GENERAL EXTENDERS IN SPANISH INTERACTIONS: FREQUENT FORMS, PRAGMATIC FUNCTIONS Y TODO ESO

The majority of Spanish discourse analysts do not consider general extenders as a type of discourse marker (a position this paper adheres to) and as a result, not many studies have been devoted to determining their characteristics. This paper presents a description of the principal forms, semantic nature and discourse functions performed by general extenders in European Spanish. The analysis is based on a large amount of data (3 700 000 tokens from two spoken corpora, Corpes XXI and Val.Es.Co. Corpus) and considers different text types and discourse genres. The research focuses on two types of pragmatic functions attributed to general extenders that are the result of semantic bleaching, i.e., they have lost their original referential meaning in order to highlight common ground in interactions: interactive functions (politeness, hedging, intensification) and metadiscourse functions (delimitation, marking transitional relevance places and forward planning). Spanish data support the hypothesis that strong cross-linguistic similarities exist at both a formal and functional level. Further lines of research in the field of Spanish linguistics are highlighted in the final considerations.


GENERAL EXTENDERS: DEFINITION AND FUNCTIONAL DESCRIPTION 1
Semantic vagueness is associated with speech production in interactional contexts, especially in cases of familiarity between interlocutors. One particular type of vague element identified by linguists working in the field of pragmatics and conversational analysis are the so called "general extenders" (Overstreet 1999). General extenders (henceforth GEs) are sentence -or clause-final constructions, i.e., appended to a phrase (NP, AdjP, AdvP, or PrepP as there are no restrictions on the type of phrase) or to a clause with referential meaning. Some examples of Spanish GEs such as y tal and y eso can be observed in (1-2): (1) // mmm en una de esas entrevistas que están bueno en el libro / eh eh entrevisté a un señor en un pueblo pequeño / de la provincia de Cuenca / y la señora no hablaba ¿no? // la señora estaba en un segundo plano como escuchando y tal y no dijo absolutamente nada pero yo le preguntaba pero ella no sé / incluso llegué a pensar que era muda (Corpes XXI, 2016, RTVE, Milenium) 2 'umm in one of those interviews that are, well, in the book / err err I interviewed a man in a small town / in the province of Cuenca / and the lady didn't speak, did she? // the lady was in the background as if she were listening and everything but said absolutely nothing, and I did ask her but she, I don't know / I even started to think she was mute' (2) S: ¡aah! / yo tengo una amiga que se ha montao una pi(s)cina pequeñita↑/ qu'es para los bebés↑/ pero / a(ho)ra / aparte d'eso se han apuntao / personas mayores /// bueno / pequeña / diecisiete metros de pi(s)cina /// (3") climatizada [y eso] (Val.Es.Co. Corpus, AP.80.A.1, turnos 301-304) 'S: aah! / I've got a friend who's made a tiny swimming pool↑/ you know, for babies↑/ but / now / as well / older people have joined in /// well / small / seventeen meters of pool /// (3") heated [and everything]' The referential phrase or clause to which a GE is appended is called an "operand" (Dubois 1992: 181) or "anchoring constituent" (Ward & Birner 1993: 208). The anchoring constituent is usually a list or narrative sequence, and the GE constitutes a closing element in enumeration or narration, a value that is usually present in lexicographic definitions (see Montañez Mesas 2008: 194-195 for an overview of GE definitions in Spanish dictionaries).
GEs are also present in written language, but they are far less frequent and almost restricted to etcetera, especially in texts situated near the communicative Consequently, I agree that GEs should be regarded as discourse elements that do not belong to the same class as DMs.
GEs have been comprehensively described in languages such as English (mainly and stuff and and everything), 6 French (Andrews 1989, Dubois 1992, Secova 2014, German (Overstreet 2005, Cutting 2015, Italian (Mauri 2014, Fiorentini & Sansò 2016, Mauri et al. 2019) and other non-European languages, but the number of studies exclusively devoted to the study of Spanish GEs is still very limited (Guil 2000;Domínguez Mujica 2005;Cortés Rodríguez 2006a, 2006b, 2006cMontañez Mesas 2008;Gille 2006;Gille & Håggkvist 2006Fernández 2015). 7 Studies on Spanish GEs have proposed different taxonomies of the functions fulfilled by these elements. Gille (2006) and Gille & Håggkvist (2006) encompass GEs in a wider functional class that they call apéndices conversacionales whose main function is to mark that a conversational unit is complete from a syntactic, pragmatic, and prosodic perspective. They form a subset called apéndices de categorización generalizada and codify an instruction for the interlocutor to look for the general category to which the previous elements belong. As can be seen below, this category is not necessarily pre-existent but dynamic, particularly relevant to the given situation, an idea that is also anticipated by Gille (2006: 160). A consequence of being an extension of the reference is to lessen the relevance of the anchoring constituent which is then viewed as just one element among others belonging to a general category (Gille & Håggkvist 2006: 67). The aforementioned authors also highlight the function of GEs as vague elements that help in cases of difficulties in speech forward planning, fluidity, or lexical competence.
