Towards Designing Playful Bodily Extensions: Learning from Expert Interviews

In our work, bodily extension refers to devices attached to and extending from the body, such as a robotic arm. In particular, as noted also above, we are interested in moving, interactive and shape-changing pieces.

Our work broadly relates to the wearables field since the body extensions are worn on and attached to the body. Wearables, computational devices worn on and attached to the body, represent a wide variety of artefacts, from textile-based smart clothing [82] to interactive tattoos [68]. Thus, our work draws on and can be compared to work about wearables, but focuses more narrowly on body-worn shape-changing technologies that reach out from the body. We further focus on devices that can manipulate body image and body schema, and how they do so. Most wearables are primarily for sensing [7] and are adopted into the body schema or body image. For instance, a smartwatch does not change how we perceive our body moving or existing in space; likewise, Google’s Project Jacquard [101] makes its technology almost invisible, as do medical wearables hidden underneath clothes. Some wearables change body image significantly. For example, the Intimacy Dress [106], which becomes transparent based on the wearer’s excitement level, probably affects how users perceive their body image, but not the body schema, because the dress is not shape-changing; thus it is outside the scope of our work. On the other hand, the inflatable skirt [124], which is computational clothing, is in the scope of our work. Thus, apart from being attached to the body and shape-changing, we classify bodily extensions as artefacts that perceptibly manipulate the body image and body schema.

Our work is also related to prosthetics. Although many studies assume that prostheses replace limbs to conform to a normative body image [81, 115, 127, 134], we posit that normativity is not possible in the fluidity of bodies [85]. Therefore, having a prosthesis represents an experientially different state. Even if the prosthesis may seem to help conform the normative understanding of how a body looks to others, it creates a fundamentally different state of being for the wearer (e.g., with phantom experiences [97]), rather than resurrecting a part of the wearer’s body or functionality [85]. Prosthetics may also enhance the bodily capabilities beyond their original states, as in the case of blade-shaped running prostheses that, it was argued, helped athletes run faster and therefore rendered them super-abled [52].

We also learned from prior work around prosthetics. For example, they can be controlled via a wide range of modalities: brain signals [22, 54], motion sensors [104], myoelectric methods [31, 76] or voice-based modalities through throat microphones [77]. These control methods are inspiring, as they suggest distinct ways of communicating between the organic body and synthetic additions. Furthermore, prosthetics can enable aesthetic expressions [63, 125], such as by adorning them with jewels for a fashionable look [3] or making them resemble superhero armours to support a playful attitude [96]. We are inspired by how these works highlight how designers can use bodily extensions as a design material that transforms parts of the body beyond functional purposes and supports playfulness, as in superhero prosthetics.

We also learned from prior work that bodily extensions can add new abilities to their user’s existing skill sets. For example, the modular robotic arm by Leigh et al. [71] has been used for various scenarios, such as carrying multiple items at once. In another study, Ding et al. tried to understand the value of bodily extensions as storage areas around the body [43]. Furthermore, bodily extensions have been used for educational purposes, such as balance training [133]. Another project aimed to support remote collaboration where the extended arms were controlled by another person for creating a sense of bodily collaboration [108]. Interestingly for the purpose of our investigation, Leigh et al. used their robotic arm to help play the guitar [71], highlighting that extension designers are increasingly thinking about playful applications. These prior works demonstrate that bodily extensions serve many different purposes and interests. However, apart from instrumental benefits, two of the most critical aspects of bodily extensions are the introduction of unfamiliar experiential qualities and the modification of bodily perceptions, the main foci of our work.

Bodily extensions’ experiential qualities have been considered in many different contexts, especially in art practice. For example, the artist Stelarc created a performance piece concerned with the experiential qualities of bodily extensions regarding proprioception and agency [113]. He attached a robotic third arm to his right arm, controlled by other people through the internet. This project, apart from extending the body physically, played with the proprioceptive sense in that he had no control of the robotic arm’s movements, yet the system employed movements relevant to his own movements. His performances addressed the unusual experiential qualities of having less agency over body movements and his perceptions of altering the body.

