0% found this document useful (0 votes)

31 views62 pages

Phonological Insights in Colonial Mexico

The document discusses the phonological descriptions made by missionary linguists in New Spain, arguing that despite traditional views of their deficiencies, these grammarians made significant advancements in understanding and describing phonological systems. It highlights the contributions of various authors who identified unique sounds and phonological processes in indigenous languages, emphasizing the importance of native speakers in this work. The paper aims to reevaluate the achievements of these early linguistic efforts in light of contemporary standards, rather than focusing solely on their shortcomings.

Uploaded by

Ju Lt

We take content rights seriously. If you suspect this is your content, claim it here.

Available Formats

Download as PDF, TXT or read online on Scribd

0% found this document useful (0 votes)

31 views62 pages

Phonological Insights in Colonial Mexico

Uploaded by

Ju Lt

We take content rights seriously. If you suspect this is your content, claim it here.

Available Formats

Download as PDF, TXT or read online on Scribd

PHONOLOGICAL DESCRIPTION IN NEW SPAIN *

THOMAS C. SMITH-STARK
El Colegio de México

"no abasta escreuir ni pronunciar las tales diciones como

quiera [...] en no escreuir ni pronunciar los con las letras que
conuienen se diran otras cosas muy al reues delas que se
pensauan dezir. Y assi en lugar de dezir verdades, dezir se
an falsedades: con muy gran peligro de las animas."1
Maturini Gilberti (1558:109v-110r)

RESUMEN

Tradicionalmente se ha afirmado que las gramáticas coloniales de lenguas ame

ricanas eran especialmente deficientes en su tratamiento de la fonología. Sin em
bargo, al enfocarse en sus logros en lugar de sus fallas, y considerándolas en tér
minos de los niveles contemporáneos de conocimiento lingüístico en lugar de
nuestras expectativas modernas, se puede sostener que los lingüistas misioneros
avanzaron en la descripción fonológica de manera significativa. Con base en ma
terial de 20 autores que escribieron sobre una decena de lenguas de seis familias
lingüísticas novohispanas entre 1547 y 1767, se demuestra que identificaron nu
Copyright © 2005. John Benjamins Publishing Company. All rights reserved.

merosos tipos de sonidos novedosos no existentes en el español, que lograron

apreciar datos fonéticos en términos fonológicos, que reconocieron procesos fono
lógicos y morfofonológicos, que registraron datos fonotácticos, que utilizaron pa-

An oral version of this paper was originally presented at the First International Conference on
Missionary Linguistics, at the Voksenåsen Konferansesenter, under the auspices of the University
of Oslo. This written version incorporates additional material, including information kindly brought
to my attention by other participants at the conference. I am also grateful to Yolanda Lastra and to
Cristina Monzón for giving me their comments on a preliminary draft of this paper. This study
develops material that I have been gathering since 1998 for a projected guide to Indomexican
philology.
1
"it is not enough to write or pronounce such words in any way whatsoever [...] by not writing or
pronouncing them with the letters that are called for, one will say things that are just the opposite of
those that one intended to say. And so, instead of speaking the truth, one will speak falsehoods,
placing souls in very great danger."

Missionary Linguistics II / Lingüística misionera II : Orthography and Phonology. Selected papers from the Second International
þÿ Conference on Missionary Linguistics, São Paulo, 10 13 March 2004, edited by Otto Zwartjes, and Cristina Altman, John Benjamins Publishing Company, 2005.
Created from upenn-ebooks on 2018-04-04 [Link].
4 THOMAS C. SMITH-STARK

res mínimos para identificar oposiciones distintivas, y que justificaron sus análisis
con argumentación fonológica. El trabajo de los franciscanos Olmos, Gilberti,
Lagunas y Cárceres, y de los jesuítas Rincón y Carochi, en particular, resaltan por
su agudeza descriptiva. Es importante reconocer que los alcances de este período
fueron posibles en parte gracias a la participación de un elenco de nativohablantes
inteligentes, reflexivos, y experimentados en la manipulación de sus lenguas como
objetos de estudio.

1. Introduction
Until about fifteen years ago or so, studies of the history of linguistics generally
stressed the importance of philosophical and theoretical issues, such as, for
example, the ancient Greek controversy between anomalists and analogists about
the nature of the linguistic sign; interest in the modes of signifying and universal
grammar among Medieval grammarians; and the debates between empiricists and
rationalists during the Renaissance. They paid relatively little attention to des
criptive linguistics per se, though they might refer to the development of certain
basic grammatical notions among the Greek and Latin grammarians or the
development of new writing systems for a variety of languages. Robins (1969) is a
good example of this trend. Though by no means ignoring descriptive linguistics,
he observes that "By far the most interesting and significant development in
linguistics during the Middle Ages is the output of 'speculative grammars'"
(Robins 1969:74). The relative unimportance of empirical practice in our histories
of linguistics reflects to some degree, perhaps, the low esteem in which descriptive
linguistics has been and continues to be held with respect to theoretical linguistics.
However, the art of writing grammars and dictionaries - what I take to be the
principal tasks of descriptive linguistics -, to be done well, requires a good deal of
Copyright © 2005. John Benjamins Publishing Company. All rights reserved.

thought about the nature of language structure and presupposes a tradition of such
thought which will have provided an author with a conceptual framework within
which to operate. The history of such writing, then, merits serious study for what it
can reveal about the evolution of linguistic thinking as reflected in descriptive
practice.
This apparently self-evident observation has not in fact lead to such study.
Rather, certain preconceptions about pre-modern descriptive work have resulted in
its neglect among contemporary linguists, at least among those in the Western
tradition. In particular, older descriptions are generally conceived of as being
overly dependent upon traditional descriptions of Greek and Latin, casting other
languages into categories which are not the ones in terms of which they are in fact

organized, and ignoring those that are, but which are not found in the classical
languages of occidental culture.2 As Jorge Suárez (1983:2) observed, speaking
about the missionary period in Mesoamerica, "The grammars written within this
period are generally valued only to the extent that they offer raw materials on the
respective languages, and are usually lumped together and dismissed as merely
latinized descriptions." A good, and influential, example of such a point of view is
that expressed by Boas (1917:6; 1940:206). Speaking about the linguistic catego
ries found in American languages as opposed to those found in European lang
uages, he observes that:
No less interesting are the categories of thought that find expression in grammatical form.
The older grammars, although many of them contain excellent material, do not clearly
present these points of difference, because they are modelled strictly on the Latin scheme,
which obscures the characteristic psychological categories of Indian languages.
Likewise, Stanley Newman, referring to the early grammar of the Meso-
american language Nahuatl, published by the Franciscan friar Alonso de Molina
(C.1513-C.1579) in 1571, observes that "When he [Molina] and other gram
marians discussed the Nahuatl participles3 in -dus (i. e., Nahuatl forms functioning
somewhat like Latin amandus, etc.), they were following the humanistic precept of
measuring the Indian vernacular against the Latin standard" (Newman 1967:180).
An added factor which has inhibited work on pre-modern grammatical
traditions is the difficulty that many modern linguists have when faced with the
philological problems which the adequate interpretation of the older materials
entails, even when they have worked with the corresponding modern languages.
The language used for writing the descriptions is, at the very least, different from
the modern variety and may even be a language unknown to the modern linguist.
Copyright © 2005. John Benjamins Publishing Company. All rights reserved.

The language being described will have changed, academic practices are different,
typographic styles and conventions are not the same, a different writing system
may have been in use. In other words, one does not just pick up such studies and
use them without investing some, and perhaps a lot of, time and effort in the
enterprise.
Despite such difficulties, there is nonetheless a wealth of material to be studied.
One of the primary motives for the production of such descriptive work was, and

2
Zimmermann (2004:15) observes that already in Humboldt's work, one finds deprecatory
judgements about the descriptive studies of the missionary grammarians.
3
The original text reads principles, an obvious misprint which I have corrected as participles.

of course continues to be, religious proselytization. Missionaries driven by the

desire to convert other peoples to their own systems of belief have been among
those most motivated to study large numbers of other languages in order to
communicate their message, express their arguments, translate their scripture
and/or doctrinal texts, aquire converts, and attend to their congregations.
Missionary activity does not, of course, necessarily require descriptive linguistics.4
The first millenium and a half of Christian expansion, for example, produced
translations, liturgies in diverse languages, the invention of original writing
systems for languages such as Coptic, Syriac, Armenian, Georgian and Old
Church Slavonic, but to my knowledge little in the way of grammars and
dictionaries. However, things changed drastically with the political expansion of
Europe begun in the 15th century into Africa, Asia and America together with the
invention of the printing press by Gutenburg around 1450. For the first time,
grammars of languages which were not the native language of the writer or the
vehicle of texts considered to be sacred or part of an important cultural canon
began to be produced in significant numbers. In his useful compilation of such
works, Rowe (1974) lists 21 languages for which grammars were first published in
the 16th century and 41 more with grammars first published in the 17th century.
Of these grammars, 25 were written in Latin, 21 in Spanish, 6 in Portuguese, 3 in
English, 2 in Dutch and one each in French, White Russian, Italian, Czech and
Swedish, an impressive indication of the explosive growth of language description
in the Europe of this period and of the leading role played by writers in Iberian
languages.
One of the important loci for this activity was in fact New Spain, the Spanish
viceroyalty governed from Mexico City and covering much of present day
Copyright © 2005. John Benjamins Publishing Company. All rights reserved.

Mexico.5 During the period from the conquest of Mexico in 1521 to 1645, at least
20 grammatical descriptions produced in New Spain have survived (Smith Stark
2000b) and by 1611 some 22 dictionaries had been compiled (Smith Stark 1992a).
In this paper, I would like to examine and evaluate the phonological work found in

4
Ostler (2004) notes that, of the three major religions which have actively sought converts,
Buddhism and Islam have not been accompanied by descriptive linguistic work, and that in the case
of Christianity, the association is only found among Roman Catholics and Protestants with the
advent of European political expansion.
5
New Spain, in its widest sense, included the districts of the five audiencias of the West Indies of
New Spain: Mexico, Guadalajara, Guatemala, Santo Domingo and the Philippines (León-Portilla
1995,111:2478). Here I use the term New Spain to refer only to the first three.

such descriptions.6 It is generally thought that the Spanish grammarians were quite
poor phonologists. Newman (1967:180), for example, in the grammatical sketch of
classical Nahuatl referred to earlier, considers that:
From the point of view of modern linguistics, the early grammars of Nahuatl were most
deficient in their treatment of phonology. Sounds were conceived in terms of orthography.
Nearly all discussions of Nahuatl sounds began with a list of the letters missing in the
language of the Aztecs, and careful attention was paid to the spelling rules derived from
Spanish orthography.
Suarez, speaking in general terms about grammatical description in New Spain,
considers that:
As is to be expected for that period, the description and recognition of contrasts are weakest
for sounds; in general the friars coped rather well with simple systems like that of the
Nahuatl [...], and even with those of moderate complexity in the consonants. However, rich
vowel systems such as those found in Otomian and Chinantecan languages [...] were
markedly underdifferentiated, and tones were at most dimly recognized but in no case
marked in the orthography. (Suárez 1983:4)
In contradistinction to such opinions, I will show in what follows that the
missionary grammars made significant progress in their phonological descriptions
of the languages of New Spain. However, it is important to bear in mind that it is
possible to take two contrasting positions when evaluating their work. One can
emphasize the errors as seen from a modern perspective, or one can underscore the
achievements found in the most accomplished work as evaluated in terms of
contemporary standards. I will take this second tack.

