The contemporary autonomous region of Extremadura comprises the provinces of Cáceres and Badajoz.  Three Andalusian towns contain residents who also claim to speak extremeño:  Cuenca (Córdoba), Encinasola (Huelva), and Jara (Sevilla) (Ariza Viguera 1981; 1987a:  24).  Although sharing historical and cultural referents, the contemporary speech of the two provinces exhibits as many differences as similarities, and there is little feeling of being Extremeños among the residents of both provinces.  The speech of Cáceres bears some similarities with Castile, despite considerable phonetic reduction, while Badajoz Spanish more closely resembles the speech of southwestern Andalusia, especially Seville.  Along the border with Portugal some traces of Portuguese creep into Extremadura Spanish, while in the northwestern part of Cáceres some Leonese traits can still be found.  Although the most picturesque and regionally identifiable forms of Extremadura speech have all but disappeared, the region (particular in Cáceres) retains the memory of castúo, the Leonese-influenced vernacular speech populariized (together with its designation) by the foremost regionalist writer, Luis Chamizo (1982).  Historically, Extremadura did constitute a definable dialect zone, although most scholars would consider Extremeños to be a transitional dialect of Castilian/Spanish rather than a separate language or even self-contained regional dialect.  Thus Zamora Vicente (1967:  332-336) lumps Extremadura in the `hablas de tránsito' while Llorente Maldonado (1995:  88) groups the Spanish of Extremadura, Murcia, the Canary Islands, and Andalusia as simple continuations of Castilian:  `no son otra cosa que la continuación del antiguo dialecto castellano (hoy convertido en la lengua española), continuación que presenta algunos fenómenos fonéticos distintos de los fenómenos del español común y del español coloquial del resto de las regiones españolas.'

Surrounded, historically or currently, by Leonese, Portuguese, and the Spanish dialects of Castile and Andalusia, Extremadura has absorbed numerous influences and constitutes a variegated mosaic of highly regionalized features more than a unified dialectal profile.  In the northwestern part of Cáceres province, the traditional speech has been strongly influenced by Leonese and traces remain to the present day.  Along the border with Portugal (e.g. San Martín de Trevejo, Eljas, Valverde del Fresno) some bilingual contact phenomena are to be found, and in at least one town on the Spanish side of the border (Olivenza, in Badajoz province), Portuguese predominates over Spanish (cf. Alvarez Martínez 1996; Ariza Viguera 1980; Flores de Manzano 1988, 1992; Martínez Martínez 1974, 1983; Onís 1930; Vasconcellos 1933; Maia 1970, 1977; Viudas Camarasa 1982).

Extremadura had an early Celtic and then Roman population, and fell under Moorish dominance after the ninth century.  When the Mozarabs were expelled from Andalusia in 1125 a substantial number arrived in Extremadura (Ariza Viguera 1987b:   51), but the real linguistic reshuffling in Extremadura took place as a consequence of the Christian reconquest.  At the beginning, the northern portion of modern Cáceres province was divided between Castile and Leon.  Leonese forces recaptured Cáceres, Mérida, and Badajoz by 1229, while Trujillo held out until 1232, when Castilian armies took the town.  Since the crowns of Castile and Leon had become united in 1230 (under King Fernando III), Castile and Leon were no longer rivals in the reconquest.  Sevilla and Toledo became the two linguistic and cultural centers of attraction for Extremadura, until Madrid took over as the linguistic center of Castile in 1561.  Much of Extremadura was repopulated from Leon, whence the abundance of Leonese features in the extremeño vernacular; to the south, resettlement from Castile and Andalusia was more common.

