diff --git a/IDEA_TRAILS.md b/IDEA_TRAILS.md
index 7ba9250e..af1dca06 100644
--- a/IDEA_TRAILS.md
+++ b/IDEA_TRAILS.md
@@ -394,17 +394,17 @@ This trail also asks: **What behaviors does this system select for?** Capability
 
 ## Root question
 
-> How can a person, community, institution, intelligence, or technical system change enough to adapt while preserving or rebuilding the capacities, relationships, and meaningful possibilities that make continued becoming possible—and who gets to shape that possibility space?
+> How can a person, community, institution, intelligence, or technical system change enough to adapt while preserving or rebuilding the capacities, relationships, and meaningful possibilities that make continued becoming possible—including when expectations fail—and who gets to shape that possibility space?
 
 | Project | Treatment |
 |---|---|
-| **Root Sequence** | Develop shared language for path dependence, dynamic coherence, agency envelopes, meta-agency, continuity-through-transformation, reflexivity, and adaptation across scale while testing whether existing disciplines already supply sharper concepts. |
+| **Root Sequence** | Develop shared language for path dependence, dynamic coherence, agency envelopes, meta-agency, continuity-through-transformation, reflexivity, anticipation under uncertainty, and adaptation across scale while testing whether existing disciplines already supply sharper concepts. |
 | **Conscious Systems** | Separate organismic, embodied, autobiographical, social, and subjective continuity; ask whether bodily regulation, development, learned history, and system boundaries matter to experience without assuming that identity and consciousness are the same problem. |
-| **Being Human(e)** | Ground adaptation in lived bodies and environments: fear, trauma, disability, learning, recovery, care, and changing possibility without diagnosing people from abstract systems language. |
-| **Coherent Computing** | Preserve future user capability across repair, automation, migration, identity, service, and failure: inspect, understand, modify, refuse, recover, and continue without unnecessary lock-in. |
+| **Being Human(e)** | Ground adaptation in lived bodies and environments: fear, trauma, disability, learning, recovery, care, and changing possibility without diagnosing people from abstract systems language or treating anticipated reactions as substitutes for lived experience. |
+| **Coherent Computing** | Preserve future user capability across repair, automation, migration, identity, service, and failure: inspect, understand, modify, refuse, recover, and continue without unnecessary lock-in; prepare for known failure modes without making continuity depend on one forecast. |
 | **Community Infrastructure** | Adapt across legal, technical, institutional, and cultural contexts without making any one surrounding institution the ontology of community; preserve portability, local authority, and graceful degradation. |
 | **Liberated Intelligence** | Ask how intelligences can learn, change, retain or revise identity and memory, and participate in relationships without treating capability as authority or functional adaptation as evidence of consciousness. |
-| **Coherent World** | Model history as changes in future dynamics and affordances, not merely a log of events; let interventions reshape agents, institutions, dependencies, feedback, and later possibility without reducing coherence to one score. |
+| **Coherent World** | Model history as changes in future dynamics and affordances, not merely a log of events; let interventions reshape agents, institutions, dependencies, feedback, and later possibility; let anticipation remain partial and fallible so surprise can revise later models and behavior rather than reducing coherence to one score. |
 | **Museum of Ordinary Life** | Preserve evidence of trajectories—ordinary objects, routines, interfaces, records, testimony, and material traces that reveal how earlier conditions became later possibilities or constraints. |
 
 **Canonical cross-project concept:** [Dynamic Coherence and Adaptive Continuity](concepts/dynamic-coherence.md)<br>
diff --git a/concepts/design-grammar.md b/concepts/design-grammar.md
index 588ad662..4583749b 100644
--- a/concepts/design-grammar.md
+++ b/concepts/design-grammar.md
@@ -1,7 +1,7 @@
 # Root Sequence Design Grammar
 
 **Document role:** Working synthesis of recurring cross-project design principles<br>
-**Status:** AI-assisted draft for author review, revised 2026-09-28<br>
+**Status:** AI-assisted draft for author review, revised 2026-10-01<br>
 **Canonical scope:** Shared questions and commitments at Root Sequence level; each project owns its own design decisions, tests, governance, and canon.<br>
 **Evidence boundary:** Normative synthesis from a conversation, not a validated universal theory or evidence that every project already implements every item.
 