2. The linguistic context: Nebrija

In order to evaluate phonological work in New Spain, let's first consider the
kinds of knowledge and attitudes about phonology which the friars had at their
disposal. For this purpose, I will take the Spanish grammarian Antonio de Nebrija
(1444-1522) as my standard of comparison.7 Nebrija was deeply immersed in the

6
For an overview of Spanish missionary linguistics in America, the reader is referred to Suárez
Roca's excellent study (1992), which includes such a fine discussion of phonological aspects of
missionary grammars that it was at times difficult for me to avoid simply repeating observations
already made by this scholar.
7
See Esparza Torres and Niederehe (1999) for an excellent bibliography of Nebrija's work. For my
understanding of Nebrija's phonology, I am especially indebted to Quilis' introductory study in his
edition of Nebrija's Gramática castellana (Quilis 1984).

Latin grammatical tradition and formed part of the resurgence, emanating out of
Italy, of humanistic interest in recuperating an accurate knowledge of classical
Latin. He was an influential teacher and writer, whose Latin studies were fre
quently mentioned by New Spain grammarians and lexicographers. He also wrote
a Spanish grammar, famous for being the first published description of a European
vernacular language, though probably not as influential as his work on Latin
among New Spain grammarians.8
Nebrija had a clear idea of the difference between sound and letter. Consider
his description of the relationship between the two:
el primer inuentor de letras quien quiera que fue: miró quantas eran todas las diversidades
delas bozes en su lengua: & tantas figuras de letras hizo: por las cuales puestas en cierta
orden represento las palabras que quiso: de manera que no es otra cosa la letra: sino figura
por la cual se representa la boz. ni la boz es otra cosa sino el aire que respiramos espessado
enlos pulmones: & herido despues enel aspera arteria, que llaman gargauero: & de alli
començado a determinarse por la campanilla lengua paladar dientes & beços. Assí que las
letras representan las bozes. y las bozes significan como dize aristoteles los pensamientos
que tenemos enel anima. (1492:6v)9

[the first inventor of letters, whoever he was, looked at the number of different sounds
found in his language and made that same number of different letters, by means of which,
placed in a certain order, he represented the words that he wanted, so that the letter is
simply the figure by which the sound is represented, and the sound is simply the air which
we breathe, thickened in the lungs and then modified in the harsh artery that they call the
throat, and from there they begin to be determined by the uvula, tongue, palate, teeth and
lips. Thus, the letters represent the sounds, and the sounds signify, as Aristotle says, the
thoughts that we have in our soul.]10
Copyright © 2005. John Benjamins Publishing Company. All rights reserved.

As partially indicated by this quote, Nebrija distinguishes clearly between the

letter (letra), figure of the letter (figura de la letra) or stroke (traço) on the one
hand, and the sound (boz), pronunciation (pronunciacion) or force (juerga) of the
letter on the other.

8
Nebrija is often mentioned in descriptive studies written in New Spain, but it is usually not clear
which of his works is being referred to. Some linguists (for example, McQuown 1967:3) believe
that such references were usually to his Spanish grammar. Others (such as myself) have assumed
that they were to his Latin grammar. But it now seems to me that both were probably influential,
though this question merits further study.
9
My transcription of Nebrija (1492) is based on the facsimile published in 1946, in conjunction
with the edition of that same year and that of 1984.
10
This and all other translations are my own, unless otherwise indicated.

Nebrija was also fully aware of the diversity of sounds in the languages of the
world. He advises that:
algunas lenguas tienen ciertas bozes: que los ombres de otra nacion, ni aun por tormento no
pueden pronunciar [...] que de tal manera son proprias de cada lengua: que por ningun
trabajo ni diligencia ombre de otra nacion las puede espressa mente proferir: si desde la
tierna edad no se acostumbra a las pronunciar. (1492:6v, 7r)

[some languages have certain sounds that the men of another nation, not even by way of
torture, can pronounce [...] so that they are proper to each language and there is no way that
a man from another nation can expressly produce them if he is not accustomed to
pronouncing them from a tender age.]
For Nebrija, the sounds of a language are its minimal units of expression: "The
letter is the smallest part of the sound that can be written [...] if I say señor 'sir',
this word can be divided into two syllables, which are se and ñor, and then se into
s and e; and the s cannot be further divided" ("Que la letra es la menor parte de la
boz que se puede escriuir [...] porque si yo digo «señor», esta boz se parte en dos
síllabas, que son: «se» y «ñor»; y el «se», después, en «s» y «e»; y la «s» ya no se
puede partir", 1977:117[1517:2v])".
Nebrija recognized the distinction between vowels and consonants. Vowels
"have sound on their own, without being mixed with other letters" ("por si mesmas
tienen boz sin se mezclar con otras letras", 1492:7v). They "sound on their own
without being modified by any of the instruments which are used to form the
consonants, but rather just passing the breath through the narrowness of the throat,
forming the different ones by means of the shape of the mouth" ("suenan por si no
hiriendo alguno delos instrumentos con que se forman las consonantes: mas sola
Copyright © 2005. John Benjamins Publishing Company. All rights reserved.

mente colando el espíritu por lo angosto dela garganta: & formando la diversidad
dellas enla figura dela boca", 1492:8r).11 Consonants, on the other hand, "cannot
sound without modifying the vowels" ("no pueden sonar sin herir las vocales",
1492:7v). "For that reason it is called a consonant, because it sounds with another
vowel letter, as the b cannot sound without the aid of the e" ("por esso se llama
consonante, porque suena con otra letra vocal, como la «b», no puede sonar sin la
ayuda de la «e»", 1517:2v[1977:118]).
The Latin consonants are divided into two groups, mutes (mudas) - which
include b, c, ch, d, f, g, p, ph, t, th, i u - and semivowels (semivocales) - which

11
Robins (1969:118-119) cites a quite similar passage from W. Holder in 1669 as evidence of "The
excellence of his phonetic theory and the conciseness of his expression."

include l, m, n, r, s, z, and perhaps consonantal i -. This distinction seems to be

based on sonority and degree of constriction, although it does not correspond to
any that we would currently recognize as valid. The mutes, which come close to
corresponding to stops, are divided into three types: those free from aspiration and
more tightly articulated (c, t, p); those pronounced in an intermediate way, more
loosely than the mutes, but more subtly than the third group (g, d, b);12 and those
which are thicker, looser, articulated less tightly (ch, th, ph) (cf. 1492:8r).
Evidently Nebrija considered the third series to be somewhat like fricatives, rather
than true aspirated sounds. For Nebrija, Latin h was not a letter, but rather a sign
of breath (1492:7r), though in Spanish he suggested that it could be considered a
letter, as it is in Arabic and Hebrew (1492:9v).13
Nebrija (1492:8r) clearly distinguished Latin sounds in terms of five points of
articulation: the uvula (c, ch, g), the teeth (t, th, d), the lips (p, ph, b, m), the upper
teeth with the lower lip (f and consonantal v) and the palate (/, n, r, s, z). In
addition, for Spanish (1492:9v), he recognizes a sixth point of articulation, the
throat (h).
The syllable is described as having three properties: number of letters, length in
time, and height of accent. The number of letters simply refers to the number of
segments in a syllable. Nebrija distinguishes short and long syllables, though he
observes that the distinction does not apply to Spanish. He distinguishes five kinds
of accent: two on simple vowels (acute and grave), two on diphthongs (deflex and
inflex), and one on triphthongs (circumflex). Each syllable has its own accent, but
the number of possible combinations is tightly constrained: "any word whatsoever,
not only in our language, but in any other as well, has a high syllable, that
dominates the rest, and which we pronounce with an acute accent, and all the
Copyright © 2005. John Benjamins Publishing Company. All rights reserved.

others are pronounced with a grave accent." ("cualquiera palabra no sola mente en
nuestra lengua mas en cualquiera otra que sea: tiene una silaba alta: que se
enseñorea sobre las otras: la cual pronunciamos por acento agudo: & que todas las
otras se pronuncian por acento grave", 1492:18r). That is, Nebrija identifies the
acute accent and high pitch with the stressed syllable. Although his treatment of

12
The voiced-voiceless distinction is not described explicitly. Robins (1969:118, 142) observes that
William Holder (1616-1698) was the first Western linguist to correctly describe this phonetic
feature in 1669, though it was well understood by ancient Sanskrit grammarians.
13
The curious distinction between h as a sign of breath versus a true letter, repeated by various
authors in New Spain, evidently is due to the fact that [h] is represented by a diacritic, the spiritus
asper ('), in Greek orthography, rather than by an independent letter.

stress and pitch together is perhaps adequate for Spanish, it is certainly not
adequate for all languages as Nebrija seems to claim.
Nebrija is quite clear about how a language should be written: "we must
presuppose what everyone who has written about orthography presupposes: that
we should write like we pronounce, and pronounce like we write; otherwise letters
would have been discovered in vain." ("avernos aqui de presuponer: lo que todos
los que escriven de orthographia presuponen: que assi tenemos de escrivir como
pronunciamos: & pronunciar como escrivimos: por que en otra manera en vano
fueron halladas las letras." 1492:8v). The ideal relationship between letter and
sound should be one-to-one. That is, each sound should be represented by a single
letter and each letter should represent a single sound. Nebrija fully endorses
Quintilian's recommendation for setting a language to writing, which he explains
as follows and which I refer to as Quintilian's method: "he who would reduce to
art some language should first know if some of the letters in use are superfluous,
and, on the contrary, if some are lacking." ("el que quiere reduzir en artificio algun
lenguaje: primero es menester que sepa: si de aquellas letras que estan enel uso:
sobran algunas: & si por el contrario faltan otras", 1492:7r).14
One last point about Nebrija's practice seems to me to be worth mentioning.
When discussing Spanish orthography, he criticizes having a single grapheme with
more than one interpretation. In particular, he wants to avoid using <ch> for both
/c/, as in muchedumbre "crowd", and for /k/, as in monarchia "monarchy", and
using <x> for both /s/, as in xabon "soap", and for /ks/, as in Latin (cf. Nebrija
1492, libro 1, cap. 6, ff. 10v-llv). In each case he advocates a new letter for the
innovative values, not found in Latin - <ch > for Icl and <x> for /š/- though his
actual practice (or that of his editors) is inconsistent in this regard.15 He even seems
Copyright © 2005. John Benjamins Publishing Company. All rights reserved.

14
Nebrija refers to a passage in book 1 of Quintilian's Institutio oratoria (chapter iv, paragraphs 7
and 9 in Butler's edition), in which the author recommends that teachers of literature should discuss
which letters are absent from the Latin alphabet but should be there -he illustrates with the
digamma, for instance- and which are superfluous -he mentions h, k, q, and x as possible
candidates-: "desintne aliquae nobis necessariae litterarum [...] an rursus aliae redundent"
("whether certain necessary letters are absent from the alphabet [...] Again there is the question
whether certain letters are not superfluous", Butler 1920:65-66). As far as I know, the extension of
this pedagogical principal for the teaching of Latin to the setting down in writing of new languages
is of Nebrija's own concoction.
15
I say inconsistent because there are clear cases of <ch> used for /č/ instead of <ch> (for example,
f. 7v 1. 23 <ocho>) and of <x> for /š/ instead of <x> (for example, f. 7v, 1. 27 <diximos>).
Moreover, in his Spanish-Latin dictionary (1495), there are cases where /k/ is written as <ch>, as in

to want to avoid using the same grapheme with different values in different
languages. For example, he says that "if we were to leave cħ without a mark, we
would commit the error of pronouncing the same letters differently in Spanish and
in Latin" ("si dexassemos la .cħ. sin señal: verniamos | en aquel error: que con
unas mesmas letras pronunciaríamos diversas cosas enel castellano & enel latin",
1492:10v-llr). However, he does not apply this principle systematically.
Now, with this important antecedent in mind, let's look at how sounds were
dealt with in the languages of New Spain, especially, though not exclusively,
during the first century or so of grammar writing (1547-1645). Specifically, I will
examine seven aspects of phonological description: the graphic representation of
sounds (§3), the articulatory description of sounds (§4), the concept of a sound
system (§5), the relation between graphemes and phonemes (§6), phonotactics
(§7), argumentation (§8), and phonological processes (§9). I will then conclude
with a consideration of the role of native speakers in the descriptions of their
languages (§10).