Another project focusing on the experiential qualities of bodily extensions is “Child Hand” [98]. In it, designers deployed a miniature hand extension that allowed users to experience what it would be like to have a grip like a child. Designers have also explored emotions that might be induced by bodily extensions. For example, Xie et al. [133] gave users a tail and investigated what emotions were induced by its movements. While these studies suggest an intriguing and unconventional experiential potential for bodily extensions, the field is still underexplored, a gap this paper aims to begin addressing.

We focus on playful bodily extensions because in playful situations people can shield themselves from the ordinary rules and issues of life [62, 114]. Thus, we believe that in play people are more prone to exploring and experimenting with unfamiliar technologies than in ordinary situations [42]. Furthermore, play establishes a protective frame [9], which further supports a perceived safe environment for exploring novel technologies. This is important, as bodily extensions can pose physical risks, such as running into things while wearing them. They can also cause social harm, as wearing bodily extensions can make others suspicious, as in the case of wearing overtly technical extensions (with visible wires, for instance) in high-security environments like airports. Using play offers researchers and designers opportunities for experimentation and exploration without the onus of designing for functionality. Bodily extensions also work well in playful situations, as they can easily be removed when players no longer want to play [80].

Bodily play has been explored in distinct contexts such as social play [65, 90], distributed play [89], wearable play [27, 34] and exertion games [91]. The experience of playing bodily games has also been investigated thoroughly, both via research through design processes that yielded guidelines [92] and through philosophical engagements such as identifying the relation of play to the lived body and the material body (e.g., the German “Körper” and “Leib” [88].) The scope of play, in our work, mostly speaks to the Danish term “at lege (being bodily playful)”, meaning being playful through bodily engagement [80]. “At lege” is a bodily attitude of exploring and experimenting with bodily perceptual stimuli without a need to reach a determinate goal or outcome and where the “doing” becomes the means in itself. In contrast, “at spille,” as in “doing,” entails a determinate outcome for actions, such as kicking a ball to score a goal. If one were to “be playful” with the ball, one would instead explore the perceptual stimulation of kicking and rolling with the ball. When users are “at lege,” playful bodily extensions can support role-playing, with players exploring the feelings of different roles while they wear the extensions.

Role-playing offers a playful context for bodily identity play, allowing players to experiment with the look of their body by extending its physicality. Role-play, which Caillois calls mimicry [30], is the play genre in which players alter their own identity by either taking on another identity or playing an identifiable role, as in the cosplay community and various carnivals around the world (e.g., Venice and Brazil) [64]. In the context of “at lege,” such play is exploring the perceptual stimulation of being bodily “extended” as an internal process affecting other bodily processes. In this process lies an exploration of the players’ body images as they are contested in others’ reactions. These processes are explained by Grosz [53] as inside-out and outside-in processes of altering and creating one’s body image. Play offers people a frame that omits common discourses and lets them access these processes in a different way [42]. In that sense, play in our work refers to experimenting with bodily sensations without a determinate goal and in relation to identity change and role-play.

Our framing of playful bodily extensions also draws on the frameworks created for bodily play through wearables, such as the design framework playful wearables (DFPW) [26]. In its performative dimension, this framework explains how wearables can be designed to leave room for imagination such as role-playing (e.g., acting as an imaginary character) or creating a detailed and immersive fictional world through elaborate costumes in cosplay. The interactive dimension of DFPW situates playful wearables in an artefact-oriented (e.g., smart watches) or embodied dimension. The embodied dimension includes extended and tangible bodily devices, speaking to our focus. The third dimension, the social, shows that playful wearables can be designed either to facilitate relaxed (e.g., waving hands) or tight (e.g., hugging) social interactions. Parallel to the social dimension, the design framework for social wearables (DFSW) [35] also draws on wearable projects based on games and play [34, 36], especially in terms of agency and the distribution of information. These frameworks help frame bodily play through wearables (and in our case bodily extensions) with regards to performativity (e.g., identity transformation, role-play), social interaction (e.g, signalling to others, computer-mediated bodily social interaction) and agency (e.g., what and who controls the bodily extensions) and have been influential for our analysis.

