2. NEGOTIATIONS OF SHARED UNDERSTANDING BY MEANS OF A REACTIVE CONSTRUCTION

In a very general sense people use their conversations for – among other things – negotiating shared understandings of situations and topics. However, this is often done only implicitly and allusively; as long as the verbal interchange evolves relatively smoothly, participants do not engage in metalinguistic commentaries or repair. But when common understandings seem to be at risk, they do indulge in repair (Schegloff Reference Schegloff and Givón1979, Reference Schegloff1992; Couper-Kuhlen & Thompson Reference Couper-Kuhlen, Thompson, Hakulinen and Selting2005). In our case we will examine a particular conversational resource in Swedish which has some links to repair, the x-å-x construction. This construction occurs in situations where a linguistic expression (x) has been used in a prior utterance, either other-responsively, when the prior utterance has been produced by another speaker, or self-responsively, when that utterance was the speaker's own.Footnote ² The incoming speaker finds the expression x to be only partially acceptable in the situation at hand; this x is therefore reacted to and negotiated in a turn which is initiated by x-å-x, i.e. a unit in which two identical x’s are conjoined by the additive coordinator å [ɔ] ‘and’ (spelled <och> in the written standard), and then further elaborated with supporting arguments in a turn-part that we will call an expansion.

For a start, let us give a simple example of other-responsive x-å-x, drawn from a short narrative about a German family who lived in Finland (in a place called Träskberg) but found themselves forced to leave the country after the war because German property was confiscated (lines 1, 5). The narrator, G, uses the verb flytta ‘move’ in line 6 when he refers to the family's return to Germany:

1. (1) FLYTTA Å FLYTTA (SAM:V1)

   (Conversation between six young men who are eating and drinking together)Footnote ³

• 

The first segment of the response turn in line 7, x å x, contains the variable element x which thus occurs twice. The x in this example is the verb flytta ‘move’, which causes a controversy on how to describe the situation at hand. Speaker M responds in line 7 with a turn which is prefaced by the particle nå, which is typical of dispreferred or non-straightforward answers, and the unit flytta å flytta. This turn-initial material is followed by ja menar ‘I mean’, leading into a somewhat more substantial contribution, the affective and resigned question va fan kan du göra ‘what the hell can you do?’.

The coordinating x-å-x format suggests in an iconic way that there is more than just one understanding of x, in (1) a neutral sense of moving house of one's free will and a sense which may entail a meaning of running away because of exigent conditions. Thus, the response focuses on a possible lack of intersubjectivity and a possible breach in agreement between the participants at this specific point of interaction, and it especially stresses that M has a subjective perspective on the matter, or at least on G's word choice – a perspective which G is made aware of with the x-å-x move. However, M does not really explain why flytta is controversial, he merely hints at it by pointing out that the family did not have a choice in the situation. After line 8, the participants start talking about other things, thereby abandoning the previous topic. This abrupt closing is rather typical of x-å-x events; they do seldom evolve into lengthy argumentations.

In this paper we will discuss x-å-x with respect to (partial) intersubjectivity and (partially) shared understandings between interlocutors involved in conversations. We will focus on examples in which these aspects are made more explicit than in Extract 1, i.e. when the expansion part following the x-å-x segment itself is more elaborated and hence provides more insight into the speaker's stances. The remainder of the paper will have the following outline. In Section 3 we will summarize some main points from earlier research on x-å-x. After that (Section 4) we will account for our data and formulate our research questions in more detail. The discussion in Section 5 will attend to the fact that intersubjectivity can be partial both between and within individual participants. The study of interactivity will therefore shed light also on split ideas within individuals. In Section 6 we summarize and develop our conclusions about intersubjectivity and subjectivity.

3. PREVIOUS RESEARCH ON THE CONSTRUCTION

The x-å-x construction and its contexts have been extensively researched earlier by the authors of this paper. These studies have mainly concerned the usage in Swedish (Lindström Reference Lindström1999, Reference Lindström2006; Lindström & Linell Reference Lindström, Linell, Engdahl and Londen2007; Norén & Linell Reference Norén and Linell2007; Linell Reference Linell, Bergs and Diewald2009b; Linell & Mertzlufft Reference Linell, Mertzlufft, Günthner, Imo and Bücker2014), but the same expressive pattern has been noted in reference grammars and dictionaries of Danish, Norwegian and Finnish. Further, this conversational practice has been attested in Icelandic and Swedish-influenced Estonian (for details, see Lindström & Linell Reference Lindström, Linell, Engdahl and Londen2007). In a study of German, Finkbeiner (Reference Finkbeiner2012) showed that a corresponding pattern (x-und-x) occurs in electronic discussion forums, but the construction is not attested in spoken German data. The languages above are Scandinavian or spoken in the Circum-Baltic region, which suggests that this particular reactive construction is an areal phenomenon.

X-å-x has features of a ‘classical’, functionally specific grammatical construction (Fried & Östman Reference Fried, Östman, Fried and Östman2005), with slots for variable elements (x) and a frame containing an obligatory lexical item (å ‘and’). The meaning of this constructional pattern is not transparent from its constituent parts, which is why it could be called a formal idiom in the same manner as the English the X-er the Y-er construction, e.g. the sooner the better (Fillmore, Kay & O'Connor Reference Fillmore, Kay and O'Connor1988). The coordination of two identical copies of x, like in flytta å flytta ‘move and move’ (1), is to be interpreted in an asymmetrical, distributive way, i.e. it involves different kinds of x rather than the same kind (compare constructions like there are teachers and teachers).Footnote ⁴

As already mentioned, the construction x-å-x consists of two parts: firstly, the x-å-x segment itself and secondly, a following expansion which typically has a clausal form or consists of several turn-constructional units. The initial x-å-x segment is not integrated in any clausal syntax but stands in a pre-front-field position, sometimes preceded by a discourse marker (like nå, ja, allså ‘well’).Footnote ⁵ The expansion should be considered a more or less obligatory part of the whole construction (Lindström & Linell Reference Lindström, Linell, Engdahl and Londen2007).Footnote ⁶ Its main function is to concede that x is not completely inappropriate in the given situation but that it is not entirely acceptable either. It alludes to the fact that the meaning potential of x (as a lexical resource) has several semantic aspects, and not all of these are relevant or applicable in the given situation (Norén & Linell Reference Norén and Linell2007). Thus, the construction is not used to directly resist or challenge the wordings of a previous speaker, as some forms of repeat-prefacing answers do (see Bolden Reference Bolden2009); rather, the speaker who produces x-å-x points out that understandings regarding a specific characterization (x) may be divided and must therefore be negotiated in the interactional context of use.

The responsive characteristic of the construction means that it can only occur in second position to a prior utterance or turn. The prior unit is typically a question, or sometimes a statement like in (1), that contains the expression of x. Often it concerns a matter of which the recipient (the one who comes up with the x-å-x response) may be expected to have first-hand knowledge, i.e. from the point of the previous speaker, it is a B-event in the sense of Labov & Fanshel (Reference Labov and Fanshel1977).

The x in the construction can evidently be of virtually any grammatical category, but it is most typically just one word (Lindström & Linell Reference Lindström, Linell, Engdahl and Londen2007). The x is repeated either in exactly the same form as in the source utterance, or in an uninflected, paradigmatic base form (which may suggest a more abstract treatment of the repeated item). As already noted, x å x is often preceded by a particle, typically ja or nå (the latter in Finland Swedish) or a discourse marker like allså, which all signal that a non-straightforward response is forthcoming (compare well in English and ben in French). The transition to the expansion is usually marked with the adversative conjunction men ‘but’ and occasionally with the discourse marker ja menar ‘I mean’ which signal a contrast or a complication. In prosodic terms, the x-å-x segment and the expansion are disjunct; the two x’s are both stressed, sometimes with some extra prominence on the second conjunct (Lindström & Linell Reference Lindström, Linell, Engdahl and Londen2007: 64–66).