3. The graphic representation of sounds in the languages of New Spain

Perhaps the most direct evidence for the phonological capabilities of the friars
in New Spain comes from their practice in writing the different languages they
tried to describe. In general I would say that there was an implicit recognition of
the variety of sounds in the languages of the world. That is, there is no evidence
that they expected all languages to be expressed by the sounds of their own tongue
and, in at least one case, there is explicit evidence to the contrary. The Augustinian
friar Diego Basalenque (1577-1651), upon noting that Matlatzinca has no ƒ,
observes that:
Copyright © 2005. John Benjamins Publishing Company. All rights reserved.

no hay que levantar misterios en la falta de ésta y de otras letras, sino reducirlo a cosa
natural de que no todas las lenguas se conforman en una misma pronunciaciòn y asì
tampoco tienen unos mismos caracteres. La griega tiene màs que la latina y otras menos.
(1975[1642]:7)16

the entry for <monarchia>.

16
This passage is also referred to by Suárez Roca (1992:211). Evidently Basalenque disapproves of
the mnemonic device used by Lagunas (1574:1) to relate the absence of the letters B, D, F, G, I, L
and R to the absence of Baptism, God (Dios), Faith, Grace, Justice, Law and Government
(Regimiento) /Rule/Reason among the pre-Christian Purepechas. Altman (2003) notes the use of
this same rhetorical technique by Sousa (1938[1587]:364-365) in Brazil.

[there is no reason to create mysteries about the lack of this or other letters, but rather it
should be treated as something natural since not all languages are alike in having the same
pronunciation and so neither do they have the same characters. Greek has more than Latin
and others fewer.]
When the sounds of a new language were quite distinct from those of Spanish,
the friars had problems, but in some cases - though, of course, not all - they
achieved an orthography which was in principle adequate, though in practice not
necessarily applied consistently. For many, the correct pronunciation of the words
of a language was something to be taken quite seriously. For example, the
Franciscan Maturino Gilberti (c. 1507-1585) quotes Saint Jerome in this regard.
La orthographia y recta pronunciacion, siempre ha sido, y es muy necessaria en qualquier
lengua. La qual segun dize Sant Hieronymo en vna epistola, en tanto excede alaque no esta
assi pronunciada ni orthographiada con deuidas letras, como difiere el hombre biuo, al
pintado o muerto, por tanto como ya tengo dicho enlos auisos, despues del prologo: no se
acertaria en dezir que esta lengua se puede bien pronunciar o escreuir, con qualesquieras
letras, que poco mas o menos es intelligible, y que en escreuir esta lengua con tantas
aspiraciones, y con letras no acostumbradas, es muy gran | estonio para daprender la lengua,
todas estas y semejantes razones carescen de esperiencia enla orthographia dela lengua
segun alcançan los que bien la entienden (Gilberti 1558:109r-v)

[Orthography and correct pronunciation have always been, and continue to be, very neces
sary in any language, and, according to Saint Jerome in an epistle, exceed that which is not
so pronounced or spelled with fitting letters, as a living man differs from one that is painted
or dead, for which reason, as I have already said in the preliminary notices, which follow
the prologue: it would not be correct to say that in this language one can write and pro
nounce with any letter whatsoever, which is more or less intelligible, and that writing this
language with so many aspirations, and with unaccustomed letters is a great bother for
Copyright © 2005. John Benjamins Publishing Company. All rights reserved.

learning the language. These and all such reasons show a lack of experience in the ortho
graphy of the language according to what those who do understand it well have achieved]
In general the designers of new orthographies used letters based on their values
in Spanish or, rarely, in Latin. For new sounds, they employed five strategies:
a. they used familiar letters with novel values;
b. they created special combinations of letters;
c. they used modified forms of conventional letters;
d. they invented completely new letters; and
e. they borrowed letters from another non-Latin-based writing system.17

My thanks to Toru Maruyama for bringing this fifth possibility to my attention during the

In this way, they were able to represent most of the unfamiliar sounds
encountered in the languages of New Spain. Among the consonantal phenomena
represented successfully are aspiration, ejecrives, a preglottalized stop, the glottal
stop, an alveolar affricate, a lateral affricate, a lateral fricative, a uvular stop, and a
retroflex rhotic. Among the vocalic phenomena registered we find nasalized
vowels, non-open central vowels, and different degrees of vowel height among
mid vowels. Among the prosodic phenomena recognized were stress, vowel length
and, in incipient form, tone.

3.1 Consonants
3.1.1 Aspiration. The letter <h> was used in 16th century Spanish to represent a
glottal fricative /h/ which resulted from the weakening of an earlier /f/ and which
was later lost altogether: for example, fazer /fadzer/ > hazer /hazer/ > hacer /aser/
"to do, to make". This letter was thus a natural choice for representing sounds like
[h] and aspiration, as occurred in Otomi and Purepecha.18 For example, in the
Otomi vocabulary attributed to the Franciscan friar Alonso Urbano (c. 1529-
1608)19 (his vocabulary dates from 1605), who continued the practice of the
Franciscan Pedro de Cárceres (fl. 1580) in this regard, we find the following
graphemes (cf. Lastra 1992, Smith Stark 1992b).

ph = /ph/ th = /th/ tzh = /ch/ ch/qhu = /kh/ h = /h/

çh = /sh/
hm = /hm/ hn = /hn/

That is, an <h> represents a glottal fricative, the aspiration of a preceding

obstruent, or the preaspiration of a following nasal.

Similarly, <h> was used in Purepecha for a glottal fricative and, somewhat
irregularly, for the aspiration of stops. For example, Monzón (2001) reports that in
the grammar by the Franciscan friar Juan Baptista de Lagunas (c. 1530-1604?),
aspiration was written after the corresponding stop in initial position, but before

conference in Oslo, based on his own experience with systems of romanization for Japanese. If the
tresillo is in fact modeled on cayn (ξ), then it would illustrate this fifth strategy (see §3.1.2).
18
In the case of Purepecha, the aspirated consonants were only partially represented and in a much
less consistent and symmetrical way than in Otomi (cf. Monzón 2001).
19
Acuña (1990:lxx) is of the opinion, for not very convincing reasons, that Urbano is not the author
of the Vocabulario trilingüe, though his annotations may have influenced it.

the stop intervocalically, which corresponds to the phonetic realization of

aspiration in that language:
ph = [ph] th = [th] thz = ths = [t h]20 ch = [čh] k,kh,qhu = [kh] ku,qhu = [kwh] h = [h]
hp = [hp] ht = [ht] htz = [ht ] hts = [ht ] hch = [hč] hc/hqu = [hk] hcu,hqu = [hkw]

Some languages had more than one h-like sound. For example, in the Motul
dictionary of Yucatec Maya, probably written by the Franciscan Antonio de
Ciudad Real (c. 15457-1617) towards the end of the 16th century, there are two
sections devoted to <h>. One is for strong <h> (H. rezia, Ciudad Real 1984[16th
century] ff.l70r-202v) and the other is for weak <h> (H. simple, ff. 202v-209v).
That is, there is clear recognition of two h-like sounds, but they are not
distinguished in the orthography. However, at least one author does make an
explicit distinction; Gabriel de San Buenaventura (1996[1684]), also a Franciscan
(c.l620?-post 1695), uses <h> for the weak h and barred <h> for the strong ft,
which was probably a velar or uvular fricative.21

3.1.2 Glottalization. An important category of new sounds encountered by the

missionaries were those accompanied by glottal closure or stricture. The glottal
stop itself, [?], which is very frequent in American languages, was quite pro
blematic. For example, it is an important sound in Nahuatl, where it was simply
ignored, or only represented sporadically as an <h>, until about 1600.22 Even
today, many students of Nahuatl fail to represent this sound adequately. In 1595,
the Jesuit priest Antonio del Rincón (1555-1601) proposed the use of a diacritic
similar to a breve accent, <ˇ>, over the vowel which precedes this sound to
Copyright © 2005. John Benjamins Publishing Company. All rights reserved.

represent it, and referred to it as the saltillo, the little jump, because it seemed to
be produced by a little jump in the voice (cf. Smith Stark 1996). However,

20
The IPA symbol for a retroflex s stands here for an apico-alveolar s.
21
Such is Fox's interpretation (1978:45). The distinction is no longer made in the modern language,
but it can be verified etymologically and can be argued for on morphophonological grounds (Orie
& Bricker 2000). Of course, <j> was not considered a possible candidate for representing this
sound since it still represented /š/, /ž/, or /i/ in late 16th century Spanish, not [x].
22
Cf. Smith Stark (1995). The use of <h> was probably motivated by the fact that the glottal stop
was pronounced as [h] in many forms of Nahuatl, a value which <h> also had, at times, in 16th
century Spanish. However, <h> was also a reasonable candidate for representing a real glottal stop
since it could represent a hiatus in 16th century Spanish as well, which, after /h/ > 0 in the 17th
century, became virtually its only possible value in modern Spanish.

Rincón's proposal was not widely followed. After 1600, a grave accent was often
used for the saltillo (Smith Stark 1995). Thus, a word such as /λa?toa:ni/
"governor", can be found written in any of the following ways (and many others as
well): tlatoani, tlahtoani, tlãtoani, tlàtoani, tlàtoāni.
In Mixtec, the glottal stop has been represented traditionally by an <h>, a usage
which is still employed in some modern practical orthographies.
doho "adobe" (Alvarado 1593)
ndoho "adobe" (Stark Campbell et al, 1986)
This solution was possible since the <h> in Spanish orthography which was
derived from Latin <h> was not pronounced in Spanish and probably was
unpronounced in the Latin of Spanish speakers. Furthermore, after the Spanish
<h> mentioned above which was derived from prior /f/ lost its value as a fricative
towards the end of the 16th century, it became an even more natural choice for
representing some sort of hiatus, which is how Spanish speakers often perceive the
glottal stop.
In the Mayan family, the glottal stop never received adequate representation
during the colony. Sometimes it was indicated by writing a double vowel, but not
always. The difficulty which the missionaries had with this sound probably
demonstrates the little influence that a solid knowledge of Arabic or Hebrew,
languages in which the glottal stop is an important element, had on their linguistic
preparation or thinking.
In sharp contrast to the problems occasioned by the glottal stop, the
missionaries were rather successful at identifying and representing glottalized
consonants, which are found in many Otomanguean languages, where they are
Copyright © 2005. John Benjamins Publishing Company. All rights reserved.

usually analyzed as consonant clusters of a plain consonant plus a glottal stop, and
in Mayan languages, where they form unitary phonemes. The solutions to the
problem of how to represent these segments were both diverse and creative,
illustrating all five of the strategies mentioned above for representing new sounds.
In Otomi as written by the Franciscan friar Alonso Urbano (1605), consonants
were doubled to indicate glottalization.24
23
The Rinconian saltillo is, however, found in a few manuscripts, such as Cantares mexicanos,
where Bierhorst (1985:692-694; 702-704) interprets its use as evidence for a Jesuit presence in the
transmission of that important document.
24
Cárceres (c. 1580) used a similar system, but with <k> representing /k'/, rather than <cc/cqu>.
Notice that physical doubling of the non-glottalized grapheme is not really a complete description
of this convention.

pp = [p'] tt=[t'] ttz=[ts'] cc/cqu = [k']

In Yucatec Maya, ad hoc solutions were used for each point of articulation.25
pp/ = [p']26 th/th = [f ] /dz = [ts'] ch = [c'] k = [k']

Doubling, doubling of the vertical stroke, barring or doubling together with

barring was used for the bilabial /p7; a <t> with an added <h> or barred <h> was
used for /tV; a special symbol, , resembling a backwards <c>, or a special
digraph, <dz>,27 was used for the alveolar affricate / /; a barred <ch> was used for
/c7; and a special value was ascribed to <k> for a /k7.
For writing the Mayan languages of Guatemala, a series of special letters
known as the Parra letters, named after the 16th century Franciscan friar Francisco
de la Parra (cl500-1560),28 their inventor (Carmack 1973:22, note 4), was used to
represent the glottalized consonants. One, known as the cuatrillo, the little four, is
shaped like the number 4. Another, known as the tresillo, the little three, is shaped
like a backwards number 3 or like Arabic cayn
tt/th = [t'] 4, = [ts'] 4h=[c'] 4=[k'] = [q']

In addition to the glottalized ejectives, Otomi had a preglottalized /?d/. This

sound was distinguished by both Cárceres and Urbano, who used a barred <d> to
represent it (Smith Stark 1992b).
Copyright © 2005. John Benjamins Publishing Company. All rights reserved.

