Non-Standard Grammatical Features in Castile-La Mancha

Abstract

The Spanish spoken in contemporary Castile-La Mancha has been traditionally considered as a mere transitional variety between Northern Castilian and Southern Spanish of the Andalusian type. In the few works devoted to its description, local phonetics attracted a good part of the attention, and the characterisation of its grammar was limited to a heterogeneous list of morphological and syntactic features. Amonag them, there was the presence of leísmo and laísmo in some areas of the region, the considerable extension of sub-standard trends, commonly found in other peninsular varieties (dequeísmo, deísmo, non-standard clitic sequences me se or te se …) and other less-extended features such as the transitive use of the verbs entrar, caer or quedar. In this work, we will address the description of the local manifestations of three grammatical features (third-person clitic pronoun systems, deísmo and reduction in second-person plural desinences) that are widely distributed in Castile-La Mancha and may be considered as general specific traits. By doing so we aim to offer a better definition of the Spanish spoken in this region, beyond the transitional cliché. A temptative approach to determine internal boundaries will also be made in order to illuminate some of the historical components that lie beyond the constitution of this Southern Spanish variety.

1. Introduction

The Spanish spoken in the autonomous community of Castile-La Mancha has not received much attention in the canonical literature on Peninsular Spanish Dialectology (Zamora Vicente 1960; Alvar (dir.) 1996). It is true that its ascription to the transitional dialects that fill the gap between Old Castile varieties and Andalusian or between Northern Peninsular Spanish and Southern Peninsular Spanish represented by Andalusian can be a sufficient description for general surveys on Spanish dialectal divisions or classifications. But it is also true that they can no longer be satisfactory for specialists, as these surveys lack descriptive and explanatory ambition and details. Furthermore, they continue to overestimate phonetic differences and historical connections and tend to avoid any consideration of the grammatical facts that can now be found in recent atlases or data corpora such as the Atlas Lingüístico y etnográfico de Castilla-La Mancha or the Corpus oral y sonoro del español rural. More recent works on this matter are helping to introduce grammar into the picture and give way to a refreshed vision of divisions and boundaries within Peninsular Spanish (Fernández-Ordóñez 2016, 2023; Camus and Gutiérrez Rodríguez 2021).

The present work is therefore dedicated to this necessary task of describing the specific grammatical features of the Spanish spoken in Castile-La Mancha in order to locate them in the more general and larger picture of Peninsular Spanish dialects. Consequently, we will address its grammar after reviewing previous works on Castile-La Mancha phonetics and its main traits. More specifically, we will focus on the distribution of three features of wide extension in the region: the presence of the referential clitic pronoun system with dative le as an accusative clitic (leísmo) and accusative la as a dative clitic (laísmo), as shown in (1); the insertion of the preposition de (deísmo) in infinitive sentences with no preposition in Standard Spanish (2) and the development of reduced forms of second-person plural desinences (3):

(1)	a.	A Juani,	lei	vi	ayer	(leísmo)
		To Juan (Direct Object)	3rd SING DAT	see1st SING PERF	yesterday
		‘Talking about John, I saw him yesterday’
	b.	A Maríai,	lai	di	un regalo	(laísmo)
		To María(Indirect Object)	3rd SING ACC	give1st SING PERF	a present
		‘Talking about Mary, I gave her a present’

(2)	Me	hacen	de	reír
	1st SING ACC	make3rd PL PRES	of	laughINF
	‘They make me laugh’

(3)	quer-ís	instead of standard form:	quer-éis
	want2nd Pl PRES, verb querer 2nd conj. in -er
	‘you (plural) want’

The accomplishment of this specific goal will allow us to try to redefine the nature of the Spanish spoken in Castile-La Mancha and its internal divisions and connections with other dialects. We will be able to underline the existence of some opposite forces between innovation and conservatism that articulate this central territory, between the pressure of standard and normative Spanish and archaism and fidelity to local usages.

The work will be organised as follows: Section 2 will be devoted to the discussion on the historical and phonetic bias in Spanish Dialectology and the recent development of a grammatical turn with interesting consequences for Peninsular Spanish dialect classification. In Section 3, Castile-La Mancha-defining grammatical features will be addressed and described. Finally, Section 4 contains, as a conclusion, an approach to more complete and accurate consideration of Castile-La Mancha’s variety of Southern Peninsular Spanish resulting from the previous description of its grammar and phonetics.

2. The Characterisation of Peninsular Spanish Dialects

Ever since the scientific study of geographical linguistic differences and the beginning of Dialectology as a branch of (Historical) Linguistics in the late 19th century, its focus was lexical inventory and phonetic traits of the vernacular varieties of national languages. This innovative discipline brought special attention to spoken samples of the cultivated languages in Europe. The actual nature of powerful Western cultures was claimed to reside in the language spoken by peasants and villagers from hidden small towns. Dialectology was to be built on the description and study of these differing manifestations, contrary to the basis of old Philology and its main interest in more homogeneous written texts. The inevitable tool of the dialectological approach became the linguistic survey with a thorough questionnaire on lexical preferences and divergent pronunciation. The answers were carefully mapped, and the corresponding cartographical collection led to the classifications of dialects and subdialects within national literary languages (García Mouton 2016, pp. 30–31; Camus and Gutiérrez Rodríguez 2021, p. 149).

2.1. Phonological and Syntactic Features as a Means of Defining Spanish Dialects

As a result of this historical bias in Spanish Dialectology, traditional classifications and subsequent dialectal boundaries are usually made using mainly lexical or phonetic features and rarely considering grammatical (morphological or syntactic) characteristics. It is a consequence of the lack of a consistent grammatical theory in those first years of the 20th century when Dialectology and Geolinguistics were taking shape. As will be shown, grammar started to be taken into greater consideration in the 1980s, when the advance of Generative Grammar forced a growing interest and knowledge of this component of language.

In the case of Spain, the study of Peninsular dialects of Spanish is not only conditioned by the bias of early Dialectology. Spanish Linguistics started to develop in 1900 within the framework of Philology and Neogrammarian Historical Linguistics, particularly under the enormous influence of Menéndez Pidal. The main concern in those early days of language science in Spain was the role of the Spanish language as a way to a better knowledge of the history of Spain. Attention to vernacular varieties is thus guided by this focus on History and a primary interest in medieval peninsular dialects (Paufler 1997, p. 421).