Currently, the two provinces of Badajoz and Cáceres share few common cultural and linguistic threads, and many residents of the autonomous region of Extremadura have indicated to the present writer that they find little use in the designation as extremeños.  Ariza Viguera (1987a) explores the relatively low level of awareness of Extremadura speakers of their dialect:


... el extremeño no tiene conciencia, o mejor dicho, buena conciencia de su habla regional.  Quiere ello decir que, qun cuando muchos hablantes respondan que hablan extremeño, la idea más generalizada es que lo que ellos hablan es un mal castellano; de ahí que tiendan--en mayor o menor medida ... a "corregir" los fenómenos lingüísticos que en su conciencia o subconciencia de yhablantes consideran como "incorrectos".  Esto es claro cuando se realizan encuestas dialectales o, simplemente, cuando se observa a un determinado tipo de hablantes:  en numerosas ocasiones, el encuestado evita pronunciar como él suelo hacerlo, tendiendo a realizar la norma castellana; por otra parte, muchos estudiantes universitarios dejan los usos de sus pueblos para incorporarse al de la ciudad en la que estudian ... muy frecuentemente los estudiantes universitarios afirman que ellos hablan mal el castellano, cuando la realidad es muy otra:  hablan bien el extremeño ... (22)


Ariza Viguera clarifies that whereas not all or even most residents of Extremadura are ashamed of their way of speaking, there exists little regional awareness of speech tendencies.  Mass media present a homonogenized and standardized Spanish throughout Spain, although some regional characteristics do slip in, and in schools, `aun cuando el profesor sea extremeño---tenderá a enseñar un castellano normativo, a escribir la S cuando en el habla es /h/, es decir, una aspirada, etc.  Por un lado ese sentimiento de ser un pueblo marginado ... puede haber creado o ayudado a la "automarginación lingüística"; por otro la falta de una entidad alutinadora--Cáceres ha mirado siempre a Madrid o a Salamanca, Badajoz a Sevilla--puede también haber contribuído al hecho en cuestión' (p. 23).

The rapid disappearance of high regional vocabulary items in recent decades has also contributed to the erosion of a linguistic consciousness in Extremadura, this despite the high regard in which the regionalist poetry of José María Gabriel y Galán and especially of Luis Chamizo.  Ariza Viguera (1987a:  23) explains:  `ello no se debe tanto a sus indudables calidades literarias como a un motivo mucho más psicológico y humano al mismo tiempo:  el que ambos empleen un lenguaje con el que el hablante se siente identificado ... el habla extremeña se "consagra" como habla literaria, lo cual palia, mitifica, en el subconsciente del hablante, su también inconsciente o latente "complejo de hablar mal" ...

Many of the linguistic studies of Extremadura and other regions is characterized by the dialect atlas parameters of 19th century Europe.  The emphasis is on small towns and more frequently rural villages.  Ideal informants are elderly, nearly or totally illiterate, with no residence or even travel outside of the region.  At a time when Spain was still a patchwork of still viable rustic regional dialects, this approach may have been warranted, but by the second half of the 20th century Peninsular Spanish was increasingly dominated by urban middle-class norms, diffused by mass media, improved public education, military service, and greater social and geographical mobility.  Zamora Vicente (1943:  11-12) typifies the reliance on older, rural, illiterate informants:  `He buscado siempre gentes de la mayor auctotonía posible:  naturales, hijos de naturales, con deficiente cultura primero, a ser posible, labriegos; cuando hombres, he preferido los que no habían hecho servicio militar, que, como es sabido, contribuye mucho a la deformación lingüística ...'  Many of his subjects were over 70 years of age, although some were younger.  This means that research conducted around 1940 was actually picking up traces of regional language which became solidified as early as 1880.  This same conclusion holds for many regional monographs published in Spain even in recent decades, so that what appears to be a strikingly diverse set of contemporary dialects in reality represents the situation in isolated rural areas more than a century ago.

Scholarship on the Spanish of Extremadura

Viudas Cmarasa (1987a) provides a recent summary of studies on Extremeño Spanish.  Classic studies of Extremadura Spanish include Fink (1929), Bierhenke (1929, 1932), Zamora Vicente (1943b, 1950), Santos Coco (1936, 1940, 1941, 1942, 1944, 1952), Canellada (1941), Cummins (1974), Velo Nieto (1956).  Viudas et al. (1987) contains more recent studies.  Other recent works include Montero Curiel (1996), Cortés Gómez (1979), Indiano Nogales (1977), Lumera Guerrero (1979, 1988, 1992), Ariza Verguera (1992).  Bibliographical surveys are contained in Viudas Camarasa (1987a) and Alvarez Martínez (1996).

Contact with other languages

As a transitional area between Spain and Portugal, Extremadura once contained a not inconsiderable black population, as black slaves and workers were transferred from southern Portugal to Andalusia (Cortés Cortés 1987).  The tangible results of this community on local varieties of Spanish were probably minimal, but particularly in Badajoz, residents became accustomed early on to Africans' attempts to speak Spanish.  Diego de Badajoz is one of the first Peninsular writers to offer literary imitations of the habla de negros (cf. Barrantes (1882, 1886), in a series of farsas, including the `Farsa teologal,' `Farsa del moysen,' `Farsa de la hechicera,' and `Farsa de la ventera.'  All were apparently composed between 1533 and 1548, which, extrapolating backwards, would make these among the very first Afro-Hispanic literary texts.  In addition to being one of the first Spanish writers to employ Africanized language, Sánchez de Badajoz is one of the few Spaniards to use literary habla de negro language who did not live and work in Seville and other southern areas where the African slave population was predominant.  Diego Sánchez was born near the city of Badajoz, probably in Talavera la Real, a few kilometers away, and almost on the Portuguese border.  His date of birth is not known, but was probably in the last decade or two of the 15th century.  In all of his dramatic writings, Diego Sánchez demonstrated a keen awareness of popular speech.  The most common linguistic resource is the abundant use of sayagués by his rustic characters, the stereotypical lenguaje pastoril of Spanish Golden Age literature, combining popular traits with regional Leonese/Extremeño forms.  This language was particularly appropriate for Diego Sánchez, since the rustic speech of southern Extremadura bore a great similarity to the partially idealized sayagués.  The African characters who appear in Sánchez' plays based their pidginized language as much on this rustic vernacular as on the learned Spanish of their masters, and the `African' component of their language must be evaluated against the backdrop of these nonstandard rustic variants.  In particular, sayagués frequently interchanged /l/ and /r/ in several positions, dropped many final consonants, shortened words by dropping unstressed initial and final syllables, and employed many lexical and morphological variants which bear a greater similarity to Portuguese than to Castilian.


(1)  All of modern Extremadura is characterized by extremely high rates of aspiration and deletion of syllable- and word-final /s/ (and /θ/).  In most of the region the rates of weakening are as high as in Andalusia, but in northeastern Cáceres province the rates of weakening diminish somewhat.  Table 1 gives some comparative figures.  Although there has been no significant variational research on the sociolinguistics of Extremadura Spanish, there is nothing to suggest that weakening of /s/ is in any way stigmatized or socially stratified within Extremadura.  Salvador Plans (1987a:  26) suggests that loss of final /s/ may lead to a phonological differentiation of lax vowels (resulting from loss of /s/) to mark plurality, as opposed to tense vowels, in singular forms, much as occurs in eastern Andalusia.  I have encountered no convincing cases of this phenomenon, but the matter remains open to further study.  In some instances, a glottal stop may replace a deleted syllable-final /s/ (Cortés Gómez 1979:  29).

