7  The Alaska QUILL Network:  Fostering A Teacher Community Through Telecommunication

In Democracy and Education, John Dewey articulates some of the complex relationships among community, education, and communication:

There is more than a verbal tie between the words common, community, and communication.  Men live in a community in virtue of the things which they have in common; and communication is the way in which they come to possess things in common...  The communication which insures participation in a common understanding is one which secures similar emotional and intellectual dispositions--like ways of responding to expectations and requirements.  [Dewey, 1916, p. 4]

These arguments suggest that the environment of any educational experience can be viewed from the perspective of community:  Do students feel a bond with others students in their classroom?  Are students and teachers part of the same or different communities?  Do teachers feel connected in a substantial way to other teachers in their school?  In other schools?  The sense of being part of an educational community with shared goals, perspectives, and rituals can be as important a part of an educational experience as the content or method. 

Jaime Escalante's success in teaching advanced calculus to students from a Los Angeles barrio (documented in the movie Stand and Deliver) was due in part to the close-knit community he created in his calculus classroom; he and the students shared jokes and ritual greetings along with a desire to learn calculus.  In his description of Brazilian samba schools, an alternative educational model to traditional schools where novices and experts alike work on their samba dancing, Papert emphasizes the importance of community.  "There is a greater social cohesion, a sense of belonging to a group, and a sense of common purpose" (1980, p. 178).

Yet creating communities is a difficult proposition in today's educational institutions; we hear more about alienation - of students, of teachers, of administrators, of parents.  Given the intimate connection between community and communication, it is at least possible that the increased communication possibilities offered by computers and computer networks could play some part in creating and supporting educational communities.  Electronic networks might alter users' social networks by introducing additional resources for information, assistance, comment, or comfort.  Electronic interchange might solidify, extend, augment, or replace interpersonal contacts, leading to a more densely interconnected community.

In fact, the QUILL project in Alaska demonstrated at least one way in which telecommunications can provide a fertile environment for the evolution of a community.  The participating teachers in the Alaska QUILL project were connected through an electronic network for the 1983-84 school year in which they were implementing QUILL.  Our analysis of their use of the network demonstrated not only that a community did emerge, but identified characteristics of the network that were instrumental in nourishing the evolving community.  The mechanisms that fostered the teacher community often surprised us.  Some were merely unanticipated effects of the network setup, but others resulted from gaps in the technology that might have, under other circumstances, undermined network use. In the end, the community fluorished, and the strength of the connections among teachers turned the realization of QUILL in Alaska into a community implementation. 

Electronic Networks in Education

While electronic networks were introduced in the business world over two decades ago (Caswell, 1988), experiments with school-based networks for teachers and students were relatively rare when QUILL was introduced in Alaska in 1983.  Much of the early work in educational networking had focused on helping administrators access resources and exchange information.  This early focus derived from several factors:  administrators were more likely than anyone else in the school system to have computers; administrators controlled budgets and, therefore, the acquisition of equipment; there were fewer administrators to link than teachers or, especially, students; there was a clear function of resource-sharing to be served by a network.  One of the earlier networks, for example, was SpecialNet, set up by the National Association of State Directors of Special Education to promote communication among special education coordinators.  The two most frequent uses of the network were for information exchanges and for gathering special education data for statewide record-keeping.

But in the early 1980's, there were few examples of networks linking students or teachers on which we could model our teacher network, so we had to make design decisions based on intuition and personal experience.  As the year went on, however, our work did benefit from a project that was setting up networks to link students in widely scattered classrooms.  In parallel with our QUILL network implementation, Jim Levin and his colleagues at the University of California, San Diego began to experiment with networks that allowed students of different cultures to exchange messages (Levin, Riel, Rowe, & Boruta, 1984).  Our work and theirs intersected over the course of the QUILL project and the structure of both networks benefitted from the experiences of the other.

One of the San Diego group's first findings was that a straightforward penpal system did not work well for student communication.  Levin and Levin (1985) set up an "electronic penpal" exchange between classrooms in Alaska and San Diego.  Students in both places were initially excited by the opportunity to find out about life in a place they considered exotic.  Such a communicative set-up, however, had many limitations, as Levin and Levin describe in the following:

Most of the initial exchanges between San Diego and Alaska took the form of "electronic penpal" exchanges.  While this is a good start-up activity, it is difficult both to manage and to integrate into the rest of the curriculum.  Students generate messages like the following (written by third graders in San Diego) with great excitement.

Hi! new friend from Alaska.  Our names are Karina and Manuel.  I wish if you could come to San Diego.  I wonder if it is cold over there.  Well here its warm but some times it rains.  We have some rats.  Don't you have any pets?  Whats your computer name?  Our computer name is apple.  Karina's friend is Brenda my friend is Juan.  We take care of the disks and all the other things of the computer.  Our printer gives us what we write.  We are going to have a book fair for next week.  We like to go to Balboa Park.  Good bye!  Write us soon.

Since the messages are sent electronically, the students know the messages get to Alaska almost immediately.  However, when days go by with no response, the motivation wanes.  To understand why responses are delayed, the students would have to take the point of view of the other school.  A set of 30 such messages arrive in Alaska, somewhat overwhelming the classroom which might only have 7 students.  The Alaskan teacher has to print out the messages, distribute them to students, and then organize times for the students to get on the computer to generate responses.  The whole process might take several weeks.  Finally, when the set of electronic responses arrive back in San Diego, those students that get responses are excited, but those that don't are disappointed and don't want to participate further.  Once the messages are read, there is little motivation to reread them.  In this way, the electronic penpal activity has a limited utility in an educational setting.  (pp 8-10)

Levin and Levin dealt with these problems by setting up a different social structure -- a student newswire called The Computer Chronicles News Network (Riel, 1985; Levin, Riel, Rowe & Boruta, 1984; Riel, 1987).  Students at each of several sites (in the United States, Israel, Japan and Mexico) wrote newspaper articles that were available to everyone on the network.  Students at each site then constructed their own editions of the Computer Chronicles, selecting from both those articles written at their own site and those that came in over the newswire.  This use of the network fostered more valuable communication than the penpal situation had, as Levin and Levin explain:

[this newswire activity] is a way of providing a wide range of audiences for writing.  The challenge of writing an article that is likely to get published in the edition put together in other sites is much more motiviating for writing than simply sending a penpal letter.  Similarly, it provides a functional environment for reading, since the articles that come in have to be carefully read to help decide which to include in the local edition.  (p. 10)

In the years since the QUILL in Alaska networking experiment, knowledge about the use of networks in education has undergone a slow but steady growth.  Several projects have used long-distance networks to create different types of student communities.  The InterCultural Learning Network has grown out of the Computer Chronicles News Network, incorporating more sites and additional joint projects (Levin, et al., 1988; Riel, 1987).  Especially successful have been collaborative science and social science projects, such as comparative descriptions of water cycles and the appearance of the moon, that take advantage of the geographically and culturally diverse nature of the network participants (Levin, Riel, Miyake, & Cohen, 1987; Levin, Waugh, & Kolopanis, 1988).

More recently, an electronic network has been used to create a distributed scientific research community.  The National Geographic Kids Network™ (Lenk, 1989; Foster, Julyan & Mokros, 1988) has given students the opportunity to contribute to a real scientific collaboration by networking hundreds of classrooms into a data-collection and analysis team studying problems such as acid rain in conjunction with a professional scientist who is also on the network.  As both the InterCultural Learning Network and Kids Network illustrate, a common trend in network environments for students has been the creation of goal-oriented projects that truly benefit from the participation of classrooms in different locations.