25
Cf. Tozzer (1967[1921]:21) and Pellicer Ugalde (1997:46).
26
The barred p's should have the bar crossing the lower part of the setter, which we have been
unable to reproduce. A fourth grapheme used to represent this sound, but which we have been
unable to reproduce typographically, is a with the vertical stroke doubled.
27
Neither Tozzer (1921) nor Pellicer (1997) mentions the use of <dz> for but it is common in
place names and family names in Yucatan, such as Dzibilchaltun (an archaeological site in northern
Yucatan) or Dzib (a Mayan family name).
28
Parra was a Galician who was in Guatemala by 1542. His invention of several special letters was
considered an inspiration from God and was followed in all subsequent Franciscan writings on
languages such as Quiché, Cakchiquel, Tzutujil, Pocomchí, Pocomam and Choltí (Vázquez
1937[1714], vol. 1:126-127 and Sánchez García 1919 apud Herbruger and Díaz Barrios 1956:34).

3.1.3 The alveolar affricate. The alveolar affricate, [ts], is very common in
Mexican languages. It was usually written as <tz>, though there are early
examples in Nahuatl documents where it was represented as <t> or <tç>.29

3.1.4 The lateral affricate. The lateral affricate [X] was a particularly prominent
sound in Nahuatl (Smith Stark 1995) and was also found in Totonac (Romero
1990[lst third of 17th century]).30 In both cases it was written as <tl> and caused
little trouble.

3.1.5 The lateral fricative. Totonac also has a phonemic lateral fricative, [ł]. The
grammar attributed to Eugenio Romero (fl. 1st third of 17th century) (1990)
represents this sound as <hl>. A similar sound, [1], occurs in Nahuatl as an
allophone of HI in syllable final position. Andrés de Olmos (c. 1485-1568)
proposes recording this sound as <lh> (1972[1547]:200).

3.1.6 The uvular stop. Several Mayan languages in Guatemala have a simple
uvular stop, [q], which was generally written as a <k>. See, for example, Carmack
and Mondloch (1983:33) for Quiché and Pantaleón de Guzmán (c.1652-1708) for
Cakchiquel (1704). In Totonac in Mexico, this sound was written as <g> (Romero
1990[lst third of 17th century]).

2
Cf. Anderson, Berdan and Lockhart (1976:34) and Téllez Nieto (2004:207-208). Interestingly,
there is no unequivocal evidence in the linguistic work of the 16th century in New Spain for the
retention of an affricated pronunciation of <ç> or <z> in Spanish. As Canfield (1934) clearly
Copyright © 2005. John Benjamins Publishing Company. All rights reserved.

argued, if these letters had still been pronounced as something like [ts] in 16th century Mexico, as
many hispanists believe, one would expect them to be used in the representation of such affricates
in newly romanized Mexican languages. However, such was not the case. Monzón (2005) observes
that Molina (1571b, "Aviso segundo"), when discussing alphabetical order, says that he is going to
"also put tza, tze, tzi, tzo, tzu with the letter T [since it is written with it] even though it is
pronounced like ça, ce, ci, go, çu" ("poner tambien el. tza. tze. tzi. tzo. tzu. enla letra. T. [pues se
escriue con ella] aunque se pronuncie a manera de. ça. ce. ci. ço. çu." ). She takes this as evidence
that Molina pronounced <ç> as an affricate in Spanish or that he was at least familiar with such a
pronunciation. However, it is also possible to understand him to be saying that the pronunciation is
similar to that of <ç>, but not necessarily identical. Impressionistically, / / was compared to /s/, but
distinguished as being stronger or more forceful (cf. Carochi's description of / / in §3.1, Gilberti's
in §3.2.1, and Cárceres' in §3.2.7).
McQuown's useful edition of this text includes his reconstruction of the phonological
representation of the Totonac forms found in this grammar.

3.1.7 The retroflex rhotic. Purepecha has two rhotics. One is a simple flap and is
generally represented as <r>. The other is usually described as being a more
retroflex [ ]. The Franciscan friar Maturini Gilberti (1559) does not distinguish
them, writing them both as <r>,31 but his fellow Franciscan Juan Baptista de
Lagunas (1574:72-73) proposed using <rh> for the second, a grapheme which is
still utilized to this day in most practical orthographies for the language, as, for
example, in Monzón (1997) and Medina Pérez and Alveano Hernández (2000).

3.1.8 Interdentals. In Mixtec, Reyes (1976[1593]) and Alvarado (1962[1593]) use

the digraph dz to represent [ð]. In Huastec, Tapia Zenteno (pre 1690-post 1769)
(1985[1767]) uses a z to represent [0].

3.2 Vowels
The main oppositions found in the vowel systems of Mexican languages, in
addition to those of vowel quality, are nasalization and laryngealization. Laryn-
gealization, which is found in Totonacan and Zapotec languages, was not resolved
adequately. Nasalization was registered well in at least one case, Otomi. Vowel
qualities unlike those of Spanish were also problematical, as Suárez observed in
the passage quoted above (§1), but Otomi was a notable exception in this regard as
well.

3.2.1 Nasalization. Many Mexican languages, above all those in the Otomanguean
family, distinguish between oral and nasalized vowels. In the case of Otomi, as
registered by the Franciscan friars Pedro de Cárceres (fl. 1580) and Alonso
Urbano, nasalized vowels were distinguished by writing a small omega-shaped
Copyright © 2005. John Benjamins Publishing Company. All rights reserved.

diacritic above them, < >, which, following a suggestion by Heriberto Avelino
(personal communcation), might be called a little bat, or 'murcielaguito' in
Spanish.
In striking contrast to the success of the Franciscans with Otomi, the Domi
nicans in Oaxaca, friars Antonio de los Reyes (c.1538-1603) and Francisco de
Alvarado (c. 1553-1603), did not register the nasalized vowels of Mixtee in a
systematic way. In fact, they were only registered in a couple of special
circumstances. A /t/ followed by a nasalized vowel was written as <tn>, a

31
Cristina Monzón informs me that Gilberti uses <hr> on one or two occasions to represent this
sound.

representation of the allophone [tn] found in this context. A /k/ followed by a

nasalized or was written using the letter q alone to represent the syllable
or , as in duq "cochineal", dzuq "neck", and yeq "squash" (cf. the entries grana,
cuello, and calabaça in Alvarado 1593). The corresponding modern forms in San
Juan Colorado (Stark Campbell et al 1986) are nduk šùh and

3.2.2 Vowel quality. Various Mexican languages have vowels with qualities
distinct from the five standard vowels found in Spanish. In Zoque, there is a sixth
vowel, which is a high central unrounded vowel, /i/. In the colonial Zoque
documents published by Ruz (1997), attempts can be seen to represent this sixth
vowel with an ash, <æ>, or with a caudated e, < >.33 However, it is not adequately
represented on the whole (Smith Stark 1997).
Otomi is a language endowed with a rich system of nine vowel qualities.
i u
e o
e a
Table 1: Vowel qualities of Otomi
The Franciscan friars Pedro de Cárceres (ca. 1580) and Alonso Urbano (1605)
employ <œ> for the mid central unrounded vowel, /A/, and for the high
central unrounded vowel, /i/, although Urbano, at least, often does not maintain
this distinction, writing <œ> for both.34 Cárceres uses <é> and <è> to represent
half close and half open mid front vowels, respectively; and <o> and <o> for the
corresponding back vowels. Urbano (1605) uses the graphemes <ê> and <o>,
which might possibly be related to /ε/ y , respectively, but this usage is far from
Copyright © 2005. John Benjamins Publishing Company. All rights reserved.

clear (cf. Lastra 1992, Smith Stark 1992b). In the 18th century, Neve y Molina
(1767) uses for /e/ and for ; in addition, he uses two invented letters for
and , a whose rightmost stroke continues vertically downwards for the

32
The only other case of a syllabic sign that I am aware of, is one recently brought to my attention
by Michael Knapp; in a colonial Mazahua document which he is editing, the letter <k> is used to
represent the syllables /ka/, kha/ and /k'a/.
33
In the Latin based orthographies of New Spain, these two letters were familiar from the tradition
of writing Latin, where they both are derived from the sequence <ae>, the tail on the < > originally
being a reduced form of the vowel <a>.
34
I interpret this letter as an o-e ligature, though Acuña (Urbano 1990) reads it as an a-e ligature,
which would perhaps better explain why there are occasional uses of a caudated < > instead of
<œ>.

first, and an <e> with a line rising from the top right edge, curving leftwards, and
with another line falling from the bottom right edge and also curving leftwards for
the second. He thereby registered all nine vowel qualities in an adequate, though
antiesthetic, way.
Chinantec (Jmiih) is another language which distinguishes various vowel
qualities. Highland Chinantec, that spoken by the dzämo' (Hernández López
2000), for example, has the following inventory of nine contrastive vowel
qualities.35
i y u
e o
æ a
Table 2: Vowel qualities in Chinantec
Barreda (1960[1730]) (fl. 1730) seems to distinguish seven of these vowels
orthographically. In addition to the five vowels shared with Spanish, <i, e, a, o, u>,
he also writes <ui> for and <æ> for /Ә/. This claim is based on a superficial
examination of the text, where one finds <mui> "woman" and "water" for forms
that Rensch (1989) reconstructs as Proto-Chinantec * and * respectively,
and <næ> "now" and <quæh> "give" for Proto-Chinantec *nә'HL and *kwә...,
respectively. The vowels, /æ/ and /y/, on the other hand, do not seem to be
distinguished. For example, Barreda has <na> "I" and <ña> "five" where
Hernández López (2000, Anexo: "Lista de palabras") registers /hnæ:/ (#1) and
/' æ:/ (#105) in Yolox; and Barreda has <nu> "casa" and <ynuh> "suck" where
Yolox has /'nüh/ (#32) and /y 7 "chúpalo" (#67) according to Hernández López.36
Copyright © 2005. John Benjamins Publishing Company. All rights reserved.

3.3 Prosodic elements

The prosodic elements of stress, duration and tone were particularly difficult
areas for the colonial grammarians, especially the latter two.

3
In addition, there is a contrast of nasalization and vowel length, yielding 33 different vowels in
all (Hernández López 2000:45).
36
I have not studied Barreda's orthography in detail, nor does Cline say very much in his edition of
the text (1960:74-76), which includes useful, though brief, Spanish-Chinantec (1960:81-118) and
Chinantec-Spanish (1960:125-131) vocabularies. The orthography of this document merits a de
tailed study, now that more descriptively adequate materials are available on the language. In
Hernández López, u with dieresis is used for the high front rounded vowel and underscoring marks
nasalization.

