 233).
In contrast to positivism and relativism, scientiWc realism (a strong form of
realism) contends that science develops statements that are true at both
theoretical and observational levels of phenomena. It claims that science
continues to progress by attaining closer approximations of reality. ‘We cannot
know that our current theories are true, but they are truer than earlier theories,
and will retain at least approximate truth when they are replaced by something
more accurate in the future’ (Chalmers 1999: 238).
Major criticisms of scientiWc realism were raised by relativists who ques-
tioned the belief in absolute truth and approximation to it (Toulmin 1953;
Feyerabend 1962, 1975; Kuhn 1962, 1970; Bloor 1976; Latour and Woolgar
1986). The completeness, correctness, and progressively-truer nature of scien-
tiWc knowledge were at stake. Niiniluoto (1980: 446) states, ‘No one has been
able to say what it would mean to be ‘‘closer to the truth,’’ let alone oVer criteria
to determine such proximity.’
Several variations of realism developed in response to the criticisms of
relativists. Suppe (1977) argued that Kuhn’s view of rapid paradigm shifts was
historically inaccurate, and rejected his claims of incommensurability among
theoretical terms across paradigms. If paradigms are truly incommensurable,
how is it possible that scholars compare diVerent paradigms and communi-
cate across the paradigms? Hacking (1983) argued that relativism inappro-
priately emphasizes the distinction between observable and unobservable
entities while neglecting the scientiWc methods of experimentation that ma-
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

nipulate and control entities to reveal their eVects.

Popper (1959) and his followers developed conjectural realism, a moderate
realist position. This position emphasizes the fallibilism of scientiWc knowledge
and acknowledges the discontinuous progression of science. Chalmers states:
So the conjectural realist will not claim that our current theories have been shown to
be approximately true, nor that they have conclusively identiWed some of the kinds of
things that are in the world. . . . Nevertheless, it is still maintained that it is the aim of
science to discover the truth about what really exists, and theories are appraised on the
extent to which they can be said to fulWll that aim. (1999: 240)

Rescher (2003) also responded to the relativists by providing a pragmatic

explanation for realism and developing realistic pragmatism. Realistic prag-
matism emphasized that the aim of science is to provide a useful model of
reality. To Rescher (2003), realism is only justiWed by the fact that our
knowledge of reality is itself fallible, and we can never fully comprehend its

Van, D. V. A. H. (2007). Engaged scholarship : A guide for organizational and social research. Oxford University Press, Incorporated.
Created from uu on 2024-11-20 05:12:33.
60 ENGAGED SCHOLARSHIP

complexity. He points out that realism represents a presupposition for

inquiry, not a result of it (Rescher 2000: 126).
The commitment to a mind-independent reality is, all too clearly, a precondition for
empirical inquiry—a presupposition we have to make to be able to use observational
data as sources of objective information. We really have no alternative but to presup-
pose or postulate it. Objectivity represents a postulation made on functional (rather
than evidential) grounds: We endorse it in order to be in a position to learn by
experience. What is at issue here is not so much a product of our experience of reality
as a factor that makes it possible to view our experience as being ‘of reality’ at all. As
Emmanuel Kant clearly saw, objective experience is possible only if the existence of
such a real, objective world is an available given from the outset rather than the
product of experience—an ex-post facto discovery about the nature of things.
(Rescher 2000: 127)
Rescher (2000) develops six important reasons why a presumption of
realism is needed for scientiWc inquiry:
1. Realism is indispensable for the notion of truth as a correspondence
between our ideas and reality. A factual statement cannot be ascertained
if there is no Wnal arbiter independent of our cognizing. Rescher (2000:
130) states, ‘A factual statement on the order of ‘‘There are pi mesons’’ is
true if and only if the world is such that pi mesons exist within it.’
2. Realism is indispensable for the distinction between our subjective
thoughts and opinions of reality and reality the way it actually is. Rescher
(2000: 131) quotes Aristotle: ‘. . . that which exists does not conform to
various opinions, but rather the correct opinions conform to that which
exists.’
3. Realism is indispensable for communication and inquiry within the
scientiWc community. It is established that the scientiWc community
shares a real world where there are real objects which would ‘. . . serve as
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

a basis for inter-subjective communication’ (Rescher 2000: 134).

4. Realism is indispensable for communal inquiry within the scientiWc com-
munity. It would be absurd to have a shared focus of epistemic strivings
that imperfectly estimate reality when there is no reality. He states, ‘We
could not proceed on the basis of the notion that inquiry estimates the
character of the real if we were not prepared to presume or postulate from
the very outset a reality for these estimates to be estimates of ’ (2000: 132).
5. Realism is indispensable for the very idea of inquiry is hinged upon the
conception of an independent reality and the attempts to understand it,
albeit not fully. He states, ‘Without the conception of reality we could not
think of our knowledge in the fallibilistic mode we actually use—as
having provisional, tentative, improvable features that constitute a
crucial part of the conceptual scheme . . .’ (2000: 132).

6. Finally, realism is indispensable because our conception of causality is

dependent on our attempts to empirically understand the real world.
‘Reality is viewed as the causal source and basis of the appearances, the
originator, and determiner of the phenomena of our cognitively relevant
experience’ (2000: 133).
Bhaskar (1979, 1998a,b,c) developed a form of realism known as critical
realism. He and his supporters viewed critical realism as a middle ground
between positivism and relativism (Collier 1994; Harvey 2002; Kemp and
Holmwood 2003). From relativism, critical realism assumed an anti-founda-
tional stance by acknowledging the fallibilism of our knowledge of reality that is
conceptually mediated and theory-laden. It also rejected the existence of
axioms or synthetic a priori principles that provided epistemic knowledge of
reality (Cruichskank 2002). From positivism, critical realism emphasized
empirical experimentation. However, it denied the possibility of generalizing
its experimental outcomes because reality is an open system consisting of
underlying contingent structures. Moreover, it maintained a mind independ-
ent, stratiWed reality consisting of underlying structures and mechanisms
that determined how things come to behave (transcendental realism). It also
held that theoretical entities have referential value (i.e., theoretical entities
genuinely reXect the way the world is).
To this critical realist perspective, Donald T. Campbell added an evolution-
ary view of the development and progression of scientiWc knowledge (Camp-
bell 1989a,b, 1990a, 1991, 1995; Campbell and Paller 1989; Paller and
Campbell 1989). He replaced Kuhn’s social constructivist interpretation of
scientiWc development with a selectionist evolutionary epistemology. Scien-
tiWc progress evolves via a process of blind variation and selective retention.
Reality (as opposed to mere opinions) serves as an external arbitrator or
common referent in editing beliefs and theories for winnowing out inferior
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

theories. Campbell (1988: 447) states, ‘I am an epistemological relativist, but

I am not an ontological nihilist.’ McKelvey (1999: 384) states, ‘His [Campbell]
development of evolutionary epistemology reXects his continuing interest in
the dynamics of how sciences change in their search for improved verisim-
ilitude in observation and explanation without abandoning objectivist onto-
logical realism.’
According to Azevedo (1997), Campbell shows how the process of blind vari-
ation and selective retention of biological evolution applies to science. She states:
Campbell argues convincingly that reality plays a part in editing beliefs, particularly in
the sort of environment in which the organism’s perceptual mechanisms evolved.
Both biological evolution and scientiWc progress evolve via a process of blind variation
and selective retention. Science, seen as a problem solving activity, is continuous with
the problem solving activity of all organisms. (Azevedo 1997: 92)

Van, D. V. A. H. (2007). Engaged scholarship : A guide for organizational and social research. Oxford University Press, Incorporated.
Created from uu on 2024-11-20 05:12:33.
62 ENGAGED SCHOLARSHIP

Campbell combined his selectionist stance with a validity-seeking hermen-

eutics to justify the validity of knowledge based on a consensus among the
scientiWc community. ScientiWc communities generally do not reach consen-
sus based simply on opinions and beliefs. ScientiWc communities vary, of
course, on the standards used to reach consensus. In the social sciences, the
standards or criteria used to reach consensus typically include sound logical
arguments and empirical evidence to substantiate the claims that are made.
While social scientists debate the nature of the arguments and evidence that
they consider legitimate and persuasive, most are willing to accept that: (1)
science is a process of error correction; (2) science is based on evidence
obtained from outside of the scientists about the world; and (3) while
evidence is theory-laden and error-prone, it is nevertheless useful for dis-
criminating between plausible alternative models for understanding a phe-
nomenon in question. This basic method of comparing evidence and
arguments between alternative claims is not undertaken to achieve an ultim-
ate Truth; instead, it is to select among competing alternative claims about a
question or problem at a given time and context. The theories and models
that better Wt the problems they are intended to solve are selected, whereas
those that are less Wt are ignored or winnowed out. Campbell argued that this
successive process of comparative selection accumulates into an evolutionary
growth of scientiWc knowledge by the scholarly community.

Discussion and Implications for Engaged Scholarship

In summary, logical positivism was an extension of the Enlightenment and
modernism’s faith in objectivity, reason, and the progress of scientiWc know-
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

ledge. It emphasized sensory observation and induction as the foundation of

scientiWc knowledge. Underlying this assumption is a value-free and neutral
observer and language. It denied all metaphysical statements as having any
correspondence with reality and considered them meaningless due to their
failure to pass the veriWability theory of meaning or veriWcationism. This also
led to conXating epistemology with ontology. Positivism reduced causal
relations or explanations to a Humean constant conjunction of events and
emphasized the unity of science or the primacy of the physical sciences as the
model for all sciences.
The perspectives that we included in relativism all reacted to positivism’s
emphasis on certainty, its anti-metaphysical attitude, its reliance on sensory
observation, and its modernistic values. Relativism represents a host of
philosophical schools of which just a sampling was described including:
historical relativism, social constructivism, postmodernism, critical theory,
and hermeneutics. These schools converged on their construal of truth as

being socially constructed and theory-laden. They adopted an anti-essentialist

stance that denied science of its objectivity and empirical/rational basis and
denied any privileged way of acquiring knowledge of it.
Pragmatism developed as an alternative to the historical debates between
rationalism and empiricism although more recently variants of pragmatism
provide an alternative to positivism. It attempted to reconcile the abstractness
of rationalism with the particularism of empiricism. Pragmatism is charac-
terized by the relation of theory and praxis and speciWcally in the predeter-
mined outcomes of an inquiry. Despite Lovejoy’s criticism of the varieties of
pragmatisms, they shared a common construal of truth as the success in
guiding action and prediction. Ideas were clariWed by showing their relation-
ship to practice. Unlike, positivism’s emphasis on induction, pragmatism
embraced abduction as the mode of scientiWc discovery. Depending on the
pragmatist, they adopted an objective or subjective ontology and epistemol-
ogy or a combination of both.
Similar to pragmatism’s attempt to provide an alternative to the historical
debates between rationalism and empiricism, realism was also an attempt to
provide such an alternative. More recently, critical realism developed an
alternative between logical positivism and the more relativistic positions.
Also, similar to pragmatism and relativism, realism consisted of numerous
perspectives which shared in common an objective ontology that presup-
poses the existence of a mind-independent reality and the ability of a theory
to capture partial aspects of reality. In contrast to positivism and relativism,
more contemporary forms of realism viewed truth as being a process
of successive approximations of reality, or verisimilitude. Furthermore, it
rejected the positivistic adoption of constant conjunction and the relativistic
view of socially constructed causal relations and replaced them with a
realistic construal of causal mechanisms that exist independently of our
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

knowledge. Contemporary forms of realism also acknowledge the falliblism

of scientiWc knowledge and attempt to explain the progression of knowledge
using an evolutionary metaphor. Finally, most forms of realism adopt
some form of subjective epistemology where there are no predeWned or
predetermined methodologies or criteria that provide privileged views of
reality.
It is tempting to view the four philosophies, especially positivism and
relativism, as incommensurate and antithetical to each other. If you adopt
this view you will probably choose one philosophy that seems closest to your
own preferences and condemn the others as ‘unscientiWc,’ ‘uncaring,’ or
perhaps just ‘unrealistic.’ In contrast and like Schultt (2004: 79), we think
there are signiWcant beneWts from adopting a more inclusive research phil-
osophy that is open to and integrates some of the diVerences of alternative
philosophies of science. Engaged scholarship represents an example of such
integration. Ontologically, engaged scholarship adopts Bhaskarian critical

Van, D. V. A. H. (2007). Engaged scholarship : A guide for organizational and social research. Oxford University Press, Incorporated.
Created from uu on 2024-11-20 05:12:33.
64 ENGAGED SCHOLARSHIP

realism with its middle-ground position between positivism and relativism

and its layered/stratiWed/multi-dimensional mind-independent reality. How-
ever; it also adopts Rescher’s realistic pragmatism to provide a pragmatic
justiWcation for its realist stance. With critical realist ontology, engaged
scholarship adopts a Campbellian relativist evolutionary epistemology to
understand the macro-level accumulation of scientiWc knowledge and its
weak anti-foundationalist methodological stance where there are better war-
ranted methods depending on the phenomenon. It also adopts triangulation
across convergent, inconsistent, and contradictory data to understand the
micro-level development of more robust scientiWc knowledge. However, the
development of engaged scholarship’s philosophical underpinnings also ben-
eWted from other philosophical and metaphysical perspectives of which the
most inXuential are discussed below.

REFLEXIVITY
Postmodern and hermeneutic scholars have emphasized the interests, values,
and biases that are served by researchers. No inquiry can be objective in the
sense of being impartial and comprehensive by including a balanced repre-
sentation of all stakeholders’ viewpoints. Critical theorists point out that
meanings and interpretations of organizational life get played out in a context
of power relationships. ‘Meanings are always politically mediated’ (Putnam,
1993: 230). Pragmatic and realist philosophers also emphasized the theory-
laden nature of human perception, conceptualization, and judgment. The
empiricist view was criticized because of the impossibility of pure, unmedi-
ated observation of empirical ‘facts’ (Mingers 2004: 90). That being the case,
engaged scholars need to be far more reXexive in their studies than positivists
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

and empiricists have admitted. ReXexivity is characterized by diVerent types

of recursive turns each providing diVerent insights and perspectives (Alvesson
and Sköldberg 2000).

ABDUCTION
Peirce argued that induction serves not to initiate theory but rather to test
it. The basic process in initiating theory was what he called abduction or
hypothetical inference. As discussed in Chapter 4, this form of inference begins
by engaging with the world and encountering an anomaly or breakdown that
is inconsistent with our understanding or theory of the world. Abduction
entails creative insight that resolves the anomaly if it were true. A conjecture
developed through abductive inference represents a new plausible alternative

to the status quo explanation of a given phenomenon in question. Because it

might solve the problem, such an insight merits further development and
elaboration as a defensible theory through deductive logic, and then testing
through inductive inferences.

SCIENCE IS AN ERROR-CORRECTION PROCESS OF KNOWLEDGE

DEVELOPMENT
It is easy to ‘throw out the baby with the bathwater.’ Philosophers of science
have extensively criticized and rejected central tenets of logical positivism and
empiricism, which had become the received view of science by the 1970s.
Despite the demise of the received view, McKelvey (1999) discusses a legacy of
useful principles that withstood criticism and are clearly apparent in contem-
porary social science.
Many key ingredients of positivism nevertheless still remain in good standing among
scientiWc realists, such as theory terms, observation terms, tangible observables
and unobservables, intangible and metaphysical terms, auxiliary hypothesis, causal
explanation, empirical reality, testability, incremental corroboration and falsiWcation,
and generalizable statements. . . . The received view is ontologically strong, in the sense
that it posits an external reality and that successive scientiWc discoveries and theories
over time more and more correctly describe and explain this reality; reality acts as a
strong external criterion variable against which scientiWc theories are held account-
able. (McKelvey 1999: 386)
The most fundamental of these principles, we believe, is that science is an
error-correction process that is based on evidence from the world rather than
merely reXecting the scientist’s opinions of the world. Indeed, McKelvey
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

(2002a: 254) asserts that ‘the singular advantage of the realist method is its
empirically-based, self-correcting approach to the discovery of truth.’
However, the relativists, like the pragmatists Dewey and James, cautioned
that ‘hard-and-fast, capital-T-Truth is simply an illusion’ (Westphal 1998: 3).
Ideas and beliefs are nothing but human constructions, shaped by social
processes and procedures. Truth is that which gets endorsed and accepted in
the scientiWc community. ‘Truth resides in agreement: social consensus does
not merely evidentiate truth, but is its creator’ (Westphal 1998: 3). We
pointed out however, that social science communities do not reach consensus
based simply on opinions; they rely on standards of sound and persuasive
arguments and empirical evidence for a scientiWc claim. The persuasiveness of
an argument is a rhetorical question. Thus, in addition to logos, the other
angles of pathos and ethos of the rhetorical triangle are important consider-
ations (too often ignored) in communicating scientiWc Wndings.

Van, D. V. A. H. (2007). Engaged scholarship : A guide for organizational and social research. Oxford University Press, Incorporated.
Created from uu on 2024-11-20 05:12:33.
66 ENGAGED SCHOLARSHIP

MODELS AS MEDIATORS
A key criticism of positivism was its syntactic view of theory (consisting of
axiomatic Wrst-order logical relations among theoretical terms, and corres-
pondence rules that gave theoretical terms meaning in terms of their observa-
tional consequences). Giere (1999), Suppe (1989), and others replaced this
syntactic view with a semantic view of theories in which models (rather than
correspondence rules) provide the interpretation of social theories (Morrison
and Morgan 1999: 5). This criticism provides a key reason for including model
development (in research design) as a core activity in the engaged scholarship
process. The semantic view claims that models stand in a mediating relation-
ship between theories and data. McKelvey (2002a) emphasizes that model-
centeredness is a key element of scientiWc realism. He quotes Cartwright as
saying, ‘The root from theory to reality is from theory to model, and then from
model to phenomenological wall’ (Cartwright 1983: 4). Like Morrison and
Morgan, McKelvey views models as autonomous mediators between theory
and phenomena.
Models are viewed as being fallibilistic and perspectival. Because data
are theory-laden and error-prone, the challenge is to compare plausible
alternative models given our current understanding of the subject matter
instead of searching for an ultimate truth. As Giere (1999) explains, models
represent alternative claims about a phenomenon in question given current
understandings of it, rather than a universal objective theory of the world.
[Science] pits one model, or family of models, against rival models, with no pre-
sumption that the whole set of models considered exhausts the logical possibilities.
This means that what models are taken best to represent the world at any given time
depends on what rival models were considered along the way. And this seems,
historically, a contingent matter. So the models of the world held at any given
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

time might have been diVerent if historical contingencies had been diVerent. (Giere
1999: 77)

Azevedo (2002) provides a pragmatic extension of using models for scien-

tiWc problem solving. A scientiWc theory is operationalized as a model that is
mapped onto reality (the problem). The test of a model is practical: how well
does it serve as a map to guide action. Because the process of making and
using maps is easily understood, the use of a mapping model of knowledge
provides a powerful heuristic for determining the validity of scientiWc theor-
ies. Azevedo (2002: 725) points out that maps and models are constructed
with interests in mind. They are selective representations of the world, and
their content and format are selected according to their relevance to the
problems they are intended to solve. Because the usefulness of a map model

can only be assessed by how well it helps to solve the problem of the user, its
validity is interest-related as well.

RELEVANCE
Users of research knowledge—both scientiWc and practical—demand that it
overcome the dual hurdles of being relevant and rigorous in serving their
particular domains and interests (Pettigrew 2001). However, diVerent criteria
of relevance and rigor apply to diVerent studies because their purposes,
processes, and contexts are diVerent. Pragmatists (particularly James and
Dewey) emphasized that the relevance of knowledge should be judged in
terms of how well it addresses the problematic situation or issue for which it
was intended. Rescher (2000: 105) maintains that the relevance of knowledge
about a problematic situation being investigated may entail any (or all) of the
following questions:
. Description (answering what? and how? questions about the problematic
situation);
. Explanation (addressing why? questions about the problematic situation);
. Prediction (setting and achieving expectations about the problematic
situation);
. Control (eVective intervention in the problematic situation); and
. Emancipation (identifying the marginalized and repressed).
One criterion of research eVectiveness does not Wt all. Pragmatists have
emphasized that diVerent criteria of relevance and rigor apply to research
undertaken to examine these diVerent kinds of questions.
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

ENGAGEMENT
A fundamental tenet of critical realism is that a real world exists out there, but
our abilities to comprehend it are very limited. The ambiguous, ‘buzzing,
blooming, confusing’ nature of reality exceeds the explanatory capabilities of
any single theory or model that a researcher might devise. ‘In the absence of
unambiguous foundational truth in the social sciences, the only sensible way
forward can be conscious pluralism’ (Pettigrew 2001: S62). As discussed in
Chapter 1, pluralism requires engaging others from diVerent disciplines and
functions who can contribute diVerent perspectives and models for under-
standing the problem domain being examined. Engagement not only requires

Van, D. V. A. H. (2007). Engaged scholarship : A guide for organizational and social research. Oxford University Press, Incorporated.
Created from uu on 2024-11-20 05:12:33.
68 ENGAGED SCHOLARSHIP

a diVerent conception of the researcher’s role, but also an extension of the

philosophers’ consensus theory of truth.
Hermeneutics and relativism provide useful guidelines for engagement.
Perhaps most fundamental is for researchers to jettison their God’s Eye view
(illustrated in Figure 2.2) and adopt a participant frame of reference (shown
in Figure 2.3) to conduct their studies. In a participant role, a researcher
listens to and learns from others who have diVerent perspectives that merit
consideration for modeling or mapping a problem domain existing in the
world. Moreover, relativism stresses the salience of divergent and often con-
Xicting interests, values, and power of stakeholders in any study, and the
impossibility of serving them all. One clear implication is the need for
researchers to be reXexive in clarifying whose interests and values are served
in their research engagements.
Engaging people from diverse backgrounds and perspectives represents a
method of triangulating on a complex problem. Triangulation is the use of
multiple sources of information, models, and methods in a study. Research
knowledge advances by comparing the relative contributions and perspectives
provided by diVerent models. Azevedo (1997) discusses how the coordination
of multiple models and perspectives may reveal the robust features of reality
by identifying those features that appear invariant (or convergent) across
diVerent perspectives. Azevedo reXects the established view in philosophy of
science of developing reliable scientiWc knowledge by identifying those per-
spectives from a pluralist approach that converge on a common or consensual
view of the phenomenon.
But the engagement of diVerent stakeholders in a study often produces
inconsistent and contradictory perspectives of a problem domain being
examined. Pluralistic perspectives should not be dismissed as noise, error,
or outliers—as they are typically treated in a triangulation research strategy.
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

Chapter 9 discusses how these inconsistent and contradictory Wndings require

an expansion of a consensus theory of truth that emphasizes convergence
and agreement among investigators and reviewers in a scientiWc community
on reliable and replicable Wndings. But many real-world problems contain
inconsistent and contradictory principles. Rendering such problems as
incommensurable denies their reality. Inconsistent and contradictory Wndings
from diVerent stakeholders involved in a study often represent truly plural-
istic views of a problem domain that might be explained through methods of
arbitrage (linking divergent bits of information and views) and paradoxical
reasoning (to reconcile opposing and contradictory Wndings). As Suppe
(1977) insightfully asked of Kuhn’s claims of incommensurability, if pluralis-
tic perspectives are truly incommensurable, how is it possible that scholars
(and practitioners) can compare diVerent paradigms and communicate con-
structively across the paradigms? One implication of the practice of engaged
scholarship is that philosophers of science need to expand their traditional

explanations that emphasize convergent central tendencies to include explan-

ations based on inconsistent Wndings through arbitrage and contradictory
Wndings with methods of paradoxical reasoning.

THE SOCIAL PROCESS OF SOCIAL SCIENCE

Another important implication of the practice of engaged scholarship for
philosophy of science is the social process of conducting research. Positivists
might be excused for their admission of a separation between a scientiWc
theory to be tested and the social, psychological, and economic processes in
which such a theory might have developed. But an excuse for such a separation
between the ‘logic of discovery’ and the ‘logic of testing’ a theory should not
apply to relativism, pragmatism, and realism. Despite the calls for assessing
research with intended actions of the users of research by pragmatists, for a
pluralistic comparison of alternative models of a problem domain being
investigated by realists, and for a social constructions of the meanings of
reality from diVerent stakeholders by relativists, it is striking how little
attention philosophers from these diVerent schools of thought have given to
the social process in which these perspectives might be realized.
Perhaps philosophers of science have deferred this question to sociologists
of science. Studies of working scientists by GarWnkel et al. (1981), Knorr-
Cetina and Amann (1990), and Latour and Woolgar (1986) indicate that
improvisation underlies the process in which scientists actually construct
models, enact experimental runs, design and interpret data, report on their
methods and Wndings, and assign credit for discovery. While such studies are
useful descriptions of how scientists engage in their practices, they provide
little guidance for action, except for the conclusion that scientists engage in
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

what Levi-Strauss (1966) termed a bricolage, improvising with a mixed bag of

tools and tacit knowledge to adapt to the task at hand. The process by which
scholars might step outside of themselves and engage others to be informed
of the problem domain being examined remains a black box. Subsequent
chapters explore possible ways to open this black box by suggesting means
and ways to engage relevant stakeholders in problem formulation, theory
building, research design, and problem solving.

