Introduction: Rethinking Change Finding our way forward in a complex world

Change is often seen as an individual choice or a top-down plan. But in complex systems, this view breaks down. Change emerges through interaction, adaptation, and learning as people respond to one another and to changing circumstances. Rethinking change means paying attention not only to what we want to achieve, but to how we find our way forward together.

In this chapter, I explore a different way of understanding and approaching change. Rather than seeing change primarily as a matter of defining an outcome, creating a plan, and persuading people to follow it, we can see it as something that emerges through interaction within complex and uncertain conditions.

A useful starting point is the idea that interaction, not the individual, is the primary unit of change. Individuals matter. They can initiate ideas, challenge assumptions, and act differently. But whether those actions lead to wider change depends on how others respond and on the patterns of interaction that develop between people over time.

This changes how we think about directing change. Dave Snowden’s Vector Theory of Change suggests that in complex situations we can have a direction of travel without specifying a fixed destination. We can move towards more of what we want and away from what we do not want, while continually adjusting as circumstances change.

The adjacent possible adds another perspective. We cannot necessarily see the whole route ahead, but we can notice possibilities that are accessible from where we are now. Taking one step changes the situation and may reveal possibilities that were not previously available. We discover the path partly by moving along it.

But we still have to act. Safe-to-fail probes provide a way of doing this when we cannot reliably predict the consequences of our interventions. Rather than committing to one large course of action, we can make many small interventions in parallel, observe how the system responds, and learn from what emerges. Promising patterns can be amplified, undesirable ones dampened, and what we learn can inform what we try next.

Conversation runs through all of this. If interaction is the primary unit of change, then changing how we interact can change what becomes possible. The Vector of Great Conversations explores how a conversation can have direction and purpose without requiring a predetermined outcome.

Safe-to-fail conversations take this a step further. A conversation can itself become a small probe into a complex situation. Rather than relying on one large conversation to find the answer, we can create many small conversations in parallel, ask different questions, bring different perspectives together, and pay attention to what emerges. These conversations can reveal possibilities, difficulties, and patterns that were not visible beforehand.

Taken together, these ideas suggest a different way of thinking about change. We can have direction without knowing the destination, take action without pretending we can predict the outcome, and discover possibilities as we move.

Change becomes less about designing and implementing a future and more about interacting with the situation as it develops: probing, learning, adapting, and discovering what becomes possible next.

Chapter: Rethinking Change

1. Introduction: Rethinking Change Finding our way forward in a complex world
2. Interaction as the Primary Unit of Change Is not the individual, it’s the interaction
3. The Vector Theory of Change Navigating change in a complex world
4. Exploring the Adjacent Possible Understanding change as a series of nearby moves
5. Conversations and the Adjacent Possible How dialogue creates quiet shifts in what is possible
6. The Vector of Great Conversations A parallel between the vector theory of change and conversational leadership
7. Safe-to-fail Probes: Rethinking Change in Complexity Probing, sensing, and responding when the way forward is uncertain
8. Safe-to-fail Conversations Using conversation as a probe when the way forward is uncertain