3.3.1 Stress accent. During the 16th century, stress accent was not represented in
Spanish orthography, although Nebrija's comments (cf. §2) show that it was well
understood in Spanish at least. It is rarely represented when writing Indian
languages. In some cases, this may simply be because it was not constrastive. For
example, in Classical Nahuatl stress regularly falls on the penultimate syllable,
except in forms with the vocative suffix -e, which always receives stress: "Note
that the acute accent is never placed on the final syllable if it is not on the e of the
vocatives" ("Nota que nunca se pone accento agudo en la vltima sino es en la. e.
de los vocatiuos", Rincón 1967[1595]:64).
In Purepecha, stress was contrastive, but not written. Even so, the Franciscan
friar Maturino Gilberti (1558) clearly recognized the contrastive nature of stress.
He explains that:
En esta lengua ay muchos vocablos que en latin llamamos Æquiuocos, de dos significados,
en los quales muy facilmente se podrían engañar los menos auisados: por que su significado
solamente se conosce en el accento: el qual se ha de hechar conforme al significado de los
dichos vocablos Equiuocos (signature bi r)

[In this language there are many words that in Latin we call Equivocal, with two meanings,
by which the less well-informed can easily be fooled: because their meaning can only be
known by the accent: which must be placed according to the meaning of said equivocal
words]
He goes on to give a series of seven minimal pairs which depend solely on the
position of the stress, which in this language nearly always falls on one of the first
two syllables of the word. For example, tecándini is "incorrigible" and técandini is
"deaf. He uses an acute accent over the stressed vowel in all these pairs. In this,
Copyright © 2005. John Benjamins Publishing Company. All rights reserved.

he seems to follow Nebrija, who likewise uses an acute accent to distinguish pairs
like ámo "I love" and amó "he loved" in Spanish.37
One case where stress accent was written fairly consistently was in the
Dominican friar Juan de Cordova's Zapotec dictionary (1578; cf. Smith Stark
2003). In Zapotec, stress is predictable grammatically, but not in terms of the
surface phonological string. It generally falls on the first syllable of lexical roots.
Cordova (c. 1501-1595) marks stress with any one of the three graphic accents he
uses in his vocabulary: the acute accent ('), the grave accent ('), and the circumflex

37
Cf. Nebrija in Quilis'edition (1492 book 2, chapter 2:18r; 1984:138-139). Nebrija cites Quintilian
as his model and says that the Latin rhetorician uses an apex for signalling stress, by which he
apparently means an acute accent.

accent (^).38 The written accents in this dictionary serve an important cumulative
function as well, allowing the reader to pick out the lexical morphemes quite
easily, since spaces are often omitted at word boundaries and compounding and
incorporation are quite common.

3.3.2 Duration, The marking of vowel length was especially difficult, it seems, for
the Europeans. For example, vowel length is contrastive in Classical Nahuatl (and
most modern forms as well), but was not adequately identified as such until the
Jesuit priest Antonio del Rincón published his grammar in 1595, and was only
written consistently on rare occasions, such as when Rincón's successor and
fellow Jesuit, Horacio Carochi (1579-1662), published his grammar in 1645.
Classical Nahuatl vowel length continued to be largely ignored in modern studies
of the language, until Andrews (1975) and Launey (1979, 1986) published their
ground-breaking grammars, and Karttunen, her dictionary (1983). Even studies of
the modern dialects tend to pay inadequate attention to this feature of the language.
Rincón suggested representing vowel length with a grave accent (') in final
position and an acute accent (') elsewhere. Short vowels would be left unmarked.
Carochi, who improved upon Rincón's grammar in many ways (cf. Smith Stark
1996), used a macron (¯) to mark long vowels and an acute accent (') for short
vowels.
In the case of other languages with contrasting vowel length, for example many
Mayan languages, I know of no cases where the distinction was marked
successfully.
The difficulty caused by vowel length is somewhat surprising when one takes
into account that Latin, the major cultural language in Europe during the Middle
Copyright © 2005. John Benjamins Publishing Company. All rights reserved.

Ages and the missionaries' primary model for grammatical description, had an
opposition between long and short vowels. However, knowledge of the nature of
vowel length seems to have been largely forgotten by the 16th century. For
example Nebrija makes the following wry observation in this regard: "But the

38
According to Radin (1943-1944:8), Cordova uses both a breve (ˇ) and a circumflex (^) accent in
his grammar. Whitecotton and Whitecotton (1993:xxii), on the other hand, say that he used the
circumflex (^), the hachek ( ˇ ), and the dieresis ("). De la Cruz (2001:28-29) notes that the
circumflex and the hachek predominate, though there are also cases of the breve and, possibly, the
acute accents. The examples of the latter, however, may in fact be damaged examples of the
circumflex accent. Like Smith Stark in the case of the Vocabvlario, he claims that these diverse
accents in the Arte are all used to mark stressed vowels.

Castilian cannot sense this difference [in vowel length], not even those who
compose verse can distinguish between long and short syllables, no more than
those who composed some works in Latin verse in past centuries could sense it"
("Mas el castellano no puede sentir esta diferencia [en la longura de las sílabas], ni
los que componen versos pueden distinguir las sílabas luengas de las breves, no
más que la sintían los que compusieron algunas obras en verso latino en los siglos
passados", 1492:17r).

3.3.3 Tone. Of all the phonological characteristics of the languages of New Spain,
tone, which is particularly common in the languages of the Otomanguean family,
was the one that caused the most difficulty for the early linguists. In many cases,
the tonal nature of a language was not really recognized until the 20th century.
Such was the case with Yucatec Maya, for example, which was first identified as a
tone language by Kenneth Pike in 1946.39 However, there are some commentaries
which indicate that the friars were aware of tonal phenomena and even made some
attempts to represent them, though not with much success. For example, the
Dominican friar Francisco de Alvarado observes that "The meanings of many
words vary a lot with the accent, and some not only in having or losing an accent
mark, but even as to pronouncing the tip with softness or with fullness of voice"
("En el acento varían muchas palabras la significacion, y algunas no solamente en
tener o perder vna tilde pero aun en pronunciar el punto con blandura o con la boz
llena", Alvarado 1593, "Prologo":6r). His contemporary and fellow Dominican
Antonio de los Reyes also attempts to talk about tone:

En todos los verbos aqui puestos y en los demas de su <r>alea que propriamente son equí
Copyright © 2005. John Benjamins Publishing Company. All rights reserved.

vocos se han de mirar los acentos como van señalados, y donde no ay señal, entiendan que
la pronunciacion ha de ser llana, que de otra manera sera mucha la equiuocacion y

39
Contrast this situation with that of Vietnamese. According to Thompson (1987:56) in 1624 the
Jesuit missionary Alexandre de Rhodes (1591-1660), a native of Avignon in southern France,
arrived in Indochina and in 1651 published his Dictionarium annamiticum, lusitanum et latinum in
which the writing system was close to phonemic, including, apparently, the representation of six
tones, five by way of overt diacritics (', ~, ?, .). The Jesuits Michele Ruggieri (1573-1607) and
Matteo Ricci (1552-1610) did not represent tone in their Chinese dictionary of ca. 1580, but the
Dominican friar Francisco Varo (1627-1687) did do so systematically by way of five diacritics (",
, ,̀ ,́ ˇ) in his grammar of Mandarin (2000[1703]). It seems reasonable to suppose that traditional
Chinese language studies may have provided an important stimulus which made these early and
sophisticated analyses of tone systems possible in Asia.

confusion, y fuera cosa enfadosa pretender poner en esto reglas, de mas de que fueran sin
prouecho." (Reyes 1976[1593]:54)

[In all the verbs given here and in the rest of their kind which properly speaking are
equivocal, it is necessary to observe the accents which are marked, and where there is no
sign, understand that the pronunciation is plain, and that otherwise there would be much
error and confusion, and it would be a vexing thing to try to reduce this to rules, in addition
to which they would be of no use.]
The evidence is extremely ambiguous and sporadic, but it appears that the
Dominicans might have recognized three tones: acute (written with an acute
accent), plain (unmarked), and grave (written with a grave accent). For example,
the following possibly contrastive pairs appear in Alvarado (A) and in Reyes (R).40
a. yo+sànu÷ndi "crecer [to grow]" (R:54, A)
yo+sánu÷ndi "doblar [to bend]" (A)
yo+sánu site÷ndi "doblar la rodilla, haziendo reuerencia
[to bend the knee, bowing]" (R:54)
b. yo+sani÷ndi "soñar [to dream]" (R:54, A)
yo+sàni÷ndi "ordeñar [to milk]" (R:54)
yo+sáni÷ndi "ordeñar" (A)
c. yo+ndayu÷ndi "temblar [to tremble]" (R:55 "llano"; A)
yo+ndàyu÷ndi "llorar [to cry]" (R:55 "acento")
yo+ndayu÷ndi "llorar" (A)

d. yo+sána÷ndi "atrauesarseme el bocado en el garguero [for a bite of food

to get stuck in my throat]" (R:56)
yo+sa'na÷ndi "ahogarse [con violencia] [to choke violently]" (A)
yo+sana "reborverse los atomos en el aposento por el rayo del sol
Copyright © 2005. John Benjamins Publishing Company. All rights reserved.

que entra, y por metaphora dizen de multitud de gente [for

the small particles of dust in the air of a room to swirl
about due to the rays of sunlight which enter, and,
metaphorically, they say it when talking about a large
group of people]" (R:56)

However, the examples of this sort are too few and too contradictory to allow a
clear conclusion to be drawn at this time.
The difficulty with tone remains with us to this day. Indeed there is a
40
In my analysis of the following examples, a prefix is separated by a plus sign (+) and a suffix, by
a division sign (÷).

widespread tendency to ignore tone when writing descriptions of Mexican

languages and developing practical orthographies for them.

4. The articulatory description of sounds

The recognition of a wide variety of different sounds can be inferred from the
orthographic practice of the friars; however, such recognition is above all
pragmatic and intuitive. That is, the sounds which are functionally different are
recognized as such in practice. It is generally assumed that they will be
pronounced as their Spanish counterparts would suggest. At times, however,
reference is made to the pronunciation of sounds in other languages to help
elucidate the value of a particular grapheme (§4.1), or explicit articulatory
descriptions of sounds are essayed (§4.2).

4.1 Descriptions by analogy with other languages

On the few occasions when reference is made to the pronunciation of another
language, it is to Latin or Hebrew. There is very little or nothing said about the
other languages which some of the missionaries surely knew: Greek, Arabic,
French, Basque, Flemish, Italian, Danish etc. The two most prominent examples of
this technique are the descriptions of the geminate/11/and of the affricate /¢/ in
Nahuatl. The Franciscan friar Andrés de Olmos (1972[1547]:198) explains the
pronunciation of the double <11> - a particularly common sequence in Nahuatl
which often results from the addition of the absolutive suffix -tl(i) to deverbal
noun stems which end in /l/ — in the following way: "in the pronunciation of
some letters, it seems to be more like Latin than like our Romance. And thus, they
use two ll's as when in Latin we say villa, not like in Romance when we say
Copyright © 2005. John Benjamins Publishing Company. All rights reserved.

maravilla" ("en la pronunciacion mas parece allegarse al latin que no a nuestro

romance en algunas sillabas o letras. Y ansi usan de dos // como en latín que
dezimos villa, y no como en el romance que dezimos marauilla"). Here of course
the reference is to the fact that in Latin the double // refers to a true geminate, /ll/,
whereas in Spanish, it is a digraph which represents a palatal lateral, /λ/.41
With respect to the affricate /¢/, both the Franciscan Alonso de Molina and the
Jesuit Horacio Carochi explain its pronunciation in terms of Hebrew. Molina, for
example, says that "This language has an Hebraic letter, the tsade, which should
be written as t plus s or as t plus z, and which should be pronounced like t plus s"

41
A similar observation is also made by the Jesuit Antonio del Rincón (1595:64v[1967:68]).

("Esta lengua tiene vna letra Hebrayca, que es tsade. La qual se hade escreuir con.
t y, s, o con .t. y .z: y ase de pronunciar como. t. y. s.", 1571a:5r).42 Carochi says
that "They have another letter which is pronounced like z or c, but stronger; it
corresponds to the Hebrew letter tsade; it is written in this language as t plus z. [...]
but it is a single letter, even though it is written as two." ("Otra letra tienen
parecida en la pronunciacion à la z. y à la c. pero es de mas fuerte pronunciacion,
corresponde à la letra Hebrea llamada, Tsade; escriuese en esta lengua con f. y z.
[...] pero es vna sola letra, aunque se escriue con dos.", 1983[1645]:lv).

