7/30/2019 New Mexican Spanish -s.pdf http://slidepdf.com/reader/full/new-mexican-spanish-spdf 1/12 New Mexican Spanish: Insight into the Variable Reduction of "la ehe inihial" (/s-/) Esther L. Brown University of Colorado Abstract: For close to a century, a noted feature of the Spanish of New Mexico has been the variable aspiration and deletion of syllable-initial /s/ (Espinosa 1909), yet no empirical investigations have been undertaken to study this process within this variety of Spanish in the United States. In fact, syllable-initial /s/ world-wide is rarely quantified, with few exceptions (e.g., Garcia and Talion 1995). This relative lack of empirical analysis means that there is much to be learned from quantitative analysis of this variable phonological reductive process. The present study examines the exact nature of syllable-initial /s/ reduction using quantitative methods to challenge the notions that syllable-initial /s-/ reduction stems directly from syllable-final /-s/ reduction, and that intervocalic /s-/ reduction is found only in a limited set of lexemes. Key Words: alveolar fricative sibilants, aspiration, coronal articulation, deletion, initial sibilant, phonetic residue, word-final sibilant 1. Introduction Syllable-final /s/ holds a special place of importance as themost widely studied variable in Hispanic linguistics. This may not be surprising given the ability of /s/ to successfully demarcate social and geographic varieties of Spanish, and given the ways in which deletion of word-final /s/ has been said to interact with the morphology of the language. Focus on the Isi in this syllable position is not only copiously represented in the literature of academic research, but is also part of the consciousness of speakers in Spanish-speaking varieties around the globe. In the Spanish of New Mexico,1 this phonological weakening process occurs, unexpectedly, in syllable-initial position. Processes affecting New-Mexican syllable-initial /s/ include aspiration and deletion (deemed "reduction" in this analysis), which is contrasted in this work with reten tion of the sibilant. Aspirated tokens, cases inwhich the sibilant loses coronal articulation, are illustrated in examples la and b below, and deleted tokens, instances where the I si is removed from the string of speech leaving no phonetic residue, are exemplified in 2a and b. la) pero cahi todos (pero casi todos [m.214.16])2 "but almost all of them" lb) el trabajo de hemento (el trabajo de cemento [m. 162.5]) "the cement work" 2a) en eeos a?os (en esos a?os [m. 162.8]) "during those years" 2b) ee me hace que (se me hace que [f.317.4]) "it seems tome that" As can be seen in the examples above, reduction of this type is found inword-medial (examples 1 a, 2a) andword-initial positions (examples lb, 2b). Brown, Esther L. "New Mexican Spanish: Insight into the Variable Reduction of 'la ehe inihiaV (/s-/)" Hispania 88.4 (2005): 813-824 Esther L. Brown: "New Mexican Spanish: Insight into the Variable Reduction of "La ehe inihial" (/s-/) Hispania, Vol. 88, No. 4 (Dec., 2005), pp. 813-824.
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Syllable-final /s/ holds a special place of importance as themost widely studied variable in
Hispanic linguistics. This may not be surprising given the ability of /s/ to successfully
demarcate social and geographic varieties of Spanish, and given the ways in whichdeletion of word-final /s/ has been said to interact with themorphology of the language. Focus on
the Isi in this syllable position is not only copiously represented in the literature of academic
research, but is also part of the consciousness of speakers in Spanish-speaking varieties around
the globe.
In the Spanish of New Mexico,1 this phonological weakening process occurs, unexpectedly,in syllable-initial position. Processes affecting New-Mexican syllable-initial /s/ include aspiration
and deletion (deemed "reduction" in this analysis), which is contrasted in this work with reten
tion of the sibilant. Aspirated tokens, cases inwhich the sibilant loses coronal articulation, are
illustrated in examples la and b below, and deleted tokens, instances where the Isi is removed
from the string of speech leavingno
phonetic residue,are
exemplified in 2a and b.
la) pero cahi todos (pero casi todos [m.214.16])2
"but almost all of them"
lb) el trabajo de hemento (el trabajo de cemento [m. 162.5])
"the cement work"
2a) en eeos a?os (en esos a?os [m. 162.8])
"during those years"
2b) ee me hace que (se me hace que [f.317.4])
"it seems tome that"
As can be seen in the examples above, reduction of this type is found inword-medial (examples1a, 2a) and word-initial positions (examples lb, 2b).
Brown, Esther L.
"New Mexican Spanish: Insight into the Variable Reduction of 'la ehe inihiaV (/s-/)"
Hispania 88.4 (2005): 813-824
Esther L. Brown: "New Mexican Spanish: Insight into the Variable Reduction of "La ehe inihial" (/s-/)
Hispania, Vol. 88, No. 4 (Dec., 2005), pp. 813-824.
contrast to the plethora of theoretical and empirical analyses of syllable-final /s/ reduction, there
isa
striking paucity of empirical studies concerning reduction of/s/ in syllable-initial position. Infact, this discrepancy is highlighted by Mason (7) in his extensive bibliographic survey of
studies dealing with aspiration and deletion of/s/ in dialects of Spanish. This obvious deficiency
of investigations into syllable-initial reduction is an oversight in the field of Hispanic Linguistics.
This present work, a large-scale analysis of syllable-initial /s/ reduction in spoken Spanish,
addresses this neglect, but also is important for our understanding of New-Mexican Spanish in
general. Syllable-initial /s/ aspiration has long been a phonological characteristic noted in the
spoken Spanish of New Mexico and Southern Colorado; for brief mention ofthat fact, see?in
addition to the items cited above?Bills, Bills and Ornstein, Bills and Vigil, Canfield, C?rdenas,
Cobos, Cotton and Sharp, Guti?rrez, and Lipski, Espa?ol. Although a few empirical studies have
been conducted on other varieties ofSpanish (Brown;
Brown and Torres Cacoullos"Qu?,"
"Dif
ferent"; Fl?rez; Garc?a and Talion; Lipski "Reducci?n"; L?pez Scott), revealing as part of their
findings significant linguistic and extra-linguistic factors involved in the variable reduction of
syllable-initial /s/ in certain varieties, no such detailed information was available for the New
Mexican variety.
Specifically the present investigation will examine forNew-Mexican Spanish the following
notions: (1) that a key ingredient for syllable-initial /s/ reduction is that itbe found in varieties of
Spanish where there is acute weakening of /s/ in general (M?ndez Dosuna, "Can"), (2) that
reduction in syllable-initial position is an extension of processes which originate in syllable-final
position (Ferguson; M?ndez Dosuna, "Aspiraci?n," "Can"; Penny; Terrell "Final"), (3) that
initial /s/ aspiration is specific to intervocalic environment (Lipski, "Many"; Vaquero), and (4)
that syllable-initial Isi aspiration is limited to a small set of lexemes (Bills and Ornstein; Garcia and
Talion). Given themany linguistic insights gleaned from the study of syllable-final Is/ processes,
and the general lack of understanding regarding syllable-initial Isi reduction, much stands to be
gained from an analysis of this sort.
2.Data and Methods
To achieve the goals outlined above, an analysis was undertaken of 10,770 tokens of Isi in
the spoken Spanish of New Mexico. All of the consultants are native speakers of New-Mexican
Spanish originating from and/or living in present day New Mexico and southern Colorado. The
speech of 24 speakers (both male and female) is analyzed in this study, and themajority of the
data comes from theNMCOSS project.
Initiated in 1991, theNMCOSS project documents, via interviews with 350 native speakers,
the Spanish language spoken throughout the state of New Mexico and the sixteen counties of
southern Colorado (Bills and Vigil). The data was collected by trained field workers who tape
recorded interviews involving both controlled elicitation and guided conversation. The
interviews averaged three-and-a-half hours in length, beginning with compilation of personal
information regarding the consultant, followed by specific linguistic elicitation and free
conversation (Bills andVigil 46).
All cases of Isi (in four syllable positions: syllable-initial word-initial, syllable-initial word
medial, syllable-final word-medial, syllable-final word-final) were coded by the investigator
according to phonetic realizations. Both the voiced and voiceless alveolar fricative sibilants, [z]
and [s], are considered non-reduced variants. The voiceless glottal fricative realization of Is/
(commonly referred to as aspirated Isi and coded as [h]) and the elided or completely reduced
tokens of Is/ (coded as [0]) are reduced variants. With respect to aspiration, no phonetic
distinction is drawn between syllable-initial and syllable-final Isl realizations, although acoustic
(Mann and Soli) and perceptual (Widdison) studies suggest that Isl has two very distinct sets of
characteristics in syllable-initial, as opposed to syllable-final, position.
These tokens wereanalyzed for their rates and patterns of reduction and were also proc
essed by VARBRUL, a logistical regression analysis. Variable rule analysis is a type of
multivariate program that considers different environmental factors simultaneously and
measures their effects on the choice of variants (Rand and Sankoff), in this case, reduced versusretained realizations. The variable rule analyses conducted with VARBRUL provide a detailed
account of which factor groups contribute significantly to the reduction of syllable-initial Is/; the
analyses also rank the factor groups by the magnitude of effect they have on the variation
(indicated by the range). Additionally, within each factor group, the individual factors are ranked
according to their factor weight, which indicates the degree towhich they favor (> .50) or disfavor
(< .50) reductive processes.
3. Discussion
Findings regardingthe nature of
syllable-initialIsi
aspirationinother dialects do not seem to
apply to the New Mexico data. Specifically, studies of other dialects have concluded that
syllable-initial Isi reduction is a result of extreme Isi reduction generally in the dialect, that it is an
extension of processes occurring syllable-finally, that it is confined to intervocalic context and
that it is also lexically limited. Each of these impressions will be discussed in turn, and itwill be
shown that theNew-Mexican data support none of these assumptions. New-Mexican Spanish, it
appears, does not follow the same /s/-articulation parameters that other well-studied varieties do.
3.1. Overall/s/ reduction
Syllable-initial Isi reduction is often perceived to be a phenomenon found most typically in
dialects of Spanish where Isi reduction is extreme, such as in theDominican Republic (Vaquero).
M?ndez Dosuna ("Can") suggests that syllable-initial Isi reduction, in fact, is found "only in
*Cuba, Argentina data taken from Terrell ("Constraints," "Aspirations," Hispania); as summarized in
Bybee (140). Chihuahua data taken from Brown and Torres Cacoullos ("Qu?," "Different").
However, an interesting cross-dialectal pattern emerges in the data. The two varieties included in
Table 2 that have word-final reduction rates pre-vocalically that are superior (or roughly equal) to
reduction inpre-consonantal positions (New Mexico 61% vs. 62%; Chihuahua 42% vs. 47%), are
precisely the dialects with considerable syllable-initial Isl reduction. It is this generalization of
word-final Isl reduction to pre-vocalic contexts, coupled with amatching path of phonetic reduc
tion ([s]>
[h]) inword-initial, post-vocalic position that leads Lipski ("Many") to note that
reduction in these positions are processes which appear to be "intimately related" (198). Does the
co-occurrence of syllable-initial Isl reduction with generalized word-final, pre-vocalic reduction
imply a diachronic process extending from syllable-final to syllable-initial position? This ques
tion is addressed in the following section.
3.2. Final > Initial Continuum
If reductions in these twoword positions are related (word-final, pre-vocalic andword-initial,
post-vocalic), do linguistic factors constrain them in the same way in each context? Separatemultivariate analyses of both word-initial and word-final Isl reduction inNew Mexico, in fact,
reveal a similar ordering of constraints within factor groups. This information is summarized in
Tables 3 and 4.
Table 3
VARBRUL Analyses of Linguistic Factors Having the
Greatest Effect on Word-Initial Isl Reduction
Word-Initial
Input: .10
Total N: 2585
% Reduction Factor Weight % Data
Preceding Phonological Environment
non-high vowel 257050
high vowel 7332
pause .32 24
consonant 4243
Range 46
Following Phonological Environment
non-high vowel 236253
high vowel 6376
Range 25
[Other factor groups selected as significant: frequency, stress
(Log likelihood =-947.244, p=.000, Chi-square per cell=
characteristics in syllable-initial, as opposed to syllable-final, position.
These tokens wereanalyzed for their rates and patterns of reduction and were also proc
essed by VARBRUL, a logistical regression analysis. Variable rule analysis is a type of
multivariate program that considers different environmental factors simultaneously and
measures their effects on the choice of variants (Rand and Sankoff), in this case, reduced versusretained realizations. The variable rule analyses conducted with VARBRUL provide a detailed
account of which factor groups contribute significantly to the reduction of syllable-initial Is/; the
analyses also rank the factor groups by the magnitude of effect they have on the variation
(indicated by the range). Additionally, within each factor group, the individual factors are ranked
according to their factor weight, which indicates the degree towhich they favor (> .50) or disfavor
(< .50) reductive processes.
3. Discussion
Findings regardingthe nature of
syllable-initialIsi
aspirationinother dialects do not seem to
apply to the New Mexico data. Specifically, studies of other dialects have concluded that
syllable-initial Isi reduction is a result of extreme Isi reduction generally in the dialect, that it is an
extension of processes occurring syllable-finally, that it is confined to intervocalic context and
that it is also lexically limited. Each of these impressions will be discussed in turn, and itwill be
shown that theNew-Mexican data support none of these assumptions. New-Mexican Spanish, it
appears, does not follow the same /s/-articulation parameters that other well-studied varieties do.
3.1. Overall/s/ reduction
Syllable-initial Isi reduction is often perceived to be a phenomenon found most typically in
dialects of Spanish where Isi reduction is extreme, such as in theDominican Republic (Vaquero).
M?ndez Dosuna ("Can") suggests that syllable-initial Isi reduction, in fact, is found "only in
Ingeneral terms, the low reduction rates found in these contexts reveal that previous pause or
consonant does not favor reduction, as is also indicated by theVARBRUL results (Tables 3 and
4).
Although syllable-initial Isl reduction is greater in the intervocalic context, these data from
New Mexico suggest that syllable-initial reduction extends to phonological environments
beyond the narrowly delimited contexts proposed for variation found in other studies. More im
portantly, an examination of the data reveals that this reductive process is not limited to a small
set of lexemes that the literature frequently cites as being found inother locales. The lexically dif
fuse nature of this phonological-reductive process inNew Mexico is the focus of the next
section.
3.4 Lexical Extension
Canonical examples of syllable-initial Isl reduction include words such as nosotros "we" and
casa "house/home" (Bills and Ornstein) orHi, he?or "Yes, sir." Indeed, in other Spanishspeaking regions, such as South Texas, this reductive process appears to be lexically limited
(Garcia and Talion). The reduction of syllable-initial Isl inNew Mexico, however, seems to be
productive in present-day usage.
The 5633 tokens of syllable-initial Isl described here are found in 385 different words. Of
these 385 types containing a syllable-initial Isl, there is variation in 127, thus affecting almost 33%
of the /s/-initial lexemes in the study. However, excluding types with only one token in the data
which due to their lone status logically cannot demonstrate variation within this corpus, themag
nitude of the syllable-initial Isl variation is somewhat greater. Of the remaining 258 types with two
ormore tokens in the corpus, 112 show variation. This corpus of New Mexican Spanish thus re
flects reduction ofsyllable-initial
Isl in 43% of all /s/-initial lexical items. What ismore,
such
reduction clearly extends to an array of lexical items, in contrast towhat has been found in the
Spanish of other regions.
Forthose words showing variation, syllable-initial /s/' s rate of reduction ranges from as little
as 2% reduction to 100%. Plainly, not all types are affected to the same degree, and some lexical
effects are apparent. Table 8 highlights lexemes occurring three or more times in the corpus with
higher than average reduction rates (>16% word-initially and >30% word-medially) for the syl
lable-initial Isl. (This represents a selection of all words of three or more occurrences whose rate
of reduction exceeds the average for that position. Words with just one or two tokens may not be
representative due to the limited number of occurrences.)Table 8
Syllable-Initial Isl Types With Reduction Rates Higher Than Average Per Word Position
Word-Initial % Reduction Total N Word-Medial % Reduction Total N
If the higher rates of reduction were a result of synchronie morpheme-boundary processes,
one could postulate comparable rates of reduction for nos and for nos + otros (nosotros), since
the rule could apply equally to the form nos inboth instances. This isnot what we find in theNew
Mexico data. The word-final Isi o? nos reduces at a rate of 79% (N=
108), and the rate of word
medial Isi reduction for nosotros is 90% (N =110). The rates of reduction for the two words are
significantly different, and the reduction in the word nosotros does not seem to stem from a
process occurring in the rhyme of themorpheme nos (p=
0.0215, Chi-square=
5.282608).5 Conse
quently, as in other varieties where nohotros is considered a fossilized form (e.g., Garcia and
Talion), the reduced tokens of the first-person-plural subject pronoun inNew Mexico do not
seem tied tomorpheme-final Isi reduction of nos, but rather reflect processes affecting a separate
lexical item, nosotros.6
4. Conclusion
This study is the first to address with quantitative data the as-yet ill-understood phenomenon of syllable-initial Isi aspiration and deletion in the Spanish of New Mexico. As noted
Bills, Garland, and J. Ornstein. (1976). "Linguistic Diversity in Southwest Spanish." Studies in Southwest
Spanish. D. Bowen and J. Ornstein (Eds.). Rowley: Newbury House Publishers. 4-16.
Bills, Garland, and N. Vigil. (1999). "Ashes to Ashes: The Historical Basis for Dialect Variation in New Mexican
Spanish." Romance Philology 53: 43-67.
Brown, Dolores. (1993). "El polimorfismo de la /s/ explosiva en el noroeste de M?xico." Nueva Revista de
Filolog?a Hisp?nica1: 159-76.
Brown, Esther L., and Rena Torres Cacoullos. (2002). "?Qu? le vamo(h) a(h)er?: Taking the Syllable Out of
Spanish /s/ Reduction." University of Pennsylvania Working Papers in Linguistics: Papers from NWAV 30.
D. Johnson and T. S?nchez (Eds.). Philadelphia: U of Pennsylvania P. 17-32.
?.(2003). "Spanish /s/: A Different Story From Beginning (Initial) to End (Final)." A Romance Perspective in
Language Knowledge and Use. Selected Papers from the 31st Linguistic Symposium of Romance Lan
guages (LSRL). R. N??ez-Cede?o, L. L?pez and R. Cameron (Eds.). Amsterdam: John Benjamins. 21-38.
Bybee, Joan. (2001). Phonology and Language Use. Cambridge, MA: Cambridge UP.
Canfield, D. Lincoln. (1981). Spanish Pronunciation in the Americas. Chicago: U of Chicago P.
C?rdenas, Daniel. (1975). "Mexican Spanish." El lenguaje de los Ch?canos: Regional and Social Charac
teristics Used by Mexican Americans. E. Hern?ndez-Ch?vez, A. Cohen, A. Beltramo (Eds.). Arlington, VA:
Center for Applied Linguistics. 1-5.
Cobos, Rub?n. (1983). A Dictionary of New Mexico and Southern Colorado Spanish. Santa Fe: The Museum of
New Mexico P.
Cotton, Eleanor, and John M. Sharp. (1988). Spanish in the Americas. Washington, DC: Georgetown UP.
Espinosa, Aurelio. (1909). Studies in New Mexican Spanish: Part I: Phonology. Albuquerque, NM: U of New
Mexico P.
Ferguson, Charles A. (1990). "From esses to aitches: Identifying Pathways of Diachronic Change." Studies in
Typology and Diachrony: Papers Presented to Joseph H. Greenberg on His 75th Birthday. W. Croft, K.
Denning and S. Kemmer (Eds.). Philadelphia, PA: John Benjamins. 59-78.
Fl?rez, Luis. (1951). La pronunciaci?n del espa?ol en Bogot?. Bogota: Publicaciones del Instituto Caro y
Cuervo.
Garc?a, MaryEllen, and M. Talion. (1995). "Postnuclear /s/ in San Antonio Spanish: Nojotros no aspiramos."
Georgetown Journal of Languages and Linguistics 3.2?4: 139-62.
Guti?rrez, John. (1981). "An Analysis of the Phoneme /s/ in New Mexico Spanish." Proceedings of the Ninth
Annual Southwestern Area Language and Linguistics Workshop. C. Elerick (Ed.). El Paso: University of
Texas at El Paso. 234-39.
Harris, James. (1993). "Integrity of Prosodie Constituents and the Domain of Syllabification Rules in Spanishand Catalan." K. Hale and S. Keyser (Eds.), The View from Building 20: Essays in Linguistics inHonor of
Sylvain Bromberger. Cambridge, MA: MIT Press. 177-93.
Hualde, Jos? Ignacio. (1991). "Aspiration and Resyllabification in Chinato Spanish." Probus 3.1: 55-76.
Lipski, John. (1984). "Reducci?n de /s/ en el espa?ol de Honduras." Nueva Revista de Filolog?a Hisp?nica 32:
272-88.
?.(1986). "Instability and Reduction of /s/ in the Spanish of Honduras." Revista Canadiense de Estudios
Hisp?nicos 11.1: 21-M.
?.(1994). El espa?ol de Am?rica. Madrid: C?tedra.
?. (1999). "The Many Faces of Spanish /s/-weakening: (Re)alignment and Ambisyllabicity." Advances in
Hispanic Linguistics. Papers from the Second Hispanic Linguistics Symposium. Vol. I. J. Guti?rrez-Rexach
and F. Martinez-Gil (Eds.). Somerville, MA: Casacadilla P. 198-213.
L?pez Scott, Alma. (1983). A Sociolinguistic Analysis of Isi Variation in Honduran Spanish. Doctoral
Dissertation, U of Minnesota.
Mann, Virginia, and Sigfried D. Soli. (1991). "Perceptual Order and the Effect of Vocalic Context on Fricative
Perception." Perception and Psychophysics 49: 399-411.
Mason, Keith William. (1994). Comerse las eses: A Selective Bibliographic Survey of Isi Aspiration and Deletion
in Dialects of Spanish. Doctoral dissertation, U of Michigan.
M?ndez Dosuna, Juli?n. (1987). "La aspiraci?n de /s/ como proceso condicionado por el contacto de s?labas."
Revista espa?ola de ling??stica 17: 15-35.
?.(1996). "Can Weakening Processes Start in Initial Position?" Natural Phonology: The State of the Art. B.
Hurch and R. Rhodes (Eds.). New York: Mouton de Gruyter. 97-106.
Penny, Ralph. (2000). Variation and Change in Spanish. Cambridge, MA: Cambridge UP.
Rand, D., and D. Sankoff. (1990). GoldVarb. A Variable Rule Application for the Macintosh, Version 2.
Montreal: Centre de Recherches Math?matiques, Universit? de Montr?al.S?nchez, Rosaura. (1982). "Our Linguistic and Social Context." Spanish in the United States: Sociolinguistic
Aspects. J. Amastae and L. Elias-Olivares (Eds.). Cambridge: Cambridge UP. 9-46.
Silva-Corval?n, Carmen. (2001). Socioling??stica y pragm?tica del espa?ol. Washington, DC: Georgetown UP.
Terrell, Tracy D. (1977). "Constraints on the Aspiration and Deletion of Final Isi in Cuban and Puerto Rican
Spanish." The Bilingual Review 4: 35-51.
?.(1978). "La aspiraci?n y elisi?n de Is/ implosiva y final en el espa?ol de Puerto Rico." Nueva Revista de
Filolog?a Hisp?nica 27: 24-38.
?.(1979). "Final Isi in Cuban Spanish." Hispania 62: 599-612.