Two weeks ago I visited the University of Twente for a conference on Human-Technology Relations: Postphenomenology and Philosophy of Technology. Next to meeting inspiring people from a wide range of academic fields and attending some great presentations, I also presented some of my own thoughts and findings from a project I did last year. I edited my presentation into the text below.
The image shows someone wearing Bagsight, a backpack that guides a blindfolded person around a space with obstacles towards a destination. Before I get into more detail on this artefact and its consequences I will introduce three topics: Phenomenology of perception, guide dogs, and objects with artificial intelligence.
As might have been clear from the introduction, the context or design setting of this project is visual impairment. One of Maurice Merleau-Ponty’s most famous quotes describes a way of perceiving a busy city street: The blind man’s stick has ceased to be an object for him, and is no longer perceived for itself; its point has become an area of sensitivity, extending the scope and active radius of touch, and providing a parallel to sight 1. His point here is that, phenomenologically speaking, our mind and our image of self are not limited to our fleshy body, but this self can incorporate artefacts. In other words, the perception of the visually impaired person does not happen at the handle, but his body and mind end only at the tip of the cane.
In this context of visual impairments we quickly encounter animal companions. Both dog and handler have to be able to take the initiative and to respond obediently to the other. The task is to become coherent enough in an incoherent world to engage in a joint dance of being that breeds respect and response in the flesh, in the run, on the course. 2 Donna Haraway is discussing another dog, but her observation is all the more relevant with regard to guide dogs. I myself have done a set of interviews with people who navigate public space with a guide dog, and a trainer of these guide dog teams. Insights from that data include the agency in both the dog and its human companion. The dog is not used as a tool to scan the environment. Neither is the human an object that is dragged along by the dog. Both have their own agency in a ‘joint dance’. Following from this, I also observed a continuous negotiation between the two on where the next step will lead. Finally there is a division of tasks in the team. Contrary to popular belief, the guide dog is not responsible for navigation of the whole team, that is done by the human. The dog only takes care of local danger avoidance.
The third topic of introduction is a current development in the design of products. This project comes from the faculty of Industrial Design Engineering at TU Delft. In this world of product design we now encounter products with their own agency. Very crudely speaking: We can now create things, artifacts, physical entities that can display behaviour. This is not entirely new, factory robots have been around for a while, as well as the humble thermostat. What is new, is that the inner workings of these things are getting more complex, and their behaviour is becoming harder to understand. We are even inclined to say these objects have their own intentions. An illuminating example is the self-driving car that displays its intentions to the bystanders by blinking before making a right turn, or slowing down before a pedestrian crossing. Similar design considerations go into giving form to the behaviour of a Roomba vacuum robot.
There are voices in the design world to explicitly design these artifacts as agents. This would require us as designers to consciously think about and decide on the intentions and agency that is realised when the product is interacting with users and the environment.3 This could be a productive approach to giving form to the behaviour and would allow users and bystanders to make sense of what is going on. The research presented here follows in this line but is particularly interested in what it means from a phenomenological perspective to interact with a thing with agency.
We have some insights on the perception of one’s physical environment, we know a few things about perception through another agent, a guide dog. I also described the case for designing agency in objects. These three topics lead us to our question and a preliminary answer. How does one experience perception through another nonhuman agent?
To answer that question I designed and build an activated backpack for visually impaired people that I called Bagsight. This backpack, taking cues from guide dog team interactions and Braitenberg vehicles 4, moves on the back of the wearer. Following the analogy of the guide dog it is designed as an embodied agent that is continuously present in the activity of navigating through the environment. By tightening the cords it can move itself over the surface of the wearers’ back. Sensory information is taken from two distance sensors and two light sensors worn on the front of the shoulders. The moving behaviour is designed in such a way that Bagsight can be seen as being afraid of obstacles (moving away from them) and towards a given goal (moving towards this goal represented by a source of light).
Using Bagsight as a stimulus I conducted an experiment with 16 participants. These (non-visually impaired) participants were blindfolded and asked to move through a place with carton board obstacles towards a certain goal. During this experiment the participants were wearing the backpack and interacting with this artefact. A post-experiment interview on their experience was conducted. To put this data into perspective, Bagsight was introduced to these participants as a generic smart object for blind people. They did not know any of the design rationale, or the designers’ thought on agency or intentions. Finally, in the analysis, several themes were developed from this data.
So how does one experience perception through another nonhuman agent? In other words: What does it mean when we perceive the world not only through our own senses, not only through an extension of our own senses, but through the intentful interaction of an object with the environment and me?