4.2 Explicit articulatory descriptions

The attempts at the articulatory description of sounds tend to be vague and
impressionistic. Perhaps the most explicit presentation in this regard is that of the
Franciscan Juan Baptista Bravo de Lagunas (1574:75-76), who simply quotes from
the third book, De arte grammatica, of a 4th-5th century Medieval treatise by the
North African Martianus Capella, which describes the articulation of the sounds of
Latin, presented in alphabetical order, as follows (cf. Arias 2001; Norton 1995,
vol. 2:828):
A. Sub hiatu oris congruo solo spiritu memoramus.
B. Labris per spiritus impetu reclusis edimus.
C. Super molaribus linguę extrema appulsis exprimitur.
D. Appulsu linguae circa superiores dentes in nascitur.
E. Spiritus facit lingua paululum pressiore
F. Dentibus, labrum inferius deprimentibus lingua palatoque dulcescit.
G. Spiritus cum palato.
H. Contractis paulum faucibus ventus ex halat.
I. Spiritus prope dentibus pressis.
K. Faucibus palatoque formatur.
Copyright © 2005. John Benjamins Publishing Company. All rights reserved.

L. Lingua palatoque dulcescit.

M. Labris imprimitur.
N. Lingua dentibus apulsa colliditur.
O. Rotundioris spiritu comparatur.
P. Labris spiritus erumpit.
Q. Appulsu palati ore restricto.
R. Spiritu lingua crispante corraditur.

42
This description is quite disturbing since it seems to evidence confusion between apical and
laminal in Molina's Spanish. However, Suárez Roca (1992:84) notes that Molina corrects this
statement in the 1576 edition of his grammar, where he states that "one should write t. plus z. and
not t plus s, and one should pronounce t plus z" ("se ha de escreuir con [Link].y no con t. y s, y hase
de pronunciar como t y z.", Arte 1576:133).

S. Sibilum facit dentibus verberatis.

T. Appulsu lingue dentibus impulsis excudit.
V. Ore constricto labrisque promulis exhibet.
X. Quicquid. c. & .S. formauit exibilat.
Y. Apressis labris spirituque procedit
Z. Vero iccirco Appius claudius detestatur: quod dentes mortui, dum exprimitur: imitator.43
However, there are occasional comments about articulation which are useful or
at least understandable with a bit of hind sight.
4.2.1 Apico-alveolar vs. lamino-alveolar. Consider, for example, the observations
made by the Franciscan Matutino Gilberti about the pronunciation of <ts> and
<tz> in Purepecha. As was mentioned earlier, many Mexican languages have an
alveolar affricate, which was usually written as <tz>. However, Gilberti dis
tinguishes two different such affricates, which he writes as <tz> and <ts>. He
describes the first as follows:
Muchos vocablos ay enesta lengua derezia pronunciacion, para la qual pronunciacion es
menester añadir .tz. [...] ha de ser poniendo el pico dela lengua entre los dientes delanteros
de arriba, abriendo vn poquito los labios: y hiriendo rezio el paladar con la lengua: como en
la pronunciacion de tza y debetzayda (1558:109v)
[There are many words in this language of forceful pronunciation, for which it is necessary
to add tz [...] it should be made by placing the tip of the tongue between the upper front
teeth, opening a bit the lips: and articulating forcefully against the palate with the tongue: as
in the pronunciation of tza and of Betzayda44]

43
Translation (based on Lucas 2002, with help from Beatriz Arias, Aurelia Vargas and Bulmaro
Reyes; but of course I am accountable for any errors): A "we mention [it] with a single and
convenient puff of air immediately upon opening the mouth"; B "we exhale, the lips opened by the
Copyright © 2005. John Benjamins Publishing Company. All rights reserved.

force of the breath"; C "it is pronounced with the contact of the back of the tongue on the molars";
D "it comes from the contact of the tongue around the upper teeth"; E "the breath makes a bit of
pressure on the tongue"; F "it is sweetened with the tongue and the palate, the teeth pressing
against the lower lip"; G "the breath with the palate"; H "air is exhaled, constricting a bit the
throat"; I "the breath with the teeth almost closed"; K "it is formed by the throat and the palate"; L
"it is sweetened with the tongue and the palate"; M "it is made with the lips"; N "the tongue is
struck in contact with the teeth"; O "it is compared to a more rounded sound"; P "the breath bursts
forth from the lips"; Q "by contact with the palate, with the mouth narrowed"; R "the curling
tongue is bunched up in the breath"; S "it makes a whistling sound, dashing against the teeth"; T "it
is produced by the contact of the tongue against the teeth"; V "it is produced with the mouth tight
and the lips in contact"; X "whatever formed c and s makes a whistling sound"; Y "tightening with
the lips the breath that goes forth"; Z "truly for this reason, Appius Claudius is detested, because
when it is pronounced, it sounds like the teeth of a dead man".
44
Betzayda is apparently his rendition of the Biblical name Bethsaida, an ancient town of Israel

Here, by forceful pronunciation, Gilberti is describing the difference between a

laminai fricative and an affricate /¢/ (= [t ]). The second affricate is described
as follows: "There are other words for whose pronunciation it is necessary to make
as if to pronounce ts, because their first syllable should be pronounced rather
forcefully, articulating against the palate with the tongue, and then relaxing the
tongue for the rest of the word, as if whistling" ("Otros vocablos ay: para cuya
prolacion es menester fingir .ts. por que su primera silaba seha de pronunciar algo
rezio, hiriendo el | paladar con la lengua, y luego afloxando la lengua para lo de
mas de la dicion: como echando vn silbeo", 1558:110v-111r). This description
seems to confirm the fact that <s> represented an apico-alveolar fricative, [ ], like
that still heard in Castilian Spanish, which has a whistling type of assibilation. The
two affricates described by Gilberti thus seem to be [ ] and [t ].45

4.2.2 An interdental fricative. Spanish sibilants underwent several important

changes in the 16th and 17th centuries. Two hundred years after Gilberti described
the affricates in Purepecha, the priest Carlos Tapia Zenteno described <z> in
Huastec in the following way: "Z is pronounced properly with the tongue
somewhat protruding from the teeth and touching them." ("La z se pronuncia con
todo rigor, con la lengua algo fuera de los dientes, pegada à ellos.", 1767:2). From
this description, from the examples given, as well as from knowledge of the
present day pronunciation of the language, we can surmise that the author was
referring to the interdental fricative [9], which is still found in Huastec. This
example is particularly interesting because it suggests that the grapheme <z> in
Spanish was a natural candidate for representing this sound, a peninsular
pronunciation probably never found in American Spanish, where <z> has always
Copyright © 2005. John Benjamins Publishing Company. All rights reserved.

represented

near the shores of the Sea of Galilee, mentioned various times in the New Testament of the
Christian Bible (for example, in Matthew 2:21; Mark 6:45, 8:22; Luke 9:10; John 1:44).
45
A parallel difference was also surely found between <ç> and <s> . The apical sibilants
<ts, s> only occurred before , and so were in complementary distribution with the
corresponding palatal sibilants <ch, x>, which never occurred before (cf. Monzón 2001, 2005)
(note that the symbol stands here for an apico-alveolar s).

4.2.3 Glottal stop. Another good example of a successful phonetic description is

the Jesuit Antonio del Rincón's characterization of the sound known as saltillo, the
little jump, in Nahuatl. He was the first to mention this term, which is still used
today in Mexican linguistics to describe the glottal stop.
Accento del saltillo es, quando la syllaba breue se pronuncia con alguna aspereza como,
tlaçolli.46
Esta aspereça no es del todo. H. hablando propriamente porque en la prouincia de
Tlaxcalla, y en algunas otras apartadas de Mexico pronuncian con este Spiritu aspero muy
affectadamente de manera que no solo es. H. mas aun pronunciada con mucha aspereça, y
tuerça verbi gracia, tlacohtli, tlahtoani, y por esta causa con mucha raçon algunos han
llamado, a este espiritu aspero el saltillo, porque ni del todo a de ser. H. como en Tlaxcala ni
suspension de la syllaba, como algunos han dicho.
Accento suspenso: porque ese es officio del accento agudo que es largo y suspende la
syllaba, y assi no es mas de vna manera de salto, o singulto, que se haze en la syllaba, y
esto solamente se halla en la syllaba breue. (Rincón 1595:63v-64v; cf. 1967[1595]:61-62)

[The little jump accent is when a short syllable is pronounced with some roughness, as in
tlahçolli "trash".
This roughness is not properly speaking an h in all regards, because in the province of
Tlaxcala, and in some other parts of Mexico, they pronounce this spiritus asper with much
affectation, so much so that not only is it an h, it is even more roughly and strongly
pronounced, as in, for example, tlacohtli "slave", tlahtoani "great lord". And therefore, for
good reason, some people have called this spiritus asper the saltillo, the little jump,
because it should not really be an h as in Tlaxcala, nor the suspension of the syllable, as
some have claimed.
The suspended accent: because this is the function of the acute accent which is long and
which suspends the syllable, and thus it is no more than a kind of jump, a jiggle, in the
syllable, and this is only found on short syllables.]
Copyright © 2005. John Benjamins Publishing Company. All rights reserved.

Here Rincón refers to two different realizations of the saltillo, both of which are
still present in modern varieties of Nahuatl. One was a glottal fricative, a spiritus
asper as he calls it, which may even have had some velar friction as pronounced in
Tlaxcala. The other was a true glottal stop. He rejects the description of this accent
as the suspension of the voice, which he feels more properly describes vowel
length.
Similarly, the Dominican Antonio de los Reyes, in his grammar of Mixtec
published in 1593, gives a fairly clear description of the glottal stop.
Solo se puede notar acerca de ¿numerables vocablos que se hallaran en esta lengua que por

46
Karttunen (1992[1983]:271) gives this form as tlahzolli "trash".

mas claridad se escriben con h antes de la vocal, y no por eso se ha de entender, que es
aspiracion, sino que la pronunciacion ha de ser de tal | suerte que parezca dexar la vocal
sola como yosini mani ñahandi es amar, la pronunciacion ha de ser como esta aqui escripto:
yosi inimani ña andi, que haga diuision, pero como en nuestra lengua Castellana se escribe
desta manera, muchos dias ha, sin ser aquella ha aspiracion sino que al parecer se declara
mejor ansi lo que se escribe, se puede tambien escriuir en la Mixteca con la h para las
vocales, y comunmente se vsa esto entre los curiosos de la lengua. (1976[1593]:2-3)

[Just one thing can be noted about the innumerable words that are found in this language
that for clarity's sake are written with an h before the vowel, which should not be
understood to mark aspiration but rather that the pronunciation should be such as to appear
to leave the vowel alone, as in yosini mani ñahandi "love", which should be pronounced as
follows: yosini mani ña andi,47 with a division between ña and andi. But since our Castilian
language has been written this way for some time now, where the aitch is not aspiration,
but rather it appears that it is clearer to write it this way, it can be done so in the Mixtec
area as well, and commonly those interested in the language do so.]
4.2.4 Nasalization. Nasalization is described as such by various authors. For
example, the Dominican Antonio de los Reyes (1972[1593]:2) explains that "the
pronunciation of q with n or with h is pronounced through the nostrils" ("la
pronunciacion de la q con n o con h se pronuncia por las narizes"). Reyes' fellow
Dominican Francisco de Alvarado, in a rare display of ethnocentrism, speaks
rather disparagingly of the same phenomenon: "this language goes so far as not to
be content with what nature has given us for pronouncing, but rather it rises to the
nostrils: and these are used in some pronunciations which would otherwise be
incomplete" ("llega a tanto esta lengua: que no se contenta con la que nos dio
naturaleza para pronunciar sino que sube a las narizes: y dellas se vale en algunas
pronunciaciones; que sin este socorro quedan faltas", 1962[1593], "Prologo":6r).
Copyright © 2005. John Benjamins Publishing Company. All rights reserved.