In their study delimited to disjunctive GEs (see below), Gille & Håggkvist (2010) propose the following functions: a) delimiting conversational units, for example units conveying different speech acts, or types of discourse, for example the narrative frame and reported direct speech; b) mitigating by offering alternatives and not imposing a claim or a request; c) approximation, a discourse strategy to compensate for a lack of knowledge or a problem with speech fluidity; d) the so called rhetoric function: a persuading strategy in argumentation by which the speaker refers to arguments that support their claim but without specifying, and e) intensification. 6 Dines (1980) could be considered a pioneer work for English language in a sociolinguistic framework. Overstreet (1999), by adopting a pragmatic and discourse perspective, has become one of the leading authors in this field after publishing her PhD and many other studies. The research on GEs in different varieties of English and in contrastive perspective with other languages has increased exponentially over the last twenty years. 7 In Spanish, some of these units have been characterized as phraseological units by experts in the field. See the studies by Alvarado, Corpas, and Ruiz Gurillo quoted in Montañez Mesas (2008). However, they are not studies focusing on GEs but on formulaic constructions. Cortés Rodríguez (2006aRodríguez ( , 2006b proposes an exhaustive list of functions for Spanish GEs. He divides the functions into two groups: interactive and textual. Interactive functions are classified as follows (Cortés 2006b): a) functions focusing on the interlocutor: shortening the message (to avoid superfluous information), indicating mutual understanding, and b) functions relevant for the introduction of a discourse topic: intensification, mitigation, showing doubt, uncertainty, or contempt. Regarding the textual functions (Cortés 2006a), he then distinguishes between primary functions and secondary functions. The primary functions are categorial implication (see also Fernández 2015) and set-closing, while the four secondary functions are avoiding excessive explicitness (shortening the message), mitigation, three-part list-construction, i.e., the tendency to use tripartite lists and series in discourse highlighted by Jefferson (1990) in her study on list constructions, and filling in for speech forward planning.
These functional descriptions are interesting, but they lack a systematic approach in order to distinguish core functions and contextual functions, i.e., functions that may arise in particular contexts but are not especially frequent or representative of these units (such as showing contempt or making a request, see Fernández 2015). The aim of this paper is to describe the pragmatic functions of elements that generally function as GEs in European Spanish by establishing their most representative and essential functions. Discourse functions are usually classified in three macro functions -argumentative, interactive and metadiscourse-but only the latter two are relevant for the description of GEs. Both will be considered here in a more essentialist way as is the case in the literature to date, in order to derive the pragmatic functions from the semantic description of GEs, correlate functions and types, and highlight differences and similarities with previous descriptions in other European languages. The description is based on the analysis of two spoken language corpora which will be briefly introduced below (Section 3). The paper opens with a presentation of the formal characteristics of GEs cross-linguistically ( § 2). Some reflections upon their semantic nature ( § 4) precede the description of five different pragmatic functions performed by Spanish GEs: politeness, hedging and intensification, which are considered interactive functions ( § 5), delimiting information chunks, marking transitional relevance places, and helping forward discourse planning, which are considered metadiscourse functions ( § 6). The final sections are devoted to identifying the presence of GEs in reported speech ( § 7) and to highlighting some differences with GEs in other languages ( § 8).
Studies on language vagueness have associated this type of discourse strategy with specific social classes (Dines 1980) and age groups, mainly adolescents and young adults (Winter & Norrby 2000, Norrby & Winter 2002, Secova 2014. Although more research is needed from a sociolinguistic perspective, I have not observed any strong correlation between GEs and youth language given that GEs seem to be pervasive in all age groups and social classes in specific communicative situations, i.e., informal interactions (cf. Guil Povedano 2000: 587). However, given that sociolinguistic issues are not dealt with in this paper, further research should be performed to determine whether there are any preferences for particular GEs in specific classes, dialects, or age groups.

FORMAL CHARACTERISTICS 8
GEs are prototypically introduced by a copulative or disjunctive conjunction and accordingly classified in adjunctives (Sp. y tal, y todo eso; Eng. and everything) and disjunctives (Sp. o algo; Eng. or whatever). 9 In some languages adjunctives, GEs can also appear without a conjunction (Sp. tal, esas cosas, It. cose così, Fr. machin), although the latter seems to be less frequent than the other two. 10 The core of GEs can be simple or complex. Simple GEs are formed by nouns (Eng. stuff), indefinites (Eng. everything; Sp. algo; It. altro), quantifiers (Fr. tout), demonstratives (Sp. eso, tal; Eng. that), adverbs (Eng. so; Ger. so; Sp. así) or interrogatives (Eng. what; Ger. was; Sp. qué). All these elements have in common a non-referential meaning: they function as pro-forms and their meaning is determined by previous information in the same utterance to which they refer as a kind of associative anaphors.
Complex GEs are formed in one of the following ways: lo que sea, qué sé yo) or clauses with a general noun or a hypernym in the object position (It. Dio sa cosa, Sp. Dios sabe qué más). Elements in such clauses have a certain freedom to change positions: Sp. y qué sé yo, y yo qué sé; y Dios sabe qué más, y sabe Dios qué más. Moreover, in all these languages the Lat. GE etcetera (originally an adjunctive GE: conjunction + indefinite) has maintained this function, sometimes with an adjectival modifier (Sp. y un largo etcétera). Secova (2014: 282, 284) argues that shorter GEs seem to be more grammaticalized than longer ones. According to this author, due to frequent use Fr. et tout or Eng. and stuff show a higher degree of grammaticalization than Fr. et tout ça and Eng. and stuff like that. In Spanish, a similar case is represented by y tal which is far more common than other longer variants such as y tal y cual. However, further historical research is needed to verify that short forms are not attested earlier or simultaneously to longer forms.
