1. Hammond's Cognitive Continuum — Functional Recap

This section does not attempt to recover CCT in its full historical scope, only to fix the functional architecture that Sections 2 and 3 need. Hammond organizes the theory around five premises.

First: modes of cognition — not merely the judgments they produce — can be ordered along a single continuum, with pure intuition at one pole and pure analysis at the other, in direct contrast to the traditional dichotomy between the two. Second: the modes situated between the poles combine elements of both and are termed quasirationality — the empirically most common mode of human judgment, close to what a layperson would call "common sense" and related to Simon's bounded rationality. Third, and central to what follows: the properties of the cognitive task itself — not merely the judge's preference or training — can be ordered along a parallel continuum, according to their capacity to induce intuition, quasirationality, or analysis. Fourth: cognitive activity is not static during task performance — it moves along the continuum; success tends to inhibit that movement, and failure tends to stimulate it, suggesting an adaptive mechanism drawing cognition toward the mode best fitted to the task. Fifth: the capacity to recognize patterns and functional relations in environmental cues is itself a variable capability of the judge, not a fixed trait.

From the conjunction of the second and third premises follows CCT's central thesis, and the reason this article needs it: the functional validity of a judgment — its achievement, in terminology already familiar from this trilogy — is more likely when there is a match between the mode of cognition actually invoked and what the task's structure demands. An analytical judge operating on a task that induces intuition, or an intuitive judge operating on a task that demands explicit analysis, does not fail from general incompetence — it fails from a mismatch between mode and task. This distinction between incompetence and mismatch is what makes CCT operationally useful rather than merely descriptive.

The two poles have reasonably well-documented behavioral markers. Under intuitive cognition, cue processing is fast, simultaneous, and holistic — combining many cues at once, often redundant with one another — and the judge's confidence in the method by which the judgment was reached tends to be lower than their confidence in the judgment itself: they know they are right without being able to fully explain why. Under analytical cognition, processing is sequential, explicit, and decomposable into justifiable steps, the response is slower, and the confidence relationship reverses — the judge trusts the method followed more than necessarily the outcome that method produced. A measurable side effect of this difference appears in the distribution of errors: analytical judgments tend to produce more concentrated (leptokurtic) error distributions than intuitive judgments — more consistent, but also more brittle when the task does not, in fact, lend itself to analytical decomposition.

Quasirationality should not be read as an inferior compromise state, a dilution of analytical purity by intuitive contamination. Hammond treats it as the dominant mode, and often the most appropriate one, for the generality of specialized judgment tasks — clinical diagnosis, weather forecasting, public policy decisions — precisely because most real tasks do not sit at either extreme of the task continuum. It is this same observation, applied to AI-IoT systems, that will motivate Section 5.

Finally, it is worth noting an adjacent distinction that Hammond also formulated, and that resonates — without being identical to — a tension already explored in Article 1: the difference between theories of judgment centered on correspondence (the accuracy of a judgment against an external, ecological criterion — exactly the tradition in which the Lens Model Equation belongs) and theories centered on coherence (the internal logical consistency of the judgment process, independent of its outcome in the world). Article 1 distinguished Popperian falsifiability from epistemological pragmatism in discussing AI evaluation; Hammond's distinction operates at a different level — the psychology of judgment, not the philosophy of science — but the structural echo is notable: in both cases, an internal, formal criterion is separated from an external, consequential one. It is not the same distinction, but two independent traditions stumbling onto the same fault line.

CCT also comes equipped with two quantitative instruments that Section 4 will return to: an index on the task side (the Task Continuum Index, which locates a task on the continuum based on its properties) and an index on the judge's side (the Cognitive Continuum Index, which locates the cognitive mode actually enacted based on markers such as response time and the kurtosis of the error distribution). It is the existence of these two paired indices — one for the task, one for the judge — that makes the correspondence thesis empirically testable, rather than merely a heuristic metaphor.

This section deliberately leaves unenumerated the eleven task properties that operationalize the task continuum — that is the subject of Section 2 — and does not yet attempt any mapping onto the AI-IoT domain, the task of Section 3. It fixes only the architecture: two poles, a dominant and non-degenerate middle ground, a task→cognition induction mechanism, an oscillation dynamic sensitive to success and failure, and two paired indices that make all of this measurable.

(Bibliographic note: the five-premise structure followed here corresponds to the standard synthesis of CCT as documented in secondary literature revisiting Hammond (1996) — namely Dhami & Mumpower (2018), already verified in the trilogy's core bibliography. The behavioral markers of intuition and analysis — response speed, relative confidence in method vs. outcome, kurtosis of the error distribution — follow the formulation of Hammond, K. R., Hamm, R. M., Grassia, J., & Pearson, T. (1987). "Direct comparison of the efficacy of intuitive and analytical cognition in expert judgment." IEEE Transactions on Systems, Man, and Cybernetics, SMC-17, 753–770 — a citation now directly verified against two independent academic sources, suitable for inclusion in the trilogy's core bibliography. The paired Task Continuum Index / Cognitive Continuum Index was likewise confirmed in independent academic literature. The specific attribution of kurtosis as a differentiating marker comes via secondary literature (Dunwoody et al.) not yet verified directly; it remains a point to confirm before formal publication.)