The uses of networks for teachers have lagged behind their uses for students, perhaps because it has not been as obvious how teachers would benefit from the introduction of electronic communication facilities.  In the last few years, however, several teacher networks have been developed, so that a critical mass of experiences is beginning to accumulate.  A network for secondary science teachers organized by the Educational Technology Center (Katz, McSwiney & Stroud, 1987) using the system Common Ground (Hancock, 1985), for example, linked 75 teachers in eastern Massachusetts to communicate about science and science teaching.  The results of their study underlined the importance of personal knowledge and common goals in the evolution of networking communities.  The network organizers found many participating teachers, most of whom did not know each other personally, were reluctant to write messages, viewing the network instead as a curriculum resource.   In particular, more inexperienced users who knew no one else on the network personally were likely to respond only to messages about a narrow set of science topics that they viewed as central to their teaching.  Katz, et al. suggest that an "activity" approach, similar to the Computer Chronicles News Network, in which participants worked toward an explicit common goal, might have gone further toward fostering widespread network use and the growth of a community.  Given the increasing accessibility of the technology necessary to support networking -- for teachers, students, and administrators -- analyses such as Katz's and ours are important in informing the future design of educational networks.

Precursors of the Alaska Teacher Network

No other QUILL implementation had an electronic network for teachers, although the role of networking in QUILL had already undergone a significant change through the field-test in 1982-1983.  We had begun by including MAILBAG in QUILL, with the notion of students using it to communicate with one another and occasionally with the teacher.  The Teacher's Guide made this purpose explicit by listing three goals for the use of MAILBAG:

•Encouraging written communication to varying, but specific, audiences (for example, friends and classmates).

•Allowing different kinds of writing to occur (for example, informing, persuading, instructing, entertaining).

•Motivating students to write more by personalizing the experience.          (p. 38)

We did not include inter-classroom networking for students in the original QUILL design because of the difficulty schools had obtaining modems and classroom telephone lines.  Some teachers, however, exchanged disks containing their students' writing either by physically handing them to one another or sending them through the U.S. mail.  These extensions fit with our original idea of giving students the opportunity to write personal messages with a specific audience and purpose, so they further articulated, rather than altered, our conception of QUILL. 

Because of our focus on writing by students, we hadn't originally considered the possibility that teachers would be as eager as students to write to one another.  The field test, however, suggested that facilitating teacher communication was a worthwhile extension of the QUILL design.  During the field test in the Northeast U.S., teachers in the three experimental sites wrote letters to one another on an ever-growing "teacher communication disk," clumsily passed through the U.S. mail.  In addition, during the first two years of QUILL use, our colleagues at The NETWORK partly fulfilled the function of a teacher network by publishing QUILL Scribbles, a teacher newsletter, every few months.  While these "networks" used more primitive technology than telecommunications, their popularity demonstrated that teachers trying out an unfamiliar and challenging innovation wanted to be in touch with others in the same situation.  Implementing a teacher network fit well with our original notion of providing purposeful writing environments for students.  The implementation of our writing innovation for students was the catalyst for a writing innovation for teachers.

In addition to our field test experiences, the context in which we were about to introduce QUILL affected our implementation.  Most dramatically, we recognized the need for the Alaska teachers, widely scattered around the state in groups of one or two, to remain in contact.  Since developing methods to cope with geographical isolation is a familiar activity in Alaska, teachers welcomed a new communication channel.  Their desire was a major reason why we made an effort to surmount the equipment difficulties and other logistical barriers to establishing a teacher network.

In Alaska in 1983, assembling a teachers' network was no small task.  There were many resource restrictions.  Few classrooms had telephones that could be used for telecommunications.  The nearest telephone was usually in the school office; sometimes the only available phone was in the teacher's home.  Except for a few teachers in Juneau and Fairbanks, using the network was not free, and each school district had to decide how much to allocate for paying networking bills.  Because they were also paying a fee to be part of the project, most districts could not put aside much for this aspect of it.  Finally, even the resources that existed were not reliable; both power and telephone service in the villages were more likely to be out of service than they were in the cities.  One classroom even had to contend with the computer being unreliable because soot from the school's wood stove got inside the machine.

Despite these problems, we decided to include a long-distance network in the Alaska QUILL implementation for the following reasons:

1.      Alaska QUILL teachers lived in widely-separated villages and cities.  A computer network would increase the communication opportunities for distant participants because it would alleviate many of the problems of distance, time, and space.

2.      Most of the participants would have met at the three-day QUILL training session in October, so they would have common experiences and personal knowledge on which to base their messages.  A few had known each other for several years. 

3.      A few of the teachers had already used the University of Alaska computer system's electronic mail system in computer-mediated classes offered by the University.  Several others had had experience with audio conferences, a type of conference call commonly in Alaska, in which several separate locations are simultaneously connected via a telephone link.

4.      In 1983, Alaska school districts were beginning to acquire telecommunications equipment and subscriptions to national networks, primarily as a way of tapping into large educational data bases.

5.      Appropriate networking software that was compatible with QUILL and simple to use had become generally available ("The Network Tool" from InterLearn Inc., written by Jim Levin (Interlearn, 1983)).

In other words, there was a perceived function (communication among teachers working on a common project who had little opportunity for face-to-face discussion), a facilitating technological context (users with some telecommunications experience, hardware and software availability), and a supportive social context (users with some personal knowledge and shared experience).  The teacher network we established interacted with these pre-existing conditions to produce a strong QUILL teacher community.

The Alaska QUILL Network:  How It Worked

We had few specifications for the network when we started, other than that everyone whom we considered a network member be able to access the network in some way.  In addition to the Alaska QUILL teachers and the two network coordinators (Carol and John), our network included three school district administrators (Mike, Malcolm, and Marcia), a University administrator (Dick), the software developers (Andee, Chip, and Jim) and the University of Alaska professor who had been responsible for bringing the QUILL developers to Alaska for a seminar in the spring of 1983 (Ron). 

The network was a "Rube Goldberg" device, made up of two established computer networks, a variety of modems, and a central portaging facility that served as "communications central" for the entire network.   As we examined the features of existing networks that would allow us to exchange mail, set up bulletin boards, and do computer conferencing, we discovered a maze of options, but none that met all of our needs.  So we fashioned our own network. 

We chose the University of Alaska Computer Network (UACN) and the SOURCE as backbones for our system.  UACN is a statewide network which is available to people or programs associated with the University of Alaska.  For those QUILL teachers who lived in the larger communities of Alaska (Fairbanks, Juneau, and Bethel), UACN was a logical choice because there was no charge for using it.  However, teachers in smaller communities and participants in the network outside of Alaska (primarily the developers) could not use UACN without significant long-distance telephone charges.  For these people, we added the SOURCE, a national consumer data bank and mail system. 

The SOURCE was available to anyone in the United States and several other countries who was willing to pay "connect time" charges (which include long-distance access charges, but are far less expensive than long-distance telephone rates).  In addition, the SOURCE provided a facility called PARTICIPATE (PARTI for short) that allows users to set up special "conferences" - groups of people who were interested in a particular topic, such as the use of computers in education.  The system supported a separate exchange of messages within each of these groups, which were open to anyone in the entire SOURCE community.  Interestingly, though, our decision to use the SOURCE was based less on economic and technological concerns than on social ones--several people who we wanted to be part of the network (the University of Alaska Education Department,  the Iditarod School District, and Jim Levin, in particular) already had SOURCE accounts.  Had the decision been based solely on cost and services, we might have chosen CompuServe as our second network, but the networking decisions made by other people became the determining factor.  Because computer networks are not yet connected by a central system like that which connects our telephones, people's choices about subscribing to a network are often based on the other people who have access to it.  This was just one example of the profound effect the social context had on the network; even before the network was formally set up, our choice of technology was determined by social considerations.