Conclusion
In conclusion, the purpose of the historical review of key concepts and
principles of positivism, relativism, pragmatism, and realism has been to
identify some of the conceptual tools and frameworks to understand diVerent

Van, D. V. A. H. (2007). Engaged scholarship : A guide for organizational and social research. Oxford University Press, Incorporated.
Created from uu on 2024-11-20 05:12:33.
70 ENGAGED SCHOLARSHIP

views of science, to initiate a process of reXexivity in choosing a philosophy of

science that suits your scholarly practice, and to gain a deeper understanding
of the philosophical basis of engaged scholarship. We discussed how these
philosophies of science have inXuenced our views of engaged scholarship, and
also indicated several areas where the practice of engaged scholarship might
advance or extend philosophy of science. As stated at the beginning of this
chapter and applied in subsequent chapters, we view engaged scholarship as
based on the following key elements of a critical realist philosophy of science:
. There is a real world out there (consisting of material, mental, and emer-
gent products), but our individual understanding of it is limited. In
general, physical material things are easier to understand than reXexive
and emergent social processes.
. All facts, observations and data are theory-laden implicitly or explicitly.
Social sciences have no absolute, universal, error-free truths, or laws as any
scientiWc knowledge.
. No form of inquiry can be value-free and impartial; each is value-full.
Some methods are better warranted than others depending on the phe-
nomenon.
. Knowing a complex reality demands use of multiple perspectives.
. Robust knowledge is a product of theoretical and methodological triangu-
lation where evidence is not necessarily convergent but might also be
inconsistent or even contradictory.
. Models that better Wt the problems they are intended to solve are selected
allowing an evolutionary growth of knowledge.
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

Van, D. V. A. H. (2007). Engaged scholarship : A guide for organizational and social research. Oxford University Press, Incorporated.
Created from uu on 2024-11-20 05:12:33.
3 Formulating the Research
Problem
Model

Th
sig

eo
De

ry
ch

Bu
ar

ild
se

in
Re

g
Solution Theory

oni
at
Pr

ul
ob

rm
le
m

Fo
So

m
lv

le
in

ob
g

Reality

The formulation of a problem is often more essential than its solution,

which may be merely a matter of mathematical or experimental skill.
(Albert Einstein quoted in Getzels and Csikszentminhalyi 1975)

It is exceedingly diYcult to say something meaningful about the real

world without starting in the real world. Observation and description
of the real world are the essential points of origin for theories in
applied areas.
(Robert Dubin 1976: 18)
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

Any scientist of any age who wants to make important discoveries

must study important problems. Dull or piZing problems yield dull
or piZing answers. It is not enough that a problem should be
interesting—almost any problem is interesting if it is studied in
suYcient depth . . . the problem must be such that it matters what
the answer is—whether to science generally or to mankind.
(P. B. Medawar, Nobel Laureate in Medicine and Physiology, 1979)
Problem formulation is often the Wrst—and most important—task of the
engaged scholarship process. Problem formulation plays a crucial role in
grounding the subject or problem in reality, and directly aVects how theory
building, research design, and problem solving tasks are performed. Yet,
researchers often overlook or pay little attention to problem formulation.
Witness, for example, the glib problem statements in the introduction of

Van, D. V. A. H. (2007). Engaged scholarship : A guide for organizational and social research. Oxford University Press, Incorporated.
Created from uu on 2024-11-20 05:12:44.
72 ENGAGED SCHOLARSHIP

most research articles in social science journals. Like other human beings,
researchers tend not to be ‘problem-minded,’ and prefer instead to be ‘solu-
tion-driven’ by focusing on developing and testing models or theories for
problems that often remain unclear.
Research is often viewed as a problem solving activity (e.g., Deutsch 1997;
Azevedo 2002). ScientiWc theories are constructed and evaluated with speciWc
interests in mind, and research Wndings are used to inform decisions about
theory and practice. Viewed as a problem solving process, science aims to
increase our understanding of complex problems or phenomena that exist
under conditions of uncertainty found in the world. This process typically
involves steps in recognizing a problem, searching and screening information,
evaluating alternatives, and choosing a solution (e.g., Polya 1957; March and
Simon 1958; Halpern 1996; Deutsch 1997; McGrath and Tschan 2004). In
terms of formulating a research problem, this process translates into four
interrelated activities: (1) recognizing and situating a problem; (2) gathering
information to ground the problem and its setting; (3) diagnosing the
information to ascertain the characteristics or symptoms of the problem;
and (4) deciding what actions or questions to pursue to resolve the research
problem.
This chapter examines the process of problem formulation in terms of these
four activities of situating, grounding, diagnosing, and resolving a research
problem. These four activities overlap and are highly interdependent. Except
for highly simpliWed and stylized problems, problem formulation activities
seldom unfold in an orderly rational progression over time. Thus, instead of
portraying these activities as unfolding in a Wxed linear sequence, I view them
as a set of parallel activities that researchers undertake simultaneously in
varying degrees throughout the problem formulation process. As discussed
in this chapter, the relative eVort devoted to each of these activities depends on
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

how the problem is perceived, the context in which it exists in the world, and
the goals of the study.
Research might be undertaken to examine a wide variety of possible
problems. Some research projects might be undertaken to address a particular
practical problem, crisis or threat, such as a gap between expectations and
outcomes experienced by practitioners in implementing an organizational
program, product, or service. If these problems are clearly structured or
understood, the problem formulation process may reXect an orderly sequence
of activities in situating, grounding, diagnosing, and resolving the problem.
Most research problems are not so well-structured. They often represent
anomalies or breakdowns that a scholar encounters in the literature or in
practice that are not consistent with the scholar’s theory of the world. Scholars
often observe something that their theory did not lead them to expect—
resulting in a breakdown or anomaly. Anomalies represent diagnostic puzzles
that trigger recognition that ‘There’s something else going on here.’

Van, D. V. A. H. (2007). Engaged scholarship : A guide for organizational and social research. Oxford University Press, Incorporated.
Created from uu on 2024-11-20 05:12:44.
FORMULATING THE RESEARCH PROBLEM 73

Finally, the problem motivating a study may deal more generally with
exploring an unclear issue or phenomenon in order to Wnd out about reality
with no speciWc end in mind. In other words, the researcher may have only
vague impressions of how to situate, ground, diagnose, or resolve a problem.
As Abbott (2004: 83) states,
We often don’t see ahead of time exactly what the problem is; much less do we have an
idea of the solution. We often come at an issue with only a gut feeling that there is
something interesting about it. We often don’t know even what an answer ought to
look like. Indeed, Wguring out what the puzzle really is, and what the answer ought
to look like often happen in parallel with Wnding the answer itself.
As these illustrations suggest, a research problem is deWned as any problem-
atic situation, phenomenon, issue, or topic that is chosen as the subject of an
investigation. The problematic situation may originate in either the practical
world of aVairs, a theoretical discipline, or a personal experience or insight.
It may be perceived to represent an unsatisfying circumstance, a promising
opportunity, a breakdown or anomaly in expected arrangements, or simply a
topic of interest. However one construes the problematic situation, researchers
tend to encounter four common diYculties in situating, grounding, diagnos-
ing, and resolving a research problem.
First, a key challenge in situating a problem is deciding what persons or
stakeholder groups will be served by the research, and to describe reality from
the perspectives of those persons or stakeholders. Implicitly or explicitly, all
research is undertaken in service of someone—whether it be the researcher, a
funding agency, practitioners, academics, a profession, or any of the above. The
point is that problems do not exist objectively ‘out there;’ they are uniquely
perceived and framed by diVerent people. Knowing from whose perspective a
problem is being addressed and engaging them in problem formulation is
necessary to frame the focus, level, and scope of a research study.
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

Second, researchers—like other human beings—have limited capabilities

in handling complexity. They often use short-cuts or heuristics that produce
biased judgments. Needless to say, solving the ‘wrong’ problem with the ‘right’
methods, or what J. Tuckey referred to as a Type III error, is costly, demoralizing,
and all-too-familiar (Volkema 1995; Buyukdamgaci 2003). Unfortunately,
problem formulation is often rushed or taken for granted. As a result, import-
ant dimensions of a problem often go unrecognized and opportunities to
advance knowledge of the problem are missed (Volkema 1983).
A third diYculty is that the issues that motivate a study are sometimes stated
as imaginary pseudo-problems that lack grounding in reality. Too many social
science studies suVer from elaborating theories that are often based on an
insuYcient diagnosis of the problem and its context. As a consequence, theory
and research tend to be grounded in myths and superstitions. Those who
generalize from concrete experiences or particulars with a problem can answer

Van, D. V. A. H. (2007). Engaged scholarship : A guide for organizational and social research. Oxford University Press, Incorporated.
Created from uu on 2024-11-20 05:12:44.
74 ENGAGED SCHOLARSHIP

the questions, For example? From whose point of view? What is that point of
view? Engaging people who experience and know the problem is necessary to
answer these questions. Lacking answers to these questions often leads to
unfounded generalizations.
Merton points out that oftentimes in science as in everyday life, ‘explan-
ations are provided of matters that are not and never were’ (Merton 1987: 21).
In legal proceedings, establishing the case is mandatory for pursuing it.
Merton (1987) cautioned that an important Wrst element in the practice of
science is ‘establishing the phenomenon.’ Evidence and arguments should
clearly indicate that the phenomenon is enough of a regularity to require
and allow explanation. In this way ‘pseudo facts that induce pseudo problems
are avoided’ (Hernes 1989: 125).
A fourth diYculty is that even when problems are grounded in reality,
their diagnosis or resolution may not lead to creative theory that advances
understanding of the phenomenon or problem. Bruner (1973) points out that
a theory or model is a generic representation of the critical characteristics of
a phenomenon. For Bruner, grounding theories in reality requires going
beyond the information given so that the problem is formulated to have
applicability beyond the situation in which it is observed.
This chapter explores ways of dealing with these four common diYculties
in situating, grounding, and diagnosing a research problem, and forming a
question to study a problem domain. The next four sections discuss each of
these key activities in problem formulation. A central theme is the close
interplay between theory and reality when formulating a research problem.
Abbott reXects this interplay in stating, ‘Often one builds out from the problem
on the one hand, and from the solution on the other, until the two halves meet
in the middle like a bridge built from two banks’ (Abbott 2004: 81).
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

Situating the Problem

All problems, anomalies, or issues motivating a study begin with a perception
that something requires attention. Problems are not given by nature, but by
how, whom, and why they are perceived. No one can possibly represent all
aspects and viewpoints of a problem domain. That being the case, any formu-
lation of a problem is a partial representation reXecting the perspective and
interests of the observer. Problems do not exist ‘out there’ in an objective state
of nature. People enact reality and its problems (Weick 1979). Reality is socially
constructed (Berger and Luckmann 1966). No one studies the social world as it
is, instead, we study reality as it appears to us. A scholar must, therefore, be
reXexive and clarify whose point of view and interests are served in a problem
or model proposed to represent reality (Van Maanen 1995).

DiVerent people interpret situations in diVerent ways. People who interact

and socialize together develop shared beliefs about the nature of things and
relationships between them, and shared norms about what they should and
should not do. Over time these beliefs may become so institutionalized that
they are taken to be ‘matters of fact.’ Our individual histories are unique,
however. DiVerent people interpret situations in diVerent ways because they
bring to a situation their own particular mental ‘framework’ of personal
beliefs, attitudes, prejudices, and expectations to make sense of a situation.
As a result, Eden et al. (1983: 2–3) point out that people pay attention to
certain things, selecting those having a particular signiWcance for themselves,
and ignoring the rest.
The ‘model’ of man we have . . . is thus not of an organism responding to some
‘stimulus,’ nor ‘driven’ by internal needs of instincts, nor of a person whose thinking
and actions are socially ‘given.’ Rather, it is a human being who acts in the light of the
personal interpretations or constructions he (or she) places upon events. (Eden
et al. 1983: 3)
One implication of this perspective is that it is impossible to assume, self-evidently
and non-problematically, that the way other people interpret a situation, is the same
as, or even similar to, the way we interpret that ‘same’ situation. An event which you
or I might see as a major crisis for a particular reason may be seen as a major crisis by
someone else for completely diVerent reasons, by another person as a minor diYculty,
and yet other people may not even have noticed it at all. No situation is inherently,
‘objectively’ a problem. A problem belongs to a person; it is an often complicated, and
always personal construction that an individual (or like-minded group of people)
places on events. (Eden et al. 1983: 8)
When and how a problem is situated largely determines how it is ap-
proached and solved. For example, labeling a situation a ‘human resources’
problem means that it will be approached diVerently than if it is viewed as an
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

‘organization design’ problem or a ‘market share’ problem. A problem’s

deWnition largely determines its solution space. That being the case, when
situating the problem we need to be reXexive about whose perspectives will
take foreground and background in situating the problem domain. The
following dimensions are useful in situating the focal area, level, and scope
of the problem domain.

FOCUS AND TIMESPAN

A problem can be viewed as having a foreground and a background, a focal
area, and a context (Abbott 2004: 138). For example, studies of businesses
often put managers in the foreground and public policy makers or other
stakeholders in the background. In this case the focus would be on the

Van, D. V. A. H. (2007). Engaged scholarship : A guide for organizational and social research. Oxford University Press, Incorporated.
Created from uu on 2024-11-20 05:12:44.
76 ENGAGED SCHOLARSHIP

problems and situations experienced by managers. The concerns of employees

of businesses and other stakeholders would be treated as part of the back-
ground or context of the manager’s problem domain. So also, local politics,
external regulations, and infrastructure in the communities where the busi-
ness is located would be treated as context in the problem domain. Selecting
who is in the foreground and background signiWcantly inXuences how the
research problem is formulated.
Some people may view a given problem in fairly narrow, clear, and static
terms experienced by an isolated group of people or institutions at a certain
point in time. Others might view the same problem domain as being a general,
unclear, and dynamic process diVused among many diVerent groups over
long periods of time. These diVering perceptions of a problem domain often
reXect the diVering interests and roles of users. Management practitioners,
for example, tend to focus on the immediate and particular problems they
are experiencing in running their organizations. BeWtting their roles, policy
makers tend to be concerned with more general views of problems aVecting a
larger population of citizens or practitioners.
Ultimately, how a problem domain is speciWed depends on who is chosen
as the users or audience of the research. Most problems are too complex to
capture the diVering perspectives of all relevant stakeholders. We have no
choice but to cut down on the complexity of the problem domain by putting
some things and people in the foreground and others in the background.

LEVEL
A problem has a ‘level,’ in the sense that it may be experienced or noticed at
individual, group, organization, industry, or broader levels of analysis. In
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

addition, the factors or events that are thought to contribute to, or be the
consequence of, the problem may exist at diVerent levels of analysis. Thus, as
Abbott (2004: 138) states, some things are bigger than our focal problem,
some are part of the problem (and possibly determine it), and some things
that are smaller than it. The choice of level of analysis not only reXects the
nature of the problem, but also the disciplinary base that is used to structure
or model the problem. For example, psychologists tend to structure their
research problems at the individual level, while sociologists tend to view
problems from more macro institutional and community perspectives.
Closely related to selecting a level of analysis is the context of the problem
domain, which typically includes characteristics broader than or outside of the
level at which a problem is examined. For example, the context for studying
individual work behavior may include the group, organization, and industry
in which the person works. If the level of analysis is expanded to the organ-
ization, then many of the group and organization-level characteristics that

were before viewed as context are now folded into the problem domain, while
more aggregate industry-level factors continue to be treated as part of the
environmental context. Of course, choosing the context of a problem domain
entails more than just selecting a level of analysis. At a given level, it involves
rearranging what things to focus on in the foreground and background and
what things to exclude or place outside of your purview.

SCOPE
How deep, how broad, and how long a problem should be studied are never-
ending questions of problem scope. Ultimately, the answer is that you study a
problem until it satisWes the curiosities and needs of those engaged in the study.
Ideally, the scope of a problem should decrease and become more man-
ageable as you become familiar with it. In practice the opposite often occurs
with ‘scope creep’ where the problem becomes expansive and includes more
complex domains as you study a problem. A variety of factors contribute to
scope creep. First, it is a constructive sign of learning that the problem of
interest may be much larger than initially anticipated, or that it plays into a
much larger problem. Research advisors and others engaged in a study may
also be learning about the problem, and may suggest a study of related issues
and questions.
Another factor that contributes to scope creep is the endless nature of
problems in reality. As Rescher (1996: 131–2) states, ‘Real things are cogni-
tively opaque—we cannot see to the bottom of them. Our knowledge of such
things can thus become more extensive without thereby becoming more
complete.’ He says that in being real, one can gain new information, which
may add to or revise what one has learned earlier. A ‘real thing’ has features
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

that may exist outside of our cognitive reach. ‘As a consequence, our know-
ledge of fact is always in Xux. It is not a thing, but an ever-changing and ever
growing manifold of processes’ (Rescher 1996: 132).
The focus, level, and scope of a problem domain are often unclear when
research begins. Familiarity emerges over time by engaging relevant stake-
holders in grounding and diagnosing information about the problem.

Grounding the Problem in Reality

Situating a problem domain and gathering information about it often repre-
sent two initial overlapping steps in problem formulation. The more you can
ground a research problem in reality from a user’s perspective the more you
learn to appreciate the multiple dimensions and manifestations of a problem

Van, D. V. A. H. (2007). Engaged scholarship : A guide for organizational and social research. Oxford University Press, Incorporated.
Created from uu on 2024-11-20 05:12:44.
78 ENGAGED SCHOLARSHIP

and its solution space. Grounding a problem in reality entails an exploratory

study into the nature, context, and what is known about the problem domain.
A variety of methods are useful for undertaking this exploration. As discussed
later, they include information-gathering activities drawing on personal experi-
ences and direct observations of how a problematic situation unfolds in a
particular context, as well as talking with people who experience the problem
through casual conversations, interviews, or in group meetings. Reviewing
the literature to determine the scope, prevalence, and context of the problem
is also needed.
The purpose of these activities is to become suYciently familiar with a
problem domain to be able to answer the journalist’s basic questions of who,
what, where, when, why, and how the problem exists. Grounding a problem
requires both particular and general answers to these questions. Particular
answers provide up-close and personal descriptions of the problem based on
Wrst-hand observations of a speciWc case or two. Particular answers provide
concrete and vivid details of a speciWc problem. General answers are needed to
show that the particular case is not unique; instead it is an instance of a much
larger or pervasive problem. Typically the general answers are based on
indirect statistical evidence obtained from literature reviews of prior research
on the problem.
Daily examples of using both particular and general descriptions of prob-
lems are found in the introductory paragraphs of feature stories in many
newspapers. Usually, the Wrst paragraph in the stories provides a particular
up-close and personal answer to the journalist’s questions. A typical format is
the following:
Each week for the past six months (when), Joe Blow, a 45-year-old machinist (who)
has been seeing his psychiatrist for moods of depression (what) that have become
worse since he was laid oV from his job (why) that he held for 20 years at AMC
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

Engineering located in this industrial Midwest town (where).

The second paragraph provides general answers to the journalist’s questions
about the pervasiveness of the problem. It might read as follows:
Joe Blow is not alone. A study by University researchers (Wanberg et al. 2005)
reported that there were 8349 mass layoVs in 2001 (when) in the US (where), which
led to 1.7 million individuals losing their jobs (who). Researchers are Wnding that job
loss has a negative inXuence on most every indicator of mental and physical health.
For example, studies demonstrate that job loss is associated with increased anxiety,
depression, sleeping problems, alcohol disorders, divorce, and child abuse (e.g.,
Dooley et al. 1996) (why). Joe Blow and millions like him (who) are posing a major
question of what to do about the problems associated with job lay oVs?
At this point the story might go in several possible directions, depending on
the writer’s perspective. If the writer takes a human resource development
perspective, the story might focus on what training and counseling services

are being provided by companies to help individuals, like Joe Blow, make
the transition and Wnd a job. Alternatively, the writer may reXect a public
policy concern, and question how government might curb corporate human-
resource abuses of layoVs and their resulting pain and health care costs on
former employees and society. In addition to illustrating the journalist’s
questions, this example illustrates a point discussed later that many diVerent
stories or diagnoses of a problem are often grounded in the same data or
observations of reality.
Answering the journalist’s questions provides useful criteria for grounding
a research problem by obtaining particular and general answers to who, what,
where, when, why, and how the problem exists. When beginning a study
researchers are seldom suYciently familiar with a problem domain to be
able to answer these questions in particular and in general. And if they think
they know the answers to these questions, then it is important to determine
who may answer the questions in similar and diVerent ways and why.
Problem formulation is not a solitary exercise; instead it is a collective
achievement. Grounding a problem requires the researcher to step outside of
him/herself, and to be open to and informed by the interpretations of others
about the problem domain. As Bruner (1986: 133) states, ‘ReXection and
‘‘distancing’’ are crucial aspects of achieving a sense of the range of possible
stances—a metacognitive step of huge import.’ Most problems tend to exist in
a ‘buzzing, blooming, and confusing’ reality. The world is too rich and multi-
layered to be captured adequately by any single person. Therefore, a pluralist
approach to problem formulation is essential. It is only by obtaining and
coordinating perspectives of other key stakeholders that robust features
of reality can be distinguished from those features that are merely a function
of one perspective (Azevedo 1997: 189–90).
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

Diagnosing the Problem

Grounding a problem domain through careful observation and data collection
provides the raw materials for diagnosing a problem. Diagnosis entails a
disciplined, yet open-minded, application of models or theories in order to
ascertain the speciWc nature of the problem in context. Becoming familiar with
a phenomenon existing in reality supplies multiple opportunities to diagnose
expected and unexpected things. Expected things are those that conform to our
model of reality; we view them as well-structured problems for which known
solutions or interventions are believed to exist. Unexpected things are those
that do not conform to our model of reality. They represent anomalies or
breakdowns that lead us to recognize that we have an ill-structured under-
standing of the problem domain. Further study may indicate that existing

Van, D. V. A. H. (2007). Engaged scholarship : A guide for organizational and social research. Oxford University Press, Incorporated.
Created from uu on 2024-11-20 05:12:44.
80 ENGAGED SCHOLARSHIP

solutions to the anomaly are inadequate; a new solution needs to be created or

discovered.1
Breakdowns are instances when expectations are not met, something does
not make sense, or when one’s assumption of coherence is violated (Agar 1986:
20). Breakdowns play a central role in bringing problems to our attention.
Some anthropologists advise researchers to ‘use surprise, the unexpected, or a
sense of diVerence as cues to what to study’ (Rosenblatt 1981: 200). Alvesson
(2004) suggests that a really interesting breakdown means that an empirical
‘Wnding’ cannot easily be accounted for by available theory. The breakdown is,
thus, not an outcome of our ignorance, naivety or narrow-mindedness. The
surprise should be the reaction likely to be experienced by other members of
the research community who are supposed to be able to understand/explain
the empirical observation/construction triggering the breakdown.
When some anomaly is perceived in a given context, our repertoire of
conceptual models or perspectives limits the range of possible explanations
we might develop to appreciate the phenomena in an intelligible way. For
example, I enjoy watching birds alight on the bird feeder outside of my
kitchen window. However, I am not an ornithologist or experienced bird
watcher. As a result I am not likely to recognize or discover if a bird of a new
species might land on my bird feeder. As Louis Pasteur said, chance of
discovery favors the prepared mind. A prepared mind is not only familiar
with particulars of the problem in context, but also has a repertoire of
plausible alternative theories or models for representing and explaining the
problem. Each theory provides an expected scenario of reality.
A repertoire is important for discovery. Serendipity as described by Merton (1973),
and the recurrent theme of ‘chance’ discovery . . . implies that the scientist has an
available agenda of problems, hypotheses, or expectations much larger than the
speciWc problem on which he works, and that he is in some sense continually scanning
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

or winnowing outcomes, particularly unexpected ones, with this larger set of sieves.
(Campbell 1988: 418)
Familiarity with a problem domain increases the likelihood of identifying
deviations from normality that merit attention as being important or novel.
The likelihood of discovering new explanations for these deviations is largely
dependent upon our repertoire of alternative theories. Each theory can serve
as an alternative thought trial or conceptual experiment with the phenomena.

1 Breakdowns can occur at any time in the research process—from the initial period of scouting
around to determine ‘what’s-going-on-here’ to the writing-up phase when particular theories are
applied to particular sorts of evidence. Van Maanen (1995) suggests that breakdowns and surprises are
often retrospective matters. This does not make them any less valuable, but it does limit the extent to
which we can ‘know’ our preconceptions at the outset of a study (without experience there is nothing
to startle us to recognize them) or our theories (without trying them out on our materials).

