Re-thinking the interface: What is the meaning of Alan Kay's"Doing with Images makes Symbols"

Alan Kay created a slogan for development of user interfaces: Doing with Images makes Symbols. This reveals the essence of the Graphical User Interface (GUI) which millions of people now use all over the world. This paper examines how GUI, based on Kay's slogan, has changed our communication with the computer. In order to explore this topic, I focus on Azuma and Turkle’s argument of "interface value" intuit for thinking about the relationship between our action and new sign surfaces. Next, Kay’s slogan is considered from the viewpoint of its formation process. Finally, I refer to Sperber and Wilson's Relevance theory. I conclude that GUI demands that we, along with computers, act for ostensive-inferential communication in a mutual cognitive environment.

What You See Is What You Touch? Double description of seeing and touching in ‘Display Acts’

What you see is what you touch?

View more presentations from Masanori MIzuno.

There is, in fact, almost no formal theory dealing with analogue communication and, in particular, no equivalent of Information Theory or Logical Type Theory. This gap in formal knowledge is inconvenient when we leave the rarified world of logic and mathematics and come face to face with the phenomena of natural history. In the natural world, communication is rarely either purely digital or purely analogic.1

This is Gregory Bateson’s words. I consider that user interface has the same problem. Computer is digital. Computer is controlled by formal theory. However, We, human, is not purely digital. We, are analogic rather than digital. Therefore, user interface should be the place where purely digital computer becomes analogic one, and analogic human becomes digital one in order to communicate each other.

Display Acts

When we use a computer, what do we do? Almost all of us look at some images on an electric display, grab and move a mouse, and type on a keyboard, then our right hand holds the mouse in order to point to an image called an icon on the display. This is very 'natural' for us. If our body makes some actions with a plastic object, then the images on the electric display change. However, this relationship between our body and the image did not exist until the computer, and especially until the Graphical User Interface (GUI), appeared. I call this phenomenon 'Display Acts': action which is formed by connecting our body action with the change of images on the electric display.2 However, we do not understand this action well because we have only separately studied body actions and images in man-computer communication.

In order to make clear the relationship between our body action and images in the man-computer world, this presentation focuses on the cursor in the electric display and the mouse in the real world. We see the cursor in the electric display: The cursor is "a small mark on a computer screen that can be moved and that shows the position on the screen, where, for example, text will be added."3 Moreover, the shape of cursor is almost arrow (→) which is "used to show direction or position."4 We just see the arrow cursor in order to know where it is and point some images on the display. Then, we also hold the mouse.

Exemplification

What does the cursor show us? It shows us the position and direction, however, the cursor often points nothing. When the arrow cursor points nothing, it does not work. We don't care this. Why? Because we not only see the images in the display, but also grabbing the mouse. What is a relationship between seeing and grabbing? In order to make clear this relationship between the cursor and mouse, I refer to Nelson Goodman’s idea; Exemplification. According to Goodman;

Exemplification, though one of the most frequent and important functions of works of art, is the least noticed and understood. Not only some troubles about style but many futile debates over the symbolic character of art can be blamed on ignoring the lessons, readily learned from everyday cases of the relation of being-a-sample-of, that mere possession of a property does not amount to exemplification, that exemplification involves reference by what possesses to the property possessed, and thus that exemplification though obviously different from denotation (or description or representation) is no less a species of reference.5

Possession means that some samples show us their own properties, and we interpret their properties for our interests. The user interface is not work of art, but I consider that the images in the electrical images with the mouse or other input devices show us almost same functions of work of art.

We see the shape of cursor as a sample in order to know functions of cursor. We get a pointing function from the arrow shape, therefore, we label this pointing function to the arrow cursor. However, the arrow cursor points nothing. As a result, we think this arrow possess another functions; Maybe it is moving for pointing because the arrow is shape for pointing. Only when we see the electric display, we can understand that the functions of the cursor is only pointing. However, user interface has the mouse. We touch and grab the mouse, then we see that the cursor moves and point the image. This touching is another function of the cursor because the cursor is designed for connecting with the mouse in order to work in the display. Only when we see the display, we can just know pointing and moving functions of the cursor (fig. 1). Only when we touch the mouse, we can understand the cursor is able to touch the image and change images in the display (fig. 2). Only pointing can not change anything in the display. Changing needs touching. In short, we not only see the images on the display, but also grabbing the mouse when we use the computer. Therefore, we have to consider what a relationship between seeing and touching is.

　　　　　　　　　　　　fig.1

　　　　　　　　　　　　fig. 2

Moreover, the arrow cursor sometimes suddenly changes into the spinning wait cursor or something else. When we see transformations of the cursor's shape, the mouse we grasp does not change. However, the moment the arrow cursor changes into the spinning wait cursor, our action with the mouse changes. Before transformation, we use the arrow cursor and the mouse in order to point to an image on the display, and therefore the cursor exemplifies the function of pointing and our body action with the mouse is formed for pointing. After transformation, the spinning wait cursor exemplifies the function of, not pointing, but showing the computer status. As a result, our action with the mouse is not formed for pointing. After a moment, we realize that what we can do with the mouse in hand is just moving the cursor in the electric display. The spinning wait cursor exemplifies that the computer is still working, but it does not leave us alone; the cursor and the mouse are connecting, please wait. Throughout the whole process, we grasp the same plastic object, even though the cursor's shape is changing.

Double Description

In user interface, the image we see is changing while the object we touch is not changing. Moreover, this transformation of image is sometimes out of our control. But we naturally accept this phenomena. In order to make it clear, I refer Bateson’s Double Description. He quotes Shakespeare’s Macbeth and says that:

This literary example will serve for all those cases of double description in which data from two or more different senses are combined. Macbeth “proves” that the dagger is only an hallucination by checking with his sense of touch, but even that is not enough. Perhaps his eyes are “worth all the rest.” It is only when “gouts of blood” appear on the hallucinated dagger that he can dismiss the whole matter: “There’s no such thing.” 6

We need Double Description of seeing and touching in order to consider and prove what the user interface is. However, user interface mainly focuses on seeing. For example, What You See Is What You Get; WYSIWYG. From Wikipedia, the main attraction of WYSIWYG is the ability of the user to be visualize what he or she is producing.7 Here, ‘to be visualize’ is emphasized. One more example. Ben Shneiderman proposed a very important idea for user interface: Direct Manipulation. From Shneiderman, Direct Manipulation’s central idea is to be visibility of object of interest.8 Here, only visibility is emphasized.

I want to focus on Direct Manipulation from anther point of view. According to George Lakoff and Mark Johnson. Direct Manipulation leads us not only seeing but also touching because we see and touch the object in order to create something. And they say that the Direct Manipulation makes the concept of CAUSATION.9 Furthermore, Søren Pold connects GUI with the concept of CAUSATION. He says that WYSIWYG interface has many buttons, and buttons are essential part of controls in GUI10 and also continues that:

When pushing a button in a interface ─ that is, by movement of the mouse, directing the representation of one’s hand onto the representation of a button in the interface and activating a script by clicking or double-clicking ─ we somehow know we are in fact manipulating several layers of symbolic representation and,…… It is a software simulation of a function, and this simulation aims to hide its mediated character and acts as if the function were natural or mechanical in a straight cause-and-effect relation.11

According to Pold, pushing button in the user interface makes a straight cause-and-effect relation in the electric display, although this cause-and effect appears by hiding manipulating several layers of symbolic representation.

Another view of the cursor and mouse. A media artist, Masaki Fujihata says computer does not have a question of materiality.12 This means that computer has different principle from us because we live in the material world. Fujihata considers that if the repeated experience are provided us, through interactive experience, Images are made into object.13 According to artist’s intuition, we go through new phenomena: Image are made into object.

What You See Is What You Touch?

However, this ‘object’ in the display is not real object. According to Pold, this ‘object’ in the display shows us cause-and-effect, therefore, we feel as if we directly manipulate it by our own hands. On the other side, Fujihata says that the ‘object’ does not have its own materiality. We, of course, see the ‘object’ but can we touch it? Maybe, yes. Maybe, no. I consider that this ‘object’ is something like “switch” as Bateson defined;

We do not notice that the concept “switch” is of quite a different order from the concepts “stone,” “table,” and the like. Closer examination shows that the switch, considered as a part of an electric circuit, does not exist when it is in the on position. From the point of view of the circuit, it is not different from the conducting wire which leads to it and the wire which leads away from it. It is merely “more conductor.” Conversely, but similarly when the switch is off, it does not exist from the point of view of the circuit. It is nothing, a gap between two conductors which, themselves exist only as conductors when the switch is on. In other words, the switch is not except at the moments of its change of setting, and the concept “switch” has thus a special relation to time. It is related to the notion “change” rather than to the notion “object.” 14

Like the concept ‘switch’, the ‘object’ is related to the notion “change” rather than to the notion “object”. The cursor is given a special relation to time by the mouse.

Why is this happen? Because the computer is purely logic machine which is made by analogic human in cause-and-effect world; “the if ... then of logic in the syllogism is very different from the if ... then of cause and effect.” 15 Bateson makes a question: Can logic simulate all sequences of cause-and-effect? His answer is maybe no.16 The reason is that “the if ... then of causality contains time, but the if ... then of logic is timeless. It follows that logic is an incomplete model of causality.”17 However, we want to re-made the computer as cause-and-effect machine via Double Description of seeing and touching. Introducing the special relation to time with the mouse, user interface makes new images with coupling cause-and-effect and logic in the electric display. We are making pseudo-cause-and-effect with logic in the user interface. Therefore, we touch the same plastic object, even though the ‘object’ in the display is changing.

                                             fig.3

Returning to the cursor and the mouse. The cursor looks like showing cause-and-effect, however, it stands for the logic in the computer: there is no time. But, we live in time: cause-and-effect world. Pold tells us that the user interface make us believe there is a straight cause-and-effect relation, even though this cause-and effect appears by hiding manipulating several layers of symbolic representation. On the other hand, Fujihata says that if we see cause-and-effect behind the image, then we can touch it. Almost all us do not understand what the logic in the computer is, but we use it with the display, the keyboard, the mouse, and something else. The reason is that we believe that the computer is not only logic, nor only cause-and-effect, but just ‘something’. Therefore, when we consider about user interface, we should not confuse ‘if ... then of logic’ with ‘if ... then of cause-and-effect’. We see the change of image in the electric display; it shows us the logic in the computer but we feel something like cause-and-effect, which is made from computer’s pure logic. We touch the same object while the image is changing: our grasping mouse introduces the special relation of time to the computer and makes pseudo-cause-and-effect in the computer (fig. 3). Therefore, only seeing is not enough and only touching is neither. In order to made clear ‘something’; the pseudo-cause-and-effect with logic, we have to consider the double description of seeing and touching in the user interface. In short, the computer has changed the relationship between our seeing and touching. It means that the computer changes our notion and experience for touching object: Touching not “object” but “change”.

References
1 Gregory Bateson, “Steps to an Ecology of Mind”, The University of Chicago Press, 2000, p.291.
2 Masanori Mizuno, “The formative process of “Display Acts” on the establishment of GUI”, Doctor Thesis,2009, p.3.
3 Oxford Advanced Learner’s Dictionary (electric version), Oxford University Press, 2000
4 Ibid.
5 Nelson Goodman, “Way of Worldmaking”, Hackett Publishing Company, 1978, p.32.
6 Gregory Bateson, “Mind and Nature”, Wildwood House, 1979, p.73.
7 Wikipedia, http://en.wikipedia.org/wiki/WYSIWYG (9, April 2009 access).
8 Ben Shneiderman, ‘Direct Manipulation: A Step Beyond Programming Languages’ in Noah Wardrip-Fruin
and Nick Montfort ed., “The New Media Reader”, MIT Press, 2003, p.486.
9 George Lakoff & Mark Johnson, “Metaphor We Live By”, The University of Chicago Press, 2003, pp.75-76.
10 Søren Pold, ‘Button’ in Matthew Fuller ed., “Software Studies: a lexicon”, MIT Press, 2008, p.31.
11 ibid., p.33.
12 Masaki Fujihata, “Masaki Fujihata: The Conquest of Imperfection ─ New Realities Created with Images and Media”, Center for Contemporary Graphic Art, 2007, p.178.
13 ibid., p.178.
14 Bateson (1979), pp.108-109.
15 ibid., p.58.
16 ibid., p.58.
17 ibid., p.59.

Origina text: Proceedings of 4th International Symposium on Multi-Sensory Design

The (pseudo) causal chain

   http://zig.media.mit.edu/Work/G-stalt

In last entry, I considered why a gesture interface like G-stalt is studied in Tangible Media Group. Prof. Ishii says that:

Interfaces that control the computer with a mouse are like a black box, and it's difficult to understand what's going on inside. If gestures are used, however, one can clearly see the causal chain, as with a violinist's response to the conductor's baton. Research into gesture interfaces is attempting to hammer out a new paradigm by seamlessly merging body, objects, and space."

MIT Media Lab Tangible Media Group's Gesture Interfaces: Dynamic Interaction in Hybrid Space in AXIS 2009.8 vol.140

Although G-stalt seamlessly connects body, objects, and space, why the image is not in this relationship? The thing that is moved by the gestures is not the object itself but the image on the display, isn't it? Are 'body, objects, space and images' seamlessly merged in G-stalt? However, Prof. Ishii does not say the image. Why? Maybe, the images have already become the objects in our life. If we think that, I can smoothly connect the gesture interfaces with 'Radical Atoms' which is next vision of TMG. Now, we rarely think that the image is the 'image'. The image is something like the object, so it is tangible. Therefore, we can feel that the image is the object and the objects is the objects at the same time now. This idea makes totally new paradigm. This paradigm may be related with plasticity, I think. 