Rapp [103] understands wearables as body extensions and explains them from two different thought styles, externalist and internalist. The externalist thought style, which the author explains as the dominant understanding, sees wearables as objects, repositories, and instruments that quantify, enhance and enable bodily performances. The internalist style, which Rapp [103] advocates investigating further, "[f]ocuses on the potentialities of wearables indirectly affecting ‘the human’ and in integrating with their sensory and intellectual experience." We focus on bodily play as rooted in this internalist thought style and activities supporting such experiences. However, Rapp’s [103] externalist and internalist thought styles are not to be confused with Grosz’s [53] inside-out versus outside-in processes. While Rapp [103] focuses on design, Grosz’s [53] two processes focus on experiential factors affecting our body images and lie within Rapp’s [103] internalist thought style. Grosz [53] distinguishes between factors from personal versus social intercorporeal perception and how these interrelate. Thus, this paper is within the internalist thought style as we focus on body extensions’ sensory and intellectual experience.

While discussions around body image and body schema are manifold [47, 48, 126], we find the understanding from de Vignemont [38] the most useful for our paper. She explains her understanding as two different body representations; "the body schema for action (that is, information about the body necessary for movements, such as posture, limb size, and strength) and the body image for perception (that is, the judgment of one’s own bodily properties)." For example, prostheses both alter and affect a person’s body schema for action; at the same time, they affect body image as the prostheses alter and augment judgement and perception of the body. From this view, playful bodily extensions play with bodily identities, as in cosplay and carnivals, where people wear costumes to alter judgments of their body images.

Rapp [103] argues that body extensions are often seen as instruments for improvements or to collect body data, that view relies on the externalist thought style. In this paper, however, we adhere to the internalist style, viewing body extensions as vehicles for exploration and experimentation. That emphasis also aligns well with somaesthetic appreciation, which has gotten traction in embodied HCI and prioritises bodily and sensory experiences in design [39, 61, 122]. Thus, the bodily extensions examined in this paper are not investigated for a particular purpose but for their opportunities to play with body image body schema and playful bodily sensations created as a result of the alteration of those constructs.

Wigglears (Figure 1), which wiggles the user’s ears based on biodata, explores the intersection between biodata, ear movement and playful expression [100]. The inspiration came from the fact that humans have limited control over their ears. While most ears move, only some people can control their ears consciously, and it is usually only a very subtle movement. Interestingly, animals’ emotions are often conveyed from their ear movements (see [70, 105]). A person who can wiggle their ears “on command” can potentially be a source of great enjoyment for others. Wiggling ears also alter the body schema by introducing a new way of moving a body part, and the effect on body image is profound, especially when other people realise that the wearer’s ears move in an unconventional way.

The system consists of two small motors attached to a headband that moves the ears. Via sensors on the index and middle finger of the user, a microcontroller (Arduino) senses galvanic skin response (GSR), which controls a set of motors moving the ears. When the user exhibits a certain level of excitement, the galvanic skin response is set to increase, causing the ears to wiggle.

Wigglears was evaluated via an autobiographical study in which the designer wore the system for a minimum of three hours per day for two weeks. The designer kept records of the experience in a diary. Two prototypes were tested this way. The records were subsequently analysed. As such, the Wigglears system was studied in the wild and through the iteration of two prototypes.

Arm-A-Dine (Figure 2) explores the use of bodily extensions around the social experience of dining [83]. Arm-A-Dine [83] gives users a third robotic arm that feeds another person, and investigates the practice of sharing food and eating together through playful behaviour. It also alters body image and schema, since it affects balance, ad movement and, since the arm can be controlled by onlookers, the meaning of the body to others. Each person wears a robotic arm attached to their vest and the system works in the following way:

Arm-a-Dine was studied in the lab with 6 pairs, 12 participants total, between 21–27 years old. The participants, who knew each other, were asked to eat casually with no further instructions. The study was followed by a semi-structured interview.

The Tail (Figure 3) was initially developed for theatre play. The first version was a mechanical tail for the main character in Ibsen’s Peer Gynt. The result was an 80 cm–long tail worn on a belt around the hip. It was made from piano wires and 3D-printed joints controlled by two servos that were controlled externally via remote control. It was subsequently tried on stage in two rehearsals. However, the actor preferred to be able to control the tail himself. This was implemented in the second version of the tail, which used accelerometers and gyroscopes placed on the hip for control. The tail in that sense has been a source of bodily role-play since the body movements required to control the tail influenced body schema and how the wearer moves. This particular way of moving also affects the body image since different ways of moving might lead to voluntary identity change [66].