This explains the prevalence and enduring prestige of some works on Spanish dialects, for instance, Zamora Vicente (1960). It provided a classification of Peninsular Spanish dialects that can be considered the reference all through the second half of the twentieth century. It distinguishes (Asturian-)Leonese, Aragonese, Andalusian and transitional varieties that include the Rioja dialect, Extremaduran, Murcian and the Canary Islands dialect outside the Peninsula. This list overtly shows a historical impression and a Castilian-centred perspective. But it also focuses almost exclusively on phonetic and lexical traits, especially when dealing with Andalusian and transitional dialects. It is worth noting that these are mainly southern varieties of more recent formation, more widely known thanks to the development of Geolinguistics from the 1930s.

Things began to change in the final years of the twentieth century. Generative Grammar provided more comprehensive tools to new dialectologists. The phonetic and lexical features found in traditional atlases were gradually completed with extensive data collections centred on distinctive morphology and syntax. This crucial task corresponds largely to the development and circulation of the Corpus Oral y Sonoro del Español Rural or COSER (Fernández-Ordóñez (dir.) 2005-). This collection of surveys from all over Spain was initiated in the early 1990s at the Universidad Autónoma de Madrid and is only now coming to an end (Fernández-Ordóñez and Pato 2020). Its example has been followed by other syntax-centred online corpora and collections, for instance, the Atlas Sintáctico del Español [en línea] (ASinEs) (n.d.), developed at the Universitat Autònoma de Barcelona under the direction of Ángel J. Gallego.

As a result of this extensive renewal of perspectives and methodology, synchronic considerations and grammar are gaining ground in Spanish Dialectology. Some contemporary works, notably Fernández-Ordóñez 2016, have taken over from Zamora Vicente as the references regarding the description, ordering and classification of Peninsular dialects. Thus, the so-called historical dialects, descending as Spanish (Castilian) directly from Latin, are now discarded and replaced by contact varieties of Spanish in bilingual territories. Therefore, the attention turns to Castilian-based varieties that are classified following two different axes. First, there is a major split between Northern and Southern dialects. Secondly, Peninsular Spanish seems to be articulated, particularly in the north, into areas that run parallel to a north–south axis and give place to western, central and eastern varieties (Fernández-Ordóñez 2016, p. 387). In both cases, the defining features for each dialect are preferably of a grammatical nature, and among these, the system of third-person clitics constitutes a robust index to establish solid dialect areas (Fernández-Ordóñez 2016, pp. 388–90; 2023, pp. 165–70; Camus and Gutiérrez Rodríguez 2021, pp. 168–70).

2.2. Conservative and Innovative Features and Their Distribution in the Peninsula

One interesting consequence of this more syntactic approach to Peninsular dialects is the redefinition of innovative varieties versus conservative ones. When considering phonetic traits, it becomes clear that many successful innovations originated since 1500 in southern areas and were disseminated northward in the Peninsula or, more frequently, they travelled across the Atlantic to the Canary Islands and America. This is the case of the convergence of old sibilants or seseo, the merger of the lateral palatal [λ] with the fricative palatal [ʝ] or yeísmo and the various changes that affect coronal consonants in the syllabic coda, including the significant weakening of [s] in this position.

As a result of this, it has become commonplace to define Andalusian or Canary Islands dialects as innovative. Other Spanish varieties spoken in the southern half of the Peninsula tend to be classified together with these ones as long as they present those changes in a more or less advanced stage of development. Thus, a crucial dialect boundary would divide Peninsular Spanish between innovative dialects and conservative dialects (Chela-Flores 2023, p. 20). The former would extend from the south slope of the Central Range to the southern coastal line of Spain. The latter would then occupy the northern half of the Peninsula. The representation of one of these changes in syllabic codas in Figure 1 below, the weakening of [s], reveals the approximate line that corresponds to that boundary.

This broad characterisation of the fundamental nature of Peninsular dialects, even if it is mainly based on their phonetics, is somewhat reassuring from a traditional perspective as it would fit the historical extension of Spanish. Innovative non-standard pronunciation would be more likely found in those new territories repopulated by people coming from the north who were speakers of old Castilian varieties (and other northern varieties such as Astur-Leonese or Aragonese).

But this state of affairs no longer holds if we introduce grammatical criteria to define Peninsular Spanish dialects. For instance, the distribution of the aforementioned systems of third-person clitics overtly contradicts the conservative nature of northern dialects. Radical innovations in this regard, corresponding to the so-called referential system, are clearly found to the north of the line that splits the Peninsula in two, the Central Range. And most of the varieties to the south of that line vigorously stand by the original usage of the clitics, the one inherited from Latin and therefore called the etymological system (Fernández-Ordóñez 2016, p. 390), as shown in Figure 2:1

Some other isoglosses drawn from the distribution of grammatical features can be cited in support of this division between a northern innovative syntax versus a southern conservative one. For instance, the following Figure 3 delimits an area where subjunctive verb forms of Standard Spanish are substituted by indicative forms in many contexts:

As Figure 3 shows, this innovative preference for indicative, also found in some neighbouring Romance languages such as French, Occitan or Catalan since Medieval times, is clearly established in a northern area and totally absent in the southern half of the Peninsula.

This “syntactic shift” in Spanish Dialectology has thus resulted in a more complex definition of current Peninsular varieties, where old labels such as conservative and innovative no longer fit unequivocally any dialect. But, in addition to this idea, dealing with grammatical features to define Spanish Peninsular dialects has another interesting consequence that was already mentioned at the end of the previous subsection. As Figure 2 and Figure 3 above show, in the northern half of the Peninsula to the east and west of the central space, there are two other areas that lack the referential system of pronouns and the preference for Indicative and may be therefore individualised as different linguistic domains that were missing in traditional classifications.

2.3. Southern Peninsular Spanish and the Varieties of Central Spain

If we now turn to the southern half of the Peninsula, internal divisions are much more difficult to draw. Whereas phonetic features had already been useful to recognise this Southern Peninsular Spanish (see Figure 1), the addition of syntactic traits confirmed its distinctive nature (Figure 2). Nevertheless, contrary to what is found in the north, there seem to be no specific features, phonetic or syntactic, that can establish secondary areas inside this southern space. Southern phonetics are present all over this territory with different scopes and intensities. The southernmost region of Andalusia is the one that presents more marked features in this regard. It is also here where a specific solution (ceceo) in the evolution of old sibilants can be found together with seseo. The behaviour of southern syntactic features is almost the same. They can be found all over this area but appear to be more consistent the more we move southwards. And again, there are specific Andalusian features such as the progressive loss of the second-person pronoun vosotros.

Consequently, the traditional consideration of an easily recognisable Andalusian dialect would have to be maintained. This leaves us with considerable uncertainty regarding the definition and classification of the varieties spoken in the vast area between the Central Range (the horizontal line that separates Northern and Southern Peninsular dialects) and the Sierra Morena, that is, the Andalusian border. Traditional Spanish Dialectology, as was previously explained, used to identify two transitional dialects in this territory, one to the west, Extremaduran, and the other in the southeast, Murcian. However, these two transitional dialects were not defined by a specific list of phonetical features, let alone syntactic ones. Most of what is distinctive about them is also found in other southern areas and Andalusia. The main reason to individualise them was the Leonese and Aragonese traces found in the vocabulary of Extremaduran and Murcian, respectively. Again, the prevalent historical bias of Spanish Dialectology comes as an explanation for this division. Contemporary surveys, however, with their careful attention to synchronic facts and grammar, do not award such a special place to the Spanish spoken in these two regions and consider them instead as part of Southern Peninsular Spanish as a whole (Fernández-Ordóñez 2016, p. 393).

While there has traditionally been a plausible approach to the varieties of Extremadura and Murcia and their (phonetic and lexical) characterisation, no matter how insufficient it is considered by current standards, the rest of this southern half of the Peninsula remains a considerable enigma. The upper and middle valleys of the Tagus and Guadiana rivers, the contemporary autonomous community of Castile-La Mancha, still lacks a comprehensive description beyond the customary consideration of a transitional area where southern phonetic traits begin and lexical elements of western and eastern origin are displayed (see, for instance, Moreno Fernández 1996, pp. 25–28). Although there is a relatively recent linguistic atlas for this area, ALeCMan (García Mouton and Moreno Fernández 2003) and an important number of surveys in COSER, there is still much to be completed in order to obtain a better understanding of its language, particularly its grammar. There are useful data available online, but the description and analysis of these data are not abundant nor sufficient. Therefore, a proper characterisation of this southern territory from a geolectal point of view is still necessary.

3. Grammatical Traits of Southern Peninsular Spanish: Castile-La Mancha Spanish

As has been previously demonstrated, the Spanish spoken in Castile-La Mancha has received little attention compared with the varieties of other areas. García Mouton (2011) can serve as a reliable guide to the reasons for this lasting oversight. When referring to the boundaries within the Spanish spoken in Central Spain (Madrid and Castile-La Mancha), she refers to transitional varieties as part of a continuum between Northern and Southern dialects. But she adds an important remark: this area experiences in a particular way the strong influence of the close general Spanish norm of the cultivated classes in Madrid. This influence has been a linguistic reference for the last two centuries and it has clearly contributed to blurring the internal differences pre-existing in popular speech. Consequently, the original heterogeneity in said speech has been somewhat subsumed under sociolectal and stylistic considerations (García Mouton 2011, pp. 82–84; Villena-Ponsoda 2023, p. 277). Thus, the study of these hidden traits was delayed until evidence and data from the ALeCMan and COSER exposed them. This section will be dedicated to the distribution in the region of these peculiarities. Firstly, well-known phonetic traits and their mapping will be reconsidered in order to have a comparative view when defining the limits drawn by grammatical features.

3.1. Some Internal Boundaries of Phonetic Nature

The available works on Castile-La Mancha Spanish assign to the majority of the area three main phonetic features present in non-standard speech: weakening of [s] in syllabic coda, loss of intervocalic [δ] and yeísmo, among some other characteristics of a more limited extension (Moreno Fernández 1996, pp. 216–20; Álvarez Rodríguez 1999, pp. 30–32; Molina 2010, pp. 93–98; Tapias 2020, pp. 62–69). As mentioned above, the three of them constitute the basis for the definition of Southern Spanish pronunciation and therefore ensure the affiliation of Castile-La Mancha varieties to this major division. But a closer examination of the distribution of each of these traits leads to important adjustments of such a clear-cut statement, as will be immediately shown.

Figure 4 actually represents the area of maximum extension of the first feature, the weakening of [s] in syllabic coda, as it shows the pronunciation of [s] before the consonant [k], as in the word mosca ‘fly’, the context that favours most this change:

Unless the process is, in this case, widely distributed across the region, there remains a conservative area to the northeast, corresponding to the southern slopes of the Central Range in Guadalajara, the upper valley of the river Tagus (Cuenca and Guadalajara) and the eastern mountainous border of Cuenca.

The weakening of [s] in the syllabic coda is actually part of a more general change in Spanish that affects the coronal consonants and includes the neutralisation of [l] and [ɾ], of [s] and [θ], and weakening of [d], always in this final position of the syllable. All of these features are also found in Castile-La Mancha, although again with the exception of the conservative north-eastern corner shown in Figure 4 (García Mouton and Moreno Fernández 1994; Fernández-Ordóñez 2016, p. 395).

Secondly, the loss of intervocalic [δ] is present in most Peninsular Spanish dialects, even in the north, but it is in Southern Spain where it seems most extended (Molina 2002). For instance, the consonant is regularly lost in Southern dialects in the past participles in -ido/a. This seems to trace again a relevant phonetic boundary in Castile-La Mancha that can be seen in Figure 5 below, where the area occupied by this region in the Peninsula is represented by a black frame. While the loss of [δ] in -ido/a words is general in western and southern provinces (Toledo, Ciudad Real and Albacete), the consonant is retained in forms like dormida ‘slept’ or vestido ‘dressed’ in north-eastern areas of Guadalajara and Cuenca (Estrada 2012, p. 17; Fernández-Ordóñez 2016, p. 394):

Finally, the merger of the lateral palatal [λ] with the fricative palatal [ʝ] or yeísmo —again an innovative trait that seems to have progressed northwards from Andalusia— is extended in rural varieties throughout most of Castile-La Mancha. Nevertheless, there are some peripheral areas that maintain the original distinction.

As can be seen in Figure 6 below, where red dots correspond to the places where the lateral palatal [λ] is retained and there is no yeísmo, again the north-eastern corner of the region, most of Guadalajara and Cuenca provinces are detached from the general phonetic picture. It is true that this conservative pronunciation is also preserved in some other few points of the mountains of the north-western Toledo province, but the area does not seem so consistent and continuous.

By way of summing up, it is important to underline that there are not any specific phonetic features of significant extension in the region. The sub-standard speech of the region is defined by typical northern–southern traits shared with other southern areas in Spain. The examination of the presence in Castile-La Mancha of three of these characteristics, the most extended ones, as claimed in current linguistic descriptions, has brought us to draw a robust internal frontier. This line would separate most of central and eastern Guadalajara province and central and eastern Cuenca province from the rest. This north-eastern part of Castile-La Mancha can be therefore characterised by a conservative pronunciation and would be better ascribed to Northern Peninsular Spanish from a phonetic point of view. Once again, as so many times in Dialectology, administrative boundaries do not fit linguistic boundaries. There seems to be a conservative extension of northern phonetics to the south of the Central Range in Guadalajara and Cuenca, whereas the other three provinces of Castile-La Mancha share their innovative pronunciation with the rest of the southern half of the Peninsula.

3.2. The Grammar of Spanish in Castile-La Mancha and More Internal Boundaries

Attention will be turned now to grammatical features of Spanish varieties spoken in Castile-La Mancha in order to test and complete these initial considerations on the phonetics of the region. Again, there will be two main guidelines through the following description: the presence of specific regional features of significant extension and the existence of internal boundaries corresponding (or not) to those just found for phonetic features.

The available literature on Spanish in Castile-La Mancha lists some relevant traits regarding syntax and morphology but, contrary to phonetical descriptions, they have not been examined in detail. As mentioned before, there is of course a variable of crucial importance in Peninsular Spanish, third-person clitic pronouns, which is clearly relevant in the region (Moreno Fernández 1996, p. 225; González Pérez 2016, pp. 225–26). Another significant variable in Castile-La Mancha is the presence of a prepositional complementiser de in subordinate clauses that result in the so-called deísmo (and dequeísmo) structures (Camus 2013; de Benito and Pato 2015).

Other features do not seem useful to define geographical varieties, because their distribution appears to follow social and stylistic criteria. As a matter of fact, these features are present in all Peninsular Spanish varieties with no clear geographical pattern of distribution. It is the case of the inverted ordering of clitics, the me se/se me alternation, described in detail in Heap (2006):

(4)	a.	Me	se	seca	todo.	(non-standard ordering)
		1SING DAT	IMPERSclitic	dry3SING PRES	everything
		‘Everything gets dried’
	b.	Se	me	seca	todo.	(standard ordering)
		IMPERSclitic	1st SING DAT	dry3SING PRES	everything
		‘Everything gets dried’.

Other variables of this type can be found in the inventory of grammatical features by González Pérez (2007), for instance, the -ra/-se alternation in the imperfect subjunctive (5) or the synthetic future in -ré vs. periphrastic future with ir ‘to go’ as the auxiliary verb (6), whose sociolinguistic distribution in Castile-La Mancha cities is studied in González Pérez (2016):

(5)	Si	llegaras/llegases	hoy,	Juan	iría	a Madrid.
	If	arrive 2nd SING IMP SUBJ	today	Juan	go3rd SING COND	to Madrid
	‘If Juan arrived today, you would go to Madrid’.

(6)	Vendré	mañana	/ Voy	a	venir	mañana.
	come1st SING FUT	tomorrow	/ go1st SING PRES	to	comeINF	tomorrow
	‘I will come tomorrow’.

Finally, as Fernández-Ordóñez (2023, p. 168) points out, the absence of the preposition de required by the corresponding verb in clausal complements (queísmo), as shown in (7), can be also added to this list of general traits of sub-standard and oral Peninsular Spanish governed by social and stylistic factors.

(7)	a.	Me acuerdo	de que	no	viniste.	(standard)
		remember1st SING PRES	de + COMPLEMENTISER	NEG	come2nd SING PERF
		‘I remember you did not come’.
	b.	Me acuerdo	que	no	viniste.	(non-standard)
		remember1st SING PRES	COMPLEMENTIZER	NEG	come2nd SING PERF
		‘I remember you did not come’.

There are also important syntactic variables that are not well-represented in Castile-La Mancha Spanish. One of them is the transitive/causative use of verbs such as caer ‘to fall’ or quedar ‘to stay’, a western feature that can be found from León, in the north, to Andalusia (Jiménez-Fernández and Tubino-Blanco 2019; Camus and Gutiérrez Rodríguez 2021, p. 165). In ALeCMan, however, this use is only documented in some places of the far west of Toledo or Ciudad Real (García Mouton and Moreno Fernández 2003, map SIN-73 quedar (dejar) la cartera). On the other hand, the transitive use of entrar ‘to go in’ is, as in most of western and southern sub-standard Spanish, well attested in the west, centre and south areas of Castile-La Mancha (García Mouton and Moreno Fernández 2003, map SIN-71 entrar (meter) la leña):

(8)	a.	Cuidado,	no	caigas	el vaso.	(non-standard)
		Attention	NEG	fall 2SING IMP	the glass
		‘Pay attention, do not let the glass fall’.
	b.	Puedes	quedar	el libro	ahí mismo.	(non-standard)
		can 2PRE	stay INF	the book	over there
		‘You can leave the book over there’.
	c.	Entra	el coche	en el garaje.		(non-standard)
		Go in 2SING IMP	the car	in the garage
		‘Get the car in the garage’

Finally, most of the recorded features in inflectional morphology are confined to scattered, isolated areas or scarcely documented. This is the case for some of the alleged archaisms in verbal forms such as non-standard second and third conjugation imperfects in -iba (traíba vs. traía2), conditionals and imperfects in -íe in the south-eastern Toledo province and second-person plural imperatives in -ái, éi: jugái, hacéi vs. cantad, haced…3 (Moreno Fernández 1996, p. 224; Álvarez Rodríguez 1999, p. 34; Pato 2018). Nevertheless, not all morphological features in Castile-La Mancha are of such a limited extension. The deviant desinences of second-person plural (entreguéis, cogís instead of entregais, cogís …4) that are found in eastern and central areas, in La Mancha, are a good example of innovative morphology. They deserve further attention and will be described in detail later in this work.

