Narrative complexity is weakly coupled with affective energy, not net valence, in relationship narratives

People often assume that emotionally intense relationship experiences “leak” into language: more elaborated stories should sound more emotional. We tested this expressive-congruence assumption in 351,734 de-identified relationship narratives using an interpretable index of narrative complexity (Level of Complexity; LoC) and sentence-level sentiment summaries. LoC showed little association with net valence magnitude (the absolute value of the document-level mean sentence valence, v i ; r = − 0.061 , R 2 = 0.37 % ) but a modest positive association with cancellation-robust affective intensity (MAV: r = 0.186 , R 2 = 3.47 % ; RMS: r = 0.220 , R 2 = 4.85 % ). A residual-based discrepancy index (NADI, defined on net valence magnitude) revealed systematic expressive profiles, including highly elaborated but affectively muted narratives and brief yet affectively charged disclosures. Unsupervised clustering yielded five reproducible expressive profiles that showed structural parallels with constructs discussed in emotion-regulation and self-presentation research, without providing direct evidence of subjective regulatory intent. Because the design is correlational, these associations do not establish causal relations or individual-level regulatory processes. Narrative elaboration is therefore not a reliable proxy for surface-level net affective tone in naturalistic relational self-disclosure.