1 Introduction

This study examines question-word interrogatives followed by the vocative-based markers βρε [vre] and μωρέ [moré] in Greek talk-in-interaction, bringing together Interactional Linguistics and Construction Grammar. In this introduction, I offer some theoretical preliminaries on constructions and their relationship to discourse, and I contextualize my research topic within the framework of constructional and interactional approaches to grammar.

Construction Grammar (CxG henceforth) is a set of diverse approaches (including, i.a., Cognitive Grammar, Langacker 1987, 2008; Radical Construction Grammar, Croft 2001; and Cognitive Construction Grammar, Goldberg 1995, 2006) that share some basic assumptions about the representation of grammatical knowledge as constructions (see Imo 2015 for a detailed discussion). More specifically, constructions are understood as “conventionalized clusters of features (syntactic, prosodic, pragmatic, semantic, textual, etc.) that recur as further indivisible associations between form and meaning” (Fried 2015: 974). These symbolic units or stored form-function pairings encompass morphemes, words, and grammatical structures, may have non-compositional properties or may be highly entrenched, display varying degrees of schematicity (ranging from fully lexically filled and fixed to fully schematic), and form a network or ‘structured inventory’ (Langacker 1987: 63–76). A number of scholars, including Fried & Östman (2005), Imo (2015) and Fischer (2010, 2015), have argued that the cognitive explanation of grammatical structures endorsed by CxG is compatible with usage-based approaches such as Conversation Analysis (CA henceforth) and Interactional Linguistics, which examine interactional phenomena.

Interactional Linguistics is an inductive, empirically grounded, CA-informed approach to the study of interactional language use that draws on data collections of fully transcribed audio- or video-recorded naturally occurring talk (e.g., everyday face-to-face conversations, telephone calls, institutional talk) that comprise several occurrences of the phenomenon under study (Couper-Kuhlen & Selting 2018). In particular, Interactional Linguistics examines how speakers mobilize linguistic resources (i.e., the lexico-semantic, morpho-syntactic, and prosodic means) as practices in specific sequential contexts (i.e., courses of action that consist of adjacency pairs) to implement actions,¹ such as proposing, assessing, accepting or rejecting, agreeing or disagreeing, etc. Interactional linguists aim to uncover “the recurrent ways” in which linguistic forms are deployed (Couper-Kuhlen & Selting 2018: 29) in real conversational time and to reconstruct the participants’ procedures and methods of action formation and ascription (Levinson 2013). In other words, both interactional linguists and construction grammarians share an interest in describing observable linguistic patterns, that is, surface-oriented, form-meaning pairs that operate as “instruments of interaction” (Fillmore 1989: 32) or “routines for the fulfillment of interactional purposes” (Imo 2015: 70).

Additionally, both CxG and CA/Interactional Linguistics adopt a non-reductionist and ‘maximalist’ approach to language, taking into account form, meaning and function (Fried & Östman 2005: 1755). For example, Fried & Östman (2005: 1753) point out that CxG’s notion of construction, which goes beyond the well-formed sentence and includes “all linguistic constructs (tokens) of a language”, is close to CA’s notion of turn-type or turn-constructional unit (TCU henceforth). Turns, namely “utterances that speakers produce when they occupy the floor” (Couper-Kuhlen & Selting 2018: 34), are composed of TCU s, namely syntactic units (clauses, phrases, lexical items) that are delivered with a distinct intonational “packaging” and “constitute a recognizable action in context” (Schegloff 2007: 4).

Alongside these similarities, there are also some differences in the analytical methods used by the two frameworks. For example, CxG uses formalization to express generalizations about constructions as conventionalized “abstract grammatical patterns distinct from concrete linguistic expressions” (Fried & Östman 2005: 1756), whereas CA/Interactional Linguistics resists the idea of constructions as fixed and stable form-meaning pairs. However, Fischer (2015: 574, 575) argues that such reservations against CxG “may in fact be more due to a misunderstanding of the positions of CxG than a theoretical incompatibility”, given that many practitioners of CxG emphasize the dynamic and context-sensitive nature of grammar (see, e.g., Langacker 2008: 217), and show that variation is inherent in dialogic constructions (see, e.g., Nikiforidou et al. 2014).

