Creating a Learning Culture
MOOSE Crossing:
Creating a Learning Culture
by
Amy Susan Bruckman
Thesis Proposal for the degree of Doctor of Philosophy
at the
Massachusetts Institute of Technology
December 1994
Thesis Advisor:
Mitchel Resnick
Assistant Professor of Media Arts and Sciences
Thesis Readers:
Pavel Curtis
Member of the Research Staff
Xerox PARC
Henry Jenkins
Associate Professor of Literature
Director of Film and Media Studies
Massachusetts Institute of Technology
ABSTRACT
MOOSE Crossing is a text-based virtual world (or "MUD") designed to
support the development of a "constructionist learning culture." (1)
Children from a variety of geographic and cultural backgrounds will connect
across the Internet to collaboratively build a virtual world. As part of this
research, I am developing a new programming language (called MOOSE) and client
interface (called MacMOOSE) designed to make it easier for children to create
new places and objects. The virtual world, MOOSE Crossing, will be opened to
300-1000 children aged twelve and under on the Internet. Through analyzing
the children's learning experiences, I will be able to explore the
potential of the combination of construction and community. In what ways can
community support and enhance constructionist learning? I will study the
learning which takes place in this environment through two techniques. First,
I will analyze the artifacts the children create and logs (recorded with
informed consent) of what takes place in the virtual world. Second, I will
work closely with one local group of children, and use an ethnographic
methodology to explore their learning experiences. The goal of the MOOSE
Crossing project is to create a new type of constructionist learning culture,
and observe that culture to shed light on the power of the combination of
construction and community.
"They dance together and as they dance everyone is learning and teaching as
well as dancing. Even the stars are there to learn their difficult
parts." (Papert 1980)
1. INTRODUCTION AND BACKGROUND
1.1 A TECHNOLOGICAL SAMBA SCHOOL
In samba schools in Brazil, members of a community gather together to prepare
a performance for carnival. People of all ages learn together, as a
community. "Let me show you something," says one dancer to another. Learning
is not dictated by a set syllabus, but is spontaneous and authentically
motivated.
In Mindstorms, Seymour Papert has a vision of a "technological samba school."
It's not the content being learned in samba schools, nor the efficacy with
which it is mastered that enchanted Papert--he is not advocating dance as
subject matter, and he does not comment on whether the participants become
particularly good dancers. Rather, he is inspired by the different
relationship to learning the members develop, and the way in which learning
becomes a community process.
Could there be a technological samba school--a place where people are
meaningfully engaged with computational ideas, working as a community on
projects that they care about? Papert highlights a number of features of
samba schools he sees as important. In samba schools, learning is:
* self-motivated,
* richly connected to popular culture,
* focused on personally-meaningful projects,
* community based,
* an activity for people of all ages to engage in together,
* life long--experts as well as novices see themselves as learners, and
* situated in a supportive community.
Text-based virtual worlds on the Internet can have each of these features.
For historical reasons, these virtual worlds are called "MUDs" or "Multi-User
Dungeons." In the late 1960s, natural language processing researchers Will
Crowther and Don Woods wrote ADVENT, the first text-based adventure game
(Raymond 1991). In the late 1970s, multi-player versions were built for
the ARPAnet. (2) People from all over the world connected to these "Multi-User
Dungeons" to play a text-based form of the game Dungeons and Dragons. In
1989, a graduate student at Carnegie Mellon University named James Aspnes
decided to see what would happen if the monsters and magic swords were
removed. (3) People's chief activity became trying to extend the virtual world,
rather than conquer it. (4)
Most people do not view MUDs in terms of learning; however, the ways in which
MUDders are constructing worlds together has resonances with the
constructionist theory of education, an approach to learning first articulated
by Papert (Papert 1991). Constructionism involves two types of construction.
First, it asserts that learning is an active process, in which people
construct knowledge from their experiences in the world. (This idea is based
on the theories of Jean Piaget.) To this, constructionism adds the idea that
people construct new knowledge with particular effectiveness when they are
engaged in constructing personally-meaningful products. They might be
constructing sand castles, LEGO machines, computer programs, or virtual
objects. What's important is that they are actively engaged in creating
something that is meaningful to themselves and to others around them.
MUDs combine construction with community. In an informal ethnographic study
of twelve adults who learned to program for the first time in a MUD, I found
that the community provides motivation for learning, technical support, and
emotional support for the technophobic (Bruckman 1994). In this
"community-based learning," teaching is often a reciprocal relationship that
takes place among peers, rather than a one-sided transmission from experts to
novices. These virtual communities are promising test-beds for exploring a
new vision of learning--a vision that is more like samba schools at their
best, and less like the stiff, stifling regimens that exist all too frequently
in traditional schools.
Could MUDs become a kind of technological samba school? The potential is
there; however, most existing MUDs suffer from the poor quality of software
tools available, lack of guidance by an educational vision, and lack of
careful study. In this thesis, I hope to rectify these problems through the
creation and study of a virtual world for children called "MOOSE Crossing."
MOOSE Crossing includes a new programming language I have developed called
"MOOSE" designed to make it easier for children to program in the virtual
world, and a client program called "MacMOOSE" to make working in that language
less awkward. The initial virtual world will be seeded with coding examples,
and structured into areas for the children to extend--for example, one store
front is an invitation and context to begin to build a town. I will open
MOOSE Crossing to 300-1000 children aged twelve and under on the Internet, and
study what emerges. Using an ethnographic methodology, I will work closely
with one group of children. I will also analyze artifacts the children
create, and logs of activity and conversations in the virtual world, recorded
with informed consent. Through analyzing the children's learning experiences,
I will be able to explore the potential of the combination of construction and
community. In what ways can community support and enhance constructionist
learning?
1.2 MUDS AS A MEDIA FORM
The fantasy worlds of MUDs are closely connected to popular culture.
Researchers in cultural studies have shown that popular culture plays an
important role in children's learning. Children take stories from popular
books and other media forms, and make them their own, using them to make sense
of their daily experiences (Wolf and Heath 1992). Henry Jenkins documents the
way young children not just watch but interact with the television show "Pee
Wee's Playhouse." He writes that "children draw upon the prefabricated
characters and situations of popular culture to make sense of their own social
experience, reworking them to satisfy their own needs and desires. The
children's manipulation of these televisual materials rarely stops when the
broadcast does. Rather, program content is fragmented and dissected and the
most meaningful bits, 'the good parts,' are integrated into the child's
other play activities, into dreams and into waking thoughts." The children
"feel little compulsion to remain faithful to the original," instead
reappropriating those cultural materials to make very personal meanings
(Jenkins 1988). At its best, popular culture is a starting point for the
child's own creative process.
In addition to viewing MUDs as learning environments, we can also view them as
a new media form. MUDs can be compared not just to school, but also to
television and books. A television viewer is a recipient of commercially
produced content. A MUD player is a collaborative creator of media content.
In "Xylophones, Hamsters, and Fireworks" (Resnick 1991), Mitchel Resnick
documents how popular culture was a jumping-off point for LEGO/Logo projects.
One elementary school teacher had never built anything out of LEGO before and
described herself as feeling uncomfortable about science. However, her desire
to build a carousel like the one she rode on as a child inspired her to learn
enough about gears not only to make the carousel go around, but also make the
horses move up and down. Her love of carousels became a path to developing a
new relationship with technology.
Similarly, in the "Game Design Project," Yasmin Kafai found that she could
leverage off of children's love of video games to encourage them to be more
meaningfully engaged with computer programming, math, and science (Kafai
1993). MUDs leverage off of children's fondness for not only games but for
all forms of popular culture. Characters and places from literature, comics,
and television become starting points for children's creations in the virtual
world. Like LEGO/Logo and the Game Design Project, MUDs amplify children's
natural tendency to re-purpose stories, redirecting their enthusiasm for
popular culture from an act of consumption to one of creation, and to
engagement with both linguistic and technological creativity.
Two key points emerge from viewing MUDs as a media form. First, it is
important to note the ways in which popular culture serves as a starting point
for an individual's creativity. Second, although research in cultural studies
strongly indicates that viewers are active creators and not passive recipients
of meaning, one can draw a distinction between media forms in which the text
itself is produced by the users versus more traditional media forms in which
the text is produced by a publisher or broadcaster. In many media forms on
the Internet including many MUDs, users are creators of content.
In this respect, the emphasis that cultural studies (especially British
cultural studies) places on analyzing people's original cultural productions
(Jenkins 1992) has clear resonances with the emphasis researchers in
constructionist education place on learning through making things (Harel and
Papert 1991). Both disciplines view individuals as empowered by the media
they reappropriate, and see a person's creative productions as a site of
resistance to dominance by corporate or institutional interests--by
publishers, broadcasters, or schools.
1.3 LEARNING CULTURES
In MOOSE Crossing I hope to create a learning culture. Two definitions of the
word "culture" are relevant:
"5. the behaviors and beliefs characteristic of a particular social, ethnic,
or age group: the youth culture; the drug culture.
6. Anthropol. the sum total of ways of living built up by a group of human
beings and transmitted from one generation to another." (Flexner 1993)
I would like to define "learning culture" as the beliefs (especially about
learning) characteristic of a particular group of people at a particular time,
and the way those beliefs are embodied in the complete way of interacting of
the group.
Any environment is a form of learning culture; one could analyze the learning
culture of a particular classroom, or even a public space like a park or a bus
stop. However, I wish to use the term primarily to denote a particular kind
of learning culture--one inspired by constructionist principles. I will
sometimes use "constructionist learning culture" and "learning culture"
interchangeably.
[A Calvin and Hobbes cartoons appears here in the formatted versions. Calvin
is called to the front of the classroom, and in his imagination he is being
sent before a monster, "to appease the evil god we call Nollij."]
Figure 1: Calvin in school: an unfortunately realistic depiction (Watterson
1991).
Unfortunately, negative views of learning are common in the culture of many
educational institutions. Too often children feel imprisoned rather than
empowered, trapped in the classroom like Calvin thinking "what do they want of
me?" (Figure 1) Learning and traditional classroom learning are commonly
conflated, and the child's frustrating experiences in school foster negative
feelings about all forms of learning. Constructionist learning cultures seek
to change the child's relationship to learning, fostering self-directed and
self-motivated learning, so they instead will think "this is what I want for
me!"
