A Living Value Theory of Narrative-based Therapies

Abstract

The narrator position is not a literary convenience but a cognitive capacity: the structural operation through which the embedded self — the living agent primarily absorbed in its own life — constructs, sustains, and continuously reallocates relevance within both its own experience and the interrecursive social worlds it must navigate. This article develops the clinical implications. If the narrator-position capacity is the structure through which the embedded self narrates its own experience and social world, then psychiatric and psychotherapeutic disturbance can be understood, with new precision, as failures of specific narrator-position functions: disruptions of the dynamic oscillation between embedded living and narrator-position awareness that constitutes healthy selfhood. The article proposes a structural typology of narrator-position failure organised around the three dimensions that Living Value Theory identifies as constitutive of the narrator position: recursivity level, mediational access, and domain. Failures at different dimensions produce different clinical presentations, and the typology offers diagnostic and therapeutic specificity that existing nosologies lack. Four paradigm cases are developed in detail: depersonalisation as the breakdown of the self-model's transparency (the narrator made visible as a narrator, producing the characteristic experience of watching oneself from outside); confabulation and the unreliable self-narrator (factual, evaluative, and motivational unreliability as three distinct clinical phenomena); the underground-man pathology (L5 recursive intelligence without L1 grounding, producing the paralysis of the self that can critique everything but inhabit nothing); and auditory hallucinations as externalised narrator position (the voice-hearer's experience reread as the self's narrator function operating from an unrecognised external locus). The article then analyses three therapeutic approaches through the LVT narrator-position framework: narrative therapy as deliberate narrator-position restructuring, cognitive behavioural therapy as L4 narrator retraining, and avatar therapy as the most direct clinical deployment of narrator-position shifting in current practice. The article argues that LVT narratology does not replace existing clinical frameworks but provides them with a structural account of what they are doing and why it works, and generates predictions about which interventions should work for which narrator-position failure modes.

I. The Gap That Existing Frameworks Have Not Closed

Clinical psychiatry and psychotherapy have long recognised that something goes wrong with the patient's relationship to their own story. The depressed patient tells a story of worthlessness that does not correspond to the evidence. The psychotic patient tells a story organised around a paranoid logic that no external observer can verify. The patient with borderline personality disorder tells stories that shift dramatically across sessions, unable to maintain a coherent account of who they are or what they want. The traumatised patient either cannot tell the story of what happened or tells it in a way that keeps the past as vivid and immediate as the present. The patient with obsessive-compulsive disorder is trapped in a story of threat and contamination that no amount of disconfirming evidence can dissolve. And the patient with auditory hallucinations hears a voice that narrates their experience from outside, often with authority and malevolence, a narrator they did not choose and cannot silence.

These observations are familiar to every clinician. What is not familiar, because it has not previously been available, is a structural account of what exactly is malfunctioning in each case. The existing frameworks — cognitive behavioural models of distorted thinking, psychodynamic models of unconscious conflict, attachment models of disrupted internal working models, diagnostic nosologies of symptom clusters — each capture something real about these presentations. But none of them offers a unified structural account of the narrator position and its failure modes. The result is a clinical landscape in which different disturbances of self-narration are treated as if they were categorically different phenomena requiring categorically different interventions, when a structural analysis reveals them as specific failure modes of a single underlying operation.

Living Value Theory narratology provides that structural account. The embedded narrator — the living agent primarily absorbed in its own life, with narrator-position awareness available on demand — deploys the narrator-position capacity selectively: escalating from absorbed L1 coordination toward L3 articulation, L4 analysis, and L5 meta-reflection when situations require it, and returning to absorbed embeddedness when they do not. When this dynamic is disrupted, it is disrupted in structured ways corresponding to specific breakdowns of specific narrator-position operations. The narrator position operates at L3 minimum, characteristically shifting between L3, L4, and L5, and faces radically different challenges across the five mediational domains and the three recursivity domains (non-recursive, self-recursive, interrecursive). A structural typology of these failures is both theoretically precise and clinically actionable. It identifies what has gone wrong at the level of the narrator position's constitutive operations, and it identifies which interventions address which specific failure modes.

II. A Structural Typology of Narrator-Position Failure

The narrator position can fail in three distinct dimensions, corresponding to the three analytical frameworks that the series has applied to it. Recursivity-level failures occur when the narrator is locked at a specific level and cannot perform the recursive fluidity that healthy self-narration requires. Mediational-access failures occur when the narrator's access to specific mediational dimensions of experience is blocked, distorted, or unmanageable. Domain failures occur when the narrator misidentifies the recursivity domain of a situation, applying non-recursive narration to interrecursive processes or treating self-recursive processes as if they were interrecursive. Each dimension of failure produces a different clinical presentation, and the same diagnostic label can often contain multiple distinct narrator-position failure modes that require different therapeutic responses.

Recursivity-Level Failures

The healthy embedded narrator moves fluidly between L3, L4, and L5, escalating when the situation requires analysis or meta-reflection and descending to L3 and the absorbed coordination of L1 when the situation is resolved. Recursivity-level failures lock the narrator at a specific level, disrupting this fluidity.

L3 locking is the narrator that can only report, cannot evaluate, and cannot step outside its own perspective. The patient who describes their experience in relentless detail without ever arriving at interpretation or evaluation may be exhibiting L3 locking. In trauma presentations, the inability to move from episodic L3 description of traumatic events to L4 integration and L5 meaning-making is a specific recursivity failure: the event is narrated over and over at L3 without the narrator position being able to achieve the retrospective distance from which L4 pattern-making and L5 meaning-generation become possible. Trauma therapy, on this account, is partly the work of facilitating the upward movement from L3 episodic narration toward L4 and L5 self-recursive processing — a movement that the traumatic material has blocked by its very intensity.

L4 over-dominance is the narrator that has abstracted so thoroughly from its own experience that it can no longer access the L2 felt misalignment and L3 particular experience from which valid abstraction must draw. The intellectualising patient who produces sophisticated L4 analysis of their difficulties without ever making contact with the L2 felt reality underlying them exemplifies this failure. The L4 abstractions are not wrong in their own terms; they are disconnected from the mesocosmic coordination that would give them life. Much of what psychoanalysis identified as intellectualisation as a defence mechanism is, in LVT terms, L4 over-dominance: the narrator position has retreated to L4 as a way of avoiding the L2 felt misalignment that honest L3 narration would require.

L5 without L1 grounding is the most philosophically interesting recursivity failure and deserves its own section below. It is the narrator that can position itself outside any framework, critique any stance, recognise the contingency of any commitment, but that cannot descend from this meta-recursive position back into absorbed coordination at L1. The result is paralysis rather than insight: the L5 capacity that should enable recursive fluidity has become a trap. This is not a failure of intelligence but a failure of the downward movement — the capacity to return from narrator-position awareness back to embedded living — that defines healthy recursive fluidity.

Mediational-Access Failures

The narrator must construct and maintain access across all five mediational dimensions of the absent world it narrates — whether that absent world is a fictional storyworld or the narrator's own past experience. Mediational-access failures occur when specific mediational dimensions are unavailable, overwhelming, or distorted.

Embodied-access failure is among the most clinically significant. The patient with alexithymia — difficulty identifying and describing emotional states — exhibits a specific failure of the narrator's access to the embodied mediational dimension of their own experience. They can narrate the external events of their life with reasonable accuracy (non-recursive facts about dwelling and multimateriality are accessible) but cannot access the embodied felt states that would give those events their emotional significance. The narrator can report what happened but not what it felt like from inside. This is the clinical version of external focalization applied to the self: a narrator that observes its own behaviour without access to its inner life.

Being-with access failure takes several forms. The paranoid patient has systematic distortion in their narrator's access to interrecursive processes: they misread others' stake structures, misidentify others' intentions, and produce unreliable narration of social situations because their interrecursive modeling is systematically skewed toward threat-detection. The patient with social anxiety has a different being-with access failure: their narrator's access to interrecursive processes is overwhelmed by the magnitude of the stakes they assign to social evaluation, producing distorted narration of social situations as more threatening and more consequential than external evidence supports. In both cases, the being-with dimension of the narrator's access is compromised, not the narrator's access to non-recursive or self-recursive dimensions.

Domain Failures: Misidentification of Recursivity Domain

The third failure dimension is perhaps the most clinically underrecognised: the narrator that misidentifies the recursivity domain of the situation it is narrating. Non-recursive processes are those whose behaviour does not change in response to being observed or theorised. Self-recursive processes are those in which the agent is an object of its own influence. Interrecursive processes are those in which agents mutually model each other. Healthy self-narration requires the ability to correctly identify which domain a situation belongs to, because different domains require structurally different narrator responses.

The most consequential form of domain misidentification in clinical practice is the treatment of self-recursive processes as if they were non-recursive. The patient who says "I am depressed" rather than "depression has been working on me in specific ways" is treating a self-recursive process (their own developing relationship with a pattern of experience that their own narration partly constitutes) as if it were a non-recursive fact (an external, stable property of the world that does not respond to being theorised differently). The self-recursive process is genuinely affected by how it is narrated. If the narrator says "I am depressed" as a non-recursive fact, the narration contributes to the stabilisation of the depressive pattern. If the narrator says "depression has been claiming my mornings lately," the narration creates distance that enables a different relationship to the pattern. This is not mere semantics. It is the recognition that self-recursive processes are ontologically different from non-recursive ones: they change in response to being narrated, and therefore the narration is a constitutive intervention in the process rather than a neutral report on it.

The reverse misidentification — treating interrecursive processes as if they were non-recursive — is equally consequential. The patient who describes their social relationships in non-recursive terms, as if the dynamics of those relationships were fixed properties that do not change in response to their own behaviour and narration, is failing to recognise the interrecursive character of the social field. Therapeutic interventions that help patients recognise the interrecursive character of their social world — that others' behaviour is partly constituted by their modeling of the patient's behaviour, and vice versa — are, in LVT terms, interventions in domain misidentification: they are helping the patient apply interrecursive narration where non-recursive narration has been wrongly operating.

III. Four Paradigm Cases

Depersonalisation: The Narrator Made Visible

Depersonalisation is characterised by the experience of watching oneself from outside: the patient observes their own thoughts, feelings, and actions as if they were watching a stranger, or as if looking at the world through a pane of glass. The experience is often described as an absence of felt reality: "I know I am here but I cannot feel that I am here." It is the most directly narratological of all psychiatric presentations, because what it describes is precisely the breakdown of the narrator position's most fundamental feature: its transparency.

Thomas Metzinger's phenomenological account identified this precisely. The healthy self-model is transparent: the organism looks through it to the world rather than at it as a model. Depersonalisation is the experience of the self-model becoming opaque: the organism suddenly sees the model as a model rather than as reality. The normally invisible narrator position becomes visible as a position. The patient is not experiencing enhanced self-awareness in any productive sense; they are experiencing the breakdown of the structural condition that makes selfhood feel like first-person reality rather than third-person observation. In LVT terms, the narrator horizon — the degree of narrator-position awareness available at any given moment — has become fixed in the expanded position. The embedded narrator cannot contract back into absorbed embeddedness; it is permanently in a state of watching rather than living.

The LVT account adds precision to Metzinger's analysis. The depersonalised narrator position is one that has involuntarily shifted to external focalization of the self: it observes itself from the outside without access to its inner states. In Genette's typology, this is the shift from internal focalization (the narrator knowing what the character knows from inside) to external focalization (the narrator observing the character from outside without access to interiority). The depersonalised patient is internally focalised in the sense of continuing to have experiences; they are externally focalised in the sense that their narrator position cannot access those experiences from inside them. The result is the characteristic dissociation between knowing that one is having experiences and feeling those experiences as one's own.

The therapeutic implication is specific. Interventions for depersonalisation that work by reducing the patient's anxious attention to the experience of depersonalisation — the CBT approach of reducing hypervigilance to depersonalisation symptoms — are, in LVT terms, interventions that restore the transparency of the self-model by reducing the self-monitoring that makes the model visible as a model. They work not by changing the content of the self-narrator's narration but by restoring the structural condition (transparency, absorbed non-self-monitoring) under which the narrator position operates without becoming visible to itself. Grounding techniques that direct attention toward embodied L1 experience — sensory engagement with the immediate environment, physical movement, contact with temperature and texture — are attempts to reconnect the externally focalised narrator with the L1 and L2 embodied intelligence from which it has become detached, and to restore the narrator horizon's capacity to contract.

The Underground Man: L5 Without L1 Grounding

Dostoevsky's Notes from Underground gave the Western literary tradition its first sustained portrait of a specific and clinically recognisable pathology: the self that can critique everything but inhabit nothing. The underground man is a narrator of extraordinary intelligence and devastating self-awareness. He sees through every social convention, every comfortable assumption, every form of self-deception available to ordinary consciousness. His L5 meta-recursive intelligence is formidable. And it is precisely this intelligence that produces his paralysis, because his capacity for meta-recursive repositioning has outrun his capacity for absorbed coordination at L1. He cannot act, because every proposed action is immediately subjected to the L5 critique that reveals its contingency, its pretension, its susceptibility to the same irony that has dissolved every previous commitment. The result is a narrator position that is everywhere and nowhere: capable of occupying any meta-position and incapable of inhabiting any L1 ground.