The designer wore the tail on several occasions and demonstrated it at a CHI conference [117, 118]. As such, the tail was studied in the wild, though the user study was not published. Nevertheless, the tail was originally developed and used for a theatre performance, which also led to several iterations of the tail including a remote control - and the removal of it. Despite there being no official records of a comprehensive user study except the report from the theatre performance for which the tail was first designed, the tail has now been developed commercially and is available worldwide.

The commercial version of the tail is described on the website like this cosgear.co²: ”The Costail moves according to your movements which means that if you move a lot; the tail moves a lot, and if you slow down, the tail will calm down with you. You are in total control, and with practice, you can become quite the tail whisperer. As a second option, you can loosen the chord that makes the tail move. This will reduce the movement as well. You can of course change this back later.

Besides telling how the Tail works and moves, the above quote indicates the commercial potential for these extensions for play, including adult play [74]—confirming the need to establish best practices for designing such extensions on an informed basis [42].

Monarch (Figure 4) is a set of shoulder pads that explores how wearable technologies can physically extend the wearer’s body and allow them to play with personal expression. In that sense, although Monarch affects the body schema by making the shoulders bigger, the focus is on altering the body image for expression. When relaxed, Monarch looks like regular shoulder pads; when activated, they expand and reveal a colourful, complex pleated interior. When expanded, the shoulder pads frame the wearer’s face, emphasising any difference in moods. The shoulder pads expands and contracts in response to the movement of the wearer’s muscles via kinetic textile that is activated by muscle movement [57].

Technically, the shoulder pads contain two servo motors that each rotate a wire arm inside the textile—all connected to a microcontroller and muscle sensors using electromyography (EMG)—to detect muscle movement of the biceps. A laser-cut acrylic textile was used for the first version [58]. However, it was too inflexible. For the second version, the designers used digital fabric printing on cotton poplin. This resulted in reduced production time, more freedom in visual design and colour choice, and lighter weight, easing the manipulation by the servo motors. Furthermore, the system included a control panel containing an on/off switch, a calibration knob for adjusting sensitivity, and a switch that gives the user a higher degree of control.

Monarch was studied by the designers who wore it during several exhibitions [57, 58]. As such, the reports of Monarch in this paper are based on the designer’s own experiences of wearing the system in the wild, not lab testing. Monarch was refined into a version 2.0 which is more comfortable and sustainable for everyday life.

We conducted two 1-hour interviews with the designers of Wigglears and the Tail, and a 2.5-hour session in a group interview setting [46] with the designers of Arm-A-Dine and Monarch together. Our aim was to gather all the designers to prompt discussion, however, due to the time zone differences this was not possible. We invited our participants using neutral language and did not promote them as “experts,” to prevent any influence. Each interview was online and started with a 10-minute presentation by us that communicated the aim of the project and explained the concepts of body image, body schema, and play. We described all projects included in this study to the participants so that they had a shared understanding of the designs. We also gave participants the chance to ask questions regarding these topics; none of the participants disputed our definitions of body schema, image and play. The only questions we got were about learning the details of projects (e.g., their working mechanism or interaction style). The interview questions were aimed at unpacking the design process. Specifically, we asked participants to 1) describe the most important highlight of their design process, 2) the key challenges they faced, 3) the parts of their prototypes which they would like to design differently, 4) examples of occasions on which their prototypes induced playful experiences and 5) recommendations for designers who would like to design playful bodily extensions. At the end of the interviews, we left time for brainstorming about possible future playful bodily extensions that the designers would like to explore by asking them to ideate freely on the playful scenarios that playful bodily extensions can create.

We included five designers. Designers varied in their experience of designing playful bodily extensions and their backgrounds, which helped us to gain different perspectives on the topic (Table 1).

