Context Dynamic Algebra, Verified at Runtime: Intention Space as the Execution Model CDA Was Waiting For

Draft skeleton — structural outline with load-bearing claims marked. Sections to be expanded into full prose.

Abstract (to draft last)

State the thesis directly: CDA (Pal, 2020) formalized requirement as a closed dynamical algebra of Attention Phrases, but was explicit that it did not — and structurally could not, from within its own scope — specify how that algebra unfolds during execution. This paper argues Intention Space / CPUX is that missing execution model: a category-theoretic runtime whose objects (execution loci, human and machine) and morphisms (CPUX paths) give CDA's equilibrium/non-equilibrium dynamics an actual place to happen. We show the correspondence is not merely translational — several CDA operators (φ, {}, δ/Δ, μ, Λ, σ, τ, ρ) map onto specific IS structures with checkable, non-trivial properties — and that the mapping surfaces a genuine addition CDA's original one-directional model did not anticipate: execution-time loop-back from machine locus to human locus.

1. CDA's Own Stance: A Specification Waiting to Be Unfolded

• Open with CDA's explicit self-limitation, quoted directly: "the paper does not cover the technical implementation details, nor does it go to the detail treatment of operators in their execution steps."
• Establish that this was a principled exclusion, not an oversight — CDA's stated goal was requirement as "a system... independent of variety of implementation characteristics, at the code, language, and hardware platform."
• Frame the central historical fact: CDA modeled its own dynamics on equilibrium/non-equilibrium (Δ/δ), borrowing the general Wikipedia definition of a dynamical system (state evolving through time via an evolution rule) — but never built or committed to that evolution rule itself. It predicted the shape of execution without specifying the mechanism.
• State the paper's core methodological move: rather than treat this gap as something to patch informally, treat CDA as making implicit, checkable predictions about what any valid execution model of it must look like — then test those predictions against Intention Space.

2. Intention Space / CPUX as the Verifying Runtime

• Introduce the CPUX stack (Perception → Pulse → Signal → Field → FieldBoard → CPUX) at the level needed for this paper, not the full framework.
• Introduce locus of execution as the missing base structure identified in the course of this alignment work — split explicitly into human locus and machine locus, with different structural properties (machine: addressable, persistent, deterministic; human: transient, embodied, inferred rather than queried).
• State the resolved answer on directionality: loop-back is real — human locus does not only precede machine locus; machine-locus execution can generate new human-locus attention (e.g., a system output creating new human intention). This is stated as a design commitment, not left open.
• Introduce CPUX formally as a category: objects = loci (human and machine), morphisms = CPUX paths, composition = the + operator (CPUX1 + CPUX2 = CPUX3), defined only when endpoints match (partial composition), associative, with local identity morphisms (zero-duration CPUX at a locus), and no general inverses — chronological sequence does not reverse.
• Note explicitly what this is not: not a group, not a total monoid. Name the correct classification (category, specifically resembling a path category over a directed graph of loci) and state why getting this classification right matters for anything built on top of it later (proof obligations, tooling, composition guarantees).

3. Operator Correspondences, Reframed as Predictions

Present the mapping not as a dictionary but as: "CDA's operator X, if genuinely unfolding at runtime, predicts IS structure Y should exist and behave like Z." Each entry gets a prediction, a check, and a verdict.

4. Λ: The Central Hinge

• State old and new definitions side by side:
  • 2020: "Λ is the only operator that maps to a computation facility... which shall make that resultant context available in the domain."
  • Revised: "Λ is a morphism realized at a machine locus, composed via CPUX +, whose output may re-enter the requirement space as a new Context — potentially at a different locus than it began, including a human locus."
• This is the paper's actual new contribution, not a restatement: CDA's 2020 model has Λ's output returning to the same domain it started in. The loop-back finding means Λ's output can cross from machine-locus execution back to human-locus attention — something the original one-directional formulation structurally could not express.
• Concrete example to develop: a machine-locus Λ resolves a non-equilibrium point (e.g., a system computation completes) and its output becomes a new human-locus non-equilibrium point (e.g., a notification that redirects human attention) — closing the loop CDA's original text did not model.

5. RM = ρ: A Fully Resolved Correspondence

Document the resolution process itself, since it demonstrates the paper's method (prediction → check → refine) concretely:

1. Initial hypothesis: ρ (set-algebra on Scopes) ↔ RM (per-pair reflection matrix)
2. Apparent conflict: a single-Pulse relabeling example initially suggested RM might instead correspond to φ (name substitution within one context)
3. Resolution: RM's true domain is a whole Signal (Scope), not a single Pulse — confirming ρ, not φ
4. Full characterization reached: RM, for a given reflector object, may invoke any of ρ's five modes (selection, insertion, join, intersection, subtraction) over the Pulses in a Signal, under one invariant — the Response instance [R] carried by each Pulse is never mutated by any mode.
5. State why this invariant matters structurally: it is what makes RM/ρ a coherent family across all five modes rather than five unrelated behaviors — restructuring of which Pulses are present, and how their labels/TVs combine, is permitted; mutation of the Response itself is not, under any mode.

6. Domain as Constrained Path-Space

• State the resolved definition: a CDA Domain is the set of all CPUX paths (locus sequences) that are admissible under the requirement's own constraints — a subcategory (or constrained path-space) of the full CPUX category, not identical to a single locus and not competing with the locus concept.
• Note what this buys the paper: Domain becomes a falsifiable claim — "this requirement permits these execution paths and not others" — checkable against actual CPUX traces, which is precisely the "verification" the paper's title promises, not merely an analogy.

7. What Remains Open (state honestly, do not overclaim)

• τ's correspondence needs a worked numerical example from an actual CPUX trace before being presented as fully closed, not just structurally plausible.
• The human-locus formalization itself (transient, embodied, non-addressable) is asserted here rather than fully specified — a rigorous account of what a human locus is, formally, is future work this paper should flag rather than resolve.
• Whether every RM mode (join, intersection, subtraction) has been observed in an actual reflector object, or only reasoned about structurally, should be stated plainly.

8. Conclusion (to draft last)

Restate the thesis with the loop-back finding as the headline result: CDA predicted a shape of execution it could not itself specify; Intention Space specifies that execution and, in doing so, reveals that CDA's original one-directional model of Λ was incomplete in a way only visible once a genuine runtime was built to test it against.

References

• Pal, P. "The Context Dynamical Algebra: A Pragmatic Model of Computation." April 2020.
• [Intention Space / CPUX foundational references — to be added]