Social Dance Experiences in Human-Product Interaction Design

The interaction between humans and products is an important part of how products are used and, as a result, being designed. Appropriated from a long tradition of cognitive psychology, designers are known to use mental representations in their interaction designs (van Dijk & van der Lugt & Hummels, 2014). However, there is upcoming more interest to not look at the purely cognitive aspects of interaction, but design from the living human body in interaction (Kirsch, 2013). More specifically, the focus is to design for experiences, seeing the body and how it moves as an essential part of being in the world (Höök, 2010; van Dijk & van der Lugt & Hummels, 2014). How designers can make use of embodiment in the design of human-product interaction is, however, a question that is not easily answered. Offered possibilities arise from the field of embodied cognition and focus on using felt-experiences through first-person-methodologies, movement, dance and choreography (Loke & Robertson, 2008; van Dijk & van der Lugt & Hummels, 2014). In this paper, this existing research will be extended by offering an account of the experience of socially partner dancing, social dance in short, specifically to answer the question of how concepts from social dance can be used in the design of physical human-product interaction using the case study of the interaction with a vacuum cleaner.

In this case, the practice of social dance is chosen for two reasons: (1) the author has a passion and expertise in social dancing, (2) an interesting parallel can be found between social dancing and human-product interaction: both are based on two entities interacting by using non-verbal communication. This is also what makes the use of social dance different than previous literature which considered dancing as having expert knowledge of movement and not knowledge of interaction. Social dances have been formulating and studying the ways in which people can move and communicate together naturally through movement for years. In addition, the redesign of the vacuum cleaner is chosen as an example to focus and apply the exploration of social dance in design on to be able to have a tangible example which is both in favor of the exploration itself and the readability of the paper. The used method is, therefore, following a research-through-design approach (Findeli, 2010; Frayling, 1993) in which the vacuum cleaning redesign process is the method of researching the broader question of the use of social dance in design.

Methodology
In the first part of the paper, the background on embodied design, and as an extension, first-person-methodologies, will be elaborated. In part II, the first-person-methodology is applied to the case of redesigning the vacuum cleaning interaction using concepts of social dance. Then finally, in the third part of the paper, the results will be discussed and extrapolated into the broader scheme of physical product interaction design.

PART I | BACKGROUND OF DANCE IN DESIGN

Cognition, bodies and movement in design
In recent years, a new way to think about minds, bodies and technologies is formulated in the theory of embodied cognition. According to embodied cognition, cognition is seen as an ongoing and situated activity (De Jaegher & Di Paolo, 2007; Suchman, 1987). ‘Activity’ is a keyword here, since it is argued that cognition is active, a verb. De Jaegher & Di Paolo (2007) prefer, for that reason, to use the term ‘sense-making’. They argue that an autonomous being is the centre of its experienced world and by acting with their body they can create meaning, or rather, make sense of the world (De Jaegher & Di Paolo, 2007). Using this phenomenological perspective, a perspective of experience and activity, on cognition means that people think by using their environment and the tools around them, they think with their bodies and know more by doing than by seeing (Kirsch, 2013). Technologies become, then, one with the body, because they help to perceive and understand the world, just as the stick of the blind become an extension of their body with which they can understand the world better (Kirsch, 2013; Merleau-Ponty, 1962). This view has implications on the design of the technologies that help people to make sense of the world. Using the theory of embodied cognition, designers can find a new perspective and designspace for the interaction between human and technology, because technologies are not anymore mere tools, but a way to experience and express.

More specifically, a growing field of expertise within embodied interaction design is movement-based-design. It entails a variety of different approaches to design for and from the moving body. There are cases, for example, where designers are making use of dancers or choreography in the design process to design better interactions and movements (Loke & Robertson, 2008). All approaches have a shared commitment to using their personal experiences as sensing, feeling and moving bodies in their designs. And so, interaction is designed from movement, instead of taking technology as a starting point (Loke & Robertson, 2008). This focus on the movement abilities of users in interaction with technology allows an interaction that fits more naturally with the way people are accustomed to move their bodies in the physical world (Larssen et al, 2004).

