I'm currently wading through some difficult mathematical territory in search of a way of articulating why Stafford Beer's Viable System Model (see http://en.wikipedia.org/wiki/Viable_system_model) can be highly effective in stimulating reflexivity within complex organisations. My thinking is that the VSM contains a powerful combination of a simple set of categories for looking at the function of an organisation, with an equally simple explanation as to how these categories relate to one another in a recursive way. It is the implicit explanatory aspect of the VSM, which draws on the cybernetic theories of Ashby (the law of requisite variety: see http://en.wikipedia.org/wiki/Variety_(cybernetics)), which I think does the trick.
What does the explanatory aspect of the VSM do? In essence, I think its function is to manipulate the probability distribution for the mapping of categories onto aspects of organisational experience. Thus, whilst the categories (which are numbered by Beer as 'system 1', 'system 2', 'system 3', etc) may be almost arbitrarily applied to different aspects of the organisation in the first instance, the explanation that accompanies them means that other categories cannot be applied arbitrarily, or rather their mapping depends on some initial choice. In this way, logically, the choice of mappings is restricted. This has a number of effects.
Firstly it means that the assignation of categories is not completely arbitrary. The assignation of an initial category (x is System 1) changes the probability distribution of the assignation of other categories.
Secondly, and related to this, it means that the scope for agreement between a number of different stakeholders in an organisation is increased, because not only are there a restricted number of mappings, but also the explanation of one mapping or another quickly reveals to other colleagues a deeper meta-understanding of the organisational structure which can be critiqued. In other words, people do not argue about the assignations of categories, but argue about their understanding. This is at a meta level.
It is interesting to compare the VSM to deeper theoretical cybernetic work which also relates to organisation. For example, I am very interested in Luhmann's communication theory. But Luhmann is not practical like Beer, and his work can be challenging. However, deep inspection reveals much that is of value, particularly relating to the necessity for rich ecologies of communication in organisations and the pathologies that can set it if communication is constrained (through fear, power, managerial isolation, etc). At the same time, Luhmann's work can be usefully applied to the analysis of communications data from the internet, and since the internet forms a key component of the environment of all businesses (particularly Universities) applying this analysis ought to carry some practical benefit.
To some extent, I think that this is what the learning analytics agenda is about. But there is a disconnect between the pretty pictures of data analysis and the real everyday political concerns of employees of an organisation. There is even a danger that clever data analysis could lead to a kind of managerial hubris which claims a 'higher intellegence' than the people on the ground doing the actual work! It is Beer's work which is with the people on the ground, yet Beer's work does not lend itself to the kind of deep analysis of communications that Luhmann's work affords.
So what can be learnt from Beer that might help in the practical realisation of the benefits of data analysis? Developing the notion of the role of explanation in Beer, and the consequences it carries for agreement, I am starting to look again at Nigel Howard's work on 'Drama Theory' (see http://en.wikipedia.org/wiki/Drama_theory). This is a game-theoretical approach which seeks to articulate the complex ways in which individuals behave in organisational situations as they act in coalitions with colleagues with ideas about the strategies of those colleagues, or sometimes for their own personal gain. Howard's work can be seen as a mathematical articulation of institutional political behaviour.
As I try to understand the maths, I'm seeing some resonances between the dynamics of understanding each other's strategies (which is an anticipatory dynamic) and the dynamics of looking at immediate gain. Leydesdorff makes a distinction between those dynamics which rely on the previous state, and those dynamics which anticipate future states. The latter, it seems to me, are related to the understanding of strategies, and the extent to which strategies are agreed upon, and strategies depend on the articulation of explanations.
If the goal of an effective organisation is maintaining rich reflexive communications as a means of maintaining profitable interactions with its environment, then the role of explanation and anticipation are fundamental. If the way that explanation works within things like the VSM can be better understood, then we can create a relationship between the data analysis of internet communications and frameworks for articulating explanations of that data. In other words, the technology can serve as an environment for nurturing explanations amongst staff and in the process creating environment for rich cooperation.
What does the explanatory aspect of the VSM do? In essence, I think its function is to manipulate the probability distribution for the mapping of categories onto aspects of organisational experience. Thus, whilst the categories (which are numbered by Beer as 'system 1', 'system 2', 'system 3', etc) may be almost arbitrarily applied to different aspects of the organisation in the first instance, the explanation that accompanies them means that other categories cannot be applied arbitrarily, or rather their mapping depends on some initial choice. In this way, logically, the choice of mappings is restricted. This has a number of effects.
Firstly it means that the assignation of categories is not completely arbitrary. The assignation of an initial category (x is System 1) changes the probability distribution of the assignation of other categories.
Secondly, and related to this, it means that the scope for agreement between a number of different stakeholders in an organisation is increased, because not only are there a restricted number of mappings, but also the explanation of one mapping or another quickly reveals to other colleagues a deeper meta-understanding of the organisational structure which can be critiqued. In other words, people do not argue about the assignations of categories, but argue about their understanding. This is at a meta level.
It is interesting to compare the VSM to deeper theoretical cybernetic work which also relates to organisation. For example, I am very interested in Luhmann's communication theory. But Luhmann is not practical like Beer, and his work can be challenging. However, deep inspection reveals much that is of value, particularly relating to the necessity for rich ecologies of communication in organisations and the pathologies that can set it if communication is constrained (through fear, power, managerial isolation, etc). At the same time, Luhmann's work can be usefully applied to the analysis of communications data from the internet, and since the internet forms a key component of the environment of all businesses (particularly Universities) applying this analysis ought to carry some practical benefit.
To some extent, I think that this is what the learning analytics agenda is about. But there is a disconnect between the pretty pictures of data analysis and the real everyday political concerns of employees of an organisation. There is even a danger that clever data analysis could lead to a kind of managerial hubris which claims a 'higher intellegence' than the people on the ground doing the actual work! It is Beer's work which is with the people on the ground, yet Beer's work does not lend itself to the kind of deep analysis of communications that Luhmann's work affords.
So what can be learnt from Beer that might help in the practical realisation of the benefits of data analysis? Developing the notion of the role of explanation in Beer, and the consequences it carries for agreement, I am starting to look again at Nigel Howard's work on 'Drama Theory' (see http://en.wikipedia.org/wiki/Drama_theory). This is a game-theoretical approach which seeks to articulate the complex ways in which individuals behave in organisational situations as they act in coalitions with colleagues with ideas about the strategies of those colleagues, or sometimes for their own personal gain. Howard's work can be seen as a mathematical articulation of institutional political behaviour.
As I try to understand the maths, I'm seeing some resonances between the dynamics of understanding each other's strategies (which is an anticipatory dynamic) and the dynamics of looking at immediate gain. Leydesdorff makes a distinction between those dynamics which rely on the previous state, and those dynamics which anticipate future states. The latter, it seems to me, are related to the understanding of strategies, and the extent to which strategies are agreed upon, and strategies depend on the articulation of explanations.
If the goal of an effective organisation is maintaining rich reflexive communications as a means of maintaining profitable interactions with its environment, then the role of explanation and anticipation are fundamental. If the way that explanation works within things like the VSM can be better understood, then we can create a relationship between the data analysis of internet communications and frameworks for articulating explanations of that data. In other words, the technology can serve as an environment for nurturing explanations amongst staff and in the process creating environment for rich cooperation.
No comments:
Post a Comment