Sequence, gaze, and modal semantics: modal verb selection in German permission inquiries

Extract 1

‘can one build at any time’ PECII_DE_Game_20151113_1603916^[1]

Extract 2

‘may one do that’ PECII_DE_Game_20210602_1448889

In Extracts 3 and 4, one and the same person, Carola, uses first kann, and a moment later, darf, to inquire about doing a “U-turn” with a car.

Extract 3

‘can I U-turn right here then’ PECII_DE_Car_20171031_1_366206 (6 mins 26 secs) (simplified)

Extract 4

‘may I U-turn here now’ PECII_DE_Car_20171031_1_417620 (6 mins 57 secs) (simplified)

We will analyze these data in more detail later. For now, just note the following:

1. Both the turns with kann and the ones with darf are taken by participants to concern possibility in a deontic mode – next speakers confirm the permissibility of the relevant action (1 and 2) or point out that there are rules prohibiting it (3 and 4);

2. These interaction data suggest that können and dürfen can be used in very similar ways to talk about deontic possibility.

It is well known that können and dürfen can both express deontic meaning. However, the systematic differences between the two verbs when used to accomplish deontic meaning are not well understood. We will argue that we can better understand how speakers select between different modal verbs by considering where the modal utterance is placed in a course of action in conversation, in short its sequential context. Interactional-linguistic research has found that language structures at different levels of organization are adapted to and rely for their core meaning on very specific contextual “home environments”. For example, it has been shown that choices in sentence mood are informed by the connection of a speech action to the wider course of activity within which they occur (in relation to requests: Raymond et al. 2021; Rossi 2012; Zinken and Ogiermann 2013; in relation to offers: Curl 2006). Here, we extend this approach to the study of lexical items. But first, we turn to some background on deontic modality and modal verbs.

Deontic modality is traditionally understood as relating to obligation or permission stemming from some authority, which can be another person or an impersonal body of rules and norms (e.g., Palmer 2001: 10). It has often, and sometimes critically (Nuyts et al. 2010), been noted that “obligation” and “permission” refer to social relations and social actions. Hence, we can initially characterize the meanings of utterances such as the first turns in the extracts above as “asking for permission”. However, on closer inspection, the associations between deontically used modal verbs and speech acts such as asking or giving permission are not straightforward. First, of course, it is well known that modal verbs such as können or dürfen can express various modal “flavors” (see discussion below). Second, modal expressions are not necessary to ask for permission (“Do you mind if I …”). And finally, when speakers do (ostensively) inquire about the acceptability of their actions using a modal verb, these moves are sometimes not well glossed as “requesting permission”: often, people already start doing the focal action (say, make a game move) as they ask, which casts doubt on the notion that the action achieved with such modal inquiries is always well glossed as requesting permission (Deppermann and Gubina 2021b). In sum, there is no necessary link between modal verbs and either expressing deontic meaning or accomplishing actions of seeking permission. The question therefore arises: what exactly do the verbs können and dürfen contribute to social actions that pave the way to some behavior? Before we tackle this question, we introduce some background on modality in general and the modal verbs können and dürfen in particular.

Extract 5

‘may one/is it allowed to eat in your car’ PECII_DE_Car_20180907_1095567

The crucial observation for us here is that Roland’s permission inquiry has an interactional pre-history. In fact, even within his turn-at-talk, it emerges out of something prior: his out-loud saying that ‘maybe I need to ea-’ (vielleicht muss ich was (0.4) ess- [line 5]). This ostensive self-talk is hearable as connecting back to prior matters. Indeed, Roland’s move to eat something, and his darf-inquiry, emerge out of a problem that has been the subject of talk between Lara and Roland in the minute or so leading up to Extract 5. Extract 6 shows some of this interaction. The relevant pre-history of Roland’s darf-inquiry begins when Lara notices that Roland seems to have fallen asleep. As it turns out, Roland isn’t sleeping. Responding to Lara, he connects his sleepiness with the fact that it is “mega cold” in the car (line 6). He thus articulates a problem he experiences. He moves to turn up the heating, but notices that Lara has actually turned it further down. This leads to their interchange about temperature preferences. Roland now explicitly names the problem that “when it’s cold I fall asleep” (line 20).