Although it is not very frequently the case, GEs can be combined as in (3-4) or repeated as in (5). Combinations and repetitions have a specific pragmatic function, as can be seen below (Section 4.3): ( Regarding their position, GEs typically appear appended to a phrase or clause as in previous examples, but it is also possible to find GEs interrupting a NP, as y tal in (6) inserted in the NP fotocopias de artículos (photocopies of articles). Some authors have observed that GEs are becoming more flexible in their position, and this has been interpreted as a feature of grammaticalization (Overstreet 2005 = see↓ what I wanted was to make some photocopies and stuff of articles / but// they won't be any use to me either, will they?' In Spanish, there is a third type of GE that we might call "reported discourse extender" because it occurs when a speaker reports speech from other speakers and always indicates an extension of discourse and not of a single set or list, as in (7): ya nos metemos / y el recepcionista casi sin hablarnos / pues mi marido le explica que tenemos allí la reserva // y patatín y patatán // y cuando volvemos a coger las maletas // para irnos a la habitación // pues nos giramos (Corpes XXI, 2001 CORALES Buenos días, 05/04/01, RNE, Radio 1) 'we went in / and the receptionist almost without speaking to us / well, my husband explained to him that we had a booking there // etcetera etcetera // and when we went back to pick up our suitcases // to go to our room // well, we turned around' There are three main types of reported discourse extenders: a) an adjunctive extender introduced by a copulative conjunction and a noun which is meaningless but whose phonetic structure based on repetition of phonemes evokes a long discourse (Iglesias & Aliaga 2018), especially when the noun is repeated twice with slight variants as in (7): y patatín y patatán. Onomatopoeic elements may play the same role: y bla bla bla. As in the case with other adjunctives, reported discourse extenders can also appear without a conjunction (patatín, patatán; bla bla bla). 12 b) a subordinate clause introduced by a saying verb (not necessarily adjacent to the clause) and a complementizer (que), as is usual in indirect reported speech, followed by a conditional conjunction (si) and either the same type of noun as in (a) or a demonstrative (esto, tal). This structure is often repeated with a different demonstrative (aquello, cual) as in (8) or a second noun or demonstrative is added after a coordinative conjunction (que si esto y aquello, que si tal y cual). 11 The question of the position of GEs deserves a more in-depth study that will not be approached here. A model of discourse segmentation (for example, Val.Es.Co. 2014) seems crucial in order to offer a systematic description and correlate functions and positions. Some steps in this direction can be found in Cortés (2006a) and Montañez Mesas (2012). 12 Repetition, particularly morpheme repetition, is one of the linguistic mechanisms used in different languages to form GEs and, in general, to mark ad hoc categories, a concept explained in § 4 (see Mauri 2014 no // ellos te enseñan / y tú te tienes que ceñir a lo que te enseñan si haces algo distinto // a ver / juegan también con el que / tú no sabes / es decir tú empiezas y realmente conoces poco te dicen que si la textura de esto no es la misma que si esto que si aquello el plato es el mismo (Corpes XXI, 2014 PRESEGAL: SCOM_H11_051) 'no // they teach you / and you have to stick to what they teach you if you do something different // let's see / they also play with what / you don't know / in other words you start and you really know little they tell you that if the texture of this is not the same this, that and the other then the dish is the same' c) a clause with an epistemic verb introducing an interrogative pronoun or two clauses of this type with a strong parallelistic structure: no sé qué, no sé cuántos. The interrogative pronoun may contain a diminutive suffix: no sé cuantitos. These clauses can also be prefaced by coordinating and disjunctive conjunctions.
Adjunctive and disjunctive GEs can also be used in reported speech contexts. However, what I call "reported discourse extenders" rarely occur outside these settings.
The prosody of Spanish GEs has not yet been empirically described, but some scattered observations can be found in the main dictionaries of discourse particles: there is no relevant pause or inflexion between the last anchoring constituent and the GE, but the last is normally followed by a pause (Santos 2002, Fuentes 2009). Generally, GEs have falling intonation and are pronounced in a speedy rhythm when compared to previous elements (Cortes 2006a). They do not appear independently of their anchoring constituent.

DATA
To determine the most frequent GEs and their functions in European Spanish, data were collected from two different spoken language corpora that comprise the corpus of my research: a) the Val.Es.Co. Corpus, compiled in the late nineties in the city of Valencia by a group of researchers at the University of Valencia coordinated by Antonio Briz (Briz & Val.Es.Co. 2002). It contains 19 spontaneous conversations (93 000 tokens) in informal contexts (among friends or family members) registered with a hidden voice recorder to avoid the observer paradox. The advantage of a small corpus like this is that it is possible to read the whole corpus and to perform in-depth queries; b) the Corpes XXI (Spain) spoken language sub-corpus. A corpus compiled in the early twentieth century ( 13 I am aware of the limitations of conducting research by lexical queries, but for the purposes of this study, which focuses on establishing the most frequent forms and main pragmatic functions, the sample is sufficiently rich.
Many of these GEs can be prefaced either by the coordinating conjunction y or the disjunctive conjunction o. I have not marked this alternance in the following lists, but I have marked it in Tables 1 and 2 when present in the corpora.
14 Only three adjunctive GEs appear in the Diccionario de partículas del español: y tal (Montañez Mesas 2008-2022), y toda la pesca (Martí & Ruiz 2008-2022, y todo (Gras 2008(Gras -2022. Fuentes (2009: 366-368, 374) briefly describes y demás, y eso, y tal, y todo eso and (y) yo qué sé. The only function acknowledged is that of closing an enumeration and, consequently, they are considered as structural discourse makers. A more detailed semantic analysis can be found in Santos (2002). This author considers etcétera, y demás, y eso, y tal,y todo, y todo eso, (y) yo qué sé and the disjunctive o tal (the lexical entry for o algo refers back to algo but no further information is available). 15 Contrary to previous studies on Spanish GEs (cf. Gille & Håggkvist 2010), I do not consider o por ahí as a GE. It functions as an approximator, indicating that what has explicitly been mentioned may not be exact but without implicating that further information must be inferred, as can be seen in the following example: "C: § treinta y dos mil o por ahí [C: § thirty-two thousand or thereabouts]" (Val.Es.Co. Corpus, XP.48.A.1, 449-450). Some uses of o algo así have the same function and should be carefully distinguished from the GE. 16 These GEs have also been called marcadores de enumeración lexicalizada (Loureda 2002: 137). This author considers that some of these forms have a pejorative value (for example, que si tal, que si cual) but in my data I have not observed systematic differences in the subjective values conveyed by reported speech GEs. However, this type of GEs is relatively infrequent in both corpora (see Table 1 and 2) and my conclusion may be corpus biased. For a detailed analysis of syntactic, semantic, and pragmatic functions of this type of GE, see Aliaga & Iglesias (2011).
17 I consider only reported discourse extenders that have a certain degree of intersubjectivity, because the same function can be performed by meaningless repeated syllables: "a mí la primera vez cuando fui // pues me dice mi hija mamá que te dio dije yo bueno / tanto hablan tanto hablan y esto no hace nada fururú eh / muy valiente / tátara tatarata [the first time I went // well, my daughter said to me mum what did s/he give you I said well / so much talk so much talk and it  (Corpes XXI, 2010 PRESEGAL SCOM_M21_041). 18 GEs with no occurrences have been omitted. 19 In the case of y tal and y eso, I have not taken into account the variant without a conjunction due to the high number of occurrences in this corpus. As is the case in other languages, short GEs are more frequent than long ones (with one exception: o algo así vs o algo), which can be interpreted as a sign of a high degree of grammaticalization. 23 GEs seem to function as highly formulaic expressions and users treat them as units that do not require any further processing (Overstreet 2005(Overstreet : 1849

SEMANTIC DESCRIPTION
Semantic descriptions of GEs consider that they are semantically related to anchoring constituents for belonging to a set or list (Dines 1980, Gille 2006. "Extenders are expressions serving to extend the set of referents announced by the previous word or phrase, or by a group of words or phrases" (Secova 2014: 282). This explains tags such as list-completers or set-marking tags (Norrby & Winter 2002). Whereas the anchoring constituent has a referential nature, the GE expresses other elements of the sets without further specification. For example, in (9) the speaker lists two items that one might leave on the floor before taking a shower (dirty pants, socks) and the GE represents any other element that might be in this list (dirty shirt, trousers, t-shirt).
In using a general extender, the speaker is attempting category-implication, so that the hearer can infer additional or alternate instances of the category or set that the speaker has in mind. (Overstreet 2005(Overstreet : 1851 This reference to an implicit set or list of elements that may form a class in a particular communicative context is considered by some researchers as an instance of ad hoc categories construction (Mauri 2014, Sansò & Fiorentini 2016, i.e., no stable categories, highly context-dependent, and constructed spontaneously to achieve communicative goals. One or two members of this ad hoc category are explicitly mentioned but the category is not necessary a wellknown pre-established one. These expressions provide a way of talking about groups of entities or actions that spontaneously need to be referenced together when no established referring expression for the group is known. (Overstreet 1999: 43) Categories introduced this way in discourse are dynamic and relevant only for the interaction in progress (Gille 2006: 160) and it is assumed that the interlocutors will be able to fill in the missing categories (Fernández 2015: 6). 24 As highlighted by Overstreet (1999: 115), the GE can either suggest that there are other items that should be added to it (adjunctive GE), as in (9)  there are different alternatives belonging to this class but not all of them cooccurring at the same time (disjunctive GE) as in (10): the speaker talking about garnishing a cod-fish dish suggests peppers o algo, i.e., any other vegetable that could serve as garnish for the fish instead of peppers. In this case, the alternative is reinforced by por ejemplo (for example).
However, when the focus of analysis is shifted from the semantic relationship with the referential anchoring constituents to the pragmatic functions (Cheshire 2007), another more relevant semantic characterization comes to light. Corpus-based studies have shown that it is not always possible to complete the list and the GE does not always signal further elements of a welldefined or ad hoc constructed set (Winter & Norrby 2000), as in (11), especially when the anchoring constituent is not a NP, as in (12) speaker may have no additional items of a category in mind and thus prefer to use a vague expression.
These cases require a new semantic interpretation of GEs based on common knowledge shared by the speaker and the listener; common knowledge needed to complete the list (Guil Povedano 2000: 583, Cortés 2006a. This common ground allows the addressee to identify the information behind a GE by drawing on pragmatic information and not in set-internal semantic relationships. In fact, this information is not only category related as we have seen in (8), but according to Overstreet (1999), it can be related to shared perspectives or experiences rather than a piece of shared knowledge. This stock of knowledge is either socially derived or based on individual experience (Overstreet & Yule 2001: 48).
In contrast, GEs are very effective in communication as they manage to convey implicit information about experiences when the speaker assumes that they are already known by the addressee. This largely contributes to linguistic economy and optimal communication (Guil Povedano 2000: 581, Montañez Mesas 2008. Sometimes information can be too complex or too long to make it explicit and a GE functions as an instruction for interpretation. Therefore general extenders seem better suited to a function that is the opposite of one requiring us to construct 'a well-defined set'. They signal that further processing in referential terms is not required. (Overstreet 2005(Overstreet : 1851 In this respect, GEs have a procedural meaning and share a basic discourse function with DMs, i.e., they function as processing instructions guiding listeners to the most relevant interpretation of an utterance. That speakers use GEs based on implied assumption of shared knowledge rather than on its actual existence (Secova 2014: 289) is easy to see in cases like (13), where the speaker talks about a situation which may be unknown for the addressee: (13) eran los años de la movida // pero yo tuve un hijo prontísimo / y las mujeres de mi generación no tenían hijos en general y menos en un ambiente como el de la radio y todo eso / bastante cercano a la movida madrileña etcétera no tenían hijos tan pronto (Corpes XXI, 2007  In this interview, a famous writer is talking about her youth in the '80s when she worked as a journalist on the radio. She argues that having children was unusual in the kind of environment (la radio y todo eso) linked to the countercultural movement called La Movida Madrileña. However, not everyone in the audience would have had direct experience of working on the radio or of living in the '80s so the GEs y todo eso and etcetera cannot refer to actual knowledge but to the presupposition on the part of the speaker that the audience would be able to gather some information about those situations from other sources. She cannot be certain that the knowledge is effectively shared by her audience. By using a GE, the speaker invites her audience to behave as if their experiences were similar, whether or not they actually were. In this case, GEs behave like DM Eng. you know or Sp. sabes. In this regard, the role played by other epistemic marks in the same turn reinforces the stance adopted by the speaker: see imagino ('I imagine') in (11).