Diagnosing problems as being expected or unexpected (breakdowns) lead

to diVerent, yet related patterns of problem diagnosis. The former entails
solving a problem by selecting from a set of pre-enumerated solutions, while
the latter leads to constructing a new solution (Clancey 1985). Research on
problems with known solutions tends to represent theory testing or evalu-
ation research projects. In contrast, research on problems with unknown
solutions is more challenging because it requires the construction or discovery
of new theory. These two related patterns of problem diagnosis are now
discussed in greater detail.
Diagnosis is a process of classifying observations of a phenomenon into
known categories that are amenable to problem solving. In simple classiWca-
tion, data may directly match solution features or may match features after
being abstracted. For example, to identify what kind of bird I see alighting on
my bird feeder, I look at the pictures and features of birds in my guidebook to
identify the kind it resembles. The essential feature of this simple classiWcation
is that I select from a set of pre-enumerated solutions. This does not mean, of
course, that I have the ‘right answer;’ it is just that I have only attempted to
match the data against known solutions, rather than construct a new solution.
I may have made errors in observing the bird or in matching its features with
the most similar one in my guidebook, so my conclusion is a hypothesis (that
I will probably not bother to test, given the situation).
As this example illustrates, in simple classiWcation, data elements tend to
directly match solution features. For more typical ill-structured problems, the
process of diagnosis is more complex. Simon (1973: 181) examined these
kinds of ill-structured problems, ‘deWned as a problem whose structure lacks
deWnition in some respect.’ He proposed a strategy for classifying complex
systems of ill-structured problems into well-structured problems at micro or
modular levels of system architectures. Simon observed that the tricks
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

that have worked in relatively well-structured domains are often extended

to ill-structured domains.
He used the example of an architect, who from long term memory and
experience knows the basic end states of designing a house subject to client
constraints. The design of a house acquires structure by being decomposed
into various problems of component design, which converts an ill-structured
problem into well-structured sub-problems. Problems are well-structured in
small components, but ill-structured overall. Expert specialists (or subcon-
tractors) can be consulted to perform the component tasks in a well-structured
manner. This may create problems of neglecting interrelations among various
well-structured sub-problems. The danger of such inconsistencies is mitigated
by a reliance on the architect’s long-term memory or design blueprints; certain
ways of dividing the whole task into parts will do less harm to interactions than
other ways of dividing the task.

Van, D. V. A. H. (2007). Engaged scholarship : A guide for organizational and social research. Oxford University Press, Incorporated.
Created from uu on 2024-11-20 05:12:44.
82 ENGAGED SCHOLARSHIP

Heuristic Match
Data Abstractions ⇔ Solution Abstractions

Data Refinement
Abstraction

Data Solution

Figure 3.1. Inference structure of problem diagnosis

Source: Clancey (1985: 296).

An initial stage of laying down general (and tentative) speciWcations is followed by

stages in which experts are called up (‘evoked’) to introduce new design criteria
and component designs to satisfy them. At a later stage, there is attention to inconsist-
encies of the component designs, and a search for modiWcations that will continue to
meet most of the criteria, or decisions to sacriWce certain criteria in favor of others. Each
small phase of the activity appears to be quite well-structured, but the overall process
meets none of the criteria set down for well-structured problems. (Simon 1973: 194)
Simon notes that when the problem space remains unchanged during
problem solving, assimilating new information is not an issue. But when
the problem space is subject to change and unanticipated events emerge
during the process, then it is necessary to consider how solution models are
modiWed or reconstructed as the problem diagnosis process unfolds.
In practice, diagnosis tends to be solution-driven. Its steps involve
classifying data into problematic symptoms or categories, aggregating the
classiWcations to infer the problem, heuristically selecting a solution that is
known to be appropriate for the problem, and then reWning its application to
the case at hand. Clancey (1985) illustrates and describes the general structure
of inference in diagnosing problems with Figure 3.1.
Problems tend to start with objects in the real world, so it makes sense that
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

practical problem solving knowledge tends to associate problems with types

of objects—people, patients, products, programs, or organizations. For example,
in medical diagnosis, basic observations about the patient are classiWed into
symptoms and abstracted to patient categories, which are heuristically linked to
disease categories, and then reWned to prescribe a treatment for the disease of a
particular patient. The steps in this diagnostic process include: (1) data classiWca-
tion and aggregation; (2) heuristic matching of a problem and a solution; and (3)
reWning the solution to Wt the case. Each of these steps are now discussed.

DATA CLASSIFICATION AND AGGREGATION

Classifying observations into conceptual categories is perhaps the most
central and important inference that is made in problem formulation (as

well as theory building, research design, data analysis, and problem solving
activities discussed in later chapters). In the medical example, classifying
patient data into symptoms and aggregating the symptoms into a patient
disease category involves a reasoning process of deWning and generalizing data
elements from subclasses to higher abstractions of classes. The very process of
classifying terms and aggregating them into more general and abstract con-
cepts changes what we know and how we view a phenomenon. As Hanson
discussed, researchers do not merely inspect the world and receive data about
problems, we interact with the world and interpret the data in ways that Wt
our understanding of the world.
The language in which we speak and think and the circumstances in which we Wnd
ourselves speaking and thinking in that language contribute to the formation and
constitution of what and how we think and hence, what and how we actually perceive.
This is not to say that our language produces what we think about, or produces what
we perceive, anymore than the plaster mold produces the bronze statue, or the recipe
produces the cake. It is rather just to suggest that perhaps the form of language
exercises some formative control over our thinking and over our perceiving, and over
what we are inclined to state as the facts (and indeed how we state those facts). What
we call ‘facts’ are almost always stated in that clauses, that some linguistic element we
encounter in seeing when we consider Seeing That. (Hanson 1969: 184)
In deWning terms, seldom do problem solvers follow the Aristotelian notion
of concept deWnition in terms of its necessary constitutive properties as dis-
cussed in the next chapter. Instead, general schema are used that include
incidental and typical manifestations or prototypes of behaviors. The deWni-
tional links are often non-essential, ‘soft’ descriptions. The meanings of concepts
depend on what we ascribe to the links that join them. Thus, in practice we
jockey around concepts to get a coherent network. Complicating this is a
tendency to use terms that confound causes, eVects, and indicators of concepts
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

without understanding the links in a principled way. For example, deWnitions of

problems in organizations are typically based on deviations from normal pat-
terns. But the meaning of ‘normal’ depends on everything else happening to the
organization, so inferring a problem always involves making some assumptions.

HEURISTIC MATCHING OF PROBLEM AND SOLUTION

Unlike data aggregation, a heuristic inference makes a great leap. For example,
it is one thing to aggregate indicators of work teams having diYculties
in problem solving and decision making, and quite another to leap to the
inference that the problem is team leadership. Such an inference is often
uncertain, based on assumptions of typicality, and is sometimes just a poorly
understood correlation. An essential practical characteristic of heuristics of
this type is that they reduce the time and eVort spent searching for information

Van, D. V. A. H. (2007). Engaged scholarship : A guide for organizational and social research. Oxford University Press, Incorporated.
Created from uu on 2024-11-20 05:12:44.
84 ENGAGED SCHOLARSHIP

and diagnosing a problem by skipping over intermediate means–ends or causal

relations. Clancey points out that the disadvantage of this problem diagnosis
heuristic is the likelihood of error.
[Heuristic inferences] are usually uncertain because the intermediate relations may
not hold in the speciWc case, or may not be observable, or may simply be poorly
understood. While not having to think about intermediate connections is advanta-
geous, this sets up a basic conXict for the problem solver—his inferential leaps
may be wrong. . . . There are unarticulated assumptions on which the interpretation
rests. . . . Yet, we might know enough to relate data classes to therapy classes and save
the patient’s life! (Clancey 1985: 307, 311)

REFINING SOLUTION TO THE CASE

Once a general solution is selected for the problem category, it must be reWned
or adapted to Wt the particular patient or case. Several solutions shown by
evidence and expert consensus to be correct for a given problem are often not
implemented in Welds of medicine (Denis and Langley 2002), human resources
(Anderson et al. 2001; Rynes et al. 2002), and management (Rogers 2003;
TranWeld et al. 2003). Chapter 9 discusses in greater detail the error-prone
process of deducing particular solutions for individual cases from general
solutions or models that are based on statistical evidence of a population or
sample. For example, the patient may have been classiWed into the wrong
subclass or archetype of a disease category. The classiWcation system may
be too general and not specify the boundary conditions or contingencies of
membership. The speciWc context or position of the individual case in
the distribution of the population sample may not be understood. As a conse-
quence, a solution category that is correct for a patient subclass or archetype may
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

not apply or may be incorrect for the individual patient. Action research that
diagnoses and implements a solution to solve the problem of a client is plagued
with these kinds of particularistic problems of diagnosis and intervention.

RELATIONS AMONG DIAGNOSTIC STEPS

Thus far the Xow of inference among the diagnostic steps has proceeded from
data to conclusions. However, the actual order of search and inference
between models of solutions and data are often reciprocal. This is evident
in the following kinds of questions that problem diagnosis often entails:
. Are data about a problem supplied or must they be requested?
. If the data are requested, what alternative solution models and question-
asking strategies should be used?

. If new data are received, how should they be used to make inferences?
. If, as is typical, alternative inferences of problems and solutions exist, how
does one decide which inference path to believe?
Clancey (1985: 324) notes that a ‘triggering’ relation between data and
solutions is pivotal in almost all descriptions of heuristic classiWcation inference.
We say that ‘a datum triggers a solution’ if the problem solver immediately thinks
about that solution upon Wnding out about the datum. However, the assertion may be
conditional (leading to an immediate request for more data) and is always context
dependent (though the context is rarely speciWed [or clearly understood]). A typical
trigger relation is ‘Headache and red painful eye suggests glaucoma’—red, painful eye
will trigger consideration of headache and thus glaucoma, but headache alone will not
trigger this association. . . . In general, speciWcity—the fact that a datum is frequently
associated with just a few solutions—determines if a datum triggers a solution
concept (‘brings it to mind’) in the course of solving a problem.
Heuristic triggers facilitate three kinds of non-exhaustive search techniques
between data and solutions:
1. Data-directed search, where one works forwards from data to abstractions,
matching solutions until a satisfactory or plausible set of alternative
inferences have been made.
2. Solution- or hypothesis-directed search, where one works backwards from
solutions, collecting evidence to support them.
3. Opportunistic search, where one combines data and hypothesis-directed
reasoning. Here heuristic rules trigger hypotheses, which lead to a focused
search, and new data may cause refocusing. Opportunistic search is not
exhaustive because the reasoning tends to be limited to a Wnite set of
plausible connections between data and solution classes.
The solutions developed with these heuristic methods may represent noth-
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

ing more than conjectures that require empirical research and testing. In
medical care, those solutions that are substantiated by research are known as
‘evidence-based’ interventions, while those that are selected for adoption by
panels of professional experts are typically referred to as ‘best practice’ guide-
lines. Since these best practice guidelines are based on expert opinions rather
than scientiWc evidence, they are often the subject in ‘calls for research’ through
clinical trials or evaluation research in order to empirically test the eYcacy of
the guidelines.

GOING BEYOND THE DATA GIVEN

As noted in the introduction, even when problems are grounded in reality, a
common diYculty is that their diagnosis may not lead to creative theory that

Van, D. V. A. H. (2007). Engaged scholarship : A guide for organizational and social research. Oxford University Press, Incorporated.
Created from uu on 2024-11-20 05:12:44.
86 ENGAGED SCHOLARSHIP

advances understanding of the phenomenon or problem. Bruner (1973)

argues that creativity requires going beyond the data given. He discusses the
cognitive learning problem of encountering an anomaly or breakdown that is
inconsistent with what we know. He points out that a theory or model is a
generic representation of the critical characteristics of a phenomenon. For
Bruner, grounding theories in reality requires going beyond the information
given so that the problem is formulated as having applicability beyond the
situation in which it is observed. This kind of creative problem formulation
involves an ‘emptying operation’ in which the scholar strips or abstracts away
idiosyncratic details of the situation observed in reality. In so doing he/she
learns something generic about the problem that generalizes to a broader
set of situations existing in reality. Bruner observes that initial descriptions
of a problem tend to be much too complex. In the beginning there is often not
a strong classiWcation scheme for distinguishing the ‘wheat’ from the ‘chaV.’
As a result, one may not see a blight in the forest because of the trees.
Henderson (1967) provides a good example of the ‘emptying operation’
that Bruner discusses by reviewing how Hippocrates, known as the father of
medicine, described the sickness and death of Philiscus.
Philiscus lived by the wall. He took to his bed with acute fever on the Wrst day and
sweating. Night uncomfortable.
Second day. General exacerbation; later a small clyster moved the bowels well.
A restful night.
Third day. Early and until midday he appeared to have lost the fever; but towards
evening acute fever with sweating, thirst, dry tongue, black urine. An uncomfort-
able night, without sleep; completely out of his mind.
Fourth day. All symptoms exacerbated; black urine. A more comfortable night, and
urine of a better colour.
Fifth day. About midday slight epitasis (nosebleed) of unmixed blood. Urine varied,
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

with scattered, round particles suspended in it, resembling semen; they did not settle.
On the application of a suppository the patient passed, with Xatulence, scanty excreta.
A distressing night, snatches of sleep, irrational talk; extremities everywhere cold, and
would not get warm again; black urine; snatches of sleep towards dawn; speechless; cold
sweat; extremities livid. About midday on the sixth day the patient died. The breathing
throughout, as though he were recollecting to do it, was rare and large. Spleen raised in
a round swelling. Cold sweats all the time. The exacerbations on even days.
Henderson (1967) notes that Hippocrates made three kinds of observations
to describe the process of death. First, there are simple descriptive observa-
tions made in the Wrst part of the illness. These observations are condensed to
the very limit and uncolored with diagnostic abstractions. Then, there are
repetitive patterns observed over time in the case that are used to diagnose
uniformities over time in the development of the patient’s illness. Finally,
(and not described here) are uniformities that Hippocrates observes across

cases, which represent recurrences in diVerent cases of single events or event

sequences. These uniformities are generalized into more abstract representa-
tions of patient categories, such as ‘Facies Hippocrates’ or the face at death:
‘Nose sharp, eyes hollow, temples sunken, ears cold and contracted with their lobes turned
outward, the skin about the face hard and tense and parched, the colour of the face as a
whole being yellow or black.’
Through cumulative observations of his patients, Hippocrates moves step-
by-step toward the widest generalizations within his reach. Henderson (1967)
concludes that such methodical descriptions of reality are necessary to develop
a science that deals with similarly complex and various phenomena.

