The Object of Inquiry and Organization Studies

Object Lost . . . ?

Say that we observe a restaurant serving food for dining guests. It would seem reasonable to assume that this is an instance of organization – an organizational phenomenon – and thus an object of inquiry for organization scholars. An inquiry into the restaurant may help better understand this instance of organization and possibly add to our understanding of organization more generally. Observations are likely to confirm the assumption of organization, and we may report what we found in terms of coordination, decision-making, communication, organizing, processes, or whatever else the concepts of our theoretical vernacular bring into focus. Despite the quirks and challenges commonly encountered when undertaking an inquiry, we may find it relatively unproblematic to take an empirical stance towards organization.

This fairly straightforward approach towards taking an empirical stance is, somewhat surprisingly, in stark contrast with the precarious character of organization that comes to the fore in debates on ‘organization lost’. It is argued, for example, that an attention to order in general has led to inattention to organizations as well as to organization outside organizations (Ahrne & Brunsson, 2011, 2019), that a lack of attention to concepts of organizations has led us to ‘in effect talk “around” the organization rather than about it’ (King, Felin, & Whetten, 2010, p. 290), and that metaphysical stances ‘disappear’ organizations by rendering them as something else (du Gay, 2020). ‘Inattention’, ‘talking around’ and ‘disappearing’ all focus on ‘the organization’, but are also relevant problems for studies of organizing because they point to the involvement of the inquiry in making something an object of inquiry in a study of organization. This question of what organization studies study comes to the fore as internal to organization studies. The phenomenon studied does not tell us how to make it an object of inquiry – and does not protect us from assumptions that imply a metaphysical stance. In other words, it’s not straightforward to take an empirical stance. There are good reasons to look at how we deal with assumptions of organization when we inquire into organization.

Say that we observe a number of people eating in a park. It would seem reasonable to assume that this is not an instance of organization, but that a number of individual groups have come out spontaneously to dine. This assumption may not hold. We may, for example, find that the groups are part of a festival organization. As it was assumed that the situation was not an instance of organization, the study would need to disconfirm assumptions. Thereby the inquiry passes a test: in order to counter the assumption, the study must make explicit what is understood by organization and how organization is identified empirically. This test is not made in the restaurant example. Yet, because the empirical stance is a recurrent rebellion (van Fraassen, 2002) against metaphysical stances and assumptions, a recurrent test of assumptions of organization is necessary. The restaurant example, if it undergoes this test, may show assumptions to be metaphysical. But given that assumptions of organization are part and parcel of our interest in studying organization, how can such assumptions at the same time be tested as part of the recurrent rebellion against metaphysics?

In this essay I suggest that the presumption of non-organization reinforces the empirical stance and contend that it should inform studies that make organization an object of empirical inquiry. The presumption of non-organization takes as its paradigm the presumption of innocence, a normative principle found in legal practice. By explicitly shifting the ‘burden of proof’, the presumption of non-organization pointedly articulates the empirical stance; it calls for making explicit what comes to count as an instance of organization in the inquiry and in what sense.

Throughout this essay, the phrase ‘object of inquiry’ is used to talk about that which is studied when organization studies study organization. In the same way as someone or something can be said to be an ‘object of attention’ and a firm, a policy process and much else can be an ‘object of criticism’, ‘organization as object of inquiry’ implies a broad notion of the ‘object’ – not delimited to ‘moderate-sized specimens of dry goods’ (Austin, 1962, p. 8) – and shares much with, for example, the notion of epistemic object (Knorr Cetina, 2001) and with the hetero-reference of an inquiry (in Luhmannian jargon). In this essay, ‘organization’ is used without any article and as part of the phrase ‘an instance of organization’ as ways to discuss ‘organization’ as used in the expression ‘organization studies’. At issue is the discussion of what is studied when organization studies study organization. Sometimes that which is studied is an organization, the organizations and the like. At other times, that which is studied is organizing, the organization of something and so forth. To address this variety within organization studies, and at the same time focus on organization, the essay discusses organization (without an article) and instances of organization.