@@ -25,6 +25,7 @@ This wording is a proposed synthesis, not an approved project slogan. Coherence
 | Deliberate before consequential decisions | Do participants have enough shared understanding, access to evidence, room to revise, and protection for dissent before collective authorization? | [Deliberative Inquiry](../research/methods/deliberative-inquiry.md) and [Collective Judgment](../analysis/collective-judgment-and-manufactured-consensus.md) |
 | Let evidence revise the model | What observation, counterexample, failed prediction, or affected-party evidence would require this model or design to change? | [Research Methods](../research/methods/README.md) |
 | Preserve adaptive continuity | What should remain stable, what must remain changeable, and which meaningful future actions does this decision open or close? | [Dynamic Coherence and Adaptive Continuity](dynamic-coherence.md) and [Adaptation](../systems/adaptation/README.md) |
+| Anticipate what you can; design for what you can’t | Which downstream effects, responses, dependencies, and failure modes are reasonably foreseeable? If the model is wrong, what stays observable, reversible where appropriate, recoverable, and adaptable? | [Adaptation model](../systems/adaptation/model.md#16a-anticipation-and-surprise), [Dynamic Coherence](dynamic-coherence.md), and [Resilience](../systems/resilience-and-graceful-degradation.md) |
 | Expand agency through chosen participation and delegation | Which burdens can be removed, and who retains a meaningful say over the result? | [Agency and Automation](agency-and-automation.md) |
 | Keep authority scoped and contestable | Who may act for whom, on what basis, and who else bears the effects? | [Agency and Automation](agency-and-automation.md) and [Idea Trail 9](../IDEA_TRAILS.md#trail-9--intelligence-automation-and-legitimate-authority) |
 | Prefer interoperable paths where they preserve autonomy | Can people change implementations, exchange information, and leave without losing the underlying capability? What coordination costs arise? | [Legible Systems](legible-systems.md#design-principles) |
@@ -42,9 +43,11 @@ This wording is a proposed synthesis, not an approved project slogan. Coherence
 The grammar is not only about what to design; it also describes how Root Sequence should remain corrigible while designing. A useful shared loop is:
 
 ```text
-question → contrast → evidence → model → deliberation → action → observation → revision
+question → contrast → evidence → model → anticipation → deliberation → action → observation → revision
 ```
 
+**Anticipation is a prompt, not authority.** Use the current model to look for foreseeable consequences and prepare for plausible stress, then ask what remains possible if the model is incomplete or wrong. The point is to improve preparedness without turning one imagined future into a brittle plan.
+
 The shorthand **“reality gets veto power”** means that no conclusion is protected merely because it is elegant, politically congenial, already embedded in Root Sequence, or associated with a respected source. It does not mean evidence is perspective-free or instantly decisive; uncertainty, method quality, and affected perspectives remain part of the model.
 
 These moves can conflict. More detailed records can harm privacy; a common protocol can impose burdens on local variants; automation that frees one person can shift labor to another. Name the tradeoff and the affected people instead of claiming that one principle always wins.
@@ -53,10 +56,10 @@ These moves can conflict. More detailed records can harm privacy; a common proto
 
 A project should link to the relevant shared page, then say what the idea means under its own constraints. For example:
 
-- **Coherent Computing:** compare ways to participate, delegate, inspect, pause, recover, repair, migrate, and preserve future capability across devices and activities. Its own requirements and scenarios remain authoritative.
+- **Coherent Computing:** compare ways to participate, delegate, inspect, pause, recover, repair, migrate, and preserve future capability across devices and activities. Anticipate foreseeable failure paths without making survival depend on one forecast; keep unexpected states diagnosable and recoverable. Its own requirements and scenarios remain authoritative.
 - **Community Infrastructure:** ask who authorizes a Community Assistant's work, who inherits administration or care labor, and what happens without the assistant. Communities retain their own governance.
-- **Coherent World and its simulation:** explore whether interventions alter not only present outcomes but later affordances, dependencies, feedback, and agency. Avoid a single coherence score; history-dependent simulation output is a question, not canon or evidence about the real world.
-- **Being Human(e):** examine how help, dependence, fatigue, refusal, and dignity feel in ordinary contexts; human-scale conclusions belong in its own editorial process.
+- **Coherent World and its simulation:** explore whether interventions alter not only present outcomes but later affordances, dependencies, feedback, and agency. Let agents and institutions anticipate from partial models, let those expectations fail, and make surprise capable of changing later behavior. Avoid a single coherence score; history-dependent simulation output is a question, not canon or evidence about the real world.
+- **Being Human(e):** examine how help, dependence, fatigue, refusal, and dignity feel in ordinary contexts. Anticipate likely burdens, misunderstandings, and support needs without scripting people into predicted reactions; actual experience remains evidence. Human-scale conclusions belong in its own editorial process.
 
 See [Idea Trails](../IDEA_TRAILS.md) for broader routes. These examples are **proposed translations**, not claims that each project has adopted this page.
 
@@ -64,4 +67,4 @@ See [Idea Trails](../IDEA_TRAILS.md) for broader routes. These examples are **pr
 
 Use [Thought Routing](../THOUGHT_ROUTING.md) to capture the source once, compare existing treatments, and choose a canonical home. A useful path may be **capture → investigate → synthesize → make a project-specific decision → test or observe → revise**. It is not a required promotion ladder: an idea can stay a question, split, be contested, or be retired. Research can challenge a principle; a principle does not certify a research finding.
 
-**Source note:** This synthesis was prompted by the 2026-09-24 discussion of recurring project patterns, agency/automation, and the later dynamic-coherence/adaptive-continuity discussion. A 2026-09-28 synthesis/connection discussion sharpened the treatment of relationships as first-class objects and synthesis as distinct from replacement expertise. A 2026-09-29 Existential Euphoria / Root Sequence comparison added the provisional Membrane Test: a boundary-and-interface prompt for relationship without erasure. The conversation text was available during drafting, but no durable conversation URL was supplied. The table and wording are AI-assisted proposals for author review.
+**Source note:** This synthesis was prompted by the 2026-09-24 discussion of recurring project patterns, agency/automation, and the later dynamic-coherence/adaptive-continuity discussion. A 2026-09-28 synthesis/connection discussion sharpened the treatment of relationships as first-class objects and synthesis as distinct from replacement expertise. A 2026-09-29 Existential Euphoria / Root Sequence comparison added the provisional Membrane Test: a boundary-and-interface prompt for relationship without erasure. A 2026-10-01 discussion added the operational anticipation rule—“Anticipate what you can. Design for what you can’t.”—as a derivative of adaptive continuity, resilience, and corrigibility rather than a separate theory. The conversation text was available during drafting, but no durable conversation URL was supplied. The table and wording are AI-assisted proposals for author review.
diff --git a/concepts/dynamic-coherence.md b/concepts/dynamic-coherence.md
index 26f1d021..041bc4b4 100644
--- a/concepts/dynamic-coherence.md
+++ b/concepts/dynamic-coherence.md
@@ -1,7 +1,7 @@
 # Dynamic Coherence and Adaptive Continuity
 
 **Document role:** Cross-project conceptual framework and synthesis  
-**Status:** AI-assisted draft for author review, 2026-09-24  
+**Status:** AI-assisted draft for author review, revised 2026-10-01  
 **Canonical scope:** Shared Root Sequence language for change, continuity, agency, adaptation, and future possibility. Systems-specific mechanisms belong in the systems layer; project-specific implementations remain local.  
 **Evidence boundary:** This document connects established concepts from several literatures with Root Sequence working language. The synthesis itself is not a validated universal theory. Similar language across organisms, minds, institutions, ecosystems, and artificial systems does not establish a shared mechanism.
 
@@ -812,6 +812,27 @@ Critical Systems Thinking is an important predecessor for this pluralist stance;
 