The construction occurs across the board in colloquial or relatively colloquial registers in Swedish spoken language but also in the written medium, especially in interactional electronic discourse and literary dialogue. Previous studies of x-å-x have accounted for this construction's interactional-grammatical properties. For example, Lindström (Reference Lindström1999) deals with x-å-x in relation to other reduplicative expressions, Linell (Reference Linell, Bergs and Diewald2009b) argues for the status of x-å-x as a grammatical construction (in a technical sense), Lindström & Linell (Reference Lindström, Linell, Engdahl and Londen2007) accounts for the grammar of x-å-x and its other- and self-responsive variants in conversation, Norén & Linell (Reference Norén and Linell2007) view x-å-x as a participants’ resource for exploring the meaning potentials of words (the ones occurring as x), and Linell & Mertzlufft (Reference Linell, Mertzlufft, Günthner, Imo and Bücker2014) discusses x-å-x in relation to other reactive constructions in Swedish and German. However, the topic of this paper, i.e. the ramifications of x-å-x for issues of intersubjectivity, has not been highlighted anywhere else.

4. DATA AND RESEARCH QUESTIONS

Three of the four excerpts to be analyzed in this paper are drawn from a series of talk-shows (called Min sanning ‘My truth’) broadcast by the Swedish (State) Television in 2014, and they come from one single show. The series was marketed with the slogan en gäst, ett samtal, en timme (‘one guest, one conversation, one hour’). Each show is devoted to one interviewee, usually a well-known public figure, often a politician, artist or business person, who has withdrawn from his or her previous post(s). The interviewer asks quite a number of provocative questions, which might serve as elicitors of problematizing responses, sometimes involving x-å-x with comparatively long expansions. Clearly, the talk-shows belong to a genre (Luckmann Reference Luckmann, Linell and Aronsson2002) or activity type (Levinson Reference Levinson1979) that is different from many other speech exchanges, both private conversations and broadcast news interviews (note, however, that the shows are marketed as ‘conversations’).Footnote ⁷ (Our fourth example, (5), also comes from a televised interview within the same genre.) Seen in relation to our entire collection of about 50 recorded authentic occurrences of x-å-x in their sequential contexts (see other publications cited above), which is mainly drawn from private conversations (e.g. Extract 1 above), professional–client exchanges (e.g. police interrogations and school contexts), and brief radio and television interviews, the talk-show data tend to be slightly deviant, in that they involve expanded sections of arguments and concessions after the x-å-x segment, something which is a minority feature in the collection as a whole. It is the expansions that give us an opportunity to analyze the participants’ navigations in the intersubjective problematic.

We look upon the present article as a study of how intersubjectivities are achieved (or not achieved) in interactivities in conversation. A more specific research question concerns the role of x-å-x as a discursive resource for making inferrable understandings more precise, but in particular for showing that understandings are oftentimes divided and can only be partially, rather than completely, shared.

Our study is based on a qualitative analysis of a few episodes from the same kind of interview genre. One may therefore raise doubts about generalizability. However, we have extensive evidence from other studies of x-å-x (that for obvious reasons could not be repeated here),Footnote ⁸ and we can testify to the fact that the talk-show data are not different from other data, as far as x-å-x is concerned, except for one possible point, namely, that they involve longer expansions of arguments. This feature was the main reason for choosing these data for this paper on intersubjectivity, as the longer argumentations open a wider window to the interactional dynamics of the x-å-x events.

5. ANALYSIS

The first three of our examples come from a recording of one and the same talk-show, in which a well-known Swedish politician is interviewed about his professional and personal life. Despite the fact that the show is part of public media, we have chosen to anonymize some aspects of the data. For example, the main character in Extracts 2–4 has been given the fictive name Lars Svensson.

The talk-show is about 57 minutes long, and Svensson is of course the person who talks for most of the time (the other participant is a well-known female political journalist whom we here call Lotta Rosenberg). During the show, Svensson uses x-å-x three times. In two of the cases (2, 3) the problematized expressions deal with verbs for activities of speaking, tala ‘speak, talk’ and prata ‘talk, speak’. Just before entering in Extract 2, the parties have talked about the fact that S is constantly active, among other things, with authoring books. In line 1 then, the interviewer (R) presents yet another example of S's many interests: ‘you can speak a host of languages’. (In the transcribed extracts (2) and (3), a series of commas (,,,,,,) indicates averted gaze.)

1. (2) TALA Å TALA

   (TV talk-show with a well-known politician, S, and an interviewer R)

• 

In line 2, Svensson reacts to the proposal that he can ‘speak’ many languages, producing the verb tala in the format x-å-x. His response is that he can ‘use’ about ten languages, which arguably is a relativization tuning down the suggestion; presumably, he means that he can use his ten languages for certain purposes, but that it would be an exaggeration to claim that he could ‘speak’ them across the board of all possible situations. What S means exactly is of course difficult to tell; the expansion is characteristically brief and semantically underdetermined (what does ‘being able to use a language’ mean more precisely?). The function of x-å-x is reduced here only to relativizing and diminishing certain aspects of the verb tala ‘speak’. Note however that the expansion following the initial x-å-x part acknowledges the second component of the interviewer's claim, namely, the number of languages S masters (to some degree). The following lines 4–13 are topically related to the first question–answer pair, but they do not pursue the discussion of the situated meaning of x. Participants seem to rely on having reached a sufficient level of intersubjectivity to continue on with the discussion.

Whereas in Extract 2 the participants are basically (and implicitly) talking about how knowledgeable one has to be in order to be qualified for the description ‘speaking a language’, in the next example they talk about the extent and depth with which one should talk with another person for the communication to count as ‘talking in depth about a given topic with a particular interlocutor’. In the background there is the question whether Svensson, who is still the main character, is nowadays at all ‘on speaking terms’ with another person (called Dahl below).

The immediate co-text for the talk in Extract 3 is that Rosenberg, the interviewer, has just played back an old recording from a discussion between Svensson and another very well-known politician, a former prime minister (Dahl). This discussion took place in the autumn of 1994, the night after the Swedish referendum about joining the EU. Svensson and Dahl were the leading figures on the two opposed sides. During this nightly exchange, the two had fallen out on each other in a rather remarkable way, considering that their conversation was broadcast on the radio. At one moment (played back in the talk-show of 2014) Svensson roared at Dahl: ‘shut up, Jens Dahl’ (Sw. håll truten, Jens Dahl), which was responded to by Dahl by recommending Svensson to go home and go to bed to get some rest. Thereby Dahl had insinuated, Svensson argues, that he (Svensson) had had a hangover. This is the topic of Svensson's ironic remark in lines 1–2 (‘a great rhetorician’). This move is followed by Rosenberg's (the interviewer) exclamation som du ‘(just) like you’ (line 3), followed by the quote håll truten ‘shut up’ (in the imperative), the words Svensson had used in the debate with Dahl.

1. (3) PRATAT OM Å PRATAT OM

   (TV talk-show with the same participants as in (2) above)

• 

Rosenberg's question in line 5 deals with whether Svensson and Dahl have talked about their public verbal fight at some later occasion. Svensson's x-å-x retort is unusual in that it cites a part – pratat om ‘talked about’ – which is not a constituent in the source utterance (line 5); there, pratat is the tensed predicate verb and om ‘about’ is a preposition taken from another constituent, i.e. what is usually analyzed as a prepositional object (here: om de här ‘about this’). While x-å-x has quite a wide applicability with regard to what kinds of source x:s it can take, this is an unusual example also from this point of view. The reason for Svensson to use it in this way is arguably that the problem lies in the combination of the verb and its implied object. He has been on speaking terms with Dahl after the infamous verbal fight (line 6), and ‘can have respect for’ him (lines 7–11), but the two have apparently not spoken about the specific event, at least not in any depth.

This episode is arguably delicate for Svensson to handle; after all, he has told a former prime minister to ‘shut up’ in a non-private discussion broadcast on the radio, hardly a very respectful conduct. And now this has been played once again to a TV audience. It is therefore not a surprise that Svensson is keen on ensuring (twice) that he ‘has/can have respect’ for Dahl (lines 8 and 11), although he also suggests that the latter is a ‘great rhetorician’ (line 1), which was presumably not meant as a praise. The keenness to smoothe over a problem may also come to the fore in Svensson's initial, straightforward response to Rosenberg's question, ‘indeed we have’ in line 6. It is only after this acknowledgement that Svensson initiates a concession with x-å-x, as if trying to correct the first, immediate version of his reply which could have been interpreted as stating that Svensson and Dahl have in fact been talking about the verbal fight. There are thus many indications that the situation at hand involves mixed feelings, a good breeding-ground for x-å-x.

The delicacy of this particular episode is also shown in Svensson's non-verbal conduct. One should know that he is a self-confident person who speaks his mind without hesitations in the interview. Like any other speaker, he vacillates between gaze contact and looking away, but here he averts gaze only for very brief moments. However, during the production of lines 6–11 in Extract 3, Svensson avoids gaze contact with the interviewer for most of the time. These lines belong to the expansion of the x-å-x construction, where the speaker is balancing the two interpretations at stake. During the very start of his response turn (beginning of line 6), where the immediate reaction is produced, he looks at his interlocutor, and he resumes gaze contact when he has arrived at the location where the ‘corrected version’ is produced, pratat me varandra ‘spoken with each other’ in line 7.Footnote ⁹ Svensson's behavior may indicate that he would rather avoid talk about the 1994 radio event, and instead tries to ‘scale it down’ (see Section 6 below) from ‘talking about’ the delicate issue to ‘talking with each other’ (line 7).

The third example (Extract 4) from the Svensson–Rosenberg talk-show follows on a fairly long sequence that we have omitted for reasons of space. The gist of this sequence was that Svensson, who was a young liberal during the 60’s, was back in those times publicly arguing against marriage and for free love. However, his own two marriages broke up, partly as a consequence of infidelities that the spouses could not handle.

In (4), the interviewer asks about the ‘political idea’ behind ‘free relationships’ (lines 1–2), which of course invokes various interpretations of ‘politics’ and ‘political’. Svensson's lengthy response (lines 3–16) revolves around the relation of marriage (a personal relation between two individuals) and politics, something that concerns the relationship between a private sphere and a public or societal policy. Marriage is both (i) a political issue for governments, parliaments, the church, etc. for regulating how individuals share lives together, especially as regards handling sexuality, cohabiting etc. in personal relations, and (ii) an individual, everyday issue: parties (spouses) must regulate their life together, take a common stance to cohabitation with or without marriage, monogamy vs. polygamy, extramarital relations, attitudes to homosexuality and other forms of sexual relationships.

1. (4) POLITISK Å POLITISK

   (TV talk-show with the same participants as in (2) and (3) above)

• 

Svensson uses x-å-x to argue for the need for freedom of choice for the persons involved, the need for personal agreement between spouses (or individuals in local communities), while there is no need for legal regulations. Hence, the reactive construction points to a difference between officially formulated, regulated policies and more covert, grass-root policies.

What happens in Extracts 2–4 is that one of the parties, the interviewer, proposes a characterization of circumstances in a first turn, and the other party, the interviewee, reacts to a specific part x of the characterization by using the format x-å-x. The reaction indicates problems with the item x, and implicates tensions in the intersubjective understanding regarding that item in a specific interactional micro-situation. By using a given label x, such as politiska ‘political’ in (4), the interviewer proposes the terms for the ensuing talk, but the interviewee then decides that these terms are only partly acceptable, displayed in the format x-å-x.

For speakers of Swedish, the construction x-å-x is a resource through which the conditions of ensuing talk can be called to question by focusing on the issue that is specifically problematic in them. In such problematizing events, the speaker makes a metalinguistic move (Norén & Linell Reference Norén and Linell2007), drawing attention to different meaning potentials associated with a certain word or phrase, and then, in the expansion of x-å-x, the speaker chooses an aspect of meaning which stands out as relevant in the situation for himself and thereby serves as an acceptable point of departure for further talk on the topic. Hence, there is both a subjective and an intersubjective angle to the x-å-x construction. To raise an issue of meaning potentials is to refer to the (mainly) shared linguistic knowledge within a speech community – a knowledge that should be accessible for both the speaker and listener(s). Hence, whatever the problem with x is, all the parties involved could be able to access the possible problem with x in a given situation. At the same time, the use of x-å-x signals that a certain understanding of x is preferred. What is left open, in the end, is whether this subjective stance to x can accommodate the prior stance of the producer of the source x; nevertheless, it is evident that some possible meanings and stances that can be associated with x need to be cancelled out in the interactional micro-situation. X-å-x is a means of making the co-participant aware of this need.

Let us finally attend to another aspect of intersubjectivity, namely the participation frameworks in which x-å-x tends to occur. Despite the fact that it embodies a reaction to a stance that cannot be accepted in its entirety, the construction is typically used in turns that are themselves monologically organized, almost as if to block off any further response from the other party (who originally lauched x). It is the same speaker who both introduces and completes the x-å-x turn, and the situated appropriateness of x is seldom pursued any further. Thus, the speaker normally gets the last word in and through the turn in which x-å-x is used, and therefore x-å-x could be heard as a marker of a subjective stance. On the other hand, the construction is a reaction to prior talk, and the speaker voices (usually two) different perspectives that mirror what occurs in the language community at large. These latter aspects reflect, of course, intersubjectivity.Footnote ¹⁰

There are partial exceptions from the characterization just proposed. In Extract 5, also from a dyadic talk-show televised in Sweden, in this case in 2015, we have a journalist (I) interviewing a retired financial and industrial leader, whom we here call Carl Anderson (A). The interviewer, in line 1, suggests that A's stressful life led him to be careless (slarva) with sleeping and eating:

1. (5) SLARVADE Å SLARVADE

   (TV talk-show with a retired industrial leader (A) and an interviewer (I))

• 

The use of the verb slarva ‘be careless’ may, in this case, be heard as accusing the addressee of having a weak moral feeling of responsibility for his own health. This is probably what A is on the verge of objecting against, when his x-å-x expansion (‘that sounds almost as’) in lines 3–4 is interrupted by the interviewer. What the interviewer does in line 5, and later, is to pursue his own version, despite A's attempts to modify the picture. Anderson defends himself by pointing out that many individuals end up in the same predicament (provided, presumably, they have A's kind of job) (lines 11–12), and that we are faced with a ‘medical’ outcome (line 17), both arguments that serve to diminish his personal responsibility. Anderson also tries to ward off the interviewer's position by suggesting that his metaphor of ‘burning one's candle at both ends’ is just a ‘popular phrase’ (line 14), thus arguably trying to bring the exchange back to serious topics, such as the kind of hard work he has shared with many. We have brought this excerpt up, because it shows that the issue of situational appropriateness of meaning (here: the justification for the use of the verb slarva) may indeed sometimes be pursued by both participants in an interactive fashion – not least in a debating interview genre. In (5), speaker A is able to put an end to the discussion only in line 17, which is far beyond his x-å-x turn in line 3. Accordingly, the expansion part of (5), roughly lines 5–17, is co-constructed, including insisting moves from the producer of the original x. In Excerpts 2–4, by contrast, we have seen one single speaker constructing the expansions by himself, but the expansions embody an ‘internal dialogue’ between two different voices of his mind. (Note, however, that in Excerpt 2, the two speakers are both involved in mutual alignment in a sequence following the expansion, lines 11–13.)