A uniform empathy deficit cannot explain the mixed clinical picture
The paper starts from a mismatch. Accounts built on impaired theory of mind, reduced emotional understanding, or an imbalance between cognitive and affective empathy each capture part of the picture, but the authors argue none fully explains why some patients show poor perspective-taking and empathic accuracy while also reporting elevated personal distress or emotional over-involvement. A single capacity reduction would predict uniformly blunted empathy, not this combination of under-responsive and over-involved behaviour.
The proposed alternative is a three-factor neurobehavioral framework organised around self–other distinction regulation. The claim is not that patients have less empathy overall, but that the boundary processes deciding what belongs to the self and what belongs to the other person are less stable. Because the framework is a review-level model rather than a new experiment, its value lies in how it partitions the problem and what it predicts, not in a reported effect size.
Three stages separate access, representation, and source attribution
Spontaneous empathic access asks whether cognitive and affective cues from another person can enter the observer's processing system under low task demands. This is the earliest stage: before any deliberate reasoning, does the other person's expression or viewpoint register at all? The paper links this stage to evidence that patients show reduced intentionality and appropriateness when describing social interactions without explicit mental-state prompts, and to spontaneous visual perspective-taking, where another's spatial viewpoint can be registered without instruction.
Inferential empathy is the representation of others' beliefs, intentions, perspectives, and emotional states as distinct from the self. This is where classic theory-of-mind difficulty sits: the paper notes meta-analytic evidence that patients perform worse than healthy controls on false-belief, hinting or irony, and story- or cartoon-based mental-state inference tasks. Crucially, the framework places these difficulties mainly within one stage rather than treating them as the whole of empathy.
Attributional empathy concerns whether activated emotions, thoughts, or perspectives are assigned to the self or to another person. A feeling can be triggered correctly and still be mislabelled as one's own; this stage is what the authors use to explain emotional over-involvement and personal distress that coexists with weak perspective-taking. The three processes are described as functionally distinguishable but closely interdependent, so disruption early on can propagate to later judgments without implying an independent deficit at every stage.
Interpersonal synchrony can help or hurt depending on its level
The framework adds interpersonal synchrony as a modulator: the real-time coordination of self- and other-related signals during social interaction. The paper is explicit that the available evidence does not support a simple claim that stronger synchrony always improves empathy. Instead it treats synchrony as context-sensitive — insufficient synchrony may weaken interpersonal access and updating, while excessive or poorly regulated synchrony may increase self–other merging or personal distress.
Applied to the three stages, synchrony could shape whether other-related signals enter self-related processing, whether changing social cues are organised into an accurate representation of another's emotion or intention, and whether an elicited response stays other-directed or shifts toward self-oriented distress. The paper presents this as a hypothesis to be tested directly in schizophrenia by combining motion tracking, physiological coupling, empathic accuracy tasks, and self–other source-judgment measures within the same dyadic paradigm, not as an established finding.
Each stage maps onto a different set of brain systems
At the neural level the framework assigns the stages to interacting systems. Spontaneous access and affective sharing are mapped onto salience, affective-sharing, and mirror networks, including the anterior insula, anterior/mid-cingulate cortex, and fronto-parietal mirror regions such as the inferior frontal gyrus and inferior parietal lobule. Inferential empathy is mapped onto mentalizing and frontotemporal social-inference circuits, including the temporoparietal junction, medial prefrontal cortex, superior temporal sulcus, and precuneus/posterior cingulate cortex.
Attributional empathy is mapped onto self–other distinction and source-attribution systems. This mapping is a organising proposal that links each behavioural stage to candidate circuitry; the review does not report new neuroimaging data testing whether these systems dissociate in patients. Readers should treat the neural assignment as the framework's scaffolding for generating predictions rather than as a demonstrated segregation.
The framework relocates theory-of-mind and disequilibrium findings
The paper compares its model with two existing accounts. Theory-of-mind models explain whether patients can infer others' beliefs, intentions, perspectives, and mental states, but the authors argue they less clearly specify whether empathy difficulties arise during early access, representation, or source attribution. In the new framework, theory-of-mind difficulty is located mainly within inferential empathy, especially the construction and maintenance of other-related representations.
Empathic disequilibrium theory holds that deficits need not be a uniform reduction but may reflect an imbalance between cognitive and affective empathy; the paper cites evidence that patients may show lower perspective-taking and empathic concern but higher personal distress, and that relatively dominant affective empathy may predict schizophrenia diagnosis more strongly than cognitive or affective empathy levels alone. The framework accepts this insight but asks where the imbalance arises — during early reception of others' emotions, representation of their states, or attribution of emotions and perspectives to the correct source.
The predicted phenotypes differ accordingly. Theory-of-mind accounts predict poor performance on false-belief, Reading the Mind in the Eyes, cartoon inference, explicit perspective-taking, and complex social-inference tasks. The disequilibrium account predicts weaker perspective-taking or empathic concern with preserved or elevated personal distress and emotional contagion. The three-stage framework predicts weak spontaneous responsiveness to others' cues, unstable emotional perspective-taking or empathic accuracy, and difficulty attributing activated emotions or perspectives to the correct source.
Different disrupted stages should produce different clinical patterns
The framework's practical payoff is a stage-by-stage reading of patients. If spontaneous empathy is insufficient, patients may show reduced responsiveness to others' emotional cues, potentially expressed as social withdrawal, indifference, or low empathic concern — a pattern the paper connects to links between negative symptoms and less stable spontaneous empathic processing. If spontaneous empathy is partly preserved but inferential empathy is impaired, patients may be emotionally affected by others yet have difficulty forming an accurate representation of what the other person feels or intends.
This is where the model departs most from a single-deficit view: the same observable empathy problem can arise from different stages, and the stage involved should change what a clinician or researcher expects to see. The paper frames this as a mechanistic basis for identifying which stage of empathic processing is most affected in different patients, rather than assigning all patients to one empathy profile.
The framework stands or falls on dissociating the three stages
The authors are direct about the main requirement: empirical testing with tasks that dissociate access, representation, and attribution. Because the stages are proposed as partially separable and interdependent, a study that measures only overall empathy accuracy cannot tell the framework apart from a theory-of-mind or disequilibrium account. The review also notes that the available synchrony evidence does not establish that stronger synchrony always improves empathy, so the modulator's role remains to be tested in schizophrenia specifically.
For intervention research, the framework generates stage-specific predictions: emotional cue training might primarily improve spontaneous empathy, theory-of-mind training might mainly improve inferential empathy, and source-monitoring and metacognitive training might be more likely to affect attributional empathy. The paper presents these as hypotheses for future trials, noting that existing social cognitive training improves affective and cognitive theory of mind, that metacognitive training improved overall social cognition and theory of mind but not emotion processing, and that empathy training effects differ across dimensions and are not consistently sustained.
The priority this paper sets is therefore not to ask whether patients are less empathic, but which of the three regulatory stages is unstable in a given patient. That stage-level question is what the framework claims theory-of-mind and empathic-disequilibrium accounts cannot answer, and it is the comparison readers should carry forward: a patient with preserved spontaneous access but impaired attribution requires a different explanation, and potentially different training target, than one whose other-related cues never register in the first place.
Sources used in this answer
When “I” and “You” blur: self–other distinction failure and empathy dysregulation in schizophrenia
Proposes a three-factor neurobehavioral framework in which empathy difficulties in schizophrenia reflect less stable self–other distinction regulation across spontaneous empathic access, inferential empathy, and attributional empathy, rather than a uniform reduction in empathic capacity, with interpersonal synchrony as a context-sensitive modulator.