We conducted an inductive reflexive thematic analysis (TA) on the transcriptions [20]. TA has different approaches, such as codebook, inter-rater coding or reflexive thematic analysis [21]. Although mashups between different types are possible, epistemological groundings of the combined types should not contradict each other. For example, while the inter-rater coding method prioritises the objectivity of the analysis by aiming to eliminate the bias of the analyser, reflexive TA emphasises deep and organic engagement with the data, heavily affected by the stance of the analyzer. Thus, mashing up these two methods without justification indicates a problematic analysis process.

In reflexive thematic analysis, the perspectives of the researchers who analysed the data and contributed to the creation of knowledge are of utmost importance [21]. The first author, who conducted the thematic analysis, has been working on bodily technologies and playful wearables for nine years. The second and third authors, who participated in the formation of design themes, are respectively a PhD candidate and a game designer who has worked on the bodily play for the last 10 years and is the director of a research lab that focuses on movement-based games and bodily technologies. Therefore, the analysis was driven by knowledge regarding playful wearables and bodily play.

We mostly followed an inductive reflexive thematic analysis. We did not create a prior codebook or have more than one coder. The coding benefited from the capabilities of Atlas.ti ³;and through Atlas.t, functions, our process borrowed steps from a codebook thematic analysis (in which a codebook is created prior to the analysis [21]). During the coding, by using the association functions of Atlas.ti, we already started to form relations among codes, which produced a network graph of codes and sub-codes. Therefore, although we did not create a codebook in advance, we generated a coding structure during the analysis.

The process started with uploading the interview audio files to a transcription service⁴; these were pseudonymised and deleted from the database following transcription. As suggested by the reflexive TA by Braun & Clarke [20], the first author went through the transcriptions along with the recordings and corrected the text when needed. After transcriptions were checked, the text files were transferred to Atlas.ti. The coding process followed open, axial and selective coding [131]. First, all data were inductively coded by the first author, as suggested for the open coding phase. At the end of this phase, a total of 227 codes were produced. The axial coding was done by using the Network Graph function of Atlas.ti. This process allowed us to see the main relationships between codes and demonstrated the recurring topics. The eight main recurring topics (not themes),—namely, “social interaction,” “body image,” “body schema,” “interaction modalities,” “context,” “design process,” “wearability” and “bodily feelings”—were transferred to Miro⁵ with the sub-codes for the selective coding process. For example, the “Context” topic included 29 sub-codes. During the selective coding process, we further grouped topics to create more nuanced sub-topics, namely “intrinsically playful,” “socially acceptable,” “daily life,” “serious” and “social.” Thus, before generating the themes, we had a three-level affinity diagram that included main topics, subtopics, and sub-codes.

After this multi-level coding process, we started theme generation. In the theme generation process, the codes that occurred across interviews were marked along with their frequency to understand their prevalence. In Miro, each main topic was examined by the first author based on the prevalence of the codes. While generating the themes, the first author reflexively examined the affinity diagram and re-checked the quotes assigned to codes for further sense-making. The sense-making process led to takeaways from the main topics that were summarised into a presentation that shows the possible practical outcomes and the meanings. This presentation was discussed among co-authors and the comments of each author were noted as comments in the presentation. After this meeting, by using the comments and suggested readings by co-authors, the first author worked on the main findings and the revisions on these findings led to themes. The generated themes were again discussed with co-authors in another meeting and the co-writing process of themes started. The themes were written by all co-authors (with the lead of the first author) in an online collaborative text editor.

The “network graph” of Atlas.ti and the “thematic map” we created based on it can be examined in the supplementary material.

Our conversations with the designers (D1, D3, D4, D5) revealed that bodily extensions could be seen as designed “organs facilitating social interaction.” As such, the extensions can free up, replace or extend existing bodily social interaction mechanisms such as social gestures or physical contact. They may also facilitate or augment the social signals one can produce, such as by producing movements that correspond to reactions such as blushing. In this way, bodily extensions can augment wearers’ social interaction capabilities and provide an additional mode of expression and even a new bodily language. While social communication mostly concerns body image, due to its inherent focus on being perceived and seen by others [128], here, bodily extensions also afford social capabilities that emerge through their impact on the body schema, for example, by exaggerating sensations of the body or becoming limbs reacting to social cues. The impact on body image and schema brings distinct ways of social play [65], either through self-expression (as in roleplaying) [27, 30], proximity [95] or embodied coordination [45]. To elaborate on this, we present three sub-themes: Exaggerating expressions, sensing others’ social cues and exclusive (non-verbal) modes of communication.

