Call activities: Are they all the same?

Call activities: Are they all the same?

S\strm. Vol 22, No. I, pp 33-45. 1994 Copyright 0 1994 Elsevm Science Ltd Pnnted in Great Britam All nghts reserved 0346-251X/94 $6.00 + 0.00 Pergamo...

957KB Sizes 0 Downloads 101 Views

S\strm. Vol 22, No. I, pp 33-45. 1994 Copyright 0 1994 Elsevm Science Ltd Pnnted in Great Britam All nghts reserved 0346-251X/94 $6.00 + 0.00

Pergamon 0346251X(93)EOOll-F

CALL ACTIVITIES:

ARE THEY ALL THE SAME?

CAROL A. CHAPELLE Iowa State University, Ames, IA, U.S.A. For effective use and study of CALL, teachers and researchers need to be able to assess the degree of similarity among CALL activities as well as the significance for language learning of any apparent differences among activities. This paper explains how the concept of genre is useful for investigating similarities among the types of language produced in CALL actitivities (i.e. CALL texts). Examples of CALL texts are provided to demonstrate how their functional elements can be analyzed and how their significant features might be identified in light of classroom research results. On the basis of this analysis, I conclude that apparent differences in the example CALL texts are not significant. Three relevant levels of analysis for CALL activities-text, genre, and context-are clarified and their implications for CALL are explained. CALL ACTIVITIES:

SAME OR DIFFERENT?

CALL activities are often referred to as if they are all similar. For example, administrators ask whether or not the computer facilitates language development, teachers ask whether students like the conzputer, and researchers report results of meta-analytic studies to sum up the effects of the computer on students’ learning. Implicit in these perspectives is the assumption that language learning activities in which the computer plays a role are similar to one another in important ways. This “technocentric” (Papert, 1987)’ assumption is contrary to the views of some instructional computing researchers [e.g. Falbel (199 1) and Clark (1985)], who warn that it is not the computer per se which effects learning and attitudes. In concert with this warning, many CALL developers, users, and researchers consider the differences among CALL activities. For example, some have asserted differences between the following pairs: learnercontrolled vs program-controlled, game vs tutorial, and intelligent vs unintelligent software. These labels, used to indicate types of programs, imply that CALL activities are believed to be different from one another. But are the cited differences really significant for second language learning? For example, is a CALL activity designed to allow maximum learner-control inherently better for language development than one that guides the learner? Is a game better than a tutorial? Is a program which uses grammar rules to parse learners’ input (sometimes called “intelligent”) better than one that lists anticipated errors (sometimes considered “unintelligent”)? In other words, do these “different” CALL activities affect the quality of the language learning experience for the learners who use them? To begin to address this question, CALL researchers must investigate CALL activities with methods similar to those used by other second language classroom researchers, who investigate the communicative events in language classrooms by documenting and analyzing 33

34

CAROL A. CHAPELLE

the language produced by learners and teachers. This research has yielded descriptive data about the nature of interactions occurring in various classroom activities, an evolving metalanguage for describing relevant aspects of interaction, hypotheses about classroom interactions which may be beneficial for language acquisition, and an agenda for fruitful second language acquisition research. CALL researchers can work toward similar positive outcomes by documenting and analyzing the language used in CALL learning activities. To conduct such research, we must view the computer as a participant in a conversation [e.g. Chapelle (1990), Luff et al. (1990) and Hirst (1991)]. CALL activities, then, are classroom communicative events in which interaction occurs among participants including the computer, student(s), and instructor. Viewing CALL activities as classroom communicative events, this paper suggests an approach to research which investigates the similarities among CALL activities as well as the significance of observed differences for second language learning. First, I will illustrate how the concept of genre is useful for investigating similarities among texts in general. I will then provide examples of CALL texts to demonstrate how their functional elements can be analyzed and I will argue that genre analysis of such texts, whose purpose is language instruction, requires hypotheses about the aspects of texts believed beneficial for language acquisition. I therefore review some relevant results of classroom research and hypothesize their implications for CALL activities. On the basis of these hypotheses, I conclude that the observed differences in the example CALL texts are not significant. In conclusion, I clarify the relevance of three levels of analysis-text, context, and genre-for CALL activities.