Extract 6

‘may one eat in your car’ PECII_DE_Car_20180907_1073360

The situation so far then is that Roland has reported a problem – he feels cold, and this makes him sleepy – and he has attempted a solution, turning up the heating, which, however, has been frustrated because Lara’s preferences run in the opposite direction. In this context, his bending down towards his bag, accompanied by his comment that vielleicht muss ich was ess- (‘maybe I need something to ea-’) is recognizable as articulating another potential solution to his problem. When Roland breaks off this emerging utterance and instead launches his darf-inquiry, he orients to the relevance of getting the driver’s (and car owner’s) permission (Rossi and Stivers 2021). At the same time, his inquiry nominates the relevant action as a solution to a problem (a need or wish) that has already been made public. Notwithstanding all the unique quirks of this particular fragment of interaction, this observation holds for darf-inquiries across our collection: darf-inquiries nominate an action as sequentially occasioned by what was made public before, specifically: as a solution to a known problem.

Here is another example, a fragment of interaction that is very different from the first one in many ways, but with a darf-inquiry that, once again, nominates an action as a solution to a problem that is out in the open. Four friends are playing the board game Catan. They are in an early phase of the game, where each player in turn can place a game piece, obeying certain rules of placement. After Anabell chooses a spot on the board for her game piece (line 1), it is Marvin’s turn. Marvin gazes at the board, his game piece in hand (line 2, Figure 2). At line 3, Jaron addresses a piece of information about the game to Marvin, and while the other two players both respond to this information, Marvin does not (lines 3–9). Instead, he keeps screening the game board. He then raises the hand that holds his game piece above the board, further considering it for another full 20 s (line 10, Figure 3). Finally, he moves his hand to a spot where he might place his figure. Hovering just above the chosen spot, he then produces a darf-inquiry, addressing it to Lena by gaze.

Extract 7

‘may one do this’ PECII_DE_Game_20210602_1448889

Although there is no talk here about Marvin’s task of finding a spot for his game piece, the fact that he is having difficulties finding one becomes public in several ways: (1) The known-in-common turn structure of the game, with Marvin’s turn beginning after Anabell has made her choice (on pre-allocation of turns in games, see Hofstetter [2021]); (2) the fact that Marvin “ignores” Jaron’s information addressed to him (lines 3–4), keeping his focus on the game board; and (3) the lengthy period of consideration by Marvin, made visible by his hand hovering above the board, all display the non-straightforward nature of Marvin’s task to his co-players. When it comes, then, Marvin’s darf-inquiry is hearable as nominating an action (“doing that”, i.e., placing the figure on the chosen spot) as a possible solution to a publicly accessible problem, while also soliciting the others’ approval of this action as a legitimate move in the game.

Let us now compare this to the sequential home environment of kann-inquiries. Our first case here comes from another group of friends playing Catan. It is Katharina’s turn, and she announces that she will build two roads (lines 1–3). As she is making her game move, Moritz asks about a game rule: kann man jederzeit (.) äh bauen eigentlich (‘can one, er, build at any time actually’ [line 5]). Gerald first answers in the affirmative (line 7). However, as Moritz then moves to actually build something, it turns out that there was a misunderstanding, and his move is forestalled.

Extract 8

‘can one build at any time’ PECII_DE_Game_20151113_1603916

Note first that Moritz’s kann-inquiry addresses a deontic possibility: just like Marvin in the previous case, he asks whether a particular game move is licensed by the rules of the game. What is different about this case is that the notion of Moritz building something has not been “prepared” by the current context. It is not Moritz’s turn at play, he is not even the next in line, and Moritz did nothing to make public his wish to build something. Moritz himself builds this out-of-the-blueness into the design of his turn: he begins with an audible in-breath, thus orienting to a context in which a turn by him may not be expectable (on audible inhalation in the context of turn-taking, see Robinson [2023]). At a point of possible completion (kann man jederzeit bauen, ‘can one build at any time’), he incrementally extends his utterance with eigentlich, a modal particle that indicates that the question does not emerge out of the current interactional context (König 1977: 123) but introduces a new matter (Thurmair 1989: 176) (for an interactional treatment, see Clift [2001], on English ‘actually’). Moritz’s kann-inquiry then nominates the action of him “building” something not as a solution to a publicized problem, but as a new matter. However, it seems that Moritz’s inquiry is sequentially so disjunct from what happened before that his fellow players, or Gerald at least, mistakenly try to make sense of it by connecting it to earlier talk. Consider the misunderstanding that comes into the open at line 10. What is it that Gerald initially understood and that led him to a (cautious) affirmative response at line 7? It seems that he took Moritz’s kann-inquiry as questioning the validity of Katharina’s game move (announced at lines 1–3). Gerald’s cautious affirmation was not, then, a go-ahead for any building plans of Moritz’s, but a defense of Katharina’s move, and Moritz’s actual intention only becomes apparent when he picks out the relevant resource cards on his hand and announces that he will, in that case, build something (line 8).

Throughout our collection, kann-inquiries bring up new action plans. Here is one further illustration. During a family breakfast that has lasted for about half an hour, five-year-old Timon asks: kann ich aufstehen (‘can I get up’).

Extract 9

‘can I get up’ PECII_DE_Brkfst_20160424_2110402 (simplified)

Again, we see that this request to leave the table comes without “warning”: the last topic of conversation has been the cheese that Timon’s dad, Florian, is spreading on his bread roll (lines 1–3). After that, there are a few seconds of silence, filled with the babbling of Timon’s little sister, Emma. To be sure, the sight of a child sitting in front of an empty plate may suggest to the alert observer that this child may want to get up soon. However, Timon does not do anything that would make public any problem to which getting up would be the solution – say, ostentatiously displaying boredom. Instead, during the 6 s of silence (line 5), he observes his dad’s cheese-spreading for a second or so, then turns his gaze to his babbling sister, before launching his inquiry. In a word, Timon’s kann-inquiry comes out of the blue. It does not connect back to any problem that he would have made public.

Having illustrated the different sequential environments in which speakers select darf or kann to make a modal inquiry, we now turn to consider the forms of addressee participation that the two forms of inquiry invite. We begin, again, with darf-inquiries.

Extract 10

‘may I take a close look’ PECII_DE_Game_20160207_1549512

We can see that Antonia’s darf-inquiry, just like the cases in Extracts 5 and 7, emerges out of a problem that she has (seeing the picture on a game card) and that has already become public, both in Antonia’s embodied conduct (adjusting her body position, line 2) and in her talk (line 4). In her darf-inquiry, Antonia nominates the action she is carrying out as a solution to this problem.

We suggest that darf-inquiries target only authorization, because they come in a sequential context in which they nominate an action as a viable solution to a publicly known problem. In such a situation, all that is left to do for a person who is asked about permissibility is to give a go-ahead (or block the action). Indeed, we find that next speakers regularly provide token approval in turn-initial position. Carola grants Antonia’s request with a yes (line 7). Gisela provides her approval in a marked format, repeating the modal verb (darfst du, ‘you may’ [line 8]). She thereby orients to Antonia’s inquiry as specifically requesting a locally and contingently needed discretion (“regardless of what the codified rules of the game may have to say about this, I approve that you deal with your problem in this way”). This darf-inquiry, while seeking the co-players’ authorization, can be seen to lie on the border to a bouletic understanding, in the sense that not rules or norms, but needs and desires form the relevant background. In any case, Gisela’s marked response constitutes no more than a polar answer (Enfield et al. 2019). Looking back to the other darf-inquiries examined so far, we can see that all response speakers (Lena and Jaron in Extract 7, Lara in Extract 5) begin their responses with a polar answer (ja, ‘yes’, or a head nod in the case of Lena in Extract 7), even if they go on to expand their respective turns. Let’s compare this with responses to deontic kann-inquiries.

