Evidence-based dialectics

Abstract

‘Evidence-based policy’ and ‘evidence-based management’ are increasingly popular ways of describing the relationship between research and practice. The majority discussing the evidence-based approach have tended to be in favour: here, ‘believers’. Yet this approach has also attracted critics: ‘heretics’. Understanding of such a division can be enhanced by dialectics: a process which tries to destabilize, reconcile or transcend apparent opposites. This divide is not simply a consequence of differences relating to epistemology, but also aesthetics: a set of reactions to the world seen as art. So, to analyse this divide requires a correspondingly rich model of dialectic. Nietzsche’s Birth of Tragedy offers this in its account of Apolline and Dionysian responses to the world. Dialectics supports a move beyond synchronous critique, and allows speculation as to the future development of the evidence-based approach.

A great many social scientists now advocate ‘evidence-based management’ and ‘evidence-based policy’ (e.g. Ashkanasy, 2007; Ashworth et al., 2010; Davies et al., 2000; Lawler, 2007; Pfeffer and Sutton, 2006; Reay et al., 2009; Tenbensel, 2004; Tranfield et al., 2003; Walshe and Rundall, 2001; Young et al., 2002). ¹ Their inspiration comes from the success and popularity of evidence-based medicine, ‘the conscientious, explicit and judicious use of current best evidence’ (Sackett et al., 1996: 71). The Sackett et al. definition is often echoed and adapted in management and policy studies, for instance, ‘evidence-based methods concern research synthesis—looking at a body of literature or empirical studies [using] systematic review or metaanalysis’ (Shillabeer et al., 2011: 7).

‘Using evidence-based medicine as an exemplar’, Rousseau’s 2005 Presidential Address to the Academy of Management, attributed great power to the evidence-based approach, and drew a remarkably close comparison between medicine and management:

If you are wondering what physicians did before [evidence-based medicine], the answer is what managers are doing now, but without medicine’s added advantages from common professional training and malpractice sanctions. (2006a: 258)

Rather than physics envy, Rousseau and other advocates of evidence-based approaches seem to have contracted physician envy (Morrell, 2008). This article explores the relationship between those enchanted by the evidence-based approach and their critics.

The evidence-based movement and its critics

The most readily apparent problem for advocates of evidence-based management and policy is that they do not enjoy the same consensus about ‘best evidence’ as those practising evidence-based medicine. Champions of evidence-based management do not even agree about the role of the most fundamental technique in evidence-based medicine: the systematic review. The best known book on evidence-based management (Pfeffer and Sutton, 2006), makes no reference to ‘systematic review’. In contrast, Briner et al. (2009: abstract) are clear systematic reviews are ‘a cornerstone’ of evidence-based management. There is considerable diversity among both advocates for evidence-based principles (as cited above) and critics (Clegg, 2005; Evans and Pearson, 2001; Fox, 2003; Hammersley, 2001; Hewison, 2004; Hope, 2004; Learmonth, 2008, 2009; Learmonth and Harding, 2006; McLaughlin, 2001; Morrell, 2008). Advocates acknowledge some problems, and engage in criticism, but are more interested in improving ‘how’ the evidence-based approach works than entertaining radical doubt. ²

To depict and summarize the division between advocates of evidence-based approaches in management and policy, and their critics, I will label advocates ‘believers’ and critics ‘heretics’. This choice of terms is partly in a spirit of play and irreverence. Play provides a counterpoint to the dry, utilitarian tone of much of the debate. Irreverence is consistent with Organization’s tradition as a journal whose point is, ‘to be provocative, and critical of power in its varied manifestations’ (Parker, 2010: 6). The labels are also consistent with how both advocates and critics describe the phenomenon. Advocates extol the evidence-based ‘movement’ (Pfeffer and Sutton 2006; Young et al., 2004), occasionally with religiosity, ‘turning evidence-based management from a practice of a prophetic few into the mainstream’ (Rousseau, 2006a: 266). A ‘heretic’ has characterized the fixation on procedure and standardization as, ‘science’s equivalent of a religion’s holy book’ (Fox, 2003: 85). The labels denote ideological aspects to a project typically presented by its advocates as simply raising epistemological problems (e.g. Tranfield et al., 2003). Commitment to an evidence-based approach is not simply dispassionate choice, but an expression of faith, or even enchantment with medical progress. Also, in terms of published work, ‘heretics’ is apt since these are definitely in the minority.

‘Heretics’ are not heretical against evidence, but they doubt the claims of ‘believers’ who promote an orthodoxy that is presented as superior to all other modes of knowledge production, and suitable to all social problems. This is not simply an epistemological debate since there is an interrelation between knowledge and power (Foucault and Gordon, 1980), and to claim ownership of the method for determining ‘best evidence’ shapes processes of privilege and exclusion (Cooke, 2008). As Rousseau (2006b: 1091) acknowledges (though only after being criticized by Learmonth, 2006), ‘[p]olitics are real and pervasive. It would be naive to think otherwise. Developing and implementing an evidence-based approach to managing and organizing can affect how funding, legitimacy, and influence are allocated for scholars, practitioners, and teachers’. Since this movement is, and will remain influential, and since it clearly has such political consequences, it is valuable to explore this divide and speculate about the character and likely trajectory of developing arguments relating to research and evidence.

An appropriate way of doing this is through dialectics, used here to describe an attempt to see whether apparently opposing ideas can be reconciled, or whether dichotomies can be transcended or destabilized. It is beyond the scope of the article to give even a partial history of dialectics, it is sufficient to say there are many nuances to the term, and many theorists associated with its use (Bhaskar, 2008 and Fuchs, 2009 offer recent reviews). In search of a specific, suitable model of dialectic, I draw on a richly influential account of two basic kinds of response to the world. These are different ways of knowing, whose essential character is aesthetic and political as well as epistemological. This is the dialectic between Apollo and Dionysus in Nietzsche’s Birth of Tragedy. This dialectic is useful to develop an account of evidence-based ‘believers’ and ‘heretics’ that throws into relief, and at times destabilizes some of the contradictions across and within both categories of response. It allows a reinterpretation, and rich depiction of both kinds of reaction. This contributes to the literature by moving beyond considerations of technique or epistemology, or empirical recounting of the successes or failures of evidence-based approaches, to a broader aesthetic and ethico-political assessment. In doing this, it tries to go beyond retrospective or synchronous critique to give a sense of the likely character of this unfolding debate.