INVESTIGATING

SIMILARITIES

AMONG TEXTS

Researchers in second language classrooms [e.g. Day (1986), Gass and Madden (19&7), and Chaudron (1988)] study the texts produced as learners and teachers communicate because the meanings relevant to students’ learning are expressed in these texts (Kress, 1989). An analytic question these researchers (and other text linguists) face is how to define types of texts which can be considered similar in important ways. One means of defining texts as similar is to refer to them as belonging to a genre. I will begin by defining and exemplifying text, and explaining how the similarities and differences among texts can be characterized through genre analysis. Texts Text is a term with multiple technical and popular meanings here.

so let’s define it as it will be used

A text is “language that is doing some job in some context.” It may be “spoken or written, or indeed in any other medium of expression that we like to think of.” Text is “made of “Text is a form of exchange; and the meanings.” It is both “product and process.” fundamental form of a text is that of dialogue, of interaction between speakers” (Halliday and Hasan, 1989: pp. 10-l 1). The following are examples of some dialogue texts. The first is from a doctor’s office, the second from a classroom, and the third from another classroom-a language classroom. Each text is accompanied by an analysis of the meanings it contains, in particular, the functional

CALL ACTIVITIES

35

moves (e.g. initiation)

which comprise each text are identified using the terminology of classroom discourse researchers (Sinclair and Coulthard, 1975). In the third move the meanings are described at a more delicate level by functional acts (e.g. evaluation and comment).

Text A: from a doctor’s office Participant Doctor: Patient: Doctor:

Language

Functional move

How long have you had those for? Well, I had’m a week last Wednesday. A week last Wednesday (Coulthard and Ashby, 1975: p. 80).

Initiation Response Follow-up [repetition]

Text B: from a classroom Teacher:

Pupil: Teacher:

Can you tell me why do you eat all that food? Yes. To keep strong. To keep you strong. Yes. To keep you strong (Sinclair and Co&hard, 1975: p. 21).

Initiation

Response Feedback [evaluation] [repetition]

Text C: from an ESL classroom Teacher: Student: Teacher:

What kind of time are we talking about here? Happening now. No, we don’t know if it is raining right now (Johnson, 1992: p. 507).

Initiation Response Feedback [evaluation] [comment]

We can summarize the functional elements of the three texts using capital letters for the moves and lower-case letters for the acts: Text A: I R FU-r Text B: I R F-e-r Text C: I R F-e<

Once we have identified some of the meanings represented in these texts by analyzing their functional elements, we can consider their similarities and differences. For example, the three texts are similar in two of their three functional moves: initiation and response. If we consider the meanings of the third move, texts B and C appear more similar to each other than they do to text A because their third move is feedback. Text C, however, is different from texts A and B because the final act is a comment rather than a repetition. Because these texts display degrees of similarity in various ways, deciding which should be considered similar and which should be different is not a clear-cut process. Some of the principles of genre study can help.

36

CAROL A. CHAPELLE

Genre There exist multiple perspectives on the definition and study of genre [see Swales (1990)] but, for our purposes, genre will be defined, consistent with Halliday and Hasan (1989) and Swales (1990), as a set of texts with the same communicative purpose. As the above examples illustrate, one view of communicative purposes comes from identifying the functional moves and acts which comprise a text. The real problem, as we saw above, lies in defining suflciently similar (i.e. in defining which meanings have to be the same in order to consider the texts to belong to a genre). We will approach the definition of sujfkiently similar through a two-part process: first by defining the texts at an abstract level and then by examining similarities in the abstract definitions in light of other features of context. Halliday and Hasan’s “generic structure potential” (GSP) allows us to define texts at an abstract level. GSP refers to the meanings “which help to structure the overall message form of texts” (Halliday and Hasan, 1989: p. 103). We infer the GSP on the basis of the texts we observe. In a real GSP analysis, we would want to obtain multiple samples from each setting (i.e. the doctor’s office, the classroom, and the ESL classroom), as illustrated in Fig. 1. The GSP associated with a given setting would define its typical texts at an abstract level.