Here is an example of a kann-inquiry from another family breakfast. Not for the first time, Sibille and Christof are trying to get their daughter Nora to put away her mobile phone (lines 1–11). While this interaction takes its course, the family’s older daughter Lissi first briefly glances at her mum (line 5), then straightens up on the bench where she has been crouching, and produces a kann-inquiry: kann ich mir eben- kann ich mir die pferde vom (pappel)hof angucken (‘can I just – can I go look at the horses from the [pappelhof] stables?’ [lines 8–9]).

Extract 11

PECII_DE_Brkfst_20160213_3039443

As in earlier cases, this kann-inquiry comes out of the blue. There is no publicly known problem of Lissi’s that the nominated action would solve. Instead, her inquiry launches a new project: leaving the breakfast to watch horse videos or pictures.

Her mum’s response is largely affirmative, but consider the details. It does not begin with a polar answer token. Sibille’s response instead explores the conditions that would make a granting possible. She begins with a tentative approval (ich glaub ja, ‘I think yes’ [line 12]), then articulates a judgment-cum-proposal that would put Lissi’s horse-watching into the context of a more general change of activity on behalf of all the family (wir lösen dann die runde auch langsam auf oder, ‘we slowly bring proceedings to a close, right?’ [lines 12–13]), and she then formulates further conditions that the children should fulfill before turning to other pastimes (clearing away plates, closing the chocolate spread jar [lines 15–16]). In sum, Lissi’s kann-inquiry has created a context not simply for authorization, but rather for exploring deontic “possibilities” by formulating conditions that would make her wish acceptable.

Looking back at Extract 9, we now see that the situation was similar. Timon’s kann-inquiry about getting up from the table was met with a return question from Florian that inserts a sequence clarifying the conditions for approval (bist du satt, ‘are you full’ [line 8]) (see Schegloff [2007], on such “pre-second” insertions). Only after Timon confirms this question does Florian give his approval (with a nod and a soft ja [lines 10 and 12]), although Timon’s mum then further interrogates Timon (line 11) before he finally “can” get up. In this context, Extract 8 from the game of Catan represents an outlier: it is the only kann-inquiry that we have seen that does receive a version of ja (joa [line 7]) in response – but recall that Gerald’s response was based on a misunderstanding.

Extract 12

‘can I buckle up there now’ PECII_DE_Car20160924_13377

Note first that this kann-inquiry lives in the same contextual habitat as earlier ones: it comes “out of the blue”. Before Julica’s turn at line 8, there is nothing that could alert her friends that Julica is having a problem with the seatbelt. In fact, with her turn-initial achso, Julica herself frames her inquiry as concerning a problem that she encountered “just now” (on “achso” in a different context [Golato and Betz 2008]).

Note also that it is certainly possible to imagine giving this kann-inquiry a deontic meaning: Julica might be asking whether Carola minds her buckling up, in case this might ruin the camera set up. However, neither Julica herself nor her friends treat her kann-inquiry that way. Julica does not look to a would-be authorizer (say, Carola, since she is the one who mounted the cameras and is recording the event). Instead, Julica looks at the area where the camera might be blocking the seat belt. Also, Julica does not stay immobile and wait for permission – she pulls the seat belt towards her, thus testing whether she “can” in fact buckle up. And of course, in terms of the social roles in this group of friends, it is just implausible that Carola would prioritize the success of her recording over her friend’s safety, and thus prohibit buckling up. In sum, it seems clear enough that Julica’s kann-inquiry addresses not permissibility, but “circumstantial” possibility. Fittingly, it is Julica who answers her own question as she sees that it is “possible” to use the seat belt (line 8), and none of her co-passengers respond or even raise their heads, effectively treating Julica’s turn as self-talk.

We can see clearly now how context (the position of the inquiry within a sequence of actions) and visible behavior (gaze and posture orientation) enter into the construction and action of modal inquiries. This is summarized in Table 1.

Firstly, speakers create different contexts by orienting to the participation structure of the event. Speakers can treat another person as an authority whose permission is sought by looking at them and “waiting” while inquiring (e.g., Extracts 7 and 9) (see also Deppermann and Gubina 2021b). Alternatively, speakers can orient to the matter at hand, through gaze and manipulation, thereby treating others’ permission as less relevant or even irrelevant (e.g., Extracts 10 and 12). These aspects of embodied behavior create a distinct kind of context that provides an important part of the background for interpreting a particular modal “flavor”. Secondly, speakers nominate actions they want to carry out in different sequential environments: where the nominated action either presents a new matter, or where it is interpretable as a solution to a known problem. This positional kind of context informs the selection of kann or darf in modal inquiries. As the examples in Table 1 illustrate, sequential position (as indexed by modal verb selection) and orientation to an authority both contribute to the deontic quality of a modal inquiry.

Extract 13

‘can I U-turn right here then’ PECII_DE_Car_20171031_1_364820

The design of Carola’s inquiry gives away that this matter has occupied the women before. Her ‘then’ (dann [line 3]) displays a connection to something prior; the nominated choice of doing the U-turn “here directly” suggests that the matter of whether to do it at all has already been settled; and of course, the very action of doing a U-turn is something that drivers tend to do in response to a driving problem. Notice also that Carola’s turn does not include any turn-initial objects that would mark it as coming unexpectedly, such as an achso (‘oh right’ [Extract 12]) or an audible in-breath (Extract 8). Indeed, this modal inquiry does connect to a known-in-common problem: in the minute before this fragment, the women noticed that the road exit they had wanted to take is closed, and they decided to make a U-turn. Carola changed lanes accordingly and the car came to a standstill in front of red traffic lights. This is where our fragment sets in.

It would seem then that this is the “right” context for a darf-inquiry, and of course the selection of darf would be entirely possible here. But it would carry a different meaning: according to our analysis, it would nominate “doing the U-turn here directly” as Carola’s chosen solution to the problem, and would seek only authorization from her co-travelers. By selecting kann, Carola instead raises the action of “doing the U-turn here now” as a matter for consideration, opening up, as it were, a new sub-project within the global project of effecting a U-turn. In the way Carola designs her move, we see her own orientation to her modal inquiry as broaching a matter for joint consideration. She gazes to the road layout and “manipulates” it with her pointing gestures (see Figure 6). Upon possible completion of her utterance, she extends her turn to tentatively answer her own question (nee ne (.) ich muss schon da vorne nochmal (links), ‘no, right (.) I do have to go up there (left)’ [lines 4–6]). In sum, we can see some clear parallels to the way in which Julica designed her circumstantial inquiry in Extract (12), and Carola’s modal inquiry would seem to be open to be interpreted both circumstantially and deontically. Katharina’s response (lines 7–8) picks up on this openness: she first provides information about the (il-)legality of the move (das is in jedem fall illegal, ‘that is in any case illegal’ [line 7]) before suggesting how to best proceed in the given circumstances, aligning with the solution (or parts of the solution, the element vorne, ‘forward/ahead’ [line 6]) Carola has formulated in her own alternative: aber geh ruhig n stück weiter vor ja (‘but do go a bit further ahead yes’ [line 8]). Valeska joins in with yet another possible way to solve their problem (line 9), before Carola announces the solution she has settled on (lines 12–13). This case illustrates how practices of speaking (in our case: the selection of a kann-inquiry or a darf-inquiry) are not determined by context, but reflexively turn aspects of the situation into relevant context (Garfinkel 1967).

An important aspect of the situation that allows for Carola’s kann-inquiry to be heard as broaching a “new” matter for consideration is time. Carola produces her inquiry just as she has brought the car to a standstill in front of red traffic lights. The specific situation of waiting at traffic lights, with its relaxed temporality, thus sheds further light on the different trajectories that we have shown to be mobilized with kann- and darf-inquiries: while kann-inquiries invite consideration of the nominated action, darf-inquiries prefer a simple go-ahead. We see this when, a few moments later, the lights turn green and the car begins to move. Talk up to line 6 in Extract 14 (which is not a deviant case) is still occupied with banter about a previous outing. As the car approaches the spot where Carola was going to turn the car around, the multiactivity of driving and talking leads to disfluencies in her talk (lines 3–4) (Mondada 2012), until she finally abandons her in-progress TCU to produce a darf-inquiry: darf ich jetzt hier u-turnen (‘may I U-turn here now’ [lines 4–5]).