(2)  Virtually all of the province of Cáceres (except along the border with Portugal-Salvador Plans 1987a:  31) and part of Badajoz province distinguishes /s/ and /θ/  The city of Badajoz exhibits seseo, with the same alveolar [s] found in western Andalusia.  Further to the north in Badajoz province and Cáceres, some use of apicoalveolar [s] is found.  In the town of Malpartida de Plasencia, whose local dialect is known as chinato, intervocalic /θ/ has traditionally been realized as [δ]; this voicing is sometimes carried over to intervocalic /s/ (Catalán 1954, Espinosa 1935, Viudas Camarasa 1987b).  Ariza Verguera (1992b) has determined that in Malpartida de Plasencia voicing of intervocalic /s/ and / / is no longer current, but has detected this process in the town of Serradilla.  In this same area, uniform ceceo (realization of /s/ as [ ]) is also found (Salvador Plans 1987a:  31).  The town of Fuente del Maestre (Cáceres) is an island of seseo surrounded by towns which distinguish /s/ and /θ/ (Salvador Plans 1987a:  31).

(3)  Final /n/ is velarized throughout Extremadura, with loss of /-n/, leaving a nasal or oral vowel, also occurs.  In the extreme northwest of Cáceres province, velarization is less common (Cummins 1974:  64).

(4)  As in the rest of southern Spain, intervocalic /d/ is weak and often disappears, especially in the suffix -ado.  Word-final /d/ routinely falls.  In traditional castúo speech, word-initial /d/ also frequently disappeared:  icí < decir, escudiar < descuidar, etc. (Santos Coco 1936:  13).

(5)  Confusion of syllable- and word-final /l/ and /r/ is usual in Extremadura.  In the capital cities of Cáceres and Badajoz, both /l/ and /r/ normally fall phrase-finally in casual speech.  Aspiration of preconsonantal /r/ sometimes occurs (Cortés Gómez 1979:  28).  The stereotyped pronunciation of word-final /r/ as [l] is still found in rural regions, especially in Badajoz province and in the north of Cáceres.  Syllable-final / / or /s/ may also emerge as [l]; the entire suffix may also reduce to -ajo:  noviazgo > novialgo > noviajo, portazgo > portalgo > portajo (Zamora Vicente 1967:  334).

(6)  The capitals of Cáceres and Badajoz no longer distinguish /y/ and /λ/, but the later phoneme remains in some rural areas and smaller towns in Cáceres province.  In Badajoz province only a few tiny enclaves of /λ/ remain (Salvador Plans 1987:  32; Hidalgo Caballero 1977).  In some rural regions groove fricative or rehilada pronunciation of /y/ is still heard (Zamora Vicente 1943:  24) but, as in the rest of Spain, this is quite infrequent.  Sporadically, the groove fricative pronunciation [zh] of /y/-/λ/ is found in Extremadura (Salvador Plans 1987a:  33-4).

(7)  In Badajoz and southern Extremadura, the posterior fricative /x/ is a weak aspiration [h].  In Cáceres and nearer the Castilian border, a velar [x] or uvular [X] appears.

(8)  Traditional castúo Spanish, heavily influenced by Leonese, tended to realize final unstressed /e/ as [i] and unstressed /o/ as [u].  This trait appears in all folkloric texts, but is limited to certain geographical regions and sociolinguistic strata nowadays.  Salvador Plans (1987a:  28-9) describes a swath cutting diagonally from northwest to southeast in Cáceres province (not reaching the southeast border) in which raising of final unstressed vowels may still be heard

(9)  There are still a few regions of Extremadura that distinguish /b/ and /v/ (Ariza Verguera 1992a, Alvarez Martínez 1996). 

Morphological features

(1)  The traditional diminutive ending in Extremadura is -ino.  In contemporary urban speech, -ito and even -illo are making inroads.

(2)  Traditional Extremadura Spanish combines the definite article and the possessive:  el tu padre, los mis muchachos, la nuestra iglesia (Zamora Vicente 1967:  335; Salvador Plans 1987b:  40).

(3)  It is still possible to hear the archaic imperative of first-conjugation verbs in -ai:  echai, hablai, etc.

(4)  Vernacular castúo speech contains apocopated verb forms, such as estuvon < estuvieron, quisun < quisieron, dijon < dijeron, vinon < vinieron.

(5)  The analogical pronouns mos (nos), mosotros (nosotros), muestro (nuestro) can be found in rustic speech.

(6)  Loss of final /s/ has resulted in lexical restructuring of some plural forms in rustic vernacular, much as in rural Andalusia:  los árboles > lo(h) árbo, etc. (Salvador Plans 1987b:  40).

(7)  For the town of Higuera de Vargas, Cortés Gómez (1979:  37) reports that female residents when speaking to children sometimes use the definite article uno instead of un:  `eres uno potrino cerril,' `dame uno beso,' etc.

Syntactic features

(1)  Vernacular speech sometimes places proclitic pronouns before affirmative commands:  se siente usted, se vaya Ud., etc. (Zamora Vicente 1943:  43), much as occurs in some varieties of Portuguese.

(2)  The verb caer can be transitive, meaning `drop, spill, knock over':  caí un vaso de agua `I spilled/knocked over a glass of water.'  Haber can be used transitively meaning `have':  no habemos mucho trabajo `we don't have a lot of work.'  Whether this is an archaic carryover or an innovation is not clear (Salvador Plans 1987b:  43).  Similarly, quedar can be used transitively meaning `to leave behind,' entrar can mean `to place inside.'

(3)  In some areas, overt preposed subjects of infinitives can be heard:  `encima de yo haberle dicho' (Cortés Gómez 1979:  38).

In the northwestern portion of Cáceres province, Leonese traits are still found in combination with Extremadura characteristics (e.g. Cummins 1974).  The Extremeño features of aspiration or loss of syllable-final /s/, elimination of word-final /l/ and /r/, are combined with raising of final unstressed vowels and significant modification of tonic vowels.  Thus, Cummins (1974:  34-5) finds continued evidence of non-dipthongized mid vowels or of the diphthong uo:  nuobi (nueve), fuora (fuera), cuorpo (cuerpo), simpri [sjimpri] (siempre), quiri [kjiri] (quiere), etc.  Reduction of diphthongs (pacencia < paciencia, tútano < tuétano) is also found, as in other vernacular dialects of Spanish.  Metaphony, that is, raising of tonic mid vowels due to the presence of a final high vowel, is still found in northwestern Extremadura:  [tjini] < tiene, utru < otro, sombriru < sombrero, etc. (Cummins 1974:  39).  Throughout the region, final unstressed mid vowels are routinely raised:  frutu < fruto, lechi < leche, eneru < enero.  Loss of word-initial consonants is frequent in casual speech:  eθe < desde, ehkanso < descanso, ahtanti < bastante, etc.  Intervocalic /d/ is consistently lost, while /x/ is a weak aspiration and sometimes disappears (Cummins 1974:  57).  There is some vocalization of /d/ before liquids, suggesting resyllabification into the coda of the preceding syllable:  mairi < madre, mairí < Madrid, poiríu < podrido (Cummins 1974:  62).  There is considerable pronunciation of /y/ as groove fricative [ž] throughout the region.  The imperatives velahí < velo ahí, velaquí < velo aquí are sometimes used without imperative or adverbial force, sometimes to mean `and that's it' (Cummins 1974:  99-100):


aquí, pu se cultivan velaquí lechuga ...

la rahtra velahí, ¿no be uhté?

Se le echa ehtierco y velahí


In constructions involving parecer, yo is used instead of a ... me:  yo parece (paece/paí) que ...

Examples of Extramadura/castúo Spanish

Voy a contalti aqueyo, mujel mía;

te voy a icir las gielis de mi alma,

el por qué de chiquinu

siempre quería estal solu en la montaña;

cuando diva a tu casa

y estaba siempri serio, siempri tristi,

como las nubes pardas.