The advantages of using two separate networks to build our QUILL Network included financial benefits, access to the resources of two systems, and a network with wider accessibility.  The obvious disadvantage was the need to transfer messages between UACN and the SOURCE:  in other words, to create a network gateway.  The fact that the account set-ups were different on the two systems added to the logistical complexity.  On UACN, several people had individual accounts; for the project we added a group ID called "FAQUILL," which was available to everyone in the project who had access to the UACN system.  On the SOURCE there was no such group account; mailboxes on that network belonged to individuals (Carol), school districts (the Itidarod district, to which four of our participating schools belonged), or institutions (Chip and Andee shared an account at BBN).  Messages intended for the entire group always needed to be transferred from one system to the other, as did messages that were written by an individual on one system and were addressed to someone on the other.  This arrangement had the unintended effect of creating a large number of messages with varying degrees of "privateness," all of which became more public in the transfer process.

The role of gateway fell to Carol, as part of her job as coordinator.  She and her official co-portager, her son John, became quite proficient at the mechanics of transferring files between the two systems.  John described his part in portaging as follows:

...one of the most important jobs I had following the conference <training session> was the portaging of messages from one computer system to the other.  I figured out the best system for doing this and for awhile it was fun and interesting.  However, after about two months of endless courier service, it got to be quite boring so I taught Mom how to do it, and from then on she got to take care of all the messages that needed to be transferred....I guess that during the QUILL year, I just provided lots of computer support for my mother.  She didn't have the time or patience to spend several hours at a time at the computer like I did.  I was happy to do the jobs that were interesting to me, and she usually paid me by giving me a certain amount of time on the SOURCE each week.

With Carol and John serving as the link between the networks, the structure became that shown in Figure 7-1.

Teachers in Juneau, Fairbanks, and Bethel had free access to both their individual mailboxes and FAQUILL, the group mailbox.  Teachers in Shungnak, Chevak, and the Iditarod District used the SOURCE because phone charges were less expensive than with UACN.  Teachers on either system could send messages to any mailbox or set of mailboxes on their own network.  Thus, the SOURCE users could send a message to individual addresses such as Carol, BBN, or Iditarod, or to some group of these.  Similarly, the UACN users could address a message to any individual mailbox on UACN, to the FAQUILL mailbox, or to some set of these.  Messages sent to the FAQUILL mailbox were always "group" mail that their senders intended to reach the audience on both networks; as time went on, in fact, they probably forgot that an intricate forwarding process was involved in getting their message to everyone.

Due to the network's complexity, many messages took circuitous routes from transmission to receipt.  Figure 7-2 illustrates the path of a message sent by students in a classroom in Shungnak, Alaska to the QUILL developers in Cambridge.

After the students completed their message at school, their teacher took it home (next door) to the Apple she had set up for telecommunications.  From there she uploaded it using Micromodem (MM) onto the QUILL group mailbox on UACN.  When Carol read it in Fairbanks, she transferred it, again using MM, to the BBN SOURCE account.  It was Chip's turn to check the SOURCE mail that day; in addition to reading the message, he sent a copy on to Andee on the BBN internal mail system.

The portaging technology was cumbersome, but it worked.  The real challenge in portaging surfaced as Carol and John tried to develop some appropriate social protocol for exchanging the messages.  Who should get messages that were addressed to an individual, but contained generally useful information?  Were some complex technical messages to and from the developers better kept private so as not to confuse less technically adept users?  They soon discovered there was an "art" as well as an "act" to proper portaging, an art that required sensitivity to the personal and informational needs of both the community and the individuals in it.  Doing the job well required an interest in and familiarity with the information being transmitted, and an ability to provide a good deal of encouragement, humor, and moral support.  They were thus an intelligent gateway that carried out tasks no automatic device could have handled.

A substantial load of other managerial tasks also fell to Carol.  In addition to taking primary responsibility for managing and routing information, she "cleaned" the group mailbox (i.e., saving some messages in separate files, flagging some for immediate attention, deleting others); provided the users with updated information on UACN and SOURCE commands and procedures; oriented new participants; acted as liaison to UACN, requesting changes to make it more useful for the teachers' network; and monitored the billing procedures for both networks.  She developed charts to keep track of the computer-related aspects of messages (e.g., file name, account, and length), their content (subject, message), and their routing in the network (sender, sent to).

As a supplement to the computer network, Carol also sent out monthly packets of QUILL-related material to everyone on the network.  Included in each packet was a paper copy of all the computer mail, copies of students' QUILL projects (newspapers, poetry, stories, booklets etc.), a short newsletter, and copies of magazine articles that were relevant to the project.  The postal mail and the computer mail, while they were somewhat redundant, turned out to serve as valuable complements to one another.  Through the mailings, information was shared that would have been too difficult or expensive to send via electronic mail. 

The packets often helped to ease teachers into using the computer network.  Teachers who read them saw that the information being exchanged on the network was not limited to hardware and software issues.  Writing standards were informal, with errors and awkwardness generally acceptable. There were many useful ideas about classrooms in general.  Moreover, there was humor; the exchanges occurred within a social network that looked appealing.  To be part of the group one would need to join the network. 

General Characteristics of the Network's Use

The network was fully operational early in November, 1983, soon after the QUILL training ended, and was in active use throughout the school year, until the beginning of May, 1984.  Of the approximately twenty-five potential users of the network, about half of them used it regularly, while most of the other half used it occasionally.  The variable that appeared to account for most of this variation in use was simply "ease of access."  All of the people who used the electronic mail system routinely had a computer, modem, and phone line that was set up at school or at home so they could use the network without making any major changes to the hardware.  Those who used the network less frequently had to move their classroom computer into a school room with a telephone (often the principal's office), walk to a school district office, or drive to a university computer terminal location.  While these differences in access were sometimes accidental - i.e., the school district had independently set up a networking capability for projects other than QUILL - they were often a reflection of teachers' interest in and commitment to computers and networking.  In particular, most of the teachers who had computers at home had gone to the trouble to set them up before the QUILL project, so that they could figure out how to use them in their classes.

During the six months when the network was fully operational, more than 300 messages were exchanged among teachers, administrators, and developers through the network gateway[1].  Most of these messages through the gateway were "broadcast" messages addressed to all of the network participants; only a few were private messages sent from one individual to another.  Their topics ranged from straightforward discussions of QUILL software through exchanges of information on language arts teaching to personal news.  Teachers not only used the network to ask and answer questions, but also to share feelings about the project.  The positive comments were generally glowing reports of their students' writing and enthusiastic suggestions of projects they had found successful.  The negative comments were most often provoked by frustrating experiences with the hardware or interactions among the several pieces of software involved in sending a message over the network, which made the process much more complicated and time-consuming than it needed to be.  One of the more expressive messages of this type came from Ernie, late one night:

AUGHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHH!!!!!

Carol sometimes I just want to throw this computer out of the window!  I have been here at school for almost three hours trying to copy my mailbox to a file so I can keep them....If you heard a loud scream Sunday night about midnight, it didn't work! (2/6/84)

During the year, the network's use went through four overlapping phases, determined by the progress teachers were making with QUILL implementation and the rhythm of the school year.  The time line in figure 7-3 illustrates these phases.

 

Figure 7-3.  Phases of the Alaska Teachers’ Network.

The first phase was marked by messages about problems in getting QUILL started, as well as by greeting messages.  The second phase was marked with a concern for integrating QUILL into the language arts curriculum.  The third phase was marked by increased message traffic and a diversity of topics.  The fourth phase saw fewer messages, as the school year wound down, and a focus on the future.

Typical messages from the first phase (October - December) include:

Greetings from Shungnak!  QUILL is underway...  we're all hanging on and going with it.  The kids are learning quickly.  When Chip was here we also had a visit from the supt.  The students took to QUILL like a sled dog to pulling.  There are a few management problems, but no major bugs with the program other than that if a student wishes to print out a mail message that they have just sent, the printer does some wierd symbols -- Chip, you saw that when you were here.  The only other problem has been with a kid named Johnson.  For some reason part of his messages never appear when he checks mail.  If we type in his name in all caps, and the sender did it in lower case would that make a difference?

Time to go send some mailbox messages...  Bonnie (11/6/83)

Greetings from Chevak!