The close relationship between grammar and discourse/pragmatics has been examined by a number of studies within the CxG framework (see, e.g., Alvanoudi 2025; Antonopoulou & Nikiforidou 2011; Auer 2005; Fischer 2010, 2015; Fried & Östman 2005; Günthner 2011; Imo 2015 and references therein; Nikiforidou et al. 2014; Terkourafi 2010; Wide 2009; Zima & Brône 2015). Some of these studies (Fischer 2010, 2015; Imo 2015; Zima & Brône 2015) have spelled out the benefits of synergies between the constructional and interactional approaches to grammatical structures. More specifically, Fischer (2010, 2015) puts forward the hypothesis that phenomena of spoken language use satisfy the ‘defining criteria’ of grammatical constructions and proposes that a construction grammatical approach to interactional phenomena offers insights and generalizations that a purely pragmatic or CA-based approach may leave unnoticed.

To illustrate her point, Fischer (2010) discusses the use of various pragmatic markers, like ah, oh, yeah, yes, okay, in turn-initial position before but-clauses that acknowledge the prior speaker’s turn, and observes that these markers carry the pragmatic meaning of ‘signaling successful perception and understanding, acceptance of contribution, continued attention/solidary interpersonal function, and continuation of the same topic’.² Fischer (2010) argues that the functional similarities observed across the different pragmatic markers cannot be predicted from the lexical items used, but derive from the sequential structure in which these markers occur. Namely, the meaning of the sequence is to provide evidence of acknowledgement and successful understanding of the interlocutor’s turn, to signal topic continuity, and to mitigate a possible face-threat related to the content of the following but-clause. Fischer (2010) suggests that the sequential structure itself (i.e., turn-initial pragmatic marker plus a but-clause in response to the interlocutor’s turn) can be treated as a grammatical construction, namely an identifiable form-meaning association (see also Fischer 2015 on oh-plus-assessment after informings³ as a form-meaning pair). Taking a grammatical perspective on several instances of various pragmatic markers in the same structural context offers a generalization about the nature of the construction and demonstrates why CxG “seems to provide a unique opportunity to consider interactional structures in grammatical terms” (Fischer 2015: 564).

In the same spirit, the present study explores question-word interrogatives followed by the vocative-based markers vre and moré⁴ in Greek conversation, following an integrated approach that draws on the tools of Interactional Linguistics and CxG. The analysis aims to decipher the functions of question-word interrogatives-plus-vre/moré, and to provide a construction grammatical account of these formats. The paper is structured as follows. In Section 2, I give an overview of the forms and functions of question-word interrogatives in Greek as well as the functions of the vocative-based markers vre and moré. In Section 3, I outline the data and method. In Sections 4.1 and 4.2, I analyze the interactional functions of question-word interrogatives-plus-vre/moré. In Section 4.3, I propose that these interrogatives instantiate the ‘beyond information seeking’ construction and the ‘challenging’ construction, and discuss the ways in which Interactional Linguistics sheds light on their meanings. In Section 4.4, I give a construction grammatical account of the challenging construction and discuss the benefits of taking a grammatical perspective on this interactional phenomenon. Concluding remarks are in Section 5.

2 Question-word interrogatives: Forms and functions

Wh‑ or question-word interrogatives (QWI s henceforth) consist of ‘question words’, typically in clause-initial position, that request information about the person or other entity (‘who’ or ‘what’), time (‘when’), place (‘where’), manner (‘how’), and reason (‘why’) of a given state of affairs (Couper-Kuhlen & Selting 2018: 220). In Modern Greek, question words include the interrogative pronouns and determiners ποιος/ποια/ποιο [pços/pça/pço] ‘who; which’ and πόσος/πόση/πόσο [pósos/pósi/póso] ‘how many; how much’, which are inflected for gender, case and number, τι [ti] ‘what’,⁵ which is uninflected, τίνος [tínos] ‘whose’, used for the genitive singular forms of ποιος and ποιο, as well as the adverbs πότε [póte] ‘when’, πού [pu] ‘where’, πώς [pos] ‘how’ and γιατί [ʝatí] ‘why’ (Holton et al. 2012: 117–118, 506).