Context ___-

Genre ---Generic structllre

potential

Text

Observed text data

r-l

GSPA

t

f

t

Text A Text A’ Text A”

Text B Text B’ Text B”

Text C Text C’ Text C”

Fig. 1. Text, genre, and context levels of linguistic analysis

Because the GSP is an abstract definition, it does not need to look exactly like any one text. Instead, it can include both obligatory and optional elements. Any individual text the GSP describes would realize the obligatory elements but would not be required to include the optional ones. For example, if we gathered additional data from the ESL classroom (illustrated by text C’ and text C” in Fig. l), we might find that some of the three-move exchanges did not end with a comment. We would indicate the comment is optional in the GSP using this notation: I R F-e-(c).? If we also found that sometimes the final move in the ESL class was a

CALL ACTIVITIES

31

follow-up, the GSP would be written like this: I R FUIF-e-(c). For our purposes here, however, we will use the GSP analyses which are each based on one sample text and therefore the GSP of each is equivalent to the actual text. Text A: GSP = I R FL-r Text B: GSP = I R F-e-r Text C: GSP = I R F-e-c

GSPs with significant similarities form a genre.’ In other words, a genre is also an abstract representation of the potential structure of meanings within texts. But a genre. as we will use the term here, represents GSPs which for some good reason can be argued to be similar. Let’s consider possible genre analyses for texts A-C before looking at one researcher’s good reason. A genre including both texts B and C (analysis l), would specify three obligatory moves: initiation, response and feedback. The feedback move would be realized by an obligatory “evaluation” act and either a “comment” or “repetition” act: Analysis I: genre for texts B and C = I R Fd-rlc

Alternatively, we might decide that the final move of the three move exchange could be optionally a follow-up or a feedback move, and therefore define texts A-C as belonging to one genre (shown in analysis 2).

Analysis 2: genre for texts A-C = I R FUIF.

For the doctor’s office and classroom texts, Berry (1987) argues that whether the third move is a follow-up or a feedback move is a significant difference4 (i.e. that analysis 2 is incorrect). The feature of the context Berry considers to make this decision is the role of the participant who contributes the third move. Feedback moves, Berry argues, are associated with a participant who knows the answers already (e.g. teachers); follow-up moves are associated with participants who are receiving information that is new to them (e.g. doctors). The fact that Berry can identify a feature in the context associated with the two different types of moves (FU and F), causes her to argue for the solution shown in analysis 3:

Analysis 3: genre X for text A = I R FU genre Y for texts B and C =

I R F.

This crucial criterion of “a feature in context which is associated with differences in texts” is not clear-cut. It will, however, be sufficient to allow us to analyze CALL texts because, for CALL texts, there exists one focal element of context: the texts’ purpose. Because their purpose is to increase the learner’s second language ability, the significant (i.e. genre-defining) elements of such texts should be those which are believed necessary for language development. Let us turn now to some CALL texts and label their functional parts. We will then consider what may be the important differences among the texts by identifying elements of interaction relevant to their purpose-second language learning.

38

CAROL A. CHAPELLE

CALL TEXTS CALL texts are produced in any language learning context where the computer takes an interactive role. Such contexts may be comprised of learners working individually with a computer, of learners working in pairs or larger groups with a computer or multiple connected computers, or of learners working with teachers or other experts. In each of these cases, the participants-one of which is the computer-contribute to an emerging text which is affected by the nature of the context and which both affects and provides evidence for the quality of the learning experiences. To exemplify how we can identify significant similarities and differences in such texts, let us begin by looking at some examples of texts produced in CALL activities. To keep the examples simple, we consider only texts documenting individual students working on computer programs designed for language learning: Text D: student working on ESL reading comprehension

questions

The student has just completed the first two paragraphs of an interactive reading passage materials. The following exchange occurs as the computer poses comprehension questions. Computer:

Student: Computer:

in

Initiation

What is a source of power which is caused by electron flow in material? It’s called + electron xzzt Almost Try again (Language

about conductivity

Learning Laboratory,

Response Feedback [evaluation-wrong] Follow-up University of Illinois).

Text E: student working on an ESL verb-tense exercise The student has just selected phrases which form a complete sentence in which the verb needs to be edited (during the prosperous 1960s. car buyers purchased large cars). The computer displays the verb phrase, placing the cursor on the same line, inviting the student to edit. Computer: Student:

Computer:

purchase large cars

[]

(moves cursor to end of “purchase”) d The verb is correct in that sentence (Chapelle, 1990: p. 215).

Text F: student working on an ESL conversation

Initiation [offer edit]

Response [edit] Feedback [evaluation-correct]

excercise (TERRI)

There is a picture of various sizes and shapes of blocks on the computer screen and the student knows from previous instructions that she can move the blocks around by typing directions. Computer: Student: Computer:

What should I do now? set the wedge behind the spool I’m thinking You have an unclear noun phrase (Coleman, 1985: p. 250)

Initiation [open solicit] Response Follow-up [identify error]

39

CALL ACTIVITIES

Text G: student working on a French conversation The student is supposed characters. Student: Computer:

Student:

to attempt to construct

exercise

a family tree by inquiring

Que est sa addresse? There is a problem in your last sentence Would you accept this utterance (yes/no)?: (Cecile lives with her brother) Cecile vit avec son frere. yes (Chanier et al., 1992: p. 127).

about relationships

and locations

of

Initiation [request information] Follow-up [identify error] Initiation

Response

Even in these few simple examples, we can see different functional moves appearing in these various CALL activities.5 On the basis of these data alone, we might analyze the GSPs of these texts as the following:b Text Text Text Text

D GSP E GSP F GSP G GSP

= I R F-ew FU = 1-e Red F-ec = I-OS R FU-ie = I-ri F-ie I R

At this level, it appears that each text is unique. However, we must now return to the question of whether their apparent uniqueness is relevant to language learning. For example, is the I R F FU of text D significantly different from the I R FU of text F? Since the purpose of the texts is to facilitate language learning, we would like our analysis at the genre level to reflect only those differences which are believed to be important for language learning. To perform such analyses we have to consider some of the aspects of interaction that are believed to affect language learning positively. SECOND LANGUAGE

LEARNING

RESEARCH:

IMPLICATIONS

FOR CALL GENRES

Fortunately, some second language researchers have investigated the essential questions about beneficial aspects of language learning texts. This research, however, has been conducted by examining the texts produced by human participants in various language learning settings. We will therefore be tentative in applying these findings to computer-student interaction. Nevertheless, as a starting point, let’s consider the implications of what researchers have discovered about positive aspects of the input to the learner, the interaction among participants, and the output the learner produces. Input to the learner The target language input learners receive is an important source of data for second language acquisition. As Krashen (1982) has theorized, input needs to be comprehensible but must also contain linguistic material which is new to the learner. In CALL texts, then, we should be concerned with the nature of the computer’s input to the student, as Underwood (1984) pointed out. Input to the learner in CALL materials can occur in any of the I R F or FU moves as well as in the materials used for instructions, reading passages, and listening exercises. Example exchanges 1 and 2 illustrate the possible locations of comprehensible input within two types of exchanges:

40

CAROL A. CHAPELLE

Example exchanges Participant Computer Student Computer

1 and 2 Moves Ex 2 Ex 1 I I R FIFU RIFU

Ideal quality Comprehensible

input?

Comprehensible

input?

