What Constitutes a Theoretical Contribution?

Author(s): David A. Whetten

Source: The Academy of Management Review, Vol. 14, No. 4 (Oct., 1989), pp. 490-495
Published by: Academy of Management
Stable URL: http://www.jstor.org/stable/258554
Accessed: 23/09/2010 03:28

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at
http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless
you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you
may use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at
http://www.jstor.org/action/showPublisher?publisherCode=aom.

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact [email protected].

Academy of Management is collaborating with JSTOR to digitize, preserve and extend access to The Academy
of Management Review.

http://www.jstor.org
?Acaderyof Management Review, 1989, Vol. 14, No. 4, 490-495

What Constitutes a
Theoretical Contribution?
DAVIDA. WHETTEN
University of Illinois
Since becoming editor of AMR, I have tried to ment? and (c) What factors are considered in
find a simple way to communicate the neces- judging conceptual papers? The first section de-
sary ingredients of a theoretical contribution. scribes the constituent elements of a theory. The
There are several excellent treatises on the sub- second section uses this framework to establish
ject, but they typically involve terms and con- standards for the theory-development process.
cepts that are difficult to incorporate into every- The third section summarizes the expectations of
day communications with authors and review- reviewers regarding the substantive contribu-
ers. My experience has been that available tion and appropriateness of AMR papers.
frameworks are as likely to obfuscate, as they
are to clarify, meaning. Besides exposure to the What Are the Building Blocks of
works of Kaplan, Dubin, and others varies
widely across the Academy.
Theory Development?
This article is a rudimentary effort to fill this According to theory-development authorities
gap: The intent is not to create a new conceptu- (e.g., Dubin, 1978), a complete theory must con-
alization of theory, but rather to propose several tain four essential elements, which are de-
simple concepts for discussing the theory- scribed in the following paragraphs.
development process. It is a personal reflection, What. Which factors (variables, constructs,
which has emerged out of my daily editorial ac- concepts) logically should be considered as part
tivities. My motivation is to ease the communi- of the explanation of the social or individual
cation problems regarding expectations and phenomena of interest? Two criteria exist for
standards, which result from the absence of a judging the extent to which we have included
broadly accepted framework for discussing the the "right" factors: comprehensiveness (i.e., are
merits of conceptual writing in the organiza- all relevant factors included?) and parsimony
tional sciences. (i.e., should some factors be deleted because
Finally, my comments should not be inter- they add little additional value to our under-
preted either as official AMR dogma or ironclad standing?).
rules governing the evaluation process. Each When authors begin to map out the concep-
submitted paper is unique, and it is judged on its tual landscape of a topic they should err in favor
own merits; however, my thinking has clearly of including too many factors, recognizing that
been influenced by the hundreds of communi- over time their ideas will be refined. It is gener-
cations I have read during the first half of my ally easier to delete unnecessary or invalid ele-
editorship. ments than it is to justify additions. However, this
This article is organized around three key should not be interpreted as license to throw in
questions: (a) What are the building blocks of the kitchen sink. Sensitivity to the competing vir-
theory development? (b) What is a legitimate tues of parsimony and comprehensiveness is the
value-added contribution to theory develop- hallmark of a good theorist.