Extract 14

‘may I U-turn here now’ PECII_DE_Car_20171031_1_411340

Carola’s darf-inquiry comes at a moment when the nominated action is already in progress and it is all but too late to change it. Clearly Carola’s modal inquiry at this moment in time cannot lead to elaborate consideration, and her selection of darf shows that Carola only seeks the others’ approval so as to realize the action as okayed by all. The fact that Valeska in next position again points to the illegality of this maneuver in terms of the traffic rules (nie, ‘never’ [line 7]) does not change the course of events as Carola begins to turn the car around.

As a final case, we turn to a darf-inquiry that puts a challenge to our analysis that these are selected in sequential positions where the nominated action can be heard as the solution to a known-in-common problem. During a family breakfast (the same one as in Extract 9), two-year-old Emma is “playing” with a small piece of cheese, and her dad, Florian, watches her. She also has an untouched buttered bread roll on her plate. At line 2, Florian stretches out his hand to take hold of the bread roll, and he simultaneously launches a darf-inquiry: darf papa mal abbeißen hier (‘may dad take a bite off here?’). Emma blocks this action (line 3), and Florian retracts his hand as he registers his daughter’s refusal (line 4). After this, he goes back to monitoring Emma’s conduct.

Extract 15

‘may dad bite off here’ PECII_DE_Brkfst_20160424_18:35

There does not seem to be a publicly shared problem that Florian is experiencing in the run-up to his inquiry, to which taking a bite of Emma’s bread roll would be a possible solution. It rather seems that Florian just “wants” a bite of the bread roll. However, closer inspection reveals additional layers to Florian’s inquiry. Note that Florian monitors Emma’s conduct during the seconds leading up to his modal inquiry, and also that he goes back to this straight after its failure. These details might already be a sign that his inquiry does not spring simply from an appetite for bread roll. In fact, looking back in the interaction, this is not the first time Florian tries to get Emma to eat it. Some two minutes earlier (16:42), when Emma suggested she would lick the butter off the bread roll, Florian rejected this idea and instructed her to “bite off” some bread roll instead (nicht die butter ablecken, du sollst was abbeißen, ‘don’t lick off the butter, you are supposed to bite something off’). As it happens, Sarah does go on to lick butter off the bread roll, and so, 30 s later (and a good minute before our Extract 15), Florian intervenes: Emma, du sollst was abbeißen davon, nicht nur ablecken (‘Emma, you are supposed to bite something off that, not just lick it’). If we now go back to Extract 15, it is noteworthy that Florian formulates his nominated action (abbeißen hier, ‘bite off here’ [line 2]) in the same way as he had in his earlier instructions to Emma, when other formulations (such as the nominal einen biss haben, ‘have a bite’, or alternatives such as was abhaben, ‘have some’, or mal probieren, ‘try’) would have been available, and possibly more common – if the matter at hand was a request to share some food. These details of formulation suggest that Florian’s bid to bite off a piece of bread roll was instead a (maybe too subtle) parental ploy to finally get Emma’s eating of the bread roll off the marks. Note also that Florian here refers to himself in his parental role as papa, ‘dad’, while elsewhere in the interaction he refers to himself pronominally when talking to his daughter (Schegloff 1996). In sum, it seems that Florian’s selection of darf does carry an orientation to the nominated action’s connection to a known-in-common problem: that the bread role must finally be eaten (even though, of course, Florian taking a bite could at most begin to solve the problem that his daughter is making no progress in her food intake). Ultimately, this case supports our generalization that darf-inquiries place the nominated action in the context of a problem that is in common ground.