In this essay, I first articulate the empirical stance in part already found in organization studies with the help of the presumption of non-organization. Then I unfold the presumption of non-organization by help of select contributions to organization studies. The aim is not to render organization studies in full, but to point, via examples, to the pertinence of countering assumptions of organization, as well as assumptions of ‘an organization’. Following this argument, the presumption of non-organization implies that bringing together findings of organization is of limited value – if it is done without bringing together accounts of how the findings were made. The conclusion reflects on the argument.

The Empirical Stance

The empirical stance is part of ‘a recurrent rebellion against the metaphysicians’ (van Fraassen, 2002, p. 36). For organization studies as an empirical undertaking, this rebellion is also recurrent, as seen by the recent example of the claim that metaphysics has ‘disappeared’ formal organization (du Gay, 2020). Put differently: for organization studies the issue is not so much whether it is an empirical undertaking (or not), but how to undertake empirical inquiry into organization informed by the empirical stance (‘organization’ is here used in a sense that includes ‘organizing’ as well as ‘an organization’). This makes it relevant to look at what empiricism criticizes:

the targets are forms of metaphysics that (a) give absolute primacy to demands for explanation and (b) are satisfied with explanations-by-postulate, that is, explanations that postulate the reality of certain entities or aspects of the world not already evident in experience. (van Fraassen, 2002, p. 37)

A problem for this rebellion is that it too easily turns into what it rebels against. When empiricism, for example, leads to the idea that experience provides the foundation of knowledge (van Fraassen, 2002, p. 38) or the idea that matter is all there is (van Fraassen, 2002, pp. 49ff), then empiricism becomes a type of postulate about the world – and comes into contradiction with what it set out to do. Empiricism is, in other words, confronted with the problem that any ultimate or superior argument for empiricism may undergo empirical tests that may result in falsification. The solution to this conundrum is, so van Fraassen argues, to formulate ‘a stance (attitude, commitment, approach, a cluster of such–possibly including some [p. 48] propositional attitudes such as beliefs as well)’ (van Fraassen, 2002, pp. 47–48). The empirical stance, then, is ‘philosophy as a stance’ (van Fraassen, 2002, p. xviii) that is informed by admiration for science:

But this admiring attitude is not directed so much to the content of the sciences as to their forms and practices of inquiry. Science is a paradigm of rational inquiry. To take it as such is precisely to take up one of the most central attitudes in the empiricist stance. (van Fraassen, 2002, p. 63)

The empirical stance, then, is helpful for a rebellion against metaphysics, but as a philosophical stance it does not say much about how to undertake studies of organization. This is where the presumption of non-organization, I suggest, comes in as a procedural norm for how to do organization studies. After all:

How do we live in a world in which, to the best of our knowledge and belief, all our best most fundamental scientific theories are false? We live in it by the lights of science as practice, as search, as rational form of inquiry par excellence. (van Fraassen, 2002, p. 63)

Here it is worth returning to the second of the targets mentioned. How do studies of organization fit in here? Is organization ‘already evident in experience?’ Or is it rather a postulate of ‘certain entities or aspects of the world not already evident in experience?’ And how to tell one from the other?

Du Gay and Vikkelsø (2017) have forcefully and convincingly argued that the metaphysical stance informs organization studies, in so far as the research field ‘subverts or occludes understanding of organization by its development of an often highly elaborate, ornate and intricate understanding of simulacra that pass under the same basic name’ (du Gay, 2020, p. 467), and points to ‘X and Organization Studies’ – and the call for essays on overlooked or repressed topics (Holt, den Hond, & Reay, 2016) – as a case in point (du Gay & Vikkelsø, 2018). Their sweeping critique is relevant here because it draws attention to the basic yet too often overlooked issue that empirical inquiry and metaphysics can go hand in hand. This subversion of an understanding of organization is not found in a lack of empirical work. On the contrary, it is found unfolded in empirical work. The issue, thus, is not merely one of ‘empirical’ vs. ‘metaphysical’. It is one of ‘empirical inquiry informed by metaphysical stances’ to be countered with ‘empirical inquiry informed by the empirical stance’. But the focus on metaphysical stances has left the empirical stance somewhat in the background.

It remains just a first step to note that metaphysical stances cancel out the empirical stance. But the question of how to take the empirical stance is complicated by the precarious status of organization as an object of empirical inquiry. If formal organization is the case, it is, as du Gay and Vikkelsø remind us, metaphysics to disappear it and render it as something else. But if formal organization is not the case, then rendering something as formal organization makes something else disappear and we are caught in metaphysics anew. Thus, if organizing is the case, then rendering it as formal organization is metaphysics. This argument extends to organization in general (formal or not) and to appearing as well as disappearing: it is in effect a metaphysical stance to make an object of inquiry appear as organization (in whatever sense subscribed to) if it is not the case. This brings us to the presumption of non-organization: taking the empirical stance in organization studies concerns how organization (including formal organization as well as organizing) is made an object of inquiry.

The Presumption of Non-organization

The presumption of non-organization that I advocate in this essay takes the procedural norm of the presumption of innocence as its paradigm. Assumptions of guilt are a main reason to pursue a court case. Such assumptions may be widely shared, possibly by judges as well as by the defendant. The presumption of innocence counters such assumptions, not by wishing them away, nor by making them illegitimate, but by establishing the procedural norm of ‘innocent until proven guilty’. The norm is institutional in the sense that it concerns how the proceedings are undertaken, and it disregards whether actors taking part in the proceedings assume the defendant to be guilty (or not). Moreover, the presumption does not imply that the defendant is in fact innocent. Whether defendants – and actors in general – are guilty by nature is made irrelevant by the presumption of innocence. What matters is making explicit what rules and procedures are invoked to arrive at a judgement in the case at hand. This helps avoid, for example, a judgement of guilt that is arrived at before the case is taken up, as well as prejudicial statements of guilt that come to shape the judgement.

I propose the presumption of non-organization as a norm, making explicit that inquiries into organization ‘work on presumptions’ (Toulmin, 1953, chapter 5). It is based on the belief that ‘the difference between legal and epistemic presumptions has been exaggerated’ (Fuller, 1993, pp. 367ff; Toulmin, 2003). Assumptions of organization are one main reason to pursue organization studies. Such assumptions are not to be wished away, nor are they illegitimate, but coining a handy procedural norm that counters such assumptions is helpful for dealing with them. The norm is suggested as an institutional norm for the research field that makes explicit argumentative requirements and can thus leave the question open as to whether scholars, common sense, or actors in the fields inquired into assume organization to be the case (or not). Non-organization (March & Simon, 1958, p. 1) offers a contrast to organization as it suggests various counterparts, such as ‘market’, ‘environment’, ‘disorganization’ and so on. The presumption of non-organization does not imply that the object of inquiry is in fact not an instance of organization, nor does it imply that there is not organization or that organization is impossible. The presumption of non-organization contrasts with assumptions of organization by ‘shifting the burden of proof’ such that it is up to the inquiry, beyond a reasonable doubt, to establish that the situation inquired into is an instance of organization (or several instances) in the sense of organization as subscribed to in the inquiry. The presumption calls for avoiding the conclusion that ‘the object of inquiry is an instance of organization’ before the inquiry is undertaken and for not allowing pre-inquiry statements to shape the findings of the inquiry.

Coined in a short form, the presumption of non-organization is useful for articulating more clearly the empirical stance in studies of organization. I do not claim that it is a new norm, but that it can inform empirical inquiries into organization more systematically. In a legal context, it can be scandalous if in just a few cases the presumption of innocence is violated. Studies of organization, by comparison, seem more accepting of assumptions of organization. Such assumptions may work well, but current debates on ‘object lost’ call for more systematic discussions of what is assumed and what is found in inquiries into organization. The presumption of non-organization offers a way for the field of organization studies to do this and brings with it, as I will suggest, a number of consequences that require attention.

