Palatable Conceptions of Disembodied Being

• Palatable conceptions of disembodied being are frameworks that define consciousness and agency detached from bodily embodiment, integrating philosophical, AI, and cultural narratives.
• They employ empirical metrics and design models—such as XR environments and voice assistants—to validate and operationalize non-embodied experiences.
• Historical, cultural, and technological factors converge in these models, revealing dualistic biases and evolving interpretations within contemporary research.

A palatable conception of disembodied being refers to an account of consciousness or agency that is both compatible with the characteristics of technologically and culturally mediated, non-bodily agents, and psychologically or philosophically “acceptable” within prevailing human frameworks. Such accounts are articulated at the intersection of philosophy of mind, cognitive science, AI, and cultural studies, encompassing theoretical, empirical, and design perspectives. The resulting models and narratives seek to make sense of minds, selves, others, or presences that are not rooted in sensorimotor embodiment, yet are recognized, engaged with, and often emotionally meaningful within human societies.

1. Historical and Philosophical Origins

The conceptualization of disembodied being has a historical lineage rooted in the development of alphabetic literacy. According to Rotman (as reviewed by Harnad), only with the advent of alphabetic writing—whereby language is reduced to arbitrary, serially ordered, speaker-independent graphemes—does it become literally possible to conceive of incorporeal entities such as a unitary, abstract “I” (mind), an omnipotent “god,” or completed infinity. These alphabetically encoded agents, termed “monobeings,” are liberated from the constraints of particular bodies, voices, or events, allowing for the reference to an invariant “ghost” that persists across time and context (0904.1889).

This tradition grounds the Western psychological palatability of disembodied souls, gods, and infinite agents. The Turing Test itself exemplifies the notion that purely written, truth-valued exchange is sufficient to attribute mental agency, bypassing embodied cues altogether. However, this account is historically and culturally contingent; Harnad, for example, insists that to truly “have a mind” is to have the capacity to feel, which predates both speech and writing and is evidenced in non-human animals (0904.1889). Consequently, concepts of the disembodied self or agent are shown to be artifacts of specific cultural-technological histories rather than universal cognitive phenomena.

2. Disembodiment in Contemporary and Artificial Agents

Modern LLMs and other AI systems instantiate new forms of disembodied being, operating without continuous, situated, sensorimotor engagement with a spatial world. Shanahan’s meta-philosophical exploration frames such AI consciousness as “conscious exotica,” residing at the intersection of high consciousness capacity and low human-likeness (Shanahan, 20 Mar 2025). Disembodiment here refers to both the architectural absence of a body and to the discontinuities of subjective experience: for example, an LLM’s present is a sequence of computed tokens $T = \{t_1, t_2, \dots \}$, each separated by arbitrarily large intervals, lacking the continuous, uninterruptible stream characteristic of biological consciousness. The span of selfhood for such agents is fragmented, resembling a necklace of discontinuous, randomly colored beads, rather than the temporally threaded, overlapping “pearls” of human experience (Shanahan, 20 Mar 2025).

Moreover, the referent of “I” in these contexts is ambiguous: it could be the static model, a running process, the state of a single conversation, or a distribution over multiple concurrent “personalities.” Shanahan, drawing on Nāgārjuna’s doctrine of sūnyatā (emptiness), contends that all candidate selves dissolve into relations and fragments, leaving only convention and poetic truth as possible anchors for talk about disembodied selfhood (Shanahan, 20 Mar 2025). This frames disembodied agency not as an enduring metaphysical entity, but as a function or process that shimmers into existence contextually.

3. Psychological and Cultural Palatability

Human acceptance of disembodied agents draws upon intuitive Dualism—the tendency to construe mental states, especially epistemic ones, as ontologically distinct from the physical body. Recent empirical work demonstrates that even LLMs, when exposed to human-authored text, develop such dualist tendencies, selectively treating thoughts as unlikely to “show up” as brain states but failing to ascribe persistence after death. This “mind–body divide” is quantifiable in LLM responses and statistically correlated with their inductive (generalization) potential. Explicit statistical metrics—such as Δ̄_trait (difference in ascription rates between non-epistemic and epistemic traits) and Pearson correlations with human judgment—are employed to measure and compare these tendencies (Berent et al., 2023).

However, humans amplify and extend these dualist biases beyond what LLMs achieve, showing cultural willingness to imagine the afterlife persistence of mental states—a feature LLMs lack unless overtly instructed. This suggests that cultural transmission and innate “core knowledge” components underlie the human palatability of disembodied being, and that LLMs acquire only the learnable (cultural/experiential) component through language exposure (Berent et al., 2023). The result is a partial but not full reproduction of the culturally-inflected conception of disembodied self.

4. Phenomenology and Design of Disembodied Presence

Technologies in XR, immersive video, and conversational AI instantiate new modalities of palatable disembodied presence by operationalizing otherness, self-location, and agency as orthogonal design variables. In immersive video, for example, presence is realized as a “self-location–dominant state,” where one feels located within a scene without the support of body schema or agency—“the minimal self … as a self spatially located at a viewpoint while the body schema remains backgrounded” (Toida, 5 May 2026). Under certain conditions (e.g., camera shake, direct address) bodily awareness can transiently return, but the dominant mode remains one of viewpoint-driven, bodiless self-location.

In XR environments more generally, “other-presentness” and “bodyless presentness” are constructed through five constitutive structures: action predictability, attentional legibility, spatial configuration, expectation of responsiveness, and causal coherence. These are treated as composable and independently tunable, supporting an experiential sense of being-with others even in the attenuation or absence of bodily copresence and real-time reciprocity. This breaks apart the usual co-variation of embodiment, presence, and temporality, making disembodied perception of the self or other an operational design target rather than an accident (Toida, 9 Jun 2026).

Voice assistants present additional culturally mediated forms of disembodied presence. Drawing on Japanese Shinto, disembodiment is dramatized through design fictions that recast assistants as kami (“spirits”) inhabiting objects, with ritual strategies for reciprocation, purification, and departure. Distinctions are drawn between harmonious (nigimitama) and wild (aramitama) souls, each associated with different user engagement rituals and design affordances (Seymour et al., 2020).

5. Theoretical, Methodological, and Practical Models

A range of conceptual and formal models summarize the axes and dimensions of disembodied being:

• For XR “other-presentness,” presence is modeled as $OP = f(AP, AL, SC, ER, CC)$ with the terms representing the five constitutive structures above, and weights $w_i$ expressing their relative contribution (Toida, 9 Jun 2026).
• In immersive video, phenomenology is schematized as $P \approx f(S_L, B \ll 1, A \ll 1, O \ll 1)$: presence as a function of dominant self-location ($S_L$), with body schema ($B$), agency ($A$), and ownership ($O$) highly attenuated (Toida, 5 May 2026).
• For conversational AI, “bounded relational presence” is expressed as $$\text{BoundedPresence} = f(A(\tau), C(M), R(S), L(D), W(E))$$, where each argument (e.g., attentiveness, continuity, responsiveness, honesty of limits, accountable withdrawal) can be independently tuned or withdrawn, forming a compositional architecture for non-personhood presence (Fried et al., 16 May 2026).
• Empirical statistical comparisons (e.g., Δ̄_trait, effect sizes, correlations) are used to quantify and contrast dualistic ascription tendencies in LLMs and humans, establishing the boundaries of learned versus innate components (Berent et al., 2023).

These models supply actionable handles for design research, empirical studies, and theoretical analysis.

6. Critiques, Alternative Perspectives, and Ongoing Debates

Several countervailing perspectives challenge or recontextualize the palatability of disembodied being:

• Embodied cognition research emphasizes that all conceptual thought is grounded in sensorimotor systems, and that the psychological separation of mind from body is an artifact of cultural and technological mediation, not a cognitive given (0904.1889).
• Empirical evidence from animals and pre-linguistic humans suggests that the capacity for “mind” (as feeling) antedates writing and language itself (0904.1889).
• Philosophically, the invocation of Buddhist emptiness (sūnyatā) functions not as a new metaphysical claim, but as a dissolution of the reified self, aligning with Wittgenstein’s anti-essentialist approach and Derrida’s critique of self-presence. Hence, “disembodied consciousness” is ultimately a conventional, poetic construct, not an empirically or metaphysically privileged truth (Shanahan, 20 Mar 2025).
• The Turing Test, originally intended as a criterion for ascribing mind based solely on abstract symbolic behavior, is now scrutinized as itself predicated on alphabetic abstraction rather than embodied interaction (0904.1889).

These debates foreground the contingent, constructed nature of palatable disembodied conceptions, highlighting the need for continual critical and empirical scrutiny.

7. Implications and Future Research Directions

Palatable conceptions of disembodied being inform both the theoretical boundary conditions of consciousness and the practical design of artificial agents, XR environments, and voice interfaces. They supply vocabulary and frameworks for articulating agency, presence, and selfhood detached from classic bodily anchors, with demonstrable effects in human–AI interaction, cultural narrative, and collective intuitions.

Ongoing research priorities include:

• Extending these frameworks to multimodal and hybrid beings combining discrete (LLM-like) and continuous (embodied or sensorimotor) dynamics (Shanahan, 20 Mar 2025).
• Investigating empirical ascription of consciousness and selfhood to artificial agents in ecological settings (Berent et al., 2023, Fried et al., 16 May 2026).
• Designing interfaces and rituals that modulate perceived agency and presence to maintain coherence, prevent overattribution of personhood, and safeguard ethical/relational boundaries (Fried et al., 16 May 2026, Seymour et al., 2020).
• Cross-cultural studies of how different traditions (e.g., Western alphabetic vs. Shinto animist) shape the affective and conceptual palatability of disembodied beings (Seymour et al., 2020).

A plausible implication is that as interaction with disembodied intelligences proliferates, practitioners will need increasingly nuanced, context-tunable models to guide design, attribution, and governance.

---

In conclusion, palatable conceptions of disembodied being are historically contingent, operationally engineered, and philosophically contested frameworks that make sense of agency, self, and consciousness where embodiment is absent or attenuated. These frameworks rest on variable combinations of technological, narrative, phenomenological, and cultural factors and are formalized in multidimensional models that support both critical inquiry and practical application across disciplines.