Show Summary Details

Page of

Printed from Oxford Research Encyclopedias, Linguistics. Under the terms of the licence agreement, an individual user may print out a single article for personal use (for details see Privacy Policy and Legal Notice).

date: 04 December 2022

Phonological Variation and Change in European Spanishfree

Phonological Variation and Change in European Spanishfree

  • José Antonio Samper Padilla

Summary

Spanish is a language characterized on the phonetic level by a rich variation in consonantism, especially in the syllable-final position (both word-inner and word-final), whereas vocalism shows a more fixed character and a less relevant variation. Thus, it is not strange that the majority of variationist studies have focused on consonantism. Investigations addressing prosodic variation are fewer and more recent and will not be broached here because of space limitations.

In the field of consonant variation, studies focusing on the weakening effect on certain elements of the coda stand out. The most relevant among these phenomena is the one which affects /s/, and this is so for various reasons: (a) because has been considered to be one of the isoglosses that divide the two great norms of current Spanish (Castilian and Atlantic); (b) because of its geographical extent; and (c) because it has led to theoretical approaches about the possible impact on number (singular/plural) and grammatical person (second/third-person singular) differentiation that implies the frequent presence of word-final -s. Additionally, variation which affects the liquid consonants (/l/ and /r/), leading to processes of both weakening and change in these two types of consonants, has been studied quite extensively (at least in Andalusia and the Canary Islands). The weakening process affecting the final nasal consonant, with velarization as a first step and potential elision as a second one, has been less frequently studied in Spain (much less than in the case of Caribbean Spanish, for example). In the field of syllabic tension, there is another phenomenon that has often been studied due to its geographic extent: the weakening of intervocalic /d/, which yields very different data depending on dialectal variety (evidence of the greater or lesser degree of progress in that weakening process). Sociolinguistic analyses also focus on the increasing expansion of yeísmo, a phenomenon usually conditioned by age as an explanatory factor in this advanced stage of the process. In Andalusian Spanish, the alternation between different pronunciations of the phonemes /s/ and /θ/ (mainly distinction, seseo and ceceo), the defricativization of /tʃ/ and the alternation between the realizations of /x/ as [x] and [h] have also been analyzed.

In the case of vowels, as has been said before, it should be pointed out that cases of sociolinguistic variation in Spanish are not as numerous or as relevant; therefore, they have been less appealing to researchers. Among the main phenomena, we shall discuss the vocalic metaphony registered in Cantabria and Asturias.

Subjects

  • Phonetics/Phonology
  • Sociolinguistics

1. Introduction

Spanish is characterized by ample consonant variation, whereas its vowels display stability and their variation is less relevant. Therefore, sociolinguistic research centers on consonantal phenomena. Prosodic variation studies are more recent and will not be broached here because of space limitations.

The work that has been done on weakening coda consonants is striking. As Chela-Flores (1986) has noted: “[they] share the articulatory aim of facilitating pronunciation whilst concurrently leading to the potential complete suppression of the affected segments, which results in the CV structure.” These processes characterize the innovative varieties; the most relevant of these affect the /s/ phoneme because it is one of the main isoglosses, dividing its vast geographical extension into the Castilian and Atlantic norms. Its word-final position also marks the difference between nominal singular and plural and between the second and third grammatical persons. Another widely studied phonetic characteristic (at least in Andalusian and Canarian varieties) involves the liquid consonants /l, r/ with weakening processes and also exchange between them. The weakening of postvocalic /n/, with a first velarizing step potentially followed by elision, has not undergone much research, possibly because of its smaller reach in peninsular Spain (in contrast with the Atlantic varieties; Chela-Flores, 2017, pp. 133–134).

There have been a greater number of studies on weakening of intervocalic /d/, a syllabic tension phenomenon which has produced very different data in dialectal varieties, signaling the degree of the process. Sociolinguistic analyses also indicate the growing advance of yeísmo (neutralization of the /ʎ–ʝ/ opposition with lateralization loss), in which the age factor has explanatory weight at this advanced stage of the process.

We also have a significant growth of research on Andalusian Spanish (especially on the Eastern variety), which points to the changes that have led to a contrast with the state of the dialect halfway through the 20th century: these changes include the realizations of the /s/ and /θ‎/ phonemes (distinction, seseo and ceceo) (neutralizations favoring the first and the second respectively); defricativization of /tʃ/; and the alternation between the realizations of /x/ as [x] and [h]. The social distribution of these variants provides a glimpse of their future in the meridional region.

According to Villena and Vida’s comprehensive definition (2017, 2020), an intermediate variety is being formed in Eastern Andalusia combining prevocalic features closely related to the standard (/θs/split, the most prominent member, and [x]), with a very eroded coda consonantal articulation (quasicategorical /s/ elision as the most outstanding feature, or the elision of the liquid pair, favored by the younger generation). New studies in real time will verify the consolidation of these changes.

As mentioned before, the Spanish vocalic variation cases are not as numerous or relevant, thus not leading to much interest among researchers. A comment will be made on metaphony in Cantabria and Asturias, northern regions, where it is showing signs of evident regression because of lack of social prestige.

2. Variation in Consonantism: Characteristic Coda Phenomena

2.1 -/s/

Work on -/s/ has increased during recent years with the PRESEEA project researchers′ analyses on Málaga (Vida, 2015, 2016), Granada (Tejada, 2012, 2015), Madrid (Molina, 2015) and Las Palmas de Gran Canaria (Samper & Hernández-Cabrera, 2016). Published works such as Fernández de Molina (2014) on Mérida, Kapović (2017) on Ciudad Real, or Harjus (2018) on Jerez de la Frontera may be added to those carried out toward the end of the 20th and the beginning of the 21st centuries in the Canary Islands, Andalusia, and the central peninsular region, and which were collected in Samper (2011).

See Table 1, with the four variants posited as the steps in the weakening process: sibilant [s] in [esˈpa.ɲa], aspiration [h] in [ehˈpa.ɲa], assimilation [etˈta.β‎a] and elision [Ø] in [ˈan.teØ].1

Table 1. Percentages of -/s/ Variants in Various Spanish Communities

[s]

[h]

[σ‎]

[Ø]

N

A

Madrid: Salamanca district (Gil-Peña, 2004)

82

14

2

1

32868

Madrid: Vallecas district (Molina, 2015)

68

26

3

3

5377

Madrid (Momcilovic, 2009)

66

12

10

12

7035

Alcalá de Henares (Blanco, 2004)

67

26

4

2

14813

Northeast Madrid (Ruiz-Martínez, 2003)

62

14

6

18

15229

Getafe (Martín-Butragueño, 1995)

53

35

5

7

5394

Toledo (Molina, 1998)

53

34

3

10

27666

Toledo (Calero, 1993)

52

19

15

14

6819

Ciudad Real (Kapović, 2014)

49

23

3

24

81846

B

El Hierro (Pérez-Martín, 2010)

13

84

--

3

18360

La Jara (Paredes, 2001)

27

52

4

17

35316

Santa Cruz de Tenerife (Almeida, 1990)

10

80

--

10

3366

Mérida (Fernández de Molina, 2014)

9

70

0,1

21

14525

Telde-Gran Canaria (Cabrera, 2009)

2

65

6

26

19610

Las Palmas (Samper, 1990)

3

58

6

33

28003

C

Linares (Gómez-Serrano, 1994)

2

34

3

61

12450

Málaga (Vida, 2004)

1,5

31

0,2

67

33697

Granada coast (García-Marcos, 1990)

1,7

26

--

72

15703

Granada (Tejada, 2015)

0,6

20

0,8

77

10119

Melilla (Ruiz-Domínguez, 1997)

3

16

1,5

79

8798

Depending on the advance of the -/s/ weakening process, three kinds of communities can be established:

A.

The Madrid modalities and those of the two capital cities of Castilla–La Mancha (Toledo and Ciudad Real), where the sibilant prevails, can be considered conservative.

B.

The varieties in the Canary Islands, the La Jara region, and the city of Mérida are intermediate, because aspiration prevails in them. From a quantitative perspective, the second variant could be the sibilant, as in El Hierro, or elision, as in Las Palmas.

C.

Eastern Andalusia is the most advanced area, as elision is the variant used by the majority.

2.1.1 Linguistic Factors

As regards -/s/ variation, position, phonetic context, and grammatical function exert relevant influence, as does the number of syllables in words where it is in the final position. It is elided significantly less in monosyllabic words than in those of two or more syllables: 11 versus 28% in Las Palmas and 89.6 versus 96.8% in Granada (see, respectively, Samper & Hernández-Cabrera, 1995; Tejada, 2012).

2.1.1.1 The Position

As reflected in Table 2, elision is far more frequent in word-final position, whereas the inner position tends to favor aspiration.

Table 2. -/s/ Variant Distribution According to Position

[s]

[h]

[Ø]

I

F

I

F

I

F

Madrid: Salamanca district (Gil-Peña, 2004)

87

80

9

15

1

1

Madrid: Vallecas district (Molina, 2015)

67

69

32

28

Madrid (Momcilovic, 2009)

76

63

7

13

3

15

Alcalá (Blanco, 2004)

81

64

13

30

0.5

2

Northeast Madrid (Ruiz-Martínez, 2003)

62

62

29

10

0

23

Getafe (Martín-Butragueño, 1995)

62

48

35

35

1

9

Toledo (Molina, 1998)

56

52

39

32

2

12

Toledo (Calero, 1993)

64

49

23

18

4

17

Ciudad Real (Kapović, 2014)

57

59

23

17

17

22

El Hierro (Pérez-Martín, 2010)

6

16

94

80

0,5

4

Santa Cruz de Tenerife (Almeida, 1990)

2

11

96

77

2

11

Mérida (Fernández de Molina, 2014)

6

13

94

31

0

55

Telde (Cabrera, 2009)

1

3

95

56

1

34

Las Palmas (Samper, 1990)

0.5

4

94

46

3

43

Linares (Gómez-Serrano, 1994)

1

2

81

8

8

90

Málaga (Vida, 2004)

0.3

2

90

15

10

83

Granada coast (García-Marcos, 1990)

0.3

2

38

22

62

76

Granada (Tejada, 2015)

0.6

0.6

82

5

7

95

Melilla (Ruiz-Domínguez, 1997)

2

4

37

10

58

85

Position is less relevant in conservative communities than in innovative ones: In the former, the [s] indices, for example, are very close to each other in either position, whereas [h] and [Ø] show very different percentages, favoring aspiration in the inner position and elision in the final one.

The data from the conservative varieties do not in general confirm Terrell’s (1979) and Ferguson’s (1990) hypotheses that the weakening process should have begun in the inner position.

2.1.1.2 The Phonetic Context

As shown in Table 3, the sibilant is retained much more often before a vowel or pause than before a consonant; in this last context, [s] realization almost disappears in the advanced communities.

Table 3. Percentages of Sibilant [s] According to Phonic Context

-C

-V

-//

Madrid (Momcilovic, 2009)

50

89

78

Northeast Madrid (Ruiz-Martínez, 2003)

45

87

91

Getafe (Martín-Butragueño, 1995)

41

81

78

Toledo (Calero, 1993)

20

83

76

Ciudad Real (Kapović, 2014)

16

82

78

El Hierro (Pérez-Martín, 2010)

4

18

60

La Jara (Paredes, 2001)

17

55

36

Santa Cruz de Tenerife (Almeida, 1990)

1

27

6

Mérida (Fernández de Molina, 2014)

3

19

13

Telde (Cabrera, 2009)

0.4

7

3

Las Palmas (Samper, 1990)

0.3

11

5

Málaga (Vida, 2004)

0.5

4

5

Granada coast (García-Marcos, 1990)

1

2

1

Melilla (Ruiz-Domínguez, 1997)

2

8

6

Table 4 reveals that, in the conservative modalities, elision is most favored by the preconsonantal context, while in the advanced varieties, it is the prepausal one that reaches high percentages. The Eastern Andalusian communities, which are already in the final stage of the process, do not show a great variation in indices between the three contexts.

Table 4. Elision Percentages According to Phonic Context

-C

-V

-//

Madrid (Momcilovic, 2009)

19

5

14

Northeast Madrid (Ruiz-Martínez, 2003)

35

9

8

Getafe (Martín-Butragueño, 1995)

5

7

13

Toledo (Calero, 1993)

21

9

20

Ciudad Real (Kapović, 2014)

37

11

17

El Hierro (Pérez-Martín, 2010)

4

3

3

La Jara (Paredes, 2001)

23

8

10

Santa Cruz de Tenerife (Almeida, 1990)

7

3

33

Mérida (Fernández de Molina, 2014)

1

3

55

Telde (Cabrera, 2009)

27

25

71

Las Palmas (Samper, 1990)

34

31

80

Málaga (Vida, 2004)

58

90

94

Granada coast (García-Marcos, 1990)

64

72

71

Melilla (Ruiz-Domínguez, 1997)

77

84

80

Jaén (Moya, 1979)

68

97

89

In the advanced communities (see Table 5) -/s/ has a sibilant realization much more frequently before a stressed vowel, which does not occur in the conservative varieties of the peninsular center.

Table 5. Sibilant Percentages in the Prevocalic Context

[unstressed V]

[stressed V]

Madrid: Salamanca district (Gil-Peña, 2004)

97

98

Northeast Madrid (Ruiz-Martínez, 2003)

85

91

Getafe (Martín-Butragueño, 1995)

81

81

Toledo (Calero, 1993)

82

84

Ciudad Real (Kapović, 2014)

81

84

El Hierro (Pérez-Martín, 2010)

10

41

Santa Cruz de Tenerife (Almeida, 1990)

6

77

Mérida (Fernández de Molina, 2014)

5

44

Telde (Cabrera, 2009)

1

26

Las Palmas (Samper, 1990)

2

41

Málaga (Vida, 2004)

2

9

Melilla (Ruiz-Domínguez, 1997)

6

14

Some changes are produced which affect specific contexts in different varieties:

1. In the Canarian varieties, aspiration is beginning to gain ground on sibilance when /s/ precedes a stressed vowel. As shown in Table 6, the early 21st-century PRESEEA-Las Palmas materials highlight a significant change regarding two research results from the end of the 20th century: aspiration is already the majority realization in the community (see Samper & Hernández-Cabrera, 2016).

Table 6. Percentages of -/s/ Variants in the Stressed Prevocalic Context in Three Research Results on Las Palmas de Gran Canaria

All sociolects

High norm

PRESEEA, high level

[s]

41

46

21,6

[h]

29

38

68

[Ø]

30

16

11

2. Another process of change involves the -st- consonant cluster and takes divergent paths in certain Andalusian communities and in Ciudad Real. Vida (2015, 2016) presents significant data on this interesting process, which produces resyllabification and two allophones of medial [s] when followed by a voiceless dental stop: a postaspirated realization [ˈpatha] <pasta> and an emerging affricate [ˈpatsa]) in Málaga. Table 7 shows the stages of the process.

Table 7. Distribution of Inner -/s/ Variants Before a Voiceless Dental Stop in Málaga

N

%

[s]

85

4

[h]

50

2

[th]

981

46

[ts]

733

34

[Ø]

295

14

N

2144

According to Torreira (2012), postaspiration could be interpreted as the consequence of an articulatory overlap between the glottal gestures of the aspiration and the supraglottal ones of an occlusion. On the other hand, the most recent stage, the affricate, has also been registered in Granada (Tejada, 2015), Antequera (Moya, 2007) and Seville (Ruch, 2013). In Málaga, it is clearly favored by the younger generation and by women with a lower level of education. Consequently, it is a change from below.2

Ciudad Real shows a completely different situation from the Andalusian one (Kapović, 2014), since the dental favors the [s] solution, more expected in the northern norm. The distribution of the variants before voiceless stops in word-final position is shown in Table 8:

Table 8. Allophone Percentages of Implosive /s/ Before Voiceless Stops in Ciudad Real

s#p

s#t

s#k

s

10

67

12

h

53

19

77

Ø

37

13

11

The very significative differences, according to age, gender, instruction level, and style, clearly reflect a change from above.

2.1.1.3 The Function

The results of most studies (see Table 9) do not confirm the functional hypothesis, which posits that /s/ should be elided less when it is the grammatical marker of plurality (libros) and of the verbal second person (tienes) than when it is just a monomorphematic element (entonces).

Table 9. Elision of /s/ According to Grammatical Status

[-gramm]

[+gramm]

Getafe (Martín-Butragueño, 1995)

7

11

Toledo (Calero, 1993)

21

20

El Hierro (Pérez-Martín, 2010)

3

4

Santa Cruz de Tenerife (Almeida, 1990)

8

14

Mérida (Fernández de Molina, 2014)

37

29

Las Palmas (Samper, 1990)

40

45

Telde (Cabrera, 2009)

25

40

Linares (Gómez-Serrano, 1994)

86

92

Málaga (Vida, 2004)

84

83

Melilla (Ruiz-Domínguez, 1997)

88

90

However, the validity of the functional hypothesis is maintained when it is reinterpreted as a function of the information transmitted by grammatical /s/: The consonant is elided much more frequently when it presents redundant information (niños estudiosos) than when it is the first mark of plurality (niños estudiosos), as shown in Table 10.

Table 10. Elided Variants According to the Feature [± Redundant] of the Mark of Plurality in the Complex NP

Modifiers

Nuclei

[-red]

[+red]

[-red]

[+red]

Las Palmas (Samper, 1990)

15

56

23

63

Telde (Cabrera, 2009)

12

53

25

64

El Hierro (Pérez-Martín, 2010)

4

5

1

3

Mérida (Fernández de Molina, 2014)

1

63

33

40

2.1.2 Social Factors
2.1.2.1 Sociocultural Level

This factor has a very important incidence on -/s/ variation. The highest sociocultural levels tend to favor the more prestigious variants ([s] in the conservative varieties and [h] in the advanced), whereas the low groups present the highest percentages of [h] in the dialects placed in the initial stages of weakening and of [Ø] in the radical ones.

2.1.2.2 Sex

There is no uniformity in the behavior of men and women as regards the weakening process. Women are more conservative in the peninsular central zone (Madrid as a whole and its districts of Salamanca and Vallecas, Toledo, the northeast of Madrid, Getafe, Ciudad Real, and La Jara), as they tend to retain the sibilant more. On the other hand opposite side, there is a greater degree of elision among women in Mérida and in Melilla.

Lastly, this conditioning factor has no relevant incidence in the Andalusian communities of Linares and Jerez de la Frontera or in any of the Canarian varieties (Las Palmas, Telde, Santa Cruz de Tenerife, and El Hierro).

2.1.2.3 Age

Although at present, changes can be detected in specific contexts in certain communities, the general situation of the weakening process is relatively stable (see Gimeno, 2008). However, within that stability a certain regression of the weakening phenomenon can be perceived. The data from the majority of Spanish communities show that the third generation has higher indices of elision in the advanced varieties, as in the Canary Islands (with the exception of El Hierro), Mérida, or Melilla and of aspiration in the more moderate ones (Toledo, Getafe, or the Salamanca district). However, young speakers show more aspiration in the first group of communities and more sibilants in the second one, which means they do not favor weakening.

Exceptionally, in El Hierro, a change is produced favoring aspiration to the detriment of sibilance, implying that its young speakers are approaching the regional norm. The young generation in Málaga and Linares is also the group that elides the more, but the differences are not large.

2.2 The Liquid Consonants

On broaching these segments, two coda processes must be distinguished: a weakening one that can lead to elision, [koˈme] ‘comer’ [faˈta] ‘fatal’; and one of neutralization in both senses, [ˈkal.ta] ‘carta’; [ˈfaɾ.ða] ‘falda’, differing in intensity depending on the community. Both phenomena are characteristic of the innovative Spanish dialectal areas (see Moreno, 2009, p. 81).

2.2.1 -/ɾ/

The /r/ coda variation has been studied sociolinguistically on the island of Gran Canaria; in the central peninsular modalities; in Mérida;3 and in Andalusia, where work has been done on Linares, Alcalá de Guadaira, and Granada, to which Melilla can be added.

Table 11 gathers the results of the three most relevant variants from a quantitative perspective:4 (a) the normative one, articulated as a tap [ɾ] or as an approximant [ɹ]); (b) the pronunciation as a lateral [l]; or (c) the elision of the consonant [Ø].

Table 11. Distribution of /r/ Variants in Various Spanish Modalities

[l]

[ɾ] + [ɹ]

[Ø]

N

Alcalá de Henares (Blanco, 2004)

3

95

1,5

5196

Getafe (Martín-Butragueño, 1995)

4

93

3

3000

Northeast Madrid (Ruiz-Martínez, 2003)

1,5

80

14

5378

La Jara (Paredes, 2001)

5

85

2

6116

Alcalá de Guadaira (Ruiz-Sánchez, 2009)

0

64

32

5772

Linares (Gómez-Serrano, 1994)

10

54

20

3680

Granada (Fernández de Molina, 2018)

0,2

62

31

1249

Melilla (Ruiz-Domínguez, 1997)

-

62

34

3042

Telde (Cabrera, 2009)

0

73

25

6901

Las Palmas (Samper, 1990)

3

65

29

11357

Lambdacism is much less relevant quantitatively than elision, since it only reaches higher percentages than loss of /ɾ/ in the Madrid regions of Alcalá de Henares and Getafe and in La Jara, only in very limited proportion. Andalusia shows very small percentages,5 except in Linares. Elision, infrequent in the peninsular center, with the exception of the towns in Northeast Madrid, could be considered characteristic of the peninsular south and Gran Canarian dialects. In Granada, Alcalá de Guadaira, and Melilla elision accounts for more than 30% of the occurrences of the vibrant; its presence is also high in the two Gran Canarian communities and in Linares.

2.2.1.1 Linguistic Factors in the Variation of -/ɾ/

1. A common denominator of all the zones under study is that the word-final position favors elision. In Table 12 one observes the interesting growth of elision in those communities where the loss is relevant. The percentage of [Ø] in Mérida, the highest in peninsular Spanish, is striking.

Table 12. /r/ Elision Percentages According to Position

Inner

Final

Northeast Madrid (Ruiz-Martínez, 2003)

3

24

Mérida (Fernández de Molina, 2014)

-

70

Alcalá de Guadaira (Ruiz-Sánchez, 2009)

12

55

Granada (Fernández de Molina, 2018)

4

57

Linares (Gómez-Serrano, 1994)

16

24

Melilla (Ruiz-Domínguez, 1997)

9

60

Telde (Cabrera, 2009)

15

34

Las Palmas (Samper, 1990)

12

45

Lambdacism does not present results as homogeneous as those for weakening. While in the Madrid case, the proportion of [l] slightly increases in word-final position, in the Canarian communities and Linares, lateralization is somewhat more frequent in inner-word position.

2. Grammatical conditioning is highly significant in the explanation of the postnuclear vibrant. The elision is clearly favored as the infinitive marker, as reflected in Table 13 with Andalusian and Canarian data.

Table 13. Percentages of /r/ Elision According to Grammatical Category in Innovative Communities

Other categories

Infinitive

Alcalá de Guadaira (Ruiz-Sánchez, 2009)

18

66

Granada (Fernández de Molina, 2018)

4

56

Melilla (Ruiz-Domínguez, 1997)

19

65

Telde (Cabrera, 2009)

23

43

Las Palmas (Samper, 1990)

32

53

As is known, the loss of the infinitive marker does not lead to ambiguity because of its predictability.

3. The contextual factor is also relevant in almost all communities. In general, the prepausal markedly favors consonantal elision in all the innovative modalities and the prevocalic favors the occurrence of the canonical articulations. Gran Canarian varieties present that situation, as do Melilla and Alcalá de Guadaira, Granada being the only exception with 53.8% of [Ø].

4. Frequency of use is also an important factor in explaining the elision of /ɾ/. The most frequent words favour the loss of the consonant: 36% of elision compared to 24% of low frequency words in Alcalá de Guadaira.

2.2.1.2 Social Factors in the variation of -/ɾ/

1. The sociocultural level is also a key factor behind the most important differences in that lower sociolects reach higher indices in the nonnormative realizations—laterals and those that are elided—evidence of their low prestige in the different communities. See Table 14.

Table 14. Percentages of -/ ɾ / Elision According to Sociocultural Level

High

Low

Mérida (Fernández de Molina, 2014)

49

83

Alcalá de Guadaira (Ruiz-Sánchez, 2009)

23

34

Granada (Fernández de Molina, 2018)

24

37

Linares (Gómez-Serrano, 1994)

17

23

Melilla (Ruiz-Domínguez, 1997)

20

42

Telde (Cabrera, 2009)

21

59

Las Palmas (Samper, 1990)

9

61

Undoubtedly, the very high elision of the two Mérida levels is striking. On the other hand, in Alcalá de Guadaira the loss is more frequent at the intermediate than at the low level, a result that, according to Ruiz-Sánchez, indicates that elision may be gaining prestige in the community.

2. The sex factor is not significative in any of the analyzed communities and in Granada, for example, the ANOVA analysis confirms this. In Las Palmas, women had a relevant role in the lateralization and elision rules, with slightly higher probabilistic indices than men, a similar result to that in Alcalá de Guadaira. In La Jara, however, men neutralize somewhat more.

3. The results derived from the age variable are uneven. In some communities, elision and lateralization show significantly lower indices among the youth, showing a progressive change from above to the detriment of the less prestigious variants. This tendency is clearly perceptible in Las Palmas and Getafe, but in Mérida and Linares this change only affects elision.

On the other hand, recent studies on Alcalá de Guadaira and Granada show that elision is favored by the younger generation. Future research will confirm whether there is currently a change in favor of the loss of final /ɾ/ in Andalusia.

2.2.2 -/l/

Studies on the fate of the lateral consonant are fewer. It is known that in some communities, coda variation of /l/ has little relevance, but more quantitative studies are needed on Andalusia, where rhotacism has been shown to be one of the region’s characteristic features. As Narbona et al. argue, “The majority result, geographically and demographically of the -r/-l confusion in Andalusia is -r” (2011, p. 214)).6 It is possible that as the elision of /l/ does not have a grammatical effect, its study has become less attractive than research on /s/, /n/, or /r/ in the same syllabic position.

Table 15. Distribution of -/l/ Variants in Various Spanish Modalities

[ ɾ- ɹ]

[l–l]

[Ø]

N

Getafe (Martín-Butragueño, 1995)

4

95

1

2038

La Jara (Paredes, 2001)

7

90

2

5629

Mérida (Fernández de Molina, 2014)

-

61

39

280

Granada (López-Moreno, 2018)

18

67

15

724

Linares (Gómez-Serrano, 1994)

26

59

14

2740

Melilla (Ruiz-Domínguez, 1997)

12

63

25

1941

Telde (Cabrera, 2009)

1

93

6

4877

Las Palmas (Samper, 1990)

11

78

11

6949

Note: The Mérida data refer only to the word-final position; López-Moreno (2018) presents the data on the elided and assimilated variants jointly; Gómez-Serrano works with 15% of vibrants plus 11% of intermediate realizations [l/ɾ].

As can be seen in Table 15, he weakening process is strongly present in two Andalusian regions, in Melilla and even more in Mérida (word-final position); in Canarian Spanish it is at an intermediate stage, while Getafe and La Jara represent the more conservative stage. In these two locations a significant index of rhotacism is found, which is higher in Las Palmas and Melilla and strikingly higher in Granada and Linares. Andalusia, then, is the dialectal area where rhotacism is most frequent.7

The above data show that in general, rhotacism is more frequent than lambdacism in the Spanish communities. On the contrary, the Caribbean shows high percentages of lateralization, and rhotacism is practically inexistent (see López-Morales, 1992, pp. 100–119). This is one feature that clearly distinguishes the two dialectal areas.

2.2.2.1 Linguistic Factors in the Variation of -/l/

In Gran Canaria, Melilla, and Linares the change toward [r] is driven by the inner position, while the word-final position produces an important increase of elisions. In La Jara, position has no effect.

As far as context goes, the two Canarian locations, Granada and Melilla, show that the change toward a vibrant consonant is clearly favored by the following consonant, while elision of /l/ is promoted by the pause.

In all existing studies on the stress factor, the loss of /l/ is more frequent in stressed words.

2.2.2.2 Social Factors in the Variation of -/l/

The distribution of variants according to sociocultural level clearly indicates that rhotacism and elision are associated with the lowest social levels. This has been verified in Las Palmas, Mérida, and Melilla. Granada and Jerez de la Frontera also present eloquent indices of rhotacism.

The same homogeneity was not found in the results for sex differences. In La Jara and in Jerez de la Frontera, men favor rhotacism more than women do, and in Mérida and Melilla, men elide more than women as well. On the contrary, in Las Palmas women favor rhotacism and elision. Differences are minimal in Linares and are not significant in Granada.

As regards age, older men elide and neutralize more in Las Palmas and Linares, and they also promote rhotacism in La Jara. Thus the data indicate that these phenomena are in regression in these communities as a result of pressure from normative forms. Mérida shows a more stable situation, with greater conservatism in the second generation. However, younger speakers in Granada and Jerez de la Frontera stand out in that they clearly favor [ɾ] realization, with a distribution that reveals a trend toward a solution that will be quite distinct from the normative form.

2.3 -/n/

The nasal weakening process has drawn little attention from Spanish linguists.8 Most studies have been carried out in the Canary Islands (Almeida, 1990; Hernández-Cabrera & Samper-Hernández, 2011; Pérez-Martín, 2010; Samper, 1990). There is also research on Melilla (Ruiz-Domínguez, 1997), and work has been done on elision in Ayamonte, Huelva (López de Aberasturi, 1997) and on velarization in Ucieda, Cantabria (Holmquist, 1988).

Table 16 gathers the data produced by the studies carried out in the Canary Islands and Mellilla. In all of these studies, three variants stood out (alveolar or assimilated [n], velar [ŋ] and elision [Ø], with or without nasalization of the preceding vowel), and these reflect how the process is evolving; this process has been effectively explained by Chela-Flores (1986) by means of his Progressive Weakening Theory.

Table 16. Distribution of the -/n/ Variants in Various Spanish Communities

[n]

[ŋ]

[Ø]

N

El Hierro (Pérez-Martín, 2010)

89

8

3

13650

Las Palmas (Samper, 1990)

82

11

6

25352

Sta. Cruz-Tenerife (Almeida, 1990)

72

24

5

2850

Melilla (Ruiz-Domínguez, 1997)

70

12

18

6337

The Canarian data points to decidedly lower elision than found in Melilla and turn out to be close to the results found in San Juan de Puerto Rico (López-Morales, 1983).

2.3.1 Linguistic Factors

1. The position of the nasal segment in the word is very important: When it occurs at the end, a marked increase of the velar variant is always produced, which for instance, goes from 0.1% in the inner position to 25% in word-final position in Las Palmas, and from 0 to 17% in El Hierro.

2. Context is also highly relevant in this process. As can be seen in Table 17, the velar variant shows poor frequency before a consonant, a high increase before a vowel, and is the majority realization in the prepausal context.

Table 17. Velar Realization Indices of -/n/ According to Context in Different Communities

_C

_V

_//

El Hierro (Pérez-Martín, 2010)

1

25

77

Las Palmas (Samper, 1990)

10

35

80

Santa Cruz de Tenerife (Almeida, 1990)

27

43

84

Melilla (Ruiz-Domínguez, 1997)

9

13

52

3. With regard to the functional conditioning, in Santa Cruz and Melilla elision decreases when the nasal is a verbal marker of the third person in the plural. On the other hand, in Las Palmas the frequency of elision without nasalization of the preceding vowel is much higher when the consonant is [+gramm]; as in the case of /s/, there are always resources that can be used against possible ambiguity (Samper, 1990, pp. 233–238). This pattern also occurs in El Hierro, but with much more moderate elision.

2.3.2 Social Factors in the Variation of -/n/

The analyzed communities coincide in terms of the small significance of the sex factor in the process. However, this is not the case when other variables such as age and sociocultural level are taken into account. For example, in Las Palmas a change is in progress in the prevocalic context: The velar variant is replacing the alveolar one under the pressure from younger speakers and those from the middle sociocultural level (see the curvilinear distribution shown in Figure 1). In El Hierro the growth of velarization in the same context can be observed, although there it is the higher sociocultural levels that drive the phenomenon.

Figure 1. Realizations of [ŋ] in the prevocalic context according to sociocultural level in Las Palmas de Gran Canaria (1: middle-high, 2: middle, 3: middle-low, 4: low).

Hernández-Cabrera and Samper-Hernández’s study (2011) on the prevocalic nasal in Canarian TV news programs shows that in spite of these programs’ very formal style, a significant number of velar realizations (65%) are present, which according to these researchers, confirms that this change could be taking place all over the archipelago.

3. Variation in Consonantism: Syllabic Tension Phenomena

3.1 -/d/-

Weakening and elision of the intervocalic obstruent is a widely present phenomenon in the Hispanic world. It is a very old process, which had already fully developed by the 16th century in both Andalusia and Madrid, as Molina (2006) reminds us.

The elision of -/d/- has spread to a differing extent between the Spanish communities, as shown in Table 18.

Table 18. Maintenance and Elision Percentages of -/d/- in Different Spanish Communities

[ð]

[Ø]

N

El Hierro (Pérez-Martín, 2010)

85

15

2688

Alcalá de Henares (Blanco, 2004)

81

18

4015

Toledo (Molina, 1998)

80

20

5735

La Jara (Paredes, 2001)

79

21

5240

Madrid: Salamanca district (Gil-Peña, 2004)

78

21

4637

Northeast Madrid (Ruiz-Martínez, 2003)

77

23

3597

Telde (Cabrera, 2009)

74

26

2699

Getafe (Martín-Butragueño, 1995)

68

32

1791

Las Palmas (Samper & Hernández-Cabrera, 1995)

62

38

4479

Málaga (Villena, 2012)

62

38

14526

Melilla (Ruiz-Domínguez, 1997)

52

48

1593

Mérida (Fernández de Molina, 2014)

41

59

1874

Jaén (Moya, 1979)

35

65

2727

Linares (Gómez-Serrano, 1994)

22

78

1546

Elision reaches very high percentages in Eastern Andalusia and in Mérida; Castilian varieties show themselves as more conservative; and Canarian localities hold an intermediate position, although a low percentage of loss must be highlighted in El Hierro.

More recent results have been obtained by the PRESEEA project teams, and in Figure 2, the elision indices of five cities can be seen.9

Figure 2. Elision indices of /d/ in five Spanish cities (PRESEEA corpus).

Source: Author. Data for Madrid come from Molina and Paredes (2014, 2015); for Valencia, from Gómez-Molina (2012, 2013); for Granada, from Moya et al. (2012) and Villena and Moya (2016); for Málaga, from Villena and Moya (2016); for Las Palmas, from Samper et al. (2011) and Samper and Samper-Hernández (2020). There are also results from the low level in Seville, with a very high elision percentage of 47.17% (Jiménez-Fernández, 2015).

These results fully confirm the diatopic distribution shown in previous work: the northern conservatism of Valencia and Madrid, with moderate elision; the intermediate position of Las Palmas, with an index close to the previous cities, and the more advanced position found in the Eastern Andalusian capitals.

3.1.1 Linguistic Factors in the Variation of -/d/-

This segment’s variation shows the more significative conditioners are, as expected, the grammatical status of the word containing /d/ and the phonic context.

1. Scholars highlight the grammatical status of the word as a fundamental factor in order to explain the loss of the dental phoneme. The highest elision indices are found in the participial forms in all the communities. As an example, consider the very different percentages in the cities studied in the PRESEEA project: 33.1% in Madrid, 31.9% in Valencia, 75.1% in Granada, 70.1% in Málaga, and 28.8% in Las Palmas (see Figure 2).

Intelligibility is not impeded by the elision of participial /d/ because the obligatory presence of the auxiliary and the following back vowel convert the elided consonant into a completely redundant mark, which only transmits already known information (see López-Morales, 1983). According to Bybee (2002), the greater participial elision of /d/ could be explained as a function of inventory frequency (type frequency).

2. The incidence of preceding vowels with grammatical status in the weakening of -/d/- is similar in all varieties: there is more elision in -ado than in -ido. When /d/ is not part of the participle, /o/ and /a/ are the vowels that lead in the promotion of its loss and /u, i/ the ones that favor its permanence. Among following vowels, /o/ fosters it to a greater degree than /a/ in all Spanish varieties.

3.1.2 Social Factors in the Variation of -/d/-

1. The results from many communities show more conservative behavior in women, which is clear in the PRESEEA data from Las Palmas, Málaga, and Madrid. The work on different zones in Madrid (the Salamanca district, Alcalá, and Getafe), Toledo, Mérida, and Linares also reveals lower elision indices in women. This is a phenomenon that, as Moreno points out (2004, p. 1001), reflects the greater degree of feminine linguistic conscience of forms of evident prestige.

There are of course exceptions, like Valencia, Granada, and El Hierro, where the elision percentages are practically equal for both sexes.

2. On analyzing the generational variable, it is observed that the orientation of the process differs among the communities. Three different situations may be distinguished:

2.1

The regression of the phenomenon shows inferior elision indices in the youngest group. This is what happens in Las Palmas, Valencia, and Getafe.

2.2

The advance of the process toward elision is observed among the young in Granada and Málaga, who favor it to a greater extent. These data show that in Eastern Andalusia the process keeps up its growing pace, while the foreseeable regression is present in other Spanish communities as predicted by Molina (2006).

2.3

There is stability in those communities where age difference is not significant. This is what PRESEEA data show for Madrid, a result that coincides with previous findings in the Salamanca district, Alcalá, and Toledo. El Hierro Island also shows a stable profile.

3. In general, elision is more intense at the lower sociocultural levels, as has been shown in previous research (see Moreno, 2004, p. 1001 and Samper, 2011, p. 114). In Table 19, the results of extreme groups in the analyzed communities within the PRESEEA project are shown.

Table 19. Elided Variants of -/d/- in Communities Studied in the PRESEEA Project According to Sociocultural Level

The differences between the percentages of high and low sociocultural levels are very marked in Eastern Andalusia, but, on the contrary, in Madrid there is little significant separation, and Valencia is an exception due to the conservatism of the lower level in this process.

3.2 The Delateralization of /ʎ/

Yeísmo, the dephonologization of the opposition between the palatal consonants /c/ and /ʝ/ to the detriment of the first, is a change in progress in the Hispanic world with different rates of development according to geographical area, sociolect, and style (see Gómez & Molina, 2013; Moreno, 2004). There is generalized consensus that the scarce functional load of the opposition /ʎ/ ~ /ʝ/ is the main cause of its disappearance. Additionally, its urban character has given it prestige, which would provide a satisfactory explanation of its rapid propagation (Molina, 1997; Salvador, 1964). Its advance is so firm that Moreno has stated that “the Iberian Peninsula moves toward a generalised yeísmo with zones with advanced solutions (Andalusia) and more conservative zones (in Old Castile)” (2004, p. 988). According to Rost (2014), the change is in the third of the phases distinguished by Moreno (2004, p. 985: “yeismo without remains of lateralization”) in the greater part of Spain.

The Madrid community has been converted, as indicated by Molina (2013), into a transition zone for the change. The results of different rural areas show a change in progress with the innovative foci in the south and the conservative ones in the north. In 16 Madrid towns studied between 2000 and 2005, about 7% of the inhabitants still systematically maintain /ʎ/. In the northeast, Ruiz-Martínez (2003) found 17% of [ʎ] from the total of occurrences, the vast majority by speakers of the oldest generation. There has, then, been an evident advance of yeísmo in the Madrid province throughout the 20th century.

In Madrid, specifically in the Salamanca district, yeísmo has generalized and nothing is left of the old phonological distinction (Molina, 2013). In two other urban Madrid communities (Alcalá and Getafe) yeísmo has spread triumphantly, although relevant phonic differences can be observed in the realization of /ʝ/ which reflect that the change has not been truly concluded.

In Burgos, yeísmo has advanced significantly: in Covarrubias 47% of speakers have [ʝ] (Chapman et al., 1983; apud Silva-Corvalán, 2001) and 61% in the city of Burgos (Martínez, 1983); these figures reflect the greater progress of the phenomenon in the urban zones. In the Extremaduran-Castilian region of La Jara, Paredes (2013) finds 48% maintenance of [ʎ], with marked differences between the various localities. The highest distinction data come from El Hierro, a Canarian island, where yeísmo is still at the beginning stage (3% of [ʝ]) facing 97% of [ʎ]) (Pérez-Martín, 2010), in open contrast with the regional capitals. The islands do not show yeísmo to be as generalized as it is in Andalusia, where the distinction remains only in some isolated nuclei (see Narbona et al., 2011, pp. 200–202).

In the bilingual communities the distinction remains in Valencia, at a high 75%.10 Gómez-Molina and Gómez-Devís (2016) distinguish three groups of speakers: “the distinguishers (41.7%), the distinguishers who also maintain some sequences containing yeísmo (33.3%) and others considerably faithful to yeísmo (25.0%).” The data provided by Torres et al. (2013) for Barcelona City indicate that yeísmo has advanced much further there than in Valencia: /ʎ/ is realized as [ʝ] in a very high 74%.

3.2.1 Explanatory Variables

In the explanation of this phenomenon, the social variables are more relevant than the linguistic ones. As it is easy to understand, the age factor becomes very relevant given the condition of a change in progress. Thus in El Hierro the first generation presents a higher index of [ʝ] (10%) than the two older generations (3% and 1.2%). The Burgos yeísmo indices are equally enlightening with regard to the conversation style: 34% (III) ➔ 57% (II) ➔ 85% (I). In La Jara “the young speakers realise the lateral sound once out of four times, while the ones older than sixty realise the same sound twice out of three” (Paredes, 2013, p. 75). The age difference also shows itself in the probabilistic indices which the generations contribute to the retaining of [ʎ] in Covarrubias. In the rural areas of the Madrid province, younger speakers have already been totally converted to yeísmo, whereas those older than 55 still retain the distinction. The propagation of yeísmo has been very intense in Valencia: the realization of [ʎ] spans from an almost general 96.5% in the third generation to 47% in younger speakers; similar results are obtained in Barcelona and Gandía.

The sex factor is not as much of an influence as age. It has no relevance in El Hierro, Burgos, and La Jara. However, in Covarrubias women show a greater conservatism as a sign of the stronger social pressure exerted on them to maintain “correct” pronunciation (Silva-Corvalán, 2001, p. 259); the same motivation explains why they do not lead the change in the Madrid region (the phonological distinction is preserved by women of the older generation). Taking into account the situations in Barcelona, Gandía, and Valencia, yeísmo has advanced more among women.

As regards the sociocultural level, while yeísmo is supported in El Hierro by speakers of the highest level, this conditioning factor does not affect the degree of conservation of the lateral in La Jara, Burgos, or the Madrid community. In Valencia the higher indices of yeísmo are present in the lower level.

3.3 Distinction Between [s] and [θ‎], Seseo, Ceceo, and Heheo

The distinction between /s/ and /θ‎/ and the realizations in which they converge (seseo, ceceo, and infrequent heheo or jejeo) have been the object of recent research in Andalusia, with novel results in relation to previous studies.11 The great attention given to this distinction has been motivated by its “radical and striking” character (Moya & Sosiński, 2015).

The studies on the shift towards the distinction between /s/ and /θ‎/ in Eastern Andalusia (for Málaga, Ávila, 1994; Villena, 1997a, 1997b, 2008, 2001; Villena & Vida, 2017, 2020, and on Granada, Melguizo, 2007a, 2007b; Moya & García-Wiedemann, 1995; Moya & Sosiński, 2015; Salvador-Salvador, 1980) are complemented by other recent research in Western Andalusia: García-Amaya (2008) and Harjus (2018) on Jerez de la Frontera, Regan (2017) on Huelva, and Santana (2016, 2017) on Seville.

These researchers show that a change is in progress, with a social ordering of the different variants and with linguistic attitudes that make it possible for scholars to predict the future of the phenomenon on a firmer basis. According to Villena (2001), it is a recent phenomenon, which could have started in the middle of the 20th century. In this sense, Moya and García-Wiedemann ascertained a radical change in Granada from the situation the Atlas lingüístico y etnográfico de Andalucía (ALEA) had observed 40 years before, and Melguizo verified that tendency in the small locality of Pinos Puente. Figure 3 (Moya & Sosiński, 2015) clearly shows the evolution of the indices of the distinction over a short period in Granada.

Figure 3. Evolution of the s/θ‎ distinction in Granada in recent years.

In Western Andalusia, Harjus’ research (2018) in Jerez de la Frontera confirms García-Amaya’s (2008) data in the sense that the observed change drives Jerez de la Frontera nearer to the Madrid norm rather than to that of Seville.

In the city of Huelva, Regan (2017) describes a change from above in that the ceceo, the predominant pronunciation there, is being replaced by the Castilian standard distinction, driven by the women of all socioeconomic groups and by the men of the middle-class districts of the city. This move has converted ceceo into an accepted marker for men of working-class districts with a low educational level. This evolution is related to the social and economic changes in the city.

Santana’s research (2016, 2017) clearly indicates that the predominant seseo is being replaced in Seville by the distinction-based solution (having reached a very high 75% at the higher social level). This also places Seville on the path of the convergent pattern, which is prestigious because it is led by highly educated speakers and first-generation women. Thus, seseo continues to be more frequently used among men with lower educational levels, particularly younger ones.

As Villena (2001) has indicated, the introduction of the /s/:/θ‎/ opposition in communities where it had not been used systematically previously is a solid example of the importance of social prestige in linguistic change, since it is favored by the inhabitants with stronger educational levels, by the young, and frequently by women. On the other hand, this change is an exception to Garde’s Principle of “the irreversibility of fusions” by linguistic means (see Labov, 1994, pp. 485–487).

The relevance of these recent studies is that the shift toward the distinguishing solution does not seem to be exclusive to Eastern Andalusia, because this move has been documented in relevant urban nuclei in the western region (see Table 20). As Villena and Ávila (2014, p. 215) have correctly indicated, “the real centre of the innovative way of the spoken Spanish (norma sevillana) would be endangered.”

Table 20. Maintenance (seseo + ceceo) and Split of the Dental Fricative in Andalusia

Málaga

Huelva

Granada

Seville (A level)

Jerez

Seville (B level)

Split

82

82

79

74.5

27

26

Reduction

18

18

21

25.5

73

74

Note: The data in Table 20 come from Villena and Vida (2017, 2020), Regan (2017), Moya and Sosiński (2015), Santana (2016, 2017) and García-Amaya (2008).

There is no doubt that we are facing a phonological variable of great interest which, due to its theoretical and practical implications, will be the object of new studies in the Andalusian communities.

3.4 Defricativization of /tʃ/

In some Eastern Andalusian communities, such as Granada and Málaga, progressive change back to [tʃ] from the traditional Andalusian [ʃ] is taking place. This is the same convergence process as that highlighted by Villena and Vida (2017) and Moya (2018) for other changes affecting the tensive consonantism already mentioned.

According to Moya and García-Wiedeman (1995), the affricate is the majority variant in Granada with a high level of 80%, in contrast with the already reduced 20% share of [ʃ], the predominant variant in the 1950s, when the ALEA was produced. In Málaga (Cuevas, 2001; Villena, 1997b), the fricative offers more resistance than in Granada, because its percentage keeps to a somewhat higher level (25%).

The incidence of the social variables related to this change from above keeps to what is expected: the young, those of higher educational level, and women favor the affricate. The fricative variant is perceived, according to Moya and García-Wiedemann, as a masculine feature.

The Eastern Andalusian regional change has no parallel in the western zone. Harjus (2018, pp. 156–160) points out that in Jerez de la Frontera, the [ʃ] index is 81%, close to the 83% found by Carbonero et al. (1992, p. 29) in the same city in the 1990s. Here as well, the young, women, and speakers with a higher level of education present more standard Spanish realizations.

3.5 Realizations of /x/

There is another change in certain Eastern Andalusian communities which leads to the loss of the traditional variant [h] in favor of the [x] variant, the most general in the peninsular center and north. Moya et al. (2014) show some results from Granada which contrast openly with those in the ALEA. In fact, they are parallel to the changes in the realizations of [tʃ] and [ʃ] in the same urban area.

At present in Granada, the aspirated variant still slightly predominates (47%), although the fricative nears it (45.6%), with the addition of 7.4% of the elided variant, which is totally without social prestige. The fricative realization is favored by the highly educated, the second generation, women, and speakers from the third life-mode (Højrup, 1983). Women with a higher educational level are clearly the leaders of the change process, especially the younger ones. This change from above, according to Moya et al., is not only evidence of another move in the convergence toward the national standard, but also an approach to the forms in other zones of Eastern Andalusia (Jaén and part of Granada and Almería), where the fricative variant has traditionally predominated.12

The data from Eastern Andalusia contrast openly with those of the western part. Harjus (2018, p. 163) quantifies the information on Jerez as 95% of aspiration, a percentage close to Carbonero’s 96% (1992, p. 22). Payán (1988) also found a similar index (97%) in the city of Cádiz. In Jerez the socio-educative level is relevant: speakers with university qualifications present an aspiration index (76%) that is much lower than those of other educational levels.

4. Variation in Vocalism

4.1 Metaphony in Northern Peninsular Areas (Cantabria)

In some northern dialects (Asturian and Cantabrian), the closing of a stressed vowel is variably produced by the influence of a final closed vowel. Thus /ˈo/ becomes [ˈu] (tsubu <lobo> ‘wolf’) and /ˈe/ moves to [i] (pirru <perro> ‘dog’); the more open vowel could close toward palatality (zapetu <zapato> ‘shoe’) or velarity (gotu <gato> ‘cat’). Alarcos (1964) defined this phenomenon as “a partial distant assimilation to the stressed vowel.”