Despite the fact that comprehensible input can occur in multiple types of moves in an exchange (and in CALL materials in general), not all input qualifies as comprehensible for all learners at a given point in time. The most important property of comprehensible input-that its basic meaning be understood (i.e. that it be “intake”) even though it contains linguistic material the learner does not know-is a feature that depends not only on the input but also on the language ability of the individual student. The best way for the researcher to assess whether input is comprehensible for an individual learner is to observe how the learner responds to it. In other words, it is necessary to observe sequences of interaction. Interaction among participants Long (1985) has hypothesized that the process of interaction with the target language input to make it comprehensible is a key factor in second language acquisition. The learner needs to receive the input, be confused about some aspects of its meaning and therefore request a repetition, clarification, restatement, or some other type of modification.’ “Modification of the interactional structure of conversation or of written discourse during reading . . . is a [good] candidate for a necessary (not sufficient) condition for acquisition. The role it plays in negotiation for meaning helps to make input comprehensible while still containing unknown linguistic elements, and, hence, potential intake for acquisition” (Larsen-Freeman and Long, 1991: p. 144). In CALL texts, then, we would like to see learners’ moves which request modifications of the input they receive. We would like to see, for example, learners’ initiations which request clarifications, restatements, definitions, and explanations-all of which are intended to help them understand the meaning of the input they receive. Example exchanges 3 and 4 illustrate the types of adjustments that we might look for in CALL texts: Example exchanges 3 and 4 Participant Computer Student Computer

Student Computer Student Computer

Moves’ Exchange 3 I MREQ MI Exchange 4 I R

M=Q MRSP

Ideal quality Comprehensible input Modification request Modified input

Comprehensible input Modification request Modified input

In text D, for example, let us say the computer’s initiation, “What is a source of power which is caused by electron flow in material?” was too difficult for the student because of the use of the passive. If the student had the option to receive a simplified version of the question (e.g. “What is a source of power which electron flow causes in material?“), and chose that option, we would have what is illustrated in exchange 3. It would provide evidence for comprehensible input for the particular learner requesting the modification.

CALL

ACTIVITIES

41

The learner’s output A third piece of interaction that is believed to be important for acquisition, particularly acquisition of grammatical competence, is what Swain calls comprehensible output. Because of the numbers and dynamics involved in classroom learning, “comprehensible output is, unfortunately, generally missing in typical classroom settings, language classrooms and immersion classrooms being no exceptions” (Swain, 1985: p. 252). Is comprehensible output present in CALL texts? Texts in which learners initiate or respond by producing language can illustrate production of comprehensible output. Again, the value of the comprehensible output is observed in a sequence of interaction, in this case consisting of the learner’s unsuccessful attempt to express something, the other participant’s communication of misunderstanding, and the learner’s correction. In some CALL texts, then, we might look for the sequences illustrated in exchanges 5 and 6. Example exchanges 5 and 6 Participant Computer Student Computer Student

Moves Exchange I R FIFU MRSP

Student Computer Student Computer

Exchange I FU MI R

Ideal quality 5 Input Output with error Error identification Comprehensible output 6 Output with error Error identification Comprehensible output Response

Exchanges l-5 illustrate GSPs representing student-computer interaction; we might look for similar desirable features of student-student and student-teacher interactions in CALL contexts. The point is that we need to make some hypotheses about significant aspects of interaction to perform an analysis which reveals significant similarities. Having identified some of the features of texts believed to be good for language acquisition, we can now consider whether the differences we identified in the example CALL texts are significant. We want to use the features believed to facilitate language acquisition as the obligatory elements in the genre(s) we define. In the example CALL texts, however, we see no requests for modifications (MREQ), and therefore no evidence for comprehensible input or modified input. With respect to comprehensible output, we can see the beginning of sequences that might provide evidence if the texts were longer. Each time the learners’ output (either an initiation or response) is evaluated as incorrect or an error is identified (texts D, F, and G), we have the beginning of an exchange in which comprehensible output is possible; however, the example texts, consisting of only three or four moves do not include the crucial move with the corrected output. As a consequence, the example texts, although superficially different, do not provide evidence of belonging to different genres on the basis of the criteria from SLA research.Y TOWARD

IMPROVING

CALL: TEXTS, GENRES, AND CONTEXTS

Analysis of computer-mediated communicative events is an important analytic linguistic endeavor in its own right. However, because CALL is intended to have particular consequences for language learners, the most important question ultimately to be addressed by

42

CAROL A. CHAPELLE

this type of research is how CALL activities can be designed and used to facilitate language learning. In addressing this question, we need to clarify three distinct but interrelated levels of analysis-text, genre, and context-and understand the relevance of each to effective CALL. Text The level of text, as illustrated above, refers to the actual linguistic data that result when participants interact. When our concern is second language acquisition, the study of actual texts produced in language learning contexts is essential because, as second language classroom researchers [e.g. Long (1980)] have pointed out, these texts document the instances of language use which are supposed to facilitate acquisition. CALL research has begun to investigate elements of the texts produced as learners work on CALL individually and in groups. For example, in a study of adult ESL learners working on listening comprehension CALL lessons, Jamieson and Chapelle (1987) used data such as the time elapsed between particular moves and some types of editing acts as evidence for strategies such as advanced planning and monitoring. In other types of activities, Chapelle and Mizuno (1989) and Doughty (1987) used learners’ requests for help as evidence for resourcing strategies and type of linguistic processing, repectively. Hsu et al. (1993) used some types of editing acts and the numbers of exchanges as indicators of an exploration strategy. Bland et al. (1990) used learners’ requests for dictionary definitions as indicators of level of vocabulary knowledge. Investigating learners working in groups at the computer, Piper (1986) and Abraham and Liou (1991) described the numbers of the various acts produced orally by students using different types of programs. As a result of this work, it is apparent that technical capabilities for data collection in CALL environments far exceed our current theoretical and analytical capabilities for their description and interpretation. Future descriptive research should benefit from analytic methods documenting the textual context in which particular instances of interaction occur and from adopting discourse terms and concepts used by classroom researchers. The context in which particular acts occur is important because it gives meaning to these instances of interaction. There is need to further investigate the use of classroom discourse analysis principles to document the textual contexts of moves made in CALL activities [e.g. Chapelle (1990)] as well as to explore the capabilities and limitations of discourse analysis methods for CALL.“’ Empirical, descriptive work is important as a first step in CALL research. To address the real issue, however, we must eventually ask whether we see evidence for significant elements within texts-those believed to facilitate language learning. To provide an analytic means for distilling the significant similarities among texts, we have sharpened Halliday and Hasan’s definition of genre to refer to texts containing similar important elements. Genre

Genre is an important level of analysis because it summarizes CALL texts to allow for meaningful generalizations based on actual data. Moreover, genre analysis underscores the need for a common terminology for describing CALL texts. Generalizations about CALL activies expressed using a lingua franca can offer guidance to future CALL designers, users, and researchers who wish to produce, provide, and investigate what we believe may be effective CALL. By investigating the effects of different CALL genres, we can begin to test in CALL contexts the hypotheses of classroom researchers about the positive effects of particular types of interaction. Here we have considered the only relevant (i.e. genre-defining) aspect of

43

CALL ACTIVITIES

context to be the texts’ purpose. As we pursue consider other features of CALL contexts as well.

such research,

however,

we may wish to

Context The context refers to features of the activities, topics, participants, and language that comprise the text and in which the text is embedded. Features of context should be used in the genre analysis as motivation for choosing obligatory elements. Context, however, plays a more obvious and practical role in CALL design, use, and research because multiple features of context-not just the computer program-have a marked impact on learning (Papert, 1987). Some features of context are beyond the control of the instructor. They typically include the learners, their purposes, and the setting for instruction. Other features, however, are created by the instructor. These include the types of CALL materials and the parameters for using the materials. Both types of contextual features affect the text produced in CALL environmentsthe texts which are crucial to language learning. Having distinguished three levels of analysis, we can see that the questions that are typically posed about similarities among CALL activities refer to one aspect of the context level: is a CALL activity designed to allow maximum learner-control inherently better for language development than one that guides the learner? Is a game better than a tutorial? Is a program which uses grammar rules to parse learner’s input better than one that lists anticipated errors? The problem with these questions is that at the context level there are other features that create the potential for particular texts to be created (Halliday, 1977; Halliday and Hasan, 1989). Moreover, we can-and should-make comparisons at two other relevant levels of analysis: text and genre. Labels such as “intelligent” serve as a short-hand device to refer to potentials of software but it is the actual texts that must be examined to make judgements about the quality of a language learning experience. In other words, to begin to address the comparative questions about CALL, it is necessary to specify the other relevant features of CALL contexts, as well as to document and analyze the texts resulting from those contexts so elements of the texts believed to be significant for language learning can be identified.