Thus, once again vagueness responds to communicative needs and is not a sign of laziness, uncertainty, lack of vocabulary, or linguistic incompetence, but a sign of how implicit meaning is referred to when a certain (effective or assumed) common ground is widely shared between participants. Speakers frequently resort to vague expressions in order to perform a wide range of pragmatic functions and it is not infrequent to combine several vague expressions in the same utterance: (14) oye ¿que a lo mejor me gustaría tener tres o cuatro trajes que tenía / uno / que tenía mm / uno para las fiestas para los domingos y otro para la semana? sí / pero bueno después en plan de vivir y eso / pues bien / bien / (Corpes XXI, 2009 PRESEGAL SCOM_M31_037) 'hey, maybe I'd like to have three or four suits that I had / one / that I had umm / one for the holidays for Sundays and another one for weekdays? yes / but well later like for living and stuff / well good/ good /'

INTERACTIVE FUNCTIONS
Pragmatic functions which have been described for GEs in other languages are also found in Spanish. In line with Overstreet's (1999Overstreet's ( , 2005Overstreet's ( : 1847 argument that GEs' functions are primarily interpersonal and not referential, I have classified the functions of general extenders into interactive and metadiscourse functions: politeness, hedging and intensification, on the one hand, and discourse delimitation and forward planning, on the other. 25

Politeness markers
GEs are considered cues of solidarity and positive politeness, which derives from the fact that they are used when there is effective or assumed shared knowledge between interlocutors. No misunderstandings arise from their use, as interlocutors never request clarification after a GE is used (Dines 1980). Interactions using GEs are, in general, very cooperative, and turn-taking with 25 Aijmer (1985) was the first to list different functions for GEs: set-marking, hedging, softening and approximation. Other authors have added functions to this list (see § 2). Here, I propose a division of functions in two macro categories, interactional and metadiscourse, and examine different functions within these encompassing categories. Functions are not mutually exclusive, i.e., the same GE can perform an interactional and a metadiscourse function in the same occurrence. GEs may trigger turn-taking with DMs to signal cooperation, agreement, and understanding (claro, normal, pues sí) or even turn-taking in co-construction as in (15).
(15) S: yaa-NO↓ voy a ir a un sitio que van un-unos amigos míos J: y ya está § S: § y así↑ pues/ con la excusa de que también van ellos↑/ pues [iré=] J: [mejor] S: = porque // como tenga que IR yo solo/ a un sitio quee está un poco lejos y tal↑ ya no voy J: claro /// [es mejor ir con gente] (Val.Es.Co. Corpus, AP.80.A.1, 568-575) 'S: not-NOW↓ I'm going to go to a place where a-some friends of mine are going. J: and that's it § S: § and so↑ well/ with the excuse that they're also going↑/ so [I will go=]. J: [better] S: = because // if I have to go by myself/ to a place that's a bit far away and stuff↑ I won't go The use of GEs presupposes a certain degree of complicity between interlocutors. GEs save time not only by not making explicit some information that the speaker assumes is already known or shared by the listener but also because explaining in detail what is deemed as presupposed knowledge may offend the addressee. In this respect, GEs are in line with Grice's maxim of quantity (try to be as informative as one possibly can, and give as much information as is needed, and no more; see Grice 1980).

Hedges
GEs also function as hedges to mitigate the directness of the utterance. This function is particularly evident in the case of disjunctive GEs which offer an alternative to a situation or fact (Gille & Håggvist 2010: 128). This is a case of negative politeness in which the speaker tries to save the addressee's face by presenting alternatives to what they have just said and not imposing a universal truth on the situation. When used to indicate a lack of commitment to something mentioned and to suggest possible alternatives, these forms may be seen as strategies of negative politeness. (Overstreet 2005(Overstreet : 1856 This is exactly what the doctor in (17)  It is not infrequent that GEs with this function are preceded or reinforced by other epistemic elements with a similar hedging function. This type of disjunctive GE is also very frequent in yes-no questions and requests. They contribute to mitigating their perlocutionary force by offering an alternative and not imposing an answer on the addressee. In my corpus, this function is usually performed by o algo and o algo así as can be observed in (18-20): (18) ¿tienes miedo a volar o algo? (Corpes XXI, 2007 PRESEGAL SCOM_H13_013) 'are you afraid of flying or something?' (19) y y el problema que me acabas de contar de la policía y todo eso / cuando estuviste en Inglaterra o en Barcelona ¿lo viste? / ¿y grabaste o algo? (Corpes XXI, 2007 PRESEGAL SCOM_H13_014) 'and and the problem you just told me about with the police and all that / when you were in England and Barcelona, did you see it? / and did you record it or something?' (20) ¿nunca hizo un curso de cocina o algo así? (Corpes XXI, 2007 PRESEGAL SCOM_H33_002) 'Did you ever take a cooking course or something similar?' There is, however, a difference between o algo ('or something') and o algo así ('or something similar'). O algo opens a wider range of alternatives because it marks a high degree of uncertainty. In (18)  possibilities that are presented as alternatives in questions and assertions, which, consequently, implies a lower degree of mitigation. The higher the vagueness and lack of commitment to the accuracy of the expression, the higher the degree of mitigation in discourse. 26 Another GE which offers alternatives is o lo que sea ('or whatever'). There is a subtle difference between o algo and o lo que sea with regards to the attitude of the speaker: o lo que sea suggests that there are many alternatives but in this case vagueness or lack of accuracy are regarded as unimportant to the speaker as s/he is not willing or able to name them. "The use of this general extender 27 seems to signal a dismissive attitude toward potential alternatives that the speaker may be unable to name" (Overstreet 2005(Overstreet : 1860. (21) y si o sea / el problema de de estar viviendo con con gente en un piso es que claro entras / en conflicto con / con cosas / o sea / a uno le toca fregar o que ha dejado la cocina un poco desastre o el cuarto de baño / cosas así / lo típico / o que / o que a lo mejor / está una persona se está duchando o lo que sea / y te tienes que / meter tú para ir a trabajar y no puedes (Corpes XXI, 2007 PRESEGAL SCOM_H13_013) 'and if, I mean / the problem of living with people in a flat is that of course you get / in conflict with / with things / I mean / you have to do the washing up or you have left the kitchen in a bit of a mess or the bathroom / things like that / typical / or that / or that maybe / someone is taking a shower or whatever / and you have to / shower to go to work but you can't.'