Problem Solving: The Research Question

In practice, the solution to a problem solving process is the application of
a particular intervention that solves the problem identiWed. In research,
however, the solution to a problem formulation process is often a research
question that merits scientiWc investigation to better understand the problem
and its resolution. As stated before, grounding and diagnosing a problem
domain typically reveal many interesting and important research issues and
questions that might be studied. For example, in terms of the structure of
inference in Figure 3.1 a diagnosis of a given problem may trigger any one or
more of the following research questions.
1. Problem classiWcation and aggregation questions:
. What kind of problem is this; does it Wt known problem categories?
. What are the deWning characteristics or symptoms of a particular
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

problem or disease? How do they cause the problem or disease?

. Is a better coding system available for identifying and classifying the
components or symptoms of a particular problem or disease?
. In what contexts, situations, or contingencies do diVerent kinds of
problems arise?
. How do the elements or symptoms of problems originate and grow
into disease categories?
2. Heuristic matching of problem and solution questions:
. What speciWc solutions are appropriate for this problem, or variations
of this problem?
. What are the strengths and weaknesses of alternative solution models
for this kind of problem?
. How and why does a solution solve a problem; what are the causal
relationships among its components?

Van, D. V. A. H. (2007). Engaged scholarship : A guide for organizational and social research. Oxford University Press, Incorporated.
Created from uu on 2024-11-20 05:12:44.
88 ENGAGED SCHOLARSHIP

.Would a new solution address the problem more eVectively than the
status quo?
3. ReWning the solution to the case at hand:
. What are the relative merits of alternative solutions for the problem
exhibited in this particular case?
. Why are evidence-based solutions to this problem not adopted or
implemented?
. How should solutions be modiWed or adapted to Wt the local situation?
. What are the particular contexts or contingencies in which a solution
is beneWcial or harmful?
Obviously, in any given study it is impossible to examine all of these and
other questions that may emerge during problem diagnosis. Priorities need to
be established by formulating a speciWc question that will be addressed in a
research project. The research question often represents the end to the problem
formulation process for researchers, for it identiWes the speciWc question from
among a host of other possible questions that will be the focus of an empirical
investigation. The research question not only narrows the focus of a study to
manageable dimensions, it also establishes a pragmatic criterion for evaluating
the relevance and quality of a research project. A research study is successful
to the extent that it answers the question it was intended to address.
Selecting the research question is a key decision in focusing a research
project. Seldom is the research question selected at one time and in a once-
and-for-all fashion. Instead, the problem formulation activities of situating,
grounding, and diagnosing the problem provide numerous trials and oppor-
tunities to formulate, reframe, and modify research questions. Honing in on
the research question entails a clariWcation of the focus, level, and scope of the
problem domain from the perspective of the research users. It should be
grounded in the sense that the research question directly addresses a critical
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

aspect of the problem as it was observed in reality. And the question should be
important in identifying a critical gap, assumption, or anomaly that requires
further theory building and testing.
A number of common-sense suggestions (too often ignored) merit con-
sideration in formulating a research question. In outline form, they include
the following:
. A research question should end with a question mark (?), not a period (.).
Too often research questions are stated in a form that implies or preor-
dains a solution. Good research questions provoke inquiry by being stated
in ways that permit and entertain at least two or more plausible answers to
the question.
. The research question should directly address a key part of the problem
observed in reality. Too often there is a disconnect between statements of
the question and the problem.

. Consider the consequences of the research question. Will it resolve a key part
of the problem from the user’s perspective? Will it substantially improve the
situation for the user? Will it advance knowledge/competence for the user?
The art of formulating good research questions is easily as important as the
art of developing clear answers. Indeed, Jerome Bruner adds that the art of
cultivating such questions and keeping them alive is crucial to the mindful
and lively process of science making, as distinguished from what may appear
in a Wnished scientiWc report. ‘Good questions are ones that pose dilemmas,
subvert obvious or canonical ‘‘truths,’’ force incongruities upon our attention’
(Bruner 1996: 127).

Problem Formulation Techniques

BIASES IN HUMAN JUDGMENTS
A researcher is exposed to many biases in human judgment when engaging
others in problem formulation. Considering others’ views taxes our limited
capacities to handle complexity and maintain attention to particulars.
Research shows that human beings lack the capability and inclination to deal
with complexity (Kahneman et al. 1982). This is true not only of the researcher,
but also the individuals a researcher talks to about a problem domain. People
tend to quickly eliminate a problem and be ‘solution-minded,’ (i.e., focus on
solutions prematurely at the expense of not adequately deWning the problem).
Techniques that extend the duration of problem-mindedness by triangulating
on multiple methods and perspectives to represent the problem decrease the
likelihood of unintended bias in interpretations.
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

Research on individual attribution and decision making has found that

individuals systematically deviate from a rational ideal in making decisions,
causal judgments, social inferences, and predictions (Bazerman 1986; Cialdini
1993). Individuals tend to rely on a limited number of heuristic principles to
reduce information complexity. These heuristics allow individuals to solve
complex problems by applying more simple judgmental operations. When
applying these heuristics, individuals Wlter information in ways that bias their
assessments and inferences, leading them to make systematic, predictable
errors in judgment. Bazerman (1986) summarizes these heuristics and their
resulting biases (see Table 3.1).
A variety of techniques have been suggested for decreasing bias and increas-
ing problem solving creativity in individual reasoning and decision making
by Polya (1957), Bransford and Stein (1993), Halpern (1996), among others.
A common theme in this literature is that individuals monitor their cognitive

Van, D. V. A. H. (2007). Engaged scholarship : A guide for organizational and social research. Oxford University Press, Incorporated.
Created from uu on 2024-11-20 05:12:44.
90 ENGAGED SCHOLARSHIP

Table 3.1. Biases in individual and group decision making

Bias Description

A. Biases emanating from the availability heuristic

1. Ease of recall Easily-recalled events, based upon vividness or recency,
are judged to be more numerous than events of equal
frequency whose instances are less easily recalled.
2. Retrievability Information that is easy to search for and obtain is more
salient than that which is difficult to retrieve.
3. Presumed associations Overestimating the probability of two events co-occurring
based upon the number of similar associations that are
easily recalled.
B. Biases emanating from the representativeness heuristic
1. Insensitivity to base rates Ignoring base rates in assessing the likelihood of events
when other descriptive information is provided—even if it
is irrelevant.
2. Insensitivity to sample size Failing to appreciate the role of sample size in assessing
the reliability of sample information.
3. Misconceptions of chance Expecting that a data sequence generated by a random
process will look ‘random,’ even when the sequence is too
short for those expectations to be statistically valid.
4. Regression to the mean Ignoring the fact that extreme events tend to regress to
the mean on subsequent trials.
C. Biases emanating from anchoring and adjustment
1. Insufficient anchor adjustment Making insufficient adjustments from an initial anchor
value (derived from past events, random assignment, or
whatever information is available) when establishing a
final value.
2. Conjunctive & disjunctive events Overestimating the probability of conjunctive events and
underestimating the probability of disjunctive events.
3. Overconfidence Being overconfident of the infallibility of judgments when
answering moderately to extremely difficult questions.
D. Two more general biases
1. The confirmation trap Seeking confirmatory information for what you think is
true and neglecting a search for disconfirmatory evidence.
2. Hindsight Overestimating the correctness of a predicted outcome
after finding out whether or not an event occurred.
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

E. Biases emanating from group decision making

1. Groupthink and conformity Group pressures to conform to others depress group
members from considering divergent views.
2. Risky shift Group decisions are either more conservative or extreme
than the average of group members’ individual decision.
3. Conflict avoidance Seeking cohesiveness and avoiding expressing ideas that
may be disagreeable or opposing to group members.
4. Falling into a ‘rut’ Focusing on evaluating an idea once expressed, rather
than introducing more new ideas in group discussion.

Sources: Bazerman, M. (1986). ‘Biases’, in B. M. Staw (ed.), Psychological Dimensions of Organizational

Behavior, 2nd edn. Engelwood Cliffs, NJ: Prentice Hall, pp. 199–223. Delbecq, A., Van de Ven, A.,
and Gustafson, D. (1975). Group Techniques for Problem Solving and Program Planning. Reading,
MA: Addison Wesley.

biases by engaging other informants to provide information and interpret-

ations of the problem domain. This is particularly true for problems where
information from people with diVerent perspectives about the problem
is important. One-on-one interviews, group meetings, surveys, and other

techniques can be used to gather this information. DiVerent techniques

provide diVerent kinds of information. For example, the information gathered
from brainstorming or focus group meetings often provides an overall gestalt
appreciation of a problem that seldom emerges from individual interviews or
survey techniques. But use of group meetings to gather information to ground
a problem must be done with caution and with careful structuring of the
group process.
Group decision biases often skew decision making away from a rational
ideal. As the bottom rows of Table 3.1 outline, four of the most common group
decision biases are groupthink, risky shift, conXict avoidance, and falling into a
‘rut.’ In a classic study of group process, Janis identiWed the phenomenon
called ‘groupthink’ (Janis and Mann 1977). In groupthink situations, group
pressures for conformity deter the group from considering all of the views of its
members. Risky shift is another group decision bias. Researchers including
Kogan and Wallach (1967) found evidence that group decision making is
sometimes more conservative than individual decisions. But more often,
group decision making tends to be more extreme than the average of group
members’ individual decisions.
Delbecq et al. (1975) describe several causes of bias in group decision
making. Group members in many cultures may be expected to conform,
causing them to withhold opinions contrary to the dominant view. Some
individuals may be hesitant to express their opinions due to personality
characteristics like shyness. Group members may lack the communication
skills to adequately present their ideas. Others skilled in impression manage-
ment techniques may be able to dominate group discussion, even though they
lack substantive expertise on the topic at hand. Some group members may be
particularly egocentric and unwilling to consider the views of others. Status
and hierarchy may aVect group discussion, with more junior or low-level
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

members deferring to the opinions of senior or high-level members.

COGNITIVE MAPPING
One technique that is particularly useful for representing a problem as
perceived by an individual or group is cognitive mapping. The technique
is not necessarily intended to reXect a logical or rational analysis of a
situation. Instead, cognitive mapping is a modeling technique designed to
represent a person’s ideas, beliefs, values, attitudes, and their inter-relation-
ships in a form that is amenable to study and analysis. Bryson et al. (2004)
and Eden et al. (1983) describe the technique for constructing a problem as
perceived by an individual. The technique could also be used in a group
setting. In outline form, the technique consists of the Wve steps illustrated in
Figure 3.2.

Van, D. V. A. H. (2007). Engaged scholarship : A guide for organizational and social research. Oxford University Press, Incorporated.
Created from uu on 2024-11-20 05:12:44.
92 ENGAGED SCHOLARSHIP

High absenteeism . . .
reasonable attendance
Decrease in the quality +
of raw materials . . . Poor . . . decent labour force
+ Q4.
+
High wastage . . . +
Q1.
production output dropping quickly
Q2. . . . steady output
+
Not meeting orders on time +
. . . meeting deliveries + Q3.
Boss questioning my ability
- . . . boss leaves me alone
Peace and quiet . . .
customers complaining

Figure 3.2. Steps in developing a cognitive map of a problem

Source: Eden, C., Jones S., and Sims, D. (1983). Messing About in Problems: An Informal Structured
Approach to their Identification and Management. Oxford: Pergamon Press, fig. 4.2, p. 42.

