1
What Would John Dewey Say About the Fallacies of Grounded Theory?
There is no âgroundâ and no âtheoryâ in Grounded Theory (GT)! Most every dissertation or qualitative study article I review claims to do GT, but does not notice that GT has changed radically over the years. Now I ask, what GT wave are you doing? GT has come under widespread criticism in every top-tier journal of management and organization research. Each of the first three GT waves has an epistemic fallacy. Here, I would like to show how John Deweyâs pragmatism philosophy offers a way forward to GT. I will briefly review how GT has developed in three waves (or paradigm shifts) and propose a 4th wave that is ontological, rather than merely epistemological as is 1st wave GT, or positivistic coding as is 2nd wave GT, or merely social constructivist as is 3rd wave GT. In 4th wave GT, Dewey is among several scholars who bring an ontological grounding and a theory combined with praxis (Heidegger, Bhaskar, Deleuze, Ĺ˝iĹžek, Barad, to name a few).
What would Dewey say about Grounded Theory? I believe he would say that it is not grounded in Nature or in the paradigm shift from Cartesian duality and Newtonian mechanistic physics, and that its manner of reflexivity separates theory from praxis. Deweyâs (1925) Experience & Nature grounds pragmatism in material nature. Deweyâs (1929) Quest for Certainty grounds pragmatism in what he calls a new Copernican Revolution that Werner Heisenbergâs principle of indeterminacy brings to Cartesian dualism and to Newtonian mechanistic physics. Rorty (2010: 152) says, âPhilosophers working after âthe linguistic turnâ (no matter how it is defined) still have great deal to learn about experience and language from Peirce, James, Dewey, and Mead.â âDewey was constantly criticized, from the Platonist right, for being reductionistic and scientistic, inattentive to our needs for âobjective valuesââ (Rorty, 2010: 166). Rorty points out that Dewey (1925 Experience and Nature) was critical of Plato being âa spectator of all times and eternityâ (Rorty, 2010: 73). Rather, Dewey used pragmatist philosophy as an instrument for social change by focusing on what is observation in his naturalism.
There is no âgroundâ and no âtheoryâ in current renditions of GT research. Why? Because GT is not based on any sort of âgroundâ ontology, and what passes for âtheoryâ is rooted in social constructivist standpoints, of gathering and sifting points of view (epistemic relativity). Remedy? To develop dialectic and multiplicity sorts of ontologies that is âgroundedâ in existential meaning of Being, worked out in spatiotemporal and material mattering, a.k.a., spacetimemattering in either Hegel (negative dialectics ontology), Being-in-the-world ontology (Heidegger), agential realism (Barad), pragmatic realism (Peirce), critical realism (Bhaskar), or Deleuze multiplicity series.
Pushkala Prasad (2005: 289, 291) helps launch our current discussion points. Prasadâs (2005: 26) critique of GT is: âwhile in some ways it bears resemblance to symbolic interactionism, the method favors a far more non-theoretical bent.â This is ironic since co-founder Anselm Strauss began his academic career as a symbolic interactionist. Prasad observed that what we called 3rd wave GT remains strongly committed to social constructivism, while favoring âa much more positivist set of ontological assumptions, believing in the existence of a much more concrete reality that ultimately anchors or âgroundsâ all research effortsâ (IBID.). Our review of 4th wave ontological options for future GT work echoes this concern.
Reviewers of management and organization journals, in this still comparatively young field, oftentimes do not have access to experts in multiple kinds of qualitative research traditions. For quality control, those reviewers do serve as gatekeepers, which is âparticularly the case with the North American segment of management and organization studies, which has, until relatively recently been dominated by industrial and organizational psychology and economics.â (p. 289). In those disciplines the preference is for positivistic methods of survey-based research and some experimental lab studies with hypothesis testing.
There are major critiques of GT in just about every top-tier journal in organizational research methods. Positivist qualitative research, despite the popularity of GT, remains on the margins of leadings journals such as Academy of Management Review (e.g. Gephart, 2004; Suddaby, 2006; Eisenhardt & Graebner, 2007; Siggelkow, 2007; Pratt, 2009), Administrative Science Quarterly (Sutton & Staw, 1995; Brown & Eisenhart, 1997), and lets add Organizational Research Methods (e.g. Fendt & Sachs, 2008).