Participants responded to the experience of wearing the backpack with quotes like the following: ‘It is sort of like your eyes, you could use to find your way in this space’. Other participants said: ‘As you walk next to something, the backpack pulls into one direction’. A third person replied: ‘If there is more light, the backpack is more secure of itself’. Another description of what was happening: ‘It protected me from crashing into the wall’. Structuring this interview data in several themes we find five categories or stances towards perception of the environment.
Perceiving the environment through an incorporated object. Above I described the cane taken up into the body image of the visually impaired user. The same thing occurred in the experiment with the backpack as well. Participants described their perception of the environment without directly relating this to the behaviour of the backpack. They used it as an extension of their sensory capabilities, e.g. ‘sort of like your eyes.’
Perceiving the environment through a tool. Several quotes emphasised how Bagsight functioned as a tool in their relation to the environment. They used it to estimate distances and to move around the space.
Perceiving the environment through an independent observer. These people described how they experienced the changes in the artefact. Interpreting these changes allowed them to make sense of what was going on around them. Bagsight was described as having its own loop of perception and action of which the wearer was an observant.
Perceiving the environment as being guided by a leader. In this case the participants ascribed a limited theory of mind to the object. They used anthropomorphic descriptions emphasising how they were guided around and following commands of the object. This stance is similar a popular way of describing the interaction with a GPS navigation device. ‘It wanted me to go in this direction.’
Perceiving the environment as being accompanied by a buddy. These comments emphasised companionship. Both backpack and wearer shape each others behaviour. Lots of these comments had an highly anthropomorphic character and were similar to relationships in guide dog teams.
In the remainder of this text I will relate these different stances to different types of technology relations presented earlier by philosophers of technology. More specifically, I will make a comparison with the modes of being as described by Martin Heidegger and technology relations by Don Ihde.
Heideggers major mark on continental philosophy is Being and Time 5. Among other important matters such as the question of being, he introduces the distinction between things zuhanden and things vorhanden. Things can be ready-to-hand (zuhanden). They are found to be available as instruments, we act through these things on the world, the instrument itself fades in the background. Sometimes however, the object will occur to me as a thing in itself. It becomes present-at-hand (vorhanden). This happens most often when we are new to the use of this object or when the tool and thus our experience of zuhandenheit breaks down. My attention moves from the activity of tool-use to the self-contained thing. When I use a hammer, the hammer is not the center of my attention, instead it is the activity of fastening a piece of wood. The instrument is ready-to-hand. However, when the head of the hammer gets loose, and I miss the nail a few times, my attention shifts to the hammer. It becomes present-at-hand 6. It is easy to recognise these modes of being in the stances that participants took towards Bagsight. As an incorporated object the stimulus is ready-to-hand. In the second and even more so in the third stance however, participants perceived the environment through a tool and as an independent observer. A careful studying of the backpack becomes the primary activity. Bagsight is present-at-hand.
Don Ihde reflects on relationships between human, world and technology in his book Technology and the Lifeworld 7. He distinguishes (at least) four types of relationships. In embodied relations, technologies form a unity with a human being, and this unity is directed at the world. A classic example is my relationship with my glasses. I see through them. This can be represented as: (human-technology) → world. In hermeneutic relations we read the world through our technological instruments. We make sense of the world by carefully studying what we see on a medical scan, e.g.. Technologies form a unity with the world: Human → (technology-world). Alterity relations describe interactions with technologies where the world fades largely in the background. Examples of these interactions include getting money from an ATM. In schema: human → technology (world). The fourth type is the background relation where technologies only provide the context of human activity. The noise of a fridge is a good example. This is (usually) not experienced in itself: Human (technology/world).
Applying Ihde’s categories to the stances towards the backpack we discover that the embodied relation fits both the incorporated object and the tool use. We find a hermeneutic relation of reading the world through the backpack in the independent observer. The alterity relation can be found in both the perception as being guided by a leader and the companionship of a buddy. We do however discover that alterity relations give rise to higher degrees of anthropomorphism for which we as human beings have more specific descriptions. Our social finesse allows us to make distinctions between more hierarchical and more emotional relationships.
How does one experience perception through another nonhuman agent? To answer this question we studied a backpack that enables perception of a physical environment. We analysed how this case is a concrete example of human-technology relations. Participants had different stances towards the technology and the world and we could relate them to philosophical approaches to humans and technology. In short we can conclude that agency in objects fundamentally shapes our understanding of the world. Looking towards the future I see a challenge for both designers and philosophers of technology to answer the increasingly important question: How do we live together with technology that becomes agentive and social?
This presentation reports on my graduation project for the master’s programme Design for Interaction, Faculty of Industrial Design Engineering, Delft University of Technology. The project was carried out under supervision of Dr. Marco Rozendaal (TU Delft), Prof. Catholijn Jonker (TU Delft) and Prof. Pim Haselager (Radboud Universiteit Nijmegen)