Lack of response

As already noted in Chapter 4 a speaker in a conversation is, in principle, entitled to a turn lasting up to one turn-constructional unit. Each possible completion of such a unit, which can be a sentence, a phrase or a lexical unit, constitutes a potential transition-relevance place. At these points a change of speaker can occur, but the current speaker can also continue.

The transition is locally managed by the participants within the frame of norms and expectations they have about conversational behavior. The speaker has an obligation to yield to the next speaker if the latter starts a new turn at a transition-relevance point. Alternatively, if selected at such a point, the interlocutor is obliged to take the turn. It may, however, happen that the selected next speaker does not take the turn offered so obliging the first speaker to continue. This may happen without any problems, but there are instances where the first speaker may aiso show that the second speaker has not behaved as expected. In the present study, these instances are called lack of response or failure to respond.

This chapter focuses on breaches of this type and the speakers' conduct before and after them.

The reasons for lack of response are various in a normal conversation, but the most probable interpretations are non-hearing, misunderstanding or deliberate avoidance. In conversations with an NNS, problems of understanding constitute a plausible explanation for instances where the second speaker does not take the turn at a transition-relevance point although the first speaker has signalled that s/he has completed

her /his turn. In some instances this is obvious, for example, when the first speaker continues after a short pause but, instead of developing the topic, repeats or reformulates what he has said in the previous unit. In such cases the conduct of the second speaker during the pause is seen as failure to respond. The most unambiguous instances are those where the tum-constructional unit is a question or a solicitation that has a high degree of sequential implicativeness.

Conversations vary as to the strength of the obligation to take the tum. In a dyad there is a strong pressure on the members to participate in a way that keeps the group together (Bales & Borgatta 1965: 502). In a dyadic conversation it is very difficult not to take the tum when it is explicitly offered, whereas in a group conversation tum-taking is more varied and lack of response could be interpreted in various ways. For instance, the selected participant may accomplish a relevant action by turning to a third participant and so avoid responding to the first speaker.

Even in a dyad, lack of response can occur without being face-threatening if the utterance is not an explicit question or solicitation but a statement which has a lesser degree of sequential implicativeness and the listener can reasonably expect the current speaker to continue. Although the fragments analyzed in the present study are mostly dyadic, the expectations about taking the turn are not always the same as in 'normal' conversations. The speakers are aware of the asymmetry of the language proficiencies between them, and failure to respond is not seen as very face-threatening. As a matter of fact, the NNS is actually expected to exhibit low proficiency, and this affects the conduct of both speakers. The NSs tolerate breaches from their NS norms and the NNSs do not need to avoid displaying their inability to use L2.

In negotiating about the transition the recipient is always allowed time to prepare her /his response (see e.g. McLaughlin, 1984: 112). The length of this transition time is determined by the particular circumstances of the interaction but the speakers nonetheless have an idea of its reasonable limits. When no switch occurs the current speaker still has time to retake the floor (Sacks & Schegloff & Jefferson 1974: 704). This means that pauses of considerable length occur naturally in all conversations. When pauses at transition relevance points, however, exceed 3 seconds, especially in focused dyadic conversations, they are often called lapses (Sacks, Schegloff & Jefferson 1974: 714; McLaughlin 1984: 115). Jefferson (1984), however, argues that pauses of more than one second are long in a normal conversation. McLaughlin and Gody (1982) have found an "awkwardness limen" of 3 seconds in their study of pauses in conversation. The limen may be culture-specific. Both Finns and Swedes are said to tolerate silence more than e.g. Anglo-Americans (Daun 1989; Daun & Ehn 1988; Lehtonen & Sajavaara 1985).

The present data suggest that pauses are interpreted differently in conversations with an NNS with low language proficiency. The speakers

become uncertain of the limits of transition times. To identify lack of response is not easy in NS-NNS conversations because of the special character of the discourse. The overall pace in such conversations is slow, but that does not mean that all pauses are automatically and systematically prolonged. Words are mostly pronounced slowly and distinctively, there are pauses within utterances and the pauses between utterances may be long. On most occasions the time offered is adjusted to the overall slower speech rate, which then means long tum-allocating pauses. The transition time can, however, also be very short in instances when the NS is speaking and, anticipating difficulties, does not wait for any reaction but rushes into a paraphrase or giving further information (see e.g. Ex. 39).

Furthermore, the talk-silence structure of conversations is always in a complex way influenced by the situation and the speakers. The conversations in the present data are a part of a task the speakers are committed to since they have joined the project either as researchers or as informants. As mentioned before, the episodes analyzed are mostly dyads with time limits set before the encounter and without the presence of distracting elements. These aspects of the situation might lead the participants to make more effort to keep the conversation going than they otherwise would. On the other hand breaches occur, since, in these conversations, being non-native or having a non-native interlocutor is a valuable point of departure, and all the strategies of communication, such as long waiting and repairing times, are more acceptable than in situations where native-like language behavior is expected.

In the present study the starting point in describing lack of response is not the exact physical length of pauses but the interlocutor's reaction which indicates that a breach has occurred. Such breaches are either factual instances of non-understanding or cases interpreted by the interlocutor as such.

In the following sections, I shall present a description of the participants' behavior in fragments where the first speaker has to continue and undertake repairs after a longer lapse. Although both the difficulties and the repairs are jointly produced, the speakers have different roles when they occur. Firstly, 5.2 addresses instances where the first speaker reaction constitutes a border line case between clear-cut repairs and joint production that is slow for other reasons, thus the lapses cannot always be labeled as failures to respond. Secondly, 5.3 describes efforts to negotiate for understanding in instances where the lack of response initiates a repair. Thirdly, 5.4 deals with instances where the first speaker initiates a self-repair, but this complicates the situation and a lack of response follows. After this lapse the repair is resumed.