Eisenhardt and Graebner (2007: 29), for example, point out the challenges of GT to reviewers and readers:
Coping with the trade-off between rich story and well-grounded theory is easier to do in a multicase book or a single-case paper. But in journal articles, multicase researchers face a particularly difficult trade-off between theory and empirical richness. It can be especially challenging to satisfy readers who expect the extensive narratives of single-case research. They ask, Whereâs the rich story?
GT is in big trouble. Prasadâs claim that not much GT has been published in the flagship academy journals, because GT is both not theory, and what constitutes ground is not readily accepted, unless the authors are able to mimic positivism (p. 298), seems justified. GT is being critiqued in every top-tier management and organization research journal. Suddaby (2006: 634) contends GT does not have to be a positivist routine or formulaic application of coding in a hierarchical set of categories routing to data done in theoretical sampling. Eisenhardt and Graebnerâs (p. 30) critique of GT is that it can move from what we call waves one and two to a 3rd wave suited to intersubjective experience. The problem is how reviewers and readers can judge research quality given the wide variety of GT approaches. Their advice:
preempt misunderstanding by engaging in systematic data collection and theory development processes that are reported with transparent description, particularly regarding how the theory was inducted from the data (e.g., description of cross-case comparison techniques)âŚ. Somewhat surprisingly, single cases can enable the creation of more complicated theories than multiple cases, because single-case researchers can fit their theory exactly to the many details of a particular case. In contrast, multiple-case researchers retain only the relationships that are replicated across most or all of the cases.
(IBID)
Both critiques suggest GT is not living up to it potential, and yet, more and more GT articles are being submitted for review.
Gephart (2004), by contrast (also AMJ), observes that while most authors claim to have used GT, it is less common to actually see detailed applications of it in our field. Fendt and Sach (2008: 430) in ORM offer perhaps the most devastating critique of GT, and conclude in their review, âthe very strengths of GT ⌠run the risk of being underminedâand thus the quality of such research impairedâby an overly orthodox application of its rigorous objectification procedures.â They echo our conclusion that a new (4th) wave, more Dewey, Peirce approach to GT is needed but do not review Heideggerâs or Bhaskarâs or Ĺ˝iĹžekâs alternative approaches: âdepends upon inappropriate models of induction and asserts from them equally inappropriate claims to explanation and predictionâ; and suggest, citing Peirce, that everyday induction be called âabductionâ and expand on this critique and in particular on the distinction drawn between âlooser and tighter kinds of induction.â They also cite Suddaby (2006), who called for a turn Peircean abduction, and its difference from GT uses of it. GT is accused of being a âGrand Theoryâ pretense, a claim to build theory inductively, or to test grand theory deductively in the data, in building block coding (p. 461). They raise the question, does GT really offer theory, it is it just interpretation? Or is GT a way of making inductive argument in a new label? With regard to what we call 3rd wave GT, Fendt and Sach (2008: 447) accuse GT of a kind of inferiority complex or false pride by its proponents âan attempt to justify an essentially interpretivist method vis-Ă -vis a research world still prejudiced in favor of positivism.â
For ASQ, Sutton and Staw (1995) attempt to answer the question, what constitutes a theory? They assert that, at ASE, theory is not data, LISRL is not theory, references and lists of hypotheses are not theory, and give this notice: âIf manuscripts contain no theory, their value is suspectâ (p. 371). With regard to GT, in their section data is not a theory, they declare, just as in quantitative studies, âqualitative data must develop causal arguments to explain why persistent finding have been observed if they wish to recite papers that contain theory (Glaser and Strauss, 1967)â (Sutton & Staw, 1995: 374â375). It is not at all clear that GT is a theory that develops causal arguments in its first three waves, or in the 4th wave options we have included in this review. Undaunted by this standard for theorizing, Brown and Eisenhardt (1997: 2) publish their GT case study in ASQ, and get around the gatekeepers by saying itâs not a full-fledged âextant theory.â
Meanwhile, for qualitative work that is nonpositivistic, âthese reviewers persist in raising questions about interrater reliability, generalizability, and replication in situations in which they have no place at all because the researcherâs interest in interpretation, historical analysis, or culture critiqueâ (IBID.). The result is many authors diluting the rigor of their own qualitative work in order to placate reviewers who lack knowledge of nonpositivistic traditions. âDespite these obstacles, superior qualitative research continues to be published even in mainstream American management journalsâ (p. 290).
What about Prasadâs claim that, by comparison to American management journals, European journals are staffed by reviewers well versed in multiple qualitative traditions and diverse philosophies? I did a review of this question in several European journals: Journal of Management (Orton, 1997; Turner, 1983; Rothaermel & Sugiyama, 2001), Human Relations (Bate, 1997; Berends, Boersma, & Weggeman, 2003), Organization Studies (Wilson, 2002; Pozzebon & Pinsonneault, 2005). I found mixed results. In Scandinavian JOM, Ortonâs (1997) study makes a move from inductive GT to iterative GT, in a process theory. Rothaemel and Sugiyamaâs (2001) JMS study only mentions doing a GT case study, but does not do any kinds of 4th wave options we propose. By contrast Turner (1983) says there is no orthodox GT, and does not want to be bound by one. He uses GT to deal with large amounts of unstandardized data from three small studies of different kinds of data. In Human Relations, Bate (1997) asks whatever happened to organizational ethnography, as alternative to GT. Berends, Boersma, and Weggeman (2003) do a study they claim is GT. In Organization Studies, Pozzebon and Pinsonneault (2005) claim GT is either inductive or deductive in application to structuration theory. Wilson (2002) does a book review of Karen Lockeâs (2001) GT book, concluding that GT has become synonymous with qualitative studies in management research. Locke fortunately does not agree, and divides GT in modernist (positivist) and interpretivist (poststructuralist) approaches, in a 3rd wave move.
In sum, it would appear that both American and European journals are resistant to GT (for good reasons), and what is published in the name of GT, as Gephart concluded, is not as rigorously conducted, as for example in nursing journals (e.g. Annells, 2006) or health studies by GT founders (Glaser & Strauss, 1965a; Corbin & Strauss, 1988).
Let me put our 4th wave challenge to GT in a wider context. I am writing a new book on ontological research methods that I believe go beyond the first three waves of GT. I will introduce 10 exemplars of these 4th wave approaches to GT (Table 1.1), then return to what Dewey would say about GT.
I want to turn now to what are the key problems with GT and how Dewey would likely address them.
What Would Dewey Say About Grounded Theory?
Deweyan naturalism was shoved aside by the dogmatism of logical empiricismâs âquest for certaintyâ as Dewey called it (which decades later also infected 2nd and 3rd wave GT).
1st wave GT (1967â1993) commits âinductive fallacyâ by doing qualitative methods to generate theory propositions out of practice that go untested and ignore historical context (Glaser & Strauss, 1967). It fails Karl Popperâs critique of inductive logic for failing to do falsification or verification of inductive propositions. Glaser and Strauss (1967: 2â3) say âthe discovery of theory from data systematically obtained from social researchâ is an idea they picked up from Merton, for whom the meaning of ground is non-theoretical social practice, out of which theory can be generated. 1st wave GT adopted inductive logic mixed with ââDescartesâ spectatorial account of knowledge,â and introspective psychology, resulting in an epistemology of sensemaking (Rorty, 2010: 74). Dewey rejected the sensorium of perception (i.e. sensemaking perception) as the final inductive methodological step. Deweyâs (1929) Quest for Certainty puts his objection to what we now call Weickian sensemaking this way:
In traditional empiricism the test is found in sensory impressions. For objective idealism, reflective inquiry is valid only as it reproduces the work previously effected by constitutive thought. The goal of human thinking is approximation to the reality already instituted by absolute reason.
(p. 88)
Dewey (1910: 86) defines âScientific inductionâ as âall the processes by which the observing and amassing of data are regulated with a view to facilitating the formation of explanator...