Nasalization of vowels is also clearly described in Otomi, where the Franciscan

friar Pedro de Cárceres (1907[1580]) notes in the "Clave" at the beginning of his
grammar that "They have another [pronunciation] through the nostrils like those
among us who speak with a nasal twang" ("Tienen otra [pronunciacion] por las
narizes como los gangosos entre nosotros").

47
I have corrected the second repetition of this phrase so that the only difference is that being
illustrated. In Alvarado's dictionary, the closest expression for "love" that I can find is chidzu-
maniyosiniñahandi "amar fuerte mente" (f. 18r).

4.2.5 Aspiration. Aspiration was also clearly described in some grammars. For
example, Gilberti (1987[1558]:110r), in his description of the pronunciation of the
aspirated affricate thz, explains that "it is necessary to let out your breath
abundantly from the depths of your chest" ("menester es sacar el huelgo
abundosamente del profundo del pecho").

4.2.6 Vowel height. One of the most remarkable examples of the description of the
pronunciation of unfamiliar sounds is that found at the beginning of the Franciscan
friar Pedro de Carceres' grammar of Otomi. The notes on pronunciation are
entitled "Clave para entender la pronunciacion figurada de la lengua otomi en el
arte de Fr. Pedro de Cárceres" ("Key for understanding the notated pronunciation
of the Otomi language in Friar Pedro de Cárceres' grammar") in the printed
version, but simply "On some observations about the Otomi language" ("De
algunos apuntamientos de la lengua otomj") on the first page of the original
manuscript, which Nicolás Leon fortunately reproduces in his edition of the
grammar.48 There, Cárceres refers to many of the unfamiliar characteristics of
Otomi phonology (to a European). For example, he explains the mid and high
central vowels as follows:
Tienen otra pronunciacion que ni bien es e. ni bien es o. ni v. y usan unas veces, la
pronuncian mas obscura y apretada que otras apretando mas la garganta; y asi cuando es
media y no tan apretada la significaremos con este diptongo œ [...] y cuando es mas
apretada la significaremos con este ("Clave")

[They have a pronunciation which is not quite e nor quite o, nor u and they at times are
accustomed to pronounce it more obscurely and tightly than at others, tightening more the
Copyright © 2005. John Benjamins Publishing Company. All rights reserved.

throat. Thus, when it is only about half as tight, we will indicate it with the diphthong œ [...]
and when it is more tightly articulated, we will indicate it with this.]
Likewise, the "Clave" describes Otomi's half open [ε, Ɔ] and half closed [e, o]
mid vowels as follows: "Sometimes they open the mouth', put on top this ` which
falls to the right [...] Other times they close it, and one should put on top this ´
which falls to the left" ("Otras vezes abren la boca; ponerse encima esta ` caida a
mano derecha [...] Otras la sierran, ponerse ha esta ' caida a mano izquierda").49

48
I am grateful to Klaus Zimmermann for bringing the existence of this facsimile to my attention.
49
Toru Maruyama informs me that the Portuguese grammarians Oliveyra (1507-1681) (1536) and
Barros (1496-1570) (1540) distinguished closed and open mid vowels in their respective
orthographies. The normal vowel signs were used for the closed varieties. Oliveyra (1536) used an

4.2.7 Laryngealization. The "Clave" also does a pretty good job at describing the
glottalized sounds in Otomi. The ejectives /k'/, /t'/ and /¢'/ are described as
follows:
Tienen otras [letras] pronunciaciongutural, apretando la garganta, dando un como
castañeda. Escriuese con esta letra k [...] Otras veces pronuncian la t apretadamente
serrando la boca y hiriendo con alguna fuerza con la punta de la lengua en los dientes de
arriba. Significasse con dos tt [...] Tienen otras que ni es z ni c. Escriuese con tz. [...] Otra
tienen con mas behemencia, escriuese con ttz ("Clave")

[Other [letters] have a guttural pronunciation, tightening the throat, and producing a sort of
click. It is written with the letter k[...] At other times they pronounce ttightly, closing the
mouth and articulating with some force with the tip of the tongue against the upper teeth.
This is represented with two tt's [...] They have others that are neither z nor c. It is written
with tz. [...] They have another with more vehemence, it is written with ttz.]
Here, the glottalized sounds are described as being made by tightening the
throat and making a click, or by pronouncing forcefully, or by being produced
with greater vehemence.
The glottalized /?d/ was likewise described with some degree of acuity in the
"Clave": "They have other [letters] that are not quite a d nor a soft r, it is a d
pronounced in the throat and difficult to pronounce. It will be indicated with å "
("Otras [letras] tienen que ni es bien. d. ni r. blanda; es d pronunciada en la
garganta y difícil de pronunciar Significarse ha con esta, â "). Here, "in the throat"
apparently refers to the laryngeal activity involved in the production of this sound.

4.2.8 Retroflexion. As a further example of the relative success of some of the

articulatory descriptions achieved by the colonial linguists, consider the

description of the retroflex rhotic in Purepecha offered by the Franciscan friar Juan
Baptista [Bravo] de Lagunas, who proposes representing it as <rh>:
Y tambien es de notar que de aqui adelante para la recta pronunciacion se escreuiran
aspi|radas las. Rr. que tienen la pronunciacion entre L. y. R. por escusar algun inusitado
character de que las emprentas suelen carecer, a differencia de las demas que se escriuen era
medio de parte sin encoger, o alçar la lengua en su pronunciacion y diuerso significado [...]
Aunque en la verdad en vnas differencian muy poquito de nuestra R. [...] Y assi opongo por
nota de atencion: para que nadie de oy mas dexe la razon y lo que peculiarmente la propia
lengua pide, por seguir su sola opinion. (1983[1574]:72-73)

epsilon, <e> for open /ε/ and an omega, <ω>, for open /ɔ/. Barros (1540) used a caudated e, <ę>,
for open /ε/ and an accented o, <ó>, for open /ɔ/.

[And it is also to be noted that henceforth, for correct pronunciation, the r's that are
pronounced halfway between l and r will be written with aspiration, rather than with some
unusual character which the presses generally lack, unlike the rest [of the r's], which are
written in the middle without bunching or raising the tongue in their pronunciation and
diverse meaning [...] Although, in fact, in some they are very little different from our r. [...]
And so I raise this word of warning: so that no-one from today onward will ignore reason
and that which the very language itself requires, in order to follow his own opinion.]
Notice that even at this early date, the practical consequences of inventing
totally new symbols were considered when attempting to decide how best to
represent new sounds.

4.2.9 Liquids. The term liquid is used on at least two occasions (Olmos
1972[1547]:201, Cárceres 1907[1580]:88). Both times it refers to the use of [u] (or
the letters <u/v>) as a semivowel of some sort rather than as a full vowel. For
example, Olmos says that in the sequence <cu> "Sometimes the u becomes liquid"
("Algunas vezes [...] se haze la u liquida"), by which he means that the sequence
represents /kw/, rather than /ko/. Cárceres refers to verbs which begin with "liquid
v" ("v liquida"), by which he means verbs which begin with /w/.

4.2.10 Concluding remarks aboutarticulatorydescription. I do not mean to over

state the success of the articulatory descriptions of sounds in New Spain. They are
typically confusing, unclear, imprecise and impressionistic, and do not reveal the
systematic level of knowledge found earlier in Nebrija. I simply want to point out
that it is possible to find a number of cases which reflect fairly mature thinking
about the way sounds are produced.
Copyright © 2005. John Benjamins Publishing Company. All rights reserved.

5. The sound system, the alphabet and Quintilian's method

I find little evidence for the idea of a sound system among the colonial gram
marians of New Spain. Generally, the description of the sounds of a language
begins, following Nebrija's interpretation of the Quintilian method, with the letters
of the alphabet with which it is to be written, the letters which are not needed, and
the additional letters which need to be added in order to adequately express the
sounds of that language. For example, the Franciscan friar Alonso de Molina
(1571:5r) begins his Nahuatl grammar as follows: "First, one should know that this
language lacks six letters, which are b, d, f, g, r, s [...] In addition, this language
has a Hebrew letter, which is tsade" ("Primeramente es de saber, que esta lengua
carece de seys letras, que son. b. d. f. g. r. s [...] Allende desto, esta lengua tiene

vna letra Hebrayca, que es tsade"). In this case it is understood that the
unmentioned letters of the alphabet are shared with Spanish.
The phonological and orthographic descriptions tend to treat sounds
atomistically and in alphabetical order, with little or no sense of natural classes
beyond the common distinction between consonants and vowels. For example,
Friar Juan Baptista de Lagunas (1983[1574]:1) presents the sounds of Purepecha
as follows: "These Indians have 21 letters, without counting the h, which is a sign
of aspiration. There are 5 vowels, a, e, i, o, u and 16 consonants, b, c, d, g, h, k. m,
n, p, ph, q, r, s, t, x, y, z" ('Tienen estos Yndios veynte y vna letras, sin la H, nota
de aspiracion. Las quales son cinco vocales, a, e, i, o, u. Consonantes diez y seys.
b, c, d, g, h. k. m, n, p, ph, q, r, s, t, x, y, z."). One is struck by the little coherence
found in the texts on the matter of the alphabet. For example, Lagunas includes ph
in his alphabet, but not th, tz, thz or rh, all of which are also special combinations
discussed in his grammar.
There are, nonetheless, occasional indications of a more systematic notion of
the phonetically motivated groupings of sounds. For example, two authors refer to
the class of rough sounds ("letras asperas"), which is an evident reference to
sibilants. The Jesuit Antonio del Rincón (1967[1595]:67-68) uses this term in his
discussion of a process which affects tz, ch, x, and ç, which are all the sibilants of
Nahuatl. His successor Horacio Carochi (1983[1645]:53v) also mentions letras
asperas, and gives the two fricative sibilants, x and z, as examples.
There is also a striking case where a natural class is recognized implicitly by
the symbols chosen to represent sounds. Here I have in mind the convention used
by Alonso Urbano to represent ejective stops systematically by means of doubling
(see §2.1.2 above).
Copyright © 2005. John Benjamins Publishing Company. All rights reserved.

6. Grapheme and phoneme

Let's look now at the kind of phonological unit that the graphemes of the
different New Spain languages represented. In general, the units that are
distinguished correspond to the taxonomic phoneme, with no reference to internal
differences in pronunciation, although in a few cases there is some allusion to
allophonic variation. Of course, there are also many cases of underdifferentiation,
as well as of overdifferentiation and of what might be called displacement.

