| December 2011
Ecole Polytechnique Fédérale de Lausanne & University of Reading
Authority’s hidden network: relations of authority, obligations and roles
Keywords: authority, Archer, morphogenesis, networks, obligation, roles.
The purpose of this paper is to explore the import of an ontologically informed understanding of social roles for a study of relations of authority and of those social networks they presuppose and, in turn, constitute. Authority is analysed as a relation based on legitimacy. The latter is in turn analysed both in terms of authorisation and of obligations. Such a perspective emphasises the import of identities and in particular social identities and social roles in the constitution of relations of authority. Moreover, our ontological study indicates that those relations of authority that are observable in any given organisation are themselves rooted in a wider – and typically neglected – network of (significant) others whose expected attitudes is commonly used as a compass for agents engaging in relations of authority.
1) Beyond Weber: studying authority as a legitimized relation of power
It has become common wisdom, when undertaking a sociological study of authority, to refer to Max Weber’s works. The present paper does not offer the room needed for a detailed critique of Weber’s overall approach that combines ideal types and hermeneutics with an actualist ontology. We can, however, analyse Weber’s definition of authority and use it as an (adjustable) spring board for our own theoretical endeavour.
In Economy and Society, Weber defines authority as legitimate domination (Weber 1978: 215). This formulation deserves some unpacking as Weber’s use of the words ‘domination’ and ‘legitimacy’ is quite idiosyncratic.
A short definition of ‘power’ and ‘domination’ is proposed in the early pages of E&S:
‘A. “Power” (Macht) is the probability that one actor within a social relationship will be in a position to carry out his own will despite resistance, regardless of the basis on which this probability rests.
B. “Domination” (Herrschaft) is the probability that a command with a given specific content will be obeyed by a given group of persons.’ (Weber 1978: 53)
The definition of domination is elaborated further on p. 946. As Weber has it:
‘To be more specific, domination will thus mean the situation in which the manifested will (command) of the ruler or rulers is meant to influence the conduct of one or more others (the ruled) and actually does influence it in such a way that their conduct to a socially relevant degree occurs as if the ruled had made the content of the command the maxim of their conduct for its very own sake. Looked upon from the other end, this situation will be called obedience.’ (Weber 1978: 946).
Let us note immediately that Weber’s definition of domination as a probability or as a situation is questionable: it is not because in a situation of domination the probability of compliance is particularly high that domination is itself a probability; similarly, it is not because there exist situations of domination that domination is itself a situation. And indeed, the rest of the passage indicates that Weber implicitly treats domination as a relation. Furthermore, this definition of domination is based on a (discursive) opposition between ruler and ruled, and between power of command and obedience, thus covering the circularity of relations of power. Notice also how domination presupposes an appropriation of the command by the ruled. This appropriation, however, is formulated from a third person perspective marked by the use of an ‘as if’ clause. While this definition allows Weber to remain truthful to those positivist standards which he regarded as a token of scientific rigour, it also sheds a veil on the ruled’s reflexive powers and on the micro-sociological processes in which they engage whenever they engage in a relation of domination.
Let us now look at Weber’s conception of legitimacy. The early pages of E&S offer an enlightening characterisation which also includes a definition:
‘An order which is adhered to from motives of pure expediency is generally much less stable than one upheld on a purely customary basis through the fact that the corresponding behaviour has become habitual. The latter is much the most common type of subjective attitude. But even this type of order is in turn much less stable than an order which enjoys the prestige of being considered binding, or, as it may be expressed, of “legitimacy”’ (Weber 1978: 31. Emphasis added)
If we follow commentators such as Bullen (1987), Weber’s conception of legitimacy is thought as an obligation rather than as an authorization. Thus, according to Bullen, Weber distinguishes between actions that are “legitimate” (that should be obligatory), entities that are “non legitimate” (for which there is neither obligation nor prohibition) and entities that are “illegitimate” (that should be prohibited).
In the rest of the paper, I purport to move beyond Weber’s conception of authority and legitimacy in order to flesh out those social networks that are presupposed by a relation of authority:
First, authority is a social relation and not a substance or a personal attribute, not to mention a Weberian probability. If it makes sense to say that authority qua power would not exist without the reflexive personal powers of people, it should immediately be added that the power of authority (as opposed to, say, the power of gathering fruits or fleeing from fire) also supposes a relation to an other who takes count of ego’s attitudes and whose attitudes ego takes into account in a process of mutual adjustment. As we shall see below, this other needs not be a real, living, person. However, the existence of (a relation of) authority supposes a minima that ego be concerned with the plausible attitudes and reactions of that other (real or imagined). Foucault’s analytics of power (Foucault 1983) provide an attempt at moving away from a view of power as a substance held by persons. It is suggested, rather, that power is better conceived as a relation in which the actions of one person will influence the actions of another. If authority is a kind of power, then, just like power, it must be circular. In other words, even when relations of subordination exist (e.g. commander/soldier; doctor/patient; teacher/student), both parties are subject to the relation of power. The commander, the doctor, the teacher need the compliance of the soldier, the patient and the student. In this sense, relations of power can be contrasted with relations of sheer violence or passivity. As Foucault has it:
‘What defines a relationship of power is that it is a mode of action which does not act directly and immediately on others. Instead, it acts upon their actions: an action upon an action, on existing actions or on those which may arise in the present or the future. A relationship of violence acts upon a body or upon things; it forces, it bends, it breaks on the wheel, it destroys, or it closes the door on all possibilities. Its opposite pole can only be passivity, and if it comes up against any resistance it has no other option than to try and minimise it. On the other hand a power relationship can only be articulated on the basis of two elements which are each indispensable if it is really to be a power relationship: that the “other” (the one over whom power is exercised) be thoroughly recognized and maintained to the very end as a person who acts; and that, faced with a relationship of power, a whole field of responses, reactions, results, and possible interventions may open up.’ (Foucault 1983: 220).
Second, authority is a social relation that parties see as legitimate. This follows logically from our initial (Weberian) definition of authority as legitimate power. The legitimacy of authority raises, however questions relatively to what is meant by ‘legitimacy’, a word that bears quite some ambiguity both in everyday language and in Weberian terminology. As we have seen above, Weber defines legitimacy as a form of non-instrumental obligation. But should legitimacy be thought in terms of obligations or in terms of permissions (or authorisations)? And more fundamentally, what is the ontological relation between obligations and permissions. Are they fundamentally different sorts of social beings or can the one be reduced to the other? To this question we shall return in this paper’s second section.
Third, the relation of authority is not merely a relation between persons. Rather, it is a relation between persons doing their best to personify a multiplicity of social roles. This point is perhaps where realist and pragmatic approaches to relations of authority diverge most visibly from Network theory approaches1. We dedicate the third section of our paper to discussing how the notion of a social role is articulated within the three frameworks of network theory; pragmatist social behaviourism and realist social ontology. This third section shall also study how different approaches to social roles secrete widely differing conceptions of authority.
Fourth, the legitimacy of a relation of authority depends itself on a network of social relations that is typically wider than the specific organisations within which it can be observed on a day to day basis. The contours of this network can be grasped by asking the question of whose legitimacy is being sought? Indeed, unless we are willing to restrict the use of the word ‘legitmacy’ to those situations which every rational being would recognise as such, we are left wanting for a theory (or at least some ontological meta-theoretical ground clearing) of authority that would help us account for the common situation in which only some participants recognise a relation of authority while others would rather question it. We shall explore this further in the later parts of this paper; at this point we can, however, raise a couple of questions directly addressed to network analysis. The latter traces the network of relations between agents and seeks correlations between the existence of such relations and objective characteristics of agents and network. But how can network analysis account for those reasons that cause agents to engage into (or avoid) certain relations? And, more crucially perhaps, how can it account for agents’ adoption of certain reasons rather than others? We shall come back to this question in more detail as we move from legitimacy’s implicit obligations to authority’s hidden networks.
2) Legitimacy, obligation and authorisation
Just like so many other words, ‘legitimacy’ is not devoid of ambiguity. Indeed, when we say that X is legitimate we can mean either that doing X is permitted, or we can mean that doing X is not only permitted but ought to be encouraged, that people engaging in X ought to be looked at favourably. Weber clearly choses the second of these meanings when he distinguishes between three (rather than merely two) basic cases. For him, a social feature can be illegitimate, or it can be non-legitimate, or it can be illegitimate.
This ternary distinction can be illustrated with an example from my current fieldwork with the Occupy Geneva movement2. This movement carries regular general assemblies which are open to all and in which any question can, in principle, be discussed. During these assemblies, some participants may bring a drink such as a can of beer. Up until the day when a formal rule was voted against alcohol in general assemblies, drinking beer was deemed to be an acceptable form of behaviour, thus it was not illegitimate in Weber’s sense. However, drinking beer was not legitimate neither in the Weberian sense since it was not perceived as an essential feature of a binding social order. Since drinking beer at Occupy Geneva’s general assemblies is neither legitimate nor illegitimate, a Weberian would qualify it as a non-legitimate action. As we can see, legitimacy is characterised by obligation; illegitimacy is characterised by prohibition; non-legitimacy is characterised by authorisation.
Intuitively, however, we can sense a contradiction in saying that drinking during a G.A. is not legitimate although it is not illegitimate. This intuitive understanding begs, however, a theoretical demonstration that the two propositions are equivalent. In other words, if we aim at moving beyond the contradictions generated by the trinity of legitimacy/illegitimacy/non-legitimacy, we are prompted to reconstruct the ontological relations that link obligation, prohibition and authorisation.
When we say in common language that social feature X (say: the practice of drinking beer at a general assembly) is legitimate, we mean primarily that it is forbidden to forbid X. And, vice versa, upon hearing that it is forbidden to forbid X, we can safely conclude that X is a legitimate social feature.
X is legitimate (in common language) it is forbidden to forbid X
But the expression ‘X is forbidden’ deserves further unpacking. When one says that X is forbidden, s/he means that people should refrain from doing X (if X is a practice) or engaging in actions that involve X (if X is an institution).
X is forbidden one has an obligation to refrain from X
It follows logically that, in common language:
X is legitimate forbidding X is not legitimate refraining from forbidding X is obligatory
Y is obligatory refraining from Y is not legitimate forbidding to refrain from Y is legitimate
This reconstruction of the relation between legitimacy, illegitimacy and obligation helps us putting the finger on a number of avenues available to those willing to move beyond Weber’s categorisation – beyond the trivial fact that everyday language treats legitimacy as primarily a matter of authorisation rather than obligation.
Firstly, the ternary distinction is unnecessary and can be expressed equally rigorously by mere reference to legitimacy and obligation. Second, although legitimacy is immediately a matter of authorisation rather than one of obligation, it is possible to understand it as a matter of obligation. Authority thus appears to be a relation of power which participants have an obligation to refrain from negating. This does not mean that authority is absolutely unquestionable but rather that a relation of authority will first be questioned on those aspects which seem less unquestionably legitimate than others. Moreover, such questioning of authority will mobilise other, more fundamental, obligations which participants recognise. This point shall facilitate our analysis of the link between relations of authority and those social obligations that are constitutive of social roles in the next section (section 3).
Secondly, our reformulation of legitimacy in terms of obligations indicate network features that would remain unnoticed from a strictly Weberian perspective3. In particular, it indicates that for a social feature to be legitimate for an ego, there must be an alter (real or imagined) who does the forbidding. Obviously, it is possible to conceive a limit case in which the same person takes the role of authorising and forbidding. Yet, even in this limit case, s/he must do so in turns must consider herself as an object to herself whenever she authorises herself to do X while, by the same token, obliging herself to refrain from forbidding X. And this is a limit case. The central case is rather that of people imagining either a specific other (say, the General Assembly’s convenor) or an abstract member of their community who will express a judgement of legitimacy that reflects the basic principles of the whole community (more about this in sections 3 and 4 below).
3) Authority and social roles
a) Obligations and social roles
Obligations, just like every feature of the personal, social and cultural realms are both structured and continuously subject to change. Because such change is relatively continuous, obligations are not punctual events devoid of past and future. Rather, they stretch through a certain period of time. Thus, if I promised to meet a friend in the pub every Wednesday evening, then I should still be bound by the obligation I contracted in a week’s time. Yet, obligations are seldom eternal and their continuity is relative: some day, I may have good reasons to stop feeling obliged to attend that weekly pub meeting. Exploring these reasons gives us some insight about the articulation of obligations in the personal, social and cultural realms. For instance, I may decide to quit drinking or to take my own promises more lightly (personal realm); my relation with that friend may get colder over time (social realm); or the practice of meeting in pubs can become out-of-fashion and other places might become more conventional meeting places (cultural realm).
The short illustration above indicates that the relative continuity of obligations over time presupposes relative continuities in the personal, cultural and social realms. In the personal realm, a continuity of concern, and those values they entail, is presupposed. Indeed, if my (positive) valuation of the pub as a meeting place is transformed, then I may feel inclined to suggest a different meeting place; if my valuation of punctuality is transformed then I should feel inclined to show up late; my concern for my friend’s wellbeing withers, then I may simply not show!
In the cultural realm, a continuity of meaning relatively to the content of our promise is also presupposed. If, say in 20 years’ time, virtual meetings become the convention, then I may feel inclined to suggest an internet meeting while expecting a positive response from my friend. Similarly, if the word ‘pub’ refers to what we call cafés, then the obligation might also be affected correspondingly. In short, my obligation is internally related to cultural emergent properties.
In the social realm, obligations presuppose – and in turn constitute – continuities that can be located at the level of social roles and of the relations between them. Back to the example of meeting a friend in the pub, the social roles I occupy bear on my being subject to the obligation of attending the meeting. For instance, if it is accepted that friends ought to keep their promises and care for one another, then I would have some obligation to attend and further obligation to let my friend know promptly in case of no show. I have argued elsewhere (Al-Amoudi 2007; 2010) that social rules are actually a particular kind of rules that is characterised by its internal relation to social relations between persons occupying social roles4. The same can arguably be said of obligations: not all obligations are social. My obligation towards my friend may or may not be social. If it is only structured by a concern with her well-being, then it is inter-personal without being social. On the other hand, it becomes a social obligation as soon as it is internally related to a social role. That is, when my ability to occupy a role is threatened by my ability to respect that obligation.
Understanding how obligations and social roles are mutually constituted is of import for the present study. Firstly, because if roles are excluded from the picture, the risk is to interpret people’s normative commitments only in terms of their personal attributes or in terms of the position they occupy within a network of relations. What would be missing in such a picture is an understanding of how people’s efforts at occupying competently their various social roles brings a contribution that is irreducible to the patterns of exchange in which they engage or even with any personal data the inquirer may gather about them. In other words, if I take my role as an employee seriously and if I am running late on a piece of work, then I may feel obliged to postpone that pub meeting. This obligation stemming from my social role as an employee depends on but can’t be explained away by reference to my personal integrity or the fact that I have been meeting that friend every week over the last 2 years. Secondly, understanding the relation between obligations and social roles opens the avenue for a study of the morphogenesis of relations of authority that spans beyond mere agreements between two individuals. An ontological study of the relation between roles, obligations and authority opens the gate for empirical studies of authority that attend to the morphogenesis of social roles and of the relations between these rather than mere re-configurations of personal networks that link individuals without much concern for those roles they are occupying.
One difficulty for such an ontological study is, however, that the notion of a social role is employed quite varyingly in the literature. In the context of the present paper, I endeavour to retrace its meaning, usage and import in two broad bodies of literature: network theory and realist social ontology. The next sub-sections are dedicated to precisely that.
b) The notion of a role in network theory
Network theorists distinguish between roles and positions. In their language, a position is a set of actors who have similar relations to all other actors in the network (Lazega 2007: 59). On the other hand, a role, or a role set, is the set of those relations as opposed to the individual actors who occupy them. Network theory leaves us, however, with the following questions:
Question 1: Are positions really of the same nature as persons? Approaches reducing positions to those people who occupy them have been widely studied and criticised by Archer (2000). The gist of the argument is that, although their continuing existence and structure depends on one-another, positions, relations and people are different kinds of things with different powers.
Question 2: Is it the fact that an individual is engaged in regular transactions with a specific set of others that is constitutive of her role? If my understanding of network theory is correct, then a nurse working exclusively for a patient and a cleaner working exclusively for the same individual occupy the same roles. Conversely, two nurses working with different patients would be occupying different roles. This result is not only counter-intuitive, it also casts doubts on the relevance of NT’s choice of words. More importantly, it prompts again the question of what robust conception of a role should be adopted by network theorists.
Question 3: Can there be roles without some form of recognition by the person occupying them? In the conception of a role above, it is sufficient for two persons to engage in a similar pattern of relations to be occupying similar roles. This would imply that they may be occupying a role without even knowing. We shall see below how a plausible distinction between roles and positions might evolve precisely around participants’ recognition (Archer 2000). While persons with similar vested interests may occupy the same position, they do not hold the same role until they become conscious of their shared interests and of the obligations implied by their defence.
Question 4: Are roles immediately occupied by individuals by the mere fact of engaging into relations of exchange with others? Or is the process of occupying a role a fragile and incomplete accomplishment? And if so, how is this vulnerability exploited within relations of authority?
c) Actors and roles in realist social theorising
It is noteworthy that, although the early works on realist social ontology, such as Bhaskar’s Possibility of Naturalism (1979) paved the way to subsequent realist developments, they also treated social positions and practices as slots which individuals occupy immediately instead of problematizing and enquiring into the subtle mechanisms through which individuals engage and commit into practices, positions and roles (for a critique from a realist Foucauldian perspective, see Al-Amoudi 2007). More recent developments, such as those of Archer since Being Human (2000, see also Archer 1995, 2003, 2007), open the black box of role personification. By the same token, they also address the four questions we have directed to network theory (Cf. supra).
Archer (2000) establishes a distinction between agents and actors. While the former are characterised by their creative actions within a collectivity, the latter are characterised by their identification with particular roles. Since the same person is typically both an agent and an actor, the question is set in terms of the dialectical relation between one’s personal identity and one’s social identity. Unless we accept to fall back to the contractualist time paradox of would-be actors picking a role with no good reasons, except those she will find ex post, we must admit that would-be actors are already endowed with a personal identity – if one that is incomplete and malleable. The pre-existence of personal identity is what accounts, in Archer’s argument, for the fact that actors are capable of personifying a social role rather than animating it. Agents (qua would-be actors) pick up a new role by reflecting on their past experiences of the natural, practical and social realms. In turn, once they started occupying a role, they discover the constraints and opportunities this role imposes and offers for the development of their personal identity. Thus, little Jane who enjoys horse-riding may want to work in a stable. As she learns more about the job and starts experimenting it first hand, she discovers that it is paid poorly and leaves little time for rest and leisure. This newly discovered constraint is of import for other concerns which she also holds dearly, such as spending time with her friends or dining in fine restaurants. In this light, Jane will have to choose between abandoning some of her personal concerns (relative to her personal identity), or look for another role (another social identity), or taylor her investment into that role (thus engaging an alignment of personal and social concerns).
This exceedingly short summary of Archer’s account helps answering the first and fourth questions I have raised above. Positions are not the same kind of things as persons, though the two are related dialectically. And the occupation of roles is never immediate but is the fragile result of agents’ creative attempts at personifying it while considering the totality of their concerns.
A slightly more detailed discussion is needed to answer question 2 (whether roles are defined exclusively by those transactions they entail?) and question 3 (do roles presuppose a prior recognition from part of those who personify them?) If a lead is taken from Archer (1995; 2000), then question 2 might be answered along the following lines: although transactions and relations are vital for the persistence of roles over time, the latter can’t be defined by those regular transactions which they entail, nor by those relations which allow for such transactions to take place in the first place. Rather, roles are characterised as bundles of obligations; vested interests; penalties and rewards (Archer 1995: 187). Archer (2000: 286) also hints at an interesting distinction between social positions and roles:
‘to take an example from institutional morphogenesis, when educational control was exclusively in Church hands, this created exigencies for a number of groups and where such problems represented a clash of beliefs, an obstacle to a nascent social movement, or the exclusion of a particular category, these could only be interpreted as inpinging upon roles by over stretching that concept to turn “believer”, “radical”, or “nouveau riche” into roles.’ (Archer 2000: 286)
While ‘believer’, ‘radical’ and ‘nouveau riche’ all entail vested interests and obligations, they remain labels for groups of individuals until they are recognised as such by their own members. By contrast, the interplay of these groups led in turn to the morphogenesis of ‘an array of new roles - teachers, administrators, inspectors and Ministers.’ (Archer 2000: 286). Thus, the answer to question 3 also has to be negative: an observer looking at people from a third person perspective can identify vested interests, regularities of practices and even commonalities of obligations (for instance by paying attention to systems of rewards and punishment). And participants may well be unaware of their vested interests, or unaware that their practices follow certain patterns. They may also be unaware that geographically distant persons share their obligations. What they can’t ignore, however, is whether they are occupying a role that crystallises these interests, practices and obligations into a social identity.
4) Authority’s hidden networks
The paper so far has attempted to establish novel bases for a study of the social networks involved in relations of authority. We have established in section 1 that authority is a relation of power whose legitimacy is recognised by participants. We then examined in section 2 whether and how legitimacy can be reformulated in terms of obligations and vice versa. Authority thus appears to be a relation of power which participants have an obligation to refrain from negating. This does not mean that authority is absolutely unquestionable but rather that a relation of authority will first be questioned on those aspects which seem less unquestionably legitimate than others5. Moreover, such questioning of authority will mobilise other, more fundamental, obligations which participants recognise.
Section 3 studied how, at the micro-sociological level, relations of authority involve typically (though not necessarily) social roles. These are not to be confused with the personal identities of those agents who occupy roles. Moreover, role occupation is best described as a process of personification that involve (while in turn constraining) people’s creative powers. Key to this process of personification, are the internal conversations in which agents engage routinely.
a) Rôle-taking in pragmatic philosophy
Network theory has offered us a valuable point of entry into roles and obligations and realist social theory has offered much needed correctives. The obligations relative to authority are typically, though not necessarily, bundled into social roles. The latter are in turn personified by actors who rely on their own reflexivity in the form of internal conversations that may be carried with varying levels of aptitude (on this particular topic, see Archer 2003, 2007). Because the internal conversation represents both a central social process and a key difference between realist ontology and network-theory, it is worth clarifying it further. In particular, we shall attend to the question whether the internal conversation presupposes a network of social relations and how this can be traced empirically. To this end, we shall perform a short leap in the past and discuss George Herbert Mead’s theorisation of the morphogenesis of reflexivity through a succession of rôle-taking exercises (Mead 1934/62).
Mead does not use the notion of role as such. However, he relies extensively on the related notion of rôle-taking. The latter is a characteristic feature of human interaction (as opposed to interaction amongst insects and higher mammals) and is described as: ‘assuming the attitude of the other individual as well as calling it out in the other’ (Mead 1934/62: 254). The particular significance of rôle-taking lies in the fact that it enhances the individual’s control over his own response and thus improves co-operative activity. Indeed, even criticism – of self and others – is closely related to rôle-taking: a critique formulated by a person is always addressed from within a particular standpoint, and with the expectation of certain reactions on the part of those to whom it is addressed. Conversely, a gesture (including uttering a sentence) that would be uttered without taking the attitudes of others into consideration can’t be meaningful, or can only be so for spurious reasons6.
Birds give us some insight as to the rudimentary mechanisms involved in rôle-taking (Mead 1934/62: 360). Each note they utter stimulates them and others into producing the next one. Yet, contrary to humans, birds don’t anticipate the reaction of the other bird, not to mention their own (future) reaction. With humans, instances of rôle-taking are particularly visible in children’s play, especially when it involves dolls. These games reveal, indeed, how ready most children are for the expression of the parental attitude in situations similar to those they imagine in their games. Through the sequence of interactions with and between dolls, the child stimulates in herself the responses of adults, that is, responses that belong in a certain sense to another. This process of rôle-taking allows Mead (and other pragmatist philosophers) to account for the relative stability of the social order without recourse to direct imitation of behaviour. There is a process of imitation involved but it is an imitation of attitudes, which allows for improvisations that are truthful to the principles of those serving as rôle-models.
Mead’s notion of rôle-taking provides a crucial insight into the functioning of the internal conversation. The latter does not merely involve a self (as I) conversing with herself (as Me) but also involves a (Meadean) you towards whom the internal conversation is oriented and whose attitude and plausible reactions are taken into account as the internal conversation progresses. Furthermore, the internal conversation does not necessarily take the form of a conversation of oneself with oneself imagined in front of another. Rather, the central case (especially in the early phases of the formation of the self) takes the form of a conversation of oneself with several imagined others.
The import of rôle-taking for the present discussion of authority’s networks should now be clearer. Social roles are personified by their human actors. Following Archer, this personification necessitates an internal conversation that seeks to align the actor’s personal identity (held as an agent) with the social identity provided by the role. However, following Mead, this internal conversation also mobilises countless attitudes which the actor has encountered during her life, both in others and in herself. This prompts novel research questions and associated methodological considerations for a study of the roles and networks involved in any relation of authority. To these questions we now turn.
b) Tracing authority’s hidden networks
Occupying a role presupposes a network of relations of authority that spreads beyond the specific organisation where it is located. In this last sub-section we provide a few indications as to how this ‘hidden network’ can be traced for the purpose of an empirical study of relations of authority.
Network theorists are eminently aware of the philosophical and methodological difficulties relative to specifying the boundaries of those networks they are purporting to analyse. Indeed, we agree wholeheartedly with Lazega (2007) that
‘le choix [des frontières du réseau à étudier] est fonction du processus que l’on cherche à examiner in fine.
Certes, il n’y a pas de découpage valable pour l’examen de tous les processus et phénomènes sociaux à la fois ; chaque découpage au niveau méso assure la lisibilité de certains processus au détriment de la lisibilité des autres.’
Our ontological ground-clearing can provide some directions as to how the boundaries should be set and as to what aspects of the network should be observed.
i) Specific attention must be devoted to participants’ belief in legitimacy (or lack thereof). Since authority is a relation of power that is viewed as legitimate by participants, any study of authority should interrogate the reasons why participants comply with those relations of power in which they engage: is it a matter of calculative interest? Or do they view them as legitimate, and for what reasons? Similarly, since relations can be complex: what aspects of the relation are considered as particularly legitimate, and which are less so?
ii) Specific attention must be dedicated to social roles. Observing individuals engaging into transactions is fundamentally insufficient. Similarly, attempts at reconstructing roles on the mere basis of those transactions in which participants engage are likely to be unconvincing. These may at best indicate patterns unknown to participants, however, they can’t account on their own for the normativity of social roles. That is, for the bundle of obligations they carry and, conversely, for their significance relatively to those relations of authority between actors.
Interviews shall explore the various roles in which participants are simultaneously engaged. They shall also explore what obligations/prerogatives these roles entail, what conflicts emerge between roles, how participants accommodate such conflicts and how much of their personal identity they are willing to invest for each of their roles.
iii) The contours of the network of authority shall consider rôle taking. The visible network of authority is provided by the regular transactions of participants and, once social roles are duly recognised, by those roles that are internally related to them (for an analysis of internal relations see Lawson 2003: 227-9, see also Al-Amoudi 2010). There is, however, a much less visible network that is provided by actors’ rôle-models. If our Meadean thoughts are taken seriously, then a network analysis of relations of authority should also include those rôle-models, real or imaginary, and specify which attitudes are generalised and transposed by actors.
iv) A diachronic approach is needed to understand participants’ socialisation. Since socialisation is neither immediate nor automatic, the question is prompted as how participants learned to interact within those relations of authority they must face in their activity. Thus, interviews shall interrogate them about those past experiences in which they were confronted with similar situations. This line of questioning is expected to flesh out past interactions in participants past-professional experiences, in their social milieu, in their university training and in childhood experiences of family and friendship (as well as hardships).
I have endeavoured to offer a succinct ontological study of authority. Its purpose is to clarify the ground for future empirical studies. It has been proposed that Max Weber’s pioneering definition of authority could be improved in several ways: by considering its relationality and circularity; by attending to the subtle relation between obligation and legitimacy; by recognising the import of social roles in the morphogenesis of authority relations; and by attending closely to the micro-processes of rôle-taking.
These ontological considerations bear implications for empirical studies of authority: i) Interrogate the reasons why participants deem certain relations more legitimate than others, and certain aspects more legitimate than others; ii) Attend to the normativity constitutive of social roles; iii) study the visible network provided by interactions and internally related roles but study as well the ‘hidden network’ provided by role models; iv) interrogate participants as to their past and study how past relations concur to shaping present ones.
One obvious limitation of this proposed approach is that it has to abandon any pretension at mapping an entire network exhaustively. Moreover, extensive mathematical crunching is expected to be impossible from the outset. The hope is that such an approach, centred on ego-networks, will nonetheless provide a thick description of relations of authority in specific social contexts and that it will result in maps that may include surprising participants (especially rôle models).