Yo nunca tuvi padres,

jui hijo de la desgracia,

jui un pilongu, ya ves, un hespeciano,

un naide, un peazo e zarza (Delgado Fernández 1925:  15)


Venga usté, señol D. Recio

que presona tan lustrá mereci tos los respetos;

usté, pol sabel de pluma,

va a ser cabo furriel presto,

y queremos preparali

pal día que desaminemos;

bien mos puedi dal las gracias

por tos aquellos currelos

que le dimos pa enseñali,

que bien mos merecemos ... (Delgado Fernández 1925:  49)


Ya me voy a encerral los guarrapus,

diju tía Consuelu:

porque me paeci

que vieni el porqueru"

y, al marchalsi, quearun las tres

dali que te pegu

al "dicin que dicin"

soltandu el venenu ... (García García 1977:  149).


Example of chinato speech, by Gregoria Canelo (Viudas Camarasa 1987bL  72):


---Buenoj diaj tia Antoña.

---Buenoj moloj de Dio Ludia.  ¿Aonde ba V. tan temprano?

---Por mira boy a pol un comino de agua pol que quiero mazal unoj panedillo u aluego ilme a pladencia a molel una maquilla.

---y tu, ¿ande baj Ludia?

---Poj yo boy apol doj pera daguardiente pa Zajinto, polque le tengo malo y tiene unoj comitoj que le dejchangan tuyto.

---Poj di ¿como no ce lo didij al meico?  No cea que baya acel enfermea!  No modobligara el didilcelo ...

---Poj mira Ludia ci te baga al tarde badamicada, polque mandicho ca la tarde ay bayle en el cadino, y a mi no me jade gracia que baya aezoj bayle la mi Anicacia pol que ezoj baylej degarraoj paece que loj mozoj laj lleban abrazaj alaj mozaj y no jaden maj que illaj contando tuytaj laj zartaj del ejpinado, y mira ami no mejido farta il aloj cadino pacadalme con un buen mozo, y rico, poj noce cijuera la mi muchacha menuo zarmorejo me la madian, aciej que badami cada ici ella quiere il no la dejej tu bue baya ...


Early imitation of extremeño speech (Santos Coco 1936:  9):



---He?  ¿Qué hay?

---Ná ... m'acaban de icí que Bartó está pa lialaj pa'l otro mundo.

---Poj yo lo qu'iba sabío era que ejtaba malo.  ¿Quién s'iba e pensá que juera pa eso?

---Poj esa ej la pura.  Perico me ijo que lo vido ejta mañana, y qu'ejtaba sin abla, y que no conocía a naide.

---No semoj naína.  Cuando ejtamo maj ejcudiaoj moj da er patatúj, moj queamoj tiesoj y ...chanfli!

---An cá Perico ejtaba tamié er tío Bajtián, y dijo iba jecho un cuiciyo delantre dér, y que la ejaba una manda a su sobrino Frajco er Cegañuto.



Pronunciation of /s/-/θ/ in selected Spanish dialects(%)


Dialect                   /s/C                 /s/#C               /s/##                /s/#V               /s/#v


                                    [s] [h] [Ø]      [s] [h] [Ø]      [s] [h] [Ø]      [s] [h] [Ø]      [s] [h] [Ø]


Barcelona                    99   1   0          92   8   0          95   4   1          100   0   0          96   4   0

Madrid                                    94   6   0          69  29   2          82  12   6          92   8   0          96   4   0

Toledo                        65  31   4          21  59  20        76   4  20        84   9   7          82   8  10

Cáceres                       2  91   7          0  94   6          9   8  83        23  77   0          0  95   5

Granada                      0  82  18        0  85  15        1   2  97        0  15  85        2  50  48

Murcia                        1  70  29        0  80  20        18  11  71        36  28  38        38  41  21

Sevilla                        0  95   5          0  91   9          5   2  93        69  10  21        1  46  54

Las Palmas                  2  85  13        0  89  11        2  17  81        75  25   0          0  92   8


C = consonant; V = stressed vowel; v = atonic vowel; # = word boundary; ## = phrase boundary/pause