490
 How. Having identified a set of factors, the little value in the laboratory. The mission of a
researcher's next question is, How are they re- theory-development journal is to challenge and
lated? Operationally this involves using extend existing knowledge, not simply to rewrite
"arrows" to connect the "boxes." Such a step it. Therefore, authors should push back the
adds order to the conceptualization by explicitly boundaries of our knowledge by providing com-
delineating patterns. In addition, it typically in- pelling and logical justifications for altered
troduces causality. Although the researcher views. This requires explaining the Whys under-
may be unable to adequately test these links, lying the reconstituted Whats and Hows.
restrictions in methods do not invalidate the in- Why research is conducted has important im-
herent causal nature of theory. plications for the link between theory develop-
Together the What and How elements consti- ment and empirical research. Combining the
tute the domain or subject of the theory. The Hows and the Whats produces the typical
more complex the set of relationships under con- model, from which testable propositions can be
sideration, the more useful it is to graphically derived. (The primary difference between prop-
depict them. Not all theoretical treatises must ositions and hypotheses is that propositions in-
contain figures with boxes and arrows, but a volve concepts, whereas hypotheses require
visual representation often clarifies the author's measures.) Technically, these statements (e.g.,
thinking and increases the reader's comprehen- A is caused by B) can be tested without under-
sion. In particular, formal models aid theory de- standing the Whys underlying the model. How-
velopers and users to assess the balance be- ever, this tends to lead to empirically, rather
tween parsimony and completeness. than theoretically, dominated discussions of the
Why. What are the underlying psychological, implications of a study's results. As a field, when
economic, or social dynamics that justify the se- we have insufficient understanding of why we
lection of factors and the proposed causal rela- collectively started an investigative journey, or
tionships? This rationale constitutes the theory's what theoretical direction we are following,
assumptions-the theoretical glue that welds then our discourse tends to degenerate into
the model together. (Like Dubin, I do not distin- heated methodological debates over how fast
guish between a model and a theory.) we are traveling. To avoid vacuous discus-
The central question addressed here is: Why sions, propositions should be well grounded in
should colleagues give credence to this particu- the Whys, as well as the Hows and the Whats.
lar representation of the phenomena? The an- To summarize thus far: What and How de-
swer lies in the logic underlying the model. The scribe; only Why explains. What and How pro-
soundness of fundamental views of human na- vide a framework for interpreting patterns, or
ture, organizational requisites, or societal pro- discrepancies, in our empirical observations.
cesses provide the basis for judging the reason- This is an important distinction because data,
ableness of the proposed conceptualization. whether qualitative or quantitative, character-
During the theory-development process, logic ize; theory supplies the explanation for the char-
replaces data as the basis for evaluation. Theo- acteristics. Therefore, we must make sure that
rists must convince others that their propositions what is passing as good theory includes a plau-
make sense if they hope to have an impact on sible, cogent explanation for why we should ex-
the practice of research. If the theoretical model pect certain relationships in our data. Together
is a useful guide for research, by definition, all these three elements provide the essential ingre-
the relationships in the model have not been dients of a simple theory: description and expla-
tested. If all links have been empirically verified, nation.
the model is ready for the classroom and is of An additional comment about the use of prop-

491
ositions is in order. Not all bona fide theoretical are embedded and must be understood within a
contributions require propositions, and all pa- context. Therefore, authors of inductively gener-
pers need not follow the same format. However, ated theories have a particular responsibility for
when the purpose of a paper is to present a new discussing limits of generalizability.
theoretical position or to call into question the Although it is important for theorists to be sen-
fundamental structure of an existing theory, re- sitive to context, the Who, Where, and When of
searchable propositions are very useful. They a theory are typically discovered through sub-
force the author to think about the concrete ap- sequent tests of the initial, rudimentary theoret-
plications of new or revised thinking, and they ical statement (What, How, Why). In the process
increase the likelihood that subsequent research of testing these ideas in various settings, we dis-
will constitute valid tests of the author's core ar- cover the inherent limiting conditions. In the ab-
guments. If propositions are used, they should sence of this breadth of experimental evidence,
be limited to specifying the logically deduced we must be realistic regarding the extent of a
implications for research of a theoretical argu- theorist's foreknowledge of all the possible limi-
ment. (Some authors mistakenly use proposi- tations on a theory's applicability.
tions to summarize a body of literature.)
Who, Where, When. These conditions place
limitations on the propositions generated from a What Is a Legitimate, Value-Added
theoretical model. These temporal and contex- Contribution to Theory Development?
tual factors set the boundaries of generalizabil-
ity, and as such constitute the range of the the- Most organizational scholars are not going to
ory. Scholars who study the effects of time and generate a new theory from scratch. Instead,
context on people and events keep asking nag- they generally work on improving what already
ging questions like, Would your predictions hold exists. In that context, it is often difficult to judge
in Japan, with a blue-collar population, or what constitutes enough of a contribution to
across time periods? Unfortunately, few theorists warrant publication in a theory journal like
explicitly focus on the contextual limits of their AMR. Nevertheless, the constituent elements of
propositions. In their efforts to understand a so- social theories described in the preceding sec-
cial phenomenon they tend to consider it only in tion suggest a set of criteria for making editorial
familiar surroundings and at one point in time. judgments.
Although it is unfair to expect that theorists What and How. Although, in principle, it is
should be sensitive to all possible boundary con- possible to make an important theoretical con-
straints, clearly there is value in conducting tribution by simply adding or subtracting factors
some simple mental tests of the generalizability (Whats) from an existing model, this process sel-
of core propositions. For example, theorists dom satisfies reviewers. The additions or dele-
should be encouraged to think about whether tions typically proposed are not of sufficient
their theoretical effects vary over time, either be- magnitude to substantially alter the core logic of
cause other time-dependent variables are theo- the existing model.
retically important or because the theoretical ef- One way to demonstrate the value of a pro-
fect is unstable for some reason. posed change in a list of factors is to identify how
Sensitivity to context is especially important this change affects the accepted relationships
for theories based on experience. According to between the variables (Hows). Just as a list of
the contextualist perspective (Gergen, 1982), variables does not constitute a theory, so the ad-
meaning is derived from context. That is, we un- dition of a new variable to an existing list should
derstand what is going on by appreciating not be mistaken as a theoretical contribution.
where and when it is happening. Observations Relationships, not lists, are the domain of the-