This is Ernie writing to you with Lena in the co-pilot seat.  I have begun to use QUILL (Mailbag) in the classroom with good results.  I also have been doing a lot of pre-writing activities with the kids.  My biggest problem at the present time is how to run the computer along with a class of 24 very active 2nd and 3rd graders!  I have the kids blocked into 30 minute slots at the computer with three kids to a group.  The problem I have with that is kids don't always think in 30 minute blocks!  The kids named the computer Mr. Halloween and the machine has taken on a personality of its own.  More as it develops............

Well that is all for now, as they say in Chevak, See you on Seair!!!!  [the airline serving Chevak]

   Lena and Ernie (11/9/83)

The second phase began as teachers felt comfortable with the QUILL hardware and software and the networking procedure.  At this point, they had learned the mechanics of QUILL and were concentrating on integrating it into their writing curriculum.  Helen, from Telida, sent this message early in January: 

"...One thing happened in my classroom which seems to be a partial answer to the question about fitting in computer time and not neglecting regular classroom work.  I had been having some standard mid-year panic about all the things the kids don't know, and my reaction to it had been to break out the old workbooks, and start handing out sheets of exercises -- punctuation, capitalization, etc.  The kids' reaction was not too unexpected -- "Sick!"  "Boring!"  "Do we have to...?" etc.  I had a hard time mustering up too much enthusiasm, and within a few days, my own insecurities had passed, and we were back at work on our class newspaper, for which there was considerable enthusiasm.  When it came time to address the newspapers, it wasn't a popular task, but there was a reason for it, and it was accomplished without grumbling.   (Apart from a reasonable request that I would do "this long one" -- I had put Chip and Andee on the mailing list, and the kids balked at "Bolt, Beranek, and Newman")

Anyway, while they were addressing the newspapers, I remembered that the workbook exercise, which had been so distasteful to them, involved identifying which letters in addresses should be capitalized, and why.  Here was a perfect, natural, way to come at the same learning. 

I also suggested that they could put the mailing list on the library disk, so they wouldn't have to write it out every time, and this was accepted as a great use of QUILL.

Here there was a real reason for being as accurate as possible, and it was easy for the kids to see the need for it.

I guess I'm just trying to suggest that we let as much reality as possible into our classroom, and let the children put it together when they are ready to do so.  It seems to stick better that way, and it's easier for all of us. (1/3/84)

The next few months saw the third phase:  a flurry of messages exchanged among a large portion of the QUILL teachers.  These included requests for and offers of curriculum assistance, complaints and exultations about the software and hardware, personal information (job changes, budding relationships!), technical discussion about features of the software and general reporting about the progress of QUILL in their classrooms.  Part of such a message sent from Wilma in McGrath follows:

...We haven't been using MAILBAG much since the initial blast-off, but LIBRARY is very popular with my 5th graders.  We have used it for our newspapers, reports, haiku, Christmas stories, creative writing, and lots of miscellaneous piddling around.  We've used PLANNER for movie reviews, reports on American Indians, and now for reports and oral presentation planning on Heroes of the American Revolution...(1/31/84)

Messages also started to arrive during this period from several "late arrivals," teachers whose equipment had been delayed, who had missed training in Fairbanks (as one entire group had, due to bad weather), or who for some other reason had taken a longer time to get involved with QUILL and the network.  These messages tended to be short, intended primarily as practice with the technology - yet, they were real in the sense that an audience was truly awaiting their arrival.  Here is one such message from Judy in Holy Cross:

Hello, Carol.  I'm just here in Mike's office in McGrath and we're trying to learn about the Source and this whole system in one easy lesson!!  Do you think it will work? (1/16/84)

As April drew near, so did the Alaska Association of Computers in Education (AACED) conference where we had arranged to present a workshop on the QUILL in Alaska project.  Most of the teachers planned to attend the conference, so they regarded it as a reunion as well as an opportunity to tell a wider audience about their work.  Rather than the event generating additional communication, the number of messages on the network actually dropped somewhat during the few weeks before the conference.  Anticipating the opportunity to see each other in person, teachers were less likely to use the network to communicate; instead, they stored up questions and comments to share at the conference.  In addition, Chip and Andee visited most of the QUILL sites either before or after the conference, so communication with them did not need to rely on electronic mail.

After the AACED conference and the round of site visits that accompanied it, network use entered its fourth and final phase.  The next big event for QUILL teachers was the end of the school year.  During the last month of school, several of the rural teachers were also making arrangements for new jobs, sometimes involving a move to another village.  Teachers were less worried about "making QUILL work," since the year was almost over and their presentation in Anchorage was past.  Thus, in this phase, communication was driven by two over-riding concerns:  planning for a summer training session that would extend the QUILL Alaska project and cementing individual relationships that had developed during the year.  Many messages such as the following from Carol to Chip and Andee were involved in the establishment of a two-week summer training session at the University:

We had our "Quill Administrative Council Meeting" on Monday and all present agreed that the University ought to do whatever it can to keep Quill rolling in Alaska.  We feel that the most important role that the university can play right now is to try and provide support for training this summer.  The best avenue for doing that seems to be via the regular "Summer Session" that is offered here in Fairbanks.  (5/1/84)

The culmination of this planning was communicated to all QUILL teachers in the following message from Carol:

Good News!  We've been able to work out an arrangement for some QUILL training for this summer.  Kathleen Starr, a co-worker of Chip's and Andee's at BBN, will be coming to Fairbanks in July.  I've enclosed an announcement about the classes she'll be offering.  There will be two levels of training... one for beginners and one for "experienced" people (which naurally includes all of you!).  Hopefully, completion of the advanced course will lead to certification as a QUILL Trainer.  It's late to be putting out announcements about summer classes, so I'd appreciate your help in passing on the information about this QUILL training.

The exchange of personal messages between people who had developed an ongoing collegial/personal relationship continued during the rest of the year.  Bonnie had been developing a culturally-relevant test of writing ability for the Inupiaq children using QUILL in her school.  Two weeks after the AACED conference, she sent Chip a progress report on the testing, enlisted his assistance in scoring the children's compositions, and added the latest chapter to a long-standing joke between them about the weather in Alaska.  Chip had always been vaguely disappointed with the lack of extreme weather during his visits to Shungnak, and professed to doubt that it ever got really cold there.  Bonnie countered:

I'd tell you that after you left the temp dropped to -30 again, but you probably wouldn't believe me.  I'd tell you that right now it's snowing up a storm like we haven't seen all winter, but you probably wouldn't believe me... (4/30/84)

and Chip replied:

Your weather stories are wonderful, sometimes I almost believe them.  (5/1/84)

And in between the arrangements for summer sessions and field test evaluations, some lyrical language even appeared on the network:

Saw many cranes, and have heard spring bird songs all day.  The temp is 70 in the sun...  Folks say the caribou migration is like a river on the snow...  I saw it once, won't forget it, but can't do it justice with words...The sun only sets now for a brief time, what a place of ying and yang...(from Bonnie) (5/6/87) (117)

How A Community Was Formed

The content and frequency of network messages, people's interactions during and after the AACED conference, and the connections several of the Alaska QUILL teachers maintained several years after the project all evidence the fostering of a community by the Alaska network.  Even though the electronic network became much less active after June 1984 and essentially disappeared later that fall, several of the individual relationships that had been established continued to thrive through the more standard communication devices of mail and telephone.  In particular, some of the most interested teachers continued to be in touch with Chip, Andee, and Carol, even as they changed jobs and locations.  Letters from these teachers over the next few years provided some indication of the role the network had played in their lives.  For example, Bonnie wrote in May of 1985, after a difficult year of being principal/teacher in Kobuk, where she had been teaching during the QUILL network year:

I think a dose of the QUILL network could have helped.  I didn't realize how truly valuable that link was to my enthusiasm.  (They say phones should be in this summer-I need to be able to talk with some folks with the same frustrations and concerns.)

