PART IV: Implications and Future Directions
9.1 Possible Empirical Pathways (Without Claims)
The open questions raised in Section 7.4 point toward concrete, adaptable experimental designs already established in the adjacent literature, rather than toward a new methodology invented for this purpose.
The MEG/behavioral dissociation paradigm used by Mencke, Quiroga-Martinez, Omigie, Michalareas, Schwarzacher, Haumann, Vuust, and Brattico (2021) to separate sensory from cognitive expectations under tonal uncertainty could be extended directly to a structured microtonal, rather than atonal, stimulus set, and further extended to a full PAF configuration incorporating delay and reverberation. The dereverberation-adaptation paradigm of Ivanov, King, Willmore, Walker, and Harper (2022) offers a template for testing whether cortical filtering adapts differently to reverberation presented alongside concurrent microtonal and delay modulation than to reverberation presented alone. The attentional-oscillator estimation method of Kaya, Kotz, and Henry (2023) could be applied before and after extended PAF listening sessions to test whether entrainment flexibility itself is modulated by the field. Absorption could be measured with existing validated instruments, the Absorption in Music Scale (Sandstrom & Russo, 2013) and Vroegh’s (2019) zoning-in/tuning-in framework, administered as a pre-registered outcome measure rather than inferred post hoc.
None of these pathways requires new theoretical machinery. What they require is access to instrumentation, participant pools, and ethical oversight that exist within university and research-institute settings, precisely the kind of collaboration this section is written to invite.
9.2 Interdisciplinary Bridges
Three research communities are natural partners for the empirical pathways above. Auditory and cognitive neuroscience laboratories working on predictive coding and precision weighting in the auditory system, the kind of work represented by Malmierca and colleagues’ ongoing research program (Lao-Rodríguez, Cacciato-Salcedo, & Malmierca, 2026), are positioned to test the physiological claims directly. Music cognition centers built specifically around the predictive-coding-of-music framework, such as the Center for Music in the Brain at Aarhus University (Vuust, Heggli, Friston, & Kringelbach, 2022), have both the theoretical alignment and the experimental infrastructure to test the frequency and temporal axes. Architectural and spatial-audio acoustics groups working on reverberation perception and cortical adaptation to real and simulated rooms (Ivanov et al., 2022) are the relevant partners for the spatial axis.
A fourth, less obvious bridge is to the cross-cultural predictive-coding-of-music program itself (Savage & Fujii, 2022), since any empirical test of PAF’s frequency axis that uses only Western-trained listeners and only 12-TET-relative detuning will underdetermine the model’s claimed relational universality.
9.3 Organizational and Collective Contexts
Beyond individual listening, structured acoustic environments of the kind described here are already used, informally and without theoretical framing, in shared cultural and artistic spaces: sound installations, live immersive performance, and collective listening settings in which a group of listeners occupies the same acoustic field simultaneously.
This study makes no claim about organizational, workplace, or wellbeing outcomes in such contexts, and none should be inferred from it. What the PAF framework does offer to these settings is a vocabulary for describing, with some structural precision, what a shared immersive acoustic field is doing to perceptual organization, a description that can inform artistic and curatorial decisions about collective listening environments without extending into claims this study is not positioned to make.
There is, however, a second and entirely independent body of research that bears on why this vocabulary might matter to an organizational audience specifically, and it deserves to be named precisely, including its own internal disputes, rather than folded silently into the predictive processing account developed in Parts I through III. Attention Restoration Theory (ART) proposes that directed attention, the effortful, voluntary form of attention required to hold focus, resist distraction, and pursue a specific goal, is a limited resource that becomes fatigued with sustained use, and that recovery comes not from further effort but from exposure to a setting capable of holding attention effortlessly, through what Kaplan (1995) calls fascination, within a setting sufficiently rich and coherent to sustain engagement without demanding directed tracking of discrete items.
Two qualifications must be attached to ART before it can be used honestly. First, the theory’s own empirical record is heterogeneous rather than uniform: a systematic review and meta-analysis by Ohly, White, Wheeler, Bethel, Ukoumunne, Nikolaou, and Garside (2016), pooling 31 studies, found reliable restorative effects on some cognitive measures, such as digit span and Trail Making Test B, but not on others, with substantial heterogeneity across studies and continuing uncertainty about which components of attention are actually affected. Second, and more seriously for the specific version of the argument that follows, Kaplan and Berman’s (2010) claim that a single depletable resource is shared between executive functioning and self-regulation draws directly on the ego-depletion resource model of self-control, a model for which a large-scale, preregistered, multi-laboratory replication (Hagger et al., 2016) found essentially no evidence at all across a broad international sample. This paper does not therefore treat the shared-resource mechanism as established; it is cited to show where the organizational argument below currently gets its mechanism from, and to flag that this specific piece of the mechanism is contested at its source, independently of anything this paper adds to it.
ART and the predictive processing framework used throughout this paper are historically separate literatures, developed independently and answering different questions; this paper is not aware of prior work that connects them, and this paragraph should be read as this paper’s own proposed bridge, not as an established finding being reported, built on top of a foundation that is itself only partially secure. The bridge is structural, and can be stated precisely regardless of that uncertainty. What Section 6.3 calls attenuation of segmentation dominance describes a shift away from environments organized around discrete, individually anchored predictions and toward a field organized around sustained, coherent regularities, which resembles Kaplan’s characterization of a restorative setting, rich and coherent extent engaged through fascination rather than effortful direction, closely enough to be worth stating as a hypothesis. The proposal, stated as a hypothesis and nothing stronger, is that a PAF-configured acoustic environment may function as a restorative environment in something like Kaplan’s sense, because its three axes reduce the density of discrete, individually tracked events relative to ordinary listening, in the direction ART associates with recovery from directed attention fatigue, whatever the correct underlying mechanism for that recovery eventually turns out to be.
Testing this would require ART’s own established methodology, and specifically the cognitive measures Ohly et al. (2016) identify as showing the most reliable effects, not a new methodology invented for this paper, nor an assumption that any measure of attention will do: a directed attention fatigue induction task, a period of exposure to a PAF-configured acoustic environment as the intervention condition, and post-exposure measurement on tasks such as digit span and Trail Making Test B, compared against a matched non-restorative or silent control condition. Section 7.4 restates this as a concrete open question that is explicit about resting on two independently uncertain literatures rather than one secure one. Until such a study exists, the organizational relevance of PAF is, honestly, the least secure claim in this paper: a proposed structural analogy stacked on top of a theory with mixed empirical support, using a specific mechanism that a major replication effort has already put in serious doubt.
9.4 Instrumental Framework: The Matrix-Routed Sonic Ecosystem
PAF is not only a theoretical proposal; it is developed alongside a dedicated instrument built to realize its three axes in a live setting: the Matrix-Routed Sonic Ecosystem (MRSE), built around a matrix mixer that allows any output of the system to be repatched to any input in real time.
MRSE organizes signal routing into a fixed domain order, Input, Texture, Space, Output, that loosely tracks the frequency, time, and space axes of this paper, with two qualifications that should be stated rather than smoothed over. First, the Input stage is where a microtonal source is fed into the system, not where microtonality is generated; the routing architecture does not itself produce a tuning system, and the theoretical claims of Section 3 rest on the material chosen at that stage, not on the router. Second, pitch-shifting and granular processing, both discussed as temporal techniques in Section 4, are engineering operations on frequency content and sit within MRSE’s Texture stage for practical reasons, not because they belong exclusively to the temporal domain; this is the same axis overlap Section 4.10 already flags as a feature of the model’s operational categories rather than a claim about strict perceptual separation.
MRSE’s real-time repatchability is a deliberate design choice: PAF is theorized as a dynamically sustained field, not a static preset, and an instrument limited to fixed signal chains would misrepresent that claim in hardware. Beyond that, this section does not claim that any particular routing choice, series versus parallel, one domain order versus another, has itself been shown to produce the perceptual integration PAF proposes; that would overstate what an engineering design decision can establish. What MRSE offers concretely to the research program in Section 7.4 is narrower and more useful: a parameterizable platform on which the frequency, time, and space axes can each be switched on or off independently and combined systematically, which is exactly the factorial manipulation the superadditivity test in Section 6.7 requires. Realizing that test in practice would still require documenting and holding constant the relevant acoustic parameters, delay times and feedback levels, reverberation decay time, direct-to-reverberant ratio, and playback level and configuration, none of which this section specifies.