System Boundaries: What's Inside, What's Outside

One intersection - one fix. A smart traffic light goes in, the jam clears. A week later a jam appears at the next intersection. Widen the boundary to the neighborhood - optimize routes. Jams migrate to neighboring areas. Widen to the city - build a subway. Fewer jams overall. The boundary of the system defined the depth of the solution the entire time. Where the boundary is drawn - that's where the solution lands.

• **Business**: Optimizing a department without accounting for dependencies on other departments
• **Urban planning**: Solving one neighborhood's problem creates problems in neighboring areas
• **Ecology**: Saving one species disrupts the ecosystem
• **Personal**: Optimizing work at the expense of health and relationships

What Are System Boundaries?

A company optimizes its logistics. Costs drop by 20%. Success? A year later it turns out: suppliers went bankrupt under pricing pressure, and now the company has no components. The problem? **The system boundaries were drawn incorrectly.**

**System boundary** - a conceptual line separating the system from its environment. What's inside the boundary is analyzed in detail. What's outside is treated as 'external factors.'

**Key insight**: Boundaries don't exist in reality. They exist in the mind. One **chooses** where to draw them.

The boundary determines what is seen and what is ignored. Draw it incorrectly - and the wrong problem gets solved.

Boundary Problems

Most failures in systems analysis come from incorrectly drawn boundaries. Three common mistakes:

**Example: Traffic congestion in a city**

**Law of problem conservation**: Boundaries that are too narrow don't solve the problem - they move it out of sight.

How to Choose Boundaries

If boundaries are a choice, how does one choose correctly? A few principles:

**Purpose principle**: Boundaries must include everything that affects goal achievement. Define the goal - then draw the boundaries.

**Other boundary-selection principles:**

1. **Include feedback loops**: If action A affects B and B affects A - both must be inside
2. **Check 'external factors'**: If an 'external factor' is critical - expand the boundary
3. **Account for stakeholders**: Whose perspective matters? Their systems intersect with the one being analyzed
4. **Start wider, narrow as needed**: Easier to exclude the unimportant than to discover what's been missed

**The expert trap**: Specialists tend to draw boundaries around their own expertise. A programmer sees a 'code system,' a marketer sees a 'brand system.' The real system is broader.

Nested Systems

Systems exist inside other systems. A cell is part of an organ, an organ is part of an organism, an organism is part of an ecosystem.

**Holarchy** (Arthur Koestler's term) - a hierarchy of nested systems where each level is simultaneously a whole (for the level below) and a part (for the level above).

**Why does this matter?**

• **Emergence across levels**: Company properties can't be reduced to team properties
• **Goal conflicts**: What's good for a team can be bad for the company
• **Top-down constraints**: Higher levels set the rules of the game for lower levels
• **Bottom-up innovation**: Changes at lower levels can transform upper levels

**Level error**: Solving a problem at the wrong level from where it originated. An industry-level problem cannot be solved within a single team.

Boundaries in Practice

How does boundary thinking apply in real work? A few practical techniques.

**'Three Boundaries' technique**: For any problem, draw three boundary variants - narrow, medium, wide. Compare what is visible in each case.

**'Stakeholder boundary' technique**: Ask different participants where they see system boundaries. The differences reveal blind spots.

**Boundary is not dogma**: Start with one set of boundaries, but be prepared to revise them as more is learned. Systems analysis is an iterative process.

Key Ideas

• **System boundary** - a conceptual line separating the system from its environment
• **Boundaries are a choice**, not an objective fact. In reality everything is connected
• **Three mistakes**: too narrow, too wide, wrong level
• **Purpose principle**: boundaries must include everything that affects the goal
• **Holarchy**: systems are nested within systems - each level is both a whole and a part
• **Three boundaries**: a technique for drawing narrow, medium, and wide boundary variants

Connections to Other Concepts

Boundaries determine which feedback loops are visible, which stocks and flows are accounted for, and at what level leverage points are sought.

• Feedback Loops — Boundaries determine which loops can be seen
• Mental Models — Our model determines our boundaries
• Leverage Points — Leverage points may lie outside narrow boundaries

Вопросы для размышления

• What system is currently being analyzed? Where are its boundaries?
• What is being treated as an 'external factor'? What if it's actually part of the system?
• At what level of the holarchy does most of the work happen? What's occurring at the level above and below?
• Which stakeholders see the system differently?

Связанные уроки

• st-01-feedback-loops — feedback loops define where the system boundary lies
• st-09-ecology — ecosystems as an example of fuzzy, contested boundaries
• st-11-social — social systems have especially ambiguous boundaries
• st-21-cynefin — Cynefin reframes how system boundaries and complexity are perceived
• cc-01-dags — selection bias in DAGs mirrors wrong system boundary choice
• ct-03