# Beyond Participation: formal reconstruction and adversarial audit

This is a constructive account of divine action, not an attribution of the complete model to a historical author. It asks which additional agency and knowledge premises can extend the Participation grammar. Model-relative consistency and independence do not prove the metaphysical truth or divine instantiation of those resources.

## Starting theory and scope

Retain Part I’s factive productive causation `C(y,x,q)`, dependence `D(x,y,r)`, qualitative realization `H(x,q,m)`, proper constituency `K(z,x)`, and intrinsic profile `J(x)`. Actuality `A(x)` abbreviates `∃m H(x,e,m)`. Modes, worlds, identity, and background constraints remain disclosed. The nine Part I clauses are G, CP, U, Q, KD, KH, FS, FC, X. See [the original formal architecture](formal-architecture.md).

G sources every actual other’s positive realization. CP connects productive cause to dependence; Q connects source production to corresponding qualitative content. U denies source dependence. KD connects proper constitution to dependence. KH controls actual constituent capacities. FS/FC compare unbounded source and bounded recipient actuality in `e`. X says that the source’s intrinsic actuality is the same across admitted possible created outcomes. For divine action, God does not acquire an intrinsic actuality by creating this world rather than another. This claim is additional; it is not a theorem of having no parts.

The source profile remains conditional; source existence and unlimited intellect are not derived. The domain covers created actualities and their proposed source, not independent divine hypostases or a complete theory of Trinity, incarnation, and grace. A no-creation case means no created actuality, not the denial of uncreated divine distinctions.

The proposed follow-up report named in an earlier brief was not supplied. This reconstruction uses the actual Part I files, consulted primary sources, and current specialist scholarship; no unseen document is presumed to establish its preferred answer.

## Formal vocabulary for intentional action

Use possible-plan indices `p` and action indices `u`. These indices introduce neither independent Platonic Forms nor additional substances. `Exercise` denotes the actual exercise of a power. The proposal interprets intentional action as a real relation grounded in the agent; whether that interpretation can preserve the stronger simplicity claims is disputed.

- `Exercise(g,u,p)`: an actual source activity directed to plan p. Intentional control and rational apprehension are additional relations, not concealed in this predicate.
- `Control(g,u)`: the agent positively controls this actual action. Mere productive causation does not define this relation.
- `Rep(g,p)` and `Supports(r,p)`: the source apprehends a possibility, and reason r supports it. Neither entails unique selection.
- `Term(u,x)`: x is the actual effect of the action.
- `Access(g,u,p)`: awareness of one’s own actual choice, rather than merely awareness of one’s capacity.
- `CausalAccess(g,u,x)`: accurate access to the action’s actual effect. Accuracy is not automatically epistemic warrant.
- `Know(g,x,warrant)`: an epistemic relation interpreted separately from causation, representation, or truthmaking. Its warrant supplied by the agent’s own activity is an added condition, not a value generated by the evaluator.

An action concerns a possible expression without requiring that expression to exist already. The proposed grounding order is the agent’s power and control → actual intentional action → actual created effect. The effect does not supply the source’s actuality or determine its choice. This is explanatory order, not a temporal deliberation in God.

## Bridge families and what is actually added

U, KD, F and X in the interface retrieve earlier commitments; F abbreviates the FS/FC actuality-mode profile. New dimensions are:

| Label | Claim | Independent burden |
|---|---|---|
| M | A, B, and no creation are admitted alternatives. | Non-exhaustion and X do not supply these possibilities. Accessibility is stipulated in the sparse interpretation. |
| I | Possibilities and reasons are intelligibly apprehended. | Unlimited actuality does not establish unlimited or even operative intellect by notation. |
| A | The agent controls its actual action. | This is an agency primitive, not bare indeterminism or causation renamed “choice.” |
| T | The intended plan is realized as the actual effect. | Intellectual possibility alone does not produce realization. Transparent, effective exercise needs warrant. |
| R | Awareness discloses one’s actual choice. | Knowing possibilities or power does not disclose which exercise obtains. |
| CT | The actual action adequately discloses its effect. | Awareness of a choice is not automatically knowledge of every actual result. |
| EP | Operative access has adequate epistemic warrant and yields knowledge. | Accurate access records or covariance are insufficient. EP is not derived merely by labeling an agent perfectly self-knowing. |

The inference is therefore conditional:

```text
U + KD                         ⇒ no proper source constituents
M                              ⇒ the specified creation alternatives
M + I + A                      ⇒ the proposed intelligent control account
M + I + A + T                  ⇒ the agent intentionally produces the effect
U + KD + M + I + A + T + X      ⇒ candidate action account with the same intrinsic source actuality
R + CT + EP                    ⇒ warranted knowledge of the actual effect through one’s own activity
```

I/A/T qualify the last case as knowing one’s intentional productive exercise. They are not necessary premises for every conceivable knowledge of an effect. The interface displays the joint proposed path, not all possible proofs.

If “adequate reflexive access” is defined so strongly that it already includes full warranted knowledge of the actual effect, CT and EP may be absorbed into it. That would be a different encoding of the information, not a new reduction of knowledge to participation. Here R is deliberately narrower, separating disclosure of actual selection from causal transparency and epistemic warrant. Eleven interface commitments are not asserted to be the absolutely smallest metaphysical basis.

## Independent warrant versus conditional sufficiency

KD’s explanatory warrant is that genuinely constituting principles contribute to what the whole actually is. Their contribution is not mere correlation. Brute-unity theories reject that grounding claim; a physical example alone cannot prove it for every metaphysical structure. The relation K is interpreted before D so that the bridge is substantive rather than tautological.

For contingency, the strongest retrieved argument is rational-volitional: necessary orientation to a complete good does not necessitate every optional finite expression through which it could be shared. Aquinas ST I q.19 a.3 distinguishes an end from means unnecessary to its attainment and explicitly invokes self-determination in ad5. This depends on a theory of voluntary power. A necessitating rule of diffusion would defeat the proposed inference and must be challenged, not ignored.

For agency, intentional control is distinguished from merely being the productive source of an outcome. Anselm’s rational rectitude and Scotus’s alternatives at the act of choice illuminate different parts. Neither warrants the whole new account. Actual voluntary exercise is admitted as an irreducible agency resource. A reason can justify A without uniquely necessitating A, provided another admissible gift B is also genuinely available. This proposal does not require a deterministic contrastive explanation as the price of intelligibility.

For knowledge, the operative rationale is that cognition can guide an intentional exercise rather than be acquired by observation of its finished product. Awareness of one’s actual action and adequate disclosure of its efficacy can therefore support knowledge through that action. An artisan analogy supplies a conceptual resource, not proof that an uncreated source adequately knows its own action. The competence, scope, transparency, and warrant needed for the inference remain added conditions.

## The sparse interpretations actually checked

`scripts/beyond-model.mjs` supplies three mutually accessible situations:

| Situation | Actual entities and sourcing | Source intrinsic realization | Action / effect |
|---|---|---|---|
| A | g and finite a; C(g,a,e), D(a,g,e) | H(g,e,∞), no K-parts | plan A, effective a |
| B | g and finite b; C(g,b,e), D(b,g,e) | the same | plan B, effective b |
| N | g alone, no created entity | the same | forbearance, no created effect |

Each situation separately records an actual action, understood plans, supporting reasons, control, awareness of the choice, access to the effect, and knowledge. They are not Boolean “axiom passed” flags. The evaluator tests relation contents and compares source H/K profiles. The fact that these records are assigned does not demonstrate that intentional control or knowledge is philosophically instantiated.

The positive interpretation satisfies the original nine clauses plus the added semantic obligations. This demonstrates that the sparse encoding contains no elementary contradiction of its own. It does not prove genuine metaphysical possibility, absolute divine simplicity, or the disputed ontology of contingent facts about actions.

### Deletion witnesses

| Removed | Concrete change | Retention and limit |
|---|---|---|
| U | Add D(g,a,e) in A. | Other selected obligations remain; grounding direction is reversed. No source part is required for this witness. |
| KD | In A, add another finite sourced b and K(b,a), but not D(a,b,e). | This violates global constitution-grounding while the source remains noncomposite. It does not prove source composition is possible under M + X preserving the full constituent profile. |
| F | Give g finite mode 1 in every situation. | X is preserved; inherited unboundedness is lost. |
| M | Retain only A. | Every original clause survives; no creation alternative is present. |
| I | Retain representation of the actual plan but omit its grasp of the other alternatives. | Control and productive sourcing remain; full apprehension of the relevant alternatives is absent. |
| A | Remove the intentional-control relation while retaining sourced outputs. | A source-level spontaneous process is represented without introducing a causally independent actual mechanism. |
| T | In A, the selected plan is B but effective output is a. | Original ontology and other sparse obligations survive; intended/effective alignment fails. Unlimited actuality alone does not imply infallible execution in this signature. |
| R | Remove access to one’s actual choice. | Possibilities remain represented; no disclosure of the actual action is inferred. |
| CT | Remove causal-access relation. | Selection access remains but its actual effect is not disclosed by that route. |
| X | Add intrinsic H(g,intention-A,1), H(g,intention-B,1), etc., across respective situations. | All original local clauses including U/KD survive; proper constituency remains empty. Intrinsic variation is not itself composition or reception. |
| EP | Retain matching access records but omit the warranted knowledge relation in A. | The evaluator detects failure of the epistemic obligation rather than silently generating knowledge. |

These interpretations check clause separation in a deliberately sparse semantics. The failures of I/A/R/CT are not proofs that a richly specified intelligent perfect agent could instantiate the same deficiencies. Such perfection claims would introduce additional constraints and may change redundancy. A model must be rechecked when stronger premises are added.

### A discovered alternative derivation

Under the scoped assumption that every actual entity other than g is created, M’s no-creation situation contains no proper source constituent. If X preserves the full constituent profile across all admitted worlds, then source noncomposition follows in the other worlds as well. KD is therefore not the only route to that conclusion in the stronger combined model.

This is why the extended KD witness uses a created composite rather than pretending that a composite source can satisfy M and this strong X. The first page’s single-situation KD deletion witness remains valid for its weaker retained theory. The new result is explanatory redundancy relative to stronger assumptions, not an independently warranted shortcut to divine simplicity: M and X themselves are additional commitments.

## Rival explanations and decisive objections

1. **Necessary emanation:** retains infinity, source noncomposition, noncompletion, and all sourcing. Only asserted alternatives block it. The witness does not identify Dionysius with this model.
2. **Impersonal indeterminism:** preserves productive grounding without positing an independently actual selector. Intentional control is genuinely additional. A rationally permitted random outcome is not thereby the agent’s direct intentional choice.
3. **Intrinsic intentional variation:** retains local noncomposition and underivation while rejecting X. This is a serious rival, not an automatic acquired-part theory. A stronger identity-of-act doctrine must explain why it is excluded.
4. **Merely external denomination:** can correctly state “creator of A” but may leave the difference between causing and intentionally settling A unexplained. Denomination is not dismissed as causally idle in every possible ontology.
5. **Action as a real relation:** proposes a controlled action directed to a particular effect without an acquired source constituent. Its central burden is explaining its real contingency without making it a third independent reality or concealing intrinsic determination.
6. **Other knowledge without receptive learning:** a direct cognitive relation can be assigned without R/CT while U remains true. Thus deleting the preferred productive route does not prove ignorance or receptive learning. Its independent warrant is still needed.

The strongest truthmaking objection is this: if g’s intrinsic actuality alone necessitates the complete fact of its willing A, then A may become necessary given effective divine action. The threat is modal collapse, the loss of genuine alternatives to creation; it does not follow automatically from noncomposition. If g alone does not necessitate that fact, what additional reality or primitive fact makes this particular exercise actual? The present model answers: an actual voluntary action, admitted as irreducible. That is a clear explanatory stopping point, not an elimination of the objection. Its compatibility with stronger identity-of-divine-act accounts remains contested.

The strongest knowledge objection is parallel: if the particular intention lies entirely outside the source’s intrinsic cognitive actuality, why does knowing that actuality disclose this actual action rather than every possibility indifferently? R, CT, and EP make the needed access and warrant explicit. They are a proposed active-cognition response, not a consequence of X.

## Where this reconstruction stops

The definition chain is not circular: plans describe possibilities; action and control are separately represented; correspondence between plan and effect connects intentional action to creation; awareness and warrant connect actual action to knowledge. No definition of participation contains freedom or knowledge. Knowledge of possibilities can guide action without being inferred from the actual effect; knowledge of actual exercise is reflexive rather than a premise that creates itself.

Nevertheless, agency and knowledge are not reduced to primitive actuality. The model ends at actual voluntary action and warranted awareness of that action. Calling the topic “divine action,” or describing knowledge as productive, supplies neither the actual choice nor the cognitive adequacy requiring defense.

The result is an audited conditional architecture, with a meaningful rival retaining noncomposition while rejecting X. It does not establish full classical simplicity or omniscience, and it does not offer a historical attribution of the whole system.