QWI s typically seek information as their main action and presuppose a steep epistemic gradient between the questioner and the respondent (Heritage 2012). Namely, in requesting information, the questioner positions her/himself as the unknowing [K-] party and the respondent as the knowing party [K+] (Couper-Kuhlen & Selting 2018: 218). An information-seeking QWI anticipates a response that provides the information requested by the prior QWI, i.e., an informing, or a response that claims inability or unwillingness to provide the requested information, i.e., a non-answer or disclaiming (Couper-Kuhlen & Selting 2018: 232–237; Fox & Thompson 2010; Thompson et al. 2015). QWI-informing and -disclaiming sequences in Greek are illustrated in Extracts 1 and 2, respectively.

Extract 1

Extract 1 is taken from a telephone call between Katerina and her friend Chrisa. After the summons-answer and how-are-you inquiries (which are omitted in the transcript), Katerina uses a polar interrogative (i.e., a yes/no question coded via final rising intonation) to request information about Chrisa’s location at the time of the telephone call (sto máθima íse? ‘Are you in class?’) at line 01, and Chrisa emphatically confirms (ne:. ‘Yes.’) at line 02. The polar interrogative-answer sequence is followed by a QWI-answer sequence at lines 03 and 05, in which Katerina uses a QWI to seek information about the time the class ends (ti óra teʎónis? ‘What time do you finish?’, line 03), and after a gap (line 04) Chrisa uses a phrase to provide the requested information (katá tis eftámisi¿ ‘Around half past seven¿’, line 05). At line 06, Katerina registers information receipt with the particle a:. ‘O:h.’ and proceeds with a proposal for joint action (lines 06–07).

Extract 2 is from the middle phase of a phone call between Adriani and her friend Nasos. In the previous turns omitted in the transcript, Nasos proposed that they go to the movies, Adriani accepted the proposal and they discussed the time of their meeting.

Extract 2

At line 01, Adriani requests information about the movie they will watch via the QWI ↑pu θélis na páme? (‘Where do you want us to go?’), which is followed by an add-on TCU continuation (se pça tenía?, ‘To which movie?’) that replaces part of its host TCU and operates as same-turn self-repair (Couper-Kuhlen & Ono 2007). At line 02, Nasos does not provide the requested information (ðen ksé:ro¿ ‘I don’t know¿’), and at line 03, Adriani continues with a proposal for joint action.

Questions are turn-types with a “double duty”, in which the request for information often serves “as the vehicle or instrument for another action” (Schegloff 2007: 169). More specifically, QWI s are not always interpreted as information-seeking questions. For example, they may seek information in the service of other-initiated repair to address problems of hearing or reference (Couper-Kuhlen & Selting 2018: 158–159, 163–166) and implement invitations and proposals (Couper-Kuhlen 2014) or challenges (Koshik 2003; Pavlidou & Karafoti 2015) that do not anticipate informing as an answer. More on this in Section 4.

In this study I focus on the Greek QWI s followed by the vocative-based markers vre and moré. These markers are described as exclamatory words (Dictionary of Standard Modern Greek 1998; Holton et al. 2012: 351) and belong to the class of familiarizers, namely conventionalized vocative forms that denote simply a ‘person’ (commonly male), such as man, dude, mate in different varieties of English and hombre, huevón, güey, che in different varieties of Spanish. Familiarizers have a general meaning that does not allow them to identify a single addressed recipient and operate as discourse markers with interpersonal or textual functions (e.g., Kiesling 2004; Kleinknecht & Souza 2017; Rendle-Short 2010). The Greek familiarizer vre originates in the vocative moré, that is, the vocative singular form of the grammatically masculine adjective μωρός [morós] ‘stupid’ (see Joseph 1997; Tzitzilis forthcoming). According to Alvanoudi & Papadamou (forthcoming), the diachronic development of vre from moré consists of early grammaticalization: adjective morós > vocative moré; and cooptation and late grammaticalization: moré > vre (cf. Heine et al. 2021). Both moré and its grammaticalized form vre are used as discourse markers in Standard Modern Greek (Alvanoudi 2024, 2026), namely as positionally variable items that are relatively syntactically independent of their environment and have a metatextual function (Maschler 2009); that is, they operate as textual markers that refer metalingually to the realm of the text, as interpersonal markers that refer to the interaction among its participants, and as cognitive markers that refer to the cognitive processes of the participants.

