How the arrow of feedback links the theories of organisational change: a multi-method use of action research more

300 Int. J. Learning and Change, Vol. 4, No. 4, 2010 How the arrow of feedback links the theories of organisational change: a multi-method use of action research M.R. Kumar* 71 Diesel Locomotive Works, Indian Railways, Banaras, 221004, India E-mail: madhu_ranjan@yahoo.com *Corresponding author P. Ranjan Department of Electronics and Communication Engineering, Indian Institute of Technology, Guwahati, 781039, India E-mail: i_palash@yahoo.com Abstract: This paper shows the implementation of ‘5S’ – a Japanese concept of housekeeping – through action research methodology. The organisational issue it tackles is the cultural inhibition among the Indian population against cleaning. It uses soft systems methodology (SSM), action science and Schein’s idea of clinical enquiry to bring about an enduring change against this inhibition. On a theoretical level, this paper develops a model which integrates these three approaches to organisational change. It argues that it is possible to integrate these approaches based on the nature of feedback loop which can be termed as hard, soft, emanicipatory, and therapeutic churning. Keywords: 5S; soft systems methodology; SSM; action science; clinical enquiry; Indian cultural values; systems theory; Indian railways; therapeutic churning; cleaning; learning; action research; organisational change; feedback; multi-method research; bureaucracy. Reference to this paper should be made as follows: Kumar, M.R. and Ranjan, P. (2010) ‘How the arrow of feedback links the theories of organisational change: a multi-method use of action research’, Int. J. Learning and Change, Vol. 4, No. 4, pp.300–316. Biographical notes: M.R. Kumar is a Mechanical Engineer and Chief Materials Manager in the Indian Railways. He has been a Consultant to the World Bank, as well as many large- and medium-scale enterprises in India in the areas of TQM, organisational learning and organisational transformation. During the last four years, he has published two book chapters, eight papers in international refereed journals and has presented seven papers in international conferences. P. Ranjan is a NTSE scholar and an undergraduate student at IIT Guwahati, India. Copyright © 2010 Inderscience Enterprises Ltd. How the arrow of feedback links the theories of organisational change 301 1 Introduction This research aims to show how three approaches of organisational change – soft systems methodology (SSM) (Checkland, 1999), action science (Argyris et al., 1985) and clinical enquiry (Schein, 1990a) can be integrated into a single change intervention. This research develops a model which can be used to integrate these three approaches. In the existing literature, these approaches have generally been used in stand alone fashion (Edmondson, 1996). This research uses the implementation of 5S – a Japanese concept of housekeeping – as the change intervention to develop the model. This paper is structured as follows. It brings out the salient features of the three approaches of organisational change and the gaps which each approach leaves unattended. Since implementation of 5S is the change initiative which has been used to develop the model, this paper then provides a brief outline of 5S. This is followed by explaining the research methodology and the process of change. In analysing the process of change, this paper shows how each of the three approaches of organisational change caters only to certain aspects of organisational change and how by using all the three approaches, the change can be made enduring. This finally gives rise to the model which integrates the three approaches. 2 Literature review A substantial literature on organisational change is influenced by systems theory (McElyea, 2003). “The early understanding of systems theory conceived organisations as hard systems which really exist and which can be defined, designed and engineered objectively” [Gold, (2001), p.558]. But this was changed by Checkland’s SSM. 2.1 Soft systems methodology SSM says that instead of trying to objectively observe the characteristics of an organisation, one should attempt to understand different ways of seeing the situation within the organisation. Therefore, SSM does not consider organisation as goal seeking rational entities which aims to engineer, optimise and solve its problem by technical rationalities. Hence, SSM aims to understand a situation from different points of views. The different views, thus, generated are contributions to a debate about possible changes. The debate indicates what activities are necessary to achieve a purpose meaningful from a particular point of view. “From such different views of organization, if sufficient meaning is derived, it becomes the basis for action (or non-action)” [Gold, (2001), p.559]. Checkland started with Mode 1 approach to SSM. Mode 1 of the soft systems methodology involved a seven-stage learning system using the mnemonics CATWOE (customers, actors, transformational process, world-view, owners and environmental constraints). It recognised that there are many ways to describe a situation depending on the point of view one takes. It suggested that the view which is most insightful or most relevant in exploring the situation should be selected. Often non-written ways of expressions like story telling and rich pictures (pictorial and cartoon like pictures) aid in grasping the contentious aspect of a situation. From these, root definitions are developed which grasp the essence of the relevant system. Each root definition reflects a particular 302 M.R. Kumar and P. Ranjan worldview of the problem situation. A different way of conceiving the problem situation will give rise to a different root definition. Thus, “a prison can be taken as a punishment system, a rehabilitation system, a system for taking revenge, a system to protect society or a system that constitutes a university of crime” [Jackson, (2000), p.S5]. “Comparison between different worldviews and a perceived real situation gives rise to a set of action – action which is culturally feasible for a particular group of people in a particular situation with its own particular history” [Checkland, (1999), p.A55]. However merely trying to understand different points of views may not be sufficient to bring about effective organisational change. “The assumptions underlying the beliefs and emotions may go unexamined in SMM and therefore may hide key differences in participants’ worldviews” [Porter and Córdoba, (2009), p.333]. There could be personal and community defences which could inhibit the very process of generating realistic points of views (Argyris, 1990), therefore, negating the very foundation of SSM. Argyris et al. (1985) came up with the theory of action science to uncover such defences. 2.2 Action science – a review Argyris and Schön (1989, p.613) consider ‘action science a form of action research’. As per action science people have two theories of action – espoused theory and theories-in-use. Espoused theory of action is what a person claims he or she is following. Theories-in-use are the theories one actually follows which can be inferred from one’s action. While espoused theory may vary widely, according to Argyris et al. (1985, p.88) ‘there is almost no variance in theory-in-use’. By unconsciously but faithfully following their theories-in-use, individuals act masterfully in ways which create error and misunderstanding which Argyris calls ‘skilled incompetence’ (Argyris, 1986). Further, “human beings live in a field of many governing variables. Generally, there is a trade off among governing variables, because actions that raise the value of one may lower the value of another” [Argyris et al., (1985), pp.84–85]. Actions lead to consequences. Consequences may be intended or unintended. When the consequences are unintended, there is a mismatch between the intention and the outcome. In response a person tries to search for another action strategy that will satisfy the same governing variables. This situation where new action strategies are used to satisfy the same governing variables has been defined as single-loop learning by Argyris et al. (1985). However, if the mismatch is used to change the governing variables themselves, then it is a double loop learning system. The master programmes which inhibit double-loop learning is called Model I theory-in-use. Model 1 theories-in-use aim at unilateral control and protection in which people believe that their opinion are correct and try to impose them on others. However there can be another set of theory-in-use which contributes to mutual learning and joint control. They are called model II theory-in-use. Model II theory-in-use facilitates double loop learning. Action science aims at converting the espoused theory into theory-in-use by bringing out the defensive routines and hidden assumptions in our Model I based behaviour. In actions consistent with Model II theory-in-use, there should be less need for camouflage and deception. A criticism of action science is that it does not consider organisation specific patterns which give rise to these camouflage and deception (Edmondson, 1996). Putnam (1999, p.179) has said that “individual theories-in-use are……interdependent with the culture in How the arrow of feedback links the theories of organisational change 303 which people live”. One way to understand and change an organisational culture is through Schein’s clinical enquiry. 2.3 Schein’s clinical enquiry for changing organisational culture Schein (1990b) understands culture as consisting of three levels – observable artefacts, values and underlying basic assumptions. Artefacts are the outward manifest behaviour like the manner people address each other, the dress code etc which one comes across as one enters an organisation. Values are the norms, ideologies and charters of an organisation. By interviewing organisational members, one can understand what the organisational values are – both which are being espoused and which are being observed. The inconsistency between what is being espoused and what is being observed forms the basis for the next level of investigation about culture – the underlying basic assumptions which give rise to the values. When values get deeply engrained, they become underlying basic assumptions. According to Schein (1990a, p.14) “shared tacit assumptions are the basic units of culture and they powerfully influence behaviour in organisation”. Organisational change essentially involves enhancing certain favourable basic assumptions and ‘unlearning’ certain other assumptions that are considered dysfunctional. For Schein, organisational change is like organisational therapy wherein the job of consultants is to make employees handle stress and anxieties which this learning and unlearning of shared assumptions involve (Quick and Garvin, 2000). “This change in basic assumptions is facilitated by creation of new norms in response to some critical incidents and by leaders acting as role models for their subordinates in shaping new behaviour” [Schein, (1990b), p.115]. In order that the organisational therapy is useful, the consultants and the client organisation should jointly own the intervention that is being planned. Therefore, Schein considers an attempt to change an organisation a clinical process. To the extent the members of the client organisation is involved in the clinical process, Schein considers it akin to action research. He calls it the clinical model of action research (Schein, 1995). In spite of the different initial thrust in these three approaches to organisational change, there are areas of similarities. In his later paper, Checkland considered the methodology driven Mode 1 too restrictive for practicing managers who are more often driven by the pressures of their environment (Checkland, 1994). Thus, he developed situation driven Mode 2 of soft systems methodology. Mode 2 SSM is driven by situational logic and situational culture. The cultural stream in SSM has three types of enquiry [Jackson, (2000), p.S7]: Analysis 1 Analysis 2 Analysis 3 It considers the intervention itself and the roles of client, problem solver and problem owner. It is called the social system analysis. It looks at social roles, norms of behaviour and what values are used in judging the performance. This examines the politics of the problem situation and how power is obtained and used. The social system analysis of Checkland with its emphasis on social roles and norms of behaviour is conceptually similar to Schein’s emphasis on organisational culture. Like Mode 2 of SSM, “a clinical researcher gets the data in ‘real time’ generated in the act of 304 M.R. Kumar and P. Ranjan managing change” [Coghlan and McDonagh, (2001), p.203]. On the other hand, “Schein is very clear that clinical researchers need to be self-aware and self-reflective, questioning their own assumptions, biases, and filters” [Coghlan, (2009), p.113]. “Listening to the other is secondary to listening to the self” [Quick and Garvin, (2000), p.32; quoted in Coghlan, (2009), p.114]. This is quite like action science. Therefore, Friedman and Antal (2005, p.78) say that “action science provide a framework for making tacit, culturally generated repertoire explicit”. From the above comparison, this paper concludes that in their mature stages, all the three approaches recognise the importance of culture in advancing their specific approach to organisational change. But, in spite of the importance of culture in their mature models, no unification has yet been attempted which can integrate the three approaches into one. It is this gap that this paper aims to bridge. This paper aims to demonstrate this in the specific change situation of implementation of ‘5S’. 2.4 5S – an over view 5S has generally been considered as the first step in an organisation’s journey towards total quality management (TQM) (Ho, 1999). According to Hirano (1996), housekeeping as per 5S begins with removing all the items from the workplace which are not needed for the current operation. Then it calls for arranging the needed items so that they are easy to use and label them so that anyone can find them. Then it emphasises cleaning so that the work area is shining. Cleaning is also taken as a means of inspection because a shining area brings out any shortcoming in equipments and work conditions. After implementation of above three, they should be maintained by framing a formal work procedure. Finally, 5S calls for self-discipline on the part of the workers to internalise it so as to make it a habit. Literature on 5S shows that it can bring about improvement in housekeeping both in manufacturing environment (O’hEocha, 2000) and in non-manufacturing environment (Warwood and Knowles, 2004). However, in Indian context barring a few exceptions (for example Kumar et al., 2009), there is a lack of reported work about implementation of 5S. 2.5 Indian culture and change Since this paper has concluded that culture could provide the underlying continuity in utilising the three approaches, it throws up a question – how the Indian culture affects organisational behaviour in India? In terms of the framework developed by Hofstede (1980), low individualism, high power distance, low masculinity, high associative thinking, and moderate level of uncertainty avoidance have been identified as the cultural characteristics of Indians (Chhokar, 2000; Kanungo and Mendonca, 1996). Sinha et al. (2004) have indicated that Hofstede dimension of power distance has remained a dominant theme among Indians. While Kumar (2007) confirms that notwithstanding more than 15 years of liberalisation, Indian bureaucracy remains hierarchical, a more recent review by Kumar et al. (in press) indicates that Indian bureaucracy, besides being hierarchical is also relationship oriented and context sensitive due to corresponding Indian social values. Kumar and Sankaran (2007) have argued that unlike the literature in the West, it is possible to exploit power distance and relationship orientation of Indians How the arrow of feedback links the theories of organisational change 305 for effective quality oriented organisational change. Similarly, there are evidences that Indians are sensitive to situational cues and modify their behaviour accordingly (Ronald, 1988; Kumar and Sankaran, 2006). Further, Indians are quite sensitive to how others regard them and seem quite open to be guided and directed by an elder or superior in their hierarchical relationships (Ronald, 1988). According to Derne (1992, p.284), “North Indian Hindu men find their true selves neither in following their impulses nor in pursuing institutional goals that they have internalised, but instead in being guided by social pressures”. With this brief review of the impact Indian social values on the Indian work culture, this paper now discusses the research methodology. 3 Research methodology The organisation where the study was done is a large factory in the Indian public sector engaged in the manufacturing of locomotive located at Banaras in north India. The research was done between 2005 and 2007. A walk around the factory would show that there were lots of waste materials strewn around. The organisational problem for the improvement in work practice was thus defined as how to have an organised system for disposal of waste and cleanliness in the factory. Therefore, the research method was to be such that it not only brought about a change in organisational practice with respect to cleaning, but it could also contribute to literature by showing how the three hitherto different approaches of organisational change could be combined. A review of qualitative research methodologies showed that action research aims at solving organisational problems and also adds to scientific knowledge (Shani and Pasmore, 1985). It aims to resolve social or organisational issues together with those who experience these issues (Coghlan, 2004). Further, SSM, action science and clinical enquiry have been considered as belonging to the methodological school of action research (Argyris and Schön, 1989; Checkland, 1999; Schein, 1995). Hence action research was selected as the methodology to develop a model which could link the three approaches to change within a single change process. The change process was initiated in the logistics division (called depot) of the factory. The depot spread over an area of one square kilometre was divided into 42 sub-units. A depot material superintendent (DMS) headed each sub unit. A team of action researchers was put together. It consisted of the chief manager, two managers (Manager-1 and Manager-2) and one assistant manager. The action research consisted of six cycles in two stages. The first stage was based on soft systems methodology and the second stage was based on action science. The first author was one of the action researchers. Conduct of action research is generally cyclical. These cycles have been regarded as learning cycles (O’Brien, 1998). Each cycle has a reflection phase. In the reflection phase, the experience of undergoing a cycle is analysed to understand what happened. The learning gleaned from the reflection can then guide the next cycle of action. A progressive summation of these reflections may develop a theory which could be a contribution to literature. In line with the cyclic approach, the research used the iterative cycles of soft systems methodology for modelling the real world in the context of implementation of 5S. For 306 M.R. Kumar and P. Ranjan each of the cycles, each of the action researchers interacted with his DMSs to solicit their points of view. These different points of views were then discussed in small group meetings by the DMSs much in line with the approach suggested by Checkland to arrive at the possible action points (Checkland, 1999). The conclusion drawn and the action point undertaken were then reflected upon by the action researchers with their DMSs. For each cycle, the common understanding arrived in the reflection phase decided how to plan the next cycle. 4 The process of change 4.1 Action research cycles based on SSM At the end of cycle 1 in SSM, the essence of the existing situation was understood as: the method of disposal of the waste generated was ad-hoc. In cycle II, ‘5S’ was developed as the framework which could be used to develop an ideal system for disposal of waste. In cycle III, the ideal system was compared with the existing system so that a revised work process could be developed. Based on this, training material on 5S was developed. The concept of ‘5S’ was institutionalised by making it a part of work instruction under the ISO 9000 certification. In cycle IV, changes based on the above actions were implemented. The implementation of changes resulted in development of a systematic method for collection and disposal of waste. The removal of clutter from different areas of depot added to the feeling of ‘déjà vu’ among the staff. This system continued satisfactorily for three to four months. However, thereafter problems cropped up with respect to the element of ‘cleaning’ of the ‘5S’ concept. Audit reports indicated that the staffs were not doing the cleaning part regularly. There were reports of packing waste lying in the depot for days. If not checked, the system ran the risk of getting reverted to the old state. Small group meetings with the DMSs brought out that there were individual and social defences at work which led to the deterioration in cleanliness. Therefore, the action researchers learnt from the SSM based cycles that even a theoretically valid and participatively designed improvement system might not endure if there were ‘hidden defences’ in the minds of the participants. 4.2 The action science cycles This situation called for understanding the existing situation using the framework of action science so as to bring out the defences and assumptions of the employees. This was labelled action science cycle I. 4.2.1 Action science cycle I This cycle is schematically shown in Figure 1. Figure 1 shows that following the cultural mind set which considered cleaning a ‘menial activity’, in depot the cleaning job was delegated to the lower employees (the workers) “who were any way doing it earlier” – to quote one of the supervisors. This indicated that in spite of their professed importance to ‘5S’, the depot employees, in reality, did not consider cleaning their work place important enough when it clashed with How the arrow of feedback links the theories of organisational change 307 their own personal opinion about cleaning. To paraphrase Argyris, there was a difference between their ‘espoused theory’ and ‘theory-in-use’. The espoused theory was to keep each work area shining. The theory-in-use was that the staff considered it infra-dig to do the cleaning. This resulted in their casual approach which ultimately affected the overall cleanliness of the depot. Figure 1 Theory-in-use for the managers, supervisors and the workers Intention Implementation of 5S in the division Governing value of the managers (espoused) The staff will be interested in achieving what has been participatively discussed and decided in consultation with them Governing value of the supervisor (espoused) I will follow the instructions of 5S Governing value of the workers (espoused) I will follow the instructions of 5S Governing value (in use) of the managers • Define goals and try to achieve them Governing value (in use) for the supervisors • • • Cleaning is menial Be rational – After all the workers have been doing the cleaning in past also. This 5S is yet another management fad. Governing value (in use) of the workers: • • Cleaning is menial Minimise generating or expressing negative feelings Action strategies of the managers Design and manage the environment unilaterally: 5S defined and the staff persuaded through training to follow it. Action strategies of the workers: Action strategy of the supervisors Thus ask the workers to do it. Work in the same manner as was customary in past Consequences (outcome) • • • Withholding of feelings – no one admitted that internally, they consider it infra-dig to clean their work place by their own hand. Defensive norms (emphasis on diplomacy) – continue to pay lip service to 5S Little internal commitment – work in the same manner as in the past. Thus cleaning continued to be a secondary job in the scheme of the managers, supervisors and the workers. The cleaning standard reverted back to the earlier standard Comparison Intention Implementation of 5S 308 M.R. Kumar and P. Ranjan It now became clear to the AR team that any effort for effective implementation of cleanliness was required to address and tackle this mind-set against cleaning. The AR team also realised that all the effort till now was basically aimed at first order change which, therefore, did not change the governing values of the staff. The strategies till now contributed only to the existing norms of interaction – it can be called a behavioural approach to housekeeping. It made no attempt to discover, much less change, the underlying values which governed the behaviour of the depot managers and staff. It, therefore, inhibited learning. It was akin to what Argyris et al. (1985, p.281) have called a ‘culture of protectionism’ – strategies that contribute to norms of interaction that inhibit learning. At this stage, to understand how a few self-critical reflections brought out and finally broke down the defences, it may be desirable to give direct quotes: One of the action researchers (Manager-1) asked a question. Manager-1: “Are we ourselves following the 5S principles?” Chief manager: “I think we are.” Manager-1: “One of the 5S principles is ‘self-discipline’. It means we should make 5S a part of our habit. But our – the manager’s – habits have not really changed. While we expect our supervisors to clean their workplaces themselves, we do not clean our workplaces ourselves. We are still dependent on the office boy to do this. Thus, what is wrong if the supervisors expect the workers to clean their workplaces?” After some murmuring, Manager-1continues: “In reality, we never felt the urge to practice what we preached about cleaning and 5S. This got reflected in the elaborate sham of developing the Indian version of 5S and then handing down a ‘work instruction’ which every one was supposed to comply with and followed it up with training on 5S. We even went through the elaborate ritual of participatively designing it with the DMSs. But what we never talked about was whether people are actually willing to maintain their work area in tip-top condition. We avoided talking about it. The main issue is not 5S; the main issue is whether we are willing to soil our hands with dirt. Because we do not want to soil our hands, we avoided talking about it.” This analysis had a cathartic effect on the action research team. We felt that through this frank reflection, we were able to come face to face with our own biases related to cleaning. This issue brought the team members face to face with their own governing values: We did not want to do the cleaning ourselves because we too considered it to be menial and below our dignity. We also realised that developing the work instruction participatively was nothing but ‘opposite Model I behaviour’ – the mirror image of Model 1 [Argyris et al., (1985), p.92]. Having discovered how we were ourselves hindering the effective implementation of ‘5S’, we reasoned that the first step in ensuring the successful implementation of ‘5S’ was by demonstrating that we – the higher echelons of the division – have changed our governing values. We must demonstrate that the theory-in-use for us was that we did not consider cleaning by ourselves socially infra-dig. It now became clear to the AR team that any effort for effective implementation of cleanliness was required to address and tackle this mind-set against cleaning. Heifetz (1994, p.245) has said, “to move at the pace of logic alone, people would need an unusually high level of rationality and intellectual freedom from habit, tradition and pride”. Perhaps the strategies till now made no attempt to free employees from their How the arrow of feedback links the theories of organisational change 309 ‘habit, tradition and pride’. In the language of action science, it made no attempt to discover, much less change, the underlying values which governed the behaviour of the depot managers and staff. It, therefore, inhibited learning. Thus, the target now was to change the governing values of the depot staff which in turn affected their action strategies. In the language of action science, double loop learning was required. The existing governing value ‘cleaning is menial’ was required to be replaced with the governing value called ‘dignity of labour’. Now the issue was – how to design an intervention which could successfully address the issue of attitudinal change among the staff on this account. Reflecting over it, it gradually became clear that successful implementation of 5S had to go beyond the conventional action research methodologies of SSM or action science. For the purpose of breaking the age-old bias of depot staff and indeed of the depot managers against cleaning, some thing which could synthesise their internal values with the desired work behaviour was required. It was recalled from literature review that one of the criticisms of action science was that it lacked a systematic way to help the new interventionist discern organisation specific patterns, and instead “assumes instead a level of skilfulness that enables facile diagnosis without a template” [Edmondson, (1996), p.591]. As a solution Edmondson (1996) says that Schein offers an approach for assessing organisation specific cultural patterns. Accordingly, the following action strategy was designed: 4.2.2 Action science cycle II Borrowing from Schein’s idea of leaders acting as role models, it was decided that to set an example, the managers would dispense with the office boys who cleaned their rooms and send them to different warehouses with the instruction that they would help the ward DMSs. Further, every day in the morning, two action researchers would visit a sub unit with the respective DMS. There, they would actually spend five minutes working as per the work instruction made for implementation of 5S i.e., they would actually keep the items and files on their earmarked location, they would ensure that each file was numbered as per the plan; if not done, they would themselves do the numbering. They would then actually clean each bin and each pallet. Thereafter, they would move on to the adjoining sub unit and repeat the same. In one week they could complete visiting all the warehouses and offices once. The hierarchy-oriented staff could now see that the managers of their division were doing what was professed to be done by all of them. This changed the shared tacit assumptions of the organisation. An emotionally neutral and ‘cleaning –inhibiting’ work climate was now made emotionally charged. “If the boss could do it why shouldn’t I” – became the refrain. The shared basic assumption ‘power distance’ acting through the governing value of ‘deference for the superior’ propelled them to follow what their boss was doing. This reduced the existing importance of hierarchy which in turn reduced the salience of the value that ‘cleaning is menial’. The value ‘deference for superior’ necessitated following the dictums of 5S in spirit. Now dignity of labour became the most fundamental governing variable in which nested other governing variables such as deference for superior and social pressure (in that order) (see Figure 2). 310 Figure 2 M.R. Kumar and P. Ranjan Nesting of earlier governing values in more fundamental governing values, resulting action strategies and consequences of the action strategies as per action science cycle II Nesting of revised governing values (in use) Dignity of labour Deference for superior Social pressure Earlier governing values (in use) of the workers: Minimise generating or expressing negative feelings Earlier governing value (in use) of the managers Define goals and try to achieve them Earlier governing values (in use) for the supervisors Be rational. Revised action strategy My superior who is above me is showing a real change in his orientation towards cleaning. Since he is above me, my dharma tells me that I must also demonstrate equal commitment to cleaning Earlier Action strategy Ask (or expect) the worker to do it Intention Implementation of 5S Comparison New consequences • • No defensive interpersonal relation New behaviour learnt Note: Dotted boxes show earlier governing values and action strategies. Further, by nesting ‘deference for superior’ into ‘dignity of labour’, ‘deference for superior’ was now profitably used to increase the salience of ‘social pressure’ which was earlier dormant. It is not that the now active governing variables were absent earlier. They were present, but the earlier behaviour dynamics of the group did not affect their magnitude. For example the ‘deference for superior’ variable was there, but by not working in line with the tenets of 5S, the managers actually did nothing which could disturb the magnitude of this variable. It was only when the ‘dignity of labour’ was made a dominant governing variable, that an imbalance was created in the relative magnitude of ‘deference for superior’ and ‘social pressure’. How this change in the relative magnitude of different ‘governing variables’ (in Argyris’ terminology) or ‘values’ (in Schein’s terminology) triggered the revised action strategy is shown in Figure 2. It is to be noted that in Figure 2, the feedback loop was not used to change the action strategy. It How the arrow of feedback links the theories of organisational change 311 was instead used to change the relative salience of different governing variables. In the language of action science, due to the trade off among different governing variables, the master programmes the depot employees now used to produce action got changed. Therefore, it resulted in a new behaviour. 5 Discussion Action science considers ineffectiveness an outcome of action strategies not organisational culture [Edmondson, (1996), p.575]. But in this research, the action strategy designed in action science cycle II shows that understanding culture can help identify governing variables which inhibit learning. Figure 2 shows that all the governing variables were drawn from the general field of culture. In fact Putnam (1999, p.179) has said that ‘we learn our theories-in-use through socialisation’. This linkage between the approaches of Schein and Argyris brings into discussion the issue of synthesising different approaches to organisational change. Therefore, this paper now attempts to answer the research question about how to integrate the three approaches of organisational change discussed in this paper. This paper argues that picking up from Taiwo (2001), it is possible to link the three approaches on a continuum of ‘nature of feedback loop’. The proposed continuum categorises the feedback loop as ‘hard’, ‘soft’, ‘emancipatory’ and ‘therapeutic churning’ (see Table 1 and Figure 3). The ‘hard’ feedback loop corresponds to the hard system thinking wherein an ideal system called 5S was designed in Indian context. The ‘soft’ feedback loop based on SSM, helped to understand different ‘world views’ as publicly enunciated by the employees. To the extent, it was able to identify different versions of organisational reality; it was able to bring out more personal designs behind employees’ behaviour. Table 1 Linkage between nature of feedback loop and theoretical approaches to change Theoretical grounding Systems theory SSM Learning based on Modelling of the real world Unearthing of hidden agendas depending on one’s perception of reality Understanding and drawing out the defensive routines Varying the relative salience of the ‘shared tacit assumptions’ Consequences for the organisation in this research Development of Indian version of ‘5S’ Resistance to cleaning Nature of feedback loop Hard Soft Emancipatory Action science Learning individual and group biases against cleaning Getting rid of bias against cleaning which had a purgative impact on the group as a whole. This lead to development of Model II theory-in-use Therapeutic churning Clinical enquiry (about how our defensive routines are derived from ‘shared tacit assumptions’) 312 Figure 3 M.R. Kumar and P. Ranjan Integration between SSM, action science and clinical enquiry based on the nature of feedback loop Intention (cleaning the work place) (mis)match Shared basic assumptions Norms and values Appreciation of system Action strategies Consequences (outcome) As the feedback loop became more subtle in the action science based cycles, it began to draw out the defensive routines of the employees. The baring of the defensive routines could be considered as bringing forth deeper designs behind employees’ behaviour. To the extent, the employees were able to understand their own defences as impediments to their proposed new behaviour; this feedback loop could be termed as emancipatory. However, a mere feeling of emancipation does not guarantee initiation of new behaviour. In order to bring out enduring changes, the feedback loop at this stage should begin to hit a number of ineffective cultural assumptions of the employees. As many cultural assumptions of the employees begin to get questioned, many behavioural artefacts in the organisation would begin to lose their legitimacy. This will lead to new behaviour. However, in this feedback loop, there is no confrontative eliciting of one’s tacit assumptions as in action science. The feedback loop at this stage is more behavioural and less dialogical – the managers in this research decided to hit the ineffective cultural assumptions of the group by working in each sub unit rather than by entering into dialogical reflections with the employees. Herein lays the importance of Schein’s idea of formation of new norms and values around critical incidents and leaders acting as role models – these lead to private but profound realignment of employee’s How the arrow of feedback links the theories of organisational change 313 beliefs and values. Herein also lays the importance of ‘clinical’ aspect of the enquiry – if the organisation desires to turn over a new leaf, its members will have to purge themselves of their existing cultural assumptions. The onus for change is on them. Thus, it can be said that at this stage the feedback loop begins to ‘churn’ the cultural assumptions. To the extent this churning challenges and changes the salience of a number of cultural assumptions, it acts as if it is having a therapeutic effect on the organisation as a whole. Hence this feedback loop has been termed therapeutic churning (see Table 1 and Figure 3) which leads to enduring change. Thus, cleaning, which was hitherto – to borrow a phrase from soft systems methodology – systemically desirable but culturally infeasible (Jackson, 2000) had now become ‘systemically desirable and culturally feasible’ thanks to the therapeutic churning. This return to SSM marked the closing of the feedback loop (see Figure 3). The contribution of this paper is that it shows that it is the nature of feedback loop which can integrate the three approaches to organisational change. Figure 3 shows that as feedback loop moves from being hard and more tangible to more subtle, it can be said to correspond to the three approaches of organisational change shown in Table 1. This is how the three approaches to organisational change can be integrated into a single model. Each of these approaches provides value at different stages of organisational change. While SSM can be useful in modelling a complex world and understanding the different world views as per different players, action science helps in articulation of ‘theories of action’ at different levels of detail. In the process, it brings out the different ‘defensive routines’ which inhibit change. Schein’s emphasis on understanding ‘the shared tacit assumptions’ as the building blocks of culture is useful in understanding which governing variables are active (leading to ‘defensive routines’) and which are dormant. The linkage of governing variables to culture in turn becomes useful in developing a uniform Model II theory-in-use for a group as a whole. The uniformity in the theory-in-use leads to an action strategy which has a natural acceptance to the group members. This decreases the extent of accommodation which is otherwise required in SSM based change. Thus, the final consequence of the action is less of a compromise and more an expression of the natural will of the group members. 6 Conclusions This paper argues that in real time, it is perhaps not possible to deploy a particular approach in a stand alone fashion because a change situation demands action strategy in keeping with the dynamics of the situation and not necessarily in keeping with the heuristic of a particular approach. Therefore, developing a theoretical insight about the interactive use of the three approaches to change without being dogmatic about a particular approach will aid both practicing managers and academicians in the area of organisational change. Accordingly, this paper has shown a way to integrate the three approaches of organisational change under the overall ambit of action research depending on what kind of data emerge as the change process proceeds, what is the nature of feedback loop and how the feedback loop is used. It, therefore, addresses the criticism about there being a general lack of synthesis among different approaches to organisational change (Edmondson, 1996). Also as the feedback loop shifts from being hard to being therapeutic, it tends to become less dialogical and more introspective. This 314 M.R. Kumar and P. Ranjan emphasis on introspection in place of ‘dialogical confrontation’ – as suggested in action science – finds a natural echo in Indian’s tendency of reflecting on a situation internally and then arriving at decisions (Nakamura, 1964) much in line with Schein’s approach to dig to uncover subtle cultural beliefs but not to go public with them until the researcher has private confidence in their utility (Edmondson, 1996). Therefore, this paper also shows a way to blend Schein’s clinical enquiry with Indian’s propensity for internal reflection so as to unearth the theories-in-use. This use of cultural artefacts in place of ‘dialogical confrontation’ to get to the governing variables is a new approach which practitioners of organisation change can usefully deploy in other cultural settings too. Acknowledgements An earlier version of this paper entitled ‘Implementation of 5S through action research: an Indian context’ was presented at the ICPQR at IIT Delhi in 2005. The authors thank the anonymous reviewers who happily reviewed this paper through its many iterations. References Argyris, C. (1986) ‘Skilled incompetence’, Harvard Business Review, Vol. 64, No. 5, pp.74–79. Argyris, C. (1990) ‘Inappropriate defenses against the monitoring of