As has been claimed, most of the aforementioned grammatical features do not seem to be useful to establish significant external boundaries and internal divisions of geographical nature within Castile-La Mancha. We have just seen that they are shared with the rest of spoken Peninsular Spanish or are not sufficiently widespread in the territory. But there are at least three grammatical traits that can be identified for that purpose: They either correspond to important variables that define Peninsular Spanish (clitic pronoun systems), partially define Southern Peninsular Spanish grammar (syntax of complementation) or can be considered as specific regional features due to their considerable extension in all the five provinces of the community (reduced forms of second plural verb desinences). The following discussion will be dedicated to the description of these three traits.

3.2.1. Referential System of Clitic Pronouns in Castile-La Mancha

Spanish, like the rest of the Romance languages, has a system of unstressed or clitic pronouns that are co-referent with direct or accusative complements or indirect or dative complements. The third-person forms are inherited from the Latin ille and marked distinctively for case, gender and number. This close relation to Latin forms justifies the usual denomination of this paradigm as an etymological system of third-person clitic pronouns, and is shown in

Table 1. Etymological system of 3rd-person clitic pronouns ¹.

	Direct Object/Accusative		Indirect Object/Dative
	Masculine	Feminine	Masculine/Feminine
Singular	Lo	La	Le
Plural	Los	Las	Les

¹ Based on Fernández-Ordóñez (1999, p. 1319).

below with its relevant features and forms:

This etymological system is present throughout all the Spanish-speaking areas. It is practically the only one in America, save for some contact areas in the Andes and Paraguay. And it is also the most extended in Spain, where it is predominant in the Canary Islands, Southern Peninsular Spanish and in western and eastern areas of Northern Peninsular Spanish (see the blank spaces in Figure 2 above). The examples in 9–10 serve as an illustration of its actual use in these areas:

(9)	a.	A Juani,	loi	vi	ayer
		To Juan	3rd SING ACC MASC	see1st SING PERF	yesterday
		‘Talking about John, I saw him yesterday’
	b.	A Maríai,	lai	vi	ayer
		To María	3rd SING ACC FEM	see1st SING PERF	yesterday
		‘Talking about Mary, I saw her yesterday’
	c.	A Juan y Pedroi,	losi	vi	ayer
		To Juan and Pedro	3rd PL ACC MASC	see1st SING PERF	yesterday
		‘Talking about John and Peter, I saw them yesterday’
	d.	A María y Anai,	las	vi	ayer
		To María and Ana	3rd PL ACC FEM	see1st SING PERF	a present
		‘Talking about Mary and Ann, I saw them yesterday’

(10)	a.	A Juan/Maríai,	lei	di	un regalo
		To Juan/María	3rd SING DAT	give1st SING PERF	a present
		‘Talking about John/Mary, I gave him/her a present’
	b.	A Juan y Maríai,	lesi	di	un regalo
		To Juan and María	3rd PL DAT	give1st SING PERF	a present
		‘Talking about John and Mary, I gave them a present’

In Castile-La Mancha, where Southern Peninsular Spanish is predominant, this etymological system is the most extended one and is found in Ciudad Real and the easternmost provinces of Guadalajara, Cuenca and Albacete. This said, there exists a vast area where this system is not present and, instead, the use of third-person clitic pronouns corresponds to the innovative referential system, the same that is found in the Northern Peninsular Spanish spoken in the central strip that goes from Cantabria in the north to Toledo in the south (see Figure 2). This innovative referential third-person clitic pronoun system differs from the etymological system inherited from Latin mainly because it does not maintain the original case distinctions and generalises gender—or referential—differences. In

Table 2. Referential system of 3rd-person clitic pronouns ¹.

	Direct Object/Accusative and Indirect Object/Dative
	Masculine	Feminine
Singular	Le	La
Plural	Les/Los	Las

¹ Based on Fernández-Ordóñez (1999, pp. 1360–66).

, a simplified version of this system is represented5.

As Table 2 shows, in some of these “referential” varieties, the use of the original dative clitic pronoun has been extended to cover the reference to accusative masculine in singular and plural (leísmo). In other cases, as will be seen, los becomes the clitic pronoun for dative plural masculine nouns, thus giving way to the so-called loísmo (example 13b) below). And in yet another variety, the original accusative feminine pronoun has extended to cover feminine singular and plural references both in accusative and dative (laísmo). The following sentences in (11) and (12) exemplify these uses of clitic pronouns in the referential system, with or without loísmo:

(11)	a.	A Juani,	lei	vi	ayer	(leísmo)
		To Juan	3rd SING ACC MASC	see1st SING PERF	yesterday
		‘Talking about John, I saw him yesterday’
	b.	A Maríai,	lai	vi	ayer
		To María	3rd SING ACC FEM	see1st SING PERF	yesterday
		‘Talking about Mary, I saw her yesterday’
	c.	A Juan y Pedroi,	lesi/losi	vi	ayer	(leísmo if les is selected)
		To Juan and Pedro	3rd PL ACC MASC	see1st SING PERF	yesterday
		‘Talking about John and Peter, I saw them yesterday’
	d.	A María y Anai,	lasi	vi	ayer
		To María and Ana	3rd PL ACC FEM	see1st SING PERF	yesterday
		‘Talking about Mary and Ann, I saw them yesterday’

(12)	a.	A Juani,	lei	di	un regalo
		To Juan	3rd SING DAT MASC	give1st SING PERF	a present
		‘Talking about John, I gave him a present’
	b.	A Maríai,	lai	di	un regalo	(laísmo)
		To María	3rd SING DAT FEM	give1st SING PERF	a present
		‘Talking about Mary, I gave her a present’
	c.	A Juan y Pedroi,	lesi/losi	di	un regalo	(loísmo if los is selected)
		To Juan and Pedro	3rd PL DAT MASC	give1st SING PERF	a present
		‘Talking about John and Peter, I gave them a present’
	d.	A María y Anai,	lasi	di	un regalo	(laísmo)
		To María and Ana	3rd PL DAT FEM	give1st SING PERF	a present
		‘Talking about Mary and Ann, I gave them a present’

The referential system found in Castile-La Mancha usually includes—together with leísmo and laísmo—the selection of the pronoun los instead of les as accusative/dative plural with masculine reference (see Table 2 above), as these sentences from the COSER survey at the village of Pulgar in Toledo testify:

(13)	a.	Ahora	no	los [los cerdos]		matan	así	(COSER, Pulgar, TO, p. 4)
		Now	NEG	3rd PL ACC MASC		kill3rd PL PRES	that way
		‘They don’t kill them [the pigs] that way now’
	b.	Los [los novios]			quitemos	toda la ropa		(COSER, Pulgar, TO, p. 15)
		3rd PL Dat MASC			take away1st PL PERF	all the c`lothes
		‘We took away all their [the newlyweds’] clothes’

The following Figure 7 represents the extension of leísmo in ALeCMan. It refers to its use in a sentence where the clitic pronoun has human masculine reference, which is the most widespread variant of leísmo:

As shown in Figure 7, this maximum extension of leísmo covers most of the Toledo province, western Guadalajara and some north-western areas in Ciudad Real. It is important to mention that this human reference leísmo is part of Standard Spanish and ‘prestigious’ Madrid Spanish (Fernández-Ordóñez 1999, pp. 1386–90). Thus, the pressure exerted on Castile-La Mancha’s formal speech could account for this greater diffusion (García Mouton 2011, p. 83).

Precisely because of this, we would expect laísmo, not part of Standard Spanish, to be less widespread in the area. This is indeed the case, as Figure 8 demonstrates:

However, both Figure 7 and Figure 8 clearly draw a north-western area in Castile-La Mancha whose syntax includes northern pronominal innovations. It could be described as a kind of linguistic wedge that crosses the Central Range through the mountain passes of Gredos and Guadarrama in Ávila and Segovia and breaks into Madrid and Toledo in the southern Spanish plateau. It is another of the relevant internal divisions inside Castile-La Mancha, but one of a really different nature to those previously considered. It corresponds to an innovation that arrived from the north that leaves in the south another large area of conservative syntax that partly coincides with the area of innovative phonetics.

To resume and complete the description of the presence of the referential system in Castile-La Mancha beyond this preliminary consideration on leísmo of human reference and laísmo, we should note that leísmo in the oral varieties of this north-western corner of the region always includes the reference to all count names, both animate and inanimate, as was represented by the canonical display of referential system in Table 2. But the coincidence with the pronoun systems that are found in the north-central strip of Figure 2 does not finish here. The varieties spoken in those north-western areas present another notorious feature, the existence of a specific set of pronouns to refer to non-count nouns (Fernández-Ordóñez 1999, pp. 1360–63), as can be seen in

Table 3. Referential system with non-count reference in north-western Castile-La Mancha ¹.

	Direct Object/Accusative and Indirect Object/Dative
	Count nouns		Non-count nouns
	Masculine	Feminine	Lo
Singular	Le	La
Plural	Los	Las

¹ Based on Fernández-Ordóñez (1999, pp. 1360–66).

.

Some examples from the COSER surveys in central Toledo province serve as an illustration of this use of the form lo as the pronoun to refer to non-count nouns:

(14)	a.	La sangrei,	loi	cogían	las mujeres	(COSER, Pulgar, TO, p. 2)
		The bloodFEM	3rd SING ACC	collect3rd PL IMPERF	the women
		‘As for the the blood, the women collected it’
	b.	El caldoi,	loi	movían…		(COSER, Pulgar, TO, p. 3)
		The brothMASC	3rd SING ACC	stir3d PL IMPERF
		‘As for the broth, they stirred it…’
	c.	Loi	llaman	leche Pascuali		(COSER, Pulgar, TO, p. 8)
		3rd SING DAT	call3rd PL PRES	milk PascualMASC
		‘They call it milk Pascual (a trademark)’

This kind of pronominal agreement for count and non-count nouns is shared by all the territory where the referential system is found in Spain, the shaded area of Figure 2. The village of Pulgar in Toledo province, the provider of examples (13) and (14), is located in that area, near the border with the southern province of Ciudad Real.

3.2.2. Complementisers and Deísmo in Castile-La Mancha

The syntax of many varieties of spoken and rural/sub-standard Spanish in Castile-La Mancha incorporates a feature that is worth describing and mapping. This feature is the so-called deísmo, the insertion of the preposition de in infinitive sentences that was already presented in Section 1 with example (2). This example is again repeated below (15a) together with two more sentences with this construction:

(15)	a.	Me	hacen	de	reír	(with causative verb)
		1st SING ACC	make3rd PL PRES	of	laughINF
		‘They make me laugh’
	b.	Me	sintieron	de	venir	(with perception verb)
		1st SING ACC	hear3rd PL PERF	of	comeINF
		‘They heared me come’
	c.	Intentó	de	disparar		(with control verb)
		try3rd SING PERF	of	shootINF
		‘He tried to shoot’

As these sentences show, this non-standard de can precede different types of infinitive clauses. It is really common in causative clauses or with perception verbs (15a-b), but its original locus corresponds to typical infinitives controlled either by subjects, as the one in (15c), or by complements (Camus 2013, pp. 22–28). This latter context seems to correspond to the same construction of control verbs with de that is sometimes documented in Medieval Spanish and can still be regularly found in French, Italian or Catalan, where it is part of the standard variety (Camus and Gómez Seibane 2015). The construction in sub-standard Spanish is, thus, a continuation of this usage that was already present in the Middle Ages but has been lost in Standard Modern Spanish. It has, however, continued to live in some dialects (15c). Even more, in most of these dialects, the construction would have been extended in recent centuries to other contexts such as those with causative or perception verbs (15a,b).

The diffusion of the deísmo pattern in Castile-La Mancha can be seen below in Figure 9, built on data from García Mouton and Moreno Fernández (2003):

As Figure 9 shows, the type of deísmo most extended in the region corresponds to infinitive clauses with causative verbs (hacer de reír: blue dots), which is present even in the mountainous and isolated east areas of Guadalajara and Cuenca provinces. But, save for this case, deísmo is only common in the Toledo and Ciudad Real provinces, south-western Guadalajara and the western half of Cuenca. In the southwestern province of Albacete, however, it is rare and can be exceptionally found in some southwestern villages.

Data from COSER seem to show similar results, except for the fact that there are some more examples of deísmo from all throughout Albacete (de Benito and Pato 2015, p. 33). However, it should be noted that these and other examples from Guadalajara and Cuenca in COSER are mostly limited to infinitive clauses with control verbs (specially gustar ‘to like’). Considering data from both sources, deísmo should be described as a common feature of the Spanish varieties of Castile-La Mancha with a stronger foothold in western and central areas.

And it is worth mentioning that deísmo is also a common feature of other southern Peninsular dialects. Following de Benito and Pato (2015, pp. 33–34) and their account of COSER data, examples with de in front of infinitive sentences are regularly found in Andalusia, Extremadura and Murcia, in addition to in Castile-La Mancha. They have also been exceptionally documented in Northern Spain, where this old construction seems to have almost disappeared.

3.2.3. Reduced Variants of Second-Person Plural Desinences

With this simplified label, we will refer to the preference shown by different varieties of Peninsular Spanish for innovative forms of this desinence. Forms of second-person plural have been generally studied in relation to the types of voseo morphology in America that become an alternative to the original second singular desinences corresponding to the second singular pronoun tú (Real Academia Española and Asociación de Academias de la Lengua Española 2009, §§ 4.4d and 4.7). As opposed to these, the Peninsular Spanish non-standard desinences are still desinences of the second plural original pronoun vosotros and differ substantially from most American voseo variants, and therefore can be easily recognisable. Curiously enough, they have received little attention in the current literature (Hernando Cuadrado 2009, p. 175), and the reference grammar does not even mention them (see Alcoba 1999, pp. 4924–26; Real Academia Española and Asociación de Academias de la Lengua Española 2009, §4.4). Some of these alternative desinences, for both present indicative and present subjunctive of verbs entregar ‘to deliver’, querer ‘to want’ and salir ‘to go out’, are listed in

Table 4. Reduced desinences for 2nd-person plural verb forms. Present tenses.

	PRESENT
	INDICATIVE	SUBJUNCTIVE
1st conj.: verbs in -ar	vosotros entreg-áis > entregu-éis, entregu-ís	vosotros entregu-éis > entregu-ís
2nd conj: verbs in -er	vosotros quer-éis > quer-ís	vosotros quer-áis > quer-éis, quer-ís
3rd conj.: verbs in -ir	vosotros sal-ís > sal-éis

below:6,7

As can be seen, new forms in this Table 4 show some kind of phonetic reduction and tend to simplify the variation in this second-person plural morphology to only two possibilities (-éis and -ís) in the most frequent paradigm or just one (-ís) when even the specific desinence for verbs in -ar suffers an extreme phonetic reduction.

These new present desinences are actually also found in other tenses, such as imperfect indicative (although only in second and third conjugation verbs), conditional and future:

In these latter tenses in

Table 5. Reduced desinences for 2nd-person plural verb forms. Imperfect, conditional and future.

	IMPERFECT	CONDITIONAL	FUTURE
2nd/3rd conj.: verbs in -er, -ir	querí-ais > queri-éis	querrí-ais > querrí-eis	querr-éis > querr-ís

, desinences are derived from a second conjugation verb (haber) in the imperfect and present tense, respectively. That explains the coincidence of conditional desinences with the imperfect ones and the one between future desinences and those of the second conjugation verbs in the present indicative (see Table 4). This coincidence allows us to limit the attention from now on to just the reduced desinences of present tenses in Table 4, and particularly to the better-documented reductions, that is, the new forms in -ís that come from -éis, corresponding either to present indicative of second conjugation verbs in -er or present subjunctive of first conjugation verbs in -ar.

Data from old surveys of ALPI already showed the diffusion of these forms in -ís (querís < queréis), mainly in eastern areas of the Peninsula. The forms were found in all three provinces of Aragon except for central and northern Huesca. It was also frequent to find them in the neighbouring Castilian province of Soria and eastern Cantabria, to the north. And they were also very common in Castile-La Mancha, with the only exception of the province of Toledo.8 Regional Spanish atlases developed in the second half of the 20th century broadly confirm this distribution. Almost every form in Table 4 and Table 5 above can be found in those surveys, mainly in Aragon and Cantabria (Alvar et al. 1979–1983; Alvar 1995, respectively) but also in eastern Andalusia (Alvar et al. 1961–1973). The corresponding maps confirm an eastern distribution of this change in second-person plural desinences in Peninsular Spanish that more recent data from COSER corroborate with actual informal spoken records (Fernández-Ordóñez (dir.) 2005-). Examples like conocís (<conocéis ‘you know’), entendís (<entendéis ‘you understand’), habís (<habéis ‘you have’), querís (<queréis ‘you want’), sabís (sabéis ‘you know’), tenís (<tenéis ‘you have’)… are found in the COSER online files. They come from Aragon, eastern Cantabria, Soria and Jaén, the same places where they had been previously documented. But they also come in large numbers from Castile-La Mancha (mostly from locations in Guadalajara, Cuenca, Albacete and Ciudad Real, but also from eastern Toledo).

The abundant presence of these reduced desinences is, as expected, proved by the corresponding questions and maps in the ALeCMan (García Mouton and Moreno Fernández 2003). The following Figure 10 serves as a meaningful example:

Data of Figure 10 refer to the most common reduced variant of second-person plural forms of the present indicative of verbs in -er (coger ‘to take’: cogéis > cogís). Other questions in ALeCMan regarding these forms receive few answers and therefore are less representative, unless they show an identical distributional pattern. As Figure 10 shows, cogís—and other similar reduced forms—is present everywhere in the three eastern provinces of Guadalajara, Cuenca and Albacete and in Ciudad Real. In Toledo, these forms are not completely unknown, and they can sometimes be found in eastern areas, such as Quintanar de la Orden (point of survey code TO-606), a place belonging to the natural region of La Mancha, as shown in Figure 10.

This distributional pattern draws a new linguistic space within Castile-La Mancha that covers northern, eastern and southern mountainous borders and the great central plain of La Mancha. This might be explained by the fact that this vast area has been open to changes coming from the northeast (Soria) and east (Aragon). This kind of influence is of a very different nature to the one seen in Section 3.2.1, which explains the referential clitic system in the north-western corner of the region. The distribution across this large eastern area has to do again with medieval repopulation, but also with enduring contact through tracks of annual cattle migrations across the mountains of the Central Range to the plains of La Mancha. It is probably part of a well-known lexical area shared by Aragon and eastern Old Castile, eastern Castile-La Mancha, eastern Andalusia and Murcia (Catalán [1975] 1989). In fact, it cannot be a surprise that this common linguistic space also emerges when dealing with inventories, even if they are of a morphological nature, such as inflectional desinences.