Conversation analysis of audio-recorded naturalistic data has shown that vre and moré, when used on their own (i.e., when they do not precede other vocative forms like proper names),⁶ commonly occur in TCU-final position and less often in TCU-initial or -medial position, and fulfill various pragmatic functions (Alvanoudi 2024, 2026). More specifically, in TCU-final position, vre and moré operate as interpersonal markers that signal friendliness and solidarity in disagreement, in humorous turns, and at the beginning and closing of phone calls, among others. Also, in certain contexts, such as disagreement or answers to requests for information or confirmation, TCU-final vre and moré operate as cognitive markers that convey counter-expectation. In TCU-initial position, both vre and moré perform textual functions, such as attracting the recipient’s attention for the production of a multi-unit turn or shifting topic.⁷ Despite their parallel uses, the two items do not always occur in free variation in the same interactional contexts (Alvanoudi 2026). For example, moré fulfils interpersonal functions in a wide range of disagreement contexts, ranked from most to least face aggravating, that is, in challenges, contradictions and counterclaims (Muntigl & Turnbull 1998: 243; see also Nikiforidou et al. 2014 on the dismissive challenge performed by the éla moré construction). Unlike moré, vre downplays the face threat carried by the most aggravating forms of disagreement, namely challenges and contradictions. This indicates that vre is a stronger marker of solidarity than moré. Some of the functional similarities and differences between the two markers are illustrated by the analysis presented in Section 4.

3 Data and method

For the analysis of QWI s-plus-vre/moré in social interaction I draw on Interactional Linguistics (see Section 1). The data analyzed derive from 40 audio-recorded everyday face-to-face conversations and 145 audio-recorded telephone calls among friends and relatives from the Corpus of Spoken Greek of the Institute of Modern Greek Studies,⁸ and 6 hours of 14 audio-recorded everyday face-to-face conversations among friends and relatives from the author’s collection of naturally occurring data. The data were fully transcribed according to standard CA transcription conventions (see Appendix), and the speakers in the data were given pseudonyms. Written informed consent to record their conversations and to use the recorded data was obtained from all participants.

For this study I used vre and moré as ‘node words’ to search the data manually and I focused only on those that occur in QWI s. I compiled a corpus of the 40 occurrences of QWI s followed by the markers vre or moré found in the data, 25 followed by moré and 15 followed by vre, as seen in Table 1.

In 38 out of the 40 cases, vre and moré occur in TCU-final position (there is one occurrence of vre in TCU-medial position and one occurrence of vre in TCU-initial position). As seen in Table 2, in most cases, QWI s-plus-vre/moré implement disagreements/challenges (68.4 %), that is, responsive actions that “take up the action of an initiating action, and […] they are ‘typed’, i.e., they are specific to a particular type of initiating action that they are understood to address (Schegloff 2007)” (Thompson et al. 2015: 3). Less often, QWI s-plus-vre/moré carry out other actions (31.6 %), namely initiating actions that launch a new sequence, such as directives, word searches, or surprised initiation of repair.

Regarding their prosodic features, QWI s-plus-vre/moré are delivered as distinct intonation phrases held together by a coherent intonation contour (Szczepek Reed 2011: 43–44). More specifically, QWI s-plus-vre/moré performing initiating actions are always delivered with rising pitch, whereas QWI s-plus-vre/moré performing responsive actions are delivered with either rising (N = 15) or falling pitch (N = 11). Also, the former may include question words that are prosodically upgraded, i.e., uttered with greater intensity or at a higher pitch (in 5 out of the 12 cases), whereas in the latter there is typically some kind of prosodic emphasis either on the question word or on another item (verb, noun) used in the clause (in 17 of the 26 cases).

4 Analysis

The analysis is structured as a sort of ‘exercise’ in integrating Interactional Linguistics and CxG for the purpose of exploring QWI s-plus-vre/moré in Greek talk-in-interaction. The first part of the analysis targets the interactional functions of interrogatives followed by vre or moré when performing an initiating action (Section 4.1) or a responsive action (Section 4.2). The second part of the analysis opens up insights into the ways in which Interactional Linguistics contributes to our understanding of the forms and meanings of the QWI vre/moré constructions (Section 4.3). Finally, Section 4.4 provides a construction grammatical account of the challenging QWI s-plus-vre/moré and a discussion of the insights we gain by taking a grammatical perspective on this interactional phenomenon.