I want to focus on Prof. Ishii' word " If gestures are used, one can clearly see the causal chain". I consider that the mouse and the cursor on the display make "pseudo-cause-and-effect" between our action which is ruled by physical cause and effect and computer logic which is not ruled by one. Then, The pseudo-cause-and effect changes the images on the display into the 'object [entity]' which is something tangible. Even though the image which we see on the display has already become the 'object', it is supported by the pseudo-cause-and-effect. However, the gesture may interfaces remove  'pseudo' from the pseudo-cause-and-effect because the gesture is closely connected our body. Therefore, Prof. Ishii is maybe right. Our action will be tightly connected with the image on the display because the gesture is very close to our body. In the mouse and cursor, 'to grab and move something' is transformed into  'to point and move something' and this transformation makes the pseudo-cause-and-effect on the display. 'Grab' and 'point' are very general actions for us, then we can do many things with the combinations of these two actions on the display. As result, there is the ambiguous relationship between the image on the display and our action. However, the gesture requires one to one relationship with the image on the display. This one-to-one relationship make us think why an action is connected with the change of image on the display. Connecting our gesture with the image on the display in one to one relationship, the image may be more close to the object. Then, may the ambiguity of our action be reduced in turn?  

Atom → Bit → Pixel → Atom

I feel something strange that Tangible Media Group(TMG) studies the gestural interface which is not like the tangible. And more, TMG leaded by professor Hiroshi Ishii shows a new vision 'Radical Atoms'; Study the interface based on the material which can freely change like the pixel on the display. There is a stream of idea; Atom → Bit → Pixel → Atom. We have touched Atom for a long time. Then, we have to touch Bit like touching Atom. We have to touch Pixel like touching Atom. In the mean time, we have to touch Atom like touching Bit and Pixel. The first Atom does not change. It is just solid material. However, the second Atom can freely change. It is like fluid material. Now, we can change Bit and Pixel at our will but Atom does not. Prof. Ishii feels strange this solid Atom, therefore he give us the vision of 'Radical Atoms' for study the interface based on the idea that we can freely change the Atoms. I think that the second Atom is the new entity based on the appearance of Bit and Pixel which is always changing. Bit and Pixel is freely changing therefore we have to make a new interface in order to correspond them. It is g-stalt. This gestural interface make us re-think our actions to Bit and Pixel. This thought leads us new actions for the new fluid Atom. I think that this is 'Radical Atoms'.  In this way,　I understand why TGM studies the gestural interface like g-stalt.

You are a game controller

Project Natal Xbox 360. "You are a game controller." You don't need any controller. You can control the image on the display via your gestures. Talking of the gesture for game, I have Nintendo Wii which needs Wii remote for playing the game. We normally do actions with something physical like a racket for table tennis or a sword for Kendo. In fact, I can easily play those sports in Wii Sports resort. How about playing them without something physical to keep holding?  I think that I would have the strength for it. Although we hold Wii remote in order to play Wii games, it is different between the feelings for hitting a ball with a racket and for hitting a ball image on the display with  Wii remote which gives us sound and vibration. However, many people do not care this difference and feel as if Wii gives us 'real' feelings. We do not hold anything physical in Natal, therefore we can not feel some vibrations. We have to control our actions based on only the image on the display and sound. The sense of sight and hearing controls our whole body actions.

I consider about the feeling of swishing the air in soccer when I watch demonstration movie of Natal. It is like something empty. When try to kick the ball which is the physical entity, my body gets some tenses for a feedback from the ball.  Then, I miss my kick. Our body tenses can not go to the ball and return to my own body. If I understand that there is nothing physical contact, my body do not get any tenses. As a result, we can do our actions without any body tenses because there is nothing physical feedback in Natal. Maybe, we use our sense of sight and hearing as a substitution for our body tenses to the contact of something physical. Even though we do the actions in Natal as same as the ones in the physical world, how our brain reacts in this situation? If we keep doing this no physical feedback actions, our body will store new kind of experience and what will happen for us.

Writhing this entry, I remember that I did a practice swing without a racket many times when I belong to a tennis club in a junior high school.  Maybe, this practice teaches me 'form' for hitting the ball in the tennis. However, there is the difference between a practice swing with a racket and without racket because our body tenses change for the weight of it. 'Form' reminds me of 'NOU' or Pantomime. Natal and Pantomime. However, there is no 'weight' or physical 'resistance' in Natal in comparison with good pantomime. Anyway, we have to play long time in no physical feedback environment like Natal long time in order to understand its effects for us.

Magic Mouse Scroll with momentum.

Magic Mouse Scroll with momentum. iPhone does the scroll with momentum and it is comfortable for us. Magic Mouse, too. iPhone and Magic Mouse do the same scroll with momentum, therefore our finger action for them and the movement of the image on the display are also the same. However, I feel something different between iPhone and Magic Mouse. With iPhone, our finger move lightly on the smooth glass surface and the image on the display sticks to our finger movement. There is Smooth and Sticky. With Magic Mouse, our finger move along the gently slope on the sticky plastic surface and the image on the display sticks to our finger movement. There is Sticky and Sticky. Firstly, I think that "Smooth and Sticky" and "Sticky and Sticky" are a big different between iPhone and Magic Mouse, then the slow slope of Magic Mouse would be important for use because I feel very comfortable when my index finger go and come along it. However, a biggest different between them is that we touch the image directly in iPhone even though our finger and the image separate  in Magic Mouse.

Anyway, this scroll with momentum brings the momentum law of our physical world into the logical world of the computer, although the momentum in the display is generated from the logic. After going through many world, we feel good when the image on the display represents the law of our physical world which we are familiar with. There is something different, isn't it? The computer can not represent our whole physical world. It just cut out only the momentum from the physical world and gives it to the movement of the image. Moreover, the movement of the image with the momentum is adjusted for us in order to fit our finger action. As a result, we feel that something new happens, even though the image on the display represents just the law of physical world  which we are familiar with.

At first, I consider about a relationship between the scroll with momentum and the shape of surface; the slow slope of Magic Mouse fits our daily action for the slope because we get the momentum for going up the slope and the momentum is given us when we go down the slope. Therefore, the slope of Magic Mouse compensates the sticky plastic surface. However, I don't know whether our body feeling for the slope is reflected in our finger action for the scroll with momentum of Magic Mouse, even though the image on the display fits our finger action with the momentum which is cut out from the physical world. Although the slope of Magic Mouse may be just the shape for easy to grab in practical manner, it is the fact that the plastic surface of Magic Mouse has the slope in comparison with the glass surface of trackpad is flat. Finally, I understand that we cut out the law of physical world in order to bring it the logic world of the computer and the relationship between the material like a plastic or a glass and the shape; The glass - Smooth-Sticky-Flat. The plastic - Sticky-Sticky-Slope.

We are groping around for the courtship action to the digital world which is not able to see and touch

I think that we have aimed to bring our action in the physical world into the virtual world as it stands until now. However, PINCH of iPhone may make us know that we do not necessarily bring our action into the virtual world as it stands. Although PINCH does not make something big or small in the physical world, we control two points of a diagonal line in order to draw a quadrangle in the virtual world like Adobe Illustrator. PINCH makes a movement of diagonal, therefore we may superimpose this hand movement and the movement for drawing a quadrangle in the digital world. As result, we naturally accept the relationship between PINCH and Zoom in and out. I consider that it is big problem to explain what makes us naturally accept the relationship between our gestures and the image action or feel something strange for it. We have to be careful and aware of what our unconscious gestures mean and correspond them to the image movements and actions in the virtual world.

However, it is our contention that gesture ─ certainly descriptive or 'iconic' gesture ─ necessarily involves indexical links to the material world, even though these links are rarely established or explicated in the communicative situation itself. Rather, in conversational contexts that are detached from the talked-about world, participants must fill in encyclopedic knowledge (ranging from universal bodily experiences to highly specific cultural practices) to see and recognize gesture. Phenomenally, there are only motions of the hands. What is perceived, however, is typically not motions but actions and, simultaneously, implied objects acted upon.

Curtis LeBaron, Jürgen Streeck, 'Gestures, knowledge, and the world' In David McNeill(Ed.), Language and Gesture, pp.118-138 (2000)

Finally, I think that a big change will happen because the digital which is no material and shape intervenes in our gesture communication based on the material and shape of our body. To show our will to the digital world. Conversely, the digital world shows its own will to us. I formerly called this correlation 'Display acts' which the image on the display changes corresponding to our actions. Now, I consider that 'display acts' is, connected with 'display' means the bird's action of the courtship, a courtship dance or action between the human and the digital furthermore, the human and the human through the digital. We are groping around for the courtship action to the digital world which is not able to see and touch. I want to feel the touch of the digital world.