To summarise, the detailed attention to grammatical features of Castile-La Mancha Spanish varieties has revealed new internal divisions that do not coincide with the boundaries traditionally defined by the phonetical features considered in the previous subsection. The heterogeneous configuration of grammar-based limits in this community overthrows the reducing idea of a transitional dialect in Castile-La Mancha. These limits draw new linguistic spaces, exposing the need for new proposals that account for the Spanish spoken in this region and a closer examination of past definitions.

4. Towards an Accurate Definition of Castile-La Mancha Varieties

A renewed consideration of Spanish Peninsular dialects in general and the varieties of Central Spain and Castile-La Mancha, in particular, should be built on a balanced inventory of the relevant geolinguistic variables. This involves the inclusion of distinctive grammatical (syntactic and morphological) features, together with phonetic and lexical traits. The importance of grammatical characteristics should be properly considered, as they are at the core of everyone’s speech. As is generally accepted, grammatical change and its diffusion depend less on external factors or occasional conditions than phonetic and, especially, lexical innovations do.

Firstly, we need to question the foundations of the idea of Castile-La Mancha Spanish being defined as a transitional variety. As was shown in Section 3.1, this definition applies when only phonetic traits are considered. The three phonetic features described in Section 3.1 (weakening of [s] in coda, weakening of intervocalic [δ] and yeísmo) are known to have spread northward from Andalusia through the provinces of Castile-La Mancha since the 16th century. It is no wonder, then, that Castile-La Mancha phonetics represent a gradient between Northern Spanish pronunciation and Andalusian phonetics. In fact, Southern Spanish can be described as beginning in this region, to the south of the Central Range in Guadalajara, on the valley of river Henares, and in eastern Cuenca, where the great plains of La Mancha spread westward (see Figure 6). The more we move southward and westward in this region, the more the pronunciation incorporates the innovative traits that define Southern Spanish varieties. In this respect, describing most of Castile-La Mancha’s varieties as transitional seems appropriate. But once we introduce the essential elements of their grammar, this is no longer the case.

As has been shown in Section 3.2, Castile-La Mancha’s grammar is built on various elements whose nature and distribution conform to different explanations and criteria. Some of them have limited extension and do not serve as defining traits for the language of the region. They are actually better explained as an appendix or continuation of a neighbouring dialect that reaches some of the borders of this community. In this respect, we have referred to the presence in some western areas of Toledo and Ciudad Real of the transitivisation of verbs like caer ‘to fall’ or quedar ‘to stay’ and the presence of imperatives in -ái or -éi. They seem to correspond to the easternmost bridgeheads of Leonese or Extremaduran characteristics. After all, the influence of these western varieties is also apparent in the vocabulary of these same areas of Castile-La Mancha (Moreno Fernández 1996, p. 226).

Other elements are shared with general sub-standard Peninsular Spanish and cannot specifically characterise the Spanish spoken in the region. They are usually remnants of old uses that are still common in many rural areas in Spain. Thus, it is not surprising that they can still be found in the speech of elder people in a region like Castile-La Mancha, where peasant life was so relevant until recently. Among this group of features, we mentioned the me se/se me alternation, but also the -re/-se forms alternation in imperfect subjunctive or queísmo (García Mouton 2011).

Finally, there are grammatical variables that ideally suit our purpose of defining geolinguistic varieties on the basis of relevant extension and significant role in the configuration of geographical variation in (Peninsular) Spanish. Three features have been selected: third-person clitic pronoun systems, deísmo and reduction in second-person plural desinences. Their nature and distribution in Castile-La Mancha have been described in detail so that different linguistic spaces can be identified. As we will see more at length immediately, they point to the existence of different linguistic forces and influences of varying provenance, time and duration. And this heterogeneity in their linguistic configuration is precisely the key aspect of Castile-La Mancha Spanish varieties, aside from their prima facie ascription to Southern Peninsular Spanish.

The consideration of the third-person clitic pronouns in Castile-La Mancha splits the region in two. There exists a large north-western area in Toledo and northern Ciudad Real that adopts systematically the referential system, thus delimiting an area open to this type of syntactic innovation coming from the northern side of the Central Range. The rest of the region adheres firmly to the older etymological system that is to be found everywhere in the Peninsula, save for the central northern stripe that goes from Cantabria to north-western Castile-La Mancha. As the referential system has been in use since medieval times, the emergence of this innovative north-western area can be explained through repopulation and permanent contact due to transhumance across the mountains of Gredos (Fernández-Ordóñez 1994, p. 123).

As for the distribution of deísmo, it could be attributed to southern innovative influences on a surviving archaic feature. That would explain its residual presence to the north and west of the region as well as its robust existence to the east. It is worth noting that the modern extension of this construction with causative and perception verbs, which is particularly common in eastern Castile-La Mancha, is also common in Extremadura, Andalusia and Murcia (de Benito and Pato 2015, p. 34). That would suggest a modern southern origin with the same diffusion path as other changes with similar distribution, that is, the phonetic ones discussed in Section 3.1.

Finally, a vast eastern area can be clearly singled out by the distribution in Castile-La Mancha of the reduced forms of second-person plural desinences. Again, the explanation is similar to the one for the north-western area of the referential clitic system: repopulation and contact through shepherding movements. In this case, however, the diffusion of these innovative inflectional forms may be given a later date, because the original form of these second-person plural desinences with the in-between Latin consonant (-ades, -edes, -ides < -ATIS, -ETIS, -ITIS) was still in use until the early 16th century and even later (Girón Alconchel 2004, pp. 865–66; de Bustos Gisbert 2007). This eastern peninsular area of shared linguistic heritage, as was already noted, is then also acknowledged on the basis of grammar and not just of vocabulary, the original locus for its identification (Catalán [1975] 1989; Moreno Fernández 1996, p. 226).

These previous considerations and comments are meant to serve as a starting point in the definition of the Spanish varieties of Central Spain. In the case of Castile-La Mancha, we would like to underline the heterogeneous nature of the Spanish spoken in the rural areas of this region, the development of which has been subject to different forces. The original marks left by medieval repopulation and the lasting contact with the northern half of the Peninsula through transhumance in the east and northwest have been partially blurred by the more recent and successful influence coming from the south, from Andalusia. And, more recently, this complex scenery is experiencing the increasing pressure of the Madrid model of Standard Spanish.