II. The History of Potential: VI. When a History of Instances Became a System

An innovation may begin with a single instance. But if it is taken up, extended and used in new circumstances, it can become more than something that happened once. It can contribute to the resources through which further instances are produced.

How does this happen?

Repetition alone is not enough. An expression may be repeated without acquiring a new function, and a pattern may recur without becoming a resource that participants can use. What matters is whether the history of instances contributes to an organisation that supports further meaning-making.

Consider a new way of using an expression. At first, its meaning may depend heavily on the circumstances of its occurrence. Other speakers may then adopt it, vary it and use it in circumstances that differ from the original. Through these relations, the expression may become part of a more general resource. Speakers can use it without having to reproduce the circumstances in which it first appeared.

The system has not simply accumulated another instance. Its resources have acquired an organisation that supports further instances.

This distinction is important. A system is not a catalogue of everything that has happened within it. Nor is it an inventory of possibilities independent of the history through which they have developed. Its organisation constrains and enables what can happen next, while new instances may contribute to changes in that organisation.

In systemic functional linguistics, this relation between system and instance is central. Instantiation draws on the meaning-making potential of a system. Yet the process of instantiation can also generate resources that become available for further instantiation. What is produced in one history of meaning-making may enter the conditions of another.

We should not assume that every instance contributes to such development, or that change follows a predictable path. Some innovations disappear; others persist without transforming the system substantially. Still others alter relations among existing resources, opening up ways of meaning that were not previously available.

A system's history, then, is not simply behind it. Some of that history becomes organised within the resources through which it continues to act.

What determines which instances leave a trace in a system's potential, and which remain events in its history?