First-person-methodologies
In this paper, a new way of designing for a smooth human-technology interaction from an embodied perspective will be explored using a first-person-methodology. In these types of methodologies, the body-in-motion and its felt-experience are the source for exploration of concepts in design for successful engaging interactions between human and technology (Loke & Robertson, 2008; Schiphorst, 2007). To design a better interaction between human and technology, one needs to understand how a smooth interaction feels. The author has the ability to use first-person-experience from three different angles in the redesign of the vacuum cleaner interaction: she is (1) a dancer, (2) a vacuum cleaner user, and (3) a designer. As a user, the (unpleasant) experience of vacuum cleaning can be explored. As a dancer new input from knowledge of the pleasant experience of a dance interaction can be described, which can be implemented in the design as a designer. As this account of social dance in design uses the personal experience of the author, she will use the pronoun ‘I’ from here on when describing the subjective practicing of social dancing or vacuum cleaning. These experiences can then form the basis on which a redesign for the vacuum cleaner interaction, in this case, will be made.

The author has been dancing and teaching blues and swing dances for over five years and has met and learned from many body (movement) geeks. For purposes of this paper, the social dance concepts of the author’s main dance form, bluesdancing, will be used. The bluesdancing tradition stems from a group of people who were actually unable to communicate verbally with each other and is more focused on improvisation and conversation than swing dances. Therefore, blues fits the purposes of designing for non-verbal human-technology interaction better. However, there are more social dances, such as latin-based dances, that are often using similar concepts that can probably be used just as well.

PART II | SOCIAL DANCE EXPERIENCES & THE VACUUM CLEANER

First, a bit of background about bluesdancing is in its place to be able to understand the experience of dancing blues and, with that, how the experience can be applied in design. What we nowadays call bluesdancing, is a set of idiom dances that emerged in the 1800s in the South of the United States by the Afro-American workers on the plantations. These dances were danced in small spaces, the dancers came from different parts of Africa and were not always able to communicate verbally with each other, the music was fairly simple, since there was no money for a lot of instruments and life on the plantation was both mentally and physically hard. The dance died out for a while and so there is not much knowledge of how the dance was danced exactly at that time. However, since its revival in the 1990’s, there is increasingly more consensus on what ‘bluesdancing’ is: what is inherent to bluesdancing in practically all opinions is that it is a dance that is meant to be social, improvisational and expressive.

It means that, nowadays, when you follow a bluesdance workshop, the focus is on learning how to connect with your partner and the music and how to communicate with each other using movement. So the gross of classes is not about learning specific steps, but about learning technical concepts of posture, movement, tension, different rhythms and ways of being creative and expressive in your dance. As a bluesdancing teacher, it is, as a result, a common exercise to find words and metaphors to describe the feeling of bluesdancing.

In this paper, I would like to explore a few of these concepts as my experiences of a pleasant social dance in light of the interaction between vacuum cleaner and user with the goal to improve this interaction. I will start with describing how a bluesdance flows in its movement, after which I’ll explain how the flow is the basis of the dance being improvisational and how this, finally, ensures an interaction between two autonomous agents that are leading and following each other. Through these concepts I will come to a definition of what I would call a ‘dancing interaction’ between human and product.

Flow of movement
I’m usually not really looking forward to using the vacuum cleaner, because it is a hassle to get it out of the storage closet, plug-in and move the thing with me as I am vacuuming the room. There are a few places in my room where the vacuum cleaner usually gets stuck: behind the legs of my bed and on the doorstep into my hallway, for example. Unfortunately, my vacuum cleaner is not of high quality, which means that when this happens, I often accidentally jack the tube out of the machine. Eventually, I started to put the vacuum cleaner in the middle of my room by hand and was then vacuuming around it, moving the vacuum cleaner as little as possible. However, this new strategy brought its own problems with it: I was getting horribly stuck in the cables and tube that turned around me, the vacuum cleaner and each other.

In conclusion, vacuum cleaning your room is not only an unpleasant activity because it is a mundane household task, but also because the interaction feels unnatural, jerky and annoying. In other words, it is not smooth. This is in stark contrast to my experience as a social dancer. Even in interaction with a dance partner, my movement keeps on flowing smoothly throughout the dance both as a leader and a follower: there are no breaks or sharp corners in the movement. When looking at my center of mass from above, my movement would draw the type of patterns as shown on the left side of figure 2 in contrast to the right side, where the movement is not flowing, but jerky. When the dancers find this flow in their partnership, the dancing feels right: as if it matches the other person and enables the dancers to become one with each other and the music.

This concept of ‘flow’ is extremely important in social dancing because understanding and having the flow improves leading, following, and improvisation. All three are an extension of the flow of movement in the body. What I often see beginner leaders do is sending their followers away and then catching them, breaking all their movement before sending them out again. At the same time, beginner followers often think from move to move, so they wait to be send out and then when their movement is over they stop moving, waiting for the next move. Ideally, a leader would align its own flow with the flow of the follower to lightly direct it from there, the follower keeps on flowing to wherever it is going, thus creating a continuous and smooth movement. In turn this means that leaders do not have to be a master in remembering a whole array of dance moves, but can let the flow inspire them to go in certain directions, speeds and positions, creating improvised dance moves in the moment. In figure 3, an overview of an example partner dance from above could then look like this.

To take a side-step back to embodied design research, the idea of ‘flow’ has been examined before in different practices and with different terminology. Jensen, Buur and Djajadiningrat (2005), for example, design tangible interfaces using human skilled action instead of mental representations. One of the qualities described is ‘wayfaring’, which refers to the idea that people move along a line instead of from point to point. Like ‘flow’ in social dance, ‘wayfaring’ describes the user interaction as a constant movement and not an interaction that consists of task after task (Jensen, Buur & Djajadiningrat, 2005). In addition, Isbister and Höök (2009) introduced a use quality that describes this type of interaction and called it ‘design for suppleness’. They define ‘suppleness’ as an interaction that is based on subtle social signals, emergent dynamics and moment-to-moment experiences to create a social/emotional dance between system and user (Isbister & Höök, 2009). Just as ‘flow’, ‘suppleness’ is based on subtleties and improvised movements and dynamics. The idea of some kind of ‘flow’ in design is, thus, not completely new. However, what makes ‘flow’ more interesting for me as a designer is that it is not just a description of movement or a quality of an interaction, but that it is a starting point on how to move in interaction. From the idea of ‘flow’, I will build further into a dancing interaction.

Improvisation
Blues, like many other social dances, is an improvisational dance. It means that you can dance with anyone to any music at any moment and create a stylistic and expressive dance together. There are no fixed steps or patterns that you need to learn to be able to dance. It is, in my experience, therefore really dancing, instead of moving to music. It is completely different than other dances in which first a choreography is created and then it is executed on music. Blues is a reaction, an expression or translation, to the music and situation the dancer is in, done in a split second. To be able to do it well, you need to have this sense of becoming one with the music and the other, tapping into your immediate reactions and stop thinking about a next phrase or move. I often tell my students to stop thinking, but start feeling. It is a being in the moment, disregarding all the rules, without breaking any rules.

In embodied cognition literature, this notion of acting as reaction in specific situations is called ‘situatedness’ or ‘situated actions’ (Suchman, 1987, Van Dijk & Van der Lugt & Hummels, 2014). Suchman (1987) takes the position that all actions, even planned actions, are situated. Simply because the circumstances of our actions are never fully anticipated beforehand (Suchman, 1987). People move through life improvising from moment to moment. However, the amount of situatedness depends on the nature of the activity and the degree of expertise. In social dancing, I teach beginners first a few moves, common patterns, that they can execute according a fixed plan. Then during the course I teach them at what moments they could start the common patterns and finally I teach them to come up with their own patterns inspired by the specific situation they are in. I encourage them to become authentic in their expression of the movement, to transcend rules and patterns and start tapping into who they are as existing beings.

When using the vacuum cleaner, the same notion can be found. I am not, before getting the vacuum cleaner out of the closet, deciding what corner of the room I will start in. Perhaps the first time in a specific room I think about some difficult spots that I should not forget. Usually, I simply grab the cleaner, and start cleaning guided by the flow of my movement and the specific (dirty) corners that attract my attention. Vacuum cleaning my room is just as much as social dancing an improvised and situated activity.

Leading & following
Then what exactly can the use of social dancing in design offer to create a new kind of interaction? It is the concept of leading and following, which combines both ‘flow of movement’ and ‘improvisation’ (together with other concepts beyond the scope of this paper) into a situated and expressive dialogue between dance partners and music, which I label as conversational dancing. This non-verbal conversation is what makes social dancing into an actual interaction, that will form the basis of a ‘dancing interaction’.