Intensifiers
The third interactive function attributed to GEs is intensification: to highlight the importance of an idea by presenting something notable, surprising or excessive, and thus intensifying the effect of the preceding phrase upon the hearer. (Secova 2014: 292).
This function is normally assumed by GEs with universal quantifiers (Fr. et tout; Eng. and all, and everything; Sp. y todo). Contrary to Eng. and everything, and all, and all that and Fr. et tout, that can also have politeness and hedging functions, Sp. y todo exclusively performs the 26 Overstreet & Yule (2001) highlight the structure not X or anything but as a metapragmatic structure: the speaker anticipates and neutralizes a possible interpretation of their behaviour. However, the formulaic character acknowledged in English (the structure is tagged as a formulaic disclaimer) has no equivalent in Spanish where the structure is very infrequent. I have only found one example in my corpus but without an initial negation: "seré muy pomposo o algo pero no no / no me gusta para nada / ponerme al sol / y el sol ya me molesta en los ojos una barbaridad tengo que ir con gafas de sol" ['It might seem very pompous or something but no no / I don't like to go out in the sun / and the sun bothers my eyes so much that I have to wear sunglasses'] (Corpes XXI, 2007 PRESEGAL SCOM_H13_012). 27 Overstreet is describing the German GE oder was weiß ich that can be considered equivalent to Sp. o lo que sea (or to closer, though less frequent, versions such as o yo qué sé, o qué sé yo).  As can be observed, Sp. y todo is particularly frequent in the corpus of spontaneous interactions where the number of occurrences is higher than for other GEs. This result reinforces the hypothesis that intensification plays a key role in this kind of interaction more than any other type of interaction or monological discourse.
Y todo points back to the phrase or clause to which it is attached and highlights it as unexpected information and thus highly relevant to the speaker's argumentation (Gras 2008(Gras -2022. This function is more relevant than set-marking, identified in the first descriptions (see Ball & Ariel 1978and Ward & Birner 1993for and everything, in Overstreet & Yule 2002. However, it also points to the shared knowledge between interlocutors, even if not necessarily to a specific or enumerable set. Y todo underlines the relevance of the previous chunk of information and conveys a scalar value. This means that it evokes a scale of some type and marks the accompanying information as being at a high or extreme point on the scale, according to the speaker/writer. (Overstreet & Yule 2002: 788) For example, in (22) G talks about a woman that discovered her homosexuality after she already married a man and the fact that this happened even when she had children is underlined by y todo, because this argument is not only considered the most relevant in this situation, but it also represents a high value on the scale. As a result, other arguments that are less salient or occupy lower positions in the argumentative scale remain implicit: 28 the woman discovers her homosexuality after meeting a man, falling in love with him, having a relationship, getting married etc. and only the element at the top of the scale, having children together, is made explicit as an anchoring constituent. 'G: § no no no no but I-I mean I still / I mean I come on a bit strong, right? people who are married for ten or eleven years↑ and after that time, right? she finds out that that-she doesn't like men § L: § ah! § G: § that she likes women E: Holy shit! § G: § after having had children and everything with him↑/ she discovers that she likes women' Argumentative salience can also be observed in (23) where a whirlpool bath is presented as the most emblematic representation of luxury underlined by y todo and confirmed by A's reaction: The expected elements (big bathroom, more than one bathroom, etc.) are implicit and must be inferred by the hearer. They all comprise part of the set "amenities in a good house". Highlighting the most relevant also constitutes a way of closing an argument.
In English, this type of GE is frequently used in but-prefaces (Overstreet & Yule 2002), i.e., structures such as X y todo, pero Y (X and everything, but Y).
[T]here is an acknowledgment in X of factors that would lead to certain expectations (and everything) following from X; however (but), the speaker/writer wishes to present information in Y as a justification for thinking contrary to these expectations. (Overstreet & Yule 2002: 786) This structure seems to be more frequent in English than Spanish. I have only found 7 occurrences in my data.
