General Extenders and Syntactic Analyzability: Sp. y todo eso vs. y todo

Abstract

This study examines the historical and functional divergence between two Spanish general extenders, y todo eso and y todo, in order to determine whether the loss of syntactic analyzability can serve as a criterion for distinguishing stages of grammaticalization and identifying pragmatic shifts. Drawing on extensive diachronic and synchronic corpus data, the analysis compares the formal evolution, semantic properties and pragmatic functions of both constructions. The results show that y todo eso follows a prototypical grammaticalization path marked by a progressive reduction in its internal structure, the weakening of referential meaning, and increasing freedom from syntactic constraints, while preserving analyzability through alternation with the simpler form y eso. In contrast, y todo displays an earlier and more advanced process of grammaticalization, dating back to medieval Spanish, in which the construction undergoes semantic bleaching, loss of additive value, and reanalysis as a scalar focus marker. These findings support the view that y todo no longer functions as a general extender in contemporary Spanish, whereas y todo eso retains this status, illustrating how syntactic analyzability correlates with shifts between pragmatic categories.

1. Introduction

Studies on the historical configuration of general extenders are still scarce (Overstreet, 2014; Secova, 2014; and microdiachronic approaches in Erman, 1995; Cheshire, 2007; Tagliamonte & Denis, 2010) and this is still an emerging line of research in the case of Spanish (Borreguero, 2023). This situation contrasts with the large number of studies that offer synchronic descriptions of the formal and semantic characteristics of these structures, as well as their pragmatic functions, for languages such as English (Dines, 1980; Overstreet, 1999), French (Andrews, 1989; Dubois, 1992), German (Overstreet, 2005; Cutting, 2015) and Italian (Voghera, 2012; Mauri, 2014), to mention just a few of the pioneering studies in each of these languages. In the case of Spanish, the work of Guil (2000), Domínguez (2005), Cortés (2006a, 2006b), Montañez (2008) and Gille and Håggvist (2010) paved the way in this field.

The aim of this paper is to compare a process of grammaticalization that can be considered prototypical of general extenders (the case of y todo eso) with another that is more problematic and that, a priori, could be thought to be the result of a later stage in the process of grammaticalization (the case of y todo). The following research question is addressed: may the loss of syntactic analyzability be a criterion not only to establish different stages of the grammaticalization process but also to identify shifts between pragmatic categories?

Our hypothesis is that the loss of syntactic analyzability (Torres Cacoullos, 2015) is a sufficient criterion to consider that a construction preceded by an additive or disjunctive conjunction is not a general extender anymore. This is indeed the case of y todo in its contemporary uses. A diachronic overview of its evolution will show a gradual loss of analyzability to the point that speakers of contemporary Spanish varieties are no longer aware that it was once an additive structure.

I will first offer a brief characterisation of the main formal and functional features of general extenders (§1), then describe the prototypical features of the grammaticalization process of these constructions as manifested in the case of y todo eso (§2) and compare them with the process that y todo seems to have undergone (Herrero, 2012, 2014), based on the data available to us and previous interpretations (§3.) Finally, I will attempt to demonstrate that both processes of grammaticalization have reached different stages of development, which explains why syntactic analysability is preserved in one case but not in the other, allowing us to conclude that y todo does not function any longer as a general extender in contemporary Spanish, contrary to what most previous studies have argued.

2. Materials and Methods

The data on which this study is based belong to European Spanish and are taken from the Real Academia Española (RAE) corpus, mainly from the Corpus del Diccionario Histórico (CDH) and the Corpus del Español del Siglo XXI (CORPES). Both corpora are available on <www.rae.es>.

The Corpus del Diccionario Histórico (Corpus of the Historical Dictionary of the Spanish Language) (CDH) is a large diachronic reference corpus compiled for the Real Academia Española’s Diccionario histórico de la lengua española. It comprises approximately 355.7 million tokens distributed across three searchable layers: a core subcorpus of over 53 million occurrences, a medieval-to-1975 subcorpus of roughly 199.4 million forms, and a 1975–2000 subcorpus of about 103.2 million records. The texts originate largely from the Academy’s historical and contemporary corpora (CORDE and CREA), and they are lemmatized and morphosyntactically annotated to support historical and lexicographical research on Spanish from the 12th century through 2000.

CORPES XXI (Corpus del Español del Siglo XXI) is a pan-Hispanic reference corpus developed by the Real Academia Española in collaboration with the Asociación de Academias de la Lengua Española. It comprises 445 million tokens drawn from written and spoken texts produced since 2001, representing the different geographical areas of Spanish (Spain, the Americas, the Philippines, and Equatorial Guinea) as well as a wide range of genres and registers. The European Spanish subcorpus contains around 133.5 million tokens. Designed according to principles of balance and stratification, the corpus is lemmatized and morphosyntactically annotated, enabling complex queries, frequency analyses, and studies of lexical, grammatical, and discourse variation in contemporary Spanish.

Occasionally, examples from other oral corpora are cited, such as the Val.Es.Co. corpus (Valencia Español Coloquial), a spoken corpus of colloquial Spanish developed by the Val.Es.Co. research group at the Valencia University. It compiles spontaneous, informal conversations recorded in natural interactional settings, predominantly in the Valencian speech community. The current electronic version (Val.Es.Co. 3.0) contains 115 colloquial conversations totaling 597.057 tokens, drawn primarily from spontaneous spoken discourse and annotated for prosodic and structural units (Pons & Badía, 2024). This resource supports studies in sociolinguistics, pragmatics, discourse analysis, and spoken language phenomena in contemporary Spanish.