5 Concluding remarks

To sum up, the analysis of QWI s followed by the vocative-based markers vre and moré in Greek conversation has shed light on their functions in different sequential positions. In first position, the QWI s-plus-vre/moré perform actions that go beyond information seeking (Section 4.1), whereas in responses to a previous speaker’s turn, they implement challenges and do not anticipate informing as a response (Section 4.2). As shown in Section 4.3, an interactional linguistic approach can help us to identify a network of related constructions (cf. Nikiforidou et al. 2014), namely the ‘beyond information seeking’ and the ‘challenging’ [QWI] vre/moré constructions, and to decipher their meanings which derive from their use as practices for performing specific actions in specific sequential slots, “built to fit the particular location and occasion of their use” (Couper-Kuhlen & Selting 2018: 22). Moreover, as demonstrated in Section 4.4, by integrating spoken language data into a constructional description, we can make generalizations that go beyond individual speakers’ practices and identify the structural context as the source of the constructional meaning associated with the ‘challenging’ QWI vre/moré construction. The latter seems to participate in a ‘larger’ grammatical construction, which consists of a sequence, namely a QWI-plus-vocative-based marker following a prior speaker’s assertion about a state of affairs of mutual knowledge, and it is associated with the meaning of ‘reactive challenge, counter-expectation to the speaker, solidary interpersonal function’.

By taking an interactional linguistic and construction grammatical perspective on spoken language interaction, this study contributes to the growing body of literature on the relationship between grammar and discourse/pragmatics (see Section 1) and on the benefits of bringing together the cognitive and interactional approaches to grammatical structures. More specifically, the analysis provides further support for Fischer’s (2010, 2015) insightful hypothesis that spoken language phenomena fulfill the ‘defining criteria’ of grammatical constructions. This allows both interactional linguists and construction grammarians to interpret interactional structures as symbolic units that encode meaning, constitute “cognitive routines” that are established through “the progressive entrenchment of configurations that recur in a sufficient number of events” (Langacker 2008: 220), and form part of grammatical knowledge. In line with Fischer (2010, 2015), I propose that an integrated usage-based approach, like the one taken in this study, can enhance our understanding of the relationship between grammatical and interactional knowledge by offering insights that a purely interactional linguistic approach or a purely construction grammatical approach may miss.

I now turn to the shortcomings of the analysis, which can help us identify promising areas for future research. The methodological tools of a qualitative interactional linguistic approach offer valuable insights into the meaning of constructions (Section 4), which can be further strengthened by quantitative approaches to CxG. The analysis of quantitative data derived from larger corpora annotated for morpho-syntactic, prosodic and discourse-related features will allow us to analyze frequencies and recurrent patterns of use and to establish positively the existence of constructions.

Furthermore, the present study was restricted to QWI s followed by two vocative-based markers, vre and moré. The analysis of the QWI followed by the marker re in Extract 9 indicates that other items in TCU-final position can give rise to similar affective and epistemic interpretations of the interrogative clause. Whether QWI s occurring with other vocative-based markers, such as re, μωρή [morí], μαρή [marí], and καλέ [kalé], encode the same constructional meaning carried by QWI s-plus-vre/moré remains to be determined in the future.

As mentioned in Section 3, QWI s-plus-vre/moré in first position are always delivered with rising pitch, whereas QWI s-plus-vre/moré in responses show prosodic variation, as they are delivered with either rising or falling pitch. Also, in the latter case, question words or other linguistic items often carry some kind of prosodic emphasis. These preliminary observations indicate that the [QWI] vre/moré constructions may be multimodal constructions (Lehmann 2024). In-depth analysis, drawing on both corpus-based and experimental evidence, can unravel the prosodic resources that recur with the [QWI] vre/moré constructions, and determine whether the latter inherit their prosodic properties from Greek QWI s and exclamatives, or whether they are prosodically constructed as different kinds of objects (see, e.g., Nikiforidou et al. 2014 on the prosodic properties of constructions).

Last, the use of pragmatic/discourse markers is known to correlate with social variables, such as gender and age (Fedriani & Sansò 2017: 20–22). Future research can shed light on the contextual properties of the meaning component of QWI s-plus-vre/moré, for example by exploring whether the use of vre and moré correlates with the speaker’s gender, age and dialectal background or exclusively with informal style.