The Birth of Tragedy

The Birth of Tragedy (from now, Birth) is the earliest, and formally most conventional, book of Nietzsche’s, but it is also the hardest to characterize or summarize. It is closer to a traditional scholarly text in style than his later works for instance (Hollingdale, 1977). Still, its reception by some was all the more hostile for that: the formal similarities it shared with other academic essays jarred with its fiercely outlandish content. For an academic essay in philology it is profoundly anti-academic, disregarding the sensitivities of those in Nietzsche’s (short-lived) disciplinary home.

Birth was reviled by his fellow scholars at the time of publication and many subsequent commentators portray it as somewhat out of place even within Nietzsche’s works. Hollingdale sidelines it as ‘immature’ in his otherwise comprehensive overview (Hollingdale, 1977: 8). Tanner suggests it defies conventional analysis or discussion, ‘the most we can hope for is accounts which preserve as much of the excitement as possible while admitting the confusion’ (Tanner, 1993: xxvii). Porter (1995: 495) describes it as Nietzsche’s, ‘most impenetrable text’. Nietzsche himself in his Attempt at Self Criticism states Birth is an, ‘impossible book—badly written, clumsy and embarrassing’ (Nietzsche, [1886]/1993: 5). At the same time, he identifies within it the origins of three of his best known works, the Genealogy of Morals, Beyond Good and Evil and Thus Spake Zarathustra. In Twilight of the Idols he claims it as, ‘my first revaluation of all values’ (Nietzsche, [1888]/1990: 121) and it encapsulates his stance towards epistemology and method, later given fuller expression in The Gay Science, ‘to see science under the lens of the artist, but art under the lens of life’ ([1886]/1993: 5).

It is possible that this integration with his later works was partly an attempt by Nietzsche to rehabilitate Birth since he was certainly wounded by its hostile reception (Tanner, 1994). However, it is also clear that at the crux of Birth is the argument that we perhaps most associate with Nietzsche: the world and our experiences only make sense as an aesthetic phenomenon. Birth is, ‘a set of intense reactions to art and the world, and really to the world seen as art’ (Tanner, 1993: xiii).

Though avowedly and up front a contribution to ‘the science of aesthetics’ (Birth: 14), it is an intensely idiosyncratic account that calls science into question. As cultural history it is a ‘remarkable doctrine of antiquity’ (Losev, 1993: 27 in Rosenthal, 2004); but Nietzsche intended it to inform analysis of the contemporary political climate, since it was written during, and published shortly after, the Franco-Prussian war. He suggests of his audience in the preface that, ‘were they really to read this essay [they] would be astonished to discover the seriously German problem that we are dealing with’ (Birth: 3).

Simultaneously a work of politics, metaphysics and aesthetics, Birth offers a powerful and rich dialectic that establishes two stark categories of reactions to the world: the Apolline and the Dionysian. These categories describe simultaneously fundamental differences in, ‘genre, psychology, artistic style and ontology’ (Force, 1982: 179). Nietzsche plays with, subverts and distorts these categories, at the same time preserving them as separate, defying conventional (Socratic, Aristotelian, Hegelian, Marxist) dialectics (Clegg, 1972). All these features make Birth both fascinating and perplexing. They also make it an appropriate model for examining the division between ‘believers’ and ‘heretics’.

‘Believers’ and ‘heretics’

In discussing the ‘world seen as art’ Nietzsche differentiates between the Apolline and the Dionysian. Apollo (the God of medicine), is the God of light, order and individuation; Dionysus is the God of dance, of the dithyrambic chorus, and of sublimation to the collective Will. Applying this to the evidence-based movement, ‘believers’ worship Apollo; ‘heretics’ worship Dionysus, and yet there is an interdependence between the Dionysian and Apolline.

In terms of art forms, the Apolline is associated with (classical) sculpture, with its sharp divisions between the artwork and the world, and between different works of art. As representations every sculpture is a testimony to a view of the world populated by discrete forms, a principle of individuation. ‘Apollo … holds sway over the beautiful illusion of the inner fantasy world … the glorious divine image of the principium individuationis’ (Birth: 16). This principium individuationis is Schopenhauer’s description of the barrier we face in appreciating a collective Will, when all our experience comes ‘parcelled up’ (Tanner, 1994: 119). In following the principium individuationis, ‘believers’ worship Apollo. In search of clarity and precision, the evidence-based approach involves accumulating individual studies, and aspires to evaluating each according to the strictures of technique and to quality standards. The more precisely such studies reflect particular aspects of the social world, the greater confidence one can have in their efficacy when it comes to implementing policy. Consider for instance the following account of the systematic review (Young et al., 2002: 219–20):

the systematic review … starts with a clear question to be answered or hypothesis to be tested. The reviewer strives to locate all relevant published and unpublished studies to limit the impact of publication and other biases. Where it is not possible to locate all the research in a given area, the review should explain how studies were identified and obtained, and highlight any known gaps. In deciding which studies to include and which to exclude the reviewer confronts the issue of selection bias. The quality of these studies is then assessed by examining, in a systematic manner, the methods used in primary studies, investigating potential biases in those studies and identifying the sources of any heterogeneity among the study results. The results are then synthesised, with conclusions based only on those studies that are methodologically sound. Thus, the systematic review takes a wide-ranging and comprehensive approach to searching for relevant research. It uses the technology now available to carry out global searches of the research databases. It aims to identify all the relevant research, not just the best known, well promoted and successful.

The term ‘systematic’ is principally justified by the manner in which the reviewer proceeds, stage by stage, with full transparency and explicitness about what is (and what is not) done, typically using a protocol to guide the process.

Apolline themes here include: clarity, transparency, elimination of specific biases, use of a protocol, categorical determination of relevance, sharp boundaries of inclusion and exclusion, progression in terms of an ordered sequence of steps, bringing forth what is hidden into the light, the use of technology and separation of the researcher from the objects of their research. ‘Believers’ have faith that this process enables the identification of discrete gaps in our knowledge, and offers the promise of accounting for such gaps.

The Dionysian signals sublimation of the individual to a common, indivisible consciousness: the collective frenzy of the dance. This is a reaction to the world that sees reality as undifferentiated and where the primal truths of reality and consciousness reflect the world as Will (Schopenhauer, [1819]/1969). ‘Heretics’ challenge ideas relating to divisibility, reject any single perspective on the world, and are hostile to claims of unassailable expertise, inviolable technique, or unrivalled clarity. For instance, Learmonth and Harding (2006: 257) describe how the critical geographer Lefebvre, ‘dismissed the possibility of a transparent, pure and neutral space’, which is central to their critique of the evidence-based approach. In discussing representational or lived space (social spaces), they go on (pp. 258–261):

space is something that is actively produced through both social interactions and the impositions of discourses of space. For example, we make the lecture theatre or GP practice lived spaces by making them the places in which the performances of lecturing or consulting take place. Through this process we can call them the lecture theatre or surgery, arrange artefacts (including bodies) within them so as to conform to the lived aesthetic of each place, and direct ourselves within them to produce these spaces. Importantly for our analysis of evidence-based management, these lived spaces are highly complex—they are not simply juxtaposed but may be intercalated and in collision (1991, p. 88); or they may interpenetrate or superimpose themselves upon one another (1991, p. 85). Lefebvre’s lived space is therefore the outcome of past actions, it permits fresh actions to occur while suggesting or prohibiting others. In this sense then it is produced, but not necessarily within conditions of the participants’ own choosing; lived space is the bearer of norms and constraints, and so tends to be repressive (1991, p. 358) …

[E]vidence-based management, as it is currently constituted, claims to describe a world which does not exist in lived space: its world ‘exists’ only as a series of concepts. And, unfortunately, these conceptual spaces have little relationship to the lived, material spaces its proponents claim they represent. They claim to provide maps … But these maps are so devoid of an understanding of the very thing they claim to represent, that they are maps not of the lived world of people …

Dionysian themes here include an emphasis on collective norms and constraints, submission to the complexity, historicity and social nature of experience; and a series of concurrent concessions: to the primal, existential nature of being in the world, to the limitation of conceptual frameworks devoid of lived experience, and to the inevitable intertwining of action and space. These all deny the possibility of individuation and separateness.

Birth sheds light on the dichotomy between scientific and aesthetic responses to the world, and thus on the relationship between ‘theory’ and ‘evidence’, and the tribal reactions to appeals to evidence-based principles. Debate, certainly by ‘believers’, has been dominated by technical and epistemological considerations. The dialectic in Birth relates the archetypal response to the world as art and breathes life into some past familiar (unresolved) arguments about the nature of medical and management knowledge. It allows speculation as to the future development of the evidence-based movement because it foregrounds an aesthetic response to the world which expresses the intense, complex and unfolding relationship between two opposing categories of engagement: the Apolline and the Dionysian. To develop this, the following sections outline Apolline and Dionysian accounts of the relationship between theory and evidence.

The Apolline account of theory and evidence

The received normative account of ‘good’ theory in the sciences follows the standard science model (SSM) of conjecture and refutation (Popper, 1962). Many organizational scholars adopt this Apolline stance: good theory depends on repeated empirical tests (Eden, 2002; Neuliep, 1991) and (reading across from Popper’s criterion of refutation) on the pursuit of ‘falsifiability’ in organization theory (Bacharach, 1989: 500; Lee et al., 1999: 459; Mitchell and James, 2001: 543; Whetten 1989: 486–487; Worren et al., 2002: 1227). On this view, tests help to extend the boundaries within which theory is generalizable, so for instance an account of citizen’s ‘voice’ that was used successfully across different public services (health, education, housing, voting) would trump one which had been applied exclusively to one sector. ‘Believers’ sign up to this with a dramatic flourish, ‘single studies almost never matter in themselves’ (Briner et al., 2009: 24).

Apolline articles of faith within this account of good theory are: that we can pursue detached and objective research; that we can draw sharp delineations between initial and repeat tests of theory; that there is the eventual promise of clarity through synthesis; and that this in turn depends on processes of accumulation and commensuration of findings. Gooding (2000: 118) describes this paradigm as supporting a logic of discovery that overturns, ‘modes of knowledge based on superstition, the authority of tradition, of ancient texts, or of religious dogma’. So the attraction of these Apolline ideals, derived from the SSM, is that they banish Dionysian elements. For instance, Pfeffer advocates a paradigm in organizational studies because its absence, ‘has the undesirable property of permitting taste, virtually unconstrained by scientific norms and standards, to run rampant’ (Pfeffer 1997: 193–194).

One difficulty relating to ‘evidence’ in the social sciences is that a result of having sharply defined limits (such as which studies can be counted as evidence, or which are better evidence, or how the process of synthesis should work) is a loss of creative scope. ‘Style’ and rejection of consensus may be a source of strength and diversity (Van Maanen, 1995; Zald, 1996). Rejecting consensus can also enhance the ability to subvert potentially harmful hegemonies (Hassard and Kelemen, 2002; March, 2003; Weick, 1999). The SSM involves a rejection of the Dionysian and its association with ecstasy, the collective will and myth. It is an embracing of the rational, ordered world of Apollo where cool and detached objective analysis reigns, ‘that restraining boundary, that freedom from wilder impulses, that sagacious calm of the sculptor God’ (Birth: 16). But, as Pfeffer’s warning about the tendency of taste to ‘run rampant’ indicates, aesthetic aspects to theory building are not easily avoided. Editors (anonymously) describe replications studies as, ‘dull’, ‘boring’, stating readers ‘aren’t interested in them’, because they ‘don’t reflect cutting edge stuff’ (Neuliep and Crandall, 1991: 88). By themselves, Apolline strictures are insufficient because of the need to be appealing and enticing to a community.