The register of this essay is argumentative (Davis, 1971; Hunter, 2006): the presumption concerns mainly the form of argumentation, and only secondarily the understanding of organization that informs an argument. Davis’s ‘That’s interesting!’ (1971) is helpful for situating the presumption of non-organization. Davis argues that interesting arguments in sociology deny or attack certain assumptions of the audience and establish new assumptions. ‘Organization’ is the first example of a template for an interesting argument in the essay: ‘What seems to be a disorganized (unstructured) phenomenon is in reality an organized (structured) phenomenon’ (Davis, 1971, p. 313). This template is not difficult to discern in organization studies, for example in studies of markets (Brunsson & Jutterström, 2018; Callon & Muniesa, 2005). Davis argues that, once established, new assumptions are attacked to make interesting arguments anew. In effect, Davis suggests, the templates are reversed. The ‘reversed organization template’ is: ‘What seems to be an organized (structured) phenomenon is in reality a disorganized (unstructured) phenomenon’ (Davis, 1971, p. 313). Although this reversal is easy to grasp, it is more difficult to find examples. Davis gives Marx’s Capital as example, but in fact that example concerns things considered ‘to be organized in one way, are in fact not organized in that way (but rather organized in another way)’ (Davis, 1971, p. 313). In sum, the template ‘Organization’ may be difficult to reverse, and examples of arguments that continuously find organization (in other unexpected forms) are easy to find in contributions to organization studies.

‘That’s absurd!’ is, says Davis (1971, p. 327), a likely reaction when encountering claims of the type: ‘Everything you always thought was true is really false’ (Davis, 1971, p. 327). To claim that ‘organization cannot be found’ is likely to be found absurd by organization scholars as it ‘denies the whole assumption-ground’ (Davis, 1971, p. 327). In contrast to ‘That’s interesting!’ (1971), the argument here concerns ‘that’s what organization studies study when studying organization’. What I suggest is not a general attack on assumptions of organization, but the presumption of non-organization as a contribution to handling such assumptions within our research field.

Innocent Until . . .

This section discusses the equivalent of prejudicial statements, which we may name ‘pre-inquiry statements’. A prejudicial statement violates the presumption of innocence by attributing guilt, not as the outcome of a trial, but by prejudice. Pre-inquiry statements are statements that define the character of what is inquired into, not as the outcome of the undertaken inquiry, but in advance of or outside the scope of the inquiry. Put in plain language: ‘This is organization – let us study it.’

The problem of pre-inquiry statements is procedural. At issue is not whether the situation inquired into is in fact an instance of organization (however defined), but that the burden of proof is not placed explicitly on the inquiry/inquirer. As a result, no systematic account needs to be given of what is claimed to be an instance of organization, of what this finding is based on, or of what understanding of organization was used.

Assumptions of organization are – of course – common in studies of organization, and are, as touched upon above, widely unproblematic. An inquiry into a restaurant, for example, may inquire into how coordination takes place. The inquiry, then, brings us from a situation where something is not yet known – in this case how coordination takes place in this instance of organization – to a situation where more is known. But the presumption of non-organization highlights that a conclusion of the type ‘this instance of organization is characterized by coordination of these kinds’ is also a more basic conclusion: ‘This is an instance of organization.’ Through the presumption of non-organization, this more basic conclusion is brought into focus.

The presumption of non-organization does inform some studies of organization. Work on organizationality (Laamanen, Moser, Bor, & den Hond, 2020; Latour, 2011; Schoeneborn, Kuhn, & Kärreman, 2019) offers an example. In this work, organization is deliberately turned into an attribute: organizationality. Thereby the empirical study of ‘social collectives’ (Dobusch & Schoeneborn, 2015, p. 1006) is based on the explicit presumption that the situation studied is not organization. This presumption suggests the need for empirical identification of the degree to which the situation inquired into shows organizationality beyond a reasonable doubt (Dobusch & Schoeneborn, 2015).