This is the L5-without-L1-grounding pathology: the embedded narrator whose narrator horizon has expanded to L5 and lost the capacity for the downward movement back to absorbed coordination. LVT's concept of recursive fluidity identifies this as a specific failure mode rather than a character deficiency. Recursive fluidity is the capacity to move between recursivity levels as the situation demands: escalating to L5 when frameworks need to be questioned, descending to L1 when coordination can proceed smoothly. The underground man's failure is not insufficient intelligence. It is the loss of the downward movement: he can escalate but cannot descend. The narrator horizon is stuck at its maximum extent.

Clinically, this pathology appears in various forms across diagnostic categories. The highly intellectually capable patient with chronic depression who can analyse their depression with great sophistication but cannot achieve the absorbed engagement with daily life that would constitute recovery exhibits this structure. The patient with obsessive-compulsive disorder whose obsessive thoughts function as an involuntary L5 commentary on every proposed action, making it impossible to act without first resolving a doubt that cannot be resolved, shows the same pattern. The highly educated patient with social anxiety whose meta-awareness of their own social performance makes natural spontaneous engagement impossible is living the underground man's specific torture.

The therapeutic implication here is the most demanding: the intervention must restore the downward movement, the capacity to return from L5 meta-reflection to L1 absorbed coordination. Paradoxically, this often requires not more analysis but less: therapeutic approaches that cultivate L1 embodied engagement rather than L4 and L5 insight. Mindfulness-based interventions work partly through this mechanism: they train the capacity to maintain absorbed present-tense L1 attention without the narrator's L5 commentary overriding it. Behavioural activation in depression works by insisting on L1 engagement with the world before the L4 and L5 narrator has achieved the insight it keeps promising but never delivering. These are not anti-intellectual interventions. They are interventions that restore the recursive fluidity the underground man has lost: not by removing L5 capacity but by rebuilding the L1 foundation from which L5 intelligence can be deployed and to which it can return.

Confabulation and the Unreliable Self-Narrator

Booth's distinction between reliable and unreliable narrators, developed in the context of literary fiction, maps onto a family of clinical phenomena that existing nosology groups under heterogeneous headings but that LVT narratology reveals as variations of a single structure. An unreliable narrator is one whose values, perceptions, or reports diverge from the implied author's (or, in the clinical case, from what external evidence establishes). Booth identified three distinct modes of unreliability: unreliability of fact (the narrator is wrong about what happened), unreliability of evaluation (the narrator misjudges the significance of events), and unreliability of motivation (the narrator does not know why it does what it does). These three modes correspond to three distinct clinical phenomena with different aetiologies and different therapeutic implications.

Factual unreliability in self-narration is documented most directly in the neuropsychological literature on confabulation, the generation of false memories without conscious intent to deceive. The confabulating patient narrates a confident account of what happened that does not correspond to any actual event. But factual unreliability is also present, in less dramatic form, in the reconstructive distortions of ordinary autobiographical memory: the tendency to remember past experience in ways that fit current self-narrative, to fill in gaps with schema-consistent detail, to mis-remember the emotional valence of events in ways that confirm present attitudes. Elizabeth Loftus's research on memory reconstruction documents this comprehensively. The self-narrator is constitutively a reconstructive narrator, not a recording device, and its factual reliability is always partial and always shaped by the narrative framework within which it operates.

Evaluative unreliability is the clinical territory most directly addressed by cognitive behavioural therapy. The depressed patient who evaluates neutral events as evidence of worthlessness, who reads ambiguous social signals as rejection, who assigns catastrophic significance to minor setbacks: these are all evaluative unreliability at the L4 level of the narrator's abstraction and interpretation. The automatic thoughts that CBT identifies and targets are the narrator's L4 evaluative operations, stabilised into habitual patterns that apply the same evaluation regardless of the specific evidence each situation provides. The CBT intervention of challenging automatic thoughts is, in LVT terms, an intervention in the narrator's L4 evaluative function: it attempts to restore the narrator's capacity to evaluate specific situations on their specific evidence rather than through the pre-formed evaluative schemas that habitual L4 patterns impose.

Motivational unreliability — the narrator that does not know why it does what it does — is the clinical territory most directly addressed by psychodynamic approaches. The patient who cannot account for patterns in their own behaviour, who is surprised by their own reactions, who acts in ways that contradict their stated values: all exhibit motivational unreliability. The psychodynamic concept of the unconscious is one account of what produces this unreliability. The LVT account is different: it locates the unreliability in the gap between the narrator's L3 and L4 symbolic self-account and the L1 and L2 pre-symbolic processes that actually drive the organism's behaviour. The narrator reports what it can symbolically access, but the organism's L1 absorbed coordinations and L2 felt orientations often operate independently of and sometimes contrary to the narrator's L4 account of its own motivations. Motivational unreliability is the narrator's constitutive limitation when the organism's most consequential coordinations are operating at levels below the narrator's symbolic access threshold.

Auditory Hallucinations: The Externalised Narrator Position

The LVT account of auditory hallucinations in psychosis is the most structurally original clinical application of the narrator-position framework, and it connects directly to Julian Jaynes's historical argument about the bicameral mind — though without endorsing Jaynes's neurological claims. What Jaynes identified, in the register of a hypothesis about ancient literary conventions, is a structural possibility that is clinically real regardless of its historical status: the narrator position can be experienced as external rather than internal, as a voice coming from outside rather than as the self's own narrating capacity. Whether or not archaic Greeks were genuinely bicameral, the possibility Jaynes describes is clinically present: the narrator-position capacity operating from an unrecognised external locus is precisely what the voice-hearer experiences.

The hallucinated voice performs exactly the functions of the narrator position: it attributes stakes, evaluates the patient's behaviour, allocates relevance within the patient's social field, and provides a governing perspective from which the patient's experience is organised and interpreted. What makes it pathological is not its functional structure, which is identical to the healthy narrator's structure, but its locus and its relationship to the patient's agency over it. The healthy embedded narrator experiences the narrator-position capacity as its own: the narrating is felt as self-narrating, even if the content of the narration is partly unknown to consciousness. The hallucinated voice is experienced as other: a narrator whose authority the voice-hearer is subject to but did not choose and cannot dismiss.

This reframing has significant therapeutic implications, and they are precisely the implications that avatar therapy has arrived at empirically without having had the theoretical framework to name them. Julian Leff's avatar therapy (2008, randomised trial 2018, multisite AVATAR2 trial 2024) asks the patient to embody the hallucinated voice through a digital avatar, voiced by the therapist, with the patient engaging in direct dialogue with it. The therapy works, on the LVT account, because it achieves exactly what the LVT framework identifies as the necessary therapeutic move: it shifts the patient's relationship to the hallucinated narrator from passive recipient to active interlocutor, and over the course of therapy to something approaching the position of an embedded narrator who owns the narrating rather than being subject to it.

The specific mechanism is narrator-position shifting. The patient who has been internally focalised through the hallucinated voice — inhabiting the world as the voice describes it, experiencing themselves as the voice characterises them — is trained through avatar therapy to achieve external focalization of the voice: to observe it as a narrator position rather than being occupied by it. This is the movement that allows the embedded narrator to reclaim ownership of the narrator-position capacity that the voice has been occupying. The avatar provides the material and interactional support that makes this shifting possible for patients whose own resources to achieve it have been exhausted by the chronicity and authority of the voice.

IV. Three Therapeutic Approaches Through the LVT Lens

Narrative Therapy: Deliberate Narrator-Position Restructuring

Michael White and David Epston's narrative therapy, developed through the 1980s and formalised in Narrative Means to Therapeutic Ends (1990), is the clinical tradition most explicitly organised around the narrator position, though it does not use that language. White and Epston's core technique of externalisation — moving from "I am depressed" to "depression has been colonising my mornings lately" — is, in LVT terms, a domain-reclassification intervention: it moves the narrator's account of the problem from non-recursive (an external fixed property of the world that is simply there) to self-recursive (a pattern that the narrator's own account partly constitutes and that can therefore be narrated differently with genuine effects on the pattern itself).

The externalisation technique works precisely because self-recursive processes are not fixed by the narration of them but are partly constituted by it. When the depressed patient narrates their depression as "I am depressed," the narration stabilises the self-recursive pattern by making it a property of the self rather than a process the self is engaged in. When the narrator shifts to "depression has been doing this to me lately," the same pattern is placed in a different recursive relationship to the narrator: the narrator becomes an agent in relation to the pattern rather than an expression of it. This shift is not merely linguistic. It is a genuine restructuring of the narrator position's relationship to a self-recursive process.

White's re-authoring conversations — the therapeutic work of identifying "unique outcomes," moments in the patient's experience that contradict the problem-saturated dominant narrative, and building them into an alternative narrative — are, in LVT terms, relevance reallocation interventions. The dominant narrative is a narrator position that has stabilised a particular relevance hierarchy: certain events are highly relevant (evidence of worthlessness, failure, contamination) and others are not relevant (evidence of competence, connection, resilience). Re-authoring rebalances the relevance hierarchy by drawing the patient's attention to the evidence that the dominant narrator position has rendered invisible. The therapist in narrative therapy is functioning as a co-narrator, helping the patient to reallocate relevance within the absent world of their own past experience in ways that make a different future possible.

Cognitive Behavioural Therapy: L4 Narrator Retraining

Cognitive behavioural therapy does not describe itself in narratological terms, but the LVT framework reveals its interventions as precisely targeted at the narrator's L4 evaluative function. The automatic thoughts that CBT identifies are the narrator's habitual L4 evaluations: stabilised patterns of abstraction and interpretation that apply the same evaluation to situations regardless of their specific evidence. The cognitive restructuring that CBT practises is narrator retraining at L4: it teaches the patient to examine the L4 evaluative moves their narrator makes automatically, to check those moves against specific evidence, and to develop alternative L4 evaluations that fit the evidence better.

The Socratic questioning technique at the heart of CBT is a structured procedure for making the narrator's L4 evaluations visible and available for examination. It achieves at L4 exactly what mindfulness achieves at L1: it moves habitual, automatic processing into deliberate, examinable awareness. The automatic thought "nobody likes me" is a narrator operating habitually at L4 without checking its abstraction against specific L3 evidence. The CBT intervention of asking "what is the evidence for this?" is the therapist facilitating the narrator's descent from L4 abstraction back to L3 specific evidence, and the construction of a better-evidenced L4 from that L3 base.

What CBT does not address, and what the LVT framework clarifies, is the recursivity-level failure that often underlies the L4 distortions it targets. The reason the depressed narrator's L4 evaluations are persistently negative is often not that it lacks the cognitive tools to evaluate evidence properly, but that its L2 felt misalignment is operating independently of the L4 corrections the therapy achieves. The patient learns cognitively that they are not worthless, but the L2 felt sense of worthlessness persists because it is embedded at a level of embodied coordination that L4 retraining does not directly reach. This is why CBT often produces good short-term outcomes but variable long-term maintenance: the L4 narrator has been retrained, but the L1 and L2 ground from which the distorted L4 originally arose has not been addressed. Integrated approaches that combine CBT's L4 retraining with interventions aimed at the embodied L1 and L2 levels (behavioural activation, somatic approaches, mindfulness) address the full recursivity range and produce more durable results.

Avatar Therapy: The Purest Clinical Narrator-Position Intervention

Avatar therapy is the most direct clinical deployment of narrator-position shifting in current psychiatric practice, and its mechanism is most clearly illuminated by the LVT framework. The randomised controlled trial of AVATAR therapy (Craig et al., 2018, Lancet Psychiatry) demonstrated significant reduction in auditory hallucination severity compared with supportive counselling at twelve weeks. The larger multisite AVATAR2 trial (2024) has confirmed these results across multiple clinical settings. The therapy works. The question is why, and the LVT account is more precise than the existing mechanistic accounts.

The existing mechanistic account holds that avatar therapy works by giving patients a sense of greater power and control over the voice through direct dialogue. This is accurate but incomplete. The LVT account specifies the structural mechanism: avatar therapy works by training the patient to shift from being the recipient of the hallucinated narrator's perspective (internal focalization through the voice) to being an interlocutor with it (the voice as an entity the patient can address rather than a perspective they are addressed by). Over the course of therapy, the avatar's characterisation of the patient becomes progressively less hostile and the patient progressively more assertive, which in LVT terms is the gradual movement from a narrator-position in which the hallucinated voice has zero-focalization authority over the patient's experience toward one in which the patient's own embedded narrator position can evaluate and contest the voice's characterisation.

The therapeutic frame in which this movement occurs is also precisely the metabolic buffering condition identified in the companion articles of this series as the precondition for safe narrator-position shifting. The avatar therapy session provides the protected co-present space within which the patient can engage with the voice's narrator position without the immediate stakes of the untherapied encounter: the therapist is present, the avatar's responses are controlled, the session is time-limited, and the patient knows that the avatar is a digital construction voiced by the therapist. This buffering is what makes it possible for the patient to inhabit the voice's perspective long enough to evaluate it, which is what chronic untherapied hallucination does not allow: the patient is too continuously inside the voice's perspective to achieve the external focalization from which evaluation and ultimately resistance become possible.

Avatar therapy is, in this account, the most concentrated therapeutic deployment of a narrator-position capacity that all embedded narrators possess: the capacity to inhabit a perspective and simultaneously evaluate it from outside, to be inside a voice and see it as a voice. Sophisticated literary fiction makes this capacity more explicitly available for cultivation in ordinary readers who encounter unreliable narrators and learn to read against the narrator's grain. Avatar therapy provides, in a concentrated and structured therapeutic context, the conditions under which a capacity that chronic hallucination has blocked in a specific domain can be reactivated and retrained. The great difference between the literary reader and the voice-hearer is not in the underlying capacity — both are embedded narrators with the same structural resources — but in the specific disruption that blocks deployment of that capacity in the voice-hearer's case, and in the specific therapeutic context that avatar therapy provides to restore it.

V. What LVT Narratology Adds to Clinical Practice

The four paradigm cases and three therapeutic analyses above demonstrate what the LVT narrator-position framework contributes to clinical understanding that existing frameworks do not provide. It is worth stating these contributions directly.

First, the framework provides a unified structural account of diverse presentations. Depersonalisation, confabulation, L5-without-L1 paralysis, and auditory hallucinations are treated by existing nosology as categorically different phenomena. The LVT framework reveals them as failure modes of a single underlying dynamic: the embedded narrator's oscillation between absorbed living and narrator-position awareness, and the specific ways in which that oscillation can be disrupted. This does not mean they require identical interventions. It means their differences can be precisely specified in terms of which dimension of the narrator position has failed and in what way. Diagnostic precision requires structural accounts, not just symptom catalogues.

Second, the framework generates specific predictions about which interventions should work for which failure modes. If L5-without-L1 paralysis is the underground-man pathology, then interventions that add more L4 and L5 analysis will not help and may worsen the condition: what is needed is L1 re-grounding, not more meta-reflection. If evaluative unreliability at L4 is the CBT target, then the limit of CBT is the boundary between L4 and L1-L2, and interventions that reach below that boundary should produce better long-term maintenance. If domain misidentification — treating self-recursive processes as non-recursive — is the maintaining mechanism of depression's stability, then the re-authoring technique's effectiveness is predictable on structural grounds, not just empirically discovered. If avatar therapy works by training narrator-position shifting from internal to external focalization of the voice, then its effectiveness should be positively correlated with patients' general capacity for narrator-position flexibility (perhaps measurable through their engagement with fiction and their capacity for complex Theory of Mind tasks) and should be less effective when the hallucinated voice is too undifferentiated to be addressed as a distinct narrator position.

Third, the framework clarifies why the therapeutic relationship itself is clinically active rather than merely a vehicle for technique delivery. The therapeutic relationship is an interrecursive process, and interrecursive processes are constitutively changed by the narration of them. The therapist who co-narrates the patient's experience with the patient — who helps the patient identify unique outcomes, examine automatic thoughts, dialogue with their voices — is a co-narrator, and the co-narrating relationship is itself an intervention in the patient's narrator position. The presence of a trusted other who witnesses and helps to reallocate relevance within the patient's self-narrative is not merely supportive; it is structurally active at exactly the level the LVT framework identifies as central: the being-with dimension of the narrator's mediational access, the interrecursive domain that the patient's own isolated narrator position cannot adequately navigate without a co-present interlocutor.

Fourth, and most importantly, the framework grounds the existing clinical traditions in a structural account that explains not merely what they do but why it works. Narrative therapy externalises because self-recursive processes change in response to narration. CBT challenges automatic thoughts because L4 evaluative schemas can be retrained through deliberate examination of the evidence they abstract from. Avatar therapy shifts narrator position because the voice-hearer can learn to occupy the evaluative distance from the hallucinated narrator that chronic untherapied encounter has prevented. Each of these explanations is more precise than the accounts the traditions themselves have provided, and more precise explanations generate more precise interventions.

The narrator who cannot narrate is not suffering from a mysterious failure of psychological machinery whose mechanisms we can only empirically correlate with symptoms. The narrator who cannot narrate is an embedded narrator whose dynamic has been disrupted at a specific and identifiable point: a narrator horizon that has become fixed in the expanded position, or blocked from expanding when needed, or externalised onto a voice that the embedded narrator did not choose. Identifying the disruption with structural precision is the precondition for therapeutic interventions that are targeted rather than generic, graduated rather than one-size-fits-all, and grounded in an understanding of the embedded narrator's dynamic rather than in the empirical observation that some techniques help some patients some of the time. That is the clinical contribution of LVT narratology. It does not replace the accumulated empirical and clinical wisdom of existing psychiatric and psychotherapeutic practice. It provides that wisdom with the structural account it has lacked.

References

Aristotle. Poetics. Translated by Stephen Halliwell. Cambridge, MA: Harvard University Press, 1995.

Beck, Aaron T. Cognitive Therapy and the Emotional Disorders. New York: International Universities Press, 1976.

Booth, Wayne C. The Rhetoric of Fiction. Chicago: University of Chicago Press, 1961.

Cohn, Dorrit. Transparent Minds: Narrative Modes for Presenting Consciousness in Fiction. Princeton: Princeton University Press, 1978.

Craig, Tom K. J., Mar Rus-Calafell, Thomas Ward, Julian P. Leff, Mark Huckvale, Elizabeth Howarth, Richard Emsley, and Philippa A. Garety. "AVATAR therapy for auditory verbal hallucinations in people with psychosis: a single-blind, randomised controlled trial." The Lancet Psychiatry 5, no. 1 (2018): 31-40.

Dostoevsky, Fyodor. Notes from Underground. Translated by Richard Pevear and Larissa Volokhonsky. New York: Vintage, 1993 [1864].

Ecks, Stefan. Living Worth: Value and Values in Global Pharmaceutical Markets. Durham: Duke University Press, 2022.

Ecks, Stefan. "Living Value Theory: Core Principles." Living Value Theory, livingvaluetheory.org, 2026.

Ecks, Stefan. "Who Is I? A Living Value Theory of Narrating the Self." Living Value Theory, livingvaluetheory.org, 2026.

Ecks, Stefan. "The Embedded Narrator: Towards a Transhistorical and Transcultural Account of Healthy Selfhood." Living Value Theory, livingvaluetheory.org, 2026.

Ecks, Stefan. "The Inner Turn: A History of the Narrator Position from Gilgamesh to the Postmodern." Living Value Theory, livingvaluetheory.org, 2026.

Ecks, Stefan. "What the Narrator Knows and Why: Recursivity Levels, Five Mediations, and Three Domains." Living Value Theory, livingvaluetheory.org, 2026.

Ecks, Stefan. "Storytime: Why We Enjoy Sharing Narratives So Much." Living Value Theory, livingvaluetheory.org, 2026.

Ecks, Stefan. "What Narrators Do That Maps Cannot: Narrative as Reference to Absent Inter-Recursive Worlds." Living Value Theory, livingvaluetheory.org, 2026.

Fonagy, Peter, George Gergely, Elliot L. Jurist, and Mary Target. Affect Regulation, Mentalization, and the Development of the Self. New York: Other Press, 2002.

Genette, Gerard. Narrative Discourse: An Essay in Method. Translated by Jane E. Lewin. Ithaca: Cornell University Press, 1980 [1972].

Hunter, Elaine C. M., Mauricio Sierra, and Anthony S. David. "The epidemiology of depersonalisation and derealisation: a systematic review." Social Psychiatry and Psychiatric Epidemiology 39, no. 1 (2004): 9-18.

Jaynes, Julian. The Origin of Consciousness in the Breakdown of the Bicameral Mind. Boston: Houghton Mifflin, 1976.

Leff, Julian, Lynda Williams, Mark Huckvale, Michael Arbuthnot, and Alexander P. Leff. "Computer-assisted therapy for medication-resistant auditory hallucinations: proof-of-concept study." British Journal of Psychiatry 202, no. 6 (2013): 428-433.

Loftus, Elizabeth F. "The reality of repressed memories." American Psychologist 48, no. 5 (1993): 518-537.

McAdams, Dan P. The Stories We Live By: Personal Myths and the Making of the Self. New York: Guilford Press, 1993.

Metzinger, Thomas. Being No One: The Self-Model Theory of Subjectivity. Cambridge, MA: MIT Press, 2003.

Metzinger, Thomas. The Ego Tunnel: The Science of the Mind and the Myth of the Self. New York: Basic Books, 2009.

Segal, Zindel V., J. Mark G. Williams, and John D. Teasdale. Mindfulness-Based Cognitive Therapy for Depression. New York: Guilford Press, 2002.

Strawson, Galen. "Against Narrativity." Ratio 17, no. 4 (2004): 428-452.

White, Michael, and David Epston. Narrative Means to Therapeutic Ends. New York: Norton, 1990.

White, Michael. Re-authoring Lives: Interviews and Essays. Adelaide: Dulwich Centre Publications, 1995.

White, Michael. Maps of Narrative Practice. New York: Norton, 2007.

Zahavi, Dan. Subjectivity and Selfhood: Investigating the First-Person Perspective. Cambridge, MA: MIT Press, 2005.