Answers are grounded in this paper and cite their sources.
Eternal inflation generates unbounded spacetime volume and, for event types that recur with nonzero rate, unbounded event counts. Predictions therefore require a measure: a prescription for regulating those counts and taking a limit. But a measure can assign probabilities only after specifying what counts as an event. This paper asks what measure equivalence can establish about that underlying event ontology. We represent a measure proposal by separating four questions: what physically exists andhow occurrences are individuated; which event class is counted; what is operationally accessible relative to a reference structure such as a geodesic or causal patch; and how occurrences are regulated and weighted. This distinction separates disputes about reference classes from the prior question of whether a putative fluctuation or record constitutes an occurrence at all. On the common prediction domains of established global–local dualities, including light-cone-time/causal-patch and scale-factor/fat-geodesic relations, equality of limiting probabilities does not identify the underlying event sets. Connecting accessible structure to ontic structure therefore requires additional physics, which we call a bridge principle. Known dualities already establish that certain local measures reproduce global predictions only with specified, attractor-related initial weightings. We formulate the resulting requirement as an obligation of an ontologically local reading: it must reproduce or replace that weighting.If a constitutive bridge principle also re-individuates events, a second obligation arises—the new event set must recover the duality’s limiting ratios; the established duality does not guarantee this automatically. We then compare two decoherence-based proposals in a common framework: the Bousso–Susskind argument for causal-diamond event ontology and the Boddy–Carroll–Pollack treatment of fluctuations in stationary de Sitter space. Prior work shows that stationary or equilibrium states can nevertheless support nontrivial decoherent histories; our result concerns a narrower, explicitly defined necessary state-change condition for record formation. Within the fixed-Hilbert-space, time-independent unitary setting of the Dyson–Kleban–Susskind Boltzmann-brain problem, an exactly stationary total state fails this condition for every fixed system–environment factorization. Thus, if both decoherence criteria are represented within this record class, their different environments cannot distinguish them; any disagreement must lie outside the class. Exact stationarity cannot arise from a nonstationary total state, and trace-norm distance from the stationary set is conserved. Consequently, within this setting, late-time suppression of record formation after a nonstationary history must be carried by subsystems or coarse-grained observables, not by convergence of the total state to stationarity. As a further utility of the framework, we formulate a Conditional Realization Hierarchy for evaluating what is realized in an infinite cosmology. It separates physical admissibility from dynamical generability, event individuation, nonzero measure support, and the repeated opportunity assumptions needed for an infinite-trial inference. This yields a two-part answer to the title question. Physical admissibility alone does not guarantee realization: infinity cannot supply a production channel that the physics lacks. But where the semiclassical dynamics of an eternally inflating landscape supply a channel with nonzero rate, however suppressed, a global event set contains unboundedly many opportunities, and whether realizations exist—unlike their relative frequency—does not depend on the choice of regulator cutoff. On a global reading with semiclassical individuation, realization of every such channel is therefore the default expectation under the stated sampling idealization, and denying it requires closing a specific gate: an exact prohibition, a dynamical ceiling, a re-individuation of events, or a patch-relative ontology. The last two are the bridge-principle moves analysed in this paper
No takes yet. Share an insight, caveat, or question.
Andrew Hoffmann (2026) studied this question.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: