In many of the studies on foreign/second language (L2) listening comprehension, the listener-subject seems to be treated like a “robot” (Brown 1986, p. 287). Given a task, the robot is first supposed to intake an oral text, interpret the meaning of the text, determine what the specifications of the task are, and then complete the task. Rost and Ross (1991), for example, had their subjects listen to a story in one-on-one dyads. While listening they were instructed to respond with questions about the story at “prespecified junctures” (p. 245), and at the end of the narration, they were asked to write a short summary of the story. Few listeners in actual listening situations would be required (or constrained) to act in such a robot-like way.
In addition to being treated like a robot, the listener is sometimes regarded as a passive receiver of information. For example, Derwing (1989) claimed to have found that speaker adjustments in the output of information types had a significant effect on listener comprehension. The success of a speaker was determined by the scores of listener recall protocols. Derwing writes, “the listeners’ replies to the comprehension questions were scored to determine whether their partners had successfully imparted [italics added] eight bits of information...” (p. 162). This seems to imply that the speakers had to, independently, do most of the important work of getting their listeners to understand, and that all the listeners had to do was try to catch bits of information given by the speakers and to cast them back at the speakers.
Rejecting the view of listening as a passive or receptive activity, Anderson and Lynch (1988) argue that the listener is not just a “tape recorder” (p. 9); listeners are “selective, in terms of what they find most interesting or important or comprehensible in any particular message” (p. 11). Just as speakers have intentions and purposes in communication, listeners bring their desires and goals into any conversation. For instance, listeners can choose to listen or not listen, to listen carefully or not so carefully. Brown (1986) coined the term “scale of co-operativeness” to measure the listener’s participation in a conversation “with maximal co-operativeness at one end and maximal unco-operativeness at the other, and many possible degrees of co-operativeness in between” (p. 288). Listeners can choose to listen minimally and produce minimal replies. According to Anderson and Lynch (1988), the listener may understand that he or she has been addressed but cannot understand the speaker’s message because of, for example, background noise or the speaker’s accent. Or the L2 listener may hear a speaker’s speech adequately but fail to understand the message because of “serious problems with syntax or semantics of the foreign language” (Anderson & Lynch, 1988, p. 6). And listeners sometimes “‘switch off’ consciously or unconsciously” (Anderson & Lynch, 1988, p. 6), and they may later “‘come to’ with a guilty start as they realize that ‘their minds have wandered’” (Brown, 1986, p. 288).
We cannot always discern whether a listener is switched on or has momentarily switched off. Listening is an unobservable process; “overt signs of listening attention or inattention can be misleading” (Witkin, 1993, p. 49). It is very difficult to get inside the listener’s head to investigate the comprehension process. In order to study the process, first language (L1) and L2 listening researchers have depended on listeners’ output from tasks set by the researchers to measure their understanding of the oral text (see Brown, 1986; Witkin, 1993). Teachers also use task outputs to evaluate their students’ listening comprehension. Tasks in the classroom, for example, have included the following: listening for perception with outputs calling for repetition or identification of linguistic items, or listening for understanding with outputs consisting of answering questions about a passage or writing short summaries of a text (Ur, 1984), to name but a few.
The problem with using task outputs to measure how well the student or subject has understood a spoken text is related to the fact that that the task itself does not necessarily represent the student or subject’s understanding of the text itself (Brown, 1986). There are three separate elements involved, and difficulty may arise for the listener trying to understand any or all three elements:
- the specification of the task
- the task
- the text (Brown, 1986, p. 285).
Brown argues that a good understanding of what the task requires of the listener is needed as well as assurance that the listener understands the requirements of the task before it can be concluded that the listener has not comprehended a text.
Support for Brown’s argument comes from Dunkel, Henning and Chaudron (1993) who state that an understanding of the interplay of the two broad components (the existence of which appears self-evident) of the listening comprehension construct, person/competence/ability and text/task/difficulty, is direly needed. Dunkel and her colleagues note that, “particular texts are appropriate for particular (though differing persons), particular tasks are representative of particular (though diverse) competencies, and particular difficulty levels are identified with reference to particular (though various) ability levels” (p. 182).
Researchers, such as Hoven (1991), are seeking to determine ways to identify how “particular texts” and “particular tasks” may represent “particular difficulty levels.” The identification of these variable features is vital to the development of an effective teaching approach to listening comprehension for much of what is now called listening comprehension teaching is closer to comprehension testing (see Anderson & Lynch, 1988; Brown, 1986, 1990; Brown & Yule, 1983b). Anderson and Lynch (1988) call for graded listening comprehension tasks that assist students in understanding rather than assess how much they understand. Brown (1986) agrees that procedures are needed to help the teacher teach instead of test:
- We need to move to a position where the teacher is able to recognize particular patterns of behaviour in listening manifested by an unsuccessful listener and to provide exercises for the student which will promote superior patterns of behaviour (superior strategies). (p. 286)
However, very little yet is known about what listening comprehension actually entails. Dunkel (1991) and Rubin (1994) state that the body of L2 listening comprehension research is small and still in its infancy. Even in L1 listening research, Witkin (1990) notes that there is no “generally agreed upon” definition of listening and “no one theory about what listening is” (p. 7). Nevertheless, the body of listening research is growing. Rubin (1994) enumerates five major aspects of L2 listening that researchers are examining: (a) text characteristics, (b) interlocutor characteristics, (c) task characteristics, (d) listener characteristics, and (e) process characteristics. Rubin refers to process characteristics as “variation in the listener’s cognitive activities and in the nature of the interaction between speaker and listener” (p. 199). In studying this fifth factor, researchers should look at the listener as an active player in the negotiation of meaning. Pica states that, “negotiation unites linguistic processes with cognitive processes,” that “repetitions, rephrasings, and other linguistic elaborations are crucial to the comprehension process,” and that these elaborations “are effective only when made at the request of the listener” [italics added] (cited in Rubin, 1994, p. 215).
L2 listening research may be just beginning, but two of the critical issues which are relevant to the study of the listening process involve (a) the role of the listener, and (b) the interaction of the listener, text, and context. This exploratory study is an attempt to examine the role of the listener in interaction, by observing an actual interaction between a speaker and a listener, and specifically, by trying to examine the critical and active role the listener plays in this interaction.
The technique used in this study is a partial adaptation of the map task used in Brown’s studies (see 1986, 1995). Her research incorporated discourse analysis of the interactions between listeners and speakers. However, in this study the researcher examines (for patterns of listening problems and strategies): (a) videotapes and audiotapes of the subjects in interaction, and (b) the subjects’ retrospective verbal reports about what they were thinking about and the problems they encountered during the interaction. Verbal report methodology is being more prevalently used in second language research, and there is support for (and objection to) the use of verbal reports as data (Cohen, 1987). Smagorinsky (1994), for example, claims that, “when researchers are attentive to the potential problems caused by the procedure and take steps to control and account for them, think-aloud protocol collection and analysis can be a remarkably illuminating research methodology” (p. 16). In L2 listening research, Buck (1991) has used verbal report methodology to examine subjects’ listening processes and test taking, and to investigate “how processes not normally accessible through quantitative research methods influence test performance” (p. 67).
The assumptions underlying the map task, and the merits and demerits, according to Brown, of the map task structure and method are discussed below.
Assumptions underlying the task. The traditional view of verbal communication held that there exists the possibility of an ideal system of communication, “a potentially perfect encoding-decoding process” (Rost, 1990, p. 3), in which the meaning of the message intended and spoken by a speaker is transmitted to and understood by a listener in exactly the same way (Brown, 1995, p. 7). This view, however, is not a useful model of how humans understand language according to Brown (1990). It needs re-examination because it does not account for the ambiguousness of language and misrepresents the role of the listener (Rost, 1990). Scholars in fields such as phonetics, theoretical linguistics, sociolinguistics, social psychology, and discourse analysis have proposed that “hearers create the meanings of utterances” (Witkins, 1990, p. 22; see also Rost, 1990). Linguists, philosophers of language, and social anthropologists “have become sensitive to the possibility that a listener’s understanding of an utterance may not yield a thought that is identical to that intended by the speaker” (Brown, 1995, p. 17).
Partial understanding of a speaker’s words can be sufficient for satisfactory communication. This may be the case for young children who are learning the language and who “carry partial information about the relationship between quite familiar words and the objects that these words may appropriately be used to describe” (Brown, 1995, p. 9). Vygotsky wrote in his book, Thought and Language (1934/1986), “‘flower’ appears in a child’s vocabulary much earlier than the names of concrete flowers....If...the child learns the word ‘rose’ prior to that of ‘flower,’ he uses ‘rose’ as a general name, calling all flowers he sees ‘rose’” (p. 143).
Adults also carry around “partial” information about words. As Quine (cited in Brown, 1995, p. 9) explains, (a) the boundaries of a term are difficult to determine (e.g., Where does a mountain begin and end?), and (b) it is difficult to decide what might be included or excluded in a term (e.g., When might something called a hill also be referred to as a mountain?). There is also the possibility that an utterance may be adequately understood in one context but inadequately understood in another. Brown (1995) demonstrates this possibility with the example of an adult who may understand utterances containing words such as, beech, elm, sycamore, or aspen, but who may not be able to pick out a photograph of a sycamore from ten photographs of different kinds trees. According to Brown, “we may claim to know the meaning of a word but it does not follow that we always deploy it in a consistent and reliable manner” (p. 10).
Brown (1995) highlights the issue of the variability of interpretation in different contexts. It is possible that “the same expression may be construed as having a different meaning in different situations of use” (p. 11). There is a problem for the listener in constructing an interpretation of an utterance which is taken out of context. The listener can only work with a “thin” semantic meaning of the sentence which is taken from the vocabulary items set in a syntactic order. When the listener has access to the pragmatic context of the utterance, a “thick” meaning can be inferred.
It is also possible that different individuals arrive “at different interpretations of the same expression in the same context of use” (p. 11). Here, Brown (1995) is concerned with the often made distinction between “speaker meaning” and “sentence meaning” (p. 14). Brown claims that there is no distinction between the two, that is, “utterance types do not have a single correct interpretation which will hold in all contexts” (p. 15). Speakers often mean far more than what is explicitly expressed in their propositions. Clark (1992), on the other hand, gives primacy to the speaker’s meaning saying that it is very different from word and sentence meaning. Speaker’s meaning is “a type of intention that speakers have toward meaning” while “word and sentence meaning...are statements about conventions that hold within communities of people that enable the words and sentences to be used as a way of meaning things” (p. xv). Brown argues that there may not be proper “prototypical contexts to develop the interpretation of all sentences in isolation” and that this “under determination” or vagueness of word-meaning contributes to a necessary flexibility in language which enables the communication of new thoughts (p. 16). Communication carries a high risk of misunderstanding and requires an active and involved listener as well as speaker.
Brown (1995) proposes that the idea of “correct interpretation” should be modified to one of “adequate interpretation” (p. 22). When Brown (1995) opts for the notion of adequate understanding, she is not claiming that utterances are never correctly understood. Formulaic expressions, (e.g., “Can you tell me the time?”—“Five past four.”) appear to be “sufficiently understood for society to function with a tolerable level of efficiency” (Brown, 1995, p. 30). It is true that in everyday situations, a correct interpretation may be assigned to utterances used in such familiar contexts, but this does not mean that it holds true for all cases of communication, especially for cases of extended discourse or text. Listener understanding may be judged adequate, if not correct, and the judgement must take into account the question, “adequate for what?” (p. 23).
In his discussion of listener construction of meaning Rost (1990) adopts a similar standpoint to that of Brown suggesting four terms relating to how listeners create meaning: (a) acceptable understanding which “refers to inferences drawn by a listener that are satisfactory to both speaker and listener”; (b) targeted understanding which “refers to a specific interpretation that was intended by the speaker” (c) non-understanding which “refers to the listener being unable to draw any appropriate inference based on what a speaker has just said”; and (d) misunderstanding which “refers to a conflict between the type of inferences that the speaker had expected the hearer to draw from the speaker’s utterances and those inferences that the hearer actually has drawn” (p. 62).
Brown (1995) concludes that communication is a risky affair. Communication, when and if successful, does not end in a static interpretation of the speaker’s meaning in the listener’s mind, rather, this understanding undergoes adaptation and modification into the listener’s framework of beliefs and information which in turn and, in part, is what the construction of interpretation is contingent upon.
In the next section, a description of the map task is given and the advantages and limitations of the task are discussed.
The map task. In the task a speaker describes a route, from starting point to goal on a map of an island, to a listener. The task involves the identification and spatial relationships of the island features. The speaker and the listener have recourse to related (but different) information which needs to be shared in order to accomplish the task. The task data on which Brown (1986) bases her work1 were collected from young native speaker subjects in Scotland and England. The external features of the map-task context were:
- Pairs performed the task; the listener had to listen to only one other person, the speaker.
- The subjects, who were classroom students, chose their partners, fellow classmates who were similar in speech, background, and point of view.
- The speaker addressed the listener directly shaping the input for the listener.
- No time limit was placed on the task.
- The students did the task in their classroom, a familiar place for them.
- The subjects knew they were being recorded, which may have had some effect on the task.
- The students worked at desks across from each other with a low screen between them.
- An information gap existed between the speaker and the listener whereby they had to share information to complete the task.
Brown (1995) says the merit of such a task in which the external context of utterance and information content is constrained, “is that the analyst can identify what is being spoken of, knowing what information was potentially available to the participants, and can compare treatments of the same point of information by different individuals and different groups” (p. 2). In addition, how much listeners have understood of what the speaker says is, “revealed in what the listeners draw on their maps, by what questions they ask, and by the comments that they make as the task proceeds” (p. 2).
Brown (1995) cautiously warns, however, that every analyst, or observer, uses a procedure of interpretation necessarily based on the observer’s set of beliefs and concepts. Stich, (cited in Brown, 1995), objects to such analyses made by researchers as they “represent the psychology of ‘me and my friends’ (p. 34) and Davidson (cited in Brown, 1995) warns that one can only understand language that expresses conceptual schemes similar to one’s own. However, analysis and interpretation of the map task interaction data may be less problematic since the information content and context of the task are controlled by the analyst.
The map task interaction is constrained (but not as stringently controlled as tasks in most laboratory experiments) and will in many ways be different from interaction that occurs naturally. The task data will therefore be limited. The map task is, however, closer to “ecological” methods of research, “which study the behaviour to be investigated in environments as near to normal as possible” (Brown, 1995, p. 36). It is also important to know “whether there are aspects of the communicative setting which give rise to problems of interpretation other than those which are manipulated in typical psycholinguistic experiments” (Brown, 1995, p. 36). Brown cautions that generalizations of findings should be made very carefully and that:
A careful, qualitative, analysis of the behaviour of listeners, even in banal tasks, should contribute to a better understanding of how listeners actually go about the process of constructing an interpretation, of how they sometimes fail, and why they fail. (p. 5)
This study attempts to further the understanding of how listeners construct meaning. It extends the work done by Brown and other researchers in L1 listening to L2 listening, specifically to the listening role the English as a foreign language (EFL) learner/listener plays in collaborative discourse.
The participants, all female native speakers of Japanese, in four EFL intermediate-level classes at a private language school took part in this study. Subjects who volunteered to be listeners and speakers ranged in age from 34 to 53, and the rest of the class members, who took the roles of overhearers2, were between 31 and 71 years old. There were no overhearers in one of the classes (Class 2) because only two members were present on the day of the task. The subjects had been studying EFL at the school for one to nine years (approximately 38 to 57 hours per year). The majority had had six years of compulsory English language instruction in secondary school and from one to four years at the post-secondary level.
For the task, two maps of an island were used: a speaker map and a listener map (see Appendix A and B). The speaker map had labeled features including mountains, a castle, a village near the larger of two rice fields, a swamp, and a marked-out route from the starting point, X , to the goal, a mine. The listener map had no route marked out and less information—no castle, and no village located near the larger rice field. In addition, two features, the swamp and surf island, were not labeled.
The study was conducted in the subjects’ own classrooms and during regular class time. The desks and chairs were set up so that the speaker and listeners were sitting opposite one another. A cardboard screen was put between them so that they could not see the other’s map. The overhearers sat at the sides of the listener and speaker, far enough away so that neither they nor the listener or speaker could see each others’ maps. The researcher, who was the teacher, told the class the names of the various features of the island and inquired whether the students were familiar with the vocabulary. If any feature proved to be unfamiliar, it was explained. The speaker was told to tell the listener how to draw the route across the island. The listener was told to draw the speaker’s route on her map. The subjects were told that the speaker map and the listener map were not the same, that is, the speaker had a map with more and later information on it.
The class was told that only the listener and speaker could talk. The overhearers were to remain silent throughout the task; their job was to listen to the speaker and listener and silently draw the route on their maps (which were the same as the listener map). The interaction between the speaker and listener was videotaped and audiotaped.
After the task was completed the video camera was stopped. The students looked at each others’ maps and compared their routes to the speaker’s. The maps were then collected. The teacher told the students that she would like to probe what they, and in particular the listeners, were thinking about during the task, and to inquire about any problems the listeners had doing the task. In three of the classes the videotape was played for the students to observe and comment on about what they were thinking and any problems they encountered during the task. Since class time was used for this study, the teacher felt she could not offer the students the choice of using either Japanese or English in their verbal reports. Most students remained in the L2. When the retrospective report time exceeded the class hour, the teacher told the students that they may use Japanese if they wished, and sometimes asked some students to repeat in Japanese what they had said in English. The students’ verbal retrospective reports were audiotaped.
The listener maps were analyzed for correct or incorrect completion of the task. The videotaped and audiotaped interactions between the listeners and the speakers were analyzed for patterns of listener behavior that showed acceptable understanding, targeted understanding, non-understanding, or misunderstanding (Rost, 1990). The audiotaped retrospective verbal reports were analyzed for evidence which supported or did not support the patterns of listener behavior found in the interaction data.
The Map Outcomes
Table 1 shows the performance outcomes of the map tasks by the four pairs of listeners and overhearers. Roko in Class 2 was the only listener able to draw the route correctly on the map. There were no overhearers in Class 2. Nobu in Class 1 and Eiko in Class 3 could not complete the task correctly although three out of the five overhearers in Class 1 and the one overhearer in Class 3 could and did. Tomi and the two overhearers in Class 4 were not able to draw the route accurately. Table 1 also shows the amount of time each pair took to complete the task.
Table 1. Outcome of Map Tasks
Note. min. = minutes.
Class 1. Yoko and Nobu worked for almost half an hour, the longest amount of time taken by any the dyads. Nobu’s map shows that she had trouble at the lake area, going around the east side of lake instead of the correct way, the west side.
Class 2. Miko and Roko took the least amount of time, 14 minutes, to finish and, according to her map, Roko was the only listener to correctly complete the task.
Class 3. Kiko and Eiko took 22 minutes to finish the task. Eiko appears to have had trouble at the beginning part of the route, hugging the coastline instead of taking a route between the palm trees and the forest. And instead of drawing a route which passes in front of the castle, she drew the route behind the castle. Her map also shows that she had a problem at the end of the route. She went beyond the stopping point, the mine, to the northernmost end of the eastern mountain range where she drew in three extra mountains.
Class 4. Sami and Tomi took only 16 minutes to complete the task, but Tomi’s map shows that she had difficulty drawing the beginning part of the route. She drew a lightly marked line that followed the west coastline very closely and another darker line that approximates the correct route. Her map also shows that she had a problem in the swamp area; her route is drawn below the swamp instead of between the swamp and palm tree beach.
Analysis of the map outcomes show that each pair performed the task in different amounts of time. Each listener reached different results from the interaction with her interlocutor and had different problem areas:
- Class 1 Nobu lake
- Class 2 Roko no problem areas
- Class 3 Eiko west coast, castle, goal
- Class 4 Tomi west coast, swamp
In the next section, one problem for each dyad has been selected for description using data from the videotapes and audiotapes of both the interaction during the task and the retrospective interview following the task. The problem was selected on the basis of its cruciality or typicality. Strategies that each listener used, strategies that appeared both helpful and unhelpful, are described. Pedagogical implications which the researcher believes may be extracted from the analysis of the listeners’ problems and strategies are suggested in the Discussion section. Discussion of the role of the overhearers is beyond the scope of this paper.
The speaker and listener’s utterances are noted in italics. Utterances in Japanese were translated by the researcher into idiomatic English and are noted in parantheses.
Listener and Speaker Interactions and Verbal Reports
Class 1. The audiotapes and videotapes of the interaction between Yoko and Nobu revealed that Nobu had trouble in more than the one area (the lake) that was indicated by the route she drew on her map. As might be expected, Nobu had difficulty in the areas near the castle and the larger village; the castle and larger village were points of information not on Nobu’s map. However, she and her partner eventually worked out the correct direction of the route and correct location of the castle and the village.
The problem which Yoko and Nobu spent the longest time on was centered on the route around the lake area. Although they backtracked twice to the route near the lake during the task, this part of the route remained a problem for Yoko and Nobu. The tapes showed that part of the problem was that Nobu’s concept of going towards the right after reaching the lake was not the same as Yoko’s idea of turning right at this point. Nobu was thinking of the right side of the map as indicated by the fact that she waved her right hand toward the right as she asked her partner, we turn right? Yoko’s concept of turning right was from the viewpoint of walking along the route as indicated by the fact that she said, first turn right then turn left, and to turn toward the west side of the lake.
The tapes showed that another part of the problem was the use of the same term (forest), by Nobu and Yoko, to refer to forests in different locations. When Nobu asked Yoko, you didn’t through between lake and forest? , Yoko answered, we go through between lake and forest, yes, yes. Yoko assumed Nobu was referring to the forest to the west of the lake but Nobu’s map indicates that Nobu was referring to the forest north of the lake.
During the retrospective interview, which lasted 44 minutes and during which the participants were able to view the videotape, Nobu, the listener, remarked that she suspected she had not drawn the route around the lake correctly. When she saw the speaker’s map after the task and the correct route she said, yappari! (I thought I hadn’t gotten it correct!). Both Nobu and Yoko said they were getting upset and anxious. When asked if she had heard Yoko say the west side of the lake, Nobu said she had not. She said, omoikonde ita (my mind was set), she was convinced that turning right meant turning toward the right side of the map, and added that, I should more question, (I thought I should have asked more questions).
Nobu’s strategies. Nobu used silence and short responses such as mmm, yes, uh, and repetition of a key word such as, castle?, not city?, village?, with rising intonation. She also asked for confirmation of the speaker’s directions saying, for example, you are going down between forests and swamp? and so, you didn’t go little village? Nobu interrupted Yoko and their talk overlapped quite often, but Nobu tended not to continue her lines, for example, ending before she completed a question, did you through....
Nobu gestured with her hands much more than the other listeners; most of the gesturing seemed to help her generate needed vocabulary items. Since Yoko and Nobu looked down at the map during most of the interaction they were not able to benefit from the hand signals Nobu used, particularly the motions she made to indicate turning right at the lake.
Findings from Class 1. Several factors appeared to have contributed to the difficulty the listener encountered: (a) insufficient eye contact between listener and speaker; (b) the listener and the speaker’s affective (anxious, nervous) states; (c) the listener’s incorrect fixation on one concept (i.e., turning right); and (d) the listener and the speaker’s misidentification of referring expressions (e.g., the forest).
Class 2. Roko’s map shows that she was able to draw the correct route on the island. However, the analysis of the tapes shows that Miko and Roko did have some difficulty at the end of the route. Miko used the L1 term, sanbun no ichi (one-third), to help clarify where Roko should terminate her route. Miko said, sanbun no ichi, you can stop one third of the mine.
During the retrospective interview which took 44 minutes and included a videotape replay, Miko and Roko said that they had thought that the whole mountain range on the right side of the map was the mine. Miko said, all this part is the mine, (pointing to the mountain range), and then asked Roko, kore zentai wa mine datto omotte ta? (did you think that this whole thing was the mine?). Roko replied that she did. Miko and Roko’s concept of the word, mine, included the whole mountain range, not an excavation site located somewhere in the range. Although they misconceived the term, they had reached a mutually acceptable understanding of the term and were able to finish the task.
When asked why she used her L1 at the mine area, Miko said that towards the end of the task she had become tired, and said, I felt lack of my ability.
Roko’s strategies. Like the first listener, Nobu, Roko utilized silence, short responses, and repetition of key words with rising intonation (e.g., village?, castle?). However, Roko included more specific information in her responses to the speaker such as, in my map there is no little village and maybe castle isn’t in my map so I write down the castle. During the retrospective interview, Roko indicated that she often made guesses, I’m sorry I was guessing the route but I couldn’t guess the route going anywhere around here. When her guesses could not help her, she asked for specific information saying, first I understand where’s the little village then I asked the route.
It is important to note that the speaker, Miko, often inquired about Roko’s understanding of the route with short questions such as, can you see the river?, okay?, and where are you now? Miko and Riko worked to establish a common perspective of the map and an understanding of what features of the map they had or did not have in common.
Findings from Class 2. The listener was probably able to complete the task because (a) the listener used good guessing strategies; (b) the listener relayed specific, helpful information to the speaker; (c) the listener and the speaker worked together to build up a common understanding; and (d) the speaker was concerned about what the listener could or could not understand.
Class 3. Eiko’s map shows that this pair had three problematic areas, the west coast area, the castle and the end of the route. Kiko and Eiko’s tapes show that Eiko was deeply confused during the interaction concerning the west coast route. She drew the start of the route close to the coast of the island with the palm trees on the right of her route. She did so even though she nodded in agreement with Kiko’s description of the route going between the palm trees and the mountains, with the palms trees on the left.
During the retrospective interview, which continued for 96 minutes and included a videotape replay, Eiko said that she thought the coastline was the road that she had to follow. When asked why she drew her route near the coastline, Eiko said she could not get the idea of the coastline as the road out of her head, itte iru koto yori omoukomu koto ga yusen suru (the idea that the coastline was the road was set in my mind and took precedence over whatever Kiko said). Eiko, however, eventually concluded that she needed to draw her own route, I think this line is road so I go on the line at first but near the lake there isn’t a road so I can’t go the route, but here [area above the lake] maybe I should draw the road.
Before the start of the task Eiko said she was nervous, I get nervous beginning of the class. Although she did not appear nervous on tape and did not ask the teacher to re-explain the task, Eiko was quite unsettled saying, teepu to kamera o ishiki shite konran shite, nani o sureba ii no wakaranai (I was very conscious of the tape and the camera and felt a lot of confusion, I did not know what to do).
Eiko’s strategies. Eiko employed the strategies that Nobu and Roko used—silence, short responses, and repetition of key words with rising intonation. Unlike Roko (who correctly completed the task), however, Eiko did not give specific information to her speaker. For example, she did not clearly state that she did not have a castle, instead she asked, is there a castle near the island?
Findings from Class 3. The listener’s difficulty appeared to stem from: (a) the listener’s affective (nervous, self-conscious) state; (b) the listener’s misunderstanding of the task; and (c) the listener’s use of nonspecific responses.
Class 4. Although Tomi’s map shows that she drew in the route from start to finish, the tapes show that this pair actually failed to complete the task. Like Eiko in Class 3, Tomi thought she had to follow the coastline of the island like a road, uh I found the city but I can’t reach there, in the picture I can city but your route is not suzuku (is not connected to the city), if I can fly I reach. Unlike Eiko, Tomi continued thinking that she needed to use the coastline as a road and even asked her partner, would you read me another route? for example down we go to mountains?, suggesting that they could take the southern coastline from the start point X to the city. Twelve minutes into the task, the teacher intervened and told Tomi to draw the route on the map. Tomi pointed to the coastline and said, this is not route? oh sorry I mistake, oh I understand this is the route, aah sorry. After Tomi realized her misunderstanding of the task, Tomi was able to draw the complete route after four minutes of interaction with the speaker, Sami.
During the retrospective interview, which was limited to only 26 minutes of verbal report with no replay of the videotape, Tomi said, first I misunderstood, I’m the confusion, I image I’ll drive. Although she understood that she needed to pass in front of a castle, she said she did not care and continued to take the coastline as the path to follow; she said, ni iku shika nai kara don’t care about the castle (this coastline is the only way to go so I don’t care about the castle).
Tomi’s strategies. Tomi also used silence, short responses, and repetition of key words with rising intonation. She also employed strategies that that did not help to specify her problems, using responses or questions that gave her speaker very little information such as, we have different map, would you read me another route, where will you guide me? Her responses were also noticeably negative, for example, no, I don’t; no, there’s one way; no...; and but I can’t reach there.
It should be noted that the speaker, Sami, appeared more concerned with her own purposes then with whether or not her listener, Tomi, was able to draw the route or not. Even after Tomi indicated non-understanding, Sami remarked, really? then I don’t know what shall I do, well, uh, I have to go so, then, uh..., and she then tried to continue directing Tomi further along the route.
Findings from Class 4. The reasons why the listener could not complete the task include: (a) the listener’s misunderstanding of the task; (b) the listener’s use of nonspecific responses; and (c) the speaker’s concern for the speaker’s purposes and disregard for the listener’s.
Discussion of Findings
This exploratory study has attempted to examine the role of the listener in collaborative discourse by observing interactions between four pairs of speakers and listeners working on a task. Concentrating on one problem from each dyad, the researcher detected a variety of listener and speaker strategies and performance factors which appeared to influence the process of the listener’s construction of meaning and subsequent completion (or noncompletion) of the task.
The task was not completed or incorrectly executed when (a) the listeners misunderstood the specifications of the task, (b) the listeners (and speakers) misidentified referring expressions, (c) the listeners used nonspecific responses, and (d) the listeners were possessed with one idea. The listeners seemed to have taken positions of “high risk”--high risk takers are “less likely to look for and less likely to query discrepancies between information in the speaker’s contribution and what is already known” (Rost, 1990, p. 228; see also Brown & Yule, 1983ab).
The task was completed when (a) the listeners used good guessing strategies and (b) the listeners relayed specific, helpful information to the speakers. These kinds of “low risk” strategies are used by listeners who want “to be assured that the understanding arrived at is as close as possible to a targeted understanding” (Rost, 1990, p. 228; see also Brown, Anderson, Shadbolt & Lynch cited in Rost, 1990).
Are low risk tactics better than high risk ones? Rost (1990) advises that listeners should aim for a balanced use of high and low risk strategies at appropriate times. Eiko’s abandoning of the idea of the coastline as the route to follow and adopting the idea of drawing in her own route is an example of a listener’s timely switch in strategies. Teachers need to look closely at the kinds of strategies their students are using and teachers need to determine whether these strategies are being employed in an appropriate and timely fashion.
Task outcomes, completion or noncompletion of the map, were also influenced by the work of the speakers: (a) the speaker and the listener built up a common perspective; (b) the speaker and the listener did not utilize good eye contact; (c) the speaker, concentrating on her own goals, disregarded the listener’s; and (d) the speaker was concerned about what the listener could or could not understand. The role of the listener in conversation is a crucial one but the role of the speaker seems to be equally important. Roko, the successful listener, worked with a speaker who cared about her partner’s comprehension, while Tomi, who could not complete the task, had to deal with a less generous speaker. Listeners will often have to cope with inept speakers and their deficient messages. It is important for teachers to help students learn how to manage communication with imperfect speakers.
Task performance was probably also dependent upon the listeners’ affective states. During their retrospective interviews, the subjects reported how nervous they felt, and how self-conscious the tape recorder and camera made them feel. It was observed, however, that the subjects, although tense, remained for the most part in English during the task. Teachers must work to create learning situations and tasks with optimal levels of pressure--pressure that compels but does not cripple.
Limitations of the Study
Several limitations of the study should be noted:
1. The study used a very small homogeneous group of subjects. Younger subjects from academic settings or older male subjects may approach and perform the map task quite differently.
2. In this study, the listeners used pencils to draw their routes on the maps and were observed using erasers during the task. Some information on how the listeners were reformulating their understanding of the speakers’ directions was probably lost when the listeners erased their “mistakes.” The researcher suggests that pens be used instead of pencils in future studies.
3. The effect of videotaping and audiotaping on the subjects appeared substantial. The researcher should have begun tape recording and filming the class several weeks prior to the test date in order to give the subjects a chance to become accustomed to the equipment and taping procedures.
4. Several problems with the retrospective interviewing procedures should be noted. The subjects were not given prior training in verbal report procedures. In addition, because the study was conducted during a regular classroom period, the listener, the speaker and the overhearers were interviewed together, and for one of the classes there was no time to view the videotape of the task interaction. For the same reason, all of the subjects were not given the choice to use their L1. Using Japanese might have helped the subjects express their thoughts and ideas with greater facility. The researcher recommends that, if possible, the listener, as the focus of investigation, be interviewed individually, and be given training in verbal reporting in future studies. Adequate time should be arranged for all interviews.
5. The possible effect of the overhearers on the listener and the speakers’ behavior, although not examined in this study, cannot be discounted. Yoko, the listener in Class 1, who reported feeling very anxious and upset during the task was observed glancing at the overhearers many times. Roko, the only listener to correctly complete the task, was in Class 2, the only class without overhearers. In addition, although the overhearers did not to talk during the task, they did employ many other communicative signals such as laughing, sighing, nodding, looking up, and saying mmm or hmm.
6. In this study the procedure of listening to the audiotapes and observing the videotapes of the map task interactions was used. A discourse analysis of the interactions should be undertaken to examine the interplay of listener, text, and context more adequately and to see where the lack of smoothness, fluency, and vocabulary in L2 production may have contributed to comprehension problems.
The map task outcomes provided some information on the listeners’ problems but not all of the dilemmas they faced. The videotapes and audiotapes of the interactions gave a more complete picture of the listeners’ difficulties and some possible reasons for their problems. The retrospective verbal reports furnished the researcher with supportive and supplementary data on the listeners, such as their affective states. Researchers and teachers stand to learn a great deal from the method used in this study for examining listener behavior, and the interaction of the listener, text, and context in collaborative discourse.
1. The research projects that Brown’s work is based on include Assessing spoken language competence and Listening comprehension both supported by the Scottish Education Department, Maintaining comprehension by the Economic and Social Research Council, and Learning listening skills by the Leverhulme Trust (Brown, 1986, 1989).
- Anderson, A., & Lynch, T. (1988). Listening. Oxford: Oxford University Press.
- Brown, G. (1986). Investigating listening comprehension in context. Applied Linguistics, 7, 284-302.
- Brown, G. (1989). Making sense: The interaction of linguistic expression and contextual information. Applied Linguistics, 10, 97-108.
- Brown, G. (1990). Listening to spoken English (2nd ed.). London: Longman.
- Brown, G. (1995). Speakers, listeners and communication: Explorations in discourse analysis. Cambridge: Cambridge University Press.
- Brown, G., & Yule, G. (1983a). Discourse analysis. Cambridge: Cambridge University Press.
- Brown, G., & Yule, G. (1983b). Teaching the spoken language. Cambridge: Cambridge University Press.
- Buck, G. (1991). The testing of listening comprehension: An introspective study. Language Testing, 8, 67-91.
- Clark, H. H. (1992). Arenas of language use. Chicago: The University of Chicago Press and Center for the Study of Language and Information.
- Cohen, A. D. (1987). Studying learner strategies: How we get information. In A. Wenden & J. Rubin (Eds.), Learner strategies in language learning (pp. 31-40). Englewood Cliffs, NJ: Prentice/Hall International.
- Derwing, T. (1989). Information type and its relation to nonnative speaker comprehension. Language Learning, 39, 157-172.
- Dunkel, P. (1991). Listening in the native and second/foreign language: Toward an integration of research and practice. TESOL Quarterly, 25, 431-457.
- Dunkel, P., Henning G., & Chaudron, C. (1993). The assessment of an L2 listening comprehension construct: A tentative model for test specification and development. The Modern Language Journal, 77, 180-191.
- Hoven, D. (1991). Towards a cognitive taxonomy of listening comprehension tasks. SGAV Review, 9, 1-14.
- Rost, M. (1990). Listening in language learning. New York: Longman.
- Rost, M., & Ross, S. (1991). Learner use of strategies in interaction: Typology and teachability. Language Learning, 41, 235-273.
- Rubin, J. (1994). A review of second language listening comprehension research. The Modern Language Journal, 78, 199-221.
- Smagorinsky, P. (1994). Think-aloud protocol analysis beyond the black box. In P. Smagorinsky (Ed.), Speaking about writing: Reflections on research methodology (pp. 3-19). Thousand Oaks, CA: SAGE.
- Ur, P. (1984). Teaching listening comprehension. Cambridge: Cambridge University Press.
- Vygotsky, L. (1986). Thought and language. (A. Kozulin, Trans.). Cambridge, MA: The MIT press. (Original work published 1934)
- Witkin, B. R. (1990). Listening theory and research: The state of the art. Journal of the International Listening Association, 4, 7-32.
- Witkin, B. R. (1993). Human information processing. In A.D. Wolvin & C. G. Coakley (Eds.), Perspectives on listening (pp. 23-59). Norwood, NJ: Ablex Publishing Corporation.
Appendix A. Speaker Map
Appendix B. Listener Map