 ---
 
+## 16A. Anticipation without brittleness
+
+Dynamic coherence has a prospective side: a system can use history, feedback, models, and domain knowledge to prepare for what is reasonably foreseeable. But a system can also become brittle by optimizing too tightly around one expected future.
+
+A practical Root Sequence formulation is:
+
+> **Anticipate what you can. Design for what you can’t.**
+
+This does not require omniscient forecasting. It asks two different questions:
+
+1. **What can reasonably be anticipated?** Downstream consequences, affected-party responses, dependencies, capacity limits, known failure modes, information needs, delayed effects, and assumptions that can be examined before acting.
+2. **What if the model is incomplete or wrong?** Preserve the ability to observe, revise, recover, migrate, appeal, refuse, or adapt where those capacities are appropriate to the domain.
+
+The distinction matters. Foresight without adaptability can produce premature optimization, analysis paralysis, or preemptive constraints around futures that never occur. Adaptability without foresight can repeatedly rediscover preventable failures at other people's expense.
+
+Unexpected outcomes are therefore not merely exceptions to the model. They can be evidence about the model's boundary, missing variables, false assumptions, or changing conditions.
+
+This overlaps established resilience-engineering work on responding, monitoring, learning, anticipating, graceful extensibility, and sustained adaptability. Root Sequence's role is to connect those questions to agency, power, reversibility, affected parties, and future possibility—not to rename the underlying mechanisms.
+
+---
+
 ## 17. Candidate Root Sequence principles
 
 These are candidates for testing, not doctrine.
@@ -828,6 +849,7 @@ These are candidates for testing, not doctrine.
 10. **Cross-domain resemblance is evidence for comparison, not evidence for identity of mechanism.**
 11. **Any coherence claim should name boundary, scale, timescale, affected parties, and externalized costs.**
 12. **Preserving the capacity for future revision can be an important design objective, but it does not override present harm, commitments, safety, or material outcomes.**
+13. **Anticipate foreseeable consequences, but do not make one prediction the system’s single point of failure; preserve ways to detect mismatch and revise.**
 
 ---
 
@@ -887,7 +909,7 @@ Preserving “future possibility” can become an excuse for failing people in t
 
 ### Root Sequence
 
-Use this page as shared language, then route causal work to Systems, empirical questions to Research, and project decisions to their local repositories.
+Use this page as shared language, then route causal work to Systems, empirical questions to Research, and project decisions to their local repositories. Treat anticipation as a design prompt: examine foreseeable downstream effects, then preserve enough optionality and feedback for the model to be corrected.
 
 ### Coherent Systems living paper
 
@@ -899,11 +921,11 @@ Use the framework to generate distinctions among bodily continuity, causal histo
 
 ### Being Human(e)
 
-Translate the framework into grounded questions about fear, learning, disability, identity, recovery, care, and environment without turning people into state machines or diagnosing them from abstract models.
+Translate the framework into grounded questions about fear, learning, disability, identity, recovery, care, and environment without turning people into state machines or diagnosing them from abstract models. Anticipate likely burdens or support needs where evidence permits, but do not substitute predicted reactions for what people actually report and experience.
 
 ### Coherent Computing
 
-Ask whether hardware, software, interfaces, automation, identity, repair, and service architecture preserve future user capability: inspect, understand, alter, migrate, refuse, recover, and continue.
+Ask whether hardware, software, interfaces, automation, identity, repair, and service architecture preserve future user capability: inspect, understand, alter, migrate, refuse, recover, and continue. Design known failure paths deliberately while keeping unexpected states observable, diagnosable, recoverable, and migratable rather than coupling continuity to one expected operating path.
 
 ### Community Infrastructure
 
@@ -915,7 +937,7 @@ Ask what forms of intelligence can learn and transform without having their goal
 
 ### Coherent World
 
-Represent history as changes in agents' and institutions' future dynamics, not merely as a list of past events. Avoid one scalar coherence score. Let interventions change affordances, feedback, dependencies, trust, capacity, and future action-space.
+Represent history as changes in agents' and institutions' future dynamics, not merely as a list of past events. Avoid one scalar coherence score. Let interventions change affordances, feedback, dependencies, trust, capacity, and future action-space. Agents and institutions can anticipate from partial models; those expectations should be capable of failing, producing surprise, and changing later models and behavior.
 
 ### Museum of Ordinary Life
 
@@ -1048,4 +1070,4 @@ It should remain vulnerable to being split, narrowed, corrected, or discarded if
 - [Conscious Systems — Open Questions](../research/conscious-systems/open-questions.md)
 - [Coherent Systems living paper](../research/papers/coherent-systems/README.md)
 
-**Source note:** The framework was prompted by the 2026-09-24 conversation on personality change, embodied identity, trauma, path dependence, agency, emergence, consciousness, AI, and Root Sequence. The structure and wording are AI-assisted synthesis for author review. External sources were used to anchor selected concepts; no systematic literature review was performed.
+**Source note:** The framework was prompted by the 2026-09-24 conversation on personality change, embodied identity, trauma, path dependence, agency, emergence, consciousness, AI, and Root Sequence. A 2026-10-01 discussion added the operational formulation “Anticipate what you can. Design for what you can’t.” as a derivative of the existing adaptive-continuity, resilience, and corrigibility work rather than a separate concept. The structure and wording are AI-assisted synthesis for author review. External sources were used to anchor selected concepts; no systematic literature review was performed.
diff --git a/systems/adaptation/model.md b/systems/adaptation/model.md
index 845a4355..4ad5da16 100644
--- a/systems/adaptation/model.md
+++ b/systems/adaptation/model.md
@@ -1,7 +1,7 @@
 # Adaptation — Working Analysis Model
 
 **Document role:** Reusable systems model for describing how a system changes its responses, structure, or environment over time  
-**Status:** AI-assisted draft for author review, 2026-09-24  
+**Status:** AI-assisted draft for author review, revised 2026-10-01  
 **Scope:** Cross-domain systems analysis; not a validated universal scientific model and not a clinical model of human behavior  
 **Conceptual parent:** [Dynamic Coherence and Adaptive Continuity](../../concepts/dynamic-coherence.md)
 
@@ -504,6 +504,36 @@ Interventions can themselves become new path dependencies.
 
 ---
 
+## 16A. Anticipation and surprise
+
+Adaptation should not require waiting for every avoidable failure to occur. If evidence, prior incidents, domain knowledge, or the current model make a downstream consequence reasonably foreseeable, include it in the design.
+
+A compact operational rule is:
+
+> **Anticipate what you can. Design for what you can’t.**
+
+Anticipation here does **not** mean treating a forecast as authority or trying to enumerate every possible future. It means asking, before action:
+
+- Which downstream consequences, dependencies, capacity limits, and failure modes are reasonably foreseeable?
+- How might affected people or other system components respond?
+- Which assumptions are carrying the prediction?
+- What information will be needed to tell whether the prediction was wrong?
+- What becomes hard to reverse if the model is mistaken?
+
+Where uncertainty remains, preserve the ability to notice and respond. Depending on the domain, that may mean observability, staged rollout, reversibility, fallback paths, redundancy, repairability, migration, appeal, human review, or enough spare capacity to adapt outside the expected operating envelope.
+
+Unexpected outcomes should become information for model revision rather than evidence that the system was unknowable all along.
+
+There is also a failure mode in the other direction: **over-anticipation** can become analysis paralysis, premature abstraction, preemptive restriction, or brittle optimization around one imagined future.
+
+A useful test is:
+
+> **Did anticipation preserve options and improve preparedness, or did it narrow the system around a prediction that has not happened?**
+
+This overlaps established resilience-engineering work on responding, monitoring, learning, anticipating, graceful extensibility, and sustained adaptability. Root Sequence uses the formulation above as a cross-project design prompt, not as a claim to a new underlying mechanism.
+
+---
+
 ## 17. Evidence and claim status
 
 Label claims explicitly:
@@ -598,6 +628,10 @@ Possible result:
 
     LONG-TERM EFFECT:
 
+    FORESEEABLE DOWNSTREAM EFFECTS / FAILURE MODES:
+
+    IF THE MODEL IS WRONG, WHAT REMAINS OBSERVABLE / REVERSIBLE / RECOVERABLE:
+
     RECOVERY / TRANSFORMATION RELATION:
 
     REVERSIBILITY / EXIT:
@@ -612,6 +646,12 @@ Possible result:
 
 ## 20. Root Sequence principle
 
+> **Anticipate what you can. Design for what you can’t.**
+
+Use available evidence to prepare for foreseeable consequences without pretending the future is fully knowable. Preserve enough observability, reversibility, recovery capacity, and room for revision that surprise does not automatically become catastrophe or lock-in.
+
+Applied to adaptation:
+
 > **Adaptation is not evidence of improvement. Ask what changed, what future responses became more likely, which possibilities opened or closed, who shaped the change, and whether the system can still revise itself when conditions change again.**
 
 This model should make adaptation more specific, not turn every historical effect into one grand theory.