Computer-Assisted Language Learning: An Introduction

Until quite recently, computer-assisted language learning (CALL) wasa topic of relevance mostly to those with a special interest in that area.Recently, though, computers have become so widespread in schools and homesand their uses have expanded so dramatically that the majority of languageteachers must now begin to think about the implications of computers forlanguage learning.

This article provides brief overview of how computers have beenused and are being used for language teaching. It focuses not on a technicaldescription of hardware and software, but rather on the pedagogical questionsthat teachers have considered in using computers in the classroom. Forthose who want more detailed information on particular applications, atypology of CALL programs (Appendix A) and a list of furtherCALL resources (Appendix B) is included at the end.

Three Phases of CALL

Though CALL has developed gradually over the last 30 years, this developmentcan be categorized in terms of three somewhat distinct phases which I willrefer to as behavioristic CALL, communicative CALL, andintegrativeCALL (cf. Barson & Debski, in press). As we will see, the introductionof a new phase does not necessarily entail rejecting the programs and methodsof a previous phase; rather the old is subsumed within the new. In addition,the phases do not gain prominence one fell swoop, but, like all innovations,gain acceptance slowly and unevenly.

Behavioristic CALL

The first phase of CALL, conceived in the 1950s and implemented in the1960s and '70s, was based on the then-dominant behaviorist theories oflearning. Programs of this phase entailed repetitive language drills andcan be referred to as "drill and practice" (or, more pejoratively, as "drilland kill").

Drill and practice courseware is based on the model of computeras tutor(Taylor, 1980). In other words the computer serves as a vehiclefor delivering instructional materials to the student. The rationale behinddrill and practice was not totally spurious, which explains in part thefact that CALL drills are still used today. Briefly put, that rationaleis as follows:

* Repeated exposure to the same material is beneficial or evenessential to learning

* A computer is ideal for carrying out repeated drills, sincethe machine does not get bored with presenting the same material and sinceit can provide immediate non-judgmental feedback

* A computer can present such material on an individualized basis,allowing students to proceed at their own pace and freeing up class timefor other activities

Based on these notions, a number of CALL tutoring systems weredeveloped for the mainframe computers which were used at that time. Oneof the most sophisticated of these was the PLATO system, which ran on itsown special PLATO hardware, including central computers and terminals.The PLATO system included vocabulary drills, brief grammar explanationsand drills, and translations tests at various intervals (Ahmad, Corbett,Rogers, & Sussex, 1985).

In the late 1970s and early 1980s, behavioristic CALL was underminedby two important factors. First, behavioristic approaches to language learninghad been rejected at both the theoretical and the pedagogical level. Secondly,the introduction of the microcomputer allowed a whole new range of possibilities.The stage was set for a new phase of CALL.

Communicative CALL

The second phase of CALL was based on the communicative approach to teachingwhich became prominent in the 1970s and 80s. Proponents of this approachfelt that the drill and practice programs of the previous decade did notallow enough authentic communication to be of much value.

One of the main advocates of this new approach was John Underwood,who in 1984 proposed a series of "Premises for 'Communicative' CALL" (Underwood,1984, p. 52). According to Underwood, communicative call:

* focuses more on using forms rather than on the forms themselves;

* teaches grammar implicitly rather than explicitly;

* allows and encourages students to generate original utterancesrather than just manipulate prefabricated language;

* does not judge and evaluate everything the students nor rewardthem with congratulatory messages, lights, or bells;

* avoids telling students they are wrong and is flexible to a varietyof student responses;

* uses the target language exclusively and creates an environmentin which using the target language feels natural, both on and off the screen;and

* will never try to do anything that a book can do just as well.

Another critic of behavioristic CALL, Vance Stevens, contendsthat all CALL courseware and activities should build on intrinsic motivationand should foster interactivity--both learner-computer and learner-learner(Stevens, 1989).

Several types of CALL programs were developed and used duringthis the phase of communicative CALL. First, there were a variety of programsto provide skill practice, but in a non-drill format. Examples of thesetypes of programs include courseware for paced reading, text reconstruction,and language games (Healey & Johnson, 1995b). In these programs, likethe drill and practice programs mentioned above, the computer remains the"knower-of-the-right-answer" (Taylor & Perez, 1989, p. 3); thus thisrepresents an extension of thecomputer as tutor model. But--incontrast to the drill and practice programs--the process of finding theright answer involves a fair amount of student choice, control, and interaction.

In addition to computer as tutor, another CALL model used forcommunicative activities involves the computer as stimulus (Taylor& Perez, 1989, p. 63). In this case, the purpose of the CALL activityis not so much to have students discover the right answer, but rather tostimulate students' discussion, writing, or critical thinking. Softwareused for these purposes include a wide variety of programs which may nothave been specifically designed for language learners, programs such asSimCity,Sleuth,or Where in the World is San Diego (Healey & Johnson, 1995b).

The third model of computers in communicative CALL involves thecomputeras tool (Brierley & Kemble, 1991; Taylor, 1980), or, as sometimescalled, the computer as workhorse (Taylor & Perez, 1989). Inthis role, the programs do not necessarily provide any language materialat all, but rather empower the learner to use or understand language. Examplesof computer as tool include word processors, spelling and grammarcheckers, desk-top publishing programs, and concordancers.

Of course the distinction between these models is not absolute.A skill practice program can be used as a conversational stimulus, as cana paragraph written by a student on a word processor. Likewise, there area number of drill and practice programs which could be used in a more communicativefashion--if, for example, students were assigned to work in pairs or smallgroups and then compare and discuss their answers (or, as Higgins, 1988,students can even discuss what inadequacies they found in the computerprogram) In other words, the dividing line between behavioristic and communicativeCALL does involves not only which software is used, but also howthe software is put to use by the teacher and students.

On the face of things communicative CALL seems like a significant advanceover its predecessor. But by the end of the 1980s, many educators feltthat CALL was still failing to live up to its potential (Kenning &Kenning, 1990; Pusack & Otto, 1990; R�schoff, 1993). Critics pointedout that the computer was being used in an ad hoc and disconnected fashionand thus "finds itself making a greater contribution to marginal ratherthan to central elements" of the language teaching process (Kenning &Kenning, 1990, p. 90).

These critiques of CALL dovetailed with broader reassessmentsof the communicative approach to language teaching. No longer satisfiedwith teaching compartmentalized skills or structures (even if taught ina communicative manner), a number of educators were seeking ways to teachin a more integrative manner, for example using task- or project-basedapproaches . The challenge for advocates of CALL was to develop modelswhich could help integrate the various aspects of the language learningprocess. Fortunately, advances in computer technology were providing theopportunities to do just that.

Steps toward Integrative CALL: Multimedia

Integrative approaches to CALL are based on two important technologicaldevelopments of the last decade--multimedia computers and the Internet.Multimedia technology--exemplified today by the CD-ROM-- allows a varietyof media (text, graphics, sound, animation, and video) to be accessed ona single machine. What makes multimedia even more powerful is that it alsoentailshypermedia. That means that the multimedia resources areall linked together and that learners can navigate their own path simplyby pointing and clicking a mouse.

Hypermedia provides a number of advantages for language learning.First of all, a more authentic learning environment is created, since listeningis combined with seeing, just like in the real world. Secondly, skillsare easily integrated, since the variety of media make it natural to combinereading, writing, speaking and listening in a single activity. Third, studentshave great control over their learning, since they can not only go at theirown pace but even on their own individual path, going forward and backwardsto different parts of the program, honing in on particular aspects andskipping other aspects altogether. Finally, a major advantage of hypermediais that it facilitates a principle focus on the content, without sacrificinga secondary focus on language form or learning strategies. For example,while the main lesson is in the foreground, students can have access toa variety of background links which will allow them rapid access to grammaticalexplanations or exercises, vocabulary glosses, pronunciation information,or questions or prompts which encourage them to adopt an appropriate learningstrategy.

An example of how hypermedia can be used for language learningis the programDustin which is being developed by the Institutefor Learning Sciences at Northwestern University (Schank & Cleary,1995). The program is a simulation of a student arriving at a U.S. airport.The student must go through customs, find transportation to the city, andcheck in at a hotel. The language learner using the program assumes therole of the arriving student by interacting with simulated people who appearin video clips and responding to what they say by typing in responses.If the responses are correct, the student is sent off to do other things,such as meeting a roommate. If the responses are incorrect, the programtakes remedial action by showing examples or breaking down the task intosmaller parts. At any time the student can control the situation by askingwhat to do, asking what to say, asking to hear again what was just said,requesting for a translation, or controlling the level of difficulty ofthe lesson.

Yet in spite of the apparent advantages of hypermedia for language learning,multimedia software has so far failed to make a major impact. Several majorproblems have surfaced in regarding to exploiting multimedia for languageteaching.

First, there is the question of quality of available programs.While teachers themselves can conceivably develop their own multimediaprograms using authoring software such as Hypercard (for the Macintosh)orToolbook (for the PC), the fact is that most classroom teacherslack the training or the time to make even simple programs, let alone morecomplex and sophisticated ones such as Dustin. This has left thefield to commercial developers, who often fail to base their programs onsound pedagogical principles. In addition, the cost involved in developingquality programs can put them out of the market of most English teachingprograms.

Beyond these lies perhaps a more fundamental problem. Today'scomputer programs are not yet intelligent enough to be truly interactive.A program like Dustin should ideally be able to understand a user'sspokeninput and evaluate it not just for correctness but also orappropriateness.It should be able to diagnose a student's problems with pronunciation,syntax, or usage and then intelligently decide among a range of options(e.g., repeating, paraphrasing, slowing down, correcting, or directingthe student to background explanations).

Computer programs with that degree of intelligence do not exist,and are not expected to exist for quite a long time. Artificial intelligence(AI) of a more modest degree does exist, but few funds are available toapply AI research to the language classroom. Thus while IntelligentCALL (Underwood, 1989) may be the next and ultimate usage of computersfor language learning, that phase is clearly a long way down the road.

Multimedia technology as it currently exists thus only partiallycontributes to integrative CALL. Using multimedia may involve an integrationof skills (e.g., listening with reading), but it too seldom involves amore important type of integration--integrating meaningful and authenticcommunication into all aspects of the language learning curriculum. Fortunately,though, another technological breakthrough is helping make that possible--electroniccommunication and the Internet.