6.1 Phonemes and allophones

The colonial authors of New Spain usually seem to have a very concrete view
of phonological units. If two sounds are heard as different, then they should be
written as different; if they are written the same, then they should presumably be
pronounced the same. However, in actual fact, the units that are written tend to
correspond pretty well to the phonological units of the languages being described
(again, recall that I am talking about the best of the grammarians). In some cases
this is probably due to the influence of native speakers writing their own languages
(cf. §10); they would naturally tend to write phonological units rather than
phonetic ones. Another factor is simply the fact that the friars would not go out of
their way to make distinctions that were not functionally important.
Nonetheless, there are also cases where allophones are recognized. A
particularly clear, and also very early, case of this is found in the Franciscan friar
Andrés de Olmos' description of [g] in Nahuatl.
Pero, quanto a las letras que hemos dicho que no tenian, ay alguna dificultad, porque parece
algunas vezes pronunciar algunas dellas, y una destas es la g porque en esta diccion
uexotzinco, y aunque escriuen c, parece que pronuncian g, y lo mismo es en esta diccion
cenca; y aunque paresca a algu|nos por esta pronunciacion que sea de escriuir g y no c, a mi
me parece que ni en la pronunciacion es g, ni se deue tampoco escriuir; pues la c quando se
pone despues de la n parece que tira a pronunciacion de g quando esta en una misma
diccion, pero en la verdad no es sino pronuncacion de c. Y ansi no se ha de escriuir ni
pronunciar cenga, sino cenca. (Olmos 1972[1547], p. 197-198)

[But, with respect to the letters which we have said that they don't have, there is a problem,
because it seems that they sometimes pronounce some of them, and one of these is g,
because in the word uexotzinco, although they write c, it seems that they pronounce g, and
Copyright © 2005. John Benjamins Publishing Company. All rights reserved.

the same thing goes for the word cenca; and although it seems to some that for this reason
one should write g and not c, it seems to me that the pronunciation is not really g, nor
should it be so written; since the c, when it is placed after an n, tends to sound like a g when
in the same word, but in reality it is only the pronunciation of the c. And so, one should not
write nor pronounce cenga, but rather cenca.]
In this quote, we can see the role of the native speaker quite well. The Spanish
speakers hear a [g], but the Nahuas themselves write <c>; that is, they treat it as a
realización of /k/. Olmos identifies the role of a preceding nasal as the relevant
conditioning factor for a description of a phoneme which we would now describe
perhaps as follows in item and arrangement terms:

/k/: [g] / n_
[k] / elsewhere

He even recognizes that the voicing does not take place across word
boundaries. However, Olmos cannot quite accept that two different phonetic
sounds are phonologically equivalent, so he closes with a seeming contradiction by
saying that the apparent [g] is in reality only a pronunciation of <c> and that it
should neither be written nor pronounced like <g>.

6.2 Underdifferentiation
Of course, the phonological representations of the friars are not consistently
phonemic. There are many cases of underdifferentiation, that is, cases where
distinctive features of a language are not represented in the orthography. This is in
a sense what would be expected if the friars were truly unable to adapt to the new
sounds they encountered in the languages of New Spain. One of the main points of
this paper is that there is ample evidence that the friars were indeed able to learn to
distinguish, to represent, and to discuss explicitly, new sounds. However, they also
failed to recognize many distinctions. There is no point in listing all such failures,
since, as I announced at the beginning of this paper, my intention is to focus on the
achievements of the New Spain linguists, not to ridicule their failures. However, it
might be well to distinguish three types of underdifferentiation.
First, there are cases where certain characteristics were simply not recognized.
Here we might include tone in Otomi (Smith Stark 1992b) or the various types of
laryngealized vowels in Zapotec (Smith Stark 2003).
Second, there are cases where features are adequately recognized, but are not
Copyright © 2005. John Benjamins Publishing Company. All rights reserved.

incorporated into the writing system. Such seems to be the case of stress in
Purepecha, for example, which Gilberti (1558) recognizes quite explicitly, but
does not choose to write (cf. §3.3.1).
Third, there are features which were apparently recognized, but which were
often ignored in practice. Here, one might include features such as the distinction
between lax and tense consonants in Zapotec, which Cordova (1578) distinguished
at times, but more often than not left undifferentiated (Smith Stark 2003).

6.3 Overdifferentiation
More interesting in the present context are the cases of overdifferentiation,
where the friars register differences between sounds which are not contrastive. In
some cases, overdifferentiation can be attributed to distinctions which are pertinent
in Spanish but which are allophonic in the language being studied. A well-known
case is the use of both <o> and in Nahuatl texts. Classical Nahuatl only had
four contrastive vowel qualities, as well as an opposition between long and short
vowels.
i o i: o:
e a e: a:
Table 3: Vowel qualities and quantities in Classical Nahuatl
However, the vowel quality of/o/ and /o:/ varied phonetically between [o] and
[u]. This fluctuation is registered in many early texts, although the friars were also
aware that the distinction was not very important. The Franciscan Andrés de
Olmos (1547), for instance, author of the earliest surviving grammar of an
American language, notes the following:
Y es de notar que entre estas dos vocales o, u, hazen muy poca diferencia en la
pronunciacion y escritura, porque una misma diccion unos la pronuncian con o, y otros con
u. Ex. vnos dizen: ocelotl, y otros dizen: ucelutl. Y para esto, quales se ayan de pronunciar
y escriuir con o, y quales con u, no se podra dar regla general. (1972[1547]:198)

[And it should be noted that they make very little difference in pronunciation and writing
between these two vowels, o and u, because some will pronounce one and the same word
sometimes with o and other times with u. For example, some say ocelotl "jaguar, ocelot"
while others say ucelutl. As to which should be pronounced and written with o and which
Copyright © 2005. John Benjamins Publishing Company. All rights reserved.

with u, no general rule can be given.]

Olmos, however, did not take this observation to its logical conclusion of only
writing one vowel, though nearly a century later, the Jesuit Horacio Carochi
(1645) did do just that: "they use o at times so closed and obscure that it
approaches the pronunciation of vocalic u, but it doesn't stop being o, so that I do
not consider it to be correct to write Teūtl, but rather Teōtl, God" ("vsan de o.
algunas vezes tan cerrada, y obscura, que tira algo la pronunciacion de la u. vocal
pero no | dexa de ser o. y assi no tengo por acertado escreuir Teūtl, sino Teōtl,
Dios" 1983[1645]:lv-2r). Notice that even the clairvoyant Carochi resists saying
that /o/ has an allophone which is [u]; it only sounds a bit like [u].
Another case of overdifferentiation undoubtedly due to Spanish influence is the

distinction, referred to above in §4.2.1, that the Franciscan Matutino Gilberti made
between <s> and <x>, which were allophones of /š/, and between <ts> and <ch>,
which were allophones of /č/.
Not all cases of overdifferentiation can be attributed to Spanish influence,
however. Nahuatl provides an interesting example to the contrary. Syllable final HI
is voiceless in Nahuatl. Olmos (1547) proposes writing the voiceless /, which does
not contrast with voiced l, as <lh>, a suggestion which other authors did not
follow. However, /w/, like HI also devoices in syllable final position, a fact which
lead to the use of the digraph <uh>, as for example in the name of the last Aztec
emperor, Cuauhtemoc. In this case the overdifferentiated orthography was widely
used. The writing of these voiceless allophones cannot be attributed to Spanish
influence. Rather it simply reflects the salience of these pronunciations which
sounded somehow different and so attempts were made to represent these phonetic
impressions irrespective of their functional status.
Another interesting case of overdifferentiation is that provided by the Jesuit
analysis of vowel length and the glottal stop in Nahuatl. The Jesuit Antonio del
Rincón makes a distinction between long vowels in final syllables, which he
proposes writing with a grave accent 0), and long vowels in other syllables, which
he would mark with an accute accent ('): "An acute accent is that which, besides
lengthening the syllable, adds a tone which raises the pronunciation with a high-
pitched sound [...] a grave accent is that which, besides lengthening the syllable,
adds a grave sound which lowers the tone" ("Accento agudo es el que fuera de
alargar la syllaba le añade vn tono que leuanta la pronunciacion con sonido agudo
[...] accento graue es el que fuera de alargar la syllaba, añade vn sonido graue con
que la abaja el tono", 1595:63v [cf. 1967[1595]:61]). This suggested orthographic
Copyright © 2005. John Benjamins Publishing Company. All rights reserved.

practice appears to respond to a phonetic difference in the pitch of long vowels

according to their position in the word, but not to a contrastive opposition.
When the Jesuit Horacio Carochi published his adaptation and expansion of
Rincon's grammar in 1645, he discontinued the orthographic distinction between
these two types of long vowels, using a macron for both, though he agreed that
"the tone of the voice lowers slightly on final long vowels" ("baxa vn poco el tono
de la voz en las finales largas", 1983[1645]:56v). In other words, he treated the
difference as allophonic. On the other hand, he introduced an orthographic
distinction between final ( ̂) and non-final (´) glottal stop, which also seems to
refer to a non-contrastive allophonic difference:

La C) es señal de la pronunciacion que suelen llamar saltillo, por que la vocal sobre que cae
este accento se pronuncia como con salto, è singulto, o reparo, y suspension [...] Del
accento (ˆ) vsaremos solamente en las vltimas vocales de todos los plurales de verbos, y
nombres, que acabaren en vocal, quando no se pronunciare immediatamente otra diccion
[...] Y esta pronunciacion tiene lugar quando tales nombres, verbos, advervios &c.
terminaren el periodo; y no se pronunciare inmediatamente otra <s>yllaba ni diccion. Por
que en tal caso [...] aquella tal vocal final tendra infaliblemente saltillo (Carochi
1983[1645]:2r)

[The grave accent f ) is a sign for the pronunciation that they are accustomed to call saltillo,
because the vowel upon which it falls is pronounced with a jump or hop or rest or
suspension [...] We will use the circumflex accent (ˆ) only on the final vowels of all the
plural forms of verbs and nouns which end in a vowel, when there is no immediately
following word [...] And this pronunciation applies when such nouns, verbs, adverbs etc.
end the period, and no other syllable or word is pronounced immediately afterwards,
because in such a case [...] the final vowel will unfailingly have a saltillo]
Notice that the final glottal stop becomes non-final if followed by another word
in the same phonological phrase.
Rincon not only distinguished between long and short vowels, but also
distinguished vowels of intermediate length which are neither long nor short
("syllaba ygual, que ni es larga ni es breue" [f. 63r]) and which he proposed
writing with a circumflex accent ("accento moderado" [f. 63v]). Carochi admits
that there are phonetically intermediate vowels, but treats them as variants of short
vowels, which he represented with an acute accent:
Por remate deste primer Capitulo, advierto, que si bien todas las syllabas no finales, sino
iniciales e intermedias, ö tienen saltillo, ö sin el son largas, o breues: con todo esto quando
vn bocablo es polisyllabo y ninguna de sus syllabas es larga parecen todas medias, ni
Copyright © 2005. John Benjamins Publishing Company. All rights reserved.

breues ni largas [...] Pero no se puede negar que ay algunas syllabas que parecen medias
entre largas, y breues; y tales son ordinario las que despues de su vocal tienen dos
consonantes que en latin se dice ser largas positione (Carochi 1983[1645]:3r-v)

[To conclude this first chapter, let it be noted that although all syllables which are not final,
but rather initial or medial, either have a glottal stop or, being without a glottal stop, are
long or short: nonetheless, when a word is polysyllabic and none of its syllables is long,
they all appear to be of intermediate length, neither short nor long [...] But it cannot be
denied that there are some syllables which appear to be intermediate between long and
short, and such syllables are ordinarily those that have two consonants after their vowel,
which in Latin are said to be long by position.]

Rincón (1595) Carochi (1645)

long vowel (final)
long vowel (non-final)
moderate vowel (=short) ˆ
short vowel (not marked)
glottal stop (final)
glottal stop (non-final)
Table 4: Prosodic overdifferentiation in Rincón and in Carochi
Purepecha offers two more cases of overdifferentiation which are sufficiently
complex that they continue to characterize many modern treatments of that
language. As mentioned earlier (§3.1.1), this language has an opposition between
aspirated and non-aspirated stops. In initial position, aspirated stops are realized
with aspiration accompanying the release of the stop, but between vowels, they are
realized with preaspiration. Monzón (2001) has shown how these two types of
allophones are distinguished orthographically by Lagunas and they continue to be
so distinguished in many modern representations of the language, such as that in
Medina Pérez and Alveano Hernandez (2000).
Following nasals and before a vