Abstract
This article is exploratory and experimental. It starts from the premise that leadership scholarship is a site of disagreement, where mainstream claims are challenged by critical scholars. Some criticism focuses on conceptual clarity, and incorporates consideration of who should be categorised as a leader, and on what basis, and whether it is helpful to refer to ‘leaders’ and ‘followers’. The ‘new wave’ of critical leadership studies generates controversial questions relating to whether leadership exists or is in fact a myth that we have reified. The bulk of criticism directed at educational leadership challenges three mainstream knowledge claims – underpinned by what I call the causality belief, the leadership dependency belief, and the conceptual belief – and which are the focus of this article's analysis. While criticism of these knowledge claims is well-rehearsed, the article breaks new ground by analysing them through an epistemic justification lens to address the question: is educational leadership (still) worth studying? Represented by these three component beliefs, the mainstream educational leadership scholarship belief system is analysed within a frame derived from the philosophy of science, and draws on BonJour's coherentist theory of epistemic justification to apply a more structured assessment than has hitherto been achieved by critical scholarship.
Keywords
The landscapes of research and scholarship fields are not set in stone. Eroded by doubt and scepticism that manifest themselves as someone's questioning or querying this or that theory, notion, belief, assumption, accepted wisdom, or dominant method(ology), they are constantly being re-contoured. Whether they are based on reasoning, compelling contradictory evidence, or a combination of the two, such doubt and scepticism – characteristics of what Kitcher (2000) calls a ‘field of disagreement’ that unsettles a scholarly community's ‘epistemic state’ – are implicitly or explicitly concerned with the worthiness or the justification of the knowledge being challenged.
Educational leadership research is the site of such a field of disagreement, where many of the claims underpinning mainstream scholarship are – and for several decades have been – repeatedly challenged by critical leadership scholars. As I discuss towards the end of this article, challenges from its critical ‘wing’ often serve as a catalyst for a field's epistemic development through its gradual assimilation of divergent perspectives, thereby making them the new mainstream. Yet if critical scholarship is having an erosive effect on the leadership field, it is doing so imperceptibly, for the challenges presented to them by ‘new wave’ critical leadership scholars (explained below) seem to go unheeded by mainstream educational leadership researchers – who, judging by their outputs, are either unaware of or choose to ignore this discourse. Thus the status quo of leaderist policy and practice prevails, prompting critical leadership scholars, such as Alvesson and Spicer (2012: 368), to bemoan that ‘many of the negative consequences implicit in leadership theory and practice … are all too often masked or even ignored’. 1
Perhaps, then, to provoke awareness of the possibility that leadership scholarship may not be all that it is purported to be, a different perspective needs to be advanced, and this article attempts to formulate one that is based on consideration of leadership scholarship's epistemic worthiness. To prelude such consideration, I clarify my interpretation of criticality.
Essentially, as I note elsewhere (Evans, 2018: 49–50): Critical leadership scholarship interprets leadership differently from how it has traditionally and conventionally been interpreted. A key point is that it represents dissatisfaction with mainstream leadership research, yet, just as there are degrees of dissatisfaction, there are degrees of criticality; so we may imagine a continuum, ranging from, at one (arguably, the ‘moderate’) end, perspectives that deviate only slightly from those dominant in the mainstream, to the paradigm-shifting perspectives located at what could be called the ‘radical’ end. (original emphasis)
Cutting across this continuum I discern, broadly, two different, but complementary, substantive interpretations of criticality that are evident in the educational leadership field. First, a social-justice-related critical discourse reflects Gunter's (2001: 40) interpretation of criticality as being ‘concerned with enduring power structures and the impact these have on the lives and work of educationalists and communities’. Immensely valuable, such criticality has established a firm foothold in the educational leadership scholarship field. It is, after all, difficult to fault and to resist committing to a social-justice-focused agenda.
Yet my interpretation of critical leadership scholarship incorporates a wider focus than social-justice-related issues. I am with Alvesson and Deetz (2021), who, explaining that ‘[o]ur take on critical research is … a bit different from researchers strongly focusing on groups they find to be the victims of oppression and injustice’ (2021: 7), argue that critical research is: oriented towards challenging rather than confirming that which is established, disrupting rather than reproducing cultural traditions and conventions, opening up and showing tensions in language use rather than continuing its domination, encouraging productive dissension rather than taking surface consensus as a point of departure. (2021: 6)
Such critical research seems more prevalent in what I call the ‘wider’ (i.e. non-context-specific) than in the educational leadership research field. Within the former field it has been referred to as ‘the new wave of critical leadership studies’ (Kelly, 2014: 907) – though Kelly's label does not seem to be widely adopted. Challenging the thinking that fuels ‘today's leadership obsessed culture’ (Alvesson and Spicer, 2012: 368), ‘new wave’ critical leadership scholarship concerns itself with epistemology, concepts and conceptual frameworks, and methods and methodologies. More specifically, it encompasses, inter alia, notions of collective leadership, of leadership as practice rather than as a person, and leadership-scepticism; its more radical proponents question whether leadership even exists. Within the wider leadership research field, ‘new wave’ critical scholarship is by definition a marginal sub-field; within the educational leadership research field it is even more marginal, since only a handful of educational leadership and management scholars knowingly, by accessing non-education-specific leadership academic literature (much of which emanates from business and management schools), engage with the discourse.
Yet by variously challenging their field's mainstream perspectives, beliefs, claims and scholarship, other educational leadership scholars articulate a critical voice that, even though they may remain unaware of its existence and unfamiliar with its outputs, reflects the spirit and is aligned with the focus of ‘new wave’ critical leadership scholarship. For discursive simplicity I use the terms ‘critical’ and ‘new wave critical’ interchangeably throughout this article to denote the wide interpretation of criticality that I share with Alvesson and Deetz (2021).
This article breaks new ground by exploring an analytical pathway that, to the best of my knowledge, has never before been applied to educational leadership scholarship
2
; it assesses the field's knowledge base and belief system through an epistemic worthiness lens. Described as a ‘thorny philosophical issue’ (Christias, 2015: 33), epistemic justification (or worthiness), as Šešelja and Straßer (2014: 3112) explain, fundamentally relates to the augmentation of knowledge and understanding: Epistemic justification is … traditionally conceived of as providing standards for the acceptability of certain beliefs in the knowledge base or the cognitive system of an intelligent agent. Applied to a scientific theory, it provides criteria for its inclusion and acceptance into the grand corpus of our scientific knowledge. It concerns the question as to whether we have good reasons to consider it as being (approximately) truthful, empirically adequate, etc.
This explanation underscores the importance of trying to distinguish ‘fact’ from ‘fiction’ in justifying the initiation or retention of a particular focus of study or research. But, to inform assessment of its fitness for purpose, such justification must be framed within the context(s) to which knowledge will be applied, and to what ends, as well as the purposes for which it was – and potentially still is – considered worth seeking.
Why study educational leadership?
‘People seek knowledge for reasons, purposes, so why would we want to study leadership?’ asked Catherine Marshall (1995: 487) at a time when the educational leadership field was establishing itself. Now, within an anniversary special issue marking 50 years’ dissemination of scholarship in Educational Management, Administration and Leadership – of which some three decades or so have included, as it gradually emerged, educational leadership scholarship – I revisit Marshall's question.
The rationale for studying educational leadership was/is its perceived capacity to influence the quality of educational provision, for, as Myran and Sutherland (2019: 683) remark, ‘[t]he deeply held belief that leadership matters for learning has driven a significant part of the educational leadership, school improvement, and school effectiveness research’. Yet this quote includes three highly significant words – ‘deeply held belief’ – which underscore a key point: as a focus of scholarship, policy, or practice, leadership represents a belief system; one believes in it or not – partially or fully. Indeed, Hunt (in Bedeian and Hunt, 2006: 195) asks: ‘[i]s leadership something real or is it little more than a term invoked to explain the inexplicable?’ In similar vein, Eacott (2018: 13) asks, ‘Who is to say, and on what grounds, that leadership … is any more than the latest fad?’ and Lakomski (2005: 3) makes the point that ‘[w]hether or not leadership exists is subject to critical examination and should not be taken for granted’, while, more decisively, critical leadership scholars Gemmill and Oakley (1992: 114) declare leadership to be ‘a social fiction’.
Eacott (2018: 14) observes that ‘[o]pening up scholarship blurs, if not breaks down, disciplinary boundaries’, and while some of the issues I raise in this article are well-rehearsed within educational leadership research, what is distinct about my examination of them is the application of an analytical frame that draws upon the philosophy of science literature relating to epistemology and the robustness of knowledge. The article thus makes an original contribution to the field through its critique of three widely held beliefs that pervade mainstream educational leadership scholarship. A key focus of my analysis is BonJour’s (1985) framework of criteria for the epistemic worthiness of belief systems. Directed at addressing the question: Is educational leadership (still) worth studying?, my analysis is of necessity selective, and aims for indicativeness rather than rigorous comprehensiveness – not only because the latter is unachievable within the space I have at my disposal, but also because, as I explain below, belief systems within the social sciences can prove difficult to subject to such criteria.
I begin by outlining BonJour's framework.
BonJour's criteria for epistemic worthiness
Belief systems, as Usó-Doménech and Nescolarde-Selva (2016: 148) interpret them, are the stories we tell ourselves to define our personal sense of Reality. Every human being has a belief system that they utilize, and it is through this mechanism that we individually, ‘make sense’ of the world around us.
Identifying 13 characteristics of them, these authors explain belief systems as ‘structures of norms that are interrelated and that vary mainly in the degree in which they are systemic. What is systemic in the Belief system is the interrelation between several beliefs’ (Usó-Doménech and Nescolarde-Selva, 2016: 147–8). BonJour (1985) uses the term ‘component beliefs’ to refer to the multiple individual beliefs that make up a particular belief system.
BonJour's framework reflects a coherence approach to epistemic justification (or a coherentist epistemology), which is explained: coherentist theories of empirical justification … reject the view that there is an absolute foundation which functions as an ‘epistemic prime mover’ of an essential hierarchical structure in our system of empirical beliefs, and hold instead that empirical justification is conducted within a network of mutually supporting beliefs, where some of them are more ‘central’ and others more ‘peripheral’ within the network, but in which all beliefs, central and peripheral alike, are justified solely on the basis of their contribution in the system's coherence as a whole. (Christias, 2015: 8)
‘What justifies beliefs’, BonJour (2009: 158) notes, ‘is the way they fit together: the fact that they cohere with each other’ (original emphasis). But, within this ‘theory of justification’ (Kuukkanen, 2007: 560), whether or not specific beliefs cohere with each other cannot simply be assessed arbitrarily or subjectively; conditions for coherence must be met, as Kuukkanen (2007: 560) explains: The customary way to characterise coherence is by stating one or two conditions: consistency or/and explanatory unity. Consistency means that no coherent set can contain both beliefs that P and not-P. Explanatory unity means that each belief P of a set is explained by other beliefs in the set.
Translating his philosophy-speak into the vernacular, what Kuukkanen's explanation means is that, to be coherent – and hence epistemically justifiable – a belief system must meet at least one of the two conditions identified above. To be consistent, it must include only component beliefs that sit easily with, and do not contradict or cancel out, each other, and for a belief system to demonstrate explanatory unity, any of its component beliefs that need explaining must be explained through another of its component beliefs, leaving no gaps in comprehensibility.
Kuukkanen (2007) expands on the nature of these conditions for coherence: We need to require that there be some positive connections between the beliefs that make up the system. If they are totally unconnected, we cannot say that the set is very coherent. BonJour writes that a natural idea is that connections are inference relations. There should be such relations that a belief or a set of beliefs can serve as the premise(s) of an argument for a further belief. (Kuukkanen, 2007: 560)
Finally, it is important to recognise that, in comprising multiple component beliefs, each of which is the ‘property’ of an individual who decides subjectively whether or not to hold it, belief systems fundamentally denote what individuals believe, and any assessment of their worthiness must incorporate consideration of the individual and her or his beliefs as being the point of departure. BonJour's comments on propositional knowledge are relevant to such consideration; ‘[m]y primary concern’, he writes (BonJour, 1985: 3) ‘… is propositional knowledge: the knowledge that something is the case, that a certain proposition is true. … What conditions must be satisfied in order for a cognitive state to constitute a genuine instance of propositional knowledge?’ (original emphasis) – in other words: what makes someone's belief(s) epistemically justified knowledge? An individual may believe, for example, that vaccination is harmful and should be avoided, but does such a perspective constitute what BonJour refers to as ‘a genuine instance of propositional knowledge’? – by which he means: is it epistemically justified, or does it constitute genuine ‘knowing’? To allow such questions to be addressed, he (BonJour, 1985: 3–4) identifies three conditions:
For a person A to know that P, where P is some propositions, three conditions must be satisfied:
A must believe confidently that P, P must be true, and A's belief that P must be adequately justified.
To clarify what is meant by some of these terms, such as: ‘believe confidently’, ‘true’ and ‘adequately justified’, BonJour (1985: 4) offers the following explanations. For his ‘condition 1’ (above) to be met, he notes that ‘[s]ubjective certainty is probably too strong a requirement, but the cognitive attitude in question must be considerably more than a casual opinion; I must be thoroughly convinced that [P]’ (emphasis added). For condition 2 to be met, he argues for adopting ‘the classical realist account of truth as a relation of correspondence or agreement or accordance between belief and world’ – or, as Guarini (2007: 132) puts it: ‘by truth he [BonJour] means a correspondence between a thought and a mind-independent reality’. And the ‘adequate’ justification required for a belief to meet condition 3, BonJour warns, must involve more than ‘a mere guess or hunch or arbitrary conviction. It cannot be merely a product of wishful thinking or something I read on the slip from a fortune cookie. Instead, there must be some sort of reasonably cogent reason or ground or warrant for my belief that the [P]’.
How, then, does educational leadership ‘knowledge’, reflected in beliefs held, measure up against BonJour's conditions, and against his criteria for belief systems’ epistemic worthiness?
The epistemic worthiness of educational leadership knowledge and beliefs: Insights drawn from the philosophy of science
Assessing educational leadership against BonJour's criteria poses challenges that stem from the nature of the scholarship field's disciplinary home: the social sciences. In many STEM fields, belief systems' core component beliefs can be expressed with the kind of clarity and precision that, whilst certainly not precluding epistemic disputes, prompt considerable consensual acceptance; for example, that an electron is a sub-atomic particle is a component belief within the particle physics belief system that, currently, all particle physicists accept as propositional knowledge. In many social science fields, in contrast, where a gold standard of objective truth is generally recognised as elusive, beliefs are highly susceptible to individualised variation. Indeed, such variation is implicit not only in Usó-Doménech and Nescolarde-Selva's (2016) expression of what they identify as belief systems’ characteristics – which takes the form of a list of diverse statements, rather than a precisely worded and uniformly presented list of nouns or adjectives – but also, and more so, in the substance of many of these characteristics, such as: ‘Belief systems vary almost infinitely in substantive content’, ‘The boundaries of a belief system are generally, although not always, undefined. Collections of beliefs do not generally have neat boundaries’, ‘The elements (concepts, propositions, rules, etc.) of belief systems are not consensual. That is, the elements of one system might be quite different from those of a second in the same content domain. And a third system may be different from each’, and ‘Beliefs can be held with varying degrees of certitude. The believer can be passionately committed to a point of view, or at the other extreme could regard a state of affairs as more probable than not’ (Usó-Doménech and Nescolarde-Selva, 2016: 148, original emphases).
Variation within a particular domain's belief systems may occur at the general level – that is, in relation to an overarching belief that shapes the nature and focus of an individual's belief system about a particular issue. To one person, for example, the leadership belief system may be focused on recognition of leadership as a form of human agency that may be carried out by anyone, and that effects change of any degree in another; to another person, it may focus on the benefits to organisations of the actions and agency of those holding designated roles within its managerial hierarchy. Variation may occur too at the specific level, reflecting subtle disagreements about the precise wording or terminology used to articulate a particular belief, or qualifications to it that some scholars impose on an expressed belief before they will accept it. What I introduce below as the causality belief, for example, I articulate concisely as Leadership is a, if not
Yet whilst such variation precludes a single educational leadership belief system that all scholars unwaveringly accept, from the field's outputs (both oral and written, and comprising both academic and grey literature) a broad consensus is discernible as beliefs that are widely shared, at the very least, ‘approximately’ – to requote Šešelja and Straßer (2014). This consensus represents the field's mainstream scholarship (for, consistent with Alvesson and Deetz’s (2021) analysis of critical social science, I identify any scepticism of or deviation from such broad consensus as indicative of critical, rather than mainstream, scholarship), and it is on what I identify as a mainstream educational leadership belief system that I focus my analysis.
Three dominant mainstream beliefs
I confine my necessarily selective analysis to what I discern to be three of this mainstream belief system's dominant component beliefs, which I label the causality belief, the leadership dependency belief, and the conceptual belief. My selection is informed by their being the key beliefs that critical leadership scholars challenge, and therefore they may be considered the most dubiously justifiable. Below, drawing on a range of philosophy of science literature relating to epistemology, I analyse the epistemic worthiness of the mainstream educational leadership scholarship belief system, as represented by these beliefs. I first consider the extent to which each belief meets BonJour's conditions for ‘genuine’ propositional knowledge (and it is worth emphasising that the term ‘propositional’ indicates acceptance that such knowledge is always contestable – at the time it is advanced, or at any point in the future when it may be superseded by new propositional knowledge), before then considering how the three beliefs together, representing a significant part of educational leadership scholarship's mainstream belief system, meet BonJour's criteria for epistemic worthiness.
The causality belief: Leadership is a, if not the, key determinant of student achievement
The causality belief is often deployed by mainstream educational leadership researchers as a rationale for examining, and the departure point for presenting findings on, the work of educational institutions’ senior management post-holders, such as headteachers and principals, as illustrated by the following statements – all appearing in the abstracts or opening paragraphs of articles in specialist educational leadership, management and administration journals: ‘School leaders are essential to the success of schools’ (Anderson et al., 2020: 1); ‘School principals play an invaluable role in schools’, teachers’, and students’ success’ (Woong Lee and Mao, 2020: 1); ‘Effective school leadership is a critical aspect to improve the quality of education. If the school leaders are unable to properly manage the school's resources, then educational institutions will struggle to survive and to develop’ (Lumban Gaol, 2021: 1); and ‘Research in the past few decades has consistently shown that principals are powerful players who can affect school effectiveness and bring about change’ (Qadach et al., 2020: 737). As their dates of publication show, such statements indicate the causality belief to remain, ostensibly at least, alive and kicking in the third decade of the 21st century, and evidently resilient in the face of extensive criticism (outlined below) that has accumulated over the last five decades of educational leadership scholarship.
Yet, whilst they are pervasive in this field's mainstream scholarly and grey literature, do such statements generally reflect beliefs derived from careful, considered and rigorous scholarship on the part of their authors, or do they in some cases represent the academic equivalent of simply retweeting chord-striking sound bites? Referring to his conditions that ‘must be satisfied in order for a cognitive state to constitute a genuine instance of propositional knowledge’, the second of these scenarios – that is, simply parroting what has become mainstream conventional wisdom – smacks of what BonJour calls ‘casual opinion’. The first scenario, however – that such statements express beliefs derived from careful, considered and rigorous scholarship – implies the conviction that is called for in the first of BonJour's conditions (‘A must believe confidently that P’ … ‘the cognitive attitude in question must be considerably more than a casual opinion; I must be thoroughly convinced that [P]’).
In the absence of explicit confirmation from them, we may reasonably infer that those scholars who manifest sustained efforts to gather robust empirical evidence to support the causality belief – such as (and perhaps most notably) Kenneth Leithwood, in collaboration with a succession of colleagues over the last two decades (e.g. Leithwood et al., 2008, 2020; Leithwood and Jantzi, 2006) – are likely to be ‘thoroughly convinced’ that leadership is a, if not the, key determinant of student achievement; such is implicit in Leithwood and Jantzi’s (2006: 201) claim, for example, that ‘A compelling, although still modest body of empirical evidence now demonstrates the significant effects of … leadership on school conditions and students’ learning’, and, more recently, in Leithwood et al.’s (2020: 571) observation that ‘[c]onsiderable amounts of evidence now indicate that school leadership matters a good deal to students’ learning’. Yet few scholars manifest Leithwood's relentless commitment to undertaking empirical research aimed at uncovering robust evidential support for the causality belief. Smacking of what BonJour calls ‘casual opinion’, the profusion of opening or justificatory statements such as those listed above suggests that – resonant of Morrison’s (2020: 12) comment that ‘[n]ovice or enthusiastic researchers make claims for what research shows when, in reality, it simply doesn't show it’ – when it comes to articulating the causality belief, it is unconsidered hearsay recycled as throwaway sound bites that dominate mainstream educational leadership literature. Evidently deaf (or oblivious) to such criticism, many scholars nevertheless seem to accept the causality belief unquestioningly, without thoroughly examining and assessing the quality of its sources and evidential bases, and without heeding the dissension choruses emanating from the critical scholarship camp.
Yet if (and where) it is indeed held, even the kind of thorough conviction that I ascribe to Leithwood and his colleagues meets only one – the first – of BonJour's conditions for propositional knowledge; the causality belief falls short of meeting conditions 2 and 3: P must be true, and A's belief that P must be adequately justified. This shortfall is implicitly exposed by those sceptics and detractors who, in contrast to the sound-bite-retweeters, do seem to have scrutinised (albeit to varying degrees) the belief's sources and evidential bases. In the 1990s Hallinger and Heck (1998), for example, noted that researchers seeking direct causal effects of leadership on student outcomes ‘have been unable to produce sound or consistent evidence’ (p. 166), while Caldwell (1994: 52) reported ‘little evidence of a direct cause-and-effect relationship between a shift to school-based management and improved outcomes for students’, and Van de Grift (1990) identified methodologically based flaws in selected studies that claim causality. Indeed, in 2008 Leithwood et al. were pointing out that: ‘A recent publication sponsored by Division A of the American Educational Research Association … claimed that research on school leadership has generated few robust claims’ (Leithwood et al., 2008: 36).
Persistent concerns (expressed by, inter alios, Angus, 1989; Bates, 1989; Courtney and Gunter, 2019; Gronn, 1996; Levačić, 2005; Murphy, 1988; Niesche, 2018; Van de Grift, 1990; Watkins, 1989; Witziers et al., 2003) about its weak evidential basis far outnumber compelling evidence-based claims of a causal link between leadership and institutional effectiveness. While it was over 30 years ago that Angus (1989: 63) was arguing that ‘the representation of leadership in recent literature … constitute[s] a “moral fiction” as it is found that none of the bases upon which claims to leadership are founded can be unequivocally supported’, and towards the turn of the millennium that Gronn (1996: 25) observed that ‘the validity of the causal role being attributed to school leaders … [has] … been called into serious question’, such criticisms have persisted into the 21st century. Levačić (2005: 198), for example, notes that ‘the actual evidence for a causal relationship is relatively sparse’, while Niesche (2018: 148), lamenting that little has changed since 1989, contends that ‘this constant search, particularly for the direct link between leadership and student outcomes, is flawed, a relation of cruel optimism’. More specifically, Eacott (2018: 16) argues that underpinned by ‘the rhetoric of agency and interaction with social structures’, educational administration literature's ‘central thread of causal vocabulary’ overlooks the social world's reciprocity: ‘[i]n other words, the ways in which the social is simultaneously shaping of, and shaped by, action’. Moreover, in a recent attack on the quality of causality-seeking leadership research, Eden (2021: 2) expresses concern ‘about the endless flood of nonexperimental, causally ambiguous, observational research that simply cannot yield actionable X→Y conclusions’, adding that ‘[f]or want of a better term, such research is causally impotent, which often uses fancy statistical procedures to try to justify causal conclusions’. Eden's overarching argument aligns with Antonakis et al.’s (2010) and Martin et al.’s (2021) identification of experimental studies as the only sure-fire mechanism for evidencing causality – ‘[u]ndoubtedly, the “gold standard” for determining causality is the classic experimental design (i.e., participants randomly allocated into treatment and control conditions …)’ (Martin et al., 2021: 1), and with Antonakis et al.’s (2010: 1113) conclusion (from their review of 110 articles) that ‘methodological practices in the causal domain of leadership are unsatisfactory’.
It is from those researchers who have joined the quest for it, that we are given the most compelling evidence of the elusiveness of what has been called the holy grail of leadership, and, by extension, of the causality belief's capacity to meet BonJour's three conditions for propositional knowledge. Pounder et al. (1995: 585), for example, report that ‘the influence of principals was not found to be associated with student achievement’, and Van de Grift’s (1990) study of 275 Dutch principals showed ‘no linear relationship between the leadership of principals and pupils’ average academic attainment’. Witziers et al.’s (2003) meta-analysis uncovered ‘no evidence for a direct effect of educational leadership on student achievement in secondary schools’ (p. 415), prompting the conclusion that ‘the tie between leadership and student achievement is weak’ (p. 418), and even Leithwood and Jantzi (2006: 223) note that ‘[o]n average, ratings of leadership effects, as in our study, are usually relatively low’ (though they qualify this finding by pointing out that their study – like others before it – ‘systematically underestimates leader effects because it invariably tests only relatively weak forms of leadership’).
Yet, reflecting Levačić’s (2005: 202) point that ‘[i]n the social world cause is rarely deterministic but is probabilistic (or stochastic). Some changes occur randomly and other changes are due to systematic relationships between variables’, some studies (e.g. Hallinger and Heck, 1996; Leithwood et al., 2008, 2020; Leithwood and Jantzi, 2006; Robinson et al., 2008) report indirect causal links. But such evidence is inevitably complex (as reported by Leithwood et al., 2020) and, often involving convoluted (potential) chains of influence that underpin multiple permutations of factors, can be messy; as Teddlie (2005: 223) observes, ‘[t]he relationships among educational leadership, teacher practices, and student outcomes are complex and highly contextualized’, while Witziers et al. (2003: 400) note that ‘studies based on the indirect effect model have not yielded unequivocal evidence supporting a relationship between leadership and student achievement’ (emphasis added).
Moreover, weaknesses in the indirect causal chain are exposed by applying findings from professional development research. Since the leadership→student achievement chain must include, as a mediating factor, a large dose of leader(ship) influence on enhancing or improving teachers’ practice – that is, contribution to teachers’ professional learning or development
3
– professional learning and development scholarship is highly relevant to educational leadership scholarship. As I note elsewhere (Evans, 2014), distinguishing it from much European scholarship, North American research into professional development typically incorporates a strong focus on generativity, to the extent of recognising professional development for teachers as effective only if it impacts upon secondary beneficiaries: students or pupils. A triple causal chain is implicit in such conceptualisation: professional development→improved teaching practice→enhanced student achievement. This generative focus has, as King (2014: 90) notes, prompted ‘an increasing number of causal impact studies carried out in the United States and elsewhere’ – several of which have, over the years, spawned claims of causality, such as Vescio et al.’s (2008: 87) conclusion, from their review, that: the collective results of these studies offer an unequivocal answer to the question about whether the literature supports the assumption that student learning increases when teachers participate in PLCs [professional learning communities]. The answer is a resounding and encouraging yes.
Yet, as with educational leadership, in the professional development scholarship field, research-informed causality claims are outnumbered by studies revealing little or no compelling evidence of causality; as Gore and Rosser (2020: 1) point out, ‘to date, few studies have shown rigorous evidence of the impact of PD [professional development] on teachers and their practice – let alone student outcomes’, while Hill et al. (2013: 478) refer to ‘null or largely null’ findings from many studies seeking the effects of professional development on student outcomes, and Yoon et al. (2007: 5) observe that ‘[t]here is more literature on the effects of professional development on teacher learning and teaching practice, falling short of demonstrating effects on student achievement’ (emphasis added). Focusing on the methodological rigour and robustness of the research it reviewed, Yoon et al.’s (2007) influential study marks something of a watershed in modifying perceptions of, and managing expectations relating to, the teacher professional development→student achievement causality chain; out of over 1300 studies ‘identified as potentially addressing the effect of teacher professional development on student achievement in three content areas’, Yoon et al. found only nine that met the What Works Clearinghouse evidence standards – which have become a benchmark for rigour in research that seeks evidence of causality. Moreover, these authors (Yoon et al., 2007: 14) point out that even the nine studies found to meet evidence standards ‘were generally underpowered’ and ‘12 effects of 20 were not statistically significant’.
Similarly failing to demonstrate teacher professional development→student achievement causality links are studies carried out by Osborne et al. (2019), Kraft et al. (2018), and Garet et al. (2016), whose research found that ‘[i]mproving teachers’ knowledge or practice did not translate into improvements in student achievement’ (Garet et al., 2016: 7). Moreover, applying ancestry searching ‘to trace backwards from policy documents, to the meta-reviews and reviews they cited, and then back a step further to the original studies that they cited’ (2016: 50), Sims and Fletcher-Wood (2021) uncovered significant flaws in the studies they examined, prompting their assessment that ‘reviews of reviews can lead to the propagation of weakly warranted findings through the hierarchy of reviews and onward into public policy, practice and research’ (2021: 57). The problem in seeking causality, Yoon et al. (2007) point out, is that, to be compelling, causality claims must be supported by very rigorous research that, meta-analyses have shown, very few studies manage to demonstrate.
In failing to meet BonJour's three conditions for propositional knowledge, then, the leadership→student achievement causality chain remains a highly contestable knowledge claim, for if improving teachers’ practice cannot be shown to impact significantly on student achievement, since the teacher professional development→student achievement causal link is not considered by methodologists to have been convincingly evidenced, it is difficult to argue for a causal chain that involves the additional link of school leadership. Yet, while it is certainly low on empirical evidential credibility, it cannot (yet) be denounced as – to quote BonJour once again – ‘a mere guess or hunch or arbitrary conviction … merely a product of wishful thinking’, nor can it be dismissed as untrue, for consensus that no compelling research-derived evidence of such causality has yet been presented simply means that researchers have not (yet) found ways of uncovering and evidencing it with the robustness and rigour that BonJour's conditions 2 and 3 calls for.
Despite the evident precarity of such knowledge claims, the leadership field has thrived on ‘unquestioned belief’ (Eacott, 2017: 19) or ‘messianic faith’ (Alvesson and Spicer, 2012: 368) in leaders and leadership; for, as Lumby (2012: 12) observes, (and this is the basis of the second component belief that I examine below): ‘we do not know how important leadership is; only that people often believe it to be so’.
The leadership dependency belief: Educational institutions cannot function effectively without effective leaders
With close ties to the causality belief, what I call the leadership dependency belief holds that, when they carry out their work ‘well’ (however that may be defined), headteachers or principals, along with others categorised as ‘(senior) leaders’, are pivotal to the effectiveness (again, however that may be defined) of, and are therefore indispensable to, their schools. This perspective is highlighted in numerous (national) policy documents and statements (e.g. Australian Institute for Teaching and School Leadership, 2019a, 2019b; Department for Education, 2020; National Policy Board for Educational Administration, 2015), and by the US-based Wallace Foundation (2013: 17) – ‘Without effective principals, the national goal we’ve set of transforming failing schools will be next to impossible to achieve’ – as well as by Leithwood et al. (2008: 29): ‘As far as we are aware, there is not a single documented case of a school successfully turning around its pupil achievement trajectory in the absence of talented leadership’.
How credible and how sound are such assertions? Certainly, they would be significantly undermined by an example of an effective school whose effectiveness was not attributable to the headteacher or other managerial hierarchy role incumbents, for such an example – and certainly multiple examples – would eat away at the notional tally of ‘true instantiations’ that Kitcher (2000: 23) considers fundamental to research: ‘I take the aim of inquiry to be the provision of a maximally unified set of explanatory schemata that will generate the largest possible set of true instantiations’. But finding such a school is a big ask, not least because, as I show above, when it comes to school effectiveness, causality is a slippery concept; non-causal agency is no more reliably evidencable than causal agency. Yet widening the net from institutional to system-level examples yields more possibilities. If there were such a place as a non-leaderist or leadership-sceptic educational environment – where the causality and the leadership dependency beliefs do not see the light of day – it would serve as a useful comparator.
There may in fact be such a place. It is called France.
The French connection
‘Leadership is seen as a catch-all solution for nearly any problem, irrespective of context’, complain Alvesson and Spicer (2012: 367–8). ‘Except in France’, they might have added, for that country, as I learned during extended residence there in 2011 as a visiting professor, does not ‘do’ leadership – at least, not as it is ‘done’ (as practice or as a focus of scholarship) in most other developed countries. With no original French word for it, leadership as a concept is evidently absent for the most part from the French imagination or psyche, as well as from policy discourses or rhetoric – certainly, as applied to the public sector (see Derouet and Normand, 2014; Normand, 2012; Progin et al., 2019; Reverdy and Thibert, 2015; Tulowitzki, 2012). Indeed, participating in a two-year-long (2014–16) programme of monthly seminars in Paris focused on addressing problems facing French schooling, I was struck by the absence of any consideration, on the part of participating French academics, practitioners and policy makers, of headteachers or school leadership as a potentially relevant factor. Progin and colleagues (2019) in fact observe that ‘le concept de leadership est plus naturellement utlisé par les chercheurs et formateurs anglo-saxons et nord-americains’ (2019: 21) and refer (2019: 18) to ‘le tabou’ (the taboo) of leadership in the context of school management: ‘parler de pouvoir ou de leadership à l’école n’a pas toujours bonne presse’ 4 (original emphases). The mainstream educational leadership belief system that is my focus here is nowhere to be seen in the French education sector – which, it seems, is populated by unbelievers.
Arguing that it fails to meet at least conditions 2 and 3, and potentially also condition 1, of BonJour's conditions for propositional knowledge, can we then dismiss as unfounded the belief statement that educational institutions cannot function effectively without effective leaders, on the basis that such institutions in France – our leadership-free comparator – seem to function perfectly well? Surely, there must be countless examples of French schools that have so far escaped Leithwood et al.’s (2008) notice – schools in which the pupil achievement trajectory has successfully been turned around, and, being in France, must have been achieved ‘in the absence of talented leadership’ (Leithwood et al., 2008: 29).
Well, no – apparently not, for the French educational system, far from exhibiting effectiveness, is recognised as being in crisis, ‘miserably failing the bottom’ (Anon, 2015). Its failure to close socio-economic inequality gaps, together with, year-on-year, low student achievement, has prompted the Macron government's suite of reforms (Anon, 2015; Fuentes/Reuters, 2018). If anything, then, the French education sector could be held up by leaderists as an example of what can happen if leadership is not recognised and afforded the precedence it deserves. (Yet it is worth noting that the United States education system, within which leaderism thrives, has also been described as being in crisis; as Blanton et al. (2020: 1015) remark, ‘[d]uring the past two decades, teachers have increasingly been placed at the forefront of a purported crisis in U.S. education’.)
Such representation of France as a leadership-free zone is, of course, rather crude and simplistic and merits far deeper scrutiny than I have the space to devote to it here (see Evans (2018) for a slightly more expansive discussion of this issue). Yet what the example of the French context does do – and this is an issue that I revisit below, in discussing anomalies – is expose a functioning alternative to the default perspective of attributing all manner of success and failure in the education sector to the quality of institutional leadership. Such exposure prompts consideration of who is getting it right: the French, or the leaderist world? Is one of the causes of the acknowledged crisis in the French education system the evident disregard of leadership's potential for increasing institutional effectiveness, or is it that in other national contexts far too much is made of leadership? Such consideration, in turn, provokes questions about the notion – the very concept – of leadership, and how it has been, and is, applied to policy, practice, and scholarship; for, as Normand (2012) and Reverdy and Thibert (2015) note, it is a concept that is largely absent from French language, culture, and ‘mentality’. Could it be here, then – at the conceptual core of educational leadership research and scholarship – that the field's epistemic structure is shakiest, and its worthiness questionable?
The conceptual belief: Leadership is what those identified as ‘leaders’ do
The concept that lies at the heart of the educational leadership research (and practice) field, and the most prominent term within the lexicon of the leadership→institutional effectiveness causality claim, is, of course, leadership. Knowing how it and its etymological derivatives are interpreted and applied is fundamental to understanding and evaluating the claims and counter claims that, along with the beliefs that they reflect, shape the educational leadership field's epistemic landscape and its mainstream belief system.
‘[W]hen people use “leadership”, more often than not they really mean “headship”’, argues Gronn (2000: 332), while Blom and Alvesson (2015: 482) similarly bemoan leadership's conceptualisation as ‘a general drift to what senior people do’, and Browne (2021: 6) contends that ‘[t]he use of the term “educational leadership” to describe those in senior positions in an organisational hierarchy has become ubiquitous’. Implicit in the statements associated with the causality and the leadership dependency beliefs examined above, such headship-specific ‘personification’ of leadership represents the field's default conceptual perspective that, despite persistent disapproval from the more critical periphery, has long framed its mainstream activity and outputs. Indeed, articulated over two decades ago, Gronn's comments have lost none of their currency, as By’s (2020: 4) recent observation confirms
5
: Leadership is something [formal] leaders do: Having previously challenged the illusion leadership=leaders … little has changed in theory and practice. Although there is some movement in peripheral areas of study such as critical leadership studies and leadership as practise [sic] … the outdated orthodoxy stands as firm as ever. I know this through reading the most recent articles in both academic and practitioner outlets specializing on leadership, and from inviting executives, students and colleagues to define ‘leadership’. Nine out of 10 times they define ‘leaders’ rather than ‘leadership’. They define the peripheral who rather than the essential what. Nothing represents the quagmire we currently find ourselves in as organizational change and leadership scholars and practitioners more than this. (original italics)
Does such terminological conflation imply ‘flawed’ conceptualisation that calls into question the epistemic worthiness of the propositional knowledge it underpins? Addressing this question takes us into a circular argument, for it depends, of course, on how one chooses to conceptualise and define ‘leadership’ – and, by extension, a ‘leader’. Yet, while subjectivity is generally an accepted feature of conceptualisation – Marshall (1995: 486), for example, ‘question[s] the assumption that there is, somewhere, a definitive and value-free definition of leadership’ – lack of conceptual clarity undermines the epistemic worthiness of educational leadership research and scholarship wherever it is unclear precisely what concept, as a form of knowledge, is intended as the main focus, and therefore whose worthiness should be assessed; is it, for example, leadership itself (and, if so, is the term intended to denote a role, a role-holder, or a feature of organisational structure and hierarchy, or an activity, practice, or a form of agency?), or is it a ‘leader’ (and, again, if so, who is, or should be categorised as, a leader, and what purpose is s/he intended to fulfil, and why?)? As Grint (2010: 3) observes, ‘we don't need to agree on the definition, but we do need to know what the definitions are’. Without such clarity, BonJour's conditions are impossible to apply.
What Gronn (2000: 332) explains as ‘the slippage in usage from the person who heads becoming cast as the person who leads’ can be traced back to the (educational) leadership field's inception. Perceptions that the promotion of leadership – which, in turn, gave rise to leaderism – lacked scientific justification are particularly significant from an epistemic worthiness perspective; as Heck and Hallinger (2005: 239) observe, ‘[i]f science has no more privilege than ideological belief, intuition, myth, or alchemy in commenting on human endeavors, then it calls into question the whole meaning of scholarship’.
The lexical shift from ‘management’ and/or ‘administration’ to ‘leadership’ 6 went hand-in-hand with – or, as some analysts interpret it, was stealthily introduced to support – (national) policy and practice shifts. O’Reilly and Reed (2010, 2011), for example, in demonstrating how, from the 1980s, the UK ‘government's core lexicon of modernization was developed’ (2011: 1082) through appropriating and adapting ‘the broader social phenomenon of leaderism’ (2010: 964), expose the promotion of a leaderist discourse and leadership as practice as features of a ‘“cascade of change’ narrative’ (2010: 967) to support an ideology that, they imply, had no research evidential basis. Expanding on this exposure, Courtney and Gunter (2019: 28–29) condemn what they call ‘corporatised fabrications’ in the form of 1990s’ to 2010s’ official UK policy rhetoric that aimed ‘to conjure “leaders” and “leadership” from contemporaneous educational professionals in senior-management roles’, and promulgated a spurious causal chain between head teachers’ role-enactment and organisational effectiveness, as indicated by student achievement: ‘the causal link between a headteacher as leader-doing-leadership and standards was unproven … little empirical evidence existed’. Such perspectives hold that, for policy reform implementation purposes, institutional heads have been ‘reinvented’ and relabelled as leaders – and, in the context of education, the concepts of leadership and leader have been promoted from the starting point of policy makers’ expectations of how, going forward, these refashioned institutional heads should be doing their jobs.
For the sake of argument, let us say that the propositional knowledge at play in this discourse relating to the origins of leaderism within the educational policy context is: first, that effective leadership is the key to effective schooling (which is more or less the articulation of, and broadly reflects, the leadership dependency belief), and, second, that headteachers and others holding posts of managerial responsibility in educational institutions are capable of, and should be charged with providing, such effective leadership. Applied to BonJour's (1985) three conditions for ‘a genuine instance of propositional knowledge', the implication of the criticism, illustrated above, of the reinvention and relabelling of school headship (and other senior managerial post-incumbency) as leadership, is that, for the instigators of it, even if condition 1 was met by their confident belief in both of these knowledge propositions, in the absence (highlighted by critics) of compelling empirical evidence of the two propositions’ truth, conditions 2 and 3 cannot be met.
Yet the conceptual belief, as I interpret and present it, goes beyond conceptualising leadership simply as what institutional heads, principals or other senior leadership team members do. In articulating this belief as leadership is what those identified as ‘leaders’ do, I intend ‘identified as’ to be interpreted widely, to incorporate consideration of two overlapping foci of study that have gained traction over the last few decades within mainstream educational leadership scholarship: teacher leadership and distributed leadership. As a sub-field or focus of educational leadership scholarship, each represents an alternative to the heroic leadership model that is predicated on the belief that leadership resides in the persons of holders of designated roles at or near the top of a management hierarchy.
Yet, by those who have examined the concept with rigorous care, distributed leadership (under which theoretical ‘umbrella’ teacher leadership sits) is recognised as being widely misunderstood and simplistically interpreted, reflecting what Harris (2013: 547) identifies as a prevalent ‘misuse of the term … as a convenient descriptor for any form of shared, collaborative or extended leadership practice’. It is within mainstream educational leadership scholarship that misapplication of distributed leadership's meaning is rife, effectively simplifying and diluting the concept as it was originally formulated: to denote leadership as a form of agency that may be found anywhere, is reciprocally exercised, and is neither bound by nor reflective of role or job designation 7 . As I discuss elsewhere (Evans, 2022a), in tending to misinterpret distributed leadership narrowly as devolved leadership, mainstream educational leadership scholarship – and hence the mainstream educational leadership belief system – either focuses on such devolution to those already included within the managerial hierarchy (albeit below the most senior levels) and holding a role or job title that denotes such inclusion, or it focuses on those to whom leader status is then ascribed, once leadership has been devolved to them, and who become identified post hoc as leaders. My point is that, so-interpreted and -applied, distributed leadership – in common with teacher leadership – offers no alternative to and no deviation from the belief that ‘leadership is what those identified as leaders do’, since all those to whom leadership is distributed are, or become, identified as leaders. Only by stepping outside the mainstream educational leadership scholarship field and looking at critical discourses on not only distributed, but also the ‘theoretical umbrella’ of collective, leadership, may such deviation be found.
A circular argument
As an abstract conceptualisation, the conceptual belief cannot be fully assessed in relation to BonJour's three conditions, for the circularity of argument that I refer to above poses the kind of problem expressed by BonJour (1985: xi) in his criticism of the foundationalist theory of epistemology: If what justifies an empirical belief in the sense of providing a reason for thinking that it is true is inference from some epistemically prior empirical belief which must itself be justified before it can serve as a justificatory premise, and if what justifies this second belief is yet a further empirical belief requiring justification in the same way, and so on, then empirical knowledge is threatened with an infinite and seemingly vicious regress of justifying beliefs.
In explaining ‘the so-called “hermeneutic circle”’, Montuschi (2003: 5) makes a broadly similar point: Our knowledge claims are interpretations (rather than representations), since they depend on context. However, a context is itself interpretative (it depends on other interpretations). So, when we try to assess our knowledge claims, we ultimately rely on yet other interpretations, rather than on independent evidence.
To believe that it is ‘what those identified as “leaders” do’ is implicitly to conceptualise leadership, and such a conceptualisation is as acceptable as any other. If the belief is confidently held, then BonJour's condition 1 is met. In relation to condition 2, while it can be challenged or rejected by those who do not share it, such a definitional or conceptually interpretative belief, formulated subjectively, cannot be considered any more untrue than true, and in relation to condition 3, the adequacy of justification for the belief is again determined subjectively; difference of opinion is the closest we may get to denying such justification – and such difference of opinion is in itself lacking adequate justification.
Yet, as with the causality and the leadership dependency beliefs, the conceptual belief should not only be assessed as a stand-alone belief, it should also be assessed in concert with these other two beliefs, all three joined together as a sample of dominant component beliefs that is representative of the mainstream educational leadership belief system. The section below presents such an assessment.
The educational leadership belief system's epistemic worthiness
In this section I shift my focus from propositional knowledge to BonJour's criteria for the epistemic worthiness of a belief system, to address the question: On the basis of its representation by three of its dominant beliefs, how does the educational leadership mainstream belief system measure up against the criteria of consistency, inferential connections, and anomalies? I re-emphasise that I do not imply that this belief system is confined to these three beliefs, yet my identifying them as dominant within it warrants my treating them as representative of the belief system.
So-represented, the mainstream educational leadership belief system incorporates the inferential connections that BonJour calls for. Inferential connectedness, as Kuukkanen (2007: 560), cited above, explains it – ‘[t]here should be such relations that a belief or a set of beliefs can serve as the premise(s) of an argument for a further belief’ – is evident in the three beliefs examined: the causality belief serves as the premise of the argument for educational institutions’ leader(ship) dependency, which, in turn, serves as the premise of the argument that leadership is conceptualised as what ascribed leaders do. The condition that BonJour calls ‘explanatory unity’, too, is met by this connectedness. In relation to consistency, again, on the basis of what I identify as its three dominant component beliefs, the mainstream educational leadership belief system meets this criterion, since none of the beliefs is at odds with either of the other two.
In relation to BonJour's third criterion, however – ‘the number of anomalous instances it exhibits’ (Kuukkanen, 2007: 560) – the case of leadership-sceptic France presents itself as an example of a potential anomaly, and hence as a potential barrier to this criterion's being fully met. ‘An anomaly is, roughly, an observation that cannot be explained from within the belief system’, explain Schubert and Olsson (2013: 37). I note above the elusiveness of evidence in the form of a (or multiple) school(s) in France whose case defies and thus undermines the leadership dependency claim; but, even if examples of such schools were found, they would not in themselves constitute anomalies. As I interpret the anomaly-related criterion within BonJour's framework, France's potentially anomalous status is not dependent upon its provision of potential physical evidence that might undermine the leadership dependency claim; rather, its anomalous status lies in France's representation of defiance of the mainstream educational leadership belief system. Such defiance lies not only in general non-adherence to, or scepticism of, the system's component beliefs, but also in the competing – and hence anomalous – beliefs about educational institutional effectiveness that permeate French culture within this context (see, e.g. Germain, 2018). And by the same token, all competing beliefs – held not only by French education sector policy makers, practitioners and stakeholders but also beyond the context of France – represent potential anomalies that, if they cannot be explained within it, undermine the coherence of the mainstream educational leadership belief system, and, by extension, its epistemic worthiness.
Yet, in liberally peppering my commentary above with the adjective ‘potential’, I intend a note of caution, for BonJour's criterion relating to anomaly is tricky to assess. The trickiness arises from the ambiguous proprietorial location of an anomaly or anomalous instance: does it in fact, and at what point does it cease to, belong to the belief system whose coherence it undermines?
In examining the nature and patterns of scientific controversies and disputes, Kitcher (2000: 27) notes that ‘complete homogeneity is frequently a very poor distribution in terms of advancing the community's epistemic state’ – in other words, it is generally through disagreement and disputes that a field's epistemic development eventually occurs. Applied to the context of the educational and wider leadership scholarship fields, the consensus practice is represented by mainstream scholarship, and the dissension from this mainstream scholarship is represented by what (to varying degrees, depending on their location on a mainstream-critical continuum) are known as critical discourses and communities – some of whose members retain an affiliation with the mainstream despite holding deviant belief systems, while some break away from it, taking their deviant belief systems with them. In neither case does the deviant belief system represent the mainstream one.
We may think of an anomaly within a belief system in similar terms, denoting a potential source of controversy. But does the anomaly then remain within the belief system, as one of its anomalies, or does its anomalous status preclude its belonging there – and, in doing so, redefine and reclassify it as no longer an anomaly? On revealing itself to be a swan, the ugly duckling, after all, ceased being an ‘anomaly’ within the duck family, becoming instead a fully fledged (no pun intended) member of the cygnine community.
This line of reasoning exposes a flaw in BonJour's anomaly-related criterion as it applies to ‘community’ belief systems, such as those representing research and scholarship fields. Moreover, it is worth mentioning that his framework, and the coherentist epistemology that it represents, have been the target of criticism on several bases (see, e.g. Christias, 2015; Guarini, 2007; Moser, 1988). Whilst I am not a philosopher, I discern from the philosophy of science literature that the field seems to embrace and then discard fads and fashions as readily as does the educational leadership field. Elgin (2021: 1504) notes that ‘[p]hilosophical progress occurs not only when questions are answered, but when philosophers realize that they have asked the questions wrongly’ – and the latter seems to have occurred in BonJour's case. In his later work, where he refers to ‘an explosion of epistemological discussion’ that has occurred since the 1980s, ‘with ever more new positions being suggested, elaborated, discussed, criticized – and then dismissed as untenable by at least large portions of the philosophical community’, BonJour (2003: 6) revises his own coherentist theory, reverting back to a form of foundationalist epistemology that his coherentist framework – applied to my analysis in this article – had been introduced to replace. Interestingly, however, even after publication of its author's revised stance, BonJour's framework remained the focus of attention from philosophers (of science) – in some cases as starting points for their own revisions to it (e.g. Šešelja and Straßer, 2014) – and coherentist epistemology seems by no means to have been written off by epistemologists and philosophers; in 2007, for example, the journal Synthese published a special issue, ‘Coherence and truth: Recovering from the impossibility results’, of papers arguing – as several texts had previously argued – that coherence, despite BonJour's own and others’ recognition of its limitations, remains an epistemologically important factor. 8
Where, then, does this experimental analysis of the mainstream educational leadership belief system leave us in relation to assessing its epistemic worthiness?
Is educational leadership (still) worth studying?
The analysis presented in this article, as I emphasise at the start, was always intended to be indicative rather than rigidly comprehensive (and the qualifications to and questionability of coherentist epistemology that I refer to above perhaps increase its potential for indicativeness over comprehensiveness). What, then, does the analysis indicate?
It indicates, at best, mainstream educational leadership scholarship's limited epistemic worthiness. While, on the one hand, the belief system is coherent insofar as it incorporates the inferential connections that BonJour's framework calls for, on the other hand, it (the belief system) is shot through with anomalies in the form of ‘observations’ – to use Schubert and Olsson’s (2013: 37) term – that cannot be explained within it (though, I question their anomalous status). These ‘observations’ take the form of counter-beliefs and challenges that emanate from deviant beliefs that prevail within specific national cultures and psyches and within specific (to varying degrees, critical) scholarly communities.
A key barrier to the scholarship's epistemic justification, or worthiness, lies in BonJour's three conditions for propositional knowledge: A must believe confidently in P, P must be true, and A's belief in P must be adequately justified. My analysis indicates little – or, more precisely, limited – evidence of confident belief in the causality and the leadership dependency beliefs, but, since such potential evidence in the field's literature is fairly elusive, we must not be too hasty to infer that no such confident belief exists. In other words, the analysis has not uncovered unbelief – nor even no confident belief – among mainstream educational leadership scholars to an extent and with the confidence that would merit ascription of low epistemic worthiness to the field; my observations relating to scholars’ predilection for unconsidered sound-bite retweeting are based on inferences drawn from the lack of evidence that the beliefs examined are true. And it is in relation to this – the second of BonJour's conditions for propositional knowledge – that mainstream educational leadership scholarship is most vulnerable in epistemic worthiness terms, for so many of its claims and the beliefs underpinning them cannot convincingly be evidenced as being true.
Such truth-elusiveness is, of course, the basis of the critical scholars’ objections to mainstream (educational) leadership scholarship – Blom and Alvesson (2015: 486) sum up the issue: ‘[t]he truth-value of leadership research and recipes claimed to be founded on evidence are often quite doubtful’ – and I doubt if my epistemic-worthiness-informed analysis here offers much by way of augmenting their vibrant discourse. Moreover, it (my analysis) is likely to cut no ice with those mainstream scholars who, despite the absence of compelling empirical evidence of it, are in fact convinced in their own way that leadership (however, they interpret the concept) is key to institutional effectiveness and effective educational provision, and that, by extension, educational leadership remains worth studying – as Thrupp and Wilmott (2003: 92) conclude, ‘it has become increasingly clear that our critical concerns are not being heard’.
But if critical leadership scholars – and I count myself among them – wish to challenge, or simply question, the epistemic justification for studying leadership (educational, or otherwise), then it is perhaps time to shift the focus of analysis more squarely onto them – onto us. Since this experiment in analysing the epistemic worthiness of mainstream leadership beliefs and claims has been only partially successful, insofar as it has yielded some problematic, and therefore inconclusive, findings, it is worth now shining a light not on the mainstream beliefs that we, the critical scholars, consider (to varying degrees) unfounded, but on those that we hold, and that we consider more epistemically justifiable than those we attack. Such a strategy poses problems, for the critical leadership discourse is, for the most part, just that: a discourse – a ‘talking shop’ that is big on ideas, and vociferous in highlighting the limitations of mainstream leadership-related scholarship, but short on specific proposals for practical ways of addressing them, through research with the potential not only to justify but also to kick-start the paradigm shift promised by well-formulated critical theoretical perspectives. Within the ideal forum of this special issue that marks Educational Management, Administration and Leadership's 50th anniversary, this article is directed at raising awareness of the potential for such a new paradigm for educational leadership research.
If we are to continue discussing and debating the epistemic worthiness of educational leadership scholarship, then – if we are serious about addressing whether educational leadership is (still) worth studying – we must, going forward, seek compelling evidence to support the discourse. The search for such evidence calls for a new research agenda: one that prioritises testing, through rigorous empirical research, the theoretical perspectives that are the mainstay of the critical leadership discourse. Since it is beyond the scope of this article, I take up discussion elsewhere (Evans, 2022a) of what such a new research agenda might look like, and how it may be pursued.
Footnotes
Declaration of conflicting interests
The author declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.
Funding
The author received no financial support for the research, authorship and/or publication of this article.
