‘Lectronic Lascaux: a Literary Salon

Literary Salon Proposal, Spring 1997

Adrian Banard, Jennifer Glos, Hannes Vilhjalmsson

Abstract

This document proposes ideas for a literary salon which uses technologies to explore, expand, and change the way people have historically interacted in such public spaces. It outlines six main projects, which cover such forms of communication as written, gestural, verbal. They explore communication with strangers, friends, self, simultaneously, and over time.

Vision Statement

Communication in order to link people:

• to themselves
• to others
• across space
• across time

Integration/Unified Design

Our design is table-centered. Each of the projects is run from or occurs at a table, different tables have different projects, and therefore, different personalities. At most cafes, one table might be small and isolated and stuck in a corner, and another would be large with a big, fluffy couch. Physically, tables are different properties and you choose one according to the experience/ambience you want that day. Accordingly, we will digitally augment the tables to match their physical personalities.

• LaughingJokes: table in the sun, bright, cheerful
• MysteryTraces: in a dark alcove, by a fireplace, lit by a candle or small glowing globe
• RecipeRemarks: at the food counter, on a stool
• ShadowTalk: 2 small tables near a wall
• WordQuilt: this is on the ceiling
• MugMarks: could be added to any table
• TableGraffiti: a large table in the center, for a big group of people. Loud, patchwork chairs.
• Ghostly Table-Mates: 2 small, well-lite tables
• Circle Writing: Tables for a single person, tucked away, for writing.

Project Specifics: Six proposals

Modularity: The literary salon is an ambitious proposal, and our ideas for what we want it to be are ambitious as well. We have, therefore, tried to break down into modules each of our six presented projects, from an immediately implementable proposal which gives some proof of concept, through what we really want each project to be. Or we have given "easier to implement" options.

1. "Laughing Jokes", "MysteryTraces," and "RecipeRemarks"

Scenario

Written language allows people to communicate through an intermediary (pen and paper) over time and distance. A women can feel close to her great-grandmother, who she never met, by coming across her diary. An author can reach millions of people he’s never met through his book. Two sisters, living across the world from each other, can keep in touch via email.

We chose to break this down into 3 projects which look at how we can enhance and play with different genres.

LaughingJokes

Sara sits by the table by the window, where the sun is shining in, and opens the joke book. It opens to a new joke about an elephant and a pickle and Sara, not expecting the ending, gasps and laughs. She then flips to a long joke, and as she reads it, she begins to hear around her laughter intermingled as invisible readers reach the punchline at different times. Sara is already smiling from the laughter and, when she reaches the end, adds her own as she laughs. She then types in a new joke of her own and looks forward to returning the following Sunday to listen to other’s laughter at her joke.

MysteryTraces

By the fireplace, there’s a table which is hidden in a dark alcove. On the bookshelves near it are mystery novels and the game of "Clue." Nearby hangs a portrait of Sherlock Holmes. Alex, who loves mysteries, sits down and pulls up the mystery he’s been reading the last few times he’s come to the lit-salon. Last night he thought he figured out who did it. He opens to his digital bookmark and clicks to link an audio segment. He speaks aloud: "The butler must have done it because of the missing shoe." He continues reading, and 2 pages later there’s an audio link from "Phantom" who Alex knows from experience is often an excellent sleuth. To his dismay, Phantom counters Alex’s claim and points out, with evidence, that the butler is being set-up by the maid...and so it continues.

• readers can choose whether or not to look at the audio files (they might not want to maybe have someone else figure it out)
• once an audio link is made, it’s permanent: readers can’t go back and delete bad ideas!

RecipeRemarks

Marc orders the veggie delight sandwich an sits down with his meal at one of the tables. To his surprise, after placing his plate on the table, the recipe for the sandwich appears on the screen. In the margins are comments written in by various people -- "try ordering it with the the honey mustard and hold the onions." There are options for Marc to add his own comments on the page, which he does after finishing the sandwich: "Great!"

Feasibility/Implementation

LaughingJokes

• Audio recording and mixing (hardware and software): audio is recorded for a certain amount of time after a joke window is opened. The audio is then mixed into a previously recorded audio file for that same window.
• Database of jokes and linked audio files: audio is played each time a joke window is opened, playing previously recorded audio for that same window.
• Note: possible to use video, but more difficult to mix and questionable benefits.
• An embedded display

MysteryTraces

• same as above, but database of mystery stories, and audio recording/playback are triggered by buttons on the screen

RecipeRemarks

• a text/drawing integrated program

Resources

• A computer connected to an embedded display
• Audio recording, mixing, and playback devices
• Database structure (scriptable)

Research Goals

• Exploring linking writer and reader through time
• LaughingJokes leaves spontaneous traces, whereas the others are purposeful

2. Shadow Talk

However, in this literary salon, there are many ways to communicate beyond written text, as the next project will describe.

Scenario

Christina sits at a table near the wall and notices a ball embedded into the tabletop in the upper right corner. She curiously touches the ball and feels how she can rotate it. As she does so, a shadow on the wall beside her springs to life, advancing towards another table further down. Christina quickly rotates the ball back into the original position, making the shadow retreat back to her table. Underneath the table, there is a keyboard. Christina types "Hello" and again, the shadow animates a talking head and a greeting gesture, this time in place. Mike, at the other table, has been observing this and reaches for the ball on his tabletop, making another shadow appear. Mike makes his shadow walk towards the first one, and types in a greeting on the keyboard. Soon, Mike and Christina are talking away, with their shadows animating their input and engagement.

Feasibility/Implementation

• This is a specialized version of "body chat"
• Runs on networked Pcs that project an overview screen on a curtain
• Open for experimentation (adding different behaviours of varying complexity)

Resources

• 2 Pcs, networked
• Projector
• 2 trackballs/Pads
• Cables
• Curtains

Research Goals

• Turns words and phrases into a control mechanism for a body. i.e. Generate appropriate body language/gestures for a given "verbal" input utterance
• model turntaking and back channel feedback

3. Word Quilt

Scenario

Jonathan is sitting alone at a table, mellow and drinking his coffee. He’s people-watching: laughing as Christina and Mike interact through their shadows, seeing two people alone at two different tables, both rapidly typing, probably to each other. He looks up and sees a quiltwork of words which are appearing and disappearing, reflecting the interactions in the room. It is almost hypotizing, watching the words change.

Words, fragments, utterances rearranged automatically into a visual quilt according to discourse rules.

Feasibility/Implementation

• A program that collects strings and rearranges them in 2D space. The program runs on a PC or UNIX station as a networked server accepting text.
• Complexity of rules can range (experimentation)
• Verbal play could be incorporated into the rules.

Resources

• A PC/UNIX station, networked
• Projector
• Overhead projection screen

Research Goals

• Create some coherence from seemingly random strings/parts of utterances.
• Try out rules from discourse analysis
• Leaving traces of interaction

4. "MugMarks" and "TableGrafiti": implementations of "active objects"

Scenario

On the popular television show "Cheers," the theme song describes the bar as the place "where everybody knows your name." This is an important element of bars/coffeehouses/salons: becoming a "regular" so that you’re remembered by others there. MugMarks proposes that rather than only being remembered by the people, a visitor is remembered by the place itself, by the physical environment. If then one returns 30 years later, and everyone they knew was gone, the place would still remember them.

Kim goes up to the lit-salon counter and orders Earl Grey tea. The bartender gives her a mug that he thinks she’ll like: a large whale mug, where the handle is of a whale’s tail. Every mug at the salon is unique, always eliciting comments from the patrons. Each mug has a history to it as well, which Kim is soon to discover. She takes her mug and sits at one of the small tables for two. She places her mug down on the mug holder and fishes out some reading material she brought. When she picks up the mug, she’s astonished to discover that it left behind a mark, like a water mark, but in this case, the mark is a round image of a person’s face which fades away as she looks at it. Intrigued, she places the mug down again for a moment and picks it up. Again, a MugMark has been left behind, this time of another person.

MugMarks augment the concept of physical traces being left on an environment (scratches on wood, stains on the tables) with digital traces which can give different types of information. MugMarks show someone a bit of history of their mug -- who has drunk from it before.

Further elements:

• A person might start asking for certain mugs -- maybe always the same one so that their own face, from previous visits, may occasionally pop up.
• A bartender might try to give certain mugs "personalities" by always, for example, giving the huge electric blue mug to young people who come in alone.
• Displaying of the photos could be random or organized somehow. Images could be linked to more than the mug -- if someone adds their laughers to the joke book and a new person is listening to it, the image they might see is of the person they’re hearing laugh. Or the bartender can easily mark what is drunk out of a mug each time, and perhaps Kim will only see other Earl Gray lovers.
• Might choose to have two of several mugs (with the same Ids), and if they are used simultaneously, they’ll display the other drinker in the room. The bartender could choose to link two people -- a mild form of social intervention.

Variations on MugMarks: some easier to implement alternatives that are, perhaps, just as appealing! Both of these remove the need for the sensing equipment (the hardest part). We also might chose to have MugMarks at one table, and TableGraffeti at another.

• TableGraffeti I: Digital equivalent of "wood carvings:" people often scratch into desks and tables things like "John + Mary" to leave a mark of themselves behind. If people sign in with their signature (on a digital pad) when they enter the salon, that signature, perhaps altered some by photoshop, is projected down onto a tabletop like a carving. Interesting if people catch on to this and begin writing things other than their names.
• TableGraffiti II: A "tablecloth" or tapestry of everyone who has ever sat at a table: a camera at the center of a table grabs images of everyone at a table and adds them to a collage of such photos which is projected down onto the table, like a tablecloth.

Feasibility/Implementation

Options (of varying difficulty) for each step:

1) Take digital photo of person

• • Set up booth where person picks up their drink. Like an ATM machine where it’s easy to get a photo of the person there (can use P-finder to adjust to height).
  • Place a Mac eye in the centerpiece in the middle of a table, facing a chair. Activated by a chair sensor to grab a photo.

2) Link the photo to the mug

• • The drink is placed by the bartender on a sensor in front of the person at the booth. The sensor activates the camera to take a picture and reads, off the bottom of the mug, the mug ID, linking the two.
  • Mug currently on table linked to new picture

3) Identify the mug on the table

• • ID sensors which the lab has upstairs
  • Bar code scanning
  • infrared and iRX boards

4) Display the photo

• • video camera above the table projects down

Resources

Mugs

Digital camera

Sensor technology

Projector

Research Goals

Exploring an active role of objects in the lit-salon environment. How can "active objects" provide: history, memory, stories, link people? How can they make the environment richer? MugMarks provide a history of a mug, and therefore, some history of the salon. It also may serve to link people, and to trigger memories of a previous visit.

5. Ghostly Tablemates

Scenario

One of the trademarks of the cafe or the bar is its sometimes private, sometimes social nature. Some groups of people in a cafe are there to talk among themselves, but other groups or individuals are there to meet new people. The physical layout reflects this confusion. There are individual tables, and while at a table, it is conventional to ignore the people at the surrounding tables, if one intends to be private. But at the same time, conversation is easily heard from one table to the next. Often a regular will overhear something of interest at another table, and will join in or comment. Single customers are especially a target for such interest. It is not considered impolite to ask a person who is sitting alone if you can sit with them. Occasionally coffeehouses encourage this, by having really large tables, which allows a number of non-affiliated parties to share a table.

Janet goes into the lit-salon carrying Discipline and Punish. She’s here to read today. She grabs some espresso and sits at a small well-lit table in the corner. Halfway down the first page, and halfway through the coffee, she realizes that there’s something funny about the table. She can see another pair of arms, and another book, on the surface of the table. It must be projected there by the light. She can’t make out the title. One arm grabs a cup and proceeds to lift it out of focus. Janet turns around, and scans the coffeehouse for the culprit, finding her at the other end, at a similar small table. Janet laughs. The other woman looks up, startled, and says "hello?". Janet can hear her as if she actually were sitting at the other side of the table. Janet asks, "What are you reading?"

The purpose of Ghostly Tablemates is to extend this concept of the "shared table", by connecting two tables through projected video and sound. As anyone who has used a speakerphone knows, a space which is connected but not actually present forms a sort of half-space, distant yet still within the conversation. Like two people can sit at a big table and strike up a conversation, two parties can sit at separate tables and choose to interact or ignore each other, engaging in a public or private presentation.

Elements:

In the scenario above, a video shot was taken from above and projected from above. Instead, it could be taken and projected from the side (if there is a convenient wall handy), allowing the tablemates to see each others’ faces in a ghostly manner.

It is not necessary to have both sound and video. Sound only, or video only, may actually be more comfortable for the participants, as it is a further remove.

If the system is set at one-person tables, it is a safe one-on-one interaction, as described above. But if it is set at two large tables, it becomes a way of joining parties. A large group could actually all share the same conversation space by sitting at two such tables, a feat impossible in normal coffeehouses.

Who determines which tables are connected? It could be a sneaky bartender, who would bring together like people as in the example above. Or it could be set, so people know which tables talk to which other tables. Or it could be automated, reacting to the situation appropriately.

Feasibility/Implementation

The main concern in this project is handling the hardware. Video cameras and projectors are needed, along with microphones, amplifiers, and small speakers. Also, if it is going to be possible to change the connections between tables, then some sort of audio/visual line switcher will be necessary. If a computer is going to control the switching, then the computer will need to be able to control the switcher and be able to sense some relevant things, like how many people are sitting at a table. There would be minimal programming necessary, except for implementing the computer/sensor and computer/switcher interfaces.

Resources

Video cameras

Video projectors

Microphones

Small speakers

Connecting cables and amplifiers

Switcher

If there is computer-controlled switching:

Sensor technology

Computer

Research Goals

How do people construct public/private conventions in a new space, like the speakerphone? What level of connection is necessary to force people to consider it rude to share the space? When there are two "halves" to a conversation, how does one address one half specifically? If only part of a body can be seen, or the body is seen at a strange angle, how are conversation dynamics affected? Does placing the technology between people make them more uncomfortable (like talking to an answering machine)?

6. Circle Writing

Scenario

Telling your friend a story and writing a story down are two very different experiences. Telling the story involves thinking fast, entertaining the listeners, and responding to comments or feedback the audience provides in the course of the story. Writing is a long exercise of composition, and feedback doesn’t happen until the story is largely written. But what if writing were the same intense experience that oral storytelling provides? And what better place to do such writing than a literary salon?

Miguel wanders into the litsalon with some friends. They order and sit down. The guys immediately go off on the Celtics. Miguel has no such interest, and looks around the cafe. The next table has a display mounted in it with a keyboard and a pen attached with a wire. This must be one of those cybercafe places. There’s a message blinking on the display: Angie is telling a story. Intrigued, Miguel goes over and clicks on the message. A long text comes up and scrolls past. At the end, Miguel sees keystrokes appearing on the display: "When I was young he introduced me to the magic of the music box. We sat on the por" He realizes that Angie is typing it in as he reads it. He scrolls back and starts reading from the beginning. It’s a story about her grandfather in India. There’s a little drawing of a music box on the first page, in the margin. There are other comments too, like "What about your grandmother?" Miguel picks up the pen and draws some notes coming out of the music box. He goes to the end of the story in progress. He laughs when he reads about Angie’s uncle, who would dance drunk outside her house. He draws a smily face in the margin and under it, "Why did he get so drunk?"

This Circle Writing system would enable others to read and comment "in the margins" of a story or essay which is being written. The storyteller can then edit or redirect their telling to include or respond to these comments, much like an oral storyteller will respond to comments by clarifying or incorporating the comment.

Elements:

This example only had the storyteller typing at the end, but our storytelling need not be so linear. If it is possible to edit the story-in-progress, then the teller could immediately respond to a comment without breaking the flow of the story to go back to the commented part.

The "listeners" don’t have to be actual people. A computer could pretend to be a listener and put its own comments in the margin.

The stories don’t have to go away when they are finished, but could be stored with the comments for reading later. The way to bring up an old story could be to associate it with the mug of the person who wrote it. Perhaps future readers could also add their comments, and the story slowly turns into the electronic equivalent of a dog-eared book full of underlinings and questions in different pen colors.

Feasibility/Implementation

The storytelling system would likely run on several networked computers. There is a significant amount of programming involved in displaying the story along with its (possibly pictorial) comments, and providing the appropriate editing mechanisms. If there is a "screen pen", interfacing with it would also take some programming.

While it is preferable to have a pen with which one can directly draw on the display, a similar effect could be achieved by placing the story in an environment similar to a bitmap editor if we cannot get the pens.

Adding a computer listener would not be difficult. SAGE and Rosebud already do similar things, and the methods they use could be adapted to this situation.

Resources

Several networked computers

Screen pens

Mounting equipment for placing display in table

Research Goals

Creating a writing situation which contains all the interactivity of oral storytelling. Seeing how people respond and adapt to a new narrative interaction. Testing out computer simulation of listeners, and the effectiveness of various programs at prompting the storyteller or asking relevant questions.