Journal of the Operational Research Society (2007) 58, 602 --613



2007 Operational Research Society Ltd. All rights reserved. 0160-5682/07 $30.00 www.palgrave-journals.com/jors

Drama theory: dispelling the myths J Bryant∗ Sheffield Hallam University, Sheffield, UK Drama theory is little more than a decade old and has already proved a versatile tool for analysing collaborations and conflicts. However, its widespread use has, it is argued here, been inhibited by a number of factors, some trivial, others more substantial. Following a resume of its emergence and a sketch of the ways that it has thus far been used, this paper sets out to examine and dispel the mist of prejudice, ambiguity and misinterpretation that has accumulated. Some specific application pathways are described, demonstrating how the strengths of the drama theory framework can be most effectively be exploited in practice. The paper includes some pointers for future development and exploitation. Journal of the Operational Research Society (2007) 58, 602 – 613. doi:10.1057/palgrave.jors.2602239 Published online 9 August 2006 Keywords: collaboration; conflict; drama theory; interaction; problem structuring; rationality

1. Introduction Drama theory made its first public appearance in a manifesto (Howard et al, 1992/1993) issued by its four co-founders following a meeting in Sheffield held in November 1991. The Manifesto—and the term was chosen in deliberate echo of the provocative and iconoclastic spirit, though not the political complexion, of Marinetti’s (1909) Futurist Manifesto, which had provided a crucial catalyst in destroying the certainties of pre Great War architecture—suggested a radical break with the traditional ‘rational choice’ paradigm underlying much of the social sciences. Like Marinetti, the authors demanded that new theory should embrace violent change and ‘universal dynamism’; and like the futurists the authors sought to depict the interaction of time, movement and emotion. The Manifesto was a creature of its time. The collapse of communism and the speedy end of the Gulf War, seemed to demonstrate America’s economic and military hegemony, expressed in the controversial phrase the ‘New World Order’. Apparently now ineluctably tied together, national, ethnic and issue groupings of drastically varying culture, power and mindset increasingly challenged each other, using political and psychological pressures rather than conventional military and economic sanctions to effect change or punish transgressions. Since then advanced technology, worldwide communications and religious zealotry have steadily transformed the mode of conflict: as Howard (1999) pointed out ‘the problem of defense in the modern world is the paradoxical one of finding ways for the strong to defeat the weak’. The interactions that take place in this global context rely more on ∗ Correspondence: J Bryant, Change Management Research Unit, City Campus, Sheffield Hallam University, South Yorkshire, Sheffield, S1 1WB, UK. E-mail: [email protected]

persuasion through credible threats and promises than they do upon the use of physical force: while Operation Desert Storm was an immediate physical success, the US lost the longerterm psychological conflict. Drama theory was put forward as a framework for making better sense of these paradoxes and for supporting considered action. A second, intellectual strand shaped the emergence of drama theory. While this can be viewed as part of the wider trend described in Rosenhead’s (2004) account of the development of problem structuring methods (PSMs), it is more helpful to identify the specific antecedents of the approach. Key to these was Howard’s (1971) metagame analysis, which explicitly recognized the frequent contradiction between the values combatants want to pursue and the credibility they need in order to pursue these preferences. This informed a structured technique for analysing conflicts that centred upon the ‘stability’ of possible scenarios: that is, upon what futures protagonists expect and, reluctantly or otherwise, accept. Hypergame analysis, conceived and subsequently developed by Bennett (1977, 1980) extended this approach by explicitly incorporating a multi-perspective construction of conflicts, thereby recognizing the plurality of ways of subjectively framing a situation. During the late 1980s, intensive private meetings and discussions between originators of some of the principal PSMs provided cross-fertilization, especially about group support processes, and led indirectly to the production of three key texts (Bryant, 1989; Rosenhead, 1989; Eden and Radford, 1990) in the broader field, the first, in its sub-title, heralding use of the drama metaphor. Added impetus for using drama as a ‘framework of inquiry’ (Bryant, 1988) rather than simply as an analytical technique was gained from the pioneering use of action methods and the techniques of sociodrama (Sternberg and Garcia, 1989) at the Community OR Unit (then linked to Sheffield City Polytechnic).

J Bryant—Drama theory: dispelling the myths

The rhetoric of the drama manifesto was developed through further regular meetings with a slightly wider membership, blossoming out into occasional conferences and public events during the early and mid-1990s. The earlier papers tended to present a summary view of the approach for new audiences, with much attention being paid to the relationship of drama theory to game theory (Howard, 1994, 1996; Bennett, 1995; Bryant, 1997). Later work shook free of this inheritance and showed how, by taking the notion of drama more comprehensively to heart, new forms of intervention involving role-play (Bryant and Darwin, 2003, 2004) might be used to stimulate and support change in organizations. Applications using drama theory as a way of exploring strategic options have continued to be made in the military arena, notably on ‘winning’ Operations Other Than War (OOTW) (Howard, 1999; Smith et al, 2001). Elsewhere, the contemporary prevalence of strategic alliances and joint ventures in the private sector and of partnership working there and across the public sector have led to a rich vein of opportunities summarised in Bryant (2003). In parallel with and often prompted by this practice the analytical base of the theory has been progressively refined and extended (Howard, 1998, 2004), while a number of related software tools have dimly appeared (Bennett and Tait, 1993; Bryant, 1998). Today the approach sits squarely alongside other PSMs in relevant texts (Bennett et al, 2001) and university courses. It might be surmised that drama theory has developed to the point where through a gentle process of diffusion its more widespread adoption is simply a matter of time. But there are some worrying signs. Although it is only a decade since the manifesto appeared, a decade is still sufficient time for a methodology to take hold. While awareness of drama theory, at least at a conversational level, is quite widespread, significant expertise with the approach has not been widely developed, and the continuing extension of the analytical framework still lies squarely in the hands of the originators. The argument of this paper is that a less passive approach to its dissemination is therefore essential. This is because, while drama theory shares with other PSMs a number of barriers to its deployment (see, eg, Bryant and Chin, 2000), its language, antecedents, application and toolbox have unfortunately given it additional special problems of its own! Until these are addressed its widespread use appears improbable. The aim of this paper is to challenge the myths that limit the potency of drama theory as a PSM. While its argument makes some reference to the generic sources of resistance to PSMs, the prime focus here is upon those that are particularly problematic for drama theory. In the next section some superficial, even frivolous, yet damaging associations and prejudgments are dispelled. This is followed by two sections that explore some of the commoner misconceptions about the focus and intent of the theory, beginning with an explanation of the role of drama theory as a theory of ‘irrational choice’. It is acknowledged in the following section that some of the ‘tools’ of drama theory have contributed to its inaccessibility

603

and some new representations are suggested here. Taking a less defensive stance, the paper then moves on briefly to describe some of the ‘success stories’ of drama theory and this develops into the final section where some opportunities for the future are presented.

2. What’s in a name? If it is familiar at all to OR practitioners, drama theory is usually thought of as an arcane branch of the strongly normative and prescriptive field of game theory—and in a sense this is perfectly fair. As ‘soft game theory’ (Howard, 1993), drama theory is a response to one of game theory’s major deficiencies, the paradox of rational action. However, while drama theory has benefited from a formal logical language inherited from game theory, it has suffered from a lack of recognition of its very different intentions. At the same time, it has made little impact on mainstream game theory and is seldom referred to in game theory journals. The strongly scientific, functionalist emphasis (Morgan, 1980) of game theory has coloured the image of drama theory, the mission of which has been misunderstood as being to create an objective and value-free representation of a real world of social relationships. It would be more accurate to characterize drama theory in Morgan’s terms as fitting (like other PSMs) within the interpretive paradigm, trying to inform processes ‘through which shared realities arise, are sustained and are changed’. Clearly, the models produced by a drama theorist could be used in a functionalist manner, generating prescriptions for action that assume a network of ontologically real relationships and describing a form of adaptive behaviour by the individuals involved. However, it is more productive to view inter- and intra-organizational life as centring upon sense-making, and drama theoretic analysis as being about understanding and supporting how people negotiate meaning. For this view to prevail, the expert analyst must take a back seat and be supplanted by the workshop facilitator. Unfortunately, drama theorists have not always helped themselves in this respect, often implying (eg Howard, 1999) that the masterly skills of an external consultant are essential for success, rather than allowing that individuals or groups might be able to use the framework to help themselves to work through a conflictual situation. With other, less specialist, audiences, the drama theorist faces different prejudices. Most obviously (but not least damagingly) referring to a situation as if it were a drama risks trivializing it (if not as badly as calling it a game). More subtly it may connote a belief that the cast list is easily bounded, that the stage is a self-contained world apart, even that characters’ moves are scripted. Confusion with other uses of the drama metaphor (Mangham, 1978; Mangham and Overington, 1987) is also understandable, but can be dispelled by noting that drama theory moves beyond being a merely linguistic or descriptive device for talking about interactions to making predictions about the feelings that characters will experi-

604

Journal of the Operational Research Society Vol. 58, No. 5

ence in an interaction and the ways that they may respond to these. In some ways the most awkward challenge for the drama theorist comes from the most superficial interpretation: for a common riposte to its mention is for a client to say ‘Oh, you mean role play’. Now one of the main ways that drama theory has so far been used is indeed to support personal and organizational development through role play (Bryant and Darwin, 2003), but its use there is as an analytical device to shape the role play so that participants will experience particular, designed challenges. This is a very different matter from creating an identity between drama theory and role play purely on account of the ‘drama’ label.

3. Misunderstanding irrationality The core concept of rationality is both the strength and the Achilles heel of game theory as a framework for understanding those situations the outcome of which depends upon the actions of several autonomous participants. For Von Neumann and Morgenstern (1944), acting rationally could be expressed by a ‘set of rules of behaviour in all conceivable situations’ that enable a participant in a social economy to achieve a maximum of utility. Establishing these rules for each player—the research agenda of game theory—is a purely technical (though non-trivial) problem of predicting others’ (not necessarily rational) moves and then identifying one’s own best course of action. Game players—consider footballers or chess masters—independently assess situations and take their ‘best’ action. But rationalizing independently can lead to trouble. Consider, for instance the story of the Jerusalem taxi driver related (but not apparently understood) by Dixit and Nalebuff (1991): Nalebuff and a colleague were returning to their hotel late at night and hailed a taxi. The driver headed off but didn’t turn on his meter. The two game theorists quickly reasoned that their best strategy was to delay any bargaining over the fare until they arrived at their destination, where the driver would surely be prepared to take whatever they offered. Sure enough at the hotel the driver demanded what his passengers felt was an inflated fare so they countered with a lower offer. He was outraged and in a fury locked the cab doors and drove them back to their starting point where he bundled them out onto the pavement.

The authors of this tale treat it as a bizarre aberration which they try to explain using atheoretic ‘common sense’: but the events are anything but bemusing to anyone except a na¨ıve game theorist! Taking the ‘game’ between the driver and passengers as fixed, delaying the fare offer is indeed rational; but fixity is a wrong assumption. The unexpected rebuff to the driver’s demand prompted him to conceive a new option (‘return to the start’) which he promptly put into practice, thereby changing the game. Now the rules of social games are not fixed. They are actually determined by the players, either through accepting

established laws, routines, customs and cultural mores or, more rarely, by explicit negotiation. Furthermore, people’s perceptions of such games are highly subjective: what is valued, what is believed, and what is considered possible are all susceptible to influence. And more drastically still, views as to who is involved or even what is going on are not necessarily shared. Players schemingly alter games so that they can achieve an outcome that represents a resolution of what is going on. Rather than merely seeking by rational choice a game-theoretic equilibrium, they transform the game in their search for an outcome that leaves them in Milton’s words ‘calm of mind, all passion spent’. Drama theory puts its emphasis upon this developmental process, seeing a game as just one ‘frame’ in a sequence that allows participants to strive for self-realization. The proposition is (Howard, 1996) that the development is driven by a desire on the part of participants to eliminate the dilemmas that frustrate their pursuit of rationality, where rationality is taken to mean ‘choosing in accordance with one’s preferences’ (Bennett and Howard, 1996). So the taxi driver who carries out his unwilling threat to forego his fare altogether does so by a reframing of how he sees things and a change in his preferences that is coupled by strong emotion. His action is irrational in terms of his (and his passengers’) prior framing of the situation, but is consistent with his new perspective on the interaction. Ever the game theorist and still not seeing this, Nalebuff’s concluding remark that the ‘passengers didn’t use their game theory analysis quite far enough: next time, they’ll get out of the taxi before discussing price’ seems doubly na¨ıve: as Howard wryly remarked (1997) ‘Someone should warn them not to try it. They may end up getting shot’. The claim that drama theory is a theory of irrational choice has caused some scepticism. If choice is indeed irrational, then how can it be captured in any systematic theory? As the illustration above demonstrates, this doubt reflects a misunderstanding of what is here meant by irrationality. In drama theory behaviour is described as irrational if a person acts against his or her preferences. This contrasts with the extreme game-theoretic position that choice reveals a person’s preferences, a view that drama theorists maintain drains the term ‘rationality’ of any substance. Crucially, drama theory does not say that irrational behaviour is arbitrary. Typically, it is in accordance with preferences that the person is under pressure to assume, and to which he or she may be in the (emotional) process of shifting. Of course, the behaviour may simply be a ploy to convince others that the person is ready to ignore his/her preferences. In either case though, the behaviour serves the purpose of establishing the credibility of a promise or threat. Preference change is a key factor in drama theory and it is taken to be associated with an emotional response to the paradoxes thrown up by seeking to act rationally. The line that has to be drawn between actions that are described as rational and those that are considered irrational is based on assessment of the extent to which the choices made are rooted in a person’s pre-existing, fundamental system of

J Bryant—Drama theory: dispelling the myths

values. So we take as rational an altruist who calmly comes to the assistance of a distressed stranger, but as irrational someone who ‘quite out of character’ and maybe momentarily moved by pity or affection dashes forward to help. This principle is of fundamental importance in an approach that has, like game theory, been criticized as amoral (or even immoral) for encouraging ruthless self-interest and the superficial pursuit of immediate gain. That such criticisms are misguided is neatly demonstrated by the last example of benevolent behaviour, since drama theory explicitly requires the tension between a person’s deep-seated beliefs or needs, and the demands or opportunities of the moment, to explain the driving force for emotion and preference change. The only sense in which the theory can be viewed as amoral is in its lack of any assumptions about people’s specific motivations. So—and this is in common with all other modelling approaches—the value judgements and motivations that, for instance, lead to selfish or altruistic actions, must come from outside the framework.

4. Understanding the moment of truth It is through interaction between ‘characters’ that events unfold. The general drama-theoretic model of this process is shown in Figure 1, where each block represents a process whose ‘output’ is stated in an arrow. The broad boundaries of the situation are established in the Scene-Setting phase where the set of possible subjective ways in which participants may frame what is going on is delimited. Through dialogue the participants reduce this set to a single ‘common reference

605

frame’ (CRF) that provides a basis for communication and negotiation. The CRF enables a definition of who is involved (the ‘characters’), what they are able to do (their ‘options’) and what they are proposing should be done (their ‘positions’), as well as the steps they say they will take (for each, its ‘fallback plan’) if their propositions are declined. Note that positions are what characters are demanding, not what they’re just prepared to accept; and that fallback plans are effectively their Best Alternatives to a Negotiated Agreement (BATNAs) (Fisher and Ury, 1982) though here we recognise that characters may make incredible threats. The CRF must be simple enough for everyone to be sure that it is common knowledge (each knows that the other knows that they know. . .what they mean), although it may not represent all participants’ real views (since there may be deceptions); nor does it represent shared values, though it can be translated into each of their terms (eg, one person’s ‘surrender’ is another’s ‘cease fire’). So there is no assumption—a common misunderstanding about the concept of the CRF—that interactors either agree or ‘see things the same way’; only that their communicated intentions are understood: since, for example, a threat that is not seen as a threat is not a threat! Indeed only a utopian would believe that a CRF is all that is required to resolve a conflict: far from it, for the CRF highlights protagonists’ differences and defines the problem jointly to be solved. Furthermore, since the CRF is essentially a model of participants’ mental models, which themselves are (arguably) manageably simplified, focussed and selective versions of ‘what’s going on’ its spare simplicity is defensible;

Figure 1 The process of interaction.

Journal of the Operational Research Society Vol. 58, No. 5

5. The need for new representations

PREDATOR Buy more of Target Cut services offered by Target Invest in alternatives for Target Integrate engineering Advance loans to Target TARGET Sell more of self to Predator Accept more loans Cut services Develop alternatives Change culture

Figure 2

Options board for business takeover.

Default Future

Fallback Future

One possible source of the confusion noted above stems from the shared reliance in drama theory and in some forms of game theory upon a tableau model of a frame/game. The tableau is used to organize the possible futures that characters’ intentions may co-create. It is a compact representation since, in contrast to metagame analysis (Fraser and Hipel, 1984) its core includes just those futures implied at the moment of truth, augmented, if desired, by any further futures that are sources of doubts (eg as temptations for someone to renege) for any character. The small example shown in Figure 2, taken from a study of a takeover battle between bus operators, includes five options for both parties and, shown as columns, only four key futures (rather than 210 as might appear in a metagame analysis). The leftmost column shows Predator’s position which is to increase control in Target and use this opportunity to leverage changes in culture and operating practices. It is easy to see how the cell labels in this tableau could lead to bafflement for problem owners, who might possibly read the options as time-dependent and read them as actions, rather than as proposals for action. The precision of the tableau display is also unfortunately matched by its opacity to many who encounter it and find it hard to read the separate choices in each cell of a column as together shaping a possible future. The difficulty is redoubled in those representations that use the visual icon of ‘cards’ that are played or not played, for this confuses people who may hold an image of a card game like poker in which choices are sequential rather than simultaneous and in which the

Target's Position

indeed expected. So characters’ options included in the CRF are just sufficient to define the positions, fallbacks and any recognized temptations. The point at which positions are frozen within the CRF is termed the ‘moment of truth’. If characters at a moment of truth have compatible positions (ie they do not disagree over the moves that anyone should make), they are ready to collaborate in resolving the situation. Provided that they can trust each other to deliver their individual commitments to the cooperative solution, then the way is clear to implementing their agreement. If however the characters do not agree on a joint position that they can trust each other to implement, they will be under pressure to change their views. However, such changes are not easily made, and drama theory asserts that emotional energy is needed to overcome the inhibiting ‘friction’. These emotions are systematically associated with the desire to overcome one or more of the six possible generic dilemmas that it can be shown (Howard, 1998) characters may experience by trying to be rational. A character may respond in four ways to any dilemma: by changing its position; by amending its preferences for the possible outcomes; by denying that the dilemma exists; or by taking irreversible unilateral action. Furthermore, all but the first of these may be feigned (by definition a position is a declared view and cannot be feigned, though it may be insincere). So there is a rich repertoire of possible thought pathways and associated interaction strategies at a moment of truth, and it is these above all with which drama theory is concerned. The focus in drama theory on the moment of truth cannot be over-emphasized. This has frequently been misunderstood by those who do not realize that drama-theoretic modelling is of the phase of ‘pre-play’ communication: it is about what happens in the complex and critical stage that occurs before any actions are implemented. Indeed, once firm commitments to act have been made, the eventual denouement is dully predictable, barring only any flunking of threats or reneging on promises made by one or more of the participants in the light of their private evaluations (which may be informed by game theoretic models) of the developments that they have said they would jointly set in train. To emphasize this distinction, consider behaviour in a broad-based community partnership led by statutory bodies but supported by voluntary sector and residents groups. The lead organizations might fear that lack of resources could result in the smaller agencies making only a token commitment and ‘free-loading’ to obtain benefits. Drama theory would explain the pressures felt and imposed by all parties to achieve a meaningful agreement and a fair basis for contribution: game theory might be used to understand why once such an agreement had apparently been reached, one party then failed to deliver what it had offered. So, in this sense drama theory and game theory are complementary, each dealing with a distinct part of the overall process.

Predator's Position

606

J Bryant—Drama theory: dispelling the myths

Figure 3 (a) Conventional tableau for interrogation interaction. (b) Tug-of-war diagram for interrogation interaction.

characters are necessarily competing. One alternative representation has been the use of ‘jigsaw pieces’ (Bryant, 2002c) on which an option taken by each character is stated: this also emphasizes that futures can be created cooperatively by a group, rather than placing the emphasis on conflict. However, while this give a holistic sense of each future, it is cumbersome to display and hard to amend, for example if fresh options are added. Alternative representations also fail to facilitate the comparison of different futures, which is an essential aspect of confrontation analysis, and is shown with great clarity in a tableau. Preference change is central to the theory, but is hard to represent in a simple tableau showing characters’ preferences for possible futures. The usual approach is shown in Figure 3(a). This depicts the archetypal situation of an ‘Interrogator’ (I) with a ‘Subject’ (S), whom we shall suppose holds information that I would like to access. Ticks and crosses signify whether or not options (rows) are taken within each outcome (column), while the numbers against each character (rows) indicate its rankings (1 = most preferred) across these outcomes. The rankings, used here purely for illustration, imply an interrogator who is prepared to wait, perhaps until a later time, rather than exert undue pressure, and a subject who is not prepared to be open except under duress. While this static picture contains all the information required for a confrontation analysis—the dilemmas facing each character can be fully established—it does not show the direction of pressures for change. This is particularly unfortunate in an approach which is inherently centrally concerned with transformations of the frame. At the same time it is over-specified, because a full ranking of all futures may not be required to

607

establish the dilemmas facing characters in a situation. Since stating a full ranking may also be problematic for users, it is both more efficient and more reliable to identify dilemmas by answering just those questions that are necessary to do so. By asking whether or not the relevant character wants to take each of its options, the pressures can be teased out. Reluctance to implement threats or promises is the sure symptom that a character faces drama theoretic dilemmas. But before the event, at the moment of truth where drama theoretic analysis is located and when this reluctance has not been put to the test, it can only be sensed as doubts by the other parties about a character’s intention to act. So the questions that must be asked to elicit the dilemmas concern these doubts. Specifically, we ask about each character if there are doubts as to whether it may carry out its expressed intentions, or whether its wishes may be achieved. So, in our example, doubts about I’s willingness to exert pressure (its fallback plan) encourage S to prefer to stay quiet: for its part, I would have doubts about S’s willingness to disclose information (S’s contribution to I’s Position) despite the possibility of duress. Respectively these point to a Threat and a Rejection Dilemma for I and a Persuasion Dilemma, also for I. I would also doubt if an agreement by S to disclose (I’s position) would be honoured, and so additionally faces a Trust Dilemma. In contrast the situation as framed poses no dilemmas for S. From the very outset, drama theorists have used diagrams to support analysis. These were initially adaptations of the strategic maps (Howard, 1989) employed in metagame analysis to show improvements and sanctions for players between scenarios. The most recent version of these is the ‘tug-of-war’ diagram (Howard, 2004) shown for our running example in Figure 3(b). Here arrows represent those dilemmas implied by eliciting doubts about each character’s intentions, which in turn imply the indicated preferences (where an arrow leads ‘into space’ this implies that there is a reachable, but here unstated, outcome that is preferred). Figure 3(b) is just a specific case of the general diagram in Figure 4 which shows all possible dilemmas in a drama involving characters A and B. Tug-of-war diagrams can be read intuitively. So the arrows that lead to S’s Position in Figure 3(b) mean that I is under pressure to accept this (the Fallback is worse for I), while S is under no pressure to accept I’s Position (since S prefers the Fallback). Drama theory asserts that each party will use emotion and reason as it attempts to ‘pull’ all the arrows towards its own Position. So if I, perhaps to meet the expectations of its own superior, can reverse its preference for S’s Position over the fallback then it will dispel its Rejection and Threat Dilemmas and create a Persuasion Dilemma for S. Such a change would predictably be accompanied by reluctant hostility by I towards S and a feeling of indignation that I has been ‘forced’ to act in this way. Reason—the need to placate I’s superiors—would be used to convince S that I’s shift of stance is in earnest. The vertical arrows in the diagram signify temptations to renege on agreements. In Figure 3(b) S’s opportunity to ‘clam up’ at any moment is a problem for I; this

608

Journal of the Operational Research Society Vol. 58, No. 5

Positioning Dilemma for A

Threat Dilemma for A Co-operation Dilemma for A

Trust Dilemma for B Persuasion Dilemma for A

A’s Position

Fallback

R ejectio n Di lemm a for A B’s Position

Future Rejection Dilemma for B

Persuasion Dilemma for B

Trust Dilemma for A

Co-operation Dilemma for B Threat Dilemma for B

Positioning Dilemma for B A’s preference B’s preference

Figure 4

Drama-theoretic dilemmas in a 2-character interaction.

can only be allayed by S helping I to believe in its willingness to disclose, perhaps by offering some commitment which would irreversibly taint S in the eyes of its own supporters, or by warm and positive emotion that would encourage I’s faith in its sincerity. An important factor in the above analysis and a usual feature in drama theoretic analyses is the interaction between a number of distinct arenas: in this case the relationship between I and its superior on the one hand and between I and S on the other. Linking of frames, whether hierarchically (summarized in the aphorism ‘every character is a drama’) or otherwise is to be expected in such modelling and can be represented visually using modern derivatives of the preliminary problem structuring diagram (Bennett et al, 1989) developed for hypergame analysis. Such pictures are also used to provide a comprehensible route into computer-based drama theoretic summary of a situation, as shown in Figure 5. Even so, feedback from those who are new to drama theory indi-

cates that present modes of representation are often confusing. Recent work (Howard, 2004) still imposes too many laminations of meaning upon the tableau as a multi-purpose communication device, and new approaches must be developed. The episodic model of drama theory neatly captures the evolution of situations as protagonists test each others’ resolution and feel their way towards the future. Unfortunately, it also suggests a staccato process, lurching from crisis to crisis, rather than underlining the steady nurturing and development of relationships between characters. This impression runs strongly counter to the practical applications of drama theory (eg Howard, 1999; Bryant and Darwin, 2004), which have been centrally concerned with the management of ongoing inter-organizational relationships. The paradox arises from different perceptions of the level of analysis involved. Drama theorists contend that there is a need for consistency between the management of short-term, local confrontations and wider strategic objectives. Indeed the notion of a computer-

J Bryant—Drama theory: dispelling the myths

Figure 5

609

Computer-based immersive drama interface.

supported command and control system has been proposed for use within the military context (Stubbs et al, 2000), precisely to ensure such drama-theoretic consistency. The coherence of strategy is taken to derive from a character’s persistent, pervasive, underlying values, which is why they are taken so seriously in defining rational behaviour in specific, immediate interactions. It is a central tenet of drama theory that characters will seek to eliminate the dilemmas that they face: in doing so they may (intentionally or otherwise) add to the dilemmas faced by others. The transformations of the frame achieved (eg changing preferences, thinking of new actions, involving or excluding other characters) amount to a confrontation strategy for the character concerned. So if, in our example, I achieves the changes suggested above, and S fearfully comes to prefer I’s Position to both the Fallback and its own initial Position, then I’s confrontation strategy will have removed all its own dilemmas (apart from an inconsequential Threat Dilemma concerning its need to exert pressure on S) while S faces no less than five distinct dilemmas! However, drama theory is not normative and so cannot prescribe a confrontation strategy for any character: all it can do is to permit the exploration of the consequences of alternative patterns of dilemma elimination. Some may be ‘better’ than the alternatives (in the sense of removing a character’s own dilemmas) but the eventual outturn depends upon the strategies of other parties which define the branches taken through the episodic ‘tree’. This explains why drama theory makes no claims to being predictive in the conventional sense. As Bennett (2004) has explained what it can do is conditionally to anticipate behaviour, rather than ab-

solutely to predict outcomes: in his words ‘[to] provide some alternative prognoses, along with some indicators of which events would favour one alternative over another’.

6. A process of managing confrontations and relationships Drama theory could be unfavourably compared with other PSMs that are portrayed as ‘iterative inquiring’ methodologies that stress the importance of individual and organizational learning. If this aspect of drama theory has not been effectively communicated then the fault probably lies in the emphasis upon one-shot analysis in many of the case studies that have been reported. In turn, this stems from the frequent use of a single tableau model (eg Bryant, 2002b) to summarize a situation, instead of showing a dynamic sequence of such models tracking its development. Fortunately, some recent publications (eg Obeidi and Hipel, 2005) have begun to correct this view while Howard’s virtuoso account of confrontation management (Howard, 1999) showed just how effective drama theory can be for supporting a structured exploration of the future. The drama-theoretic model of an episode (Figure 1) with its underpinning mathematical logic (Howard, 1998) supports more than the analysis of conflicts, since by enabling a deeper appreciation of power relations and of personal responses, it can shape interventions and deliberate action to develop relationships. While some work (Bryant, 2002a) has perpetuated the style of post hoc analysis that dominated antecedent approaches (eg Bennett and Dando, 1979), many applications

610

Journal of the Operational Research Society Vol. 58, No. 5

Figure 6

A process for managing collaboration using drama theory.

have come closer to ‘real-time’ support by offering models to serve as ‘facilitative devices’ (Eden, 1989) for managing the conversation among protagonists in conflicts. So Bennett (1995) took ‘backroom’ models of the Northern Ireland conflict to Belfast where they were used in dialogue with (though not between) politicians from the main parties. His subsequent work (Bennett, 1998) has seen it used behind the scenes to inform Department of Health policy in the UK on such issues as genetically-modified foodstuffs. Other work in the health arena (Bryant, 2002b) has used multiple analyses to address such generic challenges as the so-called ‘blame culture’. As a way of helping an individual or group to make sense of what is happening, drama theory provides a framework for inquiry (Bryant, 1988), asking a systematic, structured set of questions about a situation, and challenging those who provide the answers. Taking a stand for drama theory to be regarded as more than just an analytical tool means offering it as a rounded methodology that can guide the management of an intervention process while shaping the raw material of what is taken as the substantive content of a situation (Eden, 1990). Here, drama theory might initially seem to be at a disadvantage relative to other PSMs that have clearly articulated process frameworks (eg the ‘learning cycle’ of Soft Systems Methodology (SSM) or the four ‘modes’ of decision-making in the Strategic Choice Approach (SCA)) that are brilliantly suited to delivery through training programmes. But actually there

do exist guidelines for such a process, sketched in Figure 6, and described more fully in Bryant (2003). This is similar in status and intention to the Journey-making approach of Eden and Ackermann (1998), which also shows how what they call ‘strategic conversation’ can be enhanced by models, in their case using the cognitive mapping technique. Drama theory has quite understandably become associated with conflict analysis. Case studies from the Cuban Missile Crisis onwards have stressed this linkage. However, this tie is not altogether helpful for an approach that has as much to say about cooperation as it does about confrontation. It is particularly unfortunate in those settings where ‘conflict’ is treated as a ‘dirty word’ or is not ‘politically correct’, though as has been pointed out (Bryant, 2003) conflict is neither an aberration nor a pathology but the natural order of things when people who care about what transpires try to work together. Indeed, conflict offers many positives: it stimulates creativity, challenges groupthink and encourages commitment. Despite such arguments, drama theorists have sometimes bowed to the weight of prejudice and opportunistically submerged the language of conflict under more pacific terms: ‘confrontation analysis’ has become ‘collaboration analysis’; ‘position’ has become ‘proposal’; ‘threat’ has become ‘encouragement’; and so on. The most convincing severance from the belligerent world of conflict is represented by applications (Bryant, 2003) that focus specifically upon collaboration. In a world where the

J Bryant—Drama theory: dispelling the myths

imperatives for organizations to work together are stronger than ever (Bryant, 2003) this arena represents a major opportunity for drama theory to be used.

7. Opportunities for development In this concluding section, some of the separate proposals introduced above for the consolidation and promotion of drama theory as the basis of a more widely used PSM are brought together in a more coherent manifesto for development. However, the achievement of this will need to be complemented by other measures designed to overcome the barriers both general and specific that still inhibit change. Drama theory is a theory about how people handle interactions with others: it’s use as a PSM is just one of its applications. Nor is confrontation analysis (Howard, 1999) a methodology: it is a technique for investigating a selected interaction. The term ‘Confrontation Manager’ has been appropriated and trade-marked to represent the functionality of the most ambitious software tool (Idea Sciences, 2005) so far devised to support confrontation analysis. One response to this diversity of labels might be to press for an unambiguous and specific term by which to refer to the wider process shown in Figure 6 above: ‘interaction management’ might suit. However, such uniformity is probably of more importance for developers than it is for users or clients: the latter will be more concerned with the predicament they are trying to handle than with the name of the approach that might be used to throw light upon it. The advice given about the avoidance of jargon in using mapping approaches (Bryson et al, 2004) carries over here: the ‘road map’ provided by the process diagram is sufficient for people to log their progress and should be used as such, as is commonly done in SCA or SSM. At the same time, the word ‘drama’ carries an exciting sense of energy, emergence and emotion: used judiciously but responsibly it may engage participants in the structured exploration of the challenges they face in working with others. The mathematical base of drama theory is ‘soft’ game theory (Howard, 1998). This explains how a game may change as players seek to eliminate the dilemmas they experience. However, there is an understandable hunger for predicting the outcome of conflicts (Green, 2002) which drama theory cannot address. Research is needed to examine empirically how people choose to eliminate the multiple dilemmas that they typically face, for this will suggest how a confrontation could plausibly develop. Unfortunately, the results of such experiments as have been carried out are not yet in the public domain: until they are, scientific confidence in the theory will be limited. One of the great strengths of drama theory is that it explicitly models changing perceptions of a situation. However, despite emerging devices like the ‘tug of war’ diagram which makes the tensions in the conventional tableau more apparent, the dynamics are still too weakly represented in current user interfaces and models. Furthermore, the vital element of

611

emotion is often offered as a cosmetic lamination in analysis rather than as a fundamental component. Further work is required to create effective ways of representing and communicate the drama and possibilities of collaboration. For instance, short animations featuring computer-based avatars might vividly enact possible future pathways. The principle that ‘every character is a drama’ emphasizes the nested nature of confrontations. Yet it too is something that is only clumsily captured by contemporary models: for instance, as a link between options across a hierarchy of tableaux. Nearly 20 years ago the ‘arenas’ notation (Bennett and Cropper, 1987) was suggested as a way of representing aspects of the complexity of interlinked conflicts, and a version of this that is consistent with drama theory is now required if the complexity of real-life situations is to be convincingly captured. The ‘neutrality’ of their conceptual frameworks is one of the strengths of all PSMs. Each provides a language for representing some aspect of an experienced situation: the language is general, though the eventual model is specific. This is in contrast to the contribution of much management research, for instance in the strategy area, where the outputs are strongly normative. However, the corresponding price that has to be paid is the ‘cost’ of bespoke modelling each fresh situation. In the present context, for instance, the time cost of authoring the ‘immersive dramas’ created for the Department of Health (Bryant and Darwin, 2004) exceeded by two orders of magnitude the duration of the role-plays created. This price is largely incurred as management time in collaborative model construction, and while it is convincingly argued that such a process creates ‘buy in’ and commitment to action, it may nevertheless present too high a threshold for routine strategy-making. There is a need to explore the ground between the specific models used in OR and the generic models used elsewhere. Experience in presenting drama theory as a problem structuring framework to MBA students suggests that it readily enthuses them. They find it intuitively attractive as it meshes well with their everyday notions of politically driven change processes. Indeed, it sits neatly beside the ‘Power School’ of strategy (Mintzberg et al, 1998) and with current thinking (Buchanan and Badham, 1999) about the politics of organizational change. However, this attraction may be superficial: the driving metaphor appeals to people’s common-sense while providing them with a fresh way of thinking about strategic interactions. What is difficult is to lead novices beyond the metaphor so that the strength of the analysis is not lost. Unfortunately, few people are well-equipped to conduct such analyses that cannot yet rely upon ‘user-friendly’ software tools (as in similar areas such as SODA or Decision Conferencing) to support its process or underwrite its conclusions. This is another obvious lacuna that needs urgently to be filled. If PSMs, interaction management included, are to have the positive impact of which their proponents believe that they are capable, either singly or in combination (Mingers and Rosen-

612

Journal of the Operational Research Society Vol. 58, No. 5

head, 2001), then these will have to be overcome through demonstration, education and publication. There is still too little clear evidence available to assess the extent of promulgation of PSMs, despite some heroic attempts (Mingers and Rosenhead, 2004) to do so. The testing time will come soon as the originators of these approaches step aside and leave their fate in the hands of a fresh generation of practitioners and researchers.

References Bennett PG (1977). Towards a theory of hypergames. Omega 5: 749–751. Bennett PG (1980). Hypergames: developing a model of conflict. Futures 12: 489–507. Bennett PG (1995). Modelling decisions in international relations: game theory and beyond. Mershon Rev Int Stud 39: 19–52. Bennett PG (1998). Confrontation analysis as a diagnostic tool. Eur J Opl Res 109: 465–482. Bennett PG (2004). Confrontation analysis: prediction, interpretation or diagnosis? In: Analysing Conflict and its Resolution. Proceedings of a Conference of the Institute of Mathematics and its Applications. IMA, Southend-on-Sea, pp 7–17. Bennett P and Cropper SA (1987). Maps, games and things inbetween. Eur J Opl Res 32: 33–46. Bennett P, Cropper S and Huxham C (1989). Modelling interactive decisions: the hypergame focus. In: Rosenhead J (ed). Rational Analysis for a Problematic World. Wiley, Chichester, pp 283–314. Bennett PG and Dando MR (1979). Complex strategic analysis: a hypergame study of the fall of France. J Opl Res Soc 30: 23–32. Bennett PG and Howard N (1996). Rationality, emotion and preference change: drama-theoretic models of choice. Eur J Opl Res 92: 603–614. Bennett P, Howard N and Bryant J (2001). Drama theory and confrontation analysis. In: Rosenhead J and Mingers J (eds). Rational Analysis for a Problematic World Revisited. Wiley, Chichester, pp 225–248. Bennett P and Tait A (1993). Interact: developing software for interactive decisions. 34th Annual Convention, International Studies Association. Acapulco, Mexico. Bryant (1988). Frameworks of inquiry: OR practice across the hardsoft divide. J Opl Res Soc 39: 423–435. Bryant J (1989). Problem Management: A Guide for Producers and Players. Wiley: Chichester. Bryant J (1997). The plot thickens: understanding interaction through the metaphor of drama. Omega 25: 255–266. Bryant J (1998). SCRIPT: a dramatic interface for case briefings. In: Geurts J, Joldersma C and Roelofs E (eds). Gaming/Simulation for Policy Development and Organizational Change. Tilburg University Press, Tilburg, pp 243–250. Bryant J (2002a). A comment on Yolles (2001): viable boundary critique. J Opl Res Soc 53: 1390–1392. Bryant J (2002b). Confrontations in health service management: insights from drama theory. Eur J Opl Res 142: 610–624. Bryant J (2002c). It’s all threats and promises: understanding the pressures for effective collaboration. In: Stewart M and Purdue D (eds). Understanding Collaboration: International Perspectives on Theory, Method and Practice. University of the West of England, Bristol, pp 11–17. Bryant J (2003). The Six Dilemmas of Collaboration: Interorganisational Relationships as Drama. Wiley: Chichester. Bryant J and Chin CK (2000). Integrating approaches to revitalise a church’s mission strategy. J Opl Res Soc 51: 689–699.

Bryant JW and Darwin J (2003). Immersive drama: testing health systems. Omega 31: 127–136. Bryant JW and Darwin J (2004). Exploring inter-organisational relationships in the health service: an immersive drama approach. Eur J Opl Res 152: 655–666. Bryson J, Ackermann F, Eden C and Finn CB (2004). Visible Thinking: Unlocking Causal Mapping for Practical Business Results. Wiley: Chichester. Buchanan D and Badham R (1999). Power, Politics and Organizational Change: Winning the Turf Game. Sage: London. Dixit A and Nalebuff B (1991). Thinking Strategically: The Competitive Edge in Business, Politics and Everyday Life. Norton: New York. Eden C (1989). Using cognitive mapping for strategic options development and analysis. In: Rosenhead J (ed). Rational Analysis for a Problematic World. Wiley, Chichester, pp 21–42. Eden C (1990). The unfolding nature of group decision support—two dimensions of skill. In: Eden C and Radford J (eds). Tackling Strategic Problems: The Role of Group Decision Support. Sage, London, pp 48–52. Eden C and Ackermann F (1998). Making Strategy: The Journey of Strategic Management. Sage: London. Eden C and Radford J (eds) (1990). Tackling Strategic Problems: The Role of Group Decision Support. Sage, London. Fisher R and Ury W (1982). Getting to Yes: Negotiating Agreement Without Giving In. Hutchinson: London. Fraser N and Hipel KW (1984). Conflict Analysis: Models and Resolutions. North-Holland: New York. Green KC (2002). Forecasting decisions in conflict situations: a comparison of game theory, role-playing, and unaided judgement. Int J Forecasting 18: 321–344. Howard N (1971). Paradoxes of Rationality. MIT Press: Cambridge, MA. Howard N (1989). The manager as politician and general: the metagame approach to analysing cooperation and conflict, and The CONAN play. In: Rosenhead J (ed). Rational Analysis for a Problematic World. Wiley, Chichester, pp 239–282. Howard N (1993). The role of emotions in multi-organizational decision-making. J Opl Res Soc 44: 613–623. Howard, N (1994). Drama theory and its relation to game theory. Group Decision Negotiation 3: 187–206, 207–253. Howard N (1996). Negotiation as drama: how ‘games’ become dramatic. Int Negotiation 1: 125–152. Howard N (1997). Short cuts. Cooperation—or Conflict 11(5): 4. Howard N (1998). n-person ‘soft’ games. J Opl Res Soc 49: 144–150. Howard N (1999). Confrontation Analysis: How to Win Operations Other than War. CCRP, Department of Defense: Washington, DC. Howard N (2004). Resolving Conflicts in a Tree: Drama Theory in the Extensive Form. In: Bryant JW (ed). Analysing Conflict and its Resolution. Proceedings of a Conference of the Institute of Mathematics and its Applications. IMA, Southend-on-Sea, p 173. Howard N, Bennett PG, Bryant JW and Bradley M (1992/1993). Manifesto for a theory of drama and irrational choice. J Opl Res Soc 44: 99–103; Systems Pract 6: 429–434. Idea Sciences (2005). Confrontation Manager User Manual. Idea Sciences: Washington. Mangham II (1978). Interactions and Interventions in Organizations. Wiley: Chichester. Mangham II and Overington MA (1987). Organizations as Theatre: A Social Psychology of Dramatic Appearances. Wiley: Chichester. Marinetti FT (1909). Foundation and Manifesto of Futurism. Le Figaro, 20 February 1909. Translation in Apollonio U (1973) Futurist Manifestos. Thames and Hudson: London. Mingers J and Rosenhead J (2001). Diverse unity: looking inward and outward. In: Rosenhead J and Mingers J (eds). Rational Analysis for a Problematic World Revisited. Wiley, Chichester, pp 337–355.

J Bryant—Drama theory: dispelling the myths

Mingers J and Rosenhead J (2004). Problem structuring methods in action. Eur J Opl Res 152: 530–554. Mintzberg H, Ahlstrand B and Lampel J (1998). Strategy Safari: A Guided Tour Through the Wilds of Strategic Management. PrenticeHall: London. Morgan G (1980). Paradigms, metaphors and puzzle-solving in organization theory. Admin Sci Quart 25: 605–622. Obeidi A and Hipel KW (2005). Strategic and dilemma analyses of a water export conflict. INFOR 43: 247–270. Rosenhead J (ed) (1989). Rational Analysis for a Problematic World: Problem Structuring Methods for Complexity, Uncertainty and Conflict. Wiley: Chichester. Rosenhead J (2004). The past, present and future of problem structuring methods. J Opl Res Soc (this issue).

613

Smith R, Howard N and Tait A (2001). Confrontations in war and peace. Proceedings of the Command and Control Research and Technology Symposium. CCRP Publications: Washington, DC. Sternberg P and Garcia A (1989). Sociodrama: Who’s in Your Shoes? Praeger: New York. Stubbs L, Howard N and Tait A (2000). How to Model a Confrontation—Computer Support for Drama Theory. Dramatec: Brighton. Von Neumann J and Morgenstern O (1944). Theory of Games and Economic Behaviour. Princeton University Press: Princeton, NJ.

Received November 2004; accepted April 2006 after one revision