The sociolinguistic studies carried out in Cantabria show that this is a discredited phenomenon and that it is in regression, as shown by the incidence of the extralinguistic factors. Fernández-Juncal’s data (1998, 2000) on a Cantabrian Eastern region confirm that the speakers consider the phenomenon as rural and characteristic of a lack of education.

It is relevant to indicate that the scarcity of metaphony in the youngest generation is in clear contrast with the very high indices in the two older age groups. Therefore, age is a decisive conditioning factor in the future disappearance of this dialectal feature. The difference between the sexes is also significant because of much higher metaphony indices in men than in women. As was foreseeable, the socioeconomic variable also plays a very marked role in the difference.

The author also took into account a mixed social variable, namely the contact with the norm, with parallel results with those mentioned for the socioeconomic level. Other factors also reflect the low prestige of metaphony: (a) the demographic variable, since the municipalities with a population below 5,000 inhabitants favor the dialectal realization, while in the city of Santander the indices of the phenomenon are very low; and (b) the location of the municipalities, since those in the interior are more conservative and show higher metaphony indices than the coastal ones. The figures on the different styles confirm the negative evaluation of this phenomenon.

Fernández-Juncal’s results coincide in many aspects with Holmquist’s 1985 work in the small locality of Ucieda, in Western Cantabria, about the closing of the final vowel (jachu, jigu). This study showed that the -/u/ variant was very frequent among speakers older than 75 and decreased progressively in the intermediate generations, reaching the lowest level among young speakers in the 13–24 age group. The research also confirmed that women’s usage was closer to the normative, explaining this in the case of the older generation as resulting from their work as maids and cooks for the bourgeois families, which meant that they were thus more exposed to the standard pronunciation. Holmquist also found a political orientation (socialist–conservative) underlying the differences.

5. Conclusions

The analysis of a series of variable phenomena in European Spanish allows an understanding of their situation in the different dialectal modalities. Therefore, it can be confirmed that the Andalusian varieties (especially the Eastern ones) are those that drive furthest the weakening processes affecting /s, r, l/ as coda or postnuclear consonant phonemes. It can also be confirmed that Canarian communities present less marked advances, while, as could be expected, the Castilian–Manchegan dialects turn out to be more conservative.

As regards syllabic tension, the unstoppable advance of yeísmo is clearly drawn in practically all communities (also in the bilingual ones) of peninsular Spanish; the data from the Canarian island of El Hierro are a notable exception. In the elision of /d/, a similar dialectal distribution is perceived to that which is described for the postnuclear phenomena. The studies on three variables in Eastern Andalusia (distinction or confusion of /θ‎/ and /s/; defricativization of /tʃ/ and variation of /x/) allow an approach to very interesting situations of change—or leading to it—with a dialectal convergence movement which leads to the formation of a koineised variety.

It is accepted that the variations observed in different communities can be explained by linguistic and social factors. The former are usually more relevant and, we could say, present a greater degree of coincidence among the different varieties. Thus, it could be verified that in all of them, the word-final position openly favors consonantal weakening in the coda; the prevocalic context is usually more restrictive than the preconsonantal and the prepausal ones in the weakening processes (with the evident exception of final nasals). From a functional perspective, it is clear that elision of segments with grammatical load does not cause intelligibility problems because Spanish offers alternative markers for the expression of the content of interactional messages.

The incidence of social conditioning is not usually as homogeneous as that of linguistic conditioning. Thus, age may have an opposite influence depending on the direction of change: in the weakening of /r, l/ young speakers do not favor elision in Las Palmas, but they do in Andalusian varieties such as those of Alcalá de Guadaira or Granada. In other cases there is coincidence, as with younger Andalusians in the case of the /θ‎/ and /s/ distinction. The greater female tendency toward prestigious variants is found above all in the weakening of intervocalic /d/, but it is not general. A greater degree of coincidence is found when the sociocultural level is analyzed: low levels foster extreme solutions to coda-weakening phenomena in contrast with the conservatism of the high sociolects, which, on the contrary, prove to be very active in change situations, implying syllabic tension strengthening and changes from above.

The sociolinguistic perspective has provided a better explanation of phonic variation in Spanish modalities and makes it possible to distinguish changes that may go unnoticed from other approaches.

Acknowledgment

This publication is part of the research project “Agenda 2050: The Spanish of Seville and Las Palmas de Gran Canaria: Processes of spatial and social variation and change” (reference PID2019-104982GB-C54), financed by the Ministry of Science and Innovation (Spain).

I am grateful to Godsuno Chela-Flores for his invaluable help in the translation of this article and for his insightful comments on a previous version.

Further Reading

  • Chambers, J. K., & Schilling, N. (Eds.). (2013). Handbook of language variation and change (2nd ed.). Chichester, UK: Wiley-Blackwell.
  • Chela-Flores, G. (2009). Oclusivas, fricativas y aproximantes en el español: hacia una explicación de la marcadez y estado actual de las obstruyentes sonoras. Opción, 25(59), 98–110.
  • Chela-Flores, G. (2019). El español estadounidense y el complejo dialectal hispánico: policentralidad, periferalidad dialectal y socioasimetrías. Glosas, 9(6), 11–31.
  • Hernández-Campoy, J. M., & Villena-Ponsoda, J. A. (2009). Standardness and nonstandardness in Spain: Dialect attrition and revitalization of regional dialects in Spain. International Journal of the Sociology of Language, 196–197, 181–214.
  • Jiménez-Fernández, R. (2020). Variación fonológica de la /d/ intervocálica en el sociolecto alto de Sevilla. Rilce, 36(2), 674–707.
  • Molina, I. (2020). Between dialect and standard: Dynamics of variation and change in Madrid. Spanish in Context, 17(2), 178–199.
  • Molina, I., Paredes, F., & Cestero, A. M. (2020). Sociolinguistic patterns and processes of convergence and divergence in Spanish. Spanish in Context, 17(2), 171–177.
  • Moreno, F. (2020). Variedades de la lengua española. Abingdon, UK: Routledge.
  • Moya, J. A., & Tejada, M. (2020). Patterns of the linguistic change in Andalusia. Spanish in Context, 17(2), 200–220.
  • Narbona, A., Cano, R., Morillo-Velarde, R., & Méndez-García de Paredes, E. (2010). La identidad lingüística de Andalucía. Seville, Spain: Centro de estudios andaluces.
  • Regan, B. (2020). El [ʃ]oquero: variación de/tʃ/en Huelva capital y pueblos alrededor. Estudios de fonética experimental, 29, 55–90.

References

  • Alarcos, E. (1964). Sobre la metafonía asturiana y su antigüedad. In Symposium sobre cultura asturiana de la alta Edad Media (Oviedo, 1961) (pp. 331–340). Oviedo, Spain: Ayuntamiento de Oviedo.
  • Almeida, M. (1990). Diferencias sociales en el habla de Santa Cruz de Tenerife. La Laguna, Spain: Instituto de Estudios Canarios.
  • Ávila, A. (1994). La variación reticular e individual de s/z en el vernáculo urbano malagueño: Datos del barrio de Capuchinos. Analecta Malacitana, 17(2), 343–367.
  • Blanco, A. (2004). Estudio sociolingüístico de Alcalá de Henares. Alcalá, Spain: Universidad de Alcalá.
  • Bybee, J. L. (2002). Word frequency and context of use in the lexical diffusion of phonetically conditioned sound change. Language Variation and Change, 14, 261–290.
  • Cabrera, S. (2009). Estudio sociolingüístico del español hablado en Telde (Gran Canaria) [PhD dissertation, Universidad de Las Palmas de Gran Canaria].
  • Calero, M. Á. (1993). Estudio sociolingüístico del habla de Toledo. Lleida, Spain: Pagès editors.
  • Carbonero, P., Álvarez-Cubero, J., Casas-Gómez, J., & Gutiérrez-Sánchez, I. (1992). El habla de Jerez. Estudio sociolingüístico. Jerez de la Frontera, Spain: Ayuntamiento de Jerez.
  • Chapman, P., Dubra, A., Martínez-Gil, F., & Tritica, D. (1983). El yeísmo en Covarrubias [Unpublished manuscript]. University of Southern California.
  • Chela-Flores, G. (1986). Las teorías fonológicas y los dialectos del Caribe hispánico. In R. Núñez-Cedeño, J. Guitart, & I. Páez-Urdaneta (Eds.), Estudios sobre la fonología del español del Caribe (pp. 21–30). Caracas, Venezuela: Ediciones Casa de Bello.
  • Chela-Flores, G. (2017). Language ideologies, intervarietal conflict and their repercussions on language and society: The case of the Hispanic dialectal complex. Dialectologia et Geolinguistica, 25, 123–150.
  • Cuevas, I. (2001). Variación social y reticular de las consonantes obstruyentes dentales (/θs/) y las palatales (/c/) en el Vernáculo Urbano Malagueño (VUM) [PhD dissertation, Universidad de Málaga].
  • Ferguson, C. (1990). From esses to aithes: Identifying pathways of diachronic change. In W. Croft, S. Kemmer, & K. Denning (Eds.), Studies in typology and diachrony: Papers presented to Joseph H. Greenberg on his 75th birthday (pp. 59–78). Amsterdam, The Netherlands: John Benjamins.
  • Fernández de Molina, E. (2014). El habla de Mérida [PhD dissertation, Universidad de Extremadura].
  • Fernández de Molina, E. (2018). Estudio sociolingüístico de la pronunciación de /-ɾ/ en el español de Granada. Itinerarios, 28, 219–241.
  • Fernández-Juncal, C. (1998). Variación y prestigio: estudio sociolingüístico en el oriente de Cantabria. Madrid, Spain: CSIC.
  • Fernández-Juncal, C. (2000). Neutro de materia y metafonía en el oriente de Cantabria. Salamanca, Spain: Universidad de Salamanca.
  • García-Amaya, L. (2008). Variable norms in the production of /θ‎/ in Jerez de la Frontera, Spain. In J. Siegel, T. C. Nagle, A. Lorente-Lapole, & J. Auger (Eds.), IUWPL7, gender in language: Classic questions, new contexts (pp. 49–71). Bloomington, IN: IULC.
  • García-Marcos, F. (1990). Estratificación social del español de la Costa Granadina. Almería, Spain: Universidad de Almería.
  • Gil-Peña, N. (2004). Estudio sociolingüístico de la fonética de la ciudad de Madrid [Dissertation, Universidad de Alcalá].
  • Gimeno, F. (2008). El cambio lingüístico estable. La elisión de -s final en español. Revista Internacional de Lingüística Iberoamericana (RILI), 6(2), 141–155.
  • Gómez, R., & Molina, I. (Eds.). (2013). Variación yeísta en el mundo hispánico. Madrid, Spain: Iberoamericana/Vervuert.
  • Gómez-Molina, J. R. (2012). Variación y cambios fónicos de la /d/ intervocálica en el español de Valencia: Proyecto preseea-preseval. LEA (Lingüística Española Actual), 34, 167–196.
  • Gómez-Molina, J. R. (2013). Interacción de factores internos y externos en la pronunciación de la /d/ intervocálica. In J. R. Gómez-Molina (Ed.), El español de Valencia: Estudio sociolingüístico (pp. 31–68). Bern, Switzerland: Peter Lang.
  • Gómez-Molina, J. R., & Gómez-Devís, B. (2016). ¡Vaya valla! El yeísmo en el español de Valencia. Boletín de Filología, 51(2), 59–88.
  • Gómez-Serrano, A. (1994). Aspectos del habla de Linares (Jaén). Málaga, Spain: Universidad de Málaga.
  • Harjus, J. (2018). Sociofonética andaluza y lingüística perceptiva de la variación: el español hablado en Jerez de la Frontera. Madrid, Spain: Iberoamericana/Vervuert.
  • Hernández-Cabrera, C., & Samper-Hernández, M. (2011). Rasgos fónicos regionales en los noticiarios de la televisión autonómica canaria. La velarización de la nasal final. In R. Ávila (Ed.), Variación del español en los medios (pp. 267–288). Mexico City, Mexico: El Colegio de México.
  • Højrup, T. (1983). The concept of life-mode. A form-specifying mode of analysis applied to contemporary Western Europe. Ethnologia Scandinavica, 1–50.
  • Holmquist, J. (1985). Social correlates of a linguistic variable: A study in a Spanish village. Language in Society, 14(2), 191–203.
  • Holmquist, J. (1988). Language loyalty and linguistic variation: A study in Spanish Cantabria. Dordrecht, The Netherlands: Foris Publications.
  • Jiménez-Fernández, R. (2015). Sobre el mantenimiento y la pérdida de la /d/ intervocálica en el habla de Sevilla (estudio en el sociolecto bajo). In J. Santana, M. L.-C. Gómez, A. Zerva (Coords.) (Eds.), La variación en el español actual: Estudios dedicados al profesor Pedro Carbonero (pp. 193–210). Seville, Spain: Universidad de Sevilla.
  • Kapović, M. (2014). Análisis sociolingüístico de la variable (s) en el habla de Ciudad Real [PhD dissertation, Sveučilište U Zadru].
  • Kapović, M. (2015). El debilitamiento de la s explosiva en español: extensión geográfica y factores condicionantes. Lingüística Española Actual, 37(1), 73–100.
  • Kapović, M. (2017). Variable within variable – simultaneous stability and change. The case of syllable-final /s/ in Ciudad Real. Borealis: An International Journal of Hispanic Linguistics, 6(2), 249–283.
  • Labov, W. (1994). Principios del cambio lingüístico, volumen 1: factores internos. Madrid, Spain: Gredos.
  • López de Aberasturi, I. (1997). Estudio sociolingüístico de la elisión de /n/ final con valor gramatical en el habla de Ayamonte. In A. Narbona & M. Ropero (Eds.), El habla andaluza: Actas del Congreso del habla andaluza (pp. 593–605). Seville, Spain: Seminario Permanente del Habla Andaluza.
  • López-Morales, H. (1983). Estratificación social del español de San Juan de Puerto Rico. Ciudad de México: UNAM.
  • López-Morales, H. (1992). El español del Caribe. Madrid, Spain: Mapfre.
  • López-Moreno, M. (2018). Acercamiento sociolingüístico a las realizaciones de /-l/ en el español hablado en la ciudad de Granada. Itinerarios, 28, 243–271.
  • Martín-Butragueño, P. (1995). La variable (s) en el sur de Madrid. Contribución al estudio de la frontera de las hablas meridionales del español. Anuario de Letras, 33, 5–57.
  • Martín-Butragueño, P. (2004). El contacto de dialectos como motor del cambio lingüístico. In P. Martín-Butragueño (Ed.), Cambio lingüístico.Métodos y problemas (pp. 81–144). México: El Colegio de México.
  • Martínez, F. M. (1983). Fonética y sociolingüística en la ciudad de Burgos. Madrid, Spain: CSIC.
  • Melguizo, E. (2007a). Convergencia y divergencia dialectal: estudio de la variación lingüística granadina. Lingüística Española Actual, 29, 261–285.
  • Melguizo, E. (2007b). La variación de /θs/: estudio comparativo de dos muestras de población granadinas. ELUA, 21, 245–259.
  • Molina, I. (1997). Dos cambios fonético-fonológicos en el español peninsular: aspectos geográficos y sociales. In F. Moreno (Ed.), Trabajos de sociolingüística hispánica (pp. 69–91). Madrid, Spain: Universidad de Alcalá.
  • Molina, I. (1998). La fonética de Toledo: contexto geográfico y social. Madrid, Spain: Universidad de Alcalá.
  • Molina, I. (2006). Innovación y difusión del cambio lingüístico en Madrid. Revista de Filología Española, 86(1), 127–149.
  • Molina, I. (2013). Yeísmo madrileño y convergencia dialectal campo/ciudad. In R. Gómez & I. Molina (Eds.), Variación yeísta en el mundo hispánico (pp. 93–110). Madrid, Spain: Iberoamericana/Vervuert.
  • Molina, I. (2015). La variable sociolingüística -/s/ en el distrito de Vallecas (Madrid). In A. Cestero, I. Molina, & F. Paredes (Eds.), Patrones sociolingüísticos de Madrid (pp. 91–116). Bern, Switzerland: Peter Lang.
  • Molina, I., & Paredes, F. (2014). Sociolingüística de la elisión de la dental -/d/- en Madrid (distrito de Salamanca). Cuadernos de Lingüística de El Colegio de México, 2, 55–114.
  • Molina, I., & Paredes, F. (2015). La conservación de la dental -/d/- en el distrito de Salamanca. In A. Cestero, I. Molina, & F. Paredes (Eds.), Patrones sociolingüísticos de Madrid (pp. 63–89). Bern, Switzerland: Peter Lang.
  • Momcilovic, N. (2009). A sociolinguistic analysis of /s/-aspiration in Madrid Spanish. München, Germany: Lincom Europa.
  • Moratal, V. (2011). Estudio sincrónico y contrastivo sobre el yeísmo en Gandía: enfoque variacionista y sociolingüístico. Ianua: Revista Philologica Románica, 11, 135–154.
  • Moreno, F. (2004). Cambios vivos en el plano fónico del español: variación dialectal y sociolingüística. In R. Cano (Ed.), Historia de la lengua española (pp. 973–1009). Barcelona, Spain: Ariel.
  • Moreno, F. (2009). La lengua española en su geografía: Manual de dialectología hispánica. Madrid, Spain: Arco/Libros.
  • Moya, J. A. (1979). La pronunciación del español en Jaén. Universidad de Granada.
  • Moya, J. A. (2007). Noticia de un sonido emergente: la africada dental procedente del grupo -st- en Andalucía. Revista de Filología, 25, 457–465.
  • Moya, J. A. (2018). Sobre el equilibrado reajuste de las hablas andaluzas. Itinerarios, 28, 35–66.
  • Moya, J. A., & García-Wiedemann, E. (1995). El habla de Granada y sus barrios. Granada, Spain: Universidad de Granada.
  • Moya, J. A., García-Wiedemann, E., Montoro, E., Sosiński, M., Torres, M., Pose-Furest, F., & Melguizo, E. (2012). La /d/ intervocálica en Granada: factores lingüísticos y sociales. In E. Walluch & J. A. Moya (Eds.), Español hablado: Estudios sobre el corpus PRESEEA-Granada (pp. 95–148). Warsaw, Poland: Instituto de Estudios Ibéricos e Iberoamericanos.
  • Moya, J. A., Montoro del Arco, E. T., Manjón-Cabeza, A., Sosinski, M., Tejada, M. S., Torres M. C., Pose, F., & Sánchez, F. J. (2014). Un paso más hacia la convergencia. La /x/ en Granada: factores sociales. Revista Española de Lingüística, 44(1), 86–113.
  • Moya, J. A., & Sosiński, M. (2015). La inserción social del cambio: La distinción s/θ‎ en Granada: Análisis en tiempo aparente y en tiempo real. Lingüística Española Actual, 37(1), 33–72.
  • Narbona, A., Cano, R., & Morillo, R. (2011). El español hablado en Andalucía. Seville, Spain: Universidad de Sevilla.
  • Paredes, F. (2001). El habla de La Jara. Los sonidos (estudio sociolingüístico). Alcalá, Spain: Universidad de Alcalá.
  • Paredes, F. (2013). Las consonantes palatales sonoras en la comarca castellano-extremeña de La Jara: historia, geografía y sociedad. In R. Gómez & I. Molina (Eds.), Variación yeísta en el mundo hispánico (pp. 59–92). Madrid, Spain: Iberoamericana/Vervuert.
  • Payán, P. (1988). La pronunciación del español en Cádiz. Cadiz, Spain: Universidad de Cádiz.
  • Pérez-Martín, A. (2010). Estudio sociolingüístico del español hablado en El Hierro. Las Palmas de Gran Canaria, Spain: Universidad de Las Palmas de Gran Canaria.
  • Regan, B. (2017). A study of ceceo variation in Western Andalusia (Huelva). Studies in Hispanic and Lusophone Linguistics, 10(1), 119–160.
  • Rodríguez-Prieto, J. (2008). Distribución geográfica del “jejeo” en español y propuesta de reformulación y extensión del término. Revista Española de Lingüística, 38(2), 129–144.
  • Rost, A. (2014). Una panorámica del yeísmo: ¿un proceso acabado o en construcción? Revista Internacional de Lingüística Iberoamericana, 23, 141–163.
  • Ruch, H. (2013). Investigating gradual metathesis: Phonetic and lexical factor son /-s/ aspiration in Andalusian Spanish. Working Papers in Linguistics, 19(2), 170–180.
  • Ruiz-Domínguez, M. M. (1997). Estudio sociolingüístico del habla de Melilla [PhD dissertation, Universidad de Alcalá].
  • Ruiz-Martínez, A. (2003). Estudio fonético del nordeste de la Comunidad de Madrid. Alcalá, Spain: Universidad de Alcalá.
  • Ruiz-Sánchez, C. (2009). The variable behavior of /r/ in syllable-final and word-final position in the Spanish variety of Alcalá de Guadaíra (Seville): The role of lexical frequency. Köln, Germany: Lambert Academic Publishing.
  • Salvador, G. (1964). La fonética andaluza y su propagación social. In Presente y futuro de la lengua española (Vol. 2, 183–188). Madrid, Spain: Instituto de Cultura Hispánica.
  • Salvador-Salvador, F. (1980). Niveles sociolingüísticos de seseo, ceceo y distinción en la ciudad de Granada. Español Actual, 37-38, 25–32.
  • Samper, J. A. (1990). Estudio sociolingüístico del español de Las Palmas de Gran Canaria. Las Palmas de Gran Canaria, Spain: La Caja de Canarias.
  • Samper, J. A. (2011). Sociophonogical variation and change in Spain. In M. Díaz-Campos (Ed.), The handbook of Hispanic sociolinguistics (pp. 98–120). Chichester, UK: Wiley-Blackwell.
  • Samper, J. A., & Hernández-Cabrera, C. (1995). La variación de -/s/ en el español culto de Las Palmas de Gran Canaria. Philologica Canariensia, 2, 391–408.
  • Samper, J. A., & Hernández-Cabrera, C. (2016). Sobre un cambio en el español de Las Palmas de Gran Canaria: de [s] a [h] en el contexto prevocálico tónico. In A. López-Serena, A. Narbona, & S. del Rey Quesada (Dirs.), El español a través del tiempo. Estudios ofrecidos a Rafael Cano (Vol. 1, pp. 119–134). Seville, Spain: Universidad de Sevilla.
  • Samper, J. A., Hernández-Cabrera, C., Troya, M., Pérez-Martín, A., & Samper-Hernández, M. (2011). Estudio sociolingüístico de la variación de -/d/- en la ciudad de Las Palmas de Gran Canaria [Paper presentation]. XVI Congreso Internacional de la ALFAL, Alcalá de Henares, Spain.
  • Samper, J. A., & Samper-Hernández, M. (2020). The weakening of intervocalic /d/ in the Spanish of Las Palmas de Gran Canaria: Comparison with other speech communities. Spanish in Context, 17(2), 222–247.
  • Santana, J. (2016). Seseo, ceceo y distinción en el sociolecto alto de la ciudad de Sevilla: nuevos datos a partir de los materiales de PRESEEA. Boletín de Filología de la Universidad de Chile, 51(2), 255–280.
  • Silva-Corvalán, C. (2001). Sociolingüística y pragmática del español. Washington, DC: Georgetown University Press.
  • Tejada, M. S. (2012). Los factores lingüísticos de la /-s/ implosiva en el nivel de estudios altos de Granada. Normas: Revista de Estudios Lingüísticos Hispánicos, 2, 185–207.
  • Tejada, M. S. (2015). Contribución al estudio de los patrones sociolingüísticos del español de Granada [PhD dissertation, Universidad de Granada].
  • Terrell, T. (1979). Problemas de los estudios cuantitativos de procesos fonológicos variables: datos del Caribe hispánico. Boletín de la Academia Puertorriqueña de la Lengua Española, 7, 145–165.
  • Torreira, F. (2012). Investigating the nature of aspirated stops in Western Andalusian Spanish. Journal of the International Phonetic Association, 42(1), 49–63.
  • Torres, A., Fernández-Planas, A., Blasco, E., Forment, M., Pérez, M. Á., & Illamola, C. (2013). Estudio del yeísmo en el español de Barcelona a partir de materiales de PRESEEA. In R. Gómez & I. Molina (Eds.), Variación yeísta en el mundo hispánico (pp. 19–37). Madrid, Spain: Iberoamericana/Vervuert.
  • Vida, M. (2004). Estudio sociofonológico del español hablado en Málaga. Alicante, Spain: Universidad de Alicante.
  • Vida, M. (2015). Resilabificación de la aspiración de /-s/ ante oclusiva dental sorda. Parámetros acústicos y variación social. In A. Cabedo-Nebot (Ed.), Perspectivas actuales en el análisis fónico del habla: Tradición y avances en la fonética experimental (pp. 441–451). Valencia, Spain: Universitat de València.
  • Vida, M. (2016). Correlatos acústicos y factores sociales en la aspiración de /-s/ preoclusiva en la variedad de Málaga (España): Análisis de un cambio fonético en curso. Lingua Americana, 38, 15–37.
  • Villena, J. (1997a). Sociolingüística andaluza y sociolingüística del andaluz: problemas y métodos. In A. Narbona & M. Ropero (Eds.), El habla andaluza. Actas del Congreso del habla andaluza (pp. 277–347). Seville, Spain: Seminario Permanente del Habla Andaluza.
  • Villena, J. (1997b). Convergencia y divergencia dialectales en el continuo sociolingüístico andaluz: datos del vernáculo urbano malagueño. Lingüística Española Actual, 19, 83–125.
  • Villena, J. (2001). La continuidad del cambio lingüístico. Granada, Spain: Universidad de Granada.
  • Villena, J. (2008). Sociolinguistic patterns of Andalusian Spanish. International Journal of the Sociology of Language, (193-194), 139–160.
  • Villena, J. (2012). Patrones sociolingüísticos del español de Andalucía. In J. Villena & A. Ávila (Eds.), Estudios sobre el español de Málaga. Pronunciación, vocabulario y sintaxis (pp. 27–66). Madrid, Spain: Sarriá.
  • Villena, J., & Ávila, A. (2014). Dialect stability and divergence in southern Spain: Social and personal motivations. In K. Braunmüller, S. Höder, & K. Kühl (Eds.), Stability and divergence in language contact: Factors and mechanisms (pp. 207–238). Amsterdam, The Netherlands: John Benjamins.
  • Villena, J., & Moya, J. A. (2016). Análisis comparativo de un cambio fonológico erosivo. Variación de /d/ intervocálica en dos comunidades de habla (Granada y Málaga). Boletín de Filología: Universidad de Chile, 51(2), 281–321.
  • Villena, J., & Vida, M. (2020). Variation, identity and indexicality in southern Spanish. In M. Cerruti, & S. Tsiplakou (Eds.), Intermediate language varieties: Koinai and regional standards in Europe. Amsterdam, The Netherlands: John Benjamins.
  • Zamora-Vicente, A. (1967). Dialectología española. Madrid, Spain: Gredos.
  • Zmijanac, S. (2012). Estudio sociolingüístico y fonéticoacústico de la variación de la obstruyente velar /x/ en el español hablado en Málaga. In J. Villena & A. Ávila (Eds.), Estudios sobre el español de Málaga: Pronunciación, vocabulario y sintaxis (pp. 209–242). Madrid, Spain: Sarriá.