User control over the bodily extensions is a critical element in incorporating them into the body. Intuitively, it seems that this incorporation should be a straightforward process, and the interaction between the body and the extension should feel instinctive. Although this is true to some extent, especially when it comes to giving users a chance to operate their extensions in the way they want (D2), taking a portion of the agency can also lead to playful moments (D1, D4, D5) that can facilitate being mindful about one’s own body or the bodies of others (D3). Thus, during the design process, it is essential to explore the different experiences facilitated by different levels of agency. Different levels of agency call for the manipulation of the body schema, mostly at unexpected times, and align with bodily play that focuses on the exploration of bodily sensations (described by the Danish concept “at lege”) [80]. On the other hand, adjusting to using the extension and having a high degree of control over it might allow shifting the focus from body schema to body image, again mostly through identity change and self-expression [25, 66, 128]. We investigate this theme through three subthemes: Designing for unintentional interaction, accidental activation and high degree control for intuition.

Incorporating bodily extensions into the body schema is a process. It starts with onboarding, which is the first moment when users are acquainted with the physical entity that reaches out from their bodies. Onboarding (the beginning of the experience [14]) to play is frequently voluntary and in our study represents the intention to play. Onboarding may also be contextual; for example, the extensions can be activated when the play starts or when entering a playful context. After a moment, a learning process starts for incorporating the new body extension into the body schema and making the body interact with it; we call this the adjustment period. The final stage is concerned with removing the extension, which we call offboarding. Users may offboard when the extension is not suitable to a given context (e.g., when sleeping or when the play/game experience ends) or when the extension needs to be maintained (e.g., changing batteries). An extension might also be shared, and while one player offboards another one can onboard (and this process represents transitions among users and contexts [14]). Offboarding can disrupt the body image and the body schema, since the users may experience a deficiency after removing the extension. This theme concerns how sudden shifts in body image and schema can induce or lead to playful experiences regarding the changes in the balance of the body [29, 93] (D1, D2) while some of the effects can be mitigated with playful behaviour (D3, D4, D5). Three sub-hemes constitute this theme: Explorative joy at onboarding, feedback for adjusting to the bodily extension and feeling extensionless.

While contexts such as game settings or parties may quickly render the movements of the extensions enjoyable and attractive, more serious contexts such as a lecture or work environment may not accommodate them (D1, D4, D5). Contextual changes or discrepancies mostly concern body image, and while contexts that may afford playfulness (e.g., a work environment with close colleagues) could allow wearers to express themselves more freely, non-playful contexts might make them too conscious of their body image. On the other hand, related to body schema, the bodily feelings induced by extensions might boost the experience of playful contexts (e.g., playing a board game). Furthermore, if the lived experience of context does not fit the sensations induced by extensions (e.g., unintended activation while listening to a lecture), they may cause distractions. Thus playful experience is strongly tied to being in the magic circle of play [62, 107]. Three sub-themes were generated corresponding to this main theme: Lived experience of playful and serious contexts, multiple wearers in the same context, and sudden transition among contexts.

Identity has already been a subject for research on wearables, as they are closely connected to self-expression, fashionability, personal styles and the ability to directly alter the wearer’s bodily perception [50, 121]. It also has a close connection with role-play and character identification [27]. Designers similarly mentioned instances where their extensions led to transformations in identity (D2, D4, D5), mostly through internal bodily experiences. These transformations suggest that the extensions’ effect on the body schema was also quite influential. This modification in identity is also closely related to role-playing [27, 66] and bodily enactment [25, 67] for example through material properties [132]. This theme consists of two sub-themes: Transformation through movement, and attitude change induced by the behaviour of bodily extension.

Our investigation generated themes that associate playful bodily extensions with social play, identity transformation as in role-playing and experiences relating to different levels of agency. These themes also interplay with the contextual perception of body image and different phases of incorporation into body schema. In this section, we elaborate on how the design of bodily extensions is related to body schema, body image and playful framings.

Regarding social play, we point to the opportunity to consider playful bodily extensions for tighter social interaction—for example, to tease others physically with the extension. Tight social interaction, which was not present in the projects we analysed, has been identified as a possible dimension of wearable play [26]. The social wearables framework elaborates on social play [35], focusing on social interaction on a broader level by emphasising the differences among the controllers of actuators (e.g., self, others, both), intentionality (intentional, non-intentional) and perception by others. While the projects here can be mapped to these different dimensions, our work extends this prior work by focusing on in-depth lived social experiences of playful bodily extensions. We highlight methods such as exaggerating bodily expressions as in Wigglears (also Implication #1), creating an exclusive non-verbal bodily language as in Monarch (I#2) or positioning the extension itself as an entity that can be interacted with socially (I#4) as in Arm-a-Dine, as mentioned in Theme 1. We also add that these extensions’ exclusive communication might be exciting for the wearers but exclusionary for others (something that designers might need to take extra precautions for, as suggested by I#3).

Playing with identities was another highlight of our themes. Identity transformation is known as a valuable playful resource [27, 66], and we here point to how bodily extensions can bring about new ways of moving and proprioceptive sensations that make the wearer perceive their bodies in different ways. These experiences point to the outside-in and inside-out processes [53] of body image creation as dependent on the context (as mentioned in Theme 4), affecting other people’s reactions to the extensions. As the wearer might experience new ways to express themselves and intracorporeally communicate (inside-out) and reflect on their lived experiences of being with extensions in certain contexts (I#12) as in the Tail and Wigglears, they also feel others’ “judgement” of these extensions as different and misunderstood or funny (outside-in). When the extension is perceived and worn by others as well (as in the examples of Monarch or Arm-a-Dine), we also modify our bodily perception as an outside-in process that might then affect perceived social acceptability (I#14). The resulting differences in bodily perception can lead to the wearer altering their internal body image. As such, we can say that by altering their body image from the outside (for others), the wearer can alter their body image from the inside (seeing themselves as another being/character/avatar) [53]. For instance, wearing the Tail provides different movement possibilities that lead to an altered perception of one’s own body image, as explained in Theme 5. In other words, the body extension changes the wearer’s body schema. By doing so, it also alters the wearer’s body image and how they identify with themselves or with other imaginary characters, which also comes as a design opportunity (I#15). Previous studies have claimed that playful wearables and costumes might lead to increased identification with imaginary characters in role-playing scenarios [25], but embodied aspects of this change have rarely been explored. In our study, the identity transformation is a result of playing with the extended body, such as flapping the wings of the Monarch to induce an attitude change (I#16), adjusting to controlling the Arm-a-Dine robotic arm or exploring new ways of moving by wagging the Tail. We add to this prior work on playful wearables that playful bodily extensions can facilitate role-play through an inside-out process of proprioceptive alteration, as well as the outside-in process of changing the body image through others’ perception. We also highlight playfulness as a concept speaking to somaesthetic appreciation design [61]. The ways of extracting the playful nature of bodily extensions by altering bodily sensations correspond to several aspects of somaesthetic appreciation design. These aspects include turning inwards to make sense of any identity change with subtle guidance (as in body movements guided by the motion of the Tail) [61], and intimate correspondence in the example where the designer was bopping Monarch wings along to the rhythm of the music [61].

The design framework for social wearables also touches on different levels of agency, elaborated in Theme 2, such as who controls the wearable and how intentional it is [35]. Our systems speak to several underexplored parts of the framework. Arm-A-Dine fills an underexplored area of the design framework for social wearables since the robot arm is controlled by both the bodily movement of the wearer and the facial expressions of others (which is also one of the ways of creating the social play experiences with extensions as in I#6). This is a good example of how non-utilitarian playful explorations might yield designs that can fill underexplored design spaces. The Monarch system was also analysed by the social wearables framework and was interpreted as a wearable device that was controlled intentionally. Although this might be the purpose of the design, our interviews revealed that it is not always the case, and muscle contraction is a control mechanism that led to accidental activation. Such inadvertent activation might lead to opportunities (I#5) for play, as well as some frustration, depending on the context. Addressing this aspect might require easy on/off switches, as suggested in I#13 for not alarming players, as indicated in I#8. Previous work hinted that the different levels of agency on body-worn systems might be a source for playfulness [35], and we add to those studies by revealing the first-person experiences of our designers and communicating nuances of these experiences in Theme 2 and in I#5, I#6, I#7 I#8. We also recommend ways to create experiences of full agency over extensions (by exploring a wide range of interaction modalities such as different sound effects or haptics as in I#7, that would also help with full incorporation I#10).

We see that the incorporation process can be a source of playful interaction, but this process can also be supported by playfulness as indicated in Theme 3. The extensions support Rapp’s internalist style [103] because they explore and experiment with subjective experiences and intercorporeal communication, which has also been discussed as part of somaesthetic experiences emerging during bodily interactions with drones [69]. We corroborate this theory and extend it by adding that playful bodily extensions do not only support the internalist style, but can actively facilitate it by supporting novel bodily capabilities. While Wigglears and Monarch seek to augment the wearer’s perceptual and expressive capabilities, Arm-a-Dine explores mixing bodily expressions and reactions into one shared, interdependent experience. Designing playful behaviour for incorporating extensions and understanding the play embedded in the process are both important considerations in supporting the internalist style. In that sense, we suggest that the right amount of challenge might help with onboarding (I#9) and building up the internalist style interaction with extensions, while the possible feelings of being extensionless (i.e., breaking the internalist perception that was built up) during offboarding might be mitigated by adding playful behaviour to extensions (I#11).

Our work aims to help with designing bodily extensions by considering their experiential aspects and hopefully will lead to more awareness among designers regarding non-normative or minority bodily experiences [112] that can be induced by wearables, mechanical costumes or prosthetics. However, the merging of play and bodily extensions needs careful consideration of further unintended implications.

We acknowledge that playful bodily extensions can lead to gameful interventions. Gamefulness and gamification can have manipulative patterns [87, 120]. Related to bodily play, these deceptive patterns have been criticised in the scope of proxemic interaction and play before; that is, the reality of proxemics is ephemeral while the data about them is permanent [51]. Similarly, designing playful interactions and games around these extensions may lead to using them for extended periods, which can cause physical harm (such as strained muscles or posture problems due to the weight of extensions), similar to the mental and physical strain risks with head-mounted displays [129, 130]. Getting carried away with extensions can also cause harm to others since their movements can be unpredictable and agency over them might be limited.

Another potential risk is the creation of exclusionary games. On the one hand, designing playful activities and games with bodily extensions might seem like an opportunity for accommodating diverse body types. On the other hand, games and playful experiences designed through extensions might only be participated in by those who can afford to use those extensions bodily and economically. A similar pattern was mentioned in Theme 6.1 when it comes to creating exclusive languages through extensions, which might exclude those who do not wear them. Also, an unbalanced design of extensions in terms of capabilities might harm fair play, similar to the debates about the blade-shaped running prosthetics deemed to render athletes super-abled [52]. Possible discrimination between the people who have and do not have extensions has also been one of the main criticisms of transhuman ideals [24, 55, 59, 119].

If bodily extensions are not designed in the right way, they might hinder play at times. For instance, players may feel self-conscious about the change in their body image and thus be too embarrassed to express themselves through play. A similar effect can be observed when it comes to wearing costumes while playing tabletop role-playing games. While costuming might encourage some players to immerse themselves in the imaginary world, it can repel others who do not feel comfortable [27]. A similar negative effect might also occur with bodily extensions. Thus, designers need to understand the contexts where bodily extensions might be used; otherwise, their designs might harm the play of the experience (which was also highlighted in 6.4). Another possible disturbance to play might be caused by a sudden change in the body schema of the players. While those changes might be a resource for playful experiences as mentioned in Theme 6.3, if the activities and the surroundings of the play area are not designed carefully, large and heavy extensions such as the robotic arm in Arm-a-Dine might create fatigue (unintended fatigue needs to be avoided in bodily games [92]) and can lead to frustration rather than boosting play.