organization development practice’, Journal of Applied Behavioral Science, Vol. 26, No. 3, pp.299–312. Argyris, C. and Schön, D. (1989) ‘Participatory action research and action science compared: a commentary’, The American Behavioral Scientist (1986–1994), Vol. 32, No. 5, pp.612–623. Argyris, C., Putman, R. and Smith, D.M. (1985) Action Science, Concepts, Methods and Skills for Research and Intervention, Jossey-Boss, San Francisco. Checkland, P. (1994) ‘Systems theory and management thinking’, The American Behavioural, Vol. 38, No. 1, pp.75–91. Checkland, P. (1999) Soft Systems Methodology: A 30-year Retrospective, Wiley, Chichester. Chhokar, J.S. (2000) ‘Effective leadership in India: a multi method study’, International Journal of Psychology, Vol. 35, Nos. 3/4. Coghlan, D. (2004) ‘Action research in the academy: why and whither? Reflections on the changing nature of research’, The Irish Journal of Management, Vol. 25, No. 2, pp.1–10. Coghlan, D. (2009) ‘Toward a philosophy of clinical inquiry/research’, Journal of Applied Behavioral Science, Vol. 45, No. 1, pp.106–121. Coghlan, D. and McDonagh, J.C. (2001) ‘Research and practice in IT-related change: the case for clinical inquiry’, in Pasmore, W.A. and Woodman, R. (Eds.): Research in Organizational Change and Development, Vol. 13, pp.195–211, JAI, Amsterdam. Derne, S. (1992) ‘Beyond institutional and impulsive conceptions of self: family structure and socially anchored self’, Ethos, Vol. 20, No. 3, pp.259–288. Edmondson, A.C. (1996) ‘Three faces of Eden: the persistence of competing theories and multiple diagnoses in organisational intervention research’, Human Relations, Vol. 49, No. 5, pp.571–595. Friedman, V.J. and Antal, A.B. (2005) ‘Negotiating reality: a theory of action approach to intercultural competence’, Management Learning, Vol. 36, No. 1, pp.69–86. Gold, J. (2001) ‘Storying systems: managing everyday flux using mode 2 soft systems methodology’, Systemic Practice and Action Research, Vol. 14, No. 5, pp.557–573. Heifetz, J. (1994) Leadership without Easy Answers, Harvard University Press, Cambridge. How the arrow of feedback links the theories of organisational change 315 Hirano, H. (1996) The Sourcebook for 5S Implementation, Productivity Press. Ho, S.K.M. (1999) ‘Japanese 5-S-where TQM begins’, The TQM Magazine, Vol. 11, No. 5, pp.311–321. Hofstede, G. (1980) Culture’s Consequences: International Differences in Work-Related Values, Sage, Thousand Oaks. Jackson, M.C. (2000) ‘Checkland, Peter Bernard (1930)’, Systems Research and Behavioural Science, Vol. 17, No. S1, pp.S3–S10. Kanungo, R.N. and Mendonca, M. (1996) ‘Cultural contingencies and leadership in developing countries’, in Bacharach, Erez and Bamberger (Eds.): Research in Sociology of Organisations, Vol. 14, pp.263–295, JAI Press, Greenwich CT. Kumar, M.R. (2007) ‘Assessment of hierarchical tendencies in an Indian bureaucracy’, International Journal of Public Sector Management, Vol. 20, No. 5, pp.380–391. Kumar, M.R. and Sankaran, S. (2006) ‘From east to west: has the twain met? The linkage between context sensitivity and SSM’, Systemic Practice and Action Research, Vol. 19, No, 4, pp.365–381. Kumar, M.R. and Sankaran, S. (2007) ‘Indian culture and culture for TQM: a comparison’, The TQM Magazine, Vol. 19, No. 2, pp.176–188. Kumar, M.R., Sahay, B.K. and Ranjan, P. (in press) ‘Adapting TQM to change Indian bureaucracy: a view from inside’, Quality Management Journal. Kumar, S.P., Sudhahar, C., Dickson, J.F., Dhas, M.S., Senthil, V., Devadasan, S.R. and Murugesh, R. (2009) ‘Performance analysis of 5S teams using quality circle financial accounting executive support system’, International Journal of Management Practice, Vol. 3, No. 3, pp.241–262. McElyea, B.E. (2003) ‘Organizational change models’, Futurics, Vol. 27, Nos. 1/2, pp.57–64. Nakamura II (1964) Ways of thinking of Eastern People: India, China, Tibet and Japan, East West Center Press, Honululu. O’Brien, R. (1998) ‘An overview of the methodological approach of action research’, in Richardson, R. (Ed.): Theory and Practice of Action, Universidade Federal da Paraiba, Joao Pessoa, Brazil. O’hEocha, M. (2000) ‘A study of the influence of company culture, communications and employee attitudes on the use of 5S for environmental management at Cooke Brothers Ltd’, The TQM Magazine, Vol. 12, No. 5, pp.321–330. Porter, T. and Córdoba, J. (2009) ‘Three views of systems theories and their implications for sustainability education’, Journal of Management Education, Vol. 33, No. 3, pp.323–347 Putnam, R.W. (1999) ‘Transforming social practice: an action science perspective’, Management Learning, Vol. 30, No. 2, pp.177–187. Quick, J.C. and Garvin, J.H. (2000) ‘The next frontier: Edgar Schein on organizational therapy’, The Academy of Management Executive, Vol. 14, No. 1, pp.31–48. Ronald, A. (1988) In Search of Self in India and Japan: Towards a Cross Cultural Psychology, Princeton University Press, Princeton. Schein, E.H. (1990a) ‘Innovative cultures and adaptive organisation’, Journal of Development Administration, Vol. 7, pp.9–39, Sri Lanka. Schein, E.H. (1990b) ‘Organisational culture’, American Psychologists, Vol. 45, No. 2, pp.109–119. Schein, E.H. (1995) ‘Process consultation, action research and clinical inquiry: are they the same?’, Journal of Managerial Psychology, Vol. 10, No. 6, pp.14–19. Shani, A.B. and Pasmore, W.A. (1985) ‘Organization inquiry: towards a new model of the action research process’, in Warrick, D.D. (Ed.): Contemporary Organisation Development Current Thinking & Application, pp.438–448, Scott, Foresman and Company, Illinois. 316 M.R. Kumar and P. Ranjan Sinha, J.B.P., Bhupatkar, A.P., Sukumaran, A., Gupta, P., Gupta, R., Panda, A., Singh, S., Sinha, R.B.N., Singh-Sengupta, S. and Srinivasan, E.S. (2004) ‘Facets of societal and organisational cultures and managers’ work related thoughts and feelings’, Psychology & Developing Society, Vol. 16, No. 1, pp.1–26. Taiwo, J. (2001) ‘Systems approaches to total quality management’, Total Quality Management, Vol. 12, Nos. 7/8, pp.967–973. Warwood, S.J. and Knowles, G. (2004) ‘An investigation into Japanese 5S practice in UK industry’, The TQM Magazine, Vol. 16, No. 5, pp.347–353.