Notes

  • 1. In this table, the data collected in Samper (2011) have been supplemented with the results of new studies. Some observations were added to aid in the correct interpretation of the information in the table: (a) Momcilovic’s and Martín-Butragueño’s data correspond to the conversation style; (b) In Kapović’s results, the [r] variant percentage is not considered; (c) the Melilla figures refer to the monolingual population; (d) the Granada results do not include aspiration with assimilation, the postaspiration, and the voiceless dentoalveolar affricate in the /st/ and /kt/ combinations.

  • 2. However, Moya (2007) indicates that in Seville and Antequera the phenomenon is supported by women and the middle sociolects, and it has a growing prestige in Seville and a consolidated one in Antequera.

  • 3. The data here are restricted to word-final positions.

  • 4. Some researchers have also considered assimilation and aspiration.

  • 5. Also Harjus (2018, p. 178) states that in Jerez de la Frontera “[l] for the /r/ phoneme almost does not occur in the A corpus.”

  • 6. “The majority result, geographically and demographically of the -r/-l confusion in Andalusia is -r” (Narbona et al., 2011, p. 214).

  • 7. Jerez de la Frontera presents a high percentage of rhotacism (17%) in word-internal position; in word-final position only [r] realizations stand out in the articles el and al (Harjus, 2018, p. 181).

  • 8. There is no weakening in the peninsular central zone (see Blanco, 2004; Martín-Butragueño, 2004; Ruiz-Martínez, 2003). The ALPI showed that nasal velarization was produced in the peninsular western and southern regions. The ALEA revealed that this feature was also present in Eastern Andalusia (also see Zamora-Vicente, 1967, pp. 323–324).

  • 9. In PRESEEA the realizations of /d/ have been analyzed in an ampler context than those in previous sociolinguistic research, which, among other reasons, could explain the lower elision indices in recent studies of some communities.

  • 10. Moratal’s study of Gandía in the same province (Moratal, 2011), shows a distinction of 64.3%.

  • 11. Both terms refer to “the occasional lexically conditioned substitution of [s] by [h] in prenuclear position” (Rodríguez-Prieto, 2008, p. 130). Narbona et al. (2011, p. 226) point out that it is “a restricted [phenomenon,] all the more in the speech of very localised speakers, generally belonging to very low sociocultural levels and almost always in very careless and familiar speech.” Heheo also has been documented in many American zones (Kapović, 2015).

  • 12. The preliminary study by Zmijanac (2012) on Málaga found a fricative variant frequency much higher than that of Granada (69.8%), in contrast with the reduced level of aspiration (22.8%). Elision reaches 7.5%.