1. In the center of a piece of paper or a Xip chart, write a label for the problem.
. An agreed-upon label is usually what is noted as the problem, such as
‘production output dropping quickly.’
2. Ask ‘What is a satisfactory alternative to this circumstance?’
. Find out what the person thinks about his/her own circumstances
(rather than the oYcial or politically correct point of view).
. Ask the individual to describe an opposite alternative that may resolve
the situation, such as ‘steady output.’
3. Ask ‘Why does this matter to you? Why are you worried about it? What
are its consequences?’ How a concept is used and what it is contrasted to
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

provide the meaning in that context. It also identiWes the construct poles
(i.e., its connotative links).
. Identify the psychological (not necessarily ‘logical’) opposites for each
construct (negation).
. Use arrows with positive or negative signs to identify psychological
implications of causality among poles of constructs.
4. Ask ‘What reasons come to mind as explanations for the problem label.
What are its antecedents?’
. At this stage the problem is beginning to take shape in an explicit
model as the individual sees it. Others may see it quite diVerently. For
example, they may have the same constructs, but they diVer in the
structure of relationships or arrow.
5. Elaborate.
. Think backwards and forwards to elaborate the problem to encourage
creative expression of the issue.

. What is happening is that the nature of the problem is gradually

changing as articulation and modeling take place.
. This is not a scientiWc model of an ‘objective’ reality. It is a cognitive
mapping of an issue as one person sees it.
Bryson et al. (2004) and Eden et al. (1983) discuss numerous examples and
applications of this cognitive mapping technique for problem formulation.

GROUP PROCESS TECHNIQUES

Brainstorming techniques are useful for obtaining ideas about the character-
istics of problems from groups of individuals who experience or are know-
ledgeable informants about a problem domain. Brainstorming techniques
structure group decision processes in ways to minimize bias or render them
less problematic in group formulations of a research problem. One of the most
widely used methods of group brainstorming is the Nominal Group Technique
(Delbecq et al. 1975). Figure 3.3 outlines the structured steps of a nominal
group meeting that is conducted for the purpose of generating ideas about a
topic, problem, or issue. The meeting begins by giving group participants a
sheet of paper on which they are asked to write all of their individual ideas

1. Silent writing of ideas on question:

• No one talks, everyone thinks and writes.
2. Round-robin recording of ideas on chart:
• Leader writes an idea from each person on chart; proceeds around group;
• No one talks out of turn to evaluate ideas;
• Members listen and present new ideas when their turn comes.
3. Preliminary vote on ideas:
• Members silently list 3−5 best ideas on their own papers;
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

• Leader tabulates votes on flip chart.

4. Discussion:
• General discussion, evaluation, and debate of ideas on chart.
5. Final vote on ideas:
• Procedure is the same as step 3, and meeting concludes.

Figure 3.3. Steps in a nominal group meeting

Van, D. V. A. H. (2007). Engaged scholarship : A guide for organizational and social research. Oxford University Press, Incorporated.
Created from uu on 2024-11-20 05:12:44.
94 ENGAGED SCHOLARSHIP

A. Silent writing of ideas on question:

• provides focus, time for creativity without interruptions;
• avoids conformity, competition, & status problems;
• avoids evaluation and jumping to conclusions.
B. Round-robin recording of ideas on chart:
• forces equal sharing and participation;
• encourages more ideas through 'hitch-hiking';
• depersonalizes ideas & tolerates conflicting ideas.
C. Clarification of each idea on chart:
• clarifies each idea before jumping to conclusions;
• each idea is as important as another before vote.
D. Preliminary vote on priorities:
• allows 'trial run' & avoids premature conclusions;
• provides focus on important issues;
• silent voting forces equality & avoids influence of others.
E. Discussion of preliminary vote:
• encourages minority opinions & clarifies misunderstandings;
• promotes attacking ideas on wall (not people);
• provides preparation for decision.
F. Final vote on priorities:
• provides written ‘minutes’ of group ideas & decision;
• promotes sense of accomplishment & conclusion;
• motivates involvement in future efforts.

Figure 3.4. Reasons for steps in nominal group meeting silent writing of ideas on question
Note: For further information, see Delbecq, A., Van de Ven, A., and Gustafson, D. (1975). Group Techniques
for Program Planning, Scott-Foresman Pub.

about a question or problem. Then, in round-robin fashion each individual is

given a turn to present his/her idea. Other members are not permitted to
evaluate the idea at this time. Each person takes a turn until all ideas are
presented. Then, there is time for discussion and members to vote on the
ideas they view as being most valuable. This limits the possibility for any one
individual to ‘take over’ and dominate the discussion, allowing for a wider
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

range of possibilities for the group to consider (Delbecq et al. 1975). Additional
reasons for each step in the nominal group process are outlined in Figure 3.4.
Another technique for overcoming some of the cognitive biases of group
decision making is the Delphi technique where group members do not meet
face-to-face; instead they respond to questions that are proposed and tabulated
by a group coordinator using electronic discussion boards or email. In this
format, individual participants can engage in the process anonymously, negating
the potential eVects of status and hierarchy and possibly allowing those who are
shy or lack verbal communication skills to participate more fully in the process.
Other possible group techniques that inhibit group decision biases
include dialectical inquiry and devil’s advocacy as described by Schweiger
et al. (1989). As applied to problem formulation, brainstorming techniques
focus on generating as many problem statements or components as
possible, while dialectical methods focus on evaluating and choosing a

representation of the problem from a few competing alternatives. In dialect-

ical inquiry, a decision making group might be divided into two subgroups.
One subgroup develops an argument for a particular problem statement
supported with reasons, evidence, assumptions, and qualiWcations. Another
subgroup may do the same for a competing formulation of the problem. The
groups then present their arguments and engage in a debate by attempting to
bolster their argument and negate those of the other group. A panel of judges
or reviewers is often called upon to choose the winning argument or to
develop a synthesis that attempts to combine the strengths and minimize
the weaknesses of the two arguments. Alternatively, the groups themselves
continue debating their assumptions and conclusions until they agree on a
problem statement.
In devil’s advocacy, the second subgroup criticizes the assumptions and
recommendations of the Wrst subgroup, but does not propose new solutions.
The Wrst group revises its assumptions and recommendations and presents
them again to the second subgroup, and the process continues until the group
agrees on a problem statement. These techniques can be used to engage an
open debate and to examine a more inclusive set of ideas than a traditional
consensus technique that does not incorporate an established norm of debate.

Concluding Discussion
This chapter discussed the process of formulating a research problem in
terms of four interdependent activities: situating, grounding, diagnosing,
and resolving a problem. These four activities are highly interdependent
and typically occur in parallel throughout the problem formulation process.
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

A variety of suggestions and considerations in performing each of the

problem formulation activities were also discussed. In summary, they are
outlined below.

SITUATING THE RESEARCH PROBLEM

. Identify whose point of view and interests are to be represented in
problem formulation.
. Clarify who are the intended users, clients, and audience of the research.
. Who/what is the foreground and background in focusing on the problem?
. What is the level of analysis and context of the problem domain?
. What is the scope of the problem: how deep, how broad, and how long will
the problem be studied?

Van, D. V. A. H. (2007). Engaged scholarship : A guide for organizational and social research. Oxford University Press, Incorporated.
Created from uu on 2024-11-20 05:12:44.
96 ENGAGED SCHOLARSHIP

GROUNDING THE RESEARCH PROBLEM

. Address the journalists’ questions of who, what, where, when, why, and
how does the problem exist.
. Describe the particular problem up close; give an example, anecdote, and
an experience with a problem.
. Describe the problem in general; present evidence and studies on the
prevalence and context of the problem’s existence.
. Talk to people who experience the problem or issue.
. Conduct interviews and nominal groups with people who know about the
issue/problem.
. Review the literature to understand and situate the problem.

DIAGNOSING THE RESEARCH PROBLEM

. Classify elements or symptoms of the problem into categories.
. Aggregate categories to infer a problem.
. Heuristically match the problem and potential solutions.
. Does a solution exist for the problem or must one be created?
. ReWne the solution to the case at hand.
. What anomalies or breakdowns surfaced in your diagnosis?

SELECTING THE RESEARCH QUESTION

. What part of the problem merits research attention and focus?
. State the question in analytical terms by relating or comparing key
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

concepts.
. Connect the research question to your description of the problem.
. Ensure that you indeed have a question and not a statement.
. Permit and entertain at least two plausible answers to the question.
Alternative answers increase independent thought trials.
. Bring your question full circle by considering its consequences:
– Will the answer solve a key part of the problem from the user’s
perspective?
– Will it substantially improve the situation for the user?
– Will it advance knowledge/competence for the user?

Some years ago, the Journal of Management Inquiry featured a debate between
Paul Lawrence and Karl Weick about the merits and demerits of undertaking
problem-driven or theory-driven research. This debate included the question

of who should be the primary user or client of management research. Herbert

Simon argued that research in a professional school should both advance
knowledge in a scientiWc discipline and enlighten the practice of management.
The arguments advanced by Lawrence and Weick appear to reXect the diVer-
ing viewpoints of scholars from the reality and theory bases of the Engaged
Scholarship Diamond Model. Weick views problem formulation from a
theoretical perspective, while Lawrence examines it as a real-world phenom-
enon. But Weick and Lawrence are not just good role players; they have had
many successes in their scholarly careers by successfully completing all the
activities of the Diamond Model. Murray Davis (1971, 1986) provides an
insightful synthesis to this debate by examining what makes research inter-
esting and classic. He argues that the better we know our audience, the better
we can frame or position our research problem and question to the prevailing
assumptions of the readers/users of our research.
Paul Lawrence (1992: 140) argues that signiWcant behavioral science ori-
ginates from problem-oriented research rather than from theory-oriented
research. He states that behavioral scientists make a big mistake if they do
not ask users and research participants what needs to be studied, and how our
theories and knowledge may be inadequate. He suggests seven steps for
undertaking behavioral science research:
1. Select an important emerging human problem to study, based on careful
listening and observations and being explicit about the value choices
involved.
2. Do some initial Weld scouting of the problem to make an initial assess-
ment of key parameters.
3. Examine relevant theory for promising hypotheses and conceptualizations.
. Note: this is the third step, not the Wrst.
. Theory is a good guide for framing, not selecting, the research ques-
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

tion.
4. Select research methods after the research question and propositions are
chosen.
5. Collect data systematically.
6. Analyze data and generalize the Wndings.
7. Present results so they are useful for action by responsible problem
solvers and academicians.
Lawrence outlines seven advantages of problem-oriented research:
1. One is more likely to develop usable Wndings and Wndings that are
actually used.
2. Problems link micro and macro levels of analysis.
3. By identifying important problems, the research is practical.

Van, D. V. A. H. (2007). Engaged scholarship : A guide for organizational and social research. Oxford University Press, Incorporated.
Created from uu on 2024-11-20 05:12:44.
98 ENGAGED SCHOLARSHIP

4. It is easier to gain access to study sites and secure funding.

5. The research usually identiWes some performance measures as dependent
variables.
6. One is more likely to discover new and better organizational forms and
social inventions.
7. The chief advantage is that problems are a powerful way to identify gaps
in our theory/knowledge.
Weick (1992) argues the case for theory-driven research. All theories are
about practice and practicality, and the trick is to discover those settings
and conditions under which they hold true. He discusses three criteria
for selecting research problems: knowledge (choose problems in areas
where you have a thorough understanding), dissatisfaction (choose prob-
lems that reXect a healthy, active opposition to existing knowledge and
methodology), and generalizability (choose variables and situations that
are universal and common rather than unique and rare). A theory-based
approach is a continuing eVort to Wnd those contexts where a theory holds
true (Weick 1992: 172). This formulation emphasizes the process of diag-
nosis. The particular problem is a pretext to look for a pattern that is more
generalizable, more abstract, something that applies to people in general
(Meehl 1995); it is an exercise in sense making. In problem-focused work,
the particular is a context rather than a pretext and consists of a self-
contained story, tied together by its own logic. DiVerent logics will generate
diVerent studies, and diVerent stories will suggest diVerent remedies (Weick
1992: 172).
Because of the theory-laden nature of all observations and data, I empha-
sized that problem-driven and theory-driven research are inextricably con-
nected. Problem formulation and theory building follows an abductive form
of reasoning, which is neither inductive nor deductive. Abduction begins by
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

recognizing an anomaly or breakdown in our understanding of the world, and

proceeds to create a hypothetical inference that dissolves the anomaly by
providing a coherent resolution to the problem. As will be discussed in the
next chapter, in the social sciences this logic of discovery is typically followed
by a logic of testing in which the consequences of the hypothesis are derived
through deduction and the consequences are tested by induction.
I discussed the close connections between theory and practice in situating,
grounding, diagnosing, and solving a problem. Lawrence and Weick suggest
that the distinction between problem- vs. theory-focused research is neither
descriptive nor useful for understanding the process of problem formulation.
Theories have problems just as problems have theories (Weick 1992: 176).
Problem formulation requires theoretical categories; and theories are used to

represent reality. Theories and problems alternate as temporally lagged logics

or stories. One person’s problem is another person’s solution (theory). There-
fore, the question of whether research is problem- or theory-driven reXects a
one-sided view of the research process. Subsequent chapters will demonstrate
that the distinction between problem- and theory-driven research is also not
helpful for theory building, testing, and application.
Copyright © 2007. Oxford University Press, Incorporated. All rights reserved.