Advocates of the evidence-based approach have conceded this in terms of a need to appeal to practitioners, ‘for academic evidence to be used by managers it needs to be rendered accessible, palatable, relevant and useful’ (Denyer and Tranfield, 2009). Yet for the movement to succeed there is clearly a need to appeal to the academic community as well and ‘believers’ seem unable or unwilling to acknowledge the evidence-based approach lacks style; indeed depends on trying to drive individual style out. For instance, the claim, ‘[i]t is the collective body of evidence we need to understand’ (Briner et al., 2009: 24), denies individual style on two levels. It leaves no room for differences in interpretation: ‘we’ understand ‘the’ evidence; and furthermore evidence only comprises studies that are aggregated.

Yet the cultural disincentives to replication studies and their acknowledged paucity, suggest the call to Dionysus is hard to resist. Because of this, even if we try to understand organizational research in terms of the SSM, the idea that there are consensual norms is illusory: there are no Apolline ‘hard edges’ (Tanner, 1994: 9). If there is doubt about the contribution of replication studies, there will surely be doubt about the novelty or originality of any assessment that is based on collating other people’s work. This will influence the view many scholars take as to the extent to which such work makes a contribution, a judgment ultimately resting on appeals to a collective idea of merit.

Collins (1982: 454) suggests (in the context of natural science) that, ‘an experiment is much better disconfirmed by what appears to be an isomorphous copy’. But in the social sciences, these Apolline ideals are fantasies. There are no ‘isomorphous copies’ in the social world. This basic ontological problem has been acknowledged, but side-stepped by ‘believers’ (Morrell, 2008). Yet because it undermines the assumption that phenomena can be precisely differentiated and compared, it is a challenge to any attempt to accumulate evidence. This reveals another incoherence within a purely Apolline account. For the evidence-based approach to work there has to be a commitment to realism (Pawson, 2006). This basic assumption about the nature of the world is a matter of faith rather than deduction or experience (it is not refutable). ³ ‘Believers’ can never take radical alternatives to realism seriously and this is a Dionysian sublimation to collective Will. The following section develops this idea.

The Dionysian account of theory and evidence

For critical theorists, the relationship between theory and evidence can be understood in terms of a set of socially embedded, and culturally contingent practices, rather than as a search for transcendent knowledge (Foucault, 2002). This is apparent in several strands of the literature within the sociology of scientific knowledge (SSK), where the status of science and nature of scientific knowledge are scrutinized. Summarily, though in sufficient detail to illustrate the suitability of invoking Dionysus, three important ideas from SSK are shown below in relation to: research; consumers of research; and the researcher. Each emphasizes Dionysian elements in the relationship between theory and evidence: an un-differentiated engagement with the world that challenges claims to the possibility of detached, objective research.

Research

Instead of assuming research offers the prospect of truth, several suggest it is more appropriate to see it as a local discourse (e.g. Latour, 1993; Potter, 1996). On these terms, the meaning of ‘evidence’ is situated and its status contestable: ‘there is no realm of universal, a-social rationality’ (Collins, 1982: 452). This informs understanding of research as an activity, of its products, and how they are deployed, ‘socially and culturally as a type of resource’ (Locke, 2001: 11). Ostensibly neutral methods like ‘evidence-based’ approaches are understood as performative or rhetorical. Rather than supporting discovery, they prop up a collective narrative about the relationship between theory and the world (Morrell, 2008). This (Dionysian) critique rejects the individuation of research practices. It understands calls to accumulate ‘the’ evidence as legitimating the practices and values of a community. These calls are discursive resources, part of an unconsciously held, shared myth about status and belonging. These are offerings to Apollo that come with aligning oneself with ‘hard’ science, irrespective of its appropriateness or relevance to social problems.

Consumers of research

Research can also be understood in terms of consumers and taste (Bourdieu, 1993); ‘consumption of knowledge fuels the creation of new knowledge [which] acquires its status as ‘knowledge’ only when selected for consumption’ (Hassard and Kelemen, 2002: 333). On this view, the evidence-based approach can be seen as one way to produce bite-sized chunks of knowledge for power. Yet the academy itself comprises research consumers who sanction the approval and badging of approaches (Kuhn, 1962); which spawn further research, and self-legitimating ‘schools’ (McKinley et al., 1999; Pfeffer, 1993). Schools can be seen from an Apolline perspective as guiding and ordering future research with clearly stated principles, and sharp boundaries between rival schools. From a Dionysian perspective, schools represent tribalism (Campbell, 1979). Previous research suggests the academy’s tastes are such that repeated tests are dull and boring (Mone and McKinley, 1993; Neuliep and Crandall, 1991). If these judgments apply to evidence-based morsels, it undermines the idea that pursuing this approach is a purely Apolline endeavour. Instead it requires offerings to Dionysus. For the movement to convert followers it must appeal and excite.

The researcher

To the extent they have influence, scientists can represent secular figures of authority. They can be totems and figureheads for progress; post-theist ‘priests’, offering alternative sets of prescriptions and standards, for instance neutrality and advancement (Thorpe, 2001), or indeed ‘evidence’. Yet this status for evidence-based ‘believers’, though clearly desired, seems tantalisingly out of reach. Unlike their role models, the physicians, ‘believers’ are focused on organizations, not ills and pills. So, when they place exclusive importance on Apolline aspects (pursuit of a precise, scientific procedure) this is insufficient in two ways: it misrepresents the ‘objects’ of investigation, and fails to reflect the reality of knowledge production in policy studies and management. In these disciplines, aesthetic judgments structure relations in a community of researchers. This in turn influences the status of practices (Bourdieu, 1984).

Having explored the Apolline and Dionysian responses to evidence and theory in turn, the following sections relate them to one another in an analysis of the dialectic between ‘believers’ and ‘heretics’.

‘Believers’, ‘heretics’ and dialectic

Rather than simply reflecting epistemological difference, attitudes towards the evidence-based approach are also aesthetic reactions. ‘Believers’ display profound concern with aesthetic criteria, principally craft, in their care and passion for technique (per the Young et al., 2002 extract above). ‘Heretics’ disparage aspects of the evidence-based approach, on aesthetic grounds:

… buzzwords appear (seductively) to represent commonsense. They also provide some neat answers (or so it might seem) in the debates about the relevance and application of organizational research; answers packaged in the rhetoric of science that make them especially appealing … because they might thereby appear to be untainted by ideology. (Learmonth, 2008: 287)

The two central aspects of Birth: aesthetic, categorical response; and the Apolline / Dionysian dialectic, are both relevant to understanding relations between ‘believers’ and ‘heretics’. Winfree (2003: 61) describes the complexity of the dialectic between the Apolline and the Dionysian:

… almost nothing in The Birth of Tragedy is as it seems, not even Apollo, who may be understood provisionally as the very principle of appearance … even the delimitation of form is eventually de-formed, its delimitation de-limited, never to be fully restored … the discovery of the Dionysian … must likewise be understood in terms of a transformation of the Apollinian—the Apollinian itself, therefore, a mark of excess, form always marked by what it would need to exclude, the excluded itself that which would force the de-limiting of the delimitation that would hold it at bay.

This extract is helpful in communicating something of the complexities within Birth, and more specifically in differentiating between a Hegelian dialectic and a Nietzschean one. Hegel describes the history of thought in terms of a process of progressive realization, where contradictions between conflicting categories, or antinomies (for instance Subject/Object, Being/not-Being) are resolved and overcome. Simplifying somewhat, when antithesis confronts thesis, dialectics results in synthesis—an overcoming or sublation that renders the antinomy obsolete, unifying opposites. This results in improved understanding of the world:

… everything actual contains opposed determinations within it, and in consequence the cognition, and more exactly, the comprehension of an ob-ject [sic] amounts precisely to our becoming conscious of it as a concrete unity of opposed determinations. (Hegel, [1830]/1991: 93)

In Birth there is a more perplexing and relationally complex dialectic and that is why it makes for a better model of the division between ‘believers’ and ‘heretics’. The Nietzschean categories of Apollo and Dionysus are simultaneously stable and yet also in flux. They are each in flux because they express and carry internal, immanent sources of conflict, which is a source of comparison with the Hegelian dialectic. However, the dialectic in Birth is not the conventional triad (opposition of thesis and antithesis, with sublation or supersession into synthesis). Instead there are aspects of stability in the continuing separation and a denial of sublation. In Birth we witness a convoluted dance of eternal opposition between Apollo and Dionysus punctuated by rapprochement and subversion. Whereas for Hegel, contradictions can result in synthesis, in Birth the fundamental separateness of Apollo and Dionysus is not resolved, they remain distanced though simultaneously intertwined and perpetually in flux. This is apposite because the tension between ‘believers’ and ‘heretics’ echoes older, perhaps irresolvable, debates about positivism and anti-positivism or the (im)possibility of social progress through technology or science.

In another departure from a Hegelian dialectic, Apollo is a derivative and inferior category, rather than an exact antinomy. Dionysus (as chorus and Will) is undifferentiated reality, while the apparent brilliance and clarity of Apollo (as hero and Individual) is the illusion of a differentiated reality, ‘the ontological relationship between that which veils and that which is veiled is not the same as that between thesis and antithesis’ (Force, 1982: 181). However, as aesthetic forms, they are enmeshed because for Dionysus to appear in tragedy requires the effect of Apollo, ‘to appear Dionysus must take on a form … And the form he takes on is given by yet another form—the very form of appearance, that of Apollo’ (Porter, 1995: 481, emphasis in original). Or as Nietzsche (Birth: 52) puts it, ‘Dionysus appears in a multiplicity of figures, in the mask of a warrior hero … entangled in the net of the individual will’. Describes the rejection, acceptance, and perpetual necessity/incoherence of the principium individuationis.

The arguments of ‘believers’ and ‘heretics’ can also be understood in terms of their reactions to this principle of individuation. Individuation is a logical necessity for a project of evidence-based research: to accumulate, there have to be clear, knowable boundaries across repeated studies. At the same time, a project grounded in Apolline ideals requires a Dionysian submission to Will: collective subordination to a central article of faith—realism.

One way individuation is rejected in critique of evidence-based approaches is by challenging the notion there is ‘a’ body of evidence or such a thing as ‘the’ evidence, ‘no one fact … can be seen as “the evidence”. There are simply bodies of evidence, usually competing bodies of evidence’ (Fitzgerald et al., 2002: 1437). Any claim there is one body of evidence rests on commensuration. Commensuration is not a neutral act, but itself a mode of power (Espeland and Stevens, 1998). Indeed, there are clear parallels between the idea of assembling a body of evidence (accumulating knowledge, having commensurable measures of quality, ranking studies according to quality) and the main disciplinary functions of classification and ordering (Foucault, 1991; Foucault and Gordon, 1980). Disciplinary functions lead to power becoming internalized to the extent that its true nature is unnoticed. Power is disindividualized, it produces a reality where no one person is sovereign and there is unthinking (Dionysian) submission to established order. Part of what sustains this Dionysian submission is an Apolline fiction: that the individual ruler (politician, manager, doctor, priest) is separate from the herd.

At the same time, the scholarly space ‘heretics’ can inhabit is shrinking, precisely because of a collective premium placed on novelty and individuation:

The scale of activity by the evidence based school, and the preoccupation of academic research with ‘uniqueness’ … means it is increasingly difficult to sustain critique of fundamental problems with an evidence based approach because there has been a precursory discussion, or simply identification, of limitations or challenges. (Morrell, 2008: 627)

The Apolline principle of individuation is comforting because it appears to promise an exact place from which to view and explain the world, but it is, ‘cocooned within its logical schematism’ (Birth: 69). ‘Believers’ illustrate the effect of cocooning with unwitting irony. For instance: in ‘narrative’ reviews that set out why it is bad for others to do ‘narrative’ reviews (Tranfield et al., 2003; Walshe and Rundall, 2001; Young et al., 2002). Or, a recent ‘systematic review’ (by Reay et al., 2009) titled ‘what’s the evidence on evidence-based management’, which systematically excluded papers critical of evidence-based approaches. Rousseau (2006b: 1091) interpreted repudiation of the evidence-based approach as agreement about a ‘meta-issue’, ‘the process whereby evidence-based management (EBM) might be most effectively designed and implemented’. Inability by ‘believers’ to recognize heresy sits uneasily alongside their advising others to pay scrupulous and unbiased attention to all relevant evidence.

Apollo becomes Dionysus, Dionysus becomes Apollo

Birth’s dialectic is relevant for understanding responses to evidence-based approaches and drawing on it results in a novel analysis of the potential trajectory of arguments relating to theory and evidence. Future ‘believers’ who worship Apollo must surely pay homage to Dionysus given the collective, cultural aspects to theory building. If, for instance, systematic reviews are destined to be consigned to ghetto sections of journals, or grey literature, then there is a self-destructive seed at the heart of the evidence-based project. At the same time, the ‘heretics’ who worship Dionysus are (at least in the medium of academic texts) bound to Apollo and the principium individuationis since they must package and repackage the same essential deep-seated concerns in novel ways.

Strati and de Montoux (2002: 757) argue that the, ‘sharp distinction between science and art’ is no longer legitimate because the dichotomy ‘between scientific discourse and aesthetic experience has faded’. In Birth the relevance of ‘aesthetic’ in terms of the dialectic between ‘believers’ and ‘heretics’ is encapsulated by the question Nietzsche places in Socrates’ mouth, ‘[m]ight art even be a necessary correlative and supplement to Science?’ (Birth: 71).

Researchers working with Apolline ideals of theory testing have long recognized the influence of aesthetic elements such as interest (Hubbard et al., 1998; Neuliep and Crandall, 1991). However, ‘believers’ face a challenge because they delegitimate aesthetic responses to theory in pursuit of an ascetic ideal. Overlooking this aesthetic aspect is not simply an epistemological issue. It is an essential denial that scientific, social and aesthetic factors are intertwined in projects of accumulating evidence. ‘Believers’ write out elements of shared discourse or myth, consumption, and style when discussing evidence-based research. The clearest example of this is the antipathy towards ‘narrative reviews’ and the valorization of technique and impartiality (Morrell, 2008). Yet aesthetic factors are important in understanding the development of any research practice.

As scholars, many of us are attracted to ideas that are novel and exciting. The criterion of novelty often determines whether ideas are valued. Mone and McKinley suggest this ‘uniqueness value’ has both functional and dysfunctional consequences for the development of knowledge in management research (Mone and McKinley, 1993). One example of a functional consequence may be that the ‘uniqueness value’ encourages greater creativity and continual generation of new thinking. An example of a dysfunctional consequence may be that it prevents incremental progress and makes it harder to evaluate or compare across competing explanations of phenomena.

That divisions between ‘heretics’ and ‘believers’ will endure in their current form is unlikely and this is where dialectic is helpful. Evidence-based approaches will inevitably retain Apolline strictures of testing, rigour and comparison but if ‘believers’ are to develop their movement, and convert others, they will also need to make offerings to Dionysus: acknowledgment of issues relating to communal taste. It would be harsh to suggest ‘believers’ were not also excited by novelty, yet a problem for converting others is that many of the procedures involved in accumulating evidence strip away creative scope and indeed make a virtue of doing so, ‘a replicable, scientific and transparent process … a detailed technology, that aims to minimize bias through exhaustive literature searches … providing an audit trail of the reviewers’ decisions, procedures and conclusions’ (Tranfield et al., 2003: 209).

There is another potential trajectory in terms of the dialectic between the Apolline and Dionysian. If the strictures that underpin the evidence-based approach were to become widely accepted within a community, they would take on a disciplinary aspect. If the evidence-based approach is adopted as the best way to carry out policy or management research, then as this value judgment becomes internalized and unquestioned, Apollo will become Dionysus (principled commitment to a technique or approach becomes institutionalized as unquestioning collective adherence to an article of faith).

But can Dionysus become Apollo? Worshippers of Dionysus are not isolated from wider cultural pressures. If funding bodies remain driven by narrow constructions of evidence, then Dionysus may be forced to wear the mask of Apollo. For instance, Learmonth (2008: 288) recounts career advice he has been given, ‘apply for grants from bodies that require evidence-based research, provide them with the reports they want, and then write separate, critical academic papers from the data’. Grappling with incoherence may help the academy develop mental gymnasts, and perhaps be useful preparation for a political career, but it is harder to see how it will yield insightful research.

There are other ways in which Dionysus can relate to Apollo as this debate unfolds. Van Maanen (1995: 139) illustrates how both scientific and aesthetic criteria are relevant to building theory. Writing ‘like everyone else’ is not only aesthetically unsatisfying, we ‘bore ourselves to tears’, it also limits theory development, we ‘restrict the range of our inquiries and speculations’. Kilduff (2006: 252) suggests personal interest in research questions is ‘the route to good theory’. Weick also advocates an aesthetic sensibility, ‘[w]henever one reacts with the feeling that’s interesting, that reaction is a clue that current experience has been tested against past experience, and the past understanding has been found inadequate’ (1989: 525, emphasis in original). To acknowledge the legitimacy of these aesthetic responses to theory, and accept they can play a role in crafting work in the Apolline tradition, shows the benefits of relating the Apolline - Dionysian dialectic to the evidence-based approach.

The call to do sufficiently novel work is a challenge to ‘believers’. Although it may be used to guide contemporary decisions and actions, the evidence-based approach is in essence backward looking. ‘Evidence’ has already happened and been gathered (or created) by others. One potential source of rapprochement or temporary alliance of Apollo and Dionysus is suggested by Pfeffer and Sutton (2007) who place emphasis on ‘craft’. This suggests the possibility of combining Dionysian (recognition of social sanction and collective taste) and Apolline elements (precision). In the process it legitimizes an aesthetic criterion which can be an alternative to novelty: craft.

Yet the need for research to be interesting remains a challenge for ‘believers’. Lawler (2007: 1033) acknowledges this as a problem for ‘journalists’ and for ‘magazines’: ‘the fact that a finding is well known (even if it is only among academics) almost automatically makes it uninteresting to the people who write and edit news publications’. Yet academics are also motivated to read things they find interesting. Examining ‘current best evidence’ (Sackett et al., 1996) is both retrospective and unoriginal. Indeed these aspects to the evidence-based approach, which set it apart from most other modes of knowledge production, are taken as central and defining virtues. To combine retrospection and unoriginality with an emphasis on standardization, then in Van Maanen’s terms, this approach does seem destined to bore us to tears.

The prospects for converting followers do not seem auspicious. Briner et al. (2009: 24) suggest that, ‘for those inclined toward developing EBMgt, there are exciting roles to play’. Yet they are unable to call on the kind of intellectual excitement that characterizes the Dionysian. Instead they list a series of technocratic activities: ‘systematically collating the available evidence’, ‘consensus building’, being ‘better knowledge brokers, feeding relevant and critically appraised evidence into organizations’. To call these ‘exciting roles’ is questionable, but more troubling is the sense they support a vision of the academy as a servant of power. They continue, ‘practitioners may need to acquire, assess, adapt, and apply research evidence to their decisions; and we note that academic skills and knowledge can aid this acquisition and use’.

Blind Apollo

Clearly, epistemological questions arise from trying to apply principles of evidence-based medicine to social science: What counts as evidence? What is sufficient evidence? What should be done if there is not enough evidence? How do we determine the quality of evidence? How can such research be aggregated and compared? How can systematic reviews incorporate qualitative evidence? How should evidence be put into practice? What are the most important barriers to translating a methodology from medicine to the social sciences? and so on. Though unresolved, many of these questions have been discussed by ‘believers’. Though they concern a comparatively new phenomenon, a lot of these are not new in that they revisit longstanding problems in the social sciences: about the nature and status of empirical research, how to use the findings from such research, and how it can be accumulated (Astley and Zammuto, 1992; Beyer and Trice, 1982; Hubbard et al., 1998; Neuliep, 1991).

These have their origins in even older debates about scientism and positivism, or even ancient philosophical questions: how we gain knowledge of the external world; the problem of induction; the status of different forms of knowledge; the distinction between appearance and reality; the divide between empiricism and idealism; the relative merits of a posteriori and a priori knowledge; the relation between truth and falsehood, between error and opinion, and so on. These questions remain important and interesting of course, but a limitation with epistemological questions is that they are ahistorical. They give little sense as to the likely development of the evidence-based approach as a material phenomenon. To try to evaluate this, we need to look beyond epistemological questions to the diachronous and also the ethical/political character of developing arguments concerning research and evidence.

Though they have had far less attention, numerous ethical and political questions arise in relation to the evidence-based approach. These are ethical in the same basic way that any claims to knowledge about the social world have ethical implications, and political in the sense that judgments about evidence govern boundaries of exclusion and privilege (Cooke, 2008). The ethical and political are entwined: the ethos of ‘believers’ and ‘heretics’ and their accounts of relations between the academy and society or claims to knowledge about evidence will shape our polis (Parker, 2003). Whether in favour of, or against, this movement, the growing influence of an evidence-based approach has broad implications for the academy, for individual researchers and for the consumers and stakeholders of research. The promise of an alignment between academic research on the one hand, and policy, or management, on the other, suits the interests of a number of parties.

May (2005: 526) describes a relation between ethos and polis, ‘this imperative towards evidence-based policy in the final decades of the 20th century is one important ideological feature of the apparently post-ideological character of contemporary British politics’ (original emphasis). The UK’s Labour government was (at least rhetorically) a champion of the evidence-based approach: as early as 1999 the Home Office committed £250 million to the Crime Reduction Programme, an initiative in ‘evidence-based’ policy (Hope, 2004). What is unusual in political discourse, is that over a decade later, this language is still extremely influential.

The manifestoes of each party in the UK coalition government drew on evidence-based language: ‘It is vital that policy, especially that relating to public health, criminal justice and environmental protection, benefits from being based on the best available evidence’ (Liberal Democrats, 2010: 29); ‘We will make sure that funding decisions are made on the basis of need, and commissioning decisions according to evidence-based quality standards’ (Conservative Party, 2010: 46). Of particular interest to how academic life is governed in the UK, this language was extensively put to work by the Department of Business, Innovation and Skills (BIS, 2010: 3) in their report, The Role of Evaluation in Evidence-Based Decision-Making:

The Department currently spends around £21bn a year on a range of policy areas including skills and higher education, innovation and science, business and trade. Because these areas are crucial for the UK’s economic prospects, BIS has a pivotal role to play in rebalancing the UK economy and driving economic growth. Spending decisions often involve trade-offs. Ensuring that these are made on the basis of robust evidence, including evaluation, is key to BIS’s approach.

For politicians, or more widely, those in power, appeals to evidence-based principles are useful rhetorical buffers: attempts to divest responsibility for choices and to provide distance from the consequences of those choices. As BIS oversees unprecedented cuts to funding for UK higher educational institutions, it will be useful for it to place emphasis on evaluation and claim this as an apolitical, neutral technique. Yet ‘evaluation’ only features after the most significant ground has been conceded. In this context, the basic political choice has been made: ‘rebalancing the UK economy’ translates into drastic short-term cuts in public sector spending. An entirely contingent result is that Higher Education is one of the areas most heavily cut by this government. The evidence for this (often expressed in monolithic terms as the deficit) is clear in a sense, but the interpretation of it is not. Cloaking policy choice in ‘evidence-based’ language helps disguise an inevitable, even tautologous truth. Politicians are involved in the interpretation and construction of evidence and frequently bound by judgements that reflect ideology rather than events.

A purely Apolline account is blind to the danger of furnishing power with a convenient, rhetorical sheen. The language of ‘evidence-based practice’ can be deployed to graft ‘evidence’ retrospectively onto decisions premised on ideology. This is a concern as ‘believers’ seem so keen to be heard they are willing to have a very general definition of evidence. At the same time as providing a language readily open to abuse by power, the Apolline valorization of technique and transparency, and claims to be free of bias mean ‘evidence-based’ rhetoric leaves no space to articulate values. It is easy to see how this rhetoric could be used to try to prop up the Coalition Government’s current HE policy, its desire to increase the cost of higher education, and its assault on the humanities, but it is not a recipe for an enlightened or civilized polis. As Higgins (2010), chief arts writer of The Guardian writes:

If, under the previous government, there was an agreement that the arts were a good in themselves, though resistant to crude numerical quantification and to the banalities of political discourse, that consensus has been shattered. That the humanities in general, from history to law to literature, have a value within a civilised society, seems to have been rejected. There is a dark new philistinism abroad.

This context and case are of broader relevance for understanding the implications of arguments by ‘believers’, and the problems with a purely Apolline account of the relation between theory and evidence. To preach a shift to a new, purportedly superior form of knowledge production is imperialism. It governs exclusion and privilege, and supports the exercise of power. ‘Believers’ suggest an evidence-based approach, ‘calls for something like a revolution in social research practice’ (Young et al., 2002: 220). They rarely limit the range of problems for which they feel an evidence-based approach is suitable, instead claiming blanket superiority, as they suggest this is the way of: addressing ‘organizational problems’ (Rousseau, 2006a: 256, 266); or capitalizing on ‘better information to gain an advantage’ (Pfeffer and Sutton, 2006: 12); or enhancing ‘decision making’ (Briner et al., 2009 passim). At the same time they seem unable to recognize fundamental criticism, or articulate essential problems and limitations, and interpret any such interventions as attempts to assist the project. These blindspots, sweeping claims, and collective enchantment with medicine bring ‘believers’ closer to Dionysiac devotion than to an unbiased project of science. Widespread physician envy in organization studies means, ‘a community of unconscious actors stands before us, seeing themselves transformed’ (Birth: 43). There is of course, and for good reason, scepticism to the claim that there is any one best way of addressing social problems (Fox, 2003). Though it is partly a question of epistemology, to reject this claim is also an aesthetic choice.

Conclusion

There are wider applications of the method of dialectic outlined here and so, albeit briefly and speculatively, I will conclude with some thoughts about these. Summarily, the approach here could be adapted and applied to generate ‘heretical reviews’ rather than ‘systematic reviews’ and perhaps be called a dialectics-based approach. Such projects could seek out orthodoxy and heresy in a field and, using Nietzsche’s categories, identify aesthetic aspects that divide and unite the spirits of the Apolline and Dionysian. ‘Heretical reviews’ could be valuable if they disrupt convenient patterns of thinking which require that we deny complexity and the richness of social phenomena. They would be much more likely to yield original insights or novel treatments of phenomena than a review of evidence, and they would be less easily appropriated by power.

Many fields in organization and policy studies could be seen as dominated by the spirit of Apollo, since they exemplify, ‘faith in the explicability of nature and the universal healing powers of knowledge’ (Birth: 82). This faith reflects a fascination with individuated character rather than the transcendent experience associated with the dithyrambic chorus:

The character is no longer expected to broaden out into an eternal archetype, but rather to come across as an individual, with artificial characteristics and nuances, each trait most precisely determined, so that the spectator is no longer alive to the myth and instead focuses on the verisimilitude of the characterization [a] victory of the individual phenomenon over the universal. (Birth: 84)

The promise of apparent precision and clarity may lead followers of Apollo to confuse the study of properties (perhaps of measures, settings, populations or individuals) with the study of phenomena to which we give words like ‘ethics’, ‘leadership’, ‘culture’ or ‘organization’. Or they may wrongly take the actions of individuals, or accounts of actions by individuals, to be ‘the social’ (Latour, 2007). To question this is not, of course, to deny the possibility of evidence, or even the value of any particular method. Still, it may be that a heretical review proves valuable precisely because it makes some evidence unpalatable. Why should we make things taste sweeter to power?

The full title of Birth (in its first two editions) is, The Birth of Tragedy: Out of the Spirit of Music. ⁴ Nietzsche describes Apollo’s music as, ‘Doric architecture transmuted into sounds’ (Birth: 20); ordered, individuated and concretized, this is an architecture that is beautiful rather than terrifying, the background to action rather than constitutive of it. In contrast stands the Dionysian, dithyrambic chorus, ‘a general mirror to the universal will’ (Birth: 93), where the function of the chorus is, ‘reflection of Dionysiac man for his own contemplation’ (Birth: 42). This enshrines ecstasy, a standing apart from individual desires, choices, triumphs and consumption. From the spirit of this music, dialectics can sustain and amplify the painful, joyous discord between capital and Will.