The analysis and results provided in these pages is based on exhaustive searches of the analysed constructions, and in some cases statistical data from the corpus itself is cited. To facilitate a better understanding of the meaning and function of these structures, they are always provided in their context, which is sometimes quite broad, so we have not deemed it necessary to provide word-for-word translations. The translations are intended to help the reader understand the context of use of these constructions.

3. Formal, Semantic and Pragmatic Characteristics of General Extenders

General extenders (henceforth GEs) are constructions functioning as closing sequences at the end of a discursive fragment (called anchoring constituent) that usually takes the form of an enumeration or a narrative sequence. They are formed by a copulative or disjunctive conjunction, which allows them to be classified as adjunctive (En. and stuff, and things; Sp. y tal, y todo eso, y lo que quieras ‘and stuff, and all that, and whatever you want’)1 and disjunctive (En. or something, or whatever; Sp. o algo, o así, o lo que sea), respectively.

Depending on the element that follows the conjunction, they can be classified as simple or complex GEs. Simple GEs contain a single element after the conjunction: a demonstrative (and that), an indefinite (or something), an adverb (or so), an interrogative pronoun (or what) and, in languages such as English, even a noun (and stuff). Complex GEs, on the other hand, have a variety of structures: simple GEs combined with quantifiers (and all that), nouns preceded by a quantifier and determiner (and all that stuff), hypernyms with some kind of modifier (Eng. and things like that, Sp. y cosas de esas) and even clauses (Sp. y Dios sabe qué más, y yo qué sé ‘and God knows what else, and what do I know’), among others. (For an exhaustive list of these structures in English, see Tagliamonte & Denis, 2010, p. 337; for Spanish complex GEs, included the so-called quoted discourse GEs, which we will not deal with here, see Borreguero, 2022).

From a diachronic point of view, it has been argued that a simple GE is at a more advanced stage of grammaticalization than a complex GE (Cheshire, 2007; Secova, 2014) because the latter cooccurred more often with their expected referents. Following this logic, our initial hypothesis would lead us to assume that y todo could be at a more advanced stage of grammaticalization than y todo eso, in the same way as o algo ‘or something’ regarding o algo así (literally, ‘or something so’).

In terms of their semantics, these elements have no referential meaning of their own and function discursively as proforms (Guil, 2000 calls them allusive proforms), whose meaning is determined by their reference to an antecedent. This antecedent or anchoring constituent has traditionally been identified with a list (Dines, 1980), a set (Winter & Norrby, 2000), an ad hoc category (Mauri, 2014) or a non-exhaustive list (Mauri et al., 2019): ‘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, p. 282). This function of closing a non-exhaustive list—such as ‘occasions when gifts are given to a child’ in (1)—has led to the terms list-completers, set-marking tags (Winter & Norrby, 2000), enumerative closers (Ruiz, 1998), markers of (in)conclusion (Domínguez, 2005), elements at the end of an enumerative series (Cortés, 2006a, 2006b), etc.

• y es una vergüenza que en este país el regalo estrella de las comuniones Papá Noel o reyes magos y demás / fin de curso y tal / sean los móviles de última generación (CORPES, El Mundo, 2016).
  ‘and it’s a shame that in this country, the most popular gifts for communions, Christmas or Epiphany and so on / end of term and such / are the latest generation mobile phones.’

According to this conception, a GE refers to another element or elements that share some properties with the elements in the list. However, these lists rarely refer to stable sets of elements; rather, they are highly context-dependent, which is why they have been given the name of ad hoc categories (Mauri, 2014). For example, in (2) the places where the speaker used to walk as a child (building sites, rooftops) do not constitute an intersubjective category.

2.

tengo algunos recuerdos de ir por la / por las obras por andar por los tejados y tal (CORPES, Corpus del habla culta de Salamanca, 2003).

‘I have some memories of going to the building site, walking on the rooftops and stuff like that’.

However, several scholars have pointed out that it is not always possible to identify such lists. Indeed, GEs do not always refer to elements of a defined list or an ad hoc constructed set, especially when the anchoring constituent is not a NP (Cheshire, 2007), as seen in (3), where the clause todo el mundo tendrá comida familiar ‘everyone will have a family meal’ constitutes the anchoring constituent of the GE but it is difficult to interpret it as the first member of a list:

3.

A: § IMAGINO que todo el mundo tendrá comida familiar y todo eso (7″) (Val.Es.co. Corpus, VC.117.A.1, 214).

‘I imagine that everyone will have a family meal and all that.’

On the contrary, it is shared knowledge (common ground) that allows the interlocutor to identify the information suggested by a GE based on pragmatic information rather than on semantic relationships within a set. This information relates to shared experiences or perspectives. Therefore, this knowledge is social in nature or based on individual experiences (Overstreet, 1999; Overstreet & Yule, 2001). In (4), the speaker, who is an influencer, assumes that her interlocutor knows what professional changes are involved in opening a YouTube channel. She does not need to make them explicit but simply evokes them with the GE y tal ‘and all that’.

4.

en realidad a mí me gusta mucho la belleza / el maquillaje y la moda / pero en mi día a día yo voy cómoda / y la verdad es que no es que me arregle especialmente mucho ni ni / la verdad es que no / eeh sí que hubo un antes y un después / después de / el canal de Youtube y tal / pero si te digo la verdad en el día a día luego yo soy / mmm muy sencillita (CORPES, ratolina Q&I | ¿Por qué no voy a eventos?¿Me arrepiento?¿Hay postureo en Youtube?, 2016).

‘Actually, I really like beauty make-up and fashion, but in my day-to-day life I like to be comfortable, and the truth is that I don’t really get dressed up that much, no, not really. Well, yes, there was a before and after, after the YouTube channel and all that, but to be honest, in my day-to-day life I’m very simple.’

This function of conveying implicit information makes them carriers of a procedural instruction indicating that it is not necessary to expand the number of referents: ‘further processing in referential terms is not required’ (Overstreet, 2005). They therefore constitute a resource of linguistic economy, like all semantically vague elements, for example verba ómnibus (case, question, thing) or approximators (En. like; Sp. tipo, en plan, rollo).

Alongside this main function, GEs also perform interactional and metatextual functions (Overstreet & Yule, 2001; Winter & Norrby, 2000; Norrby & Winter, 2002). Among the former is the function of mitigating the illocutionary force of a speech act: mitigation is a pragmatic function that derives from semantic vagueness itself and is characteristic of both GEs and approximators (Jørgensen, 2009). In the case of GEs, it is particularly disjunctive GEs which, by offering alternatives (Brown & Levinson, 1987), can reduce the threat to the interlocutor’s face in questions (5) or commands:

5.

¿ tienes miedo a volar o algo? (CORPES, PRESEGAL SCOM_H13_013, Entrevista, 2007.)

‘Are you afraid of flying or something?’

On the other hand, there is also an interactional function directly related to linguistic politeness and Grice (1975)’s maxim of quantity: in order to avoid verbosity and redundancy in communication the GE helps the interlocutor to retrieve implicit information, thereby helping to speed up the pace of interaction. In (6), the speaker states that he is discouraged from going alone to places that are far away, and with the GE, he implies what it means to go to such places because it is knowledge he shares with his interlocutor, as this clearly shows by opening his turn with claro ‘of course’:

6.

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: = Because // if I must go alone to a place that’s a bit far away and stuff↑ I don’t go.

J: Of course /// [it’s better to go with people]’.

Among the metatextual functions, the main one is the segmentation of discourse into informative chunks to facilitate processing by the interlocutor (7). A particular case of segmentation is that of indicating the end of a turn with the aim of facilitating the alternation of turns between interlocutors, i.e., the GE marks a place of transitional relevance (8). Finally, as with many elements that have undergone a strong process of desemantisation (approximators, discourse markers), GEs become supporting elements in the process of discourse planning (9).

7.

yo creo que / yo creo que va a ser La La Land / básicamente porque el / el humor de Estados Unidos ahora mismo es tan // tan / o sea están tan bajos de moral / que necesitan algo que es / que que te incite a bailar y tal / y esta película la verdad es que / tiene tiene un punto de / que sales del cine / como cuando ibas a ver una película de Bruce Lee / y salías haciendo / patadas y tal / […] es lo mismo Tom Ford envió perfumes / de su línea / esta personal y tal a todos los / a miembros de la Asociación de la Prensa y tal (CORPES, Movistar+ Pool Fiction: Quiniela para los Globos de Oro y los Goya, 2016).

‘I think / I think it’s going to be La La Land / basically because the / the mood in the United States right now is so // so / I mean, people’s morale is so low / that they need something that makes you want to dance and stuff / and this film, the truth is that / it has a certain something / that when you leave the cinema / it’s like when you used to go and see a Bruce Lee film / and you’d come out doing / kicks and stuff / […] it’s the same thing. Tom Ford sent perfumes / from his line / this personal and stuff to all the / members of the Press Association and stuff.

8.

A: [por ejemplo] para administrar el Aese o administrar el Debedós pues utilizas→ pantallas de Iesepeefe / y Iesepeefe es un– es el gestor de menús para evitar ir comando a comando en Teseó que es bastante aburrido / pues te saca menús muy bonitos / llenos de colores y tal / y de ayudas y tal.

B: es una interfaz ¿no? o sea→ (Val.Es.Co. Corpus, XP.48.A.1).

‘A: [For example] to manage AS or DB2, you use → ISPF screens / and ISPF is a—it’s the menu manager to avoid going command by command in TSO, which is quite tedious / because it brings up very nice menus / full of colours and such / and help and such.

B: It’s an interface, isn’t it? I mean’.

9.

eso les / eso les permite pues a / acceder por ejemplo a la Administración / a determinados eeh / niveles / que si no tienes un título superior no puedes llegar ¿no? // entonces pues algunos han comenzado / una cierta carrera en la Administración desde auxiliar administrativo / eeh luego administrativo en fin luego yo no sé qué vi técnico luego gestión y tal / y eso mmm a a al último al último escalón de la Administración a la última escala de la Administración no puedes llegar / si no partes / de la si de una situación de licenciado ¿no? // (CORPES, Corpus de habla culta de Salamanca n° 5, 2003.)

‘That allows them to, for example, access the civil service at certain levels that you cannot reach if you do not have a higher education degree, right? // So some have started a certain career in the civil service, starting as administrative assistants, eeh then administrative staff, well then I don’t know what, I saw technician, then management and so on / and that mmm, in the end, in the end, you can’t reach the top of the civil service, the highest level of the civil service, if you don’t start from a position as a graduate, right?’

4. The Grammaticalization of GEs: The Case of y todo eso

GEs are the result of a process of grammaticalization in which sequences of categorically diverse components are reanalysed by speakers as a single chunk that can no longer be semantically decomposed. Unlike discourse markers (Garachana, 1999; Company, 2004a), GE undergo a “classic” process of grammaticalization in which a free sequence of words, with not fixed order and semantic compositionality, is transformed into a fixed construction of bounded forms, although without reaching the stage of becoming grammatical forms.

Formal changes at the morphological and syntactic levels play a fundamental role in this process of reanalysis and originate from the increased frequency of use of these forms (Bybee, 2011). As in the case of discourse markers, each GE has followed its own path of grammaticalization, which is not always easy to trace due to the limited data available in the corpora.

Some of the most characteristic features of this process are presented below (Tagliamonte & Denis, 2010; Overstreet, 2014; Secova, 2014; Borreguero, 2023):

• Erosion or phonetic reduction. This is evident in languages such as English and > /n:/ (Cheshire, 2007) but does not affect conjunctions in Romance languages, which are composed of a single phoneme.
• Integration into the intonation unit of the anchoring constituent: this feature is typical of simple GEs (10), while complex GEs tend to form an independent tonal unit and are separated from the rest of the utterance by means of a tonal inflection, as in (11) and (12), especially in the latter example where the GE appears between pauses that delimit the tonal unit in the transcription.

  10.

  Sí un chico joven que / trabaja y tal / y su / el típico tonto que nadie hace caso / que nadie quiere estar con él (CORPES, CHUS:MR08, Entrevista, 2001).

  ‘Yes, a young lad who / works and stuff / and his / the typical idiot that no one pays attention to / that no one wants to be with’.

  11.

  o sea cada carrera tiene su / su cosa / a unos nos dicen que claro / lo de letras es solo memorizar y tal y cual / digo / sí y no / pues no / no es así ¿no? (CORPES, CHUS:MR08, Entrevista, 2001.)

  ‘In other words, every degree has its own thing. Some people tell us that, of course, arts degrees are just about memorising and that. I say, yes and no. Well, no, that’s not how it is, is it?’

  12.

  las mujeres / somos más presumidas / el el no tener pelo te afecta mucho / el venga hombre quita la visera /y tal y cual // yo creo que eso es lo que más / mmm / más se se te satisface como persona // y eso (CORPES, PRESEGAL:SCOM_M23_001, Entrevista, 2008).

  ‘women / we are more vain / not having hair affects you a lot / come on, man, take off the visor /and so on and so forth // I think that’s what / mmm / most satisfies you as a person // and that’.

• Progressive reduction of the source structure: Eng. and stuff like that > and stuff, Fr. et tout ça > et tout (Secova, 2014), Sp. y todo eso > y eso.
• Independence from the original syntactic structure or cancellation of syntax (Company, 2004b): Some of the GEs come from structures that could be modified by relative clauses and perform syntactic functions as subjects or direct objects. In their diachronic evolution, they are freed from these syntactic patterns until, in many cases, they are reduced to the structure conjunction + core.
• Shifts in position from the propositional core of the utterance to peripheral positions, in appendix to other phrases, and even in more recent developments to inter- and intrasyntagmatic positions, play an important role in syntactic change. Although positions at the end of a turn or before a pause that delimits discursive units (acts in the terminology of the Val.Es.Co. Group, 2014) are very frequent, in more recent phases of the grammaticalization process, GE is found appended to syntagms that do not occupy final positions in the propositional core (13):

  13.

  Esas imágenes que vemos de la playa y tal están hechas por él, y es ella quien cobra y después le paga al fotógrafo (CORPES, El Mundo, 27 July 2003).

  ‘Those images we see of the beach and so on are taken by him, and it is she who charges and then pays the photographer’.

• Decategorization and loss of coreferentiality, which is manifested as loss of morphological agreement with the antecedent: Many of the nuclei of GEs are anaphoric in nature (demonstratives such as Sp. tal, eso, or those that appear in structures such as Sp. y cosas de esas, y todas esas cosas) and their antecedent is originally a noun with which they agree in gender and number. In the process of grammaticalization, first morphological agreement is lost but the reference to a specific, inanimate antecedent is maintained; later on, they become an appendix of NPs that introduce any type of referents, including abstract and animate referents (see Borreguero, 2023, pp. 225–226 for the case of y cosas de esas). The final step in this process is its disassociation from the NP: they are appended to VPs and other more complex syntactic structures (14).

  14.

  Clara: Ya sé que tú querías [que hiciera una carrera] y todo eso…, pero es un trabajo bien pagado… Y me deja tiempo libre… (CORPES, G. Morales, Como si fuera esta noche, 2010).

  ‘Clara: I know you wanted me [to study at university] and all that, but it’s a well-paid job, and it leaves me with free time.’

In the case of y todo eso, we will observe how some of these phenomena occur, starting with the increase in frequency shown in

Table 1. Occurrences of y todo eso in the CDH (Theatrical texts and dialogues in narrative texts).

Date	Total Number of Occurrences	Occurrences with GE Functions
16th C	6	-
17th C	8	2 (25%)
18th C	1	-
19th C	22	3 (13.6%)
1900–1949	33	4 (12.1%)
1950–1969	57	23 (40.3%)
1970–1979	76	49 (64.5%)

, which corroborates the degree of grammaticalization.

As we have already mentioned, GEs with complex nuclei evolve into GEs with simple nuclei. These types of changes have been described for British English and stuff like that > and stuff (Cheshire, 2007) and the French et tout ça > et tout (Secova, 2014), although they are not consistent with other microdiachronic approaches that compare speakers from different generations in other English varieties (see Tagliamonte & Denis, 2010 for Canadian English). Once the reduction occurs, a considerable increase in the frequency of reduced forms can be observed, as shown in

Table 2. Absolute frequencies and relative frequences (per million words) of simple vs. complex GEs in CORPES (Spain).

GEs	Number of Occurrences
Simple GEs	Y tal (962)—7.21	Y eso (456)—3.42
Complex GEs	Y tal y cual (40)—0.30	Y todo eso (51)—0.38

, which collects data from contemporary Spanish.

In the case of y todo eso, the most frequent structure until the 20th century (based on data provided by the CDH) has the following characteristics:

(a)

The conjunction functions as an additive textual connector without any link to a preceding noun phrase; it often appears after a strong punctuation mark and connects two independent utterances.

(b)

The demonstrative has its antecedent in the immediate co-text, as in (15)–(16) where eso has an anaphoric value whose antecedent (underlined) is formed by a sequence of NPs:

15.

Fulano, ¿qué hacienda tiene?—Señor, tantos pares de casas, tantas aranzadas de olivar, tantos cahíces de pan de renta, tantos mil ducados de juros; y todo es mentira, porque debe más que tiene, y todo eso está afianzado y atributado; que sacado en limpio lo que es suyo, apenas hay para comer. (CDH, Fray Alonso de Cabrera, De las consideraciones sobre todos los evangelios de la Cuaresma, a1598).

‘Hey, what property does he own?—Sir, so many pairs of houses, so many acres of olive groves, so many bushels of bread as rent, so many thousand ducats in interest; and it’s all a lie, because he owes more than he has, and all that is pledged and mortgaged; when you take away what is his, there is barely enough to eat.’

16.

Desmayarse significa nervios, voluntad contrariada, corazón, sentimientos…, y todo eso, Oshidori, acaba de quedar muerto dentro de mí (CDH, E. Jardiel Poncela, Usted tiene ojos de mujer fatal, 1932).

‘Fainting means nerves, thwarted will, heart, feelings…and all that, Oshidori, has just died inside me.’

(c)

The universal quantifier, as a determiner of the demonstrative, acquires an intensifying value; therefore, the conjunction, the quantifier and the demonstrative are syntactically and semantically analysable.

(d)

The most common syntactic function is that of subject (see 16 above) or direct object (17), whereby the NP is followed by a verbal form:

17.

[…] san Pablo, a su imitación, trai la boca abierta, como él scribe a los Corintios, diciendo: O Corintii, os meum patet ad vos, para que por allí se entren en sus entrañas y se asgan a las palabras de Cristo que por allí se derraman. Y todo eso lo pueden en Cristo, cuya virtud se les communica. (San Juan Bautista de la Concepción (Juan García López), Algunas penas del justo en el camino de la perfección, a1613.)

‘Saint Paul, following his example, opens his mouth, as he writes to the Corinthians, saying: O Corinthians, my heart is open to you, so that they may enter into his heart and cling to the words of Christ that pour forth from it. And all this they can do in Christ, whose virtue is communicated to them.’

(e)

It is also common for it to be followed by a relative clause:

18.

—Y todo eso que decís, oh valiente caballero, ¿no pudiera interpretarse como un sacrificio al genio de la vanagloria, una ofrenda pasada por llama impura para colocarla en las impías aras de Moloc? (CDH, Ramón López Soler, Los bandos de Castilla o El caballero del cisne, 1830).

‘—And all that you say, oh brave knight, could it not be interpreted as a sacrifice to the genius of vainglory, an offering passed through an impure flame to be placed on the unholy altars of Moloch?’

Three factors are necessary for the grammaticalization of this GE:

• Its appearance after a phrase or clause and the consequent change in the value of the conjunction, which ceases to function as a textual connector and begins to function as an intersyntactic copulative conjunction;
• The emancipation of the NP from any syntactic schema, in such a way that its use as an independent structure is consolidated; this involves removing the relative clauses (18) or prepositional phrases that modify eso (19).

  19.

  Mejor salud tenemos acá desde que se llevó Dios al médico—dijo la vieja, por nombre y cognomen Celedonia Recajo—, y aquí, don Quiboro, no hay mas maleficio que el no comer, y todo eso del miquiborio es enredo y trabalenguas como el nombre de usted. (CDH, B. Pérez Galdós, El caballero encantado, 1909.)

  ‘We have better health here since God took the doctor away—said the old woman, whose name and surname was Celedonia Recajo— and here, Don Quiboro, there is no greater evil than not eating, and all that miquiborio stuff is nonsense and gibberish, just like your name.’

• The combination of this structure with other lexical elements that form an enumeration or list.

At the end of the 19th century, we find enumerations closed with the form y todo eso, followed by a relative clause that provides information about the property common to all the elements that make up the list: in other words, the relative clause makes explicit the defining feature of the ad hoc category to which the listed elements belong. This is accompanied by a significant change in position after a noun phrase.2

20.

LUCRECIA.—(Con acritud.) ¡A quejarte! ¿De qué? Pues eso me faltaba. ¿Crees que tengo yo en mi mano los destinos, las fianzas, y todo eso que ambicionas? (CDH, Benito Pérez Galdós, El abuelo, 1897.)

‘LUCRECIA.—(Bitterly.) To complain! About what? That’s all I need. Do you think I hold in my hand the destinies, the bonds, and all that you covet?’

21.

—Sí que es verdad, señor Félix. Si recorriese esto como nosotros, bien se hartaría de comer con los ojos el candeal y todo eso que se le antojaba (CDH, Gabriel Miró, Las cerezas del cementerio, 1910).

‘—That’s true, Mr. Felix. If you travelled around here like we do, you’d get your fill of feasting your eyes on the wheat fields and everything else that took your fancy.’

From the first third of the 20th century onwards, this same structure is documented but with the relative clause omitted, so that the defining feature of the elements in the list must be inferred from the context, as in (22):

22.

Sí, también inventó las casas: descubrió la manera de hacer yeso y hacer mortero y ladrillos, y luego, con piedras y mármoles y tierras y todo eso, levantó casas… (CDH, Manuel Abril, Cuentos para niños, c1930).

Yes, he also invented houses: he discovered how to make plaster and mortar and bricks, and then, with stones and marble and earth and all that, he built houses…

Once again, we encounter a linguistic economy device, as the speaker considers it unnecessary to encode the meaning and trusts in the interpreter’s ability to infer other elements of a set that are not explicitly mentioned.

Once emancipated from this syntactic pattern, the anchoring element is no longer systematically a list of nominal elements and the GE appears as a closing sequence after other types of lexical units, for example, adjectives (23) or even clauses (24). In the CDH, these examples are documented from the 1950s onwards. A search in the CDH for the sequence y todo eso followed by punctuation marks yields 195 cases, of which only 10 are prior to 1950 (not all of them are GE), 12 belong to the 1950s and 1960s, 78 to the 1970s and 1980s, and 94 to the 1990s, which gives an idea that the use of this sequence as GE has become widespread in contemporary peninsular Spanish. Considering that the period 1901–2005 represents 37% of the occurrences in the corpus, the frequency is 1.77 per million words in the 20th century compared to 0.05 for the period between 1801 and 1900.

23.

—También hablaba de ti, no creas. Sabemos la clase de estudiante que eres, revoltosa y todo eso… ¿No puedes correr más? Tengo prisa (CDH, Juan Marsé, Últimas tardes con Teresa, 1966).

‘—He was talking about you too, don’t think he wasn’t. We know the kind of student you are, rebellious and all that… Can’t you run any faster? I’m in a hurry.’

24.

Estaba yo en mi despacho, muy quitado de la pena, cuando llega un muchacho, de los que vienen aquí todos los días, y que se hace pasar por amigo de Pablito; yo no lo reconocí al principio, como vienen tantos; pero él se presentó, y dijo que venía muy seguido, y todo eso… (CDH, Jorge Ibargüengoitia, Susana y los jóvenes. Comedia en tres actos, 1954.)

‘I was in my office, feeling quite downhearted, when a lad arrived, one of those who come here every day and claim to be friends of Pablito. I didn’t recognise him at first, as so many come here, but he introduced himself and said he came here often, and all that…’

The final step in the grammaticalization process is the loss of the coreference relationship between the GE and its antecedent, undoubtedly brought about by the different syntactic entity of the latter. In examples such as (22), the demonstrative functions as a textual anaphora encapsulating a discursive fragment that takes the form of an enumeration (although sometimes only the first member of this enumeration is mentioned, as in (21)) and constitutes the antecedent of the anaphora. The encapsulating function is maintained when this antecedent is no longer nominal but adjectival or predicative, as in (23) and (24), although in these cases it is much more difficult to reconstruct the implicit elements, even when the context is known. In this case, the function is not to indicate the incompleteness of a set of elements (Domínguez, 2005) or an ad hoc category (Mauri, 2014), but rather to evoke shared knowledge about what it means to be a rebellious scholar (23) or the phrases with which a conversation with strangers begins (24).

In the second half of the 20th century, we find cases in which the antecedent can be a nominal expression referring to an animate or even human entity, as in (25) and (26), contrary to expectation because the form of the demonstrative eso always privileges inanimate entities as antecedents. We are therefore faced with a clear case of loss of referentiality.

25.

—Quiero ver gente respetable —dijo Elena—. Señoras de Acción Católica y todo eso… ¿Dónde se las encuentra? (CDH, Juan Goytisolo, Señas de identidad, 1966).

‘I want to see respectable people,’ said Elena. ‘Catholic Action ladies and all that… Where can I find them?’

26.

Se pasa la vida leyendo a Tolstoy, Ibsen y todo eso… (CDH, R. Chacel, Barrio de las maravillas, 1976).

‘He spends his life reading Tolstoy, Ibsen, and all that sort of thing…’

5. The Grammaticalization Process of y todo

Following Secova’s (2014) hypothesis about the origin of fr. et tout < et tout ça, we could assume that the loss of the demonstrative represents a new step in the process of grammaticalization that gives rise to a simple GE. However, the historical study of the form y todo ‘and all/and everything’ openly contradicts this hypothesis (Herrero, 2012).

If we consider the cases in which the conjunction does not function at the textual level linking two sentences, the sequence NP(s) y todo + NP can be found as early as the 14th century (Herrero, 2012, 2014) as the closing of enumerations, preceding the combination of lo + indefinite (lo al, lo demás ‘the rest’) modified by a relative (27) and (28). In these combinations, todo has an emphatic value and can be omitted without affecting the grammaticality of the construction. For its part, the relative pronoun indicates the common denominator of the category to which it refers.

27.

E en quanto el rrey alli estaua en aquellos tres dias, fueron algunos de los de la hueste a Faro e a Laule lugares que son en esta costa de la mar, e a otros lugares que son en esta comarca, e traxeron ganados e vacas e ouejas e omes catiuos e todo lo al que fallaron (CORDE, Gran crónica de Alfonso XI, c. 1348–1379).3

‘And while the king was there during those three days, some of the army went to Faro and Laule, places on this coast, and to other places in this region, and brought back cattle and cows and sheep and captive men and everything they could find.’

28.

E por que sopieron que se apellidaua toda la tierra y ayuntauanse muy grandes gentes de moros para venjr alli por mar y por tierra y ellos tenjan el acorro muy lexos oujeron a dexar la villa y troxieron ende muchos moros y todo lo al que qujsieron traer y vinjeronse para seujlla syn njnguna contienda. (CDH, F. Sánchez Valladolid, Crónica de Alfonso X, a1340–1350).

‘And because they heard that the whole land was being ravaged and that a great number of Moors were gathering to come there by sea and by land, and they had their supplies very far away, they decided to leave the town and brought with them many Moors and everything they wanted to bring, and they came to ravage it without any fighting.’

This structure alternates with another in which the relative or even the emphasized NP has been lost:

29.

por este respecto, avido su consejo, procuraron cómo hacer junto al real y casi dentro dél una población de muy buena cerca y valuartes con sus traveses, buenas casas y todo lo demás, para que se pudiesen defender contra los enemigos (CORDE, Alonso de Santa Cruz, Crónica de los Reyes Católicos, 1491–1556).

‘In this regard, having sought his advice, they managed to build next to the royal palace and almost within it a well-fortified town with ramparts and crossbars, good houses and everything else, so that they could defend themselves against their enemies’.

From the 15th century onwards, there are enumerative closures with y todo in which this sequence functions as a GE evoking the rest of the elements of a set that is not made explicit (30). The construction is also found in some documents prior to the 15th century, but its interpretation is not without difficulty due to the rarity of the conjunction y ‘and’ in medieval texts—where the forms e, et and & predominate—which makes its presence in texts from the 12th and 13th centuries suspicious, to say the least, not to mention cases where it is confused with the adverb of place ý (cf. Herrero, 2012, pp. 158–159). Therefore, we will not take these cases into account in this study.

30.

y, por que tengáis juntadas las bestias, ombres y todo, tomad estas dos posadas (CDH, Anónimo, Cancionero de obras de burlas provocantes a risa, c1445–1519).

‘and, so that you may have your animals, men and everything, take up residence in these two inns’.

One of the bridging contexts that allows the transition to another value that is not specific to the GE is one in which y todo is linked to a single element that can evoke a set or category (ad hoc or not) but which, when only a single element is mentioned, is difficult for the interlocutor to recover, as in (31).

31.

quien roba el coraçón ha de llevar cuerpo y todo (CDH, Juan del Encina, Cancionero, 1481–1496).

whoever steals the heart must take the body and everything else’.

This is a bridging context because three interpretations are possible: (a) the interpretation as GE (“the body and everything else”), (b) the additive interpretation (“also the body”), and (c) the scalar additive (“even the body”), which is best seen in occurrences from the 16th and 17th centuries. In (32)–(34), y todo has an additive value (Castro & Gili Gaya, 1917; Sánchez, 1999, p. 1110, fn. 56): “leave the shield too”, “the chain flew away and so did the common sense”, while in (35) the scalar additive reading (König, 1981) seems clearer (“he gets up even with the chair”).

32.

PÁRMENO. ¡Huye, huye, que corres poco! ¡Oh pecador de mí, si nos han de alcanzar, deja broquel y todo! (CORDE, Fernando de Rojas, La Celestina, 1490–1502.)

‘PÁRMENO. Run, run, you’re not running fast enough! Oh, sinner that I am, if they are going to catch us, leave your buckler and everything/too!’

33.

Dato. La cadena voló, y el juicio y todo. (CDH, Augustín Moreto, El lego del Carmen, 1652.)

‘Dato. The chain flew, and the common sense too.’

34.

Doña Petronila: Estoy celosa.

Conde: Yo y todo;

Mas hay dos suertes de celos

Unos nobles y otros no. (Tirso de Molina, La huerta de Juan Fernández, 1634, in Castro & Gili Gaya, 1917.)

‘Mrs. Petronila. I am jealous.

Count. Me too;

But there are two kinds of jealousy

One is noble, and the other is not.’

35.

DAMA 2.ª No es la menor que al decirlo / de la tierra se levanta / con silla y todo (CDH, Calderón de la Barca, Entremés del convidado, c1658).

‘DAMA 2.ª It is not the least that when saying it / he rises from the ground / even with the chair’.

In the scalar additive interpretation y todo becomes a focus marker that precisely focuses on the nominal element that serves as its anchoring constituent, placing it at the top of a scale (Domínguez, 2005; Gras, 2008; Herrero, 2012; Guillén, 2022, pp. 218–219), similarly to what happen to Eng. and everything with a certain intonational pattern (Ward & Birner, 1993, pp. 212–213; Overstreet & Yule, 2002).4

The additive value of the conjunction and, therefore, the function of GE are gradually lost, something that becomes evident in contexts where y todo is not closing a list (Castro & Gili Gaya, 1917) and thus it is not possible to recover more elements of a category, especially when metaphorical images are used (36):

36.

Di dos o tres esperezos y levántome más tiesa que un ajo, dando decamino un pescozón al mochillero para sacarle el sueño con raíces y todo (CDH, F. López de Estrada, La pícara Justina, 1605).

‘I gave two or three shakes and got up straighter than a stick, giving the backpacker a slap to wake him up [lit. to pull out his dream with roots and all]’.

In this other example from the second half of the 17th century, it can no longer be read as GE and the conjunction has lost its additive value. Note that in (37) the anchoring constituent is no longer an NP, which favours the additive scalar reading (“even by me”).

37.

1.º Dejemos eso y a otra cosa vamos;/ y pues juntos aquí los tres estamos, / vamos a entretenernos de algún modo./ 2.º Por mí, vamos, ¡por Dios!/ 3.º Y por mí, y todo (CORDE, Vicente Suárez de Deza, Mojiganga de los casamientos, 1663).

‘1. Let’s leave that and move on to something else;/ and since the three of us are here together,/ let’s entertain ourselves in some way./ 2. As far as I’m concerned, let’s go, for God’s sake!/ 3. [Lit. And even as far as I’m concerned]’.

In fact, y todo behaves like a focalizer and, therefore, can be postponed to constituents of all grammatical categories since the 17th century (Herrero, 2012, p. 155), although more frequently since the 20th century.5 Therefore, we can say that, at the latest, in the 17th century, this structure lost its syntactic analyzability and was thus reanalysed and processed by the speakers as a chunk. In fact, it is detached from the sequence of NPs and even from the nominal anchoring constituent to be placed as an appendix to any category or complex structure, thereby also losing its semantic compositionality since the idea of a whole conveyed by todo is not preserved. Only the procedural meaning of emphasizing remains, allowing it to be converted into an intensifier:

38.

Ya no estamos en tiempo de Calomarde; ahora se puede hablar claro y sin rodeos todo lo que se piensa, cuando se piensa. Aquí se habla mal de muchos ministros, y se los nombra y todo: a nadie han preso todavía por eso (CORDE, Mariano José de Larra, Carta de Fígaro a su antiguo corresponsal, 1835).

‘We are no longer in Calomarde’s time; now we can speak clearly and without beating about the bush everything we think, when we think it. Here many ministers are spoken ill of, and they are even named: no one has been imprisoned for that yet’.

From a semantic point of view, we are dealing with a process of subjectivisation (Traugott & Dasher, 2002), according to which a form or sequence of forms acquires pragmatic value as a result of the bleaching of referential meaning. The process of subjectivisation is closely linked to the loss of syntactic analysability, as typically occurs in deverbal discourse markers. “There is a fairly transparent, inversely proportional relationship between the amount of syntax a form requires and the degree of subjective-evaluative meaning that form expresses” (Company, 2004a, p. 41): the speaker’s assessment, in this case the emphatic value, acquires explicit codification in the grammar of a language, which allows y todo to be used as a marker of intensification, which is the main value in contemporary Spanish.

39.

G: [….] está casada durante diez o once años↑ yy al cabo de ese tiempo ¿no? ella descubre que que—los hombres no le gustan § […] que le gustan las mujeres.

E: ¡hostia! §

G:                       § después de haber tenido hijos y todo con él↑/ descubre que le gustan las mujeres (Val.Es.Co. Corpus, L.15.A.1, 714–723).

‘G: [….] she has been married for ten or eleven years↑ and after that time, right? she discovers that she doesn’t like men § […] she likes women.

E: Wow! §

G:                §    after having even children with him↑/ she discovers that she likes women’.

It is interesting to observe how y todo ends up acquiring new functions in more recent times and evolves towards a structure with a concessive value in an incidental position (Herrero, 2012, pp. 164–167, 2014), as do other focus particles such as incluso.

Therefore, we can conclude that y todo does not function as a GE in contemporary peninsular Spanish and has not done so since the Middle Ages. Although it appears in structures with this function in some medieval texts, its grammaticalization predates that of most Spanish GEs, which occurred between the 19th and 20th centuries. By the 15th century, the construction had taken on an additive (too) and scalar additive (even) value that is not observed in other GEs. Although the former was lost in the 18th century, the latter does not seem to have undergone any changes in the last two centuries.

6. Conclusions

After examining the evolution of both structures in some detail, we can observe commonalities in the process of grammaticalization, specifically the presence of certain linguistic changes characteristic of these processes, such as the loss of the referentiality with the antecedent, the combination with anchoring constituents belonging to categories other than the noun phrase, the desemantization of the universal quantifier, or its displacement to peripheral positions in phrases and utterances.

However, in the case of y todo, we can speak of a more advanced stage in the process of grammaticalization, not only because of the loss of semantic compositionality, given that speakers no longer perceive an additive value in the conjunction, but also because of a loss of syntactic analyzability due to the impossibility of inserting any element between the conjunction and the quantifier or of modifying it in any other way.

In the case of y todo eso, however, despite the loss of referentiality of the demonstrative, we observe that this construction can be alternated with a semantically and functionally equivalent simple GE (y eso). This indicates that the construction is perceived by speakers as syntactically analyzable, since the quantifier can be extracted and continues to maintain an independent semantic value, i.e., giving greater emphasis to the construction. According to the CORPES data reflected in Table 2, this alternative form is much more frequent in contemporary European Spanish and can be considered a phenomenon of lexical replacement similar to that observed by Tagliamonte and Denis (2010) for and things versus and stuff.

Therefore, we can conclude that the loss of syntactic analyzability is a good indicator of pragmatic shifts not only regarding the shift toward the pragmatic or discursive level of lexemes or sequences of words that become lexicalized but also regarding movements between pragmatic categories. If a GE cannot be syntactically decomposed, it ceases to function as a GE and becomes another type of pragmatic element, a focus particle or an intensifier, as in the case of y todo.