To ensure a smooth situated interaction and flow in the dance, social dancers use a couple of concepts to clearly, but gently, lead and follow. First of all, it is important to note that both leading and following are an active job. The follower chooses to follow and respond to the leader (DeMers, 2012) and is not merely moved around by the leader, since that wouldn’t constitute a dance but a dragging around of a doll (De Jaegher & Paolo, 2007; Merleau-Ponty, 1962). Unfortunately, on the dancefloor I sometimes encounter leads who think that they need force to move me around, disregarding my choice to follow them autonomously. It often results in an unpleasant dance in which I am not allowed to let my movement flow and I am being jerked from side to side and with that, jerking the leader with me as well. We call those leads ‘arm leaders’ (they often use their arm muscles to move their follows around) in opposition of the principle on which social dances are based: ‘leading by example’. What is meant with this principle is that I, as a leader, use my own body direction and movement to suggest the follower where and at what speed to go. As seen in the idea of ‘partnerflow’, the leader moves with the follower to redirect it from there. In most humans this happens quite naturally in our daily lives. When we walk hand-in-hand with someone, you often automatically lead and follow each other smoothly using the principle of leading by example. In learning to dance socially, you need to become more conscious of this principle. When suggesting movements that are not as plain as walking in a certain direction, it is important to be very clear in your body language. Learning this principle, it can feel unnatural, because you are consciously exaggerating your body directions. In time, however, you learn to use more subtle movements to lead and follow, allowing the dance partners to instantly understand each other in their movements.

A social dance is, in that sense, an example of a social interaction as explained by De Jaegher & Di Paolo (2007). They define a social interaction beyond a view of a simple coincidence of two agents who influence each other. Instead a social interaction is an emergent and autonomous process that has two characteristics: (1) there is a coupling taking place which makes the interaction itself temporarily autonomous, but (2) at the same time the individuals involved are and remain autonomous as interactors in the process. Combining these two characteristics, it can be seen that an interaction is not reducible to individual actions or intentions, but creates a relational domain that has its own properties constraining and modulating individual behaviour within the interaction (De Jaegher & Paolo, 2007). A social dance is also not just a coincidental influencing of each other, but a joint sensemaking of the music and the movements. It is an interaction in which both dance partners keep their autonomy, but in interaction create an overarching autonomous domain: the dance itself.

In case of the vacuum cleaner, it is not following actively. It is, in fact, passively being dragged around by the force of the user, creating the jerky interaction. The vacuum cleaner is not an autonomous interactor in the interaction. As a vacuum cleaner user, I also don’t expect it to, because I have grown accustomed to this kind of interaction. It is, however, not pleasant, because it is a dragging around instead of an interaction. On the other side of the spectrum, there is the robot vacuum cleaner, such as the Roomba. These vacuum cleaners operate completely autonomous: at certain times of the week or day, it leaves its charging station and finds its way through the living room, vacuuming it completely. As a user, I don’t even have to be present. However, if I would be present, it would react to me. It would see me as another object to be circumvented. Hence, even though the robot vacuum cleaner is autonomous in how it reacts to me, again: there is no interaction. It is a solo dance.

A ‘dancing interaction’
An ideal ‘dancing interaction’ with a vacuum cleaner would, therefore, be like a conversational dance, as an interactive dialogue between two parties: both contributing to the overarching situated process in a flowing movement. It should feel as if both parties are active in the main task of vacuum cleaning. In a quite literal sense, it could be a compromise between the robot vacuum cleaner and the “normal” one. It is autonomously following and flowing in the interaction. However, even a non-intelligent vacuum cleaner could be more autonomous in the interaction as a pole or stick in a dance could too. A pole dancer, for example, knows its pole within the interaction. The dancer knows where the pole is, how thick and smooth it is, and it is these properties of the pole in its known unmoving stability, that makes it autonomous in the interaction. Within the interaction of the dance, the dancer knows the role of the pole and she can push herself off of it and find it and grab it without looking for it.

Hence, in a redesign of a normal vacuum cleaner that has more of a ‘dancing interaction’ with its user, it is extremely important to make sure the vacuum cleaner has a stable and smooth flow in its movement. A flowing movement of a vacuum cleaner is, in my opinion, how a non-intelligent object that is moving around can become autonomous in an interaction, because flow, together with rhythm, is what people use to predict and understand movement. What makes this objective difficult for designers is, however, that at the same time, the activity of vacuum cleaning is a situated action: it is improvised. And so, the task at hand is to create a predictable, flowing moving vacuum cleaner in a non-defined order and unknown situation. To be able to design for such a contradiction, social dance can offer a way for designers to understand these concepts and movements in their own bodies by actually dancing themselves.