NOTES ’ “Technocentric refers to the tendency to give centrality to a technical object-for example computers tendency shows up in questions like ‘What is THE effect of THE computer on cognitive development?“’ 1987: p. 23). 2 See Halliday and Hasan (1989: pp. 59968) for a more complete notation system for expressing

i The criterion for identifying genres presented here-that

they be associated more focused than those discussed by Halliday and Hasan (1989).

with significant

.This (Papert,

elements of GSP. aspects of context-is

J Berry’s (1987) paper does not argue in so many words that participant roles are genre-defining, but it uses participant roles to explain differences in the texts. I assume that she is attempting to explain differences which she sees as significant. ’ The names given to some of these moves and acts are tentative. b If we had more data from each of the CALL activities, we would undoubtedly have somewhat different GSPs for each. For example, the multiple exchanges from which text E was taken also contained some editing acts in which errors were made and identified. The GSP might have therefore been represented as follows: I R F-eclew.

’ For other indicators of conversational adjustments see Larsen-Freeman and Long (1991: p. 126). * The abbreviations in exchanges 3-6 have the following meanings: MREQ = modification request; MI = modified initiation; MRSP = modified response.

CAROL A. CHAPELLE

44

’ Of course this is not to say that significant

differences

do not exist among any CALL activities.

“I Such research is needed to help identify, for example. useful units of analysis in CALL texts such as those discussed by Crookes (I 990).

REFERENCES ABRAHAM. spoken

R. and LIOU. H.-C. (1991) InteractIon

language.

Practice.

In Dunkel,

pp. 85-109.

generated

P. (ed.), Computer-assisted

by three computer

Language

Learning

programs:

analysis of functions

and Testing: Rrseurch

of

1ssue.s and

New York: Newbury House.

BERRY, M. (1987) Is teacher an unanalyzed concept’? In Hailiday, M. A. K. and Fawcett, R. P. (eds), New developments in Systemic Linguistics. Vol. 1: Theory and Descriptior2. pp. 41-63. London: Pinter Publishers. BLAND, S. K., NOBLITT, J. S., ARMINGTON, S. and GAY, G. ( 1990) The naive lexical hypothesis: computer-assisted language learning. Modern Language Journal. 74,440-450.

evidence from

CHANIER, T., PENGELLY, M. TWIDALE, M. and SELF, J. ( 1992) Conceptual modelling in error analysis in computer-assisted language learning systems. In Swartz, M. L. and Yazdani. M. (eds), fnrelllgewtTutoring Systems for Foreign Language Learning, pp. 125-150. Berlin: Springer-Verlag. CHAPELLE, C. (1990) The discourse of computer-assisted research. TESOL Quarterly 24, 199-225. CHAPELLE, 25-47.

C. and MIZUNO,

S. (1989) Students’

strategies

CHAUDRON, C. (1988) Second Language Classrooms: University Press. CLARKE, R. E. (1985) Research, l(2), 137-148. COULTHARD, COLEMAN,

Confoundmg

language

learning:

toward

with learner-controlled

a context

CALL. CALICO Journal, 7(2),

Research on Teaching and Learning. Cambridge:

in educational

computing

research.

Journal

of Educational

M. and ASHBY, M. (1975) Talking with the doctor. Journal qfCornmunication, D. W. (1985) TERRI: a CALL lesson simulating

CROOKES, G. (1990) The utterance, Linguisrics, 11(2), 183-199.

and other basic

DAY, R. (ed.) (1986) Talking to Learn: Conversation House.

for descriptive

conversational

interaction.

units for second

language

in Second Language

Cambridge Computing

25(3), 140-147.

.Sysrem, 13, 247-252. discourse

Acquisition.

analysis.

Rowley,

Applied

MA: Newbury

DOUGHTY, C. (1987) Relating second-language acquisition theory to CALL research and application. In Smith, W. F. (ed.), Modern Media in Foreign Language Education: Theor? and Itn~)lementation, pp. 133-167. Lincolnwood, IL: National Textbook. FALBEL, Norwood,

A. (1991) The computer NJ: Ablex Publishing.

GASS, S. M. and MADDEN, Newbury House Publishers. HALLIDAY,

as a convivial

tool. In Harell, 1. and Papert, S. (eds) Constructionurn

C. G. (eds) (1987) Input in Second Lnngucrge Acquisition,

M. A. K. (1977) Explorations

analysis

Context.

JAMIESON, J. and CHAPELLE, C. (1987) Working strategies. Langunge Lemming, 37, 523-544.

KRASHEN, KRESS, G.

and Text: Aspects

have a role in computational

HSU, J. CHAPELLE, C. and THOMPSON, A. (1993) Exploratory explore? Journal of Educational Computing Research (in press).

JOHNSON, K. (1992) Learning Quarterly, 26(3), 507-535.

Rowley, MA:

in the Functions of Language. New York: Elsevier North Holland.

HALLIDAY, M. A. K. and HASAN, R. (1989) Language. Social-Semiotic Perspective. Oxford: Oxford University Press. HIRST, G. (1991) Does conversation 17(2), 212-227.

pp. 235-253.

pp 29-37.

to teach: instructional

actions

linguistics’? Computational

environments:

styles on computers

of Language

Linguistics,

what are they and do students

as evidence

and decisions

in a

of second language

of preservice

ESL teachers.

learning TESOL

S. (1982) Principles andpractice in Second Language Acqursition. Oxford: Pergamon Press. (1989)Lillguistic Processes in Socioculrural Practice. Oxford: Oxford Umversity Press.

LANGUAGE LEARNING LABORATORY (nd) ConductIon of Electricity in Matenals: an advanced-intermediate ESL Lesson for Students ofScience and Technology. Urbana, IL: University of lllinois at Urbana-Champaign.

45

CALL ACTIVITIES

LARSEN-FREEMAN, London: Longman. LONG,

D. and LONG,

M. (1991).

An Introduction

M. H. (1980). Inside the “black box”:methodological

Language

Learning,

to Second

Language

issues in classroom

Acquisition

research

Research.

on language

learning.

30, 142.

LONG, M. H. (1985) Input and second language acquisition theory. In Gass, S. M. and Madden, C. G. (eds), &wt Acquisition, pp. 377-393. Rowley, MA: Newbury House Publishers.

in

Second Language

LUFF, P.. GIL.BERT, N. and FROHLICH,

D. (eds) (1990) Computers and Conversation.

PAPERT, S. (1987) Computer criticism vs. technocentric

thinking. Educational

PIPER, A. (1986) Conversation and the computer: a study of the conversational English as a second language working in groups. System, 14, 187-198. SINCLAIR,

J. McH. and COULTHARD, R. M. (1975) Towards London: Oxford University Press.

an Analysis

London: Academic

Researcher.

Press.

January-February

spin-off generated

22-27.

among learners of

ofDiscourse:the

English

used b)

Teacher and Pupls.

SWAIN, M. (1985) Communicative competence: some roles of comprehensible input and comprehensible output in its development. In Gass, S. M. and Madden, C. G. (eds) Input in Second Language Acquisition. pp. 235-253. Rowley. MA: Newbury House Publishers. SWALES, Press.

J. I 1990) Genre An&sis:

UNDERWOOD,

_I.(1984) Linguistics,

English

in Academic

Computers,

and Research Settings. Cambridge:

and the Language

Teacher. Rowley,

Cambridge

MA: Newbury

University

House.