And a year later, she still missed the contact the network had provided, as she commented to Carol:

I wish I had the network going still - it's like a hole now - there's nothing to replace it.  I wish I could get hooked back into another network.  In Alaska when you're in small places you often don't have people who share the same philosophy.  You need someone to fuel new ideas, so you don't lose enthusiasm.  The QUILL network was great because we knew we always had an audience - someone was there to read and respond to our ideas, thoughts, and concerns.

During the 1984-1985 school year, several of the Alaska QUILL teachers worked as QUILL teacher trainers after taking the QUILL advanced course during the summer.  This gave them the opportunity to communicate with one another about substantial topics -- how to organize a training session, the relationship between QUILL and the Alaska Writing Project, etc.  A letter to Chip and Andee from Helen made clear the importance of these training sessions as a way for teachers to keep in touch:

...it is great that there are quite a number of training sessions set up, and we are managing to do them all... I really think I'll enjoy having this way of sharing some thoughts about teaching and writing, as well as the specifics of QUILL.

Because she had moved from Telida to Fairbanks in the summer of 1984 and temporarily stopped teaching, the training sessions were especially important to Helen as a way of keeping in touch with the teachers she had gotten to know the previous year.  The following excerpts from a letter she wrote to the other QUILL trainers after a session in November of 1984 illustrate both the content of her interactions with the other teachers and the emotional connections that accompanied them.

Marcia and Dick and I just finished a 3-day training session, and I thought it would be a good idea if, any time any of us does a training session, we make a note of things which went well or problems which arose...

Alex came in and gave an excellent presentation.  Since we all have experience in such a wide variety of situations, I think it's important to pay close attention to each other, so that our sessions about classroom management can be as wide ranging as possible... Parents need to know that QUILL doesn't make corrections in spelling and punctuation for the children.  [Alex] got a lot of questions about that at conferences...

Happy Thanksgiving--I'm thankful for people like you in the world!

Helen also attempted to prolong the QUILL community by trying to start a QUILL Magazine in which students could publish their best writing.  In a letter to Alaska QUILL teachers (both old and newly-trained), she noted the need for "a real reason for students to edit their best work into a final 'polished' form" and proposed the magazine as a solution. 

There is even some evidence that more long-range social connections resulted from the year of frequent communication.  During the 1987-1988 school year, three teachers who were involved in the QUILL project moved to McGrath, where there were already three other QUILL veterans.  It is at least possible that the social connections these teachers had formed via the QUILL network had some influence on the personal and career decisions that led to this concentration of former QUILL teachers in one village.

The effect of the network on the teachers' social context over the course of the year gradually transformed the QUILL project in Alaska from a twelve-classroom implementation to a community implementation.  And this change from a collection of classrooms to a community affected the way QUILL was implemented.  "Contagious" ideas for writing topics found their way from one classroom to others via the network.  Several teachers borrowed Ernie's "Describe a possible cure for baldness" idea, and newspapers appeared in many classrooms.  A series of autobiographies and autobiopoems appeared in the middle of the year.  In response to several requests delivered over the network, Bonnie shared her autobiography Planner with the rest of the group and suggested it could be used for biographies (e.g., of grandparents) as well.  The network also provided a way for frustrations that may have stopped an individual teacher to be shared, absorbed, and defused in a community setting.  Early in the year, questions about classroom management were posed and discussed over the network.  Bonnie helped several teachers who were concerned that having only one computer meant that students couldn't read each others' work on the machine:

In Shungnak QUILL classrooms, there never seemed to be any time for reading other students' work on the computer.  We have a solution to the problem.  The kids print out a copy for the QUILL book -- a "magic stick" photo album.  We have organized it so similar stories are on a page.  The book will be placed in the library at the end of the school year. (2/3/84)

The network community also provided support for teachers when they feared the effects of their hard work wouldn't show up on standardized tests.  Helen, in particular, was concerned about this, so Chip responded:

I wouldn't worry about the test scores.  We never considered looking at them in the first place as a means of evaluating QUILL because we didn't believe (nor did any one we talked to who knew about testing) that limited computer use over a short period of time would result in noticeable general achievement test score gains.  [We should look instead] at things like amount of writing, attitudes, amount and types of revision, kinds of writing, use of writing in different subject areas, and so on.

Such reassurance helped teachers go on with the process of implementing QUILL, even when they were having difficulties or temporary doubts.  By providing technical and curriculum resources, discussions of classroom management and pedagogical approaches, an outlet for frustration, and general support, the Alaska teachers' community supported and changed individual teacher's implementation.  How did the network contribute to the growth and structure of this community?  Not surprisingly, characteristics of the audience and purpose embodied in the network, the same characteristics of writing environments we emphasized in QUILL, account for much of the evolution of the community.

Writing to a Friendly Audience.  Even though messages on computer networks are almost always delivered, there is no guarantee that anyone will answer them.  One frequently-heard complaint about computer networks and conferences, in fact, is that authors of messages often get no response at all.  Even worse, they may have no idea whether anyone has even read what they have written.  The lack of response is particularly frustrating if they have spent a long time carefully crafting an argument or opinion.  After a few experiences that remind them of being "a lone voice in the wilderness," people often drop out of electronic networks. Communication with an unresponsive audience is not really communication at all.

In contrast, participants in the Alaska QUILL network knew that any message they wrote that needed to portaged between UACN and the SOURCE would be read and taken seriously by at least one person:  Carol.  In part due to the clumsy configuration of the network and the need for Carol to reroute messages appropriately, none of these messages were without an audience.  Questions that came to Carol over the University of Alaska network had to be explicitly sent on to Cambridge via the SOURCE if they were to be answered by Chip or Andee, and any message that was to be read by the entire group had to be moved from the SOURCE to UACN or vice versa.  The traces of Carol's reroutings are easy to find in the messages:

I don't know about Johnson's problem, but I will send the message on to Boston, Bonnie.  (11/7/87)

The following message was sent on The Source by Mike Baumgartner in McGrath and Don Stand from Nikolai.  Don is going to be participating in the Quill Project and when he sent the message he was in McGrath learning how to do Quill.  Some of you already know Don (from the Small High Schools Project or from graduate classes at the U of A or from his earlier participating in computer networks).  As soon as he gets all his equipment, we'll be hearing from him.

Welcome aboard, Don!  [Don's message follows] (1/19/84)

In addition to accomplishing the routine message transfers that resulted from the structure of the network, Carol became an "information broker," deciding for each message she reviewed whether there were portions of it that should have a wider audience.  Thus, some of the personal messages she received remained private, portions of others were passed on to Chip and Andee for technical comment, and others were distributed to the entire network.  Early in December, as network traffic began to accelerate, Carol faced head-on the decisions she had to make in this "triage" role.  The following is a message she sent to Chip and Andee:

I'm wondering if you do want to be barraged with ALL of the mail from up here or do you want me to screen it first and just send you the parts that you need to respond to or would you like me to send all of it to you, but indicate just those parts that I'm not able to answer.  (12/9/83)

Carol's acceptance of this role meant that she had to read and analyze each message to understand what function she was meant to serve in relation to it.  A consequence of that task was the fact -- obvious and important to everyone on the network -- that Carol was a guaranteed audience for anything they might write.  In fact, some teachers who actually wanted to send group messages but were intimidated by the prospect of their writing being seen by so many different people, relied on Carol as a knowledgeable filter who could assess how appropriate it was to make their messages public.

Carol's job did not stop at the administrative task of routing messages, but extended into being a touchstone for the teacher community.  In particular, she was aware of the teachers' feelings of both frustration and accomplishment and tried to use her moderator role to share the burden or the success.  The following is an especially straightforward attempt at sharing the burden that Carol included in one of the mail-out packages she put together.

Just wanted to share with you part of one of Ernie's late night messages.  I think his "AUGHHHHH....." is a perfect description for the feelings of frustration and exasperation that we've ALL experienced at different times (and for different reasons) with this whole computer business.  Being a pioneer in this area isn't easy, but this group of QUILL teachers sure keeps plugging away despite all of the obstacles.   [Ernie's message follows this introduction; it is included separately in this chapter on page 142.]  (2/6/84)

Because of the group's acceptance of Carol's triage role, the complex physical structure of the network and the fact that sending a message required some technical effort, messages were often addressed to several audiences - both individual and group - at once.  The following message from Sandy in Holy Cross is typical:

Just a note to let you know that the X-CED computer in Holy Cross is still sick --- in other words, we have not done much with QUILL as yet.  Judy and Joe are moving right along however.  Chip, what did you do with my flashlight when you left Holy Cross?  (12/8/83)

While the above message moves from a public audience to a private one, other messages started by being addressed to an individual (often in response to a personal message) and almost imperceptibly became more public by the end.   The following excerpt of a message from Andee to both Ernie and Lena in Chevak illustrates both the personalizing of a piece of a message (by using the direct address "Lena,") and its eventual shift from private to public conversation.

...Lena, your question about giving students blocks of time is a common one.  Teachers have found that they have to make their classrooms a little more flexible than they have been.  Some teachers structure their classrooms as activity centers in which each group in the class is doing some different piece of school work; the computer then works well as one of these activity centers.  In some classrooms, teachers have used QUILL especially when reading groups were working.  In others, teachers have always had someone on the computer, even if he or she missed something else in the classroom, on the assumption that it would even out in the end.  Does anyone else have any suggestions? (12/23/83)

Such a private to public shift appears to be a feature of electronic mail, because of the ease with which writers can send copies of messages to many people at once.  It was prevalent in the Alaska QUILL network messages because of the task-oriented nature of the network.

Occasionally, especially as experience with the network accumulated, some participants commented on the unusual assumptions about audience our messages implied.  A message from Andee to Carol in February includes several phrases that evidence an awareness of this pattern.

I've just written to Helen, Wilma, Mike, Deane and Bonnie and am finally getting around to writing to you again.....

I don't know the details yet, but it looks as if both Chip and I will be able to attend the AACED conference - I hope that convinces a great number of you to come, too, so we can have a reunion.  Let's find a really unique restaurant in Anchorage and have a rowdy time....

I realize this message is strange - started out being to Carol, ended up being to QUILL in Alaska - it just shows what happens when you're not sure who the audience for a text is!  (2/14/84)

Besides saving on message transmission time, the inclusion of multiple audiences in the same message gave network participants the feeling of "eavesdropping" on conversations that were not really addressed to them, as if they were in a large family that held its personal conversations in the kitchen.  Since network members knew their messages were likely to be read by a large group, the messages were never so personal that it was inappropriate for them to be shared, but their semi-public nature assumed a kind of familiarity and trust even before it was well-established in the group.  In fact, this consequence of the technology probably fostered the very sense of community that it presupposed. 

"Mixed messages" (including both public and private communications) were possible in part because of another characteristic of the network audience: a set of mutually shared experiences.  Almost all of the teachers on the network had attended the training session in Fairbanks, and Chip, Andee, and/or Carol had visited every classroom in the week following training.  All the teachers were working with the same software and curriculum materials; many were struggling with the same implementation problems.  The size of the group, too, was conducive to this sense of sharing; about 15 people were regular contributors to the network, and keeping track of their classroom and personal situations was possible.

Partly in recognition of this shared background, almost all of the messages on the network were broadcast to everyone; even if they were addressed primarily to one person. These communications further expanded the common background that the training session had begun, and network participants took advantage of this by writing messages that assumed a large amount of shared knowledge.  One result was several on-going public discussions, some about educational issues and others that were expressions of the shared (often humorous) themes that fed the group identity.  One such theme focused on the weather (a frequent topic in Alaska) and, in particular, on Chip's complaints that he hadn't experienced colder temperatures.  On a more serious topic, several teachers exchanged information on their use of videotape as an observational tool in the classroom.  In January, Bonnie broadcast the following message; it clearly presupposes a group whose members knew each other pretty well and had all read a message Helen had sent previously about how her reaction to mid-year insecurity was to "break out the old workbooks." (1/3/84)

Today I hooked up the video equipment and filmed us QUILLing.  I haven't had much time to look at the tape yet, but have a few impressions to pass along.  1-There is a lot of good stuff happening at the computer, but I'm too busy "teaching" to get in on it.  Although the audio is very hard to hear, from the non-verbal signs I can see kids very absorbed in what they are doing...talking over some part of the story, discussing how to move the cursor or work some command, or find a key on the board.  2-They are not afraid of the equipment at all. The kids are becoming more and more comfortable with the hardware as time passes.  I know that none of this is earth shattering news, but as Helen stated, with the coming of the second semester, there's an urge to drag out all kinds of meaningless book stuff.  What is happening with kids and QUILL is making learning worthwhile. (1/18/84)

In contrast to their active participation in their own network, the Alaska QUILL teachers took little advantage of the computer conferences on the SOURCE.  PARTI offered several conferences relevant to QUILL (e.g., Computers in Early Childhood, Computer Teacher) and, in addition, a QUILL conference set up by Sandy Levin in San Diego, primarily to establish pen-pal relationships between classrooms in San Diego and Alaska.  Carol, in an extension of her role as network triage person, sent an introductory statement about the Alaska project to the QUILL conference and sent information about PARTI to the rest of the Alaska teachers.

These exchanges, however, never became popular with the Alaska QUILL teachers.  While the PARTI community was available to Alaskan teachers as an extension of their own network, their participation in it was almost totally limited to a few messages that Carol sent, a few that Ernie sent, and an interchange Bonnie had with the Levins about her use of Planner.  It is likely that their lack of interest in PARTI stemmed from the fact that the Alaskan network met their needs in a way the larger network could not easily accomplish, and their involvement in the Alaskan network left little time or psychological space for the effort that PARTI participation would require. 

Members of the Alaska QUILL network, then, felt they were addressing a reliable and familiar audience, carefully managed and "nurtured" by Carol.  They relied on a constantly-growing shared background to support their communication.  This sense of "being listened to" and "being known" were major contributions to a feeling of community and made them willing to put in the effort necessary to become familiar with the network procedures and to contribute on a regular basis. 

Addressing Common Goals.  Closely linked to the Alaska QUILL teachers' shared background were a set of common purposes that provided a starting point for the formation of the network.  When the network became more robust, this set expanded, as participants realized that the network could satisfy some educational and personal goals they had not identified before. 

At first the participants' most immediate goal was joint problem-solving about QUILL itself.  Everyone was painfully aware that the software was less than perfect and that curriculum and classroom management details needed additional attention.  The teachers needed to be in touch with one another to discuss details of the software and classroom implementation.  Soon after the training session, a student in Ernie's class lost a MAILBAG message because a disk was too full.  Ernie's first message, sent jointly with Lena, asked for help in avoiding the situation.  Eventually, he received three separate responses to his message:  one from John promising to investigate, one from Bonnie confirming that she had run into the same thing, and one from Andee explaining the source of the problem and a temporary solution.

Bonnie took it upon herself to explain to the other QUILL teachers how to move quickly through the QUILL menus to get  to the word processor.

For some of you who had asked about getting into the actual writing faster, Chip showed us how to type more than one number for a command....If you know the number of the choices you will be making, you can type all the numbers one after another then put in the 3 returns.  The QUILL program will take all that information and do all those tasks without pausing to ask you for the information.  Good luck!  (2/3/84)

Teachers also exchanged ideas about using Planners to write autobiographies and interviews, evaluating the writing their students did both on and off the computer, using a cardboard "post office" and paper flags to indicate when individual students had mail in their mailboxes and specific topics for group writing (e.g. cures for baldness, ethnic family stories).  Whatever the specific topic, this communication was clearly and purposefully focused on getting QUILL to work smoothly and productively in the classroom; teacher participants in the network used it as a medium for tackling their own implementation problems and sharing their solutions.

Since we were also on the network, teachers viewed us as a source of rapid feedback about the software and, occasionally, of a new version with fewer bugs.  But the presence of the developers added another shared purpose:  producing a piece of software that would be successfully marketable.  This co-developer sense encouraged Mike, an administrator from the Iditarod district, to make this suggestion:

Deane is using QUILL with her Jr. high class.  We tried to teach it to her 9th graders, they rebeled because it takes so long to boot and to move from screen to screen to start writing in Library.  Isn't there something Andee and Chip can do about that?  That might be why it is billed as appropriate for grades 3-6, but it could keep the program from expanding to many classrooms. (1/30/84)

Ernie took his role as a participant in development so seriously that he wondered aloud toward the end of the year whether QUILL was really ready for the market: 

This is the third time such an error has occurred with one of my disks...I wonder if the program is ready for the market yet.  With all the problems in McGrath and the problems with my disks I wonder if more time should be given to research on QUILL when it is used for long periods of time.  I hope Chip and Andee have some suggestions.  I think that the program is fantastic, my kids love it, but maybe it needs more time to smooth out the rough edges. (3/25/84)

The teachers' influence on development meant that there was at least one purpose that all network participants shared.  While other purposes served by the network were common concerns primarily for the teachers (e.g. figuring out how a computer influenced their classroom organization), this one universal goal made all the network participants into a sociologically "flat" community.  Some teacher networks - such as the Educational Technology Center science teacher network described at the beginning of this chapter - have included "experts" whose main role was to provide information and answer questions.  In the QUILL network, on the other hand, teachers did not see their relation to the developers as strictly one of "novice user" to "expert"; rather, there was an honest sense that everyone could make important contributions to improving the program. 

Computer networks have in general been acclaimed as opportunities for participants to be less aware of the social status information obvious in face-to-face interactions (sex, skin color, ethnicity, physical handicaps) and traditional forms of written communication (letterhead, type of stationery).  Because the Alaska network's participants knew each other, an egalitarian context could not grow out of ignorance; rather, it had to grow out of a shared purpose to which everyone could meaningfully contribute.

In addition to these group goals of making QUILL work, many participants in the Alaska QUILL network were fulfilling a significant personal purpose: communicating with colleagues.  About three-quarters of the QUILL teachers taught in remote village schools alone or with only a few other teachers.  A general lack of communication facilities also characterized several of these villages; most of the telephones in Telida, for example, had been installed only a few months before the QUILL project started.  The distances between villages make teachers in Alaska more acutely aware of their needs for interaction than teachers in urban areas, whose isolation from one another is less complete, yet still significant.  Ernie asked for help on a curriculum development task:

I am going to write up or should I say attempt to write up a scope and sequence for grades K-8 dealing with writing.  I want to use the QUILL and Writer's Assistant as the main ingredient of this.  I also want to use ideas from the writing consortium which I am going to this week in Anchorage...I realize that I can't create this monster myself, but if I get the information then myself and a few other teachers in Chevak could work on it.  (1/19/84)

Messages also contained birth announcements, recipes, requests for help in locating old friends, reports on extracurricular activities, travel plans, and confessions of guilt for not doing "enough" with QUILL.  Bonnie clearly saw the network as an important social outlet, since she wrote on a visit to Fairbanks:

The phones are out in Shungnak, I'm going through networking withdrawal.  (2/7/84)

The dispersed community even provided a legitimate purpose for the writing that several teachers did during their follow-up training sessions.  Instead of practicing using QUILL and the communications program by doing empty "exercises," teachers had a real goal - to communicate with people they had started to get to know in Fairbanks.  For the teachers from Holy Cross, who had missed training because of the weather, their personal introduction to others on the system came through their initial network messages. 

A cultural bond also linked most of the teachers on the Alaska network.  Except for Lena, all of the network participants were Anglos and, for the most part, Anglos living in the midst of a culture that was not their own.  Cultural contrasts were frequent: some were serious, some amusing.  But in each village, the Anglo teachers had few people with similar backgrounds with whom to share their reactions.  The network provided the opportunity to create a community unconstrained by space, time, or the traditional (and often rigid) boundaries imposed by individual schools, districts, cultural regions and institutions.  In her description of the year's experience, Helen commented:

Being the only teacher in Telida, I really appreciated hearing from other teachers (and others involved with QUILL). ...The important thing was the respect and support we offered each other.  It's easier to put extra effort into something if you feel like someone else is interested in what you are doing.

And Bonnie's message below humorously shared one implication of being an Anglo in a non-Anglo setting:

Have heard that the church folk are looking for a Mary and Joseph...they usually like white people to do that sort of thing. --We're hiding. (12/23/83)

The line begins to blur here between audience and purpose.  The purposes served by the Alaska network reflected the needs of its participants, and those participants for whom the network was functional continued to be part of the audience.  As in any community, common interests and common needs were the glue that held the group together.  In this case, the connections were reinforced by another somewhat serendipitous parallel.  Teachers were writing about writing and, as such, experiencing the writing process at the same time that they discussed it in their messages.  Inevitably, they faced many of the same issues in their own writing that they observed in their students, and their own experiences writing with a computer gave them a generative perspective on their classrooms. 

One of these issues was their own, students', and parents' desires for "perfect" student papers, especially when they were produced on the computer.  (See section 5.3 on revision for a more complete discussion.)  As computer network writers themselves, teachers had to face a similar decision:  how much effort would they put into making their messages technically perfect?  At the beginning of the year, network users had to type their messages directly into the message system, which supported only the most rudimentary editing.  Because they were paying telephone and/or network charges while they typed, they could not afford to spend much time editing their messages.  So, their messages were full of apologies for imperfections, asides like "oops...I know that's not right" and retyped words, phrases and even sentences.

By the middle of the year, the software supported composing messages within QUILL, and then transferring them to the network.  Several teachers, breathing a sigh of relief, started to prepare carefully-edited messages as QUILL files and upload them to the network.  Simultaneously, they were struggling in their classrooms with students who worried more about correct spelling and punctuation than about meaning-centered revisions.  The relationship between the requirements they imposed on themselves as computer-using writers and those they wanted their students to adopt gave teachers unusual insight into the reactions their students had to QUILL.

But the most important example of synergy between the classrooms and the network was in their sense of "purpose fulfilled by writing."  The network reinforced for QUILL teachers the principles set forth for the QUILL program itself.  There, by design, purpose was paramount; meaning was the mainstay of the pedagogical philosophy.  So in using the network, teachers experienced themselves what they attempted to create in the classroom for their students.  In the service of making QUILL work better, we created a "macroworld" - a real world context that embodied the communicative environment we were tyring to create in the classroom.

Community as a Property of Networks

The Alaska QUILL teachers' network "worked," in the sense that it met participants' needs by fostering a community that helped them implement QUILL more successfully.  But networks don't always work; in spite of increasingly sophisticated telecommunications technology and easier-to-use interfaces, some networks never achieve a sense of self-sustaining energy.  In our own experience, this is sometimes due to technological problems, but is more often a reflection of a mismatch between personal needs and what the network offers.  Some networks never reach a critical mass because the purpose they might serve doesn't appeal to a coherent community.  More interesting, though, are networks that continue to work for some people, but, as they mature, lose members who had originally chosen to participate.  One recent example is an international discussion of pedagogy and computers that generated long philosophical discussions that effectively excluded casual users.  Another is a tennis network that began as an exchange of information on tennis and a way for players to meet others of comparable skill.  It evolved, however, into a mechanism for setting up formal weekly tennis events, and several more casual players removed their names from the distribution list.  In both cases, the original group of participants was loosely defined as "people who were interested in the topic," but the use of the network defined a more specific community that was appropriate for only a subset of the original participants.  In both cases, the critical variable that defined the evolving community was not technological capabilities, but personal needs and interests.

While the influence of personal needs and purposes on network use may be clear to the designers of general networks, they often have difficulty integrating such a perspective into their network designs.  Thus, network planning has focused on technological parameters such as baud rate (at what rate can an individual receive or transmit information), bandwidth (how much information can be sent over a link between nodes in the network), network configuration, time between  message sending and receiving, need for special equipment, and cost.  Similarly, much of network evaluation has examined technological measures such as message volume, average response time, mean time between failures, and number of active participants.

Much of this emphasis on technological considerations is historical.  The impetus for network development in the business world came from a need to access resources (e.g., programs, databases, manuals) that were not locally available.  The first networks were time-sharing networks, whose person-to-machine, rather than person-to-person, connections made them utilities rather than potential communities.  Even as person-to-person connections were added, the first problems to be tackled were, appropriately, designing efficient ways to route messages, dealing with complexity as networks grew large and complicated, and inventing monitoring procedures to diagnose and repair network breakdowns.  Early networks sometimes acted primarily as file servers, whose main function was to facilitate the transfer of data files from one workspace to another; programs such as FTP (file transfer protocol), made this process possible without involving any communication between people.  Focusing on access to remote resources made technical requirements such as high bandwidth and quick reponse time salient; users would not continue to use a utility that required them to read a large remote data base using insufficient networking power.

Although networks were used primarily as utilities, individual purposes and needs substantially began to affect how and how often a person used any particular network connection.  Electronic mail systems added new possibilities for interpersonal interaction and issues of community began to emerge.  Thus, by now, most networks in the business, academic and education worlds involve aspects of both utility and community.  The CompuServe Macinfo conference, for example, allows Macintosh enthusiasts to exchange hard-to-find tips about making the best use of their computers.  Traffic in this computer conference is heavy; more than one user has even complained about its volume.  However, the utility aspect of this network is bolstered by a sense of community; participants in Macinfo are energetic in their reactions to the Macintosh, feel a bond with others who are similarly inclined, and even grow to recognize some frequent contributors.  This sense of community even keeps some participants from giving up in frustration over the volume of mail they receive. 

The mix of community and utility is carried even further in the growing international network known as Internet, which connects many national and regional networks.  With the recent increase in availability of networking hardware, constructing such a network has become simple, creating, in essence, the potential for a world-wide electronic community of university and industry scientists and engineers involved with high technology.  If past use of the constituent networks is any indication, Internet is likely to be an amalgam of utility and community, with personal messages and asides mixed among the requests for information, and a variety of subcommunities coalescing over time.

A more explicit recognition of the importance of social structure in network use is in the growing number of "digests" to be found on large networks.  These serve the same general purpose as conferences - allowing a subgroup of users to communicate with others who share their interest in a particular topic.  In contrast to conferences, however, digests require a single organizer who - often on a rotating basis - collects all messages on the topic, organizes them, comments on them and send out a newletter-like message once every few weeks.  This structure substitutes for the deluge of messages sometimes associated with special interest discussions and works particularly well when it doesn't matter if messages are disseminated immediately.

With aspects of community showing up and even being formalized in utility-oriented business networks, analyses of the social structures that foster and are fostered by network use become even more important.  For people setting up educational networks, who have mainly business networks as models, an understanding of the communication characteristics that are most likely to lead to a strong community of teachers or students is especially important.  And business network planners will undoubtably find such information increasingly useful as their networks become more obviously community-oriented.  One study, in fact, has examined networks in both business and education to generate a list of "participant structures" most likely to lead to successful networking communities  (Levin, Kim, & Riel, 1988).  Their general guidelines specify that a successful network should involve:

•A group of people who work together or share interest in a task, but who find it difficult to meet in the same location and/or at the same time.

•A well-specified task to be accomplished by this group.

•Ease of access to a reliable computer network.

•A sense of responsibility to the group and/or task.

•Strong leadership and final evaluation of the group task.

Further investigation and elaboration of such criteria will need to continue as technology advances and networking resources become available to larger groups of people; the day when the effects of electronic networking on communities rival those of postal mail and telephones may not be far off.

Summary

Our experience with the Alaska QUILL teachers' network provides one answer to the question posed in Riel (1989):  "Can electronic networks be used to create cooperative learning conditions for teachers as well as students?"  Our answer (as is Riel's, in describing the AT&T Long Distance Learning Network) is a resounding "Yes."   The Alaska network not only created a cooperative learning experience, but beyond that a community of teachers for whom writing was the critical link.  It thus= modeled the literacy environments they were attempting to create in their classrooms.  Our analysis of the structure and strength of the network identified two categories of characteristics that help explain its success in fostering a community, even given the imperfect nature of the technology on which it was built.  Since the network was built on communication, it is not surprising that those categories were audience and purpose. 

The salient audience characteristics that led to an evolving community were 1) previous interpersonal connections, and 2) a guaranteed, interested audience.  The network participants had met and worked together during the three-day QUILL training, and some of the teachers had taken classes at the University together before.  The audience's interest was guaranteed to some extent by a common purpose, but grew as well out of the unusual network structure that made Carol a "moderator" as she transferred messages between networks.  The assurance that Carol would read and seriously consider most messages gave the network a primarily one-to-many rather than one-to-one communication structure.  The semi-public quality of the messages such a structure encouraged enhanced a "large family" community sense, although it may have contributed to the network's eventual demise by making participants too dependent on Carol.  Still, while the network was operational, everyone knew that their messages would not go unread.

Several interacting elements of purpose also supported the teacher community: 1) common individual goals and a shared group goal, 2) a flat social organization, and 3) a synergy between network and classroom activities.  From the beginning, participants on the network shared a common individual goal:  to implement QUILL in their classrooms.  Communicating about this goal was even more critical in the first few months, before hardware and software problems had been solved.  So, a sense of a common problem to be solved provided an initial boost to the network.  As the year went on, a group goal emerged as well: to prepare a session for the AACED conference.  Thus, teachers shared individual goals and a group goal, both of which relied on the network for success.  In the pursuit of these goals, one kind of cooperative learning environment for teachers was supported by the network.

Teachers also shared a purpose with the wider network community that included the QUILL developers: to critique and improve the software.  This "flat" social organization added to teachers' sense of the importance of their comments, since they were likely to be reflected in changes in QUILL.  Finally, the network created for teachers a functional learning environment which in many ways mirrored what they were creating in their classrooms.  While they were attempting to create situations for their students that made possible purposeful audience-sensitive writing, teachers had access to just such situation for themselves.  What they learned about their own writing processes in that setting provided insights that could influence their work in the classroom, as they saw some of the dilemmas they faced in their classrooms reflected in their writing on the network.

These audience and purposes characteristics contributed to an environment that prompted Carol to comment in a message to Andee in February:

It's amazing how "connected" I'm able to feel to both you and Chip, and I really do believe that this same kind of "connectedness" wouldn't occur if we were using phones or the mail system.  Certainly there would be no way to share thoughts so readily with such an extended group if one were using the traditional ways of communicating.  Access to you and Chip is a tremendous boon to all of us...both for program information and for the "pats on the back."  (2/15/84)

and Mary Goniwiecha to write to the other teachers in the network:

I must say that I have read all the letters from you all with much enthusiasm each time I've received a packet.  You bush teachers are so good about communicating with us that we city folk are put to shame!  I'm actually in more contact with you through the newsletters than I am with my fellow Fairbanks teachers.  Tsk! Tsk!

As the community developed, it turned the network from something new and somewhat strange:

This is the first time I've gotten connected on the SOURCE.  Hi Carol and Malcolm and anyone else who's listening out there in outer space.  (Helen, 12/9/83)

to a comfortable medium, quite in contrast to "outer space" in just a few days:

Hi you guys, all is going fine and I wanted to say hello and goodnight and Merry Christmas, in case I don't get on for another week or so... (Helen, 12/15/83)

---

[1]Participants could and did exchange private messages.  We have no way of knowing the content or even the number of such messages.  The count here refers to those sent through the UACN/SOURCE gateway discussed in §7.3,