Fostering Embodied Sense of "Maai" in Musical Ensembles: An Interactive Media Approach for Cultivating Collaborative Timing and Spacing

Fostering Embodied Sense of “Maai” in Musical Ensembles: An Interactive Media Approach for Cultivating Collaborative Timing and Spacing Karen Kurotaki1,KaoruMikami 2,YoshinariTakegawa 2[0000000319470021], and Masaki Matsubara1[000000031950683X] 1University of Tsukuba 2Future University Hakodate Abstract. Maai,thenuancedsenseoftimingandspacingbetweenperformers, is essential in musical ensemble performance but challenging to teach through verbal guidance alone. This study explores the potential of an interactive media system to support the experiential understanding of “maai” in a non-musical context, aiming to foster mutual coordination skills relevant to ensemble playing. The system invites users to synchronize hand movements with a projected autonomous object in a water tank, encouraging shared spatiotemporal awareness akin to ensemble interactions. An initial experiment was conducted with ten groups of string performers, analyzing video recordings and questionnaire responses before and after the experience. While further research is needed, preliminary observations suggest that such interactive media may help cultivate embodied sensitivity and coordination, offering a promising direction for developing expressive and cohesive ensemble performance skills. Keywords: Interactive Media ·Music Performance Support 1Introduction Maai is a Japanese term that refers to the subtle interplay of timing, spacing, and relational awareness between individuals engaged in a shared activity. Originally associated with martial arts and traditional performing arts in Japan, maai has broader relevance to many domains, including sports, interpersonal communication, and—most pertinent to this work—musical ensemble performance. In an ensemble setting, maai involves dynamically adjusting to others’ tempo and phrasing, recognizing cues in posture and breath, and maintaining awareness of physical distance. Although maai is a culturally specific concept, it resonates with universal challenges in human collaboration: the need to coordinate multiple senses, anticipate others’ actions, and establish a fluid collective rhythm. All rights remain with the authors under the Creative Commons Attribution 4.0 International License (CC BY 4.0). Proc. of the 17th Int. Symposium on Computer Music Multidisciplinary Research, London, United Kingdom, 2025 Proc. of the 17th International Symposium on CMMR, London, UK, Nov. 3-7, 2025 905 K. Kurotaki et al. Fig.1: The experience of using the system Despite its ubiquity in collaborative domains, maai is notoriously difficult to cultivate for three key reasons. First, it is deeply grounded in individual bodies and their movement habits; small differences in posture, muscle tension, or breathing can dramatically affect how a performer perceives and produces timing. Second, maai is not easily translated into verbal instructions. Although teachers or conductors might say, “Align your breathing” or “Watch your partner’s bow,” such directives do not always transmit the bodily intuition required to sense the right moment to lead or yield. Third, maai is highly situationdependent, making it challenging to replicate the same conditions in rehearsals or training sessions. Subtle changes in venue acoustics, ensemble size, or even interpersonal dynamics can alter the demands of maai in unpredictable ways. While technologies have been explored to support ensemble practice, most training methods still rely on verbal instructions and repetitive rehearsal, which often fail to convey the pre-verbal, bodily sensitivity needed to internalize maai. Prior work in music-related domains has investigated how interactive systems can facilitate embodied coordination through feedback mechanisms and immersive interaction [1–5]. These studies highlight the potential of technology to enhance synchrony and nonverbal awareness among performers. However, cultivating the subtle, relational quality of maai—the dynamic interplay of timing and spacing between bodies—remains a unique challenge. To address this, we explore how physically engaging, non-musical tasks can provide musicians with alternative ways of developing spatiotemporal attunement and collaborative sensitivity. To address this gap, we developed an interactive media system that invites users to engage with the principles of maai through a non-musical, physically immersive task (Fig. 1). The system projects a dynamic object following a figureeight trajectory onto the surface of warm water in a tank. As participants move their hands in the water, the object responds by changing its speed and color, prompting continuous adjustment between leading and following. This experience encourages the development of shared spatiotemporal awareness through direct bodily engagement. Proc. of the 17th International Symposium on CMMR, London, UK, Nov. 3-7, 2025 906 Fostering Embodied Sense of “Maai” in Musical Ensembles This paper reports on initial findings from an ongoing study. Our contributions are twofold: (1) We present an interactive media system designed to help musicians internalize relational timing and spacing (maai)throughshared,nonverbal coordination in a low-pressure environment. (2) We report qualitative insights from video observations, questionnaire responses, and post-experience interviews with ten groups of string performers. Additionally, quantitative gaze analysis revealed increased visual engagement during interaction, suggesting enhanced attentiveness to temporal and spatial cues. While causal relationships remain to be fully clarified, these results highlight the promise of interactive, physically engaging media for supporting embodied coordination in ensemble. 2TheoreticalBackground 2.1 Maai and Its Embodied Nature Maai,oftendiscussedindomainssuchasmartialarts,dance,andtheater, describes how participants manage both distance and timing in a fluid, coconstructed manner. Kono [6] propose that “Accordingly, maai implies ‘fitted interval’, ‘suitable distance’, or ‘adequate space’.” Masciotra et al. [7] propose that in the context of Karate, “the sense of distance as maai constitutes an integrated spatiotemporal structure, producing a dynamic connection between mutually attuned existences.” In other words, maai transcends mere physical measurement or clock-based timing; it emerges from the interplay of bodies in motion, each continuously adapting to the other. These insights translate well to musical ensembles, where bow placements, breathing cues, and subtle tempo shifts collectively shape the overall performance. Understanding maai as an embodied, relational process underlines the importance of designing learning tools that engage performers’ bodily senses rather than relying solely on verbal directives or sheet music cues. 2.2 Maai in Ensemble Music In ensemble settings, the ability to perceive and respond to maai—the dynamic interplay of timing and spacing—is essential for cohesive performance. When musicians overly focus on mechanical beat-matching, musical flow and expression can suffer. In contrast, an embodied sense of maai enables performers to flexibly alternate between leading and following, responding to each other’s microadjustments in tempo and dynamics. This “shared timing” is supported by nonverbal communication through subtle bodily cues such as gestures, gaze, posture, and breath. Studies of string quartets and vocal duos have shown that mutual gaze and breathing play a key role in maintaining expressive synchrony [8,9]. In visually restricted contexts, performers often rely on aligned inhalation to cue entries [10]. Vera and Chew [11] further demonstrated that even within sustained notes, nuanced dynamic shaping can encode metrical cues to support ensemble synchronization. Proc. of the 17th International Symposium on CMMR, London, UK, Nov. 3-7, 2025 907 K. Kurotaki et al. These findings underscore that maai emerges through embodied spatiotemporal awareness—an ongoing adjustment among visual, auditory, and kinesthetic channels. In ensemble music, shared rhythm and expressive unity emerge from the performers’ ongoing bodily and adaptive coordination. 2.3 Internalization of maai Through Hand Movements The Japanese concept of maai,involvingsubtletimingandspatialawareness among performers, is embodied and difficult to convey through verbal instruction. We suggest that hand movement provides a direct means of internalizing maai by physically engaging with spatiotemporal dynamics. Research indicates that bodily motion supports musical coordination and expressivity. Chang et al. found that ensemble musicians exhibit synchronized body sway, correlating with emotional unity and engagement [12]. Breathing has also been shown to function as a timing cue, especially when visual contact is restricted [10]. Furthermore, Lange et al. demonstrated that cardiac and respiratory rhythms synchronize during ensemble performance with physical contact [13], suggesting multi-level bodily coordination in music. Our system invites participants to synchronize hand movements with a projected object in a water tank, creating a dynamic task that requires constant calibration of timing and proximity. This immersive interaction helps develop embodied sensitivity to rhythmic flow and spatial negotiation—skills essential to ensemble maai. 2.4 Maai and Gaze In ensemble performance, maai facilitates synchronization through subtle timing and spatial awareness. Gaze plays a key role in this process by supporting nonverbal communication and joint attention [14, 15]. Eye movements help performers monitor spatial dynamics and anticipate timing. Bishop et al. found that duet pianists looked at each other more frequently during unstable passages [16], while D’Amario et al. showed that visual contact improved synchronization in vocal duos [9]. Vandemoortele et al. similarly observed gaze exchanges at phrase boundaries in trios [17], highlighting gaze as a structural and expressive cue. These studies suggest that gaze fosters shared understanding and enhances ensemble cohesion. Even brief glances during rests or transitions can reinforce collective timing and musical intent. Our system reflects this mechanism in a non-musical context. By tracking and responding to a projected moving object, participants engage in visual coordination that mirrors the anticipatory processes underlying maai,supporting embodied interaction through shared visual attention. 2.5 Prioritized Integrative Attention Keller [18] describes the type of attention necessary for achieving ensemble harmony as “Prioritized Integrative Attending.” Performers must simultaneously Proc. of the 17th International Symposium on CMMR, London, UK, Nov. 3-7, 2025 908 Fostering Embodied Sense of “Maai” in Musical Ensembles concentrate on their own part, the overall musical texture, and cues from fellow ensemble members [19]. While each musician maintains a level of focus on their instrument and part, their attention must also be distributed outwards to integrate multiple sources of information—visual, auditory, and even kinesthetic. This concept aligns well with the notion of maai, which also demands that performers continually adjust their awareness and actions based on relational cues. Mastering such attention patterns takes significant practice, often involving trial and error during rehearsals. However, relying purely on verbal instructions like “listen more carefully” or “watch the concertmaster” can be inadequate. Instead, having a physically engaging way to hone these distributed attention skills may enhance overall ensemble cohesion. 3SystemDesign 3.1 Purpose The interactive system proposed in this study helps users internalize the principles of maai—relational timing and spacing between performers—through a shared, non-verbal coordination task in a low-pressure, sensory setting. While prior systems have supported musical training, few address embodied coordination through non-musical, physically immersive activities. Our system fills this gap by engaging participants in a sensorimotor task that highlights pre-verbal aspects of ensemble interaction. Participants place their hands in warm water while a dynamic object is projected onto the surface. The object, which follows a figure-eight path, changes color when aligned with the user’s motion, and adjusts its speed in response. This real-time interaction enables users to physically explore leading and following, fostering awareness of subtle timing adjustments that underpin ensemble synergy. 3.2 Design We chose a figure-eight pattern for the object’s motion to create a smooth, calming effect that supports a relaxed, attentive state essential for sensing maai. This motion encourages sustained mutual attention and a sense of continuity, making it easier for users to engage without feeling rushed or disoriented. We also tested alternative trajectories such as linear oscillation and random motion, but they led to unstable synchronization and less consistent user engagement. The system adjusts the object’s speed to match the user’s hand movement, promoting coordinated motion. Tactile feedback through warm water enhances embodied awareness, integrating visual, tactile, and proprioceptive cues. When the user’s hand is near the object for a while, the object changes color to signal synchrony. 3.3 Setup The system setup is illustrated in Fig. 3. A transparent water tank is placed on a stable surface and filled with warm water. Lightweight, water-absorbent resin Proc. of the 17th International Symposium on CMMR, London, UK, Nov. 3-7, 2025 909 K. Kurotaki et al. Fig.2: The trajectory of the dynamic object projected onto the tank. Fig.3: System setup toys (e.g., small floating balls) are also placed in the tank. A projector mounted above the tank casts the dynamic figure-eight object onto the water surface. ADepthAIcamera(OAK-DS2,Luxonis)ispositionedtocapturethetank from above and track the user’s hand coordinates in real time. Python handles camera-based hand position tracking, while Processing controls the projected object’s movement and color changes. 3.4 Interaction with Dynamic Objects The figure-eight path is calculated with the center of the tank as the origin using the following equations: x=centerX + cos(angle)·radius (1) y=centerY + sin(2 ·angle)·radius (2) Here, centerX and centerY are the screen’s center, radius defines the figureeight size, and angle updates based on the speed difference between the user’s hand and the object. When the user’s hand is within a 150-pixel radius of the object, the object’s color changes to signal synchrony. If sustained for five seconds, the system slightly adjusts the object’s speed in intervals to encourage continuous attention and adaptation. 4 Experiment In order to evaluate the effectiveness of our interactive media system, we conducted two separate but parallel experiments. The first experiment involved five ensembles (labeled Groups A–E) who experienced the interactive media system. In the second experiment, five ensembles (labeled Groups F–J) received only verbal instructions without using the interactive system. By comparing Proc. of the 17th International Symposium on CMMR, London, UK, Nov. 3-7, 2025 910 Fostering Embodied Sense of “Maai” in Musical Ensembles results from these two conditions, we aimed to isolate how the embodied, nonverbal approach might impact spatiotemporal coordination in ensemble performance. The study was approved by the Institutional Review Board (IRB), and all participants provided informed consent. 4.1 Participants Interactive Media Condition We recruited five ensembles (Groups A–E) of string players, all of whom used the interactive media system. All performers had more than 10 years of experience on their instruments. Verbal Instruction Condition We recruited five additional ensembles (Groups F–J) of string players performing a similarly challenging repertoire. All performers had more than 10 years of experience on their instruments. These ensembles received only written and spoken instructions, without using the interactive system. This verbal-only condition served as a baseline to compare the outcomes of traditional verbal coaching with the embodied training of the interactive system. As Goolsby [20] noted, “multiple performance variables were often addressed in a single stop” (e.g., phrasing alongside airflow, articulation, and balance). This method mirrored conventional teaching practices and provided a useful reference point for evaluating the interactive system’s impact. 4.2 Procedure The core design was similar across conditions, with Step 3 varied to distinguish the interactive media group from the verbal-instruction group. 1. Baseline Recording. Each ensemble performed and was recorded under typical rehearsal conditions. To examine their usual performance, as a baseline, we asked the ensemble groups to play a piece they usually work on. For multi-movement pieces, a short excerpt (up to 5 minutes) was chosen. 2. Pre-Questionnaire & Interview. Participants completed a brief questionnaire and interview to assess initial impressions of their ensemble coordination. 3. Condition-Specific Intervention: – Interactive Media Condition (Groups A–E). Participants interacted with the projected object for 2 minutes, aiming to synchronize hand movements with the object’s figure-eight path and speed. – Verbal Instruction Condition (Groups F–J). Participants received four verbal coaching cues: (a) “Match your breathing,” (b) “Coordinate bow placement and speed,” (c) “Play confidently and broadly,” (d) “Relax and play comfortably.” These cues were noted on their sheet music before the second performance. 4. Second Recording. The ensemble performed the same piece again, followed by a video recording for comparison with the baseline. Proc. of the 17th International Symposium on CMMR, London, UK, Nov. 3-7, 2025 911 K. Kurotaki et al. 5. Post-Questionnaire & Interview. Participants completed the questionnaire again (with additional items about the intervention) and participated in a final interview to describe any observed changes in performance. 4.3 Data Collection and Analysis Performance Video Analysis Each ensemble’s performance was filmed preand post-intervention from the same camera angle. To quantify mutual awareness, we used ELAN software [21] to label how often performers looked toward each other (mutual gaze). Gaze frequency and timing are established indicators of attentional focus in ensemble contexts. Questionnaire and Interviews We had participants answer a 5-point Likert scale questionnaire (1 = Strongly Disagree to 5 = Strongly Agree), both before and after the intervention. The questionnaire included seven items designed to assess participants’ self-perceived ensemble abilities and musical awareness: listening to others’ sounds (Q1), skill in ensemble playing (Q2), riding the overall sound flow (Q3), maintaining an appropriate sense of distance from others (Q4), expressing one’s own musicality (Q5), performing in a relaxed state (Q6), and feeling a sense of unity in sound with others (Q7). For the interactive media group, additional questions addressed perceived changes after using the system and asked participants to freely describe specific effects and influential aspects of the experience. After completing the questionnaire, participants were interviewed for 5–10 minutes to further explore their subjective experiences regarding timing, relaxation, and group coordination. 5Results In this section, we present the results of both experiments side by side to facilitate direct comparison. We first analyze questionnaire responses (Section 5.1), then compare changes in mutual gaze (Section 5.2). 5.1 Questionnaire Responses We compare how participants’ self-reported experiences shifted in each condition. Key items included blending with others’ music (Q3), maintaining appropriate distance (Q4), feeling relaxed (Q6), and sensing sound unity (Q7). Figure 4 displays radar charts summarizing preand post-intervention scores. For the Interactive Media Condition (Figure 4a), notable shifts were observed in items relating to interpersonal timing (Q3, Q4) and a relaxed performance state (Q6). Several participants commented that interacting with an autonomous object heightened their sense of bodily awareness, making it easier to manage tension. The increased sense of relaxation following the system experience is likely due to the relaxation effects of hand bathing [22]. By contrast, Proc. of the 17th International Symposium on CMMR, London, UK, Nov. 3-7, 2025 912 Fostering Embodied Sense of “Maai” in Musical Ensembles (a) Interactive Media Condition (b) Verbal Instruction Condition Fig.4: Questionnaire Responses (a) Interactive Media Condition (Groups A–E) (b) Verbal Instruction Condition (Groups F–J) Fig.5: Comparison of Changes in Mutual Gaze Before and After Intervention. the Verbal Instruction Condition (Figure 4b) revealed subtler or inconsistent improvements, with interviewees frequently noting difficulty putting verbal advice into immediate embodied practice. 5.2 Changes in Gaze Behavior We focus on the frequency of mutual gaze events, a core indicator of the performers’ attentional distribution toward each other. Figure 5 compares pre/post mutual gaze counts for each ensemble under the two conditions. In the Interactive Media Condition (Figure 5a), all ensembles(A–E) showed a significant increase in gaze events. Interviews indicated that the system encouraged more frequent visual checks of each other’s playing. In the Verbal Instruction Condition (Figure 5b), ensembles F–J had smaller in gaze events. Participants noted that while the reminders—“Match breathing” and “Coordinate bows”—were conceptually helpful, physically integrating them during live performance was challenging. 5.3 Comparative Overview Overall, these findings suggest that the embodied training provided by the interactive media system is more effective in prompting behavioral and perceptual Proc. of the 17th International Symposium on CMMR, London, UK, Nov. 3-7, 2025 913