The Nature of Flexibility

veryard projects - innovation for demanding change

The Nature and Nurture of Flexibility

veryard projects > systems engineering > flexibility

we offer

the essence of flexibility

material

links

consultancy

management briefings

independent advice on tools and methods

It sometimes seems that flexibility is an illusion, for which people are asked to pay large amounts of money. Is this irrational, or what?

Seriously, it is obvious from numerous examples that we have to define what classes of change we are to provide flexibility for. However, any such definition is a restriction. There is a sense in which ‘true’ flexibility is always beyond any specific definition we may formulate.

However, some formulation of flexibility is necessary, if we are to develop any quantitative or even qualitative assessment of its value, to one or more stakeholders. The willingness of such stakeholders to incur costs in return for some level of flexibility must be understood in terms of a model of finite rationality.

On this page, we discuss the nature of flexibility, and its specification. We often need to formulate business cases and system requirements in terms of enhanced flexibility. We therefore need to understand what flexibility is, and how an enhancement in flexibility might be anticipated and measured.

on this page

Principles

Challenges

FAQ

elsewhere

Requirements

Flexibility as the Absence of Rules

A man dwelt in a very comfortable house, with a large, light, airy cellar. The river ran near by. One day the river overflowed, the cellar was flooded, and all the hens that he kept in it were drowned. The next day he bounced off to see the landlord. 'I have come,' he said,'to give you notice. I wish to leave the house.' 'How is that?' asked the astonished landlord. 'I thought you liked it so much. It is a very comfortable, well-built house, and cheap. 'Oh, yes,' the tenant replied, 'but the river has overflowed into my cellar, and all my hens are drowned.' 'Oh, don't let that make you give up the house, the landlord reasoned; 'try ducks!'

Frank Boreham, Mushrooms on the Moor, London: The Epworth Press, 1930, p. 52-53

Principles of Flexibility

veryard projects > systems engineering > flexibility > principles

	Situations / opportunities possess certain properties that demand flexibility in the solution.	uncertainty / risk variety / tension change
	Solutions are flexible to the extent that they possess certain properties that respond to the needs of the situation / opportunity.	requisite variety robustness adaptability contingency
	Flexibility is a property of a system, preserving some recognizable qualities in response to change. Change implies not-change.
	Flexibility depends on how the system is viewed.	level of abstraction system boundary

Challenges of Flexibility

veryard projects > systems engineering > flexibility > challenges

	The challenge for engineers is to match the flexibility implicit in the demand with the flexibility encapsulated in the solution.	scenario planning contingency planning
	The challenge for managers is to reason intelligently about flexibility and change.	business case

Flexibility FAQ

veryard projects > systems engineering > flexibility > FAQ

	Why flexibility? biological organism enterprise building pension plan organization information systems
	What is flexibility?
	How much does flexibility cost?
	How much is flexibility worth?
	How can we measure flexibility?
	How should we design for flexibility?

Requirement for flexibility

Flexibility is a desired characteristic of many designs. Let us consider a range of examples.

Biological organism In biology, survival involves adaptation to unknown conditions.

> Bateson on Adaptation

Also we are never quite sure whether we are interested in the survival of the individual, the species, the gene or the ecosystem. Kenwyn Smith argues that this type of confusion is a significant cause of muddled thinking in organizational change as well as biological evolution.

[K.K. Smith, 'Philosophical Problems in Thinking about Organizational Change' in Paul S. Goodman & Associates (eds), Change in Organizations (San Francisco CA: Jossey-Bass, 1984) pp 316-374]

Enterprise An enterprise should be flexible - this presumably means it should cater for the following classes of change: (i) in the competitive environment - power relationships and contracts between customers, suppliers and competitors

(ii) in the economic environment - interest rates, unemployment, energy costs, and so on

(iii) in the social environment - relationships with community, patterns of employment, consumption, and so on

(iv) in the technological environment - product, process and production technologies available to enterprise itself, customers, suppliers, competitors

(v) in the political / regulatory environment - relations with legislators and administrators (local, national, international), competition / trade policies, case law, and so on

Each of these classes of change is itself heterogeneous. Furthermore, this is by no means a complete list. We could talk about the flexibility of an enterprise with respect to a specific class of change, but then how do we sum across all changes? Is it meaningful to define a scalar addition of flexibility

Flex (E) = Si Flex (E, Di), where E is an enterprise and Di is a class of change

or should we be thinking of a multidimensional vector addition?

Building A building should be flexible - this presumably means that it should cater for at least four classes of changes: (i) in the requirements of the present occupant - for example, size, organizational changes, installation of technology

(ii) in the requirements of potential future occupants - how attractive is this building likely to be for new occupants?

(iii) in its operational efficiency - how well is it insulated, how much does it cost to heat/light/ventilate, and how much would it cost to upgrade these characteristics

(iii) in the institutional requirements - what rules may be applied by local authorities, health & safety inspectors, building societies, and what might it cost to conform to these rules?

Note: this raises the question: flexibility for whom? The present occupants, or the landlords? The identification of the stakeholders affects the assessment of flexibility. Pension plan A pension plan should be flexible - this presumably means that it should cater for at four classes of changes: (i) in the personal circumstances of the individual (health, acquisition/loss of dependants)

(ii) in the employment circumstances of the individual (identity of employer(s), full-time permanent employment versus contract)

(iii) in the economic environment - capital/interest structures of investment

(iv) in the fiscal/legal environment - changes in tax structures or other regulations

We are here assuming that the primary stakeholder is the would-be pensioner. Obviously, there are other stakeholders: the employer, the pension company, various intermediaries such as advisers and agents, the Treasury. Organization An organization (by which I mean the structure of relationships between the members of a collective entity such as an enterprise) should be flexible - this presumably means that it should cater for the following classes of change: (i) in the business strategies

(ii) in the business processes (including technology)

(iii) in the individual employees occupying each role (skill development, career progression, job rotation, staff turnover)

Information system An information system should be flexible - this presumably means that it should cater for the following classes of change: (i) functional requirements - this is known as maintainability

(ii) technical platform - this is known as portability

(iii) ???

The main benefits of open systems in general, and ODP in particular, can be expressed in terms of various forms of flexibility: portability, distributability, maintainability, xxx-ability.

What is flexibility?

Flexibility as response to change In the previous section, we saw that flexibility of a given entity was described in terms of its ability to cater for change. In order to gain a clear notion of flexibility, we need to specify which kinds of change we are talking about. Flexibility as property of a system A system is a bounded description of a portion of reality. It follows that flexibility is a mental property, belonging to the realm of mental objects. Two observers may describe the same set of entities, events and processes differently, with a different view of the boundary between the system and its environment. One may see more flexibility than the other.

Furthermore, flexibility is defined in terms of change, and change is itself a property of descriptions rather than of any underlying reality. Does this mean that flexibility is subjective? What does this question mean?

> Bateson on Change

In the evaluation of enterprises, we are provided with a conventional template model of value called the Balance Sheet. (Although this conventional model is itself subject to both interpretation and evolution, we may take it a priori to be a reasonably stable template.) This defines the scope of the enterprise (at least from the accountants’ point of view), and an objective method for evaluating a given class of changes to the enterprise.

Flexibility as not-change An entity is flexible if it maintains some stable property (description) in the face of environmental change. For example, an enterprise remains profitable despite changes in the competitive environment; a pension plan continues to satisfy a set of financial criteria, a building or information system continues to fit its purpose, even though the original designer didn’t know exactly what its purpose would become.

Flexibility is the ability to make small changes in order to avoid large (catastrophic) changes. For example, when riding a bicycle, you need to be able to make small adjustments to the front wheel to retain your balance. If the front wheel is fixed, you will fall off.

Flexibility depends on the caprice of the environment The word ‘caprice’ is used by Bateson to describe the tricks of Nature. It applies equally to the competitive environment of commercial enterprises, or to the other classes of system discussed in this paper. An enterprise is ‘encouraged’ to rely on some characteristic of the competitive environment, and then the rules of the game are changed. For example, as the result of some unanticipated political activity by a competitor. This is ‘not fair’.

> Bateson on Caprice

The latest wisdom of management gurus (such as Tom Peters) is that the only characteristic you can rely on is change itself. This sounds wiser than it actually is. The trouble is that as soon as you focus on certain classes of change, you neglect other change possibilities. Prepare yourself for turbulence of a particular kind, and you will be caught unawares by turbulence of a different logical type.

One of the ways of operating flexibly in a turbulent world is through flexible outsourcing and joint venture relationships with other companies. Many companies have been badly burned by such strategies, from IBM downwards. Recent studies suggest that companies are refusing to take advantage of interfirm coordination opportunities, for fear of losing bargaining power.

Eric K. Clemons & Michael C. Row, 'Limits to Interfirm Coordination through Information Technology: Results of a Field Study in Consumer Packaged Goods Distribution' Journal of Management Information Systems 10(1) Summer 1993, pp 141-163

Flexibility depends on the level of abstraction Let us consider a biological example. If a predator species evolves a new strategy for discovering and catching its prey, the prey species will need to evolve a new strategy for camouflage or escape. At one level, these new strategies seem to increase the flexibility of both species. At another level, however, the predator becomes all the more specialized in the capture of this particular prey species, and therefore the interdependence between the two species is increased. Flexibility for the individual may entail loss of flexibility for the species.

Similar considerations apply to commercial enterprises. Each more sophisticated marketing strategy may increase the interdependence between the enterprise and its particular customer base, making it more difficult for the enterprise to develop in new markets. For example, as a company develops ever more flexible and powerful techniques of direct mail marketing, it becomes ever more incapable of using any other marketing channel.

Flexibility depends on the system boundary If we define flexibility in terms of the responsiveness to events in the environment, then it obviously matters where we draw the line between the system and its environment. Paradoxes of flexibility Simplistic attempts to increase flexibility often result in a loss of flexibility.

Flexibility is achieved not by keeping your options open, but by making decisions. Keeping your options open is a form of neurosis, and leads to paralysis (Wiederholungszwang). Bateson calls this a narrowly homeostatic system, rigidly indecisive, and identifies it as a characteristic of schizophrenic families.

> Bateson on Being Uptight
> Lacan on Indecision

And yet making the wrong decision may also reduce flexibility. Remember IBM, who thought that outsourcing the writing of an operating system for the PC would increase its flexibility.

Cost of flexibility

In a stable and predictable environment, habit is worth more than flexibility. Habit is a way of reducing operational costs, by eliminating the need to consider certain factors. If you always use the same route to work, this saves you having to waste mental energy each morning deciding which route to use. If you have a standard procedure (habit) for handling customer complaints, then you can ensure each customer complaint is handled effectively and efficiently, provided that every customer complaint satisfies the assumptions built into the procedure. The SEI capability maturity model refers to this characteristic as process repeatability.

In a hostile world, habit makes people more vulnerable to surprise attack. Kidnappers and burglars often rely on the predictable habits of their victims. And in a competitive marketplace, an aggressive player may rely on the predictable behaviour of its competitors to poach their customers.

Organisms and organizations need habit. But which habits?

Flexibility may benefit the managers (or shareholders) of an enterprise, but may place additional (and unwelcome) pressures on staff

Reg Sell, 'Quality of Working Life and Organizational Effectiveness' in M.C. Jackson, P. Keys & S.A. Cropper (eds) Operational Research and the Social Sciences (New York: Plenum Press, 1989) pp 667-671

Flexibility will often add to the difficulty of the design and the costs of construction, as well as to the costs of operation. Infinite flexibility is likely to be technically as well as economically infeasible.

Organizational issues

We define flexibility as the ability to maintain a given state of affairs in the teeth of (a class of) environmental change. We define responsibility as an answerability (of an agent to another agent) for a given state of affairs. Therefore there is a close conceptual relationship between flexibility and responsibility.

Responsibility and Flexibility

Relationship between flexibility and responsibility

In the management literature, there have been attempts to design flexible organizations. Ackoff, for example, claims to have a method of designing an organization that can adapt without reorganizing. He posits three dimensions of the division of labour: functional, product and market, and asserts that most of the cost and disruption of reorganization is caused by changing the order of these three dimensions. He then argues that a multi-dimensional structure (incorporating the same three dimensions at all levels of the organization) will never need major reorganization.

R.L. Ackoff, 'On Flexibility and Freedom' in M.C. Jackson, P. Keys & S.A. Cropper (eds) Operational Research and the Social Sciences (New York: Plenum Press, 1989) pp 19-34

His argument is fallacious. Firstly, each of the dimensions may change: BPR may alter the arrangement of functions; a new marketing vision may radically alter the product groupings; socio-political events may reshape the markets. Secondly, he fails to recognize the extent to which reorganization is the result of self-serving action by politically astute individual managers. Thirdly, his structure is so complicated that when it does change, the cost and disruption will probably be all the greater.

Technological issues

If something becomes a commodity, its use is (perhaps) more flexible. According to Borgmann, technology is a bridge between means (device) and ends (benefit). As the technology becomes more sophisticated, there is an increasing separation between the device providing a benefit and the benefit itself. This makes the user independent of the device, makes the device partially or wholly invisible (transparent) to the user. ODP is an abstract technology in this sense.

> Borgmann on the Device Paradigm
> Notes on ODP

How do you get portability, distributability, and so on? By buying into a more abstract technology. See the following table. Flexibility at one level always drives you to a form of dependence at the next level.

Level of independence	Strategy
Hardware model independence	Select a hardware supplier with an adequate range of consistent models
Hardware supplier independence	Select a software platform supported by a range of hardware suppliers (e.g. UNIX, MSDOS)
Database independence	Select a database interface standard (such as SQL), supported by a range of DBMSs. Select a CASE tool capable of regenerating applications on a range of database management systems
Development tool independence	Select a CASE platform supported by a range of CASE tool vendors
Device independence	Select a device interoperability standard (such as ODP), supported by a range of device vendors.

In the technical literature, change management is often formalized by specifying a class of changes that are to be considered, so that it becomes a formal mathematical puzzle. This work is obviously useful at the technical level, but raises questions about the extent to which these formal models represent the actual requirements for change and flexibility in the real world.

See for example Jeff Kramer, Jeff Magee & Morris Sloman, 'Managing evolution in distributed systems' Software Engineering Journal November 1989, pp 321-329. Jeff Kramer & Jeff Magee, 'The Evolving Philosophers Problem: Dynamic Change Management' IEEE Transactions on Software Engineering Vol 16 No 11, November 1990, pp 1293-1306

Formalization of flexibility Increasing stability In order to specify flexibility in terms of the maintenance of desired states, we need to fix the following: a what exactly is the system whose flexibility is to be considered?

b what are the states (statements) that are to remain true of the system - for an enterprise, we may assume these are given by the balance sheet, or some extended version of a balance sheet?

c what are the changes (classes of change) to which the system must be able to respond without altering the truth of the statements in (b)?

d what is the perspective from which these changes / non-changes are to be assessed? who are the stakeholders?

Increasing responsiveness One way of formalizing flexibility could be in terms of the set of possible actions. An increase in flexibility can be defined as an increase in the quantity and variety of possible actions.

Enhancing Flexibility

However, this raises a number of theoretical problems.

• In order to count the number of possible actions, we have to have a criterion of identity for the class of possible actions. In other words, we have to know whether A is the same action as B.

• In order to assess the variety of possible actions, we have to have some measure of the difference between two actions. Thus (from the flexibility point of view) it is presumably better to have two radically different choices of action than to have two similar choices.

• Each possible action may be a reasonable (i.e. not bad) response to a set of possible changes in the environment. This requires a model of reasonableness, which is highly problematic.

Recommended reading unfortunately, many of the classic works in the field are out of print
Gregory Bateson, Steps to an Ecology of Mind (San Francisco CA: Chandler Publishing Company, 1972)
Gregory Bateson, A Sacred Unity: Further Steps to an Ecology of Mind (edited R.E. Donaldson, New York: Harper Collins, 1991)
Paul S. Goodman & Associates (eds), Change in Organizations (San Francisco CA: Jossey-Bass, 1984), especially contributions by Kenwyn K Smith and Karl E. Weick.
John R. Kimberly & Robert E. Quinn (eds), New Futures: The Challenge of Managing Corporate Transitions (Homewood IL: Dow Jones-Irwin, 1984), especially contributions by Kenwyn K Smith and the editors.
Gareth Morgan, Images of Organization (London: Sage, 1986), especially Chapter 8. New full edition (1997). Avoid the abridged edition (1998).

Document history

The first draft of this document was written in December 1993, as part of the Enterprise Computing Project. A major revision was done in October 1997.

We are currently working on aspects of the Adaptive Business, as part of the SITE project. Further material will be published here when it's ready.

Flexibility - Links

veryard projects > systems engineering > flexibility > links

SITE Project - The Fluid Business
Flexibility Home Page

veryard projects > systems engineering > flexibility