In (24) a woman is narrating how she felt after receiving chemotherapy for the first time. Following her daughter's suggestion, she goes to a supermarket and feels well when first going in, a feeling that is underlined by y todo. This gives rise to the expectation that she has not had any secondary effects. In order to cancel this expectation, she uses a but-construction after which she states how ill she felt a little later. The GE underlines a type of shared knowledge: when one is shopping one usually feels well. ANU.FILOL.ESTUD.LINGÜÍST., 12/2022, pp. 155-187. ISSN: 2014-1408. DOI: 10.1344/AFEL2022.12.8 (24) mamá vamos a Alcampo pues vamos // y estaba entré muy bien en Alcampo y todo pero cuando llegué así a junto donde están los cestos / noté así una cosa así por aquí // dije yo ay ~Sofía / m m ¿marchamos? y dice no voy a coger el pan / pues bueno vete a coger el pan / y ya noté como si quisiera vomitar (Corpes XXI, 2010 PRESEGAL SCOM_M21_041) 'mom let's go to Alcampo then let's go // and I was in Alcampo and everything but when I got to the baskets / I noticed something like here // I said oh ~Sofia / sh sh shall we go? and she said I'm gonna get some bread / well go and get the bread / and I felt as if I wanted to throw up.' This formula highlights an awareness of typical expectations: the speaker acknowledges that some expectations may arise on the basis of the anchoring constituent (if one goes shopping, then one does not have serious health problems) and presents Y as a justification to contradict those expectations. According to Santos (2002: 632), Sp. y todo seems to convey a concessive meaning similar to the meaning of incluso or hasta, and this could explain the absence of adversative conjunctions in the immediate co-text.
Another intensification mechanism involves repeating GEs as in (25-26) or combining different GEs (27). The type of intensification by repetition primarily occurs with adjunctive GEs (with or without a conjunction), whereas intensification by combination is more frequent (but not exclusively, cf. 34 below) with disjunctive GEs.
(25) mmm bueno pues eh te decía anteriormente / en los momentos difíciles él / bueno / pues estaba siempre al lado de los jugadores / nos defendía y // y si en alguna ocasión había que reunirse / nos reuníamos // con él y etcétera etcétera ¿no? yo creo que en ese sentido // pues fue un presidente // eeh ejemplar y sobre todo diferente a lo que había en ese momento (Corpes XXI, 2001 CORALES Lo que es la vida, 05/04/01, RNE, Radio) 'umm well, as I was telling you before / in difficult times he / well / he was always on the players' side / he defended us and // and if on any occasion we had to meet / we met // with him and etcetera etcetera, right? I think that in that sense // he was an exemplary president // err exemplary and above all different to what existed at the time.' (26) pues sí que nos juntamos gente de Erasmus pero cuando íbamos al gimnasio // yo creo que eran chicos que estaban estudiando aquí la carrera eran italianos // y al ser de fuera muchas veces los italianos o gente de fuera que van a clases de idiomas // se juntan y tal y tal con esa / tú te haces amigo de ellos y si vas a una fiesta con ellos // pues casi ge / toda la gente que conozca en vez de españoles son // extranjeros no sé por qué (Corpes XXI, 2010 PRESEGAL SCOM_H12_028) 'well, yes, we got together with people from Erasmus, but when we went to the gym // I think they were guys who were studying here, they were Italians // and being from abroad, many times Italians or people from abroad who go to language classes // get together and such and such with that / you become friends with them and if you go to a party with them // well, almost everyone I know instead of Spaniards are // foreigners I don't know why.' This repetition indicates that a lot more could be said and there is an iconicity relationship between duplicated forms and the suggestion that the implicit discourse is longer (Overstreet 2005: 1853, Cortés 2006a. In (25) the actions of the president of the soccer team to support the players were manifold: three are mentioned (estaba siempre al lado de los jugadores, nos defendía, nos reuníamos -'he was always on the players' side', 'he defended us', 'we met') but the duplicated GEs etcétera etcétera implies that many other similar actions were performed. This duplication is also found in other languages: Ger. und und und; und so, und so, und so. There is also an interrogative GE with the function of offering alternatives in a more emphatic tone which prompts the addressee to give an answer. It may be reinforced by a DM requesting confirmation such as ¿no? in (28). Thus, in this case boundaries between GEs and DMs are blurred as "the emphasis added is clearly not about another choice, but about giving an answer" (Overstreet 2005(Overstreet : 1857. Phonological prominence markers, such as loudness and rising pitch, contribute to highlighting the intensifying function which seeks to solicit the listener's agreement.

METADISCOURSE FUNCTIONS
Regarding metadiscourse functions, GEs seem to perform three main functions: they act as delimiting signals, for example between different parts of a narrative; they signal places of transitional relevance in interactions (to leave the floor in most cases); they show the interlocutor that the speaker is planning his/her discourse and is not willing to leave the floor yet.

Delimiting signals
Some speakers tend to use GEs as a way of delimiting speech chunks in their own interventions. The GE y tal seems to have specialized in this function in spontaneous speech. The text in (29)  The sequence of three GEs (yo no sé qué, y tal, y eso) accompanied by DMs (en fin), repetitions (luego), filled pauses (umm) and false starts evidence the speaker's difficulty in planning their speech. All these elements enable the speaker to keep the floor during forward planning, as they would run the risk of losing the floor with a silent pause.

GENERAL EXTENDERS IN REPORTED SPEECH
GEs are particularly frequent in reported speech. Direct and indirect speech is very frequent in oral narratives, which are primarily constructed by giving voice to their participants (Macaulay 1987, Baxauli 2000, Camargo 2006, Benavent 2022. When reproducing another's speech, it is not always necessary to be precise, exact, or to repeat literally what the other person said on a particular occasion (Gille & Håggvist 2010: 136). Sometimes it is enough to give an impression of their tone, opinion, or attitude. This is easy to achieve by mentioning just one word or clause and then using a GE to suggest that more was said in the same tenor.
The non-specific character of [a GE] also makes it well suited for use in quoted speech, where speakers seek to reproduce someone's words in an authentic manner, even when they are unable to reproduce them verbatim. (Secova 2014: 291) While there seems to be a group of GEs specialized in reported speech (see above for a list of reported speech GEs), disjunctive and adjunctive GEs can occur in both direct and reported speech (35). Reported speech GEs are not usually found outside the reported speech context.
