Towards semiotically driven empirical studies of ballet as a communicative form

The paper addresses how dance—particularly classical ballet—functions as a communicative system and how its movement can be systematically linked to discourse and narrative interpretation. Building on the recognition that dance involves both iconic (embodied) and conventional (symbolic) meaning-making, the authors aim to provide a principled framework that adapts tools from linguistics, semiotics, and multimodality to dance. The goals are to: define a semiotic mode for ballet that bridges measurable movement to discourse semantics; demonstrate how movement data (e.g., via motion capture) can be transformed into qualitative, discourse-relevant units; and establish a basis for empirical study, corpus building, and experimentation. The work is motivated by theoretical interest in the nature of movement-based communication and by practical benefits for teaching, archiving, and choreography analysis.

The study situates itself within social semiotic multimodality and systemic-functional linguistics, drawing on Bateman et al. (2017) and Maiorani’s Functional Grammar of Dance (FGD). It notes challenges in empirical multimodal research, especially addressing iconicity and movement-based semiotics. Formal approaches to non-verbal discourse, such as DRT/SDRT extensions to visual and dance narratives (Abusch 2013; Patel-Grosz et al. 2019; Asher & Lascarides 2003), inform the discourse layer. Perceptual grouping and segmentation principles for dance (Charnavel 2019) provide constraints aligning cognitive segmentation with proposed semiotic units. The authors distinguish their approach from professional dance notation systems (Labanotation, Benesh), which encode movement but not discourse-level semantics, interaction with space, or interpretative choices. Prior motion capture and movement analysis work (e.g., Camurri et al. 2002, 2016; Vincs & Barbour 2014) offers complementary technical background for capturing and analysing movement features.

The authors adopt a tri-stratal definition of semiotic mode (Bateman et al. 2017): material (measurable physical regularities), form (qualitative categories/structures), and discourse semantics (mode-specific discourse relations and coherence strategies). For ballet, the form layer is instantiated via Maiorani’s Functional Grammar of Dance (FGD), which organizes movement around two dimensions: (1) spatial displacements forming trajectories (Moves and Minimal Ballet Sequences, MBSs), and (2) projections—coordinated pointing/extension by articulators (arms, hands, legs, feet, torso, head) toward meaningful spatial regions. Segmentation: The smallest unit is the Move (a bounded displacement with a set of projections). Two Moves combine into a Minimal Ballet Sequence (MBS) that establishes either continuity (same direction) or variation (change of direction) in trajectory. Projections are defined by articulators’ directions relative to the dancer’s movement (perpendicular/vertical or horizontal, following, opposite). Mapping to discourse: Each projection configuration maps to abstract discourse predicates: CONNECTING, COMING-FROM, GOING-TO, LOCATING, ENGAGING, ADDRESSING (Table 1). These events have roles (agent, source, goal, relatum, addressee), analogous to event semantics in linguistics. Spatial semantics: Projections resolve to discourse referents by anchoring to qualitative stage regions using Freksa’s double-cross calculus (15 regions, plus vertical Top/Ground), allowing tracking of referents as dancer orientation changes. Interpretation proceeds in two stages: (i) under-specified iconic event structures from projections, (ii) contextual resolution to stage entities (e.g., King/Queen, Princes, audience) based on dancer orientation, stage layout, and narrative context. Worked example: Using Princess Aurora’s Act I solo (Sleeping Beauty), the authors segment raw movement into Moves and MBSs, derive projection-based discourse events for specific Moves (e.g., M1: CONNECTING to sky; COMING-FROM court; LOCATING on ground; ENGAGING/ADDRESSING court), and combine Moves into MBS-level structures exhibiting topics and contrasts. Direction changes between MBSs invert referent mappings, producing parallelism and mirroring across sequences. Comparative analyses: Applying the method to Bolshoi, Opéra de Paris (Nureyev), Kirov TV, and Matthew Bourne’s contemporary version reveals congruent MBS-level discourse patterns with local differences in projections and addressees, driven by choreography and staging. Empirical pipeline: Material layer via motion capture (Perception Neuron inertial suit) and a Python script to extract articulator orientations/directions; form layer via FGD-based qualitative classification; discourse layer via projection-to-event mapping and spatial resolution. Corpus annotation: Use ELAN with tiered, controlled-vocabulary annotations aligned to semiotic strata—Moves, MBSs, narrative/interactive projections, structure, and implicative dimensions—supporting pattern discovery and hypothesis generation. Experimental agenda: Propose eye-tracking to test discourse expectations and reference resolution; participant-based event segmentation to validate Move/MBS boundaries; broader evaluation of discourse relations (Narration, Elaboration, Parallel) in ballet versus language.

• A principled semiotic mode for ballet links measurable movement to discourse-level interpretations by: (a) segmenting into Moves and Minimal Ballet Sequences, (b) defining projections by articulators’ orientations, and (c) mapping projection configurations to abstract discourse events (CONNECTING, COMING-FROM, GOING-TO, LOCATING, ENGAGING, ADDRESSING).
• Worked analysis of Aurora’s Act I solo shows step-by-step derivations from motion to under-specified event structures and contextual resolution to stage referents using Freksa’s spatial calculus, yielding coherent discourse structures with parallelism and mirroring across MBSs.
• Comparative studies across traditional (Bolshoi; Opéra de Paris/Nureyev) and variant contexts (Kirov TV studio; Matthew Bourne’s modern version) reveal stable MBS-level discourse patterns despite local variation in projections and interactions, demonstrating the framework’s robustness and sensitivity to interpretative differences.
• A preliminary motion-capture pipeline (Perception Neuron + Python) correctly detects articulator orientations and projection changes needed for FGD categorization, enabling semi-automated extraction of the material/form layers.
• Corpus annotation with ELAN, using FGD-derived controlled vocabularies and tier hierarchies, exposes higher-level rhetorical structures (e.g., mirrored varied MBSs) not apparent from move-level repetition; example from Raymonda highlights mirrored MBSs (M15=M18; M16=M17) and consistent projection patterns towards on-stage participants.
• The approach imports empirical methods from linguistics and perception (eye-tracking for reference/attention, participant-based segmentation for boundary detection) to test discourse hypotheses in ballet.

By integrating FGD within a tri-stratal semiotic mode, the study addresses the central question of how ballet movements convey discourse and narrative meanings. The mapping from projections to abstract discourse predicates, together with qualitative spatial resolution, operationalizes both iconic and conventional aspects of dance communication. The example analyses demonstrate that movement trajectories and projection patterns structure discourse via MBS-level topics, contrasts, parallelism, and mirroring, aligning with discourse relations familiar from language (Narration, Elaboration, Parallel). Comparative applications show that the framework captures choreographic and contextual variation while preserving higher-level discourse congruencies, facilitating cross-version and cross-context interpretation. The motion-capture and corpus-annotation components bridge continuous movement to discrete, analysable representations, enabling hypothesis-driven empirical research. This advances multimodality studies by making dance amenable to corpus-based and experimental methods and by enabling cross-semiotic comparisons of discourse structure and reference.

The paper presents a prototype end-to-end methodology that transforms raw dance movement into abstract discourse-semantic representations suitable for empirical analysis. Contributions include: (1) defining ballet as a semiotic mode with material, form, and discourse strata; (2) operationalizing FGD constructs (Moves, MBSs, projections) and mapping projection configurations to discourse events; (3) linking projections to referents via qualitative spatial semantics; (4) demonstrating analyses on canonical and variant performances; and (5) outlining a data and tools pipeline (motion capture, Python analysis, ELAN annotation) for corpus construction and experiments. Future research directions include expanding coverage to a broader repertoire of movements and multi-dancer configurations; characterizing the full inventory of ballet discourse relations; integrating background story knowledge with on-line semantic configurations; conducting eye-tracking and segmentation studies to validate reference and boundary predictions; and exploring applications in pedagogy, archiving, choreography support, and audience guidance.

• The motion-capture and analysis pipeline is preliminary; automation focuses on capturing articulator orientations and projection changes, not full high-level semantics.
• The study does not report controlled experimental results; empirical validation (e.g., eye-tracking, segmentation tasks) is proposed for future work.
• Analyses concentrate on single-dancer sequences and a limited set of pieces; extension to multi-dancer interactions and broader movement vocabularies is pending.
• Resolution of under-specified discourse referents depends on staging and contextual knowledge, which may vary across productions and audiences.
• The framework complements but does not replace professional notation systems; it targets communicative/discursive meaning rather than exhaustive kinematic encoding.
• While the approach is generalizable, the current instantiation focuses on classical ballet and may require adaptation for other dance forms.