The capstone exercise. You work with multiple mechanisms interacting across levels, analyse how they reinforce or constrain each other, and design interventions that target specific components. This integrates everything from the earlier QDE stages.

What

Given a complex scenario, you should be able to:

1. Identify which mechanisms are involved at each level
2. Name specific phenomena at each level
3. Explain how the levels interact
4. Propose interventions targeting specific mechanism components

When to use

Only after you're comfortable with level identification, core mechanism explanation, and key phenomena application. This is the final stage of the progression.

Preparation

Allow 45–60 minutes for the main scenario. Have the mechanism reference table available. If working in pairs, one person analyses while the other probes with "which mechanism?" and "what component would you target?"

How

The Drift

A well-regarded unit deployed with strong values. Over a 9-month deployment, behaviour deteriorated: dehumanising language became normal, rules were bent, eventually atrocities occurred. Soldiers who participated were not obviously "bad people" before deployment.

Analyse across all five levels.

Pattern detector: Repeated exposure to threat and enemy normalised patterns that would have felt wrong initially. Dehumanising language built pattern associations (enemy = less than human). Arousal from combat tagged aggression as appropriate.

Lazy controller: Lazy controllers across the unit generated disengagement techniques: euphemism, advantageous comparison ("they're worse"), displacement ("following orders / everyone else is"). Moral conflict was detected but disengaged from.

Affordance auction: The operational environment afforded violence, provided weapons, placed enemy in proximity. Time pressure and threat elevated aggressive affordances. Opportunities for reflection or dissent were scarce.

Prestige engine: The prestige engine stabilised new norms. Hardness, aggression, "getting results" earned status. Soldiers who objected lost social capital. Leaders (representatives) modelled or tolerated the behaviour, signalling its acceptability. Bonding capital was used to close ranks.

Enforcement infrastructure: Institutional monitoring collapsed over distance and operational tempo. The tight culture remained — but tight around local norms (the unit's drift), not institutional norms (the Army's values). Enforcement infrastructure existed but monitored the wrong things (metrics, not conduct).

Interaction: The pattern detector primed the intuitions, the lazy controller failed to override, the affordance auction enabled action, the prestige engine rewarded it and punished dissent, enforcement infrastructure failed to correct because its monitoring was decoupled from lived behaviour.

Interventions would need to target multiple levels: external monitoring that can't be captured (enforcement infrastructure), alternative prestige pathways (prestige engine), environmental redesign (affordance auction), disengagement recognition training (lazy controller), pattern disruption through rotation/leave (pattern detector).

Generate your own

The highest-value exercise. Writing a good scenario requires understanding the mechanism well enough to construct a situation where it's the primary driver.

Protocol:

1. Pick one mechanism as the primary driver
2. Write a scenario (3–5 sentences) where that mechanism is clearly at work
3. Add at least one secondary mechanism that interacts with the primary
4. Write an answer key: primary mechanism, secondary mechanism, interaction, one intervention
5. Trade with a partner. Work each other's scenarios. Discuss disagreements — they're often more instructive than agreements

Quality check for your scenario:

• Does the primary mechanism clearly drive the behaviour?
• Is the scenario specific enough to analyse (not just "things went wrong")?
• Could a reasonable person identify the primary mechanism from the text?
• Does the answer key name mechanism components, not just labels?

Why it works

Multi-level analysis is the target skill for ethical decision-making. Real ethical failures are never single-mechanism — they're cascades where each level contributes a small bias that compounds. The Drift scenario forces you to trace the full cascade, connecting mechanisms you've practised separately.

Scenario generation inverts the skill: instead of recognising a mechanism in someone else's scenario, you construct a situation where the mechanism operates. This requires deeper understanding — you need to know the mechanism's components well enough to build a context that activates them.

Outcomes

After completing these exercises, you should be able to:

• Analyse a novel ethical scenario across all five mechanism levels
• Trace how mechanisms interact (which enables, constrains, or amplifies which)
• Design interventions that name the specific component they target
• Write scenarios that reliably surface specific mechanisms
• Explain why single-level interventions ("just train more") usually fail