Van, D. V. A. H. (2007). Engaged scholarship : A guide for organizational and social research. Oxford University Press, Incorporated.
Created from uu on 2024-11-20 05:12:44.

Mckinsey's 7s Model
No ratings yet
Mckinsey's 7s Model
14 pages
Gilbert Worthy Performance 6boxes
No ratings yet
Gilbert Worthy Performance 6boxes
6 pages
Gilbert Worthy Performance 6boxes PDF
No ratings yet
Gilbert Worthy Performance 6boxes PDF
6 pages
Van de Ven, Engaged Scholarship
No ratings yet
Van de Ven, Engaged Scholarship
43 pages
Transfering Corner Office Strategy Into Frontline Action
0% (2)
Transfering Corner Office Strategy Into Frontline Action
10 pages
Warren Strategy Implementation and Digital Twin Business Models
No ratings yet
Warren Strategy Implementation and Digital Twin Business Models
7 pages
Effective Argumentation Techniques
No ratings yet
Effective Argumentation Techniques
1 page
Saaty How To Make A Decision The Analytic Hirearchy Process
No ratings yet
Saaty How To Make A Decision The Analytic Hirearchy Process
26 pages
A Paradox of Ethics Why People in Good Organizatio
No ratings yet
A Paradox of Ethics Why People in Good Organizatio
21 pages
(IIA UK) - Position Statement - RIsk Based Internal Auditing
No ratings yet
(IIA UK) - Position Statement - RIsk Based Internal Auditing
4 pages
Creativity - The Discipline of Innovation by Drucker, Peter
100% (15)
Creativity - The Discipline of Innovation by Drucker, Peter
7 pages
Hitt, M.a., Keats, B.W., & DeMarie, S.M. 1998. Navigating in The New Competitive
100% (1)
Hitt, M.a., Keats, B.W., & DeMarie, S.M. 1998. Navigating in The New Competitive
22 pages
Corporate Finance - European Edition David Hillier - Ebook PDF PDF Download
100% (3)
Corporate Finance - European Edition David Hillier - Ebook PDF PDF Download
115 pages
Understanding Change Resistance
100% (1)
Understanding Change Resistance
12 pages
Why Transformation Efforts Fail
No ratings yet
Why Transformation Efforts Fail
25 pages
Reilly Building Business Acumen Hrwest1209
No ratings yet
Reilly Building Business Acumen Hrwest1209
3 pages
Organizational Control - Sim B Sitkin - Laura B Cardinal - Katinka Bijlsma-Frankema
100% (1)
Organizational Control - Sim B Sitkin - Laura B Cardinal - Katinka Bijlsma-Frankema
561 pages
A Review of Action Learning Literature
No ratings yet
A Review of Action Learning Literature
7 pages
Developing An Innovation Strategy: © 2009 John Wiley & Sons LTD
No ratings yet
Developing An Innovation Strategy: © 2009 John Wiley & Sons LTD
35 pages
Organizational Culture and Innovation
No ratings yet
Organizational Culture and Innovation
55 pages
PDCA
No ratings yet
PDCA
28 pages
Topic Workbook - Strategic Leadership - Managing The Strategy Making Process
No ratings yet
Topic Workbook - Strategic Leadership - Managing The Strategy Making Process
17 pages
Strategy Dynamics Essentials
No ratings yet
Strategy Dynamics Essentials
175 pages
Effectuation How Entrepreneur Thinks
No ratings yet
Effectuation How Entrepreneur Thinks
63 pages
Innovation Streams and Dynamic Capabilities
No ratings yet
Innovation Streams and Dynamic Capabilities
42 pages
Kepner Tregoe
100% (2)
Kepner Tregoe
26 pages
Understanding Business Strategy
100% (1)
Understanding Business Strategy
64 pages
Leading Your Team Into The Unknown
No ratings yet
Leading Your Team Into The Unknown
2 pages
Situational Theory of Management
No ratings yet
Situational Theory of Management
5 pages
It's Not Luck: Gower
No ratings yet
It's Not Luck: Gower
1 page
Six Sigma Metrics and DMAIC
100% (2)
Six Sigma Metrics and DMAIC
21 pages
Decision Tree Analysis
No ratings yet
Decision Tree Analysis
5 pages
Executive Pay: Rent vs. Value Debate
No ratings yet
Executive Pay: Rent vs. Value Debate
57 pages
The Strategy Process
No ratings yet
The Strategy Process
34 pages
How To Compare Six Sigma Lean and The Theory of Constraints
No ratings yet
How To Compare Six Sigma Lean and The Theory of Constraints
6 pages
IRIS Bengkel Coaching For Performance (PowerPoint Presentation)
No ratings yet
IRIS Bengkel Coaching For Performance (PowerPoint Presentation)
28 pages
Richard Murphy and The Biscuit Company
100% (1)
Richard Murphy and The Biscuit Company
8 pages
Strategic Change for Managers
100% (2)
Strategic Change for Managers
45 pages
1 Motivation Script
No ratings yet
1 Motivation Script
15 pages
Adaptive Critical Thinking For A VUCA World
No ratings yet
Adaptive Critical Thinking For A VUCA World
7 pages
TheDefinitiveDrucker PDF
100% (1)
TheDefinitiveDrucker PDF
4 pages
Target Setting - Stacey Barr
No ratings yet
Target Setting - Stacey Barr
4 pages
The Definitive Drucker - Book Review
No ratings yet
The Definitive Drucker - Book Review
2 pages
Talent Management in Malaysian Universit
No ratings yet
Talent Management in Malaysian Universit
20 pages
Strategic HR
No ratings yet
Strategic HR
12 pages
Conceptual Foundations of Balanced Scorecard
No ratings yet
Conceptual Foundations of Balanced Scorecard
37 pages
Disruptive Strategy Syllabus
No ratings yet
Disruptive Strategy Syllabus
2 pages
4 Types of Business Innovation
100% (1)
4 Types of Business Innovation
4 pages
Managementgt Excellence
No ratings yet
Managementgt Excellence
20 pages
Analytical Problem Solving Guide
50% (2)
Analytical Problem Solving Guide
7 pages
Mind Map Summary Created by Eryc Eyl. The 4 Disciplines of Execution Is A Registered Trademark of Franklincovey. All Rights Reserved by The Owner
No ratings yet
Mind Map Summary Created by Eryc Eyl. The 4 Disciplines of Execution Is A Registered Trademark of Franklincovey. All Rights Reserved by The Owner
1 page
Isb Cee GMP Brochure
No ratings yet
Isb Cee GMP Brochure
21 pages
10 Schools of Strategy
100% (1)
10 Schools of Strategy
6 pages
Evaluation 4
No ratings yet
Evaluation 4
73 pages
Teamwork Turmoil Report Puneet Agarwal EPGP 1914002
No ratings yet
Teamwork Turmoil Report Puneet Agarwal EPGP 1914002
8 pages
Learning Outcome 4: Developing Leaders 4.1: Evaluate Leadership and Fellowship 4.2: Understand Ethical Leadership 4.3. Discuss Leadership Management
No ratings yet
Learning Outcome 4: Developing Leaders 4.1: Evaluate Leadership and Fellowship 4.2: Understand Ethical Leadership 4.3. Discuss Leadership Management
2 pages
Double-Loop Learning in Education
No ratings yet
Double-Loop Learning in Education
17 pages
90763.casewrap - Indd 1 3/30/09 3:45 PM 3/30/09 3:45 PM
100% (1)
90763.casewrap - Indd 1 3/30/09 3:45 PM 3/30/09 3:45 PM
199 pages
1.engaged Scholarship in A Professional School
100% (1)
1.engaged Scholarship in A Professional School
35 pages
Wk3 Cascio 2008
No ratings yet
Wk3 Cascio 2008
15 pages
The Antonomy of Competition
No ratings yet
The Antonomy of Competition
6 pages
Competition and Business Strategy in Historical Perspective
No ratings yet
Competition and Business Strategy in Historical Perspective
38 pages
Balancing Divergence and Convergence: Stimulating Creativity Through Hybrid Thinking
No ratings yet
Balancing Divergence and Convergence: Stimulating Creativity Through Hybrid Thinking
21 pages
Stefan Jonsson and Patrick Regner - Normative Barriers To Imitation: Social Complexity of Core Competences in A Mutal Fund Industry
No ratings yet
Stefan Jonsson and Patrick Regner - Normative Barriers To Imitation: Social Complexity of Core Competences in A Mutal Fund Industry
21 pages
NEP 2020 Handbook for HEIs
No ratings yet
NEP 2020 Handbook for HEIs
33 pages
Historical Foundation of Education
No ratings yet
Historical Foundation of Education
23 pages
What Is Impact Assessment OECDImpact
No ratings yet
What Is Impact Assessment OECDImpact
7 pages
Education: For Reference, See The Easier English Dictionary For Students (0 7475 6624 0)
No ratings yet
Education: For Reference, See The Easier English Dictionary For Students (0 7475 6624 0)
2 pages
British University Enrollment Challenges
100% (1)
British University Enrollment Challenges
5 pages
Qa Handbook
No ratings yet
Qa Handbook
80 pages
Approval Oplan Maticas
No ratings yet
Approval Oplan Maticas
13 pages
Scholarships Empowering Women
100% (1)
Scholarships Empowering Women
2 pages
IIIT Pune Non Teaching Advt - Aug 2019
No ratings yet
IIIT Pune Non Teaching Advt - Aug 2019
7 pages
RA 8292 Ched Modernization Act
100% (1)
RA 8292 Ched Modernization Act
7 pages
WWW - Aimst.edu - My - Handbook - UnderGraduate Handbook (5th Printing) FINAL 160413 Rev2
No ratings yet
WWW - Aimst.edu - My - Handbook - UnderGraduate Handbook (5th Printing) FINAL 160413 Rev2
116 pages
UDUS-Students-Handbook-2019-2020-reprint 2021
No ratings yet
UDUS-Students-Handbook-2019-2020-reprint 2021
142 pages
UNIDA Gontor: A Unique Islamic University
100% (1)
UNIDA Gontor: A Unique Islamic University
85 pages
Stenhouse-1979-Research As A Basis For Teaching
0% (1)
Stenhouse-1979-Research As A Basis For Teaching
24 pages
List of Accredited Medical Schools
No ratings yet
List of Accredited Medical Schools
11 pages
Emerging Trends and Challenges in Tertiary Education
No ratings yet
Emerging Trends and Challenges in Tertiary Education
7 pages
Kuvempu University: Amount in Rs
No ratings yet
Kuvempu University: Amount in Rs
2 pages
Professor: Emad Elbeltagi, M.SC., PH.D., P.Eng
No ratings yet
Professor: Emad Elbeltagi, M.SC., PH.D., P.Eng
10 pages
Holly Drutarovsky: Education Temple University: Philadelphia, Pennsylvania 19122
No ratings yet
Holly Drutarovsky: Education Temple University: Philadelphia, Pennsylvania 19122
2 pages
BS Odl PDF
67% (3)
BS Odl PDF
104 pages
EDUCATION Handbook Updated June 28 2019
No ratings yet
EDUCATION Handbook Updated June 28 2019
377 pages
Tracer Study Enhances University Programs
No ratings yet
Tracer Study Enhances University Programs
14 pages
How Do I Write The First Email To A Professor For PHD Application - Yet Another
100% (1)
How Do I Write The First Email To A Professor For PHD Application - Yet Another
9 pages
Desales Undergrad Catalog PDF
No ratings yet
Desales Undergrad Catalog PDF
151 pages
Education in The Context of Different Societies Upon Which The School Have Been Established - 065818
No ratings yet
Education in The Context of Different Societies Upon Which The School Have Been Established - 065818
1 page
UT Dallas Syllabus For Musi2324.003 06f Taught by Randall Nye (rwn016000)
No ratings yet
UT Dallas Syllabus For Musi2324.003 06f Taught by Randall Nye (rwn016000)
5 pages
SOP Samples for Aspiring Students
No ratings yet
SOP Samples for Aspiring Students
5 pages
The History and Evolution of The Indian Education System
No ratings yet
The History and Evolution of The Indian Education System
16 pages
Construction of Objective Paper and Pencil Tests: Identification, Enumeration, Analogy, Interpretive Items
No ratings yet
Construction of Objective Paper and Pencil Tests: Identification, Enumeration, Analogy, Interpretive Items
49 pages
Curriculum Vitae Prof. Dr. Vibhuti Patel 2-12-2010
No ratings yet
Curriculum Vitae Prof. Dr. Vibhuti Patel 2-12-2010
15 pages