Results: redesign the vacuum cleaner
In the process of redesigning the vacuum cleaner, I tried to come up with ideas to improve the vacuum cleaning activity by applying what I had learned in the embodied design research and my experience of social dance. During the ideation phase I repeatedly got up from my chair and started dancing, both imagining a vacuum cleaner with me and actually grabbing the vacuum cleaner out of the closet. As a result, I could imagine better how I would move around my room with a vacuum cleaner than just picturing it, which led to new ideas I wouldn’t have come up with otherwise. In addition, and this is possibly a more personal result, a break from sitting behind my desk was welcomed by my body and enabled me to think and feel better. It felt as if my design skills were enlarged with a new layer: the layer of movement in design. It seemed foolish that I had been designing for moving bodies all this time using just pen and paper. If we learn to make models to understand three dimensional designs better, then why don’t we learn to move to understand living human bodies better?

In the redesign of this vacuum cleaner, the functionality should stay the same, but the way of interacting and, as a result, the visual design could change. I came up with different design ideas ranging from slight to radical changes in interaction (see figure 4). A simple solution would be to get rid of all kinds of tubes and cables and make the cleaner move around easily using a one piece design. In that case, the interaction would be much more flowy, but not much of a dance yet. Another obvious solution would be to make the vacuum cleaner more active and intelligent by giving it a motor and sensors so it could follow its user more autonomously. However, I soon came to the conclusion that all these solutions still only involved arm movement and could actually be identified as ‘arm leading’. I, therefore, started to look at the movement of the whole body and how I could imply ‘leading by example’ in the design. I came up with a couple of solutions, on the right side of the image, that would involve either a combination of multiple body movements or different body parts than the arms.

The two-hand-maneuvering idea has been chosen to be elaborated further, because it offers an interesting way to explore the concept of ‘leading by example’ in the vacuum cleaning interaction. Since the user needs two hands to direct the cleaner, the user is less likely to only use arm movement. This effect can even be exaggerated by making the vacuum cleaner heavy, inviting the user to push its body weight into the product to move it around. The vacuum cleaner can keep its own balance and, just as a segway, moves by being pushed out of balance by the user. Once it is moving, it keeps on moving without new input from the user. The user can then redirect the vacuum cleaner into different corners of the room. In addition, the user can move the vacuum cleaner up and down by going up and down itself with its body. In figure 5, a storyboard of vacuum cleaning a room as a dance can be seen.

PART III | EXTRAPOLATION

Discussion
The case study of the vacuum cleaner showed how the described experience of social dancing could lead to new and interesting concepts for human-product interaction. The ideas ranged from ensuring a smoother interaction to creating a complete new experience of vacuum cleaning. By interpreting human-product interactions as a dance, the designer was able to not only create a vacuum cleaner interaction that is possibly more fun for the user, but is also more in alignment with how a user moves around during the activity, creating a theoretically more pleasant interaction. Theoretically, because in this paper, the new concept idea was not actually tested with different users to ensure the actual benefits of the ‘dancing interaction’ perspective in design. However, the aim of this paper was merely to explore how social dance concepts could be a way to give fresh input for designers according to the insights of embodied cognition and not to prove this specific design method. In the future, it could be interesting to move further into using social dance in the design process by non-dancers and examine the design results more in-depth.

The redesign case of the vacuum cleaner offered an example for which a dancing interaction could quite easily be imagined, because of the nature of the vacuum cleaning activity: it is an activity in which two agents work together on one task while moving around a specific space. It would be interesting in the future to do the same for technologies for which the characteristics lay further from a dance, such as navigating a smartphone or driving a car. In my opinion, it is possible to use a dancing perspective on any type of interaction. In time, when internalizing this perspective, it becomes easier to see interactions as dances and use the knowledge in the design process. However, some creativity might be needed to scale it down to for example only finger, hand and wrist movements and to start looking at interactions as one flowing movement, instead of different linked movements. Humans are not robots, they don’t move by a script that tells them to move the thumb to position x=4,y=8 on the screen and then move to the right along the x-axis. They just swipe right and from that swiping movement, the thumb is moving back in a circle to swipe again. It is not a ‘reset’, but it follows out of the first movement and being input again for the next movement.