I experienced one innovative, constructionist learning culture first-hand as a
child. When I was a teenager, I used to say I grew up a year's worth each
summer at camp, and then stagnated over the rest of the year. From the summer
I was eleven through the summer I was eighteen, I participated in a remarkable
learning culture called Buck's Rock Creative Work Camp in New Milford,
Connecticut. (5) I will describe the features of that experience that I hope to
emulate and push further in MOOSE Crossing.
Activities at Bucks Rock center around the creative arts. Over the eight
summers I attended, I blew glass, wove a blanket, made soft sculptures of
batiqued fabric, edited and wrote stories for a magazine, acted in a play,
adopted a goat, planted and cared for a vegetable patch, and made many things
out of wood: bowls turned on the lathe, a recipe card cabinet for my mother, a
jewelry box. I spent most of my time in the wood shop, because I liked the
people there, and loved the smell and texture of wood.
A key feature of Buck's Rock is its lack of schedule. Campers must
participate in some activity each day, but it can be any activity. No one
minded that I spent most of my time at the wood shop--it is perfectly
allowable to spend an entire summer at one activity. No one tells children
where to go, or even checks where they've been. Adult counselors,
professional artisans in each of these areas, arrive at their activity each
morning, ready to help whoever happens to show up. Consequently, all the
campers at an activity want to be there--that activity is their first choice
from a rich variety of options. As a result, attitude and motivational
problems are rare. Furthermore, campers have ample time to develop a
meaningful engagement with the activities that interest them--they aren't
shuttled off to the next area when the clock strikes the hour.
Each activity or "shop" at Bucks Rock has its own culture. It was the company
of my fellow wood enthusiasts that drew me there, as much as the process of
working with wood. I can still describe the personality type attracted to
each activity. For example, one extreme type are the glass blowers. At the
glass blowing shop, only two people can use the furnace at a time. It's a
wonderful activity to watch--the shop always attracts a crowd who look on,
chat, and gossip. You have to wear dark glasses to blow glass, and many
regular glass blowers seemed to have a dark glasses sort of attitude--much of
glassblowing struck me as sitting around and acting cool. I preferred the
wood shop next door, where folks were more down to earth. At the wood shop I
found friends based on common interests and temperament, rather than being
stuck in an arbitrary grouping based on gender and age. Within most
communities are sub-communities with distinct cultural features. I should
perhaps refer to MOOSE Crossing as a collection of "learning cultures," not a
single, monolithic culture.
What would it be like to structure a school rather than a summer camp in this
way? The "Friskolen 70" school in Denmark is just such a school, and was
studied by Aaron Falbel in his dissertation "Friskolen 70: An Ethnographically
Informed Inquiry into the Context of Learning" (Falbel 1989). At Friskolen
70, there are no classes or groupings of children by age or ability. There
aren't any classrooms--most of the school building is one open space (a former
stocking factory), with wooden beer cases that can be stacked and restacked to
create areas for special purposes, as needed. Children arrive in the morning
and do what they please. Teachers are available to help the children pursue
what interests them. The character of the place is captured in this excerpt
from Falbel's field journal:
"Maria (11) and Johanne (9) are using the wood-burning set in the workshop to
make Christmas presents for their friends and relatives. Maria has just
completed a small, wooden tic-tac-toe game for a cousin of hers in Sweden.
The X's and O's are made out of wooden pieces that fit neatly into a finely
crafted playing board. She is etching an inscription to the recipient of the
gift on the back of the playing board when little Clara (6), who just started
at the school at the beginning of the month, wanders by and is drawn into the
scene. She watches Maria etch a floral design under her inscription and is
fascinated by the strangeness of this smoldering writing instrument. Maria
notices the presence of Clara and displays for her the finished product.
Clara turns over the board and feels the grooves of the inscription with her
fingers. Maria asks her, "Can you read?" Clara shyly shakes her head no.
Maria smiles and says "Come," and she motions for Clara to come sit on her
lap. Then, very slowly and sweetly, Maria sounds out the words as Clara
guides her fingers over the dark-brown letters. Clara is totally absorbed:
her face conveys an expression of rapt concentration, her mouth partly open,
and her cheek leaning against Maria's arm. The entire episode lasts not much
more than a minute. It is so effortless, natural, and unself-conscious that
to call it "peer tutoring" would be to debase the beauty of the
situation."(Falbel 1989)
It is certainly a much more radical decision to use this open,
student-directed style for a school than for a summer camp, and some open
questions remain about its effectiveness. Falbel himself is careful not to
put the school forward as a model. However, experiences at Friskolen 70 raise
a number of important issues about the complete social context or "culture" of
learning.
Friskolen 70 and Buck's Rock share a number of features I plan to incorporate
into MOOSE Crossing:
* All participation is self-motivated,
* Children are trusted to monitor themselves,
* Adults are present as facilitators,
* Self-selection leads to peer groupings based on common interests
rather than gender or age, and
* Interaction with peers is a key component of the learning experience.
These features are at the heart of constructionist learning cultures.
1.4 EDUCATIONAL MUDS
An increasing number of educational MUDs are being founded. Most are designed
to teach college composition. Examples include DaedalusMOO, Diversity
University, LitMUSH, and WriteMUSH. I know of two MUDs designed to be
learning environments for young children: MicroMUSE and MariMUSE.
MicroMUSE is the oldest and largest MUD for kids. It was founded by Stan Lim
in 1990, and leadership of the project was soon taken over by Barry Kort. Its
official charter states that "MicroMUSE is chartered as an educational
Multi-User Simulation Environment (MUSE) and Virtual Community with preference
toward educational content of a scientific and cultural nature."(Ender 1993)
Anyone with Internet access can join MicroMUSE. Both adults and children
participate.
MariMUSE is a project of researchers at Phoenix College in Arizona, working
with Longwood Elementary, a local elementary school serving a population of
at-risk primarily Native-American and Mexican-American children. While
MicroMUSE is open to anyone on the Internet, MariMUSE serves a small, local
population. During the summers of 1993 and 1994, students were bused to a
computer room at Phoenix College for three hours each day. They interacted
primarily with one another, but also with adults who volunteered to log on and
spend time with the children.
I had the opportunity to visit Longview and meet the MariMUSE students in
March of 1994. I was impressed with the creative writing and design of
virtual rooms the students had done. One fifth grade boy built an airplane
hangar. He carefully researched the details of each plane, and asked the
school librarian to order more books about airplanes. A fifth-grade girl
built a many-roomed mansion. In real life, her family lives in a homeless
shelter.
However, none of the students I met at Longview had done any programming. One
girl had designed a horse, but the "ride horse" command didn't work. We
looked at the code together, and I asked her how it worked. She replied that
she didn't know--an adult had typed it for her. MUDding without programming
misses an important learning opportunity. Programming opens up new design
opportunities for children within the virtual world, and provides a context
for developing a meaningful relationship to mathematics, computer science, and
technology in general (Soloway, Guzdial et al. 1993).
An informal exploration of MicroMUSE shows that while a wealth of creative
writing is occurring there, only a handful of exceptional children are doing
any programming. This is probably as a result of the awkwardness of the MUSE
language. A better programming language is needed.
Furthermore, no careful research has been done on what children are or are not
gaining from these projects. Critical analysis is needed.
1.5 PILOT STUDY
In October of 1992, I began working on a MUD designed to be a professional
community for media researchers which I called MediaMOO (Bruckman and Resnick
1993). MediaMOO opened to the public in January of 1993, and as of December
1994 had over 1100 active members from 29 countries. MediaMOO is based on the
MOO software, developed by Pavel Curtis of Xerox PARC (Curtis 1992; Curtis
1993; Curtis and Nichols 1993). Constructionist ideas have informed the
design of MediaMOO. The virtual world of MediaMOO is built by its
inhabitants. Unlike most MUDs, all members are given full programming
privileges, and are encouraged to learn to program. Experience designing and
managing MediaMOO has guided the design of MOOSE Crossing, serving as a pilot
study.
As part of the MediaMOO project, I conducted interviews with twelve adults who
learned to program for the first time on a MUD (eight on MediaMOO; four on
other MUDs) (Bruckman 1994). Those interviews provided significant insight
into the type of learning that takes place in MUDs. For example, one
twenty-year-old undergraduate biology major who I will call Lynn programmed a
fan, like the one she uses in Chinese folk dancing club. She designed it so
that when she uses it, the results are graceful; when someone else attempts to
use it, it falls clumsily on the floor, as her fan often did when she was
first learning this type of dance. Before Lynn discovered MUDs, she took one
programming class taught in Pascal in college, which she did not enjoy. On
MediaMOO, her projects had more significance to her:
"In Pascal it was kind of frustrating. They would give you challenging ones
and then when I finished it it was satisfying but in a different way.... In
this way [on the MOO], it was something that I wanted solve, something that I
wanted to make. Something that would give me satisfaction after I finished
it. Some of the programs I did in Pascal like the word alphabetizer or
taking a bunch of text and organizing it, I would never use that. And it
doesn't really have any significance to me. But in the MOO something that
I wanted to do--that I wanted to show people, about myself--I could program
it."
"The kind of programs I had in the class, they were kind of interesting and
it was neat to be able to make the kinds of program they gave you the problem
for. But I think in the MOO this whole idea of having a virtual world, and
having your own things, and making your own things, and setting up your
identity on the MOO--partly through using these things--I guess gave me the
drive to make it work, and whenever I ran into a bug to really ask people
"how do you fix this?" I guess maybe having a personal connection or
ownership of the objects.
Lynn felt a greater personal connection to her projects on MediaMOO, compared
to the assignments from her undergraduate computer science class. She was
inspired to work further on her projects, and supported in those efforts by
other people in the virtual world. Most MUDders remembers being new and
receiving help from others, and are happy to help new players. There is
almost always someone around who is eager to help. A number of people helped
Lynn with her project, especially a researcher at Apple Computer named Eric.
Lynn and Eric became friends. She was excited to develop a friendship with
someone older, and enjoyed asking Eric questions about his job. When Lynn
traveled to California one summer, she visited Eric at Apple. She describes
people at her college as "pretty homogeneous." Online she has developed
meaningful friendships with people she would not normally meet.
Lynn is proud of her programming accomplishments. She says, "When I got my
fan working, I went around showing everyone." In addition to providing
technical assistance and critical feedback for work in progress, members of
the community also form an appreciative audience for completed work.
Lynn's experiences online had a significant impact on her. She had previously
been half-heartedly considering applying to medical school, but now is
enthusiastically pursuing a career researching the social side of computing
and human/computer interface design. She credits her MUD experiences with
inspiring this interest. It is certainly gratifying that the virtual
community could inspire a student, particularly a woman, to pursue a career in
the computing field. However, the goal of using MUDs educationally is
certainly not vocational. Lynn's story is a good example of several positive
features of MUDs as learning environments. For Lynn, the virtual community
provided:
* Motivation for learning,
* Emotional support to overcome technophobia,
* Technical support,
* An appreciative audience for her work.
1.6 A HYPOTHETICAL SCENARIO
One main goal of the MOOSE Crossing project is to bring this sort of learning
experience to children. Extrapolating from the pilot study, one can construct
a hypothetical scenario of a child's learning experience on a MUD designed for
kids:
After school each day, Jane, who is eleven and lives in a suburb of Seattle,
spends two hours at the community center, until her parents are done with
work and can pick her up. Each day she picks which activity she'd like to
participate in. Usually, she goes to draw with the art group, or outside to
play dodge ball with the sports group if the weather is nice. Coming back
from dodge ball the other day, she passed by the computer room, where her
friend Sarah was still sitting at the keyboard, even though it was time to go
home. Sarah called Jane over, "Jane, Jane! You gotta check this out! I'm
talking to someone in Los Angeles!"
Jane looked over Sarah's shoulder as she chatted with Julio and Anna. Sarah
asked if they know any movie stars, and Julio said he was friends with Arnold
Schwarzenager. He also said he had seven pet cats. Sarah said she didn't
believe Julio really had that many cats. Soon it was time to go. They
agreed to meet again online at the same time the next day. The computer
teacher, Mr. Smith, asked if Jane wanted to try MOOSE Crossing too. Jane
said yes, so Mr. Smith gave her some information and a permissions form to
take home to her parents.
Jane returned the next day with the permissions form signed by herself and
her parents, and a week later Mr. Smith told Jane that her own character was
ready to be used. Jane chose to go to the computer area a few days a week
from then on, and always sat at a terminal next to Sarah's. Jane and Sarah
found Julio and Anna logged on, and went to join them. Julio and Anna were
in a store Anna had designed in the town of MOOSE Crossing, "Anna's Furry
Pets." Julio told Sarah that now he really does have seven cats. In the pet
store were seven virtual cats. Three of them were carefully described with
different color fur, ages, and sizes. The other four weren't finished yet,
Julio explained. Sarah showed Jane how to pet the cats. When Jane typed
"pet Fluffy," the little white kitten purred, but when she typed "pet
Raphael," the big black cat wagged its tail! Julio apologized and said that
he wasn't done making Raphael yet. He had made his cats by adapting code for
dogs from the library. Later Ranger_Rachel, an adult computer programmer who
volunteers time to work with children on MOOSE Crossing and whose son Tim
comes there occasionally, came by to say hi. She showed Julio how he could
make a "generic cat" and have all seven cats inherit from generic cat, so he
wouldn't have to make all those changes to each cat individually.
The children decided they wanted to have a pet show. Anna made two pet birds
that sing silly songs, and Jane made a pet elephant with pink and green spots
that tells elephant jokes. Sarah made an auditorium with a stage, and
ribbons for the winners. They announced the show in the newspaper, and soon
other children began making pets to enter. The children asked Ranger_Rachel
if she would judge the show. She responded that maybe it would be more fun
to have everyone vote. The children liked that idea, so Ranger_Rachel helped
Sarah write some voting software. Twelve children entered pets, and more
than a hundred came to see the show and vote for their favorite. When the
voting was over, Jane's joke-telling elephant had won first prize. Sarah is
now rewriting her voting program (with a little help from Mr. Smith and
Ranger_Rachel) so people can vote in lots of different categories--funniest,
prettiest, smartest (best programmed). Julio copied Jane's program and made
one of his cats tell cat jokes. The children are planning to have another
pet show soon.
The preceding story is hypothetical. MOOSE Crossing is being designed with
the aim of making this sort of scenario possible. There are a number of
questions one can pose about the children's experiences, starting with
concrete questions about what has been learned, and moving to broader
questions those answers shed light on:
* Have the children improved their reading, writing and programming
skills?
* Which elements of the technological environment (particularly those
unique to MUDs and MOOSE Crossing) helped learning? Which elements
hindered learning?
* Have children with greater initial strength in one area, either
writing or programming, developed in the other area?
* Is there any observable synergy in the way these skills are connected
in MUDs?
* Are the children able to express themselves in new ways in this
interactive medium?
* Which elements of the social environment (particularly those unique to
MUDs and MOOSE Crossing) helped learning? Which elements hindered
learning?
* In what ways are children supporting one another's learning
experiences?
* In what ways are adults and children supporting one another's learning
experiences?
* Have they incorporated feedback from other members of the community in
their projects?
* Have children revised and refined projects over time?
* Have they developed a sense of audience, tailoring their creations to
a target group of people?
* Has a new type of constructionist learning culture developed? How
does this culture compare to other constructionist and traditional
learning cultures?
* How do children compare their learning experiences on MOOSE Crossing
to those of school?
* Do they see reading, writing, and programming as personally useful?
* Do they have a better understanding of what reading, writing, and
programming are good for?
* Do they want to cultivate those skills further?
* Do they feel empowered by their encounter with a constructionist
learning culture? In what ways?
All of these questions can be seen as aspects of a core question:
* In what ways can community support and enhance constructionist
learning?
Not all of these questions are necessarily answerable; the list outlines the
things I hope to look for. Section 2.4 discusses the methodology for
investigating these questions.
2. THE MOOSE CROSSING PROJECT
2.1 "MOOSE CROSSING"
The "MOOSE Crossing" project includes four major components:
1. The creation of a virtual world for kids on the Internet called "MOOSE
Crossing,"
2. A new programming language called MOOSE, designed to make it easier
for children to learn to program,
3. A client interface called MacMOOSE designed to make accessing MOOSE
Crossing less awkward, and
4. A study of the experiences of children on MOOSE Crossing.
"MOOSE" stands for "MOO Scripting Environment. " The MOOSE language is built
on top of the MOO ("MUD Object Oriented") software, developed by Pavel Curtis
at Xerox PARC and Steven White. The design of MOOSE is inspired by elements
of the languages Logo, Smalltalk, Hypertalk, MOO, and MUSE.
The virtual world built with this software is called MOOSE Crossing. It is a
crossing in the figurative sense of a crossroads for children from different
socioeconomic and ethnic backgrounds, and different cultures. It is a
crossing also in a more literal sense, since when you connect to MOOSE
Crossing, you are at a country crossroads. In one direction, the road leads
to the town. In the other, to the city. A path wanders off into the woods,
where children can build different sorts of natural environments. At the
crossroads is a tree. If you climb the tree, you are in a tree house with
public notices and announcements. Up above the tree in the clouds is a place
to build areas with fantasy themes. This core is minimally developed; it is a
scaffolding to inspire the children to create places and objects. A small
forest may inspire children to build a meadow or swamp near by; a general
store may inspire a pet store or hair salon down the virtual street. The
small number of pre-constructed examples serve as coding examples,
inspiration, and context.
MOOSE Crossing will be restricted to children twelve years old and younger.
By imposing a maximum age, I hope to reduce the predominance of issues of
teenage sexuality in the culture. This will also enable me to give the
children greater freedom of speech, because there is a lessened chance of
older children introducing much younger children to inappropriate subject
matter. Adults and older children will be able to apply to be volunteer
"rangers" on the system, serving as mentors and guides. The ranger
application essay will ask them to reflect on issues of appropriate conduct,
and how and when to intervene in disputes which arise among children.
2.2 THE MOOSE LANGUAGE
One design decision has shaped much of the MOOSE language: the programming
language and the command line language should be identical, so that children
can try out a command interactively, and put it in their program if it works.
(6)
In most MUDs, it is possible to "emote." At the command line, if I type:
emote smiles.
Everyone in the room sees:
Amy smiles.
Now suppose I want to write a program for my virtual pet dog Rover to wag its
tail when you pet it. When I type, "pet rover," I see "You pet Rover," and
everyone in the room sees "Rover wags his tail." In MUSE (Wallace 1993), MOO
(Curtis 1993), and MOOSE this would be:
[MUSE] @va dog=$PET DOG:@pemit You pet Rover.; @emit Rover wags his tail.
@vb dog=$PET ROVER:@pemit You pet Rover.; @emit Rover wags his tail.
[MOO] @verb dog:pet this none none
@program dog:pet
player:tell("You pet Rover.");
this.location:announce_all("Rover wags his tail.");
[MOOSE] script dog:pet
on pet me
tell player "You pet Rover."
emote wags his tail.
end
The name "Rover" was written into the code of the above programs. Suppose you
want the dog's name and gender to be automatically substituted in. This
allows you to create a kind of "generic dog." Objects which inherit from
generic dog can respond to their own name. (Inheritance is not used in MUSE;
the program would have to be copied rather than inherited.) Code for a
generic dog in each of these languages would be:
[MUSE] @va dog=$pet *:@switch %0=dog,{@pemit %#=You pet the dog.; spoof wags
%p tail.},name(me),{@pemit %#=You pet [name(me)]; spoof wags %p
tail.},@pemit %#=I don't see any %0 here!
[MOO] @verb dog:pet this none none
@program dog:pet
player:tell("You pet ", this.name);
this.location:announce_all($string_utils:pronoun_sub("%T wags %p
tail.", this));
[MOOSE] script dog:pet
on pet me
tell player "You pet " + my name + "."
emote "wags " + my pp + " tail."
end
In general, in MOO there are two syntaxes; MOOSE syntax is a combination of
the two.
[MOO command line] verb direct_object preposition indirect_object
[MOO code] object:verb(argument1, argument2, ... argumentn);
[MOOSE] verb argument1 argument2 ... argumentn
The MOOSE language design is being revised as it is implemented and sample
programs are tried out. It will be further revised based on feedback from an
initial group of test users. A complete language specification and a
narrative of how the language was designed and revised will appear in the
written thesis. The thesis will evaluate the success of the MOOSE language
through interviews with children, and comparisons of their MOOSE programs to
other children's creations in MUSE and MOO.
2.3 THE MACMOOSE CLIENT PROGRAM
It is possible to connect to a MUD using raw telnet, but this interface is
awkward. The MacMOOSE client program adds the conveniences of a windowing
system to make MOOSE more usable. MacMOOSE:
* Separates input from output, so that text generated by others does not
interrupt what the user is typing.
* Puts help messages in a separate window, so that the information does
not scroll away as the user tries out what it suggests.
* Allows WYSIWYG ("what you see is what you get") editing of text. If
the user wants to correct "thhe" to "the" on the third line of a
script or text document, he/she can simply click at the appropriate
place in the window and then hit delete, rather than typing
"substitute /thhe/the/3" in the line editor.
* Includes an object browser to display all the properties and scripts
on an object, and climb the inheritance hierarchy to look at an
object's parents or children.
Unfortunately, users not working on Macintosh computers will not be able to
use MacMOOSE. It is still possible to access MOOSE Crossing without it. An
alpha version of MacMOOSE was made available to fifteen test users in April
1994, and a fully-functional version is planned for December.
2.4 A STUDY OF THE CHILDREN'S EXPERIENCES
A key contribution of this thesis, more important than the new technology
developed, will be a thorough study of the impact of the project on its
participants. The central question I plan to address is in what ways can
community support and enhance constructionist learning? In order to address
that top-level question, I will first investigate a number of specific,
detailed questions about the children's experiences on MOOSE Crossing. (See
section 1.6)
The study of the children's experiences on MOOSE Crossing has two major
components. First, using an ethnographic methodology, I plan to work with a
local group of children in an after school program. This will give me an
opportunity to observe children using the system, and take careful notes.
Some of the sessions will be either video or audio taped. I will interview
the children before, during, and after the program. The interviews will be
designed to assess the questions detailed in section 1.6.
Second, I plan to record all interactions in the virtual world. Written
consent from both parents and children will be obtained for this recording.
(7) This data can be used in a variety of ways. Statistical, computational
analysis of the raw data collected is possible, but I believe only marginally
interesting. More promising is the fact that I will be able to select a
child, and analyze that child's complete experiences--every conversation,
every command typed. I will analyze the complete experiences of a small group
of children, including both children randomly selected and children selected
because their backgrounds or experiences are in some way noteworthy.
3. IMPLEMENTATION
3.1 RESOURCES
As an interpreted language written in an interpreted language, MOOSE is slow.
The database is expected to be large, and the server keeps the entire database
in memory. A fast workstation with generous amounts of RAM and swap space are
needed. A Digital Equipment Corporation Alpha 3000/400 with 128 Mb of RAM has
been dedicated to this project. One to two Gb of disk storage will be
required for data collection.
A local school or community center with Macintosh computers connected to the
Internet is needed for the after school program. The Somerville Community
Computing Center has suitable facilities, and has expressed interest in
participating in the project.
3.2 PROGRESS TO DATE
The MediaMOO project served as preparation for the MOOSE Crossing project.
MediaMOO is a MUD designed to be a professional community for media
researchers. It has been open to the public since January 1993, and currently
has 1100 members from 27 countries (Bruckman and Resnick 1993). Experience
founding and managing MediaMOO has guided the design of MOOSE Crossing. In an
informal ethnographic study of twelve adults who learned to program for the
first time in a MUD (eight on MediaMOO; four on other MUDs), I gained insight
into how an online community can motivate and support learning (Bruckman
1994).
The initial design and most of the implementation of the MOOSE language are
complete. This design will be revised based on feedback from initial users.
Major components still to be implemented include better error handling, better
internal security, documentation, tutorials, and sample programs.
An alpha version of the MacMOOSE client program was released in April, and
feedback from initial users has been positive. The design has been revised
based on their comments. At present, we are almost ready to release a beta
version.
3.3 SCHEDULE
April 1994 Version 1.0a1 of MacMOOSE client program released
First simple MOOSE program run
December 1994 Submission of thesis proposal
Release of version 1.0b1 of MacMOOSE client program
January 1994 First alpha testers invited to MOOSE Crossing
Release of version 1.0 of MacMOOSE client program
February 1995 MOOSE language revised based on alpha user feedback
March 1995 MOOSE Crossing opens to kids on the Internet
After-school program begins
June 1995 Data analysis and writing begin
November 1995 Defense notification
Submission of written thesis
December 1995 Oral defense
ACKNOWLEDGMENTS
My advisor Mitchel Resnick should be credited not just with guiding my work,
but also with contributing significantly to the MOOSE language design. Pavel
Curtis' wonderful MOO software has proved to be an excellent development
platform, and he has also implemented a number of changes to MOO to make this
work possible. Henry Jenkins has inspired me to believe in the basic
creativity of users, and see media in its larger context. Thanks are due to
many other faculty members at MIT, especially Glorianna Davenport and Sherry
Turkle.
I'd like to thank the MIT Undergraduate Research Opportunities Program for
making it possible to work with so many talented undergraduates. An early
version of the MacMOOSE client program was written by Greg Hudson; the current
version is by Adam Skwersky and Steve Tamm. Albert Lin and Trevor Stricker
have contributed to the MOOSE language implementation.
Equipment has been supplied by Digital Equipment Corporation. Financial
support has come from Interval Research, the National Science Foundation, and
AT&T.
REFERENCES
Bruckman, A. (1994). "Programming for Fun: MUDs as a Context for Collaborative
Learning." National Educational Computing Conference, Boston, MA,
International Society for Technology in Education. Available via anonymous
ftp from media.mit.edu in pub/asb/papers/necc94.{ps.Z,rtf.Z,txt}.
Bruckman, A. and M. Resnick (1993). "Virtual Professional Community, Results
from the MediaMOO Project." The Third International Conference on Cyberspace,
Austin, TX. Available via anonymous ftp from media.mit.edu in
pub/asb/papers/3cyberconf.{ps.Z,rtf.Z,txt}.
Curtis, P. (1992). "MUDding: Social Phenomena in Text-Based Virtual
Realities." DIAC, Berkeley, CA. Available via anonymous ftp from
parcftp.xerox.com in pub/MOO/papers/DIAC92.{ps,txt}.
Curtis, P. (1993). "LambdaMOO Programmer's Manual." Available via anonymous
ftp from parcftp.xerox.com in pub/MOO/ProgrammersManual.{ps,dvi,txt}.
Curtis, P. and D. Nichols (1993). "MUDs Grow Up: Social Virtual Reality in the
Real World." Third International Conference on Cyberspace, Austin, TX.
Available via anonymous ftp from parcftp.xerox.com in
pub/MOO/papers/MUDsGrowUp.{ps,txt}.
Ender (1993). "The MicroMUSE Charter and Bylaws." Available via anonymous ftp
from musenet.bbn.com in /micromuse/Charter.gz.
Falbel, A. (1989). "Friskolen 70: An Ethnographically Informed Inquiry Into
the Social Context of Learning." Massachusetts Institute of Technology.
Flexner, S., Ed. (1993). The Random House Unabridged Dictionary of the English
Language. New York, Random House.
Harel, I. and S. Papert, Eds. (1991). Constructionism. Norwood, NJ, Ablex
Publishing.
Jenkins, H. (1988). ""Going Bonkers!": Children, Play, and Pee-wee." Camera
Obscura 17.
Jenkins, H. (1992). Textual Poachers: Television Fans and Participatory
Culture. New York, Routledge.
Kafai, Y. B. (1993). "Minds in Play: Computer Game Design as a Context for
Children's Learning." Massachusetts Institute of Technology.
Papert, S. (1980). Mindstorms: Children, Computers, and Powerful Ideas. New
York, Basic Books.
Papert, S. (1991). "Situating Constructionism." Constructionism Eds. I. Harel
and S. Papert. Norwood, NJ, Ablex Publishing.
Raymond, E. (1991). The New Hacker's Dictionary. Cambridge, MA, MIT Press.
Resnick, M. (1991). "Xylophones, Hamsters, and Fireworks: The Role of
Diversity in Constructionist Activities." Constructionism Eds. I. Harel and S.
Papert. Norwood, NJ, Ablex Publishing.
Soloway, E., M. Guzdial, et al. (1993). "Should We Teach Students to Program?"
Communications of the ACM 36(10): 21-24.
Turkle, S. (1984). The Second Self: Computers and the Human Spirit. New
York, Simon & Schuster.
Wallace, E. L. (1993). "MUSE Manual version 1.5." Available via anonymous ftp
from musenet.bbn.com in /MUSEman1.5c.
Wolf, S. A. and S. B. Heath (1992). The Braid of Literature, Children's Worlds
of Reading. Cambridge, MA, Harvard University Press.
NOTES
(1) See section 1.3 for a discussion of the notion of a learning culture. It
would be more accurate to say, "a collection of learning cultures," since most
cultures are composed of varied sub-cultures.
(2) Numerous multi-user games based on the Dungeons and Dragons role playing
game appeared in 1978-1979 including Scepter of Goth by Alan Klietz and MUD1
by Roy Trubshaw and Richard Bartle [email conversation with Alan Klietz, March
1992].
(3) Electronic mail conversation with James Aspnes, February 29th, 1992.
(4) As of October 1994 , there were more than 300 active, publicly announced
MUDs on the Internet. Most MUDs are still dungeons and dragons games.
(5) Bucks Rock was founded by Dr. Ernst Bulova and Mrs. Ilsa Bulova during
World War II, as a haven for British children fleeing the bombing of London.
At the time I attended in the late 1970s and early 80s, it was directed by
Sybil and Lou Simon. During the academic year, Lou was then also the
principal of the High School of Music and Art in New York City.
(6) In MOO, the programming language and command-line language are quite
different; learning one does not help you learn the other. In MUSE they are
identical, but the commands needed to write programs are filled with special
characters, and are jammed onto single lines with no line breaks. There are
no variables in MUSE--instead of using variables, you store values in
registers.
(7) The details of this project will of course be cleared with MIT's Committee
on the Use of Humans as Experimental Subjects (COUHES).