(35) y me dijo mi madre mira si estás tú por el medio / puede darse el caso // de que nos llevemos bien / tu mujer y yo y tal y tal // pero si te marchas yo qué sé / lo que puede ocurrir (Corpes XXI, 2010  When GEs are used in direct speech as in (36), the speaker may use a GE to suggest that they could go on talking about a topic for a while but that the most important information has already been presented and the rest can be inferred: (36) es una especie de road movie documental sobre la / Nacional sexta // que es la carretera por la que yo viajé toda mi vida porque mi familia es de Madrid / entonces nos vinimos aquí en el ochenta y tres y yo llevo toda la vida viajando por ahí y bueno es / un poco el reencontrarme con la carretera de mi infancia y bla bla bla bla bla (Corpes XXI, 2007 The position of GEs in reported speech is (a) either part of the reported speech itself (see 35 above) or (b) part of the narrative framework introducing the reported speech, or to be more precise, closing the reported speech fragment and thus acting as a sort of metalinguistic mark or transition cue between two different voices in the text (37) Sometimes, there is no narrative frame introducing a voice and the GE is the only element that enables a text to be identified as reported speech, as in (38): (38) bueno a mí me sacaron medio muerto ¿eh? me sacaron del desierto / tuve cinco operaciones // seis meses de recuperación / bueno increíble // hicimos aquella presentación mía / en Cuatro / eeh no / hemos fichado a un aventurero bla bla bla entonces yo me vine con mi traje y tal yo tenía unas escayolas aquí / que había tapado con la camisa y con todo y yo / mmm de repente va y me dicen que yo tengo que bajar rapelando por la fachada de este edificio // para hacer una presentación triunfal de que han fichado a un aventurero (Corpes XXI, 2015 shirt and everything and I / umm suddenly they tell me that I have to rappel down the front of this building // to make a triumphant entrance that they have signed an adventurer.' In this regard, both GEs and DMs function as polyphonic cues, but whereas DMs, like prosody (see Estellés 2015), often signal the transition to a different voice at the beginning of a discourse unit (Borreguero 2017;Rosemeyer/Posio i.p), GEs signal the end of a discourse unit.

CONCLUSIONS
Notwithstanding the structural similarity of the elements that can function as GEs in Spanish and their fixed position after a sentence or clause, these discourse elements show a high degree of multifunctionality, similar to other European languages. This multifunctionality is sensitive to communicative situations (Fiorentini & Sansò 2016: 191). Some of the most frequent pragmatic functions performed by GEs in Spanish are interactive and metadiscursive functions: those relating to politeness, mitigation, and intensification, on the one hand, and the function of delimiting speech chunks and keeping the floor, on the other. Lastly, the frequent presence of GEs in reported speech has also been highlighted, either as part of reported speech itself or as part of the narrative frame.
However, some differences have been noted, especially regarding the GE y todo. Contrary to the multifunctionality deployed by equivalent elements in other languages (e.g. Eng. and everything, Fr. et tout), Sp. y todo is not used in the same contexts as other GEs given that its main semantic nature is not only to mark an implicit shared knowledge with the interlocutor but to introduce scalarity between the explicit anchoring constituent and the implicit information. As a result, it does not perform politeness and hedging functions. Its function is limited to intensification since it highlights the relevance of previous information and may even add a concessive value. Marking the information to which it is attached as remarkable or surprising is a shared function with the abovementioned GEs in English and French. However, GEs in both these languages may also perform other interactive functions (Overstreet & Yule 2002: 788) marked by different intonational contours. In Spanish, in contrast, a rising intonation is probably common to all the occurrences of y todo (but once again empirical studies need to corroborate this impression). Another important difference with English is the limited presence of Spanish GEs in formulaic constructions, especially in but-prefaced constructions.
In turn, "the frequency with which GEs are used in discourse are often determined by the formality of the interactional context and/or text type" (Fernández 2015: 11). I have observed that GEs are particularly frequent in spontaneous conversation and informal interviews but are almost completely absent in press conferences and speeches in formal situations. There is, therefore, a clear correlation between the use of GEs and the conceptional variation (communicative immediacy vs. distance), which is, to my mind, far more significant than the correlation with age groups and social classes highlighted in previous studies. However, a quantitative analysis is needed to verify this perception.
In fact, the study of Spanish GEs has a long road to go. One interesting question is their qualitative (not only quantitative) distribution in different Spanish varieties and different social classes. Another question is how they grammaticalized and what are the differences with DMs in their path towards grammaticalization. And lastly, it would be interesting to analyze whether there have been any changes in frequency and use since oral interaction records began: do young people employ new GEs or do they show any preference for any of these forms? Are some of these elements becoming obsolete? As we have seen above, many Spanish GEs (que si esto, que si lo otro; que si patatín, que si patatán; y toda la pesca) have not been found in the consulted corpora. Moreover, elements with negative evaluative function that downgrade what might have been said (Overstreet 2005: 1854) such as Eng. and all that shit and Ger. und so Kram, which have a Spanish equivalent in y toda esa mierda, have not been found in the consulted corpora. 32 Further investigation needs to determine whether this result is corpus-biased or signals a tendency in contemporary European Spanish. In turn, further research is needed to ascertain whether there is any other word or construction assuming the function of GEs in youth speak that has not been identified yet.
This study considers a wide range of forms and provides a more exhaustive description of their discourse function than previous papers on Spanish GEs. The theoretical approach is based on descriptions of GEs from five different languages, and not exclusively on the literature in English, and from different perspectives (sociolinguistics, pragmatics, language teaching, contrastive linguistics). My conclusion is that there are important cross-linguistic similarities not only in the formal characteristics of GEs but also in the pragmatic functions performed (and probably in the original grammaticalization processes which give rise to these functions). I hope this description will constitute a solid point of departure for future research on GEs in Spanish linguistics, especially in the fields of sociolinguistics, variational linguistics, and history of language.