Finally, as stated before, the designer needs to become quite familiar with the idea of the ‘dancing interaction’ to be able to apply it. As often stated in movement-based-design literature, it is practically needed for the designer to become a dancer itself as the author of this paper is. However, it is likely to be argued that it is not a method that would be suitable for every designer. That is why this author decided to choose specifically to use a first-person-methodology using an interesting practice that she personally had affinity with. There are, however, many more similar types of practices using movement interaction concepts that can be of use in designing for movement. It could be other types of dancing, or karate or something else. For Höök (2010), it was for example horseback riding. The point is, that as designers, it is not forbidden to use knowledge from your personal passions and amateur practices outside of design, it could even be favorable when done clearly and carefully. We design not for users, but for humans, and how convenient that designers are humans too. Personal knowledge can be of great value in the process of designing for people.

Conclusion
To answer the research question posed at the beginning of this paper of how concepts from social dance can be used in the design of physical human-product interaction, it can be concluded on three levels: (1) the social dance concepts of ‘flow’, ‘improvisation’ and ‘leading & following’ can be used to understand human-product interaction as a conversational dance to ensure a more fun and pleasant ‘dancing interaction’ between user and product, (2) to better design for movement-based interaction, it is helpful for designers to understand movement themselves in their own bodies, thus, to move themselves during the design process, and (3) on a higher level, it is useful for designers to broaden their design skills using different types of practices they have affinity with; why use only one toolbox when you actually have more at hand? All three combined, this paper showed using the case study of the redesign of the vacuum cleaner, that design can use the understanding of personal experiences of social dancing to create or redesign physical products from a new perspective. A perspective in which all interactions, even the design process itself, can be seen as graceful and pleasant dances.

REFERENCES
De Jaegher, H., & Di Paolo, E. (2007). Participatory sense-making. Phenomenology and the cognitive sciences, 6(4), 485-507.

DeMers, J.D. (2013) Frame matching and ΔpTed: a framework for teaching Swing and Blues dance partner connection, Research in Dance Education, 14:1, 71-80, DOI: 10.1080/14647893.2012.688943

Findeli, A. (2010). Searching for Design Research Questions: Some Conceptual Clarifications. In Questions, Hypotheses & Conjectures: Discussions on projects by early stage and senior design researchers (pp. 286–303).

Frayling, C. (1993). Research in Art and Design. Royal College of Arts: Research Papers 1:1.

Höök, K. (2010). Transferring qualities from horseback riding to design. In Proceedings of the 6th Nordic Conference on Human-Computer Interaction: Extending Boundaries (pp. 226-235). ACM.

Isbister, K., & Höök, K. (2009). On being supple: in search of rigor without rigidity in meeting new design and evaluation challenges for HCI practitioners. In Proceedings of the SIGCHI conference on human factors in computing systems (pp. 2233-2242). ACM.

Jensen, M. V., Buur, J., & Djajadiningrat, T. (2005). Designing the user actions in tangible interaction. In Proceedings of the 4th decennial conference on Critical computing: between sense and sensibility (pp. 9-18). ACM.

Kirsh, D. (2013). Embodied cognition and the magical future of interaction design. ACM Transactions on Computer-Human Interaction (TOCHI), 20(1), 3.

Larssen, A. T., Loke, L., Robertson, T. J., & Edwards, J. (2004). Understanding Movement as Input for Interaction-a study of Two EyeToy (tm) Games. In Australian Computer Human Interaction Conference. Bluora and Cedir.

Loke, L., & Robertson, T. (2008). Inventing and devising movement in the design of movement-based interactive systems. In Proceedings of the 20th Australasian Conference on Computer-Human Interaction: Designing for Habitus and Habitat (pp. 81-88). ACM.

Merleau-Ponty, M. (1962). Phenomenology of Perception. Routledge.

Schiphorst, T. (2007). Really, really small: the palpability of the invisible. In Proceedings of the 6th ACM SIGCHI conference on Creativity & cognition (pp. 7-16). ACM.

Suchman, L. A. (1987). Plans and situated actions: The problem of human-machine communication. Cambridge university press.

Sundström, P., & Höök, K. (2010). Hand in hand with the material: designing for suppleness. In Proceedings of the SIGCHI Conference on Human Factors in Computing Systems(pp. 463-472). ACM.

Van Dijk, J., van der Lugt, R., & Hummels, C. (2014). Beyond distributed representation: embodied cognition design supporting socio-sensorimotor couplings. In Proceedings of the 8th International Conference on Tangible, Embedded and Embodied Interaction (pp. 181-188). ACM.