492
ory. As Poincare (1983) so aptly noted, "Science cient to point out limitations in current concep-
is facts, just as houses are made of stone.... tions of a theory's range of application. For ex-
But a pile of stones is not a house, and a collec- ample, discovering that a mainstream person-
tion of facts is not necessarily science." There- nel selection model has low predictive validity in
fore, theoretical insights come from demonstrat- a military setting does not by itself constitute a
ing how the addition of a new variable signifi- theoretical contribution. In addition, theorists
cantly alters our understanding of the phe- need to understand why this anomaly exists, so
nomena by reorganizing our causal maps. For that they can revise the How and What of the
example, the addition of "growth-need strength" model to accommodate this new information.
to job-design theories transformed extant views Conversely, applying an old model to a new
and altered research practice (Hackman & setting and showing that it works as expected is
Lawler, 1971). not instructive by itself. This conclusion has theo-
Important changes in a theory's What and retical merit only if something about the new
How are frequently stimulated by surprising re- setting suggests the theory shouldn't work under
search results. In the process of gathering either those conditions. In other words, it is preferable
quantitative or qualitative data, scholars are of- to investigate qualitative changes in the bound-
ten confronted with an inconsistency between aries of a theory (applications under qualita-
their observations and conventional wisdom. tively different conditions), rather than mere
Although contrary results are frequently dis- quantitative expansions. Two examples of this
counted by theorists on the basis of measure- approach are Maruyama's examination of
ment error, ongoing challenges to outmoded Western theories of management in the context
thinking about motivation (Organ, 1988) demon- of Eastern culture (1984) and Whetten's exami-
strate that sufficient data can be persuasive. nation of growth-oriented organizational theo-
Why. This is probably the most fruitful, but ries under conditions of decline (1980).
also the most difficult avenue of theory develop- The common element in advancing theory de-
ment. It commonly involves borrowing a per- velopment by applying it in new settings is the
spective from other fields, which encourages al- need for a theoretical feedback loop. Theorists
tering our metaphors and gestalts in ways that need to learn something new about the theory
challenge the underlying rationales supporting itself as a result of working with it under different
accepted theories. This profound challenge to conditions. That is, new applications should im-
our views of human nature, group develop- prove the tool, not merely reaffirm its utility.
ment, organizational transactions, and so forth, Three broad themes underlie this section:
generally precipitates a broad reconceptualiza- First, proposed improvements addressing only a
tion of affected theories. single element of an existing theory are seldom
This aspect of conceptual development is par- judged to be sufficient. Therefore, a general rule
ticularly critical, and generally overlooked. of thumb is that critiques should focus on multi-
Theories often are challenged because their as- ple elements of the theory. This approach adds
sumptions have been proven unrealistic (gener- the qualities of completeness and thoroughness
ally by work imported from other areas). Al- to theoretical work.
though it is just as difficult to build consensus Second, theoretical critiques should marshal
around paradigmatic truth as around empirical compelling evidence. This evidence can be log-
fact, nonetheless, recent macro theoretical de- ical (e.g., the theory is not internally consistent),
velopments involving ecology and economics empirical (its predictions are inconsistent with
demonstrate the salience of this approach (Han- the data accumulated from several studies), or
nan & Freeman, 1989; Ouchi & Barney, 1986). epistemological (its assumptions are invalid-
Who, When, Where. Generally, it is insuffi- given information from another field).

493
 Third, in general, theoretical critiques should itly laid out, or easily, reliably deduced)? Does
propose remedies or alternatives. Although we the paper go beyond making token statements
can think of classic critiques in the history of sci- about the value of testing or using these ideas?
ence that stood on their own merits, the typical Are solutions proposed for remedying alleged
debate in our field is less clear cut. Conse- deficiencies in current theories? These questions
quently, critics should share responsibility for are less appropriate for the rare, highly concep-
crafting improved conceptualizations. Other- tual papers aimed at changing the way organi-
wise, it is difficult to know whether the original is zational scholars think, in general. However,
indeed inferior, or simply the best we can do in the purpose of the standard theoretical paper
a very complex world. should be to alter research practice, not simply
to tweak a conceptual model in ways that are of
What Factors Are Considered in little consequence.
Judging Conceptual Papers? 3. Why so? Are the underlying logic and sup-
porting evidence compelling? Are the author's
Thus far we have examined the inherent mer- assumptions explicit? Are the author's views be-
its of a theoretical argument. In addition, re- lievable? Theory development papers should be
viewers consider other factors, including clarity built on a foundation of convincing argumenta-
of expression, impact on research, timeliness, tion and grounded in reasonable, explicit views
and relevance. of human nature and organizational practice.
The following list of seven key questions, 4. Well done? Does the paper reflect sea-
roughly in the order of frequency in which they soned thinking, conveying completeness and
are invoked, summarizes the concerns raised thoroughness? Are multiple theoretical ele-
most frequently by our reviewers. These ques- ments (What, How, Why, When-Where-Who)
tions cover both the substantive issues discussed covered, giving the paper a conceptually well-
in the first two sections as well as several for- rounded, rather than a superficial, quality? Do
matting concerns. Together they constitute a the arguments reflect a broad, current under-
summary answer to the broad question, What standing of the subject? If propositions are in-
constitutes a publishable theory paper? cluded, are they used properly? Does the argu-
1. What's new? Does the paper make a signif- ment have any glaring logical flaws? Does it ap-
icant, value-added contribution to current think- pear that the author has developed these
ing? Reviewers are not necessarily looking for thoughts over an extended period of time, in-
totally new theories. However, modifications or formed by extensive peer input?
extensions of current theories should alter schol- 5. Done well? Is the paper well written? Does
ars' extant views in important ways. Proposed it flow logically? Are the central ideas easily
changes can be calibrated in terms of scope and accessed? Is it enjoyable to read? Is the paper
degree. Scope tends to reflect the level of theo- long enough to cover the subject but short
rizing (general versus middle level), while de- enough to be interesting? Does the paper's ap-
gree reflects the radicalness of the proposal. In pearance reflect high professional standards?
general, scope (how much of the field is im- Are the paper's format and content consistent
pacted) is less important in determining the mer- with the specifications in the Notice to Contrib-
its of a contribution than is degree (how different utors?
is this from current thinking). 6. Why now? Is this topic of contemporary in-
2. So what? Will the theory likely change the terest to scholars in this area? Will it likely ad-
practice of organizational science in this area? vance current discussions, stimulate new dis-
Are linkages to research evident (either explic- cussions, or revitalize old discussions? Review-

494
ers give low marks to papers they perceive are lems. Otherwise, they are more appropriate for
redundant, unconnected, or antiquated. a discipline-based journal.
7. Who cares? What percentage of academic In conclusion, the theory-development pro-
readers are interested in this topic? A paper may cess and criteria for judging theoretical contri-
be technically adequate but inherently uninter- butions need to be broadly understood and ac-
esting to most of our broad audience. Papers cepted so that editors and contributors can com-
written on topics with narrow appeal are typi- municate effectively. Hopefully this brief article
cally held to a higher standard for Criteria 1 and will facilitate that process. I urge readers to as-
2; that is, they are expected to make a more sist in the further development of frameworks for
significant contribution to current thinking and describing and enhancing these important
research practice. In general, even highly spe- scholarly activities. Papers on the process of
cialized papers should be linked to core man- building new and improving current theories
agement or organizational concepts and prob- are always welcome.

References
Dubin, R. (1978) Theory development. New York: Free Press. Organ, D. W. (1988) A restatement of the satisfaction-
Gergin, K. (1982) Toward transformation in social knowl- performance hypothesis. Journal of Management, 14,
edge. New York: Springer-Verlag. 547-558.
Hackman, J. R., & Lawler, E. E. 11 (1971) Employee reactions Ouichi, W. G., & Barney, J. B. (1986) Organizational eco-
to job characteristics. Journal of Applied Psychology nomics. San Francisco: Jossey-Bass.
Monograph, 55, 259-286.
Poincare, J. H. (1903) La science et 1 hypothese [The science
Hannan, M. T., & Freeman, J. (1989) Organizational ecology.
and the hypothesis]. Paris: E. Flammarion.
Cambridge, MA: Harvard University Press.
Maruyama, M. (1984) Alternatives, concepts of manage- Whetten, D. A. (1980) Organizational decline: A neglected
ment: Insights from Asia and Africa. Asia Pacific Journal topic in organizational science. Academy of Management
of Management, 1, 100. Review, 5, 577-588.

I would like to thank Andrew Van de Ven, Richard

Klimoski, Kim Cameron, Huseyin Leblebici, Graham
Astley, Jack Brittain, Arthur Bedeian, Donald Hellrei-
gel, and Steven Kerr for their suggestions and com-
ments.

495