But although the presumption of non-organization does inform organization studies, the opposite is also the case. Understandings of organization studies as a variety of approaches that offer perspectives on a shared type of phenomenon may be a case in point. When organization is considered a particular type of ‘social phenomena’ (Ahrne, Brunsson, & Seidl, 2016, p. 93; Schoeneborn et al., 2014, p. 309), it is no longer just a name for a particular empirical situation, such as a restaurant, but becomes a name for a type (or a class), and comes to suggest that otherwise distinct approaches and inquiries agree that they study the same type of phenomenon: organization. But how can this be known? Is it not the same as postulating ‘the reality of certain entities or aspects of the world’ (van Fraassen, 2002, p. 37) in advance of the inquiries?

The problem is when dealing with organization as a phenomenon leads to pre-inquiry statements of the type ‘these are instances of organization’ and thereby removes the burden of proof from the inquiry. March and Simon’s opening of their book Organizations may serve as example:

This book is about the theory of formal organizations. It is easier, and probably more useful, to give examples of formal organizations than to define the term. The United States Steel Corporation is a formal organization; so is the Red Cross, the corner grocery store, the New York State Highway Department. The latter organization is, of course, part of a larger one–the New York State government. But for present purposes we need not trouble ourselves about the precise boundaries to be drawn around an organization or the exact distinction between an ‘organization’ and a ‘non organization’. We are dealing with empirical phenomena, and the world has an uncomfortable way of not permitting itself to be fitted into clean classifications. (March & Simon, 1958, p. 1)

The question of how to identify organization is handled by making lists of examples. Understanding organization as a phenomenon then comes to suggest that no definition or concept of organization can be brought forward by the inquiry when organization is identified.

But the apparent absence of a definition is instead an ostensive definition. Implicitly, concepts provide criteria of sameness and are given at the outset as part and parcel of the ostensive definition (Wittgenstein, 2009). The presumption of non-organization calls for such criteria to be made explicit. The presumption does not suggest an ‘exact distinction between an “organization” and a “non-organization”’, just as a presumption of innocence does not suggest that there is an exact distinction between guilt and innocence. By contrast, because no such clear distinction can be assumed, there are reasons to place the burden of proof on the inquiry, so that it makes explicit what criteria of identification are being relied on.

The presumption of non-organization may further show that organization is only sometimes a phenomenon, in the sense of an ‘immediate object of sensation or perception’ (OED Online, 2020). One may point to a protest on a city square or the marching of a troop of soldiers (Coulter, 2001), but arguably ‘one never can point at a bureaucracy’ (Heyck, 2015, fn 14, p. 178). Often organization seems not to be a phenomenon but rather a ‘mediated object’ that can only be observed by bringing together a number of observations of phenomena or traces (Taylor & Van Every, 1999), such as of buildings, logos, products, meetings, work, uniforms and more. Presuming non-organization helps make explicit the analytical work involved in analysing such ‘mediated objects’.

. . . Proven Guilty

If defendants in court, for example, are presented in a cage, this may be a violation of the presumption of innocence. The presumption of non-organization also draws attention to the presentation, and here the issue of ‘what organization is identified’ can become pressing as well as challenging. In a court case, the question of who the defendant is may be less challenging to answer. Say that a case on trespassing relies on witnesses, video surveillance and fingerprints to identify the individual who has trespassed. Organization is more difficult to identify, not least because several instances of organization may very well be entangled, and an overlapping of several instances of organization may be found at the same time and place. An indication or a number of indications of organization, then, may be part of several instances of organization simultaneously. Take a construction site. The work organization there may include carpenters, electricians, bricklayers, painters and more. Workers from each of these trades may, by employment, be part of their organization (firm) and, by union membership, be part of a number of unions that are also taking part in organizing the work undertaken. The carpenter apprentice may be part of a school, and the school may have a say in what work the apprentice should do to gain certain qualifications. More aspects could be taken into account, such as national regulations of working conditions, ‘foreign labour’ and construction codes. When the carpenter apprentice is found mounting plaster boards, this may, at the same time, be part of a school programme, part of a carpentry firm and part of a construction project. Although observations report organization – beyond a reasonable doubt – the observations may not show what organization is identified.

A distinction between relations and sets of relations is helpful for formulating this issue in more general terms. Studies of organization typically observe relations and suggest organization – understood as ‘an organization’ or as ‘processes of organizing’ or something else – as a set of relations. Yet, some relations may be part of several sets without much overlap found between the sets. This implies that although data is found that beyond a reasonable doubt shows that organization is the case, we may not know what organization is the case, as the data may well indicate relations that are part of several, partly overlapping instances of organization. Two examples illustrate the pertinence of this analytical issue.

The first example concerns the formal organization of a university. Chester Barnard’s suggestion that ‘always there is purpose as the coordinating and unifying principle’ is relevant here (Barnard, 1951, p. 95). If ‘education’ is the purpose found, this would include activities of staff and students in the inquiry, not least because the work and participation of students in classes, groups and so on is part and parcel of teaching: students, for example, prepare, do group presentations and undertake field work, without which the educational activities are disrupted. Formal organization is also given prominent attention by Roethlisberger and Dickson with a particular focus on ‘those patterns of interaction prescribed by the rules and regulations of the company’ (Roethlisberger & Dickson, 1939, p. 566). In so far as student activities are not prescribed by rules and regulations of the university, which would often be the case, these activities would not figure as formal organization. Comparing the two shows important differences. In the former case, students are suggested to be part of the formal organization, in the latter, often not. The identification of ‘formal organization’ is not self-evident, and the same can be expected with other concepts of organization. Types of relations quite different in kind and scope may be identified, along with multiple sets of relations, which suggests that sometimes multiple instances of organization may be identified (perhaps unsurprisingly, given that, for example, Roethlisberger and Dickson understand ‘formal organization’ to be just one of several types of simultaneous organization). How do such multiple instances of organization relate? This would, following the argument here made, be an empirical question.

The second example concerns the communicative constitution of organization (CCO). This approach, as touched upon above, makes a considerable effort not to assume organization (Cooren, 2020; Schoeneborn et al., 2014). The concept of ventriloquism (Cooren, 2010, 2020) is developed in this context and is helpful for illuminating the complexities of the issue. A ventriloquist makes a puppet speak, and ventriloquism is, in this context, ‘the phenomenon by which an actor makes another actor speak through the production of a given utterance’ (Cooren, 2010, p. 1).

The figure of ventriloquist and puppet doubles when used in identifying organization as communication:

When we speak or act on behalf of an organization, we not only speak or act as its puppet but also as its ventriloquist. A puppet because we are led to say specific things as we are supposed to represent this organization as well as its interests, expectations, and preoccupations; a ventriloquist because we eo ipso make this organization say something that is also supposed to represent its interest, expectations, and preoccupations [. . .] In other words, a ventriloquial view insists on the vacillation or oscillation that is taking place whenever something or someone communicates. Communicating therefore amounts to sharing one’s agency with others, which become the conduits through which one manages to express oneself (and reversely). (Cooren, 2020, p. 182, italics original)

Ventriloquism thus suggests a view of communication in which the representation of an organization is simultaneously part of making the organization (Cooren, 2010, pp. 151ff) – in many ways an elegant way to unfold the presumption of non-organization. This focus on relations of representation implies ‘that everything or everyone can potentially become a channel, medium, or conduit for something or someone else’ (Cooren, 2020, p. 182, italics original). Identifying organization concerns identifying oscillating relations with an empirical attention to who or what becomes a medium for who or what and how this amounts to an organization.

Yet there may not be merely ‘this organization’ (Cooren, 2020, p. 182), but several instances of organization simultaneously. Often, inquiry into organization encompasses more than two actors, such as a puppet and a ventriloquist. There could be situations involving one ventriloquist and two puppets, or two ventriloquists and one puppet, or a number of ventriloquists and puppets, some ventriloquists with several puppets, and some with an arm up the same puppet. In these more complex situations, it must be determined what organization is represented and made. The identification of relations is not the same as identifying a set. Further criteria are required to identify whether, for example, one set is studied (as in ‘this organization’), several sets (as when a puppet or ventriloquist is a boundary object), or a set of sets (as in the communicative constitution of a ‘many-headed hydra’; Teubner, 1993). The presumption of non-organization here helps make explicit the criteria for the identification of set(s) such that a focus on identifying relations does not come to assume organization by assuming the set(s).

Accounts and Findings

The presumption of innocence is also used if a case considered in one jurisdiction, for some reason, is deliberated in some other jurisdiction. Someone found guilty of, say, violating the public order in one jurisdiction may not be found guilty if the case is deliberated in some other jurisdiction. As a consequence, it is of limited value to compare findings across jurisdictions in so far as different jurisdictions define violations differently and have different requirements for arriving at the judgement that a violation has taken place. To understand to what degree the findings are comparable, even if the findings are given the same name, a comparison would have to include not only the findings but also the rules and procedures used to arrive at the findings.

How does what is identified in one inquiry into organization relate to other inquiries? A not uncommon answer is that findings about organization can be added to other findings about organization, and that one perspective can be added to other perspectives, in sum giving a fuller picture. But when adopting the presumption of non-organization, these answers become insufficient. This presumption supposes that it is of limited value to compare findings of organization. To understand the degree to which findings of organization add to other findings of organization, an account of the inquiries would need to show how, in each inquiry, organization is identified and in what sense the finding is an instance of organization. Without bringing together accounts, it cannot be known whether findings of organization agree, whether the findings are partly overlapping, or whether they, despite using the same name ‘organization’, have no overlap.

Bringing together accounts in this way is not uncommon in organization studies, but the presumption of non-organization is nevertheless helpful to bring it to more systematic attention. The issue here is only partly a conceptual issue. Relatively specified concepts, such as ‘formal organization’, are broad, in the sense that there seems to be neither agreement nor an established taxonomy of types of relations relevant for identifying, for example, formal organization as well as organizing. Even when such agreement is established, there is typically room for methodological variations. Two organization studies that focus on ‘membership’ and subscribe to the same definition may still apply significantly different criteria for identifying members empirically. A specific concept of organization is thus no guarantee of ‘a common denominator’.

Taken to its logical conclusion, the presumption of non-organization thus bars us from relying on organization as a given external point of reference – for example, as a phenomenon, as something on which various perspectives can converge, or as an object given to a naive or reflected realism. The burden of proof placed on the inquiries brings with it the burden to demonstrate the degree to which the inquiries talk about the same thing and are comparable. The presumption of non-organization posits not treating organization as ‘an epistemological primitive category’ in the sense that ‘similarities between two particulars are the basis for creating a generalization’ (Ohlsson & Lehtinen, 1997, p. 41). The finding of organization is a claim, and this finding has no direct correlative outside of an inquiry. We may wish, for example, to compare the findings of an inquiry with the organization (in some pre-inquiry sense), but to undertake such a comparison, we would have to undertake another inquiry to lift our burden of proof, and then we would need to compare the two accounts. Such accounts may, as is well known, differ significantly. In sum, studies of organization thus not only seek to understand concrete and empirical organization, but are ‘actively involved in the constitution’ (Knights & Morgan, 1991, p. 270) of the object of inquiry – and this is not in contradiction to, but inherent to, the empirical stance as informed by the presumption of non-organization.

Organization is a concept that is also widely used outside organization studies, and the circulation of ‘organization’ as a native category may be found to contradict the point just made. It may be held, for example, that as natives use the category and do organization, there is indeed a type of object given to organization studies to study. Moreover, native categories may be held to be the most convincing way (if not the only way) to establish meaningful concepts of organization, as argued convincingly by Egon Bittner (1965) and others. Yet, following the presumption of non-organization, the burden of proof remains with the inquiries, which means that native categories, like other types of observation, should be mobilized in the inquiry to establish, beyond a reasonable doubt, the claim of organization. The likely observation that there is not just one but several native categories of organization that can be observed adds to the complexity of arriving at a claim and underlines the benefits of presuming non-organization.

The argument so far thus implies some complexities in determining what organization studies study when studying organization. To reiterate, the object or objects of inquiry of the field of organization studies are not given, but are established by iteratively bringing together accounts of inquiries that compare the burden of proof across studies. As Otto Neurath famously has put it, ‘We are like sailors who have to rebuild their ship on the open sea, without ever being able to dismantle it in dry dock and reconstruct it from the best components’ (Cartwright, Cat, Fleck, & Uebel, 1996, p. 155). This point also confirms – rather than contradicts – that values are inherent to organization studies. It is not innocent to make organization an object of inquiry. Taking the empirical stance involves values, attitudes and commitments (van Fraassen, 2002, pp. 47–48) – in short, ideals may well inform the recurrent rebellion against metaphysics, as the work of Neurath and the so-called left-wing Vienna Circle testifies (Uebel, 2015). But valuing, for example, formal organization over organizing (du Gay, 2020) neither removes the burden of proof, nor does it necessarily imply that studies of organizing imply a metaphysical stance. However, it is possible that the objects of inquiry differ significantly to the degree that they have little or nothing in common.

The empirical stance targets forms of metaphysics ‘satisfied with explanations-by-postulate, that is, explanations that postulate the reality of certain entities or aspects of the world not already evident in experience’ (van Fraassen, 2002, p. 37). Returning to this quote, the presumption of non-organization contends that postulates of organization are to be replaced by claims of organization. It also stresses that the experience relevant to organization studies is the experience mobilized in making claims of organization (whether the experience concerns observations made by organization scholars or the experiences of natives). The presumption guards against general assumptions. Thus, for a theory to be compatible with the empirical stance as defined by the presumption of non-organization, the theory needs to allow for non-organization. Otherwise, the inquiry is rigged as it does not allow empirical inquiry to arrive at other outcomes.

Towards the Empirical Stance

The presumption of innocence can be criticized for slowing down judicial processes. This criticism may also be raised against the presumption of non-organization, but organization studies is not primarily oriented towards swift results, and the presumption of non-organization is unlikely to significantly change the pace of its endeavours. The presumption of innocence may also be criticized for sometimes resulting in the acquittal of defendants who have committed the crimes of which they are accused. The presumption of non-organization may analogously be criticized for sometimes barring organization studies from offering a better understanding of some instances of organization. But given that organizational scholars take part in developing not only concepts of organization but also procedures for identifying organization, this problem is likely to be handled by conceptual and methodological choices and developments.

Organization is often assumed in organization studies. At the same time, organization has a precarious status because neither what is understood by organization nor what is an instance of organization is a given. The empirical stance proposed here, inspired by the principle of ‘innocent until proven guilty’, suggests that we should presume non-organization as a way to bring more systematic attention to the work of analysing organization. I hope I have shown that it is productive to reverse the burden of proof so that organization studies may more systematically make explicit how the work of analysing makes organization into an object of inquiry. If we understand social life as marked by organization, as we obviously should, we must inquire into the onto/epistemological commitments and ideals that shape our empirical understanding of organization, since such commitments are part and parcel of each of our inquiries. The empirical stance seeks to make this explicit.

The aim has not been to formulate a theory of organization. The empirical stance invites both existing and new approaches to organization into a conversation about how organization is made into an object of inquiry, and accentuates the transversal role of concepts, methodology and data in the production of findings. When organization studies is found preoccupied with objects of inquiry other than organization (such as ‘market’) – other presumptions (e.g. of non-market) may prove helpful for articulating the empirical stance (Frankel, 2018).

Assumptions are hard to avoid. Any inquiry takes some things for granted while questioning others. A door needs hinges to work, as Wittgenstein puts it. At issue here is how assumptions are dealt with. Organization is too important – for the working life of people, for large societal problems – to be assumed without being countered systematically by the presumption of non-organization.