Seafield Research and Development Servicessystems approach
Home Up


Up Sister Sites:

Training of Trainers

Social Development

Land Reform
Social Landownership

All the one-pagers in this section are based on

Clayton AMH and Radcliffe NJ (1996)

Sustainability - a systems approach;

Earthscan: ISBN 1 85383 3193

    To download word processed versions click below
Interventions to promote sustainability Compilation G970513a.wpd
The Nature of Systems   G970509d.wpd
Defining Systems   G970509c.wpd
Systems Theory - hard systems   G970512c.wpd
Systems Theory - soft systems   G970512d.wpd
Positional Analysis   G970512e.wpd
Flexibility is necessary for Sustainability   G970512b.wpd

Interventions to promote Sustainability

It was common in the past that when development projects ended they also collapsed. It is not thus surprising  to hear the call for development interventions to be sustainable.

Most development interventions are channelled through organisations within local social systems - mainly through government  but increasingly through NGOs. This being the case it is useful to consider the nature of ‘systems’ and the history of the ‘systems thinking’ which, known to them or not,  pervades the thinking of most development professionals.

This collection of one-pagers covers the main ideas inherent in the systems approach, differentiates between hard and soft systems, and suggests some strategies for building elements of sustainability into the latter.

In ‘The Nature of Systems"  we see that the systems approach deals not only with things and events but also with the relationships between them. The key concepts of emergence, hierarchical control and communication are briefly expounded.

In ‘Defining Systems’  we note that there is no commonly accepted way of categorising the many different types of system which exist but the fundamental difference between natural, designed and human activity systems is noted.

The next two papers look in turn at the characteristics of hard and soft   systems. The shrewd reader may reflect that most development project planning assumes that intervention in social systems is a ‘hard’ process while it blatantly is not!

In ‘Positional Analysis’  there is a brief sketch of a technique to bring transparency to decision making processes - such an important requirement when dealing with soft systems.

The last paper throws down the gauntlet by saying that ‘Flexibility is necessary for Sustainability’ and that this can only be achieved through a paradigm shift in management style.

Hands up all those hard system control freaks who

(a) begin to see the error of their simplistic ways and
(b) are willing to let go softly?

Note: You might find it useful to refer to two other collections of one-pagers which relate to this issue:

bulletThe set of six one-pagers dealing with Guba & Lincoln’s concept of 4th Generation Evaluation.
bulletg970306 Stakeholder involvement for project sustainability, and the new style TCO (Technical Cooperation Officer)


The nature of systems

A systems approach involves placing as much emphasis on identifying and describing the connections between objects and events as on identifying and describing the objects and events themselves.

Systems are patterns of cause and effect relationships. These can be simple and unidirectional or they may be linked together in long chains. Any one factor can exert a control function (causing a change in another) and a dependent function (being changed by another) - this is called multi-factoriality.

Chains of cause and effect can be circular so that control function eventually becomes its own dependent function - this is called a feedback loop - and the influences can be either positive, when it causes an increase in the original effect, or negative, when it causes a decrease in the original effect.

Some of the more important defining characteristics of systems are emergence, hierarchical control and communication.


This means that at any given level of complexity, there are emergent properties that cannot readily be explained solely by reference to lower levels.


Hierarchies are levels of relative complexity within a system, and hierarchical control refers to the imposition of new functional relationships by each level on the detailed dynamics of the level below. Controls can be positive (where certain actions are promoted) or negative (where certain actions are constrained).

One of the challenges facing biological systems, in particular, is to optimize between excessive control (which gives little flexibility, a small behavioural repertoire and so a limited ability to respond to new circumstances) and insufficient control (which reduces the ability of the system to determine outcomes, and so incurs a higher risk that the internal processes themselves might drift beyond system limits and cause the system to disintegrate).


This refers to the transmission of information in some form to effect regulation and feedback. Information must flow from the regulator to the regulated in order for the regulator to exercise control. Information must also flow back from the regulated to the regulator if the regulator is to be able to monitor the compliance of the regulated and incorporate the information into its future programme. Positive and negative feedback loops are therefore the core of the process of communication.

The more recent extension of systems theory to include social and economic systems has required an evolution of the concept of communication, as it is the communication of meaning rather than information that is important in the world of human systems.

Defining Systems

There are a number of steps to go through when building a model of a system.

  1. The identification of the coherent elements of the system, and the definition of the principles of coherence. This can be quite difficult as some systems are organised on an ad hoc basis.
  2. The identification of the control mechanisms by which the system maintains its coherence, and the value ranges within which these operate. This can be complicated by the fact that biological and social systems are often characterised by redundancy, which means that they use multiple pathways of control.
  3. The delineation of the system boundary. The system boundary defines the inputs and outputs to the system. This is particularly difficult for social systems in this communications age.
  4. The identification of any subsystems of the system, or supersystems.

There is no general agreement on how systems might be classified. Some general categories are

bulletliving v non-living,
bulletconcrete v abstract,
bulletopen v closed.

Other categories include basic, operational, purposive and controlling systems. The following groups have proved useful:

bulletnatural systems                                  atoms, planets
bulletdesigned physical systems              machines
bulletdesigned abstract systems              mathematics
bullethuman activity systems                     socio-political structures
bullettranscendental systems                    systems beyond 'knowledge'

The systems in these classes are of quite different types. In particular, designed and activity systems are different from natural systems. This is because humans are self-aware, and monitor their own behaviour, and because of the ‘self fulfilling prophecy’ problem, where the knowledge of the prediction itself can become an input to the system, and alter the likelihood of the possible outcomes.

Systems Theory - Hard Systems

Systems Theory has two main approaches - the original hard systems approach (more relevant to technical, engineered systems) and the more recent soft systems approach (more relevant to human and social systems). Here we look at two hard systems approaches.

The hard systems approach is essentially about defining the problem solving sequence.

  1. Problem definition. This involves defining the problem and what has to be done
  2. Choice of Objectives. This involves deciding what would be required to reach each objective, and formulating the measures of effectiveness which then form the basis for making comparisons between strategies.
  3. Systems Synthesis. This involves identifying the various possible alternative strategies.
  4. Systems Analysis. This involves analysing and evaluating the various hypothetical systems in the light of the objectives.
  5. Systems Selection. This means choosing the most promising alternative.
  6. System Development. This entails developing the chosen alternative up to the prototype stage.
  7. Current Engineering. This consists of the realization of the system, but also includes the essential processes of monitoring the system, feeding this information back to the design stage, and then modifying the system as necessary.

Another version of this approach focuses on the costs and benefits of alternative programmes.

  1. Defining Objectives. This involves clarifying the desired aims and goals.
  2. Describing the various alternative techniques or systems available for achieving those objectives.
  3. Identifying the costs and resources required to achieve each alternative
  4. Developing systems models, in the form of a mathematical or logical framework that can show the interdependence of the objectives, the systems, the environment, and available resources.
  5. Developing the criteria for selection, and relating the objectives, the costs, and the resources to choose the optimal or otherwise preferred alternative.

A number of problems arise when these hard systems approaches are applied to soft systems, especially those systems that involve humans.

The hard systems approach starts with a basic acceptance of the objectives, problem specification, and organizational needs. Hard systems engineering aims to provide a solution to a defined problem in the terms in which the problem is posed, so these factors are generally taken as given.

With soft systems, however, there are frequent disagreements as to what the goals and objectives should be. It is very important to recognise this issue, and deal with it so that the whole process does not degenerate into what has been called ‘the quest for improved means to carelessly examined ends.’

Systems Theory - Soft Systems

Systems Theory has two main approaches - the original hard systems approach (more relevant to technical, engineered systems) and the more recent soft systems approach (more relevant to human and social systems). Here we look at the soft systems approach.

With soft system applications, system thinking should be regarded as a contribution to problem-solving, rather than as a goal-directed methodology and this applies to all situations where the task itself cannot be entirely and objectively defined.

Factors to be considered:

bulletwhen the definition of the problem depends of the viewpoint adopted then it is important to make that viewpoint explicit; the world views of the various actors provides the context within which events are given meaning, the various meaning systems have to be taken into account.
bulletthe problem itself gets redefined in the process of solving it so you must be prepared to go back to the first stage of problem definition as required ie be prepared for iteration; soft system methodology has more feedback than hard systems
bullethuman systems are always multivalued and have informal intentionalities related to establishing or modifying relationships in terms of power structures; roles, norms, values concepts and applications of power are all critical and should be openly addressed rather than hidden. This often makes it necessary to use more participative planning techniques.
bulletit is important to define the roles and functions of the humans in the system before designing the service system.
bulleta more open ended approach is required where the outcome is not seen as being an optimal solution but rather a continuous learning process

This being the case, the stages of a soft system approach are more like the following:

  1. reviewing the unstructured problem situation
  2. clarifying and expressing the problem situation
  3. defining the relevant systems and subsystems, whether these are formal of informal
  4. building conceptual models, scenarios and analogies
  5. comparing these models with the expressed situation
  6. effecting such changes as are currently both feasible and desirable
  7. taking action to improve the problem situation

Note: ‘desirability’ can be decontextualised but ‘feasibility’ cannot. What may be deemed highly desirable from some particular viewpoint may not be feasible in a given context because of a clash of meaning systems and/or the particular, informal intentionalities of key actors.

Positional Analysis

Positional analysis is a technique to illuminate and improve the decision making process within systems by making explicit all the choices being made. It uses multiple indices and accounting procedures to explicitly acknowledge the diversity of factors in complex issues and therefore make the decision making process more transparent.

When used effectively the decision maker can readily answer such questions as, ‘Why was this approach chosen?’, and ‘Which other approaches were considered and why were they rejected?’.

The specific functions of positional analysis are to:

  1. produce reliable knowledge through systematic enquiry
  2. illuminate the decision making process
  3. identify all the options and trade-offs available to decision makers
  4. make the decision making process more transparent, by, for example, identifying winners and losers, commonalities and conflicts of interest and ethical and ideological issues
  5. use complete option profiles on the key relevant dimensions, and to match these with the value profiles of the decision makers.

Positional analysis consists of the following steps:

  1. identifying the relevant conditions, factors and dimensions when dealing with a compound problem or making a strategic decision
  2. establishing procedures for quantifying, measuring, ranking or otherwise prioritizing change on each of these dimensions.
  3. measuring or rating the options or scenarios in terms of the change on each dimension
  4. identifying trade-offs on each of the relevant dimensions
  5. making a decision on the basis of the overall profile of each option

Note that the gathering of information is not enough. There is also a need for the will and a mechanism for incorporating the information into the decision making process and implementation plan. This may require modifications to the organisational structure.

If the provision of information is widely based and participatory this in itself may be an effective way of changing behaviour because:

bulletpeople normally try to incorporate new information into their actions through progressive stages of assimilation and adaptation, and
bulletinformation that an individual or group is required to report on will influence them to produce actions that will look ‘benign’ - the information inductance effect.


If changes to systems of managing information or making decisions are to be effective then they must be:
bulletunderstood and supported within the group
bulletintroduced as part of a coherent strategy
bulletaccorded a high organisational priority

Flexibility is necessary for Sustainability

It is said that the only constant thing is change. If a system is to sustain itself through time it must be able to adapt to changes in its external environment.

The key to the ability to adapt is flexibility which may be defined as follows:

flexibility = an uncommitted potentiality for change.

Any system operates according to rules, regulations and procedures. In soft, human and social systems these may be viewed as tradition and thus not open to change. ‘We have always done it this way’. Such systems are rigid and often resistant to change and will not thus be sustainable.

What is required in a sustainable system is a more scientific attitude towards rules, regulations and procedures. Science no longer talks of truth but rather of the best working hypothesis in the light of evidence presently available.

Operators in a sustainable system must be constantly scanning the external environment by way of noting changes that are taking place and relating the ongoing relevance of the existing rules, regulations and procedures. When a mismatch is found the operators must be able and willing to revise the way things are done and to establish new traditions.

In soft, human and social systems, however, there is always the danger that operators may become over enthusiastic and seek to change the rules, regulations and procedures when there is no real need to do so. This is change for change’s sake and is as great a danger to the sustainability of the system as resistance to change.

Any solution to the problem of sustainability therefore has to solve two connected problems:

  1. flexibility must be found or created
  2. the system must be prevented from immediately absorbing the new flexibility

Effective scanning of the environment is best done when all the eyes and ears of the system are used as sensors and when channels of communication are created so that the ultimate decision makers can make use of the information so gathered.

In management terms this involves a move away from a top down to a bottom up ethos and from a Theory X to a Theory Y style of leadership.

The concepts of participation and empowerment are also relevant as aspects of a move towards a commitment to staff training and development which would be operationalised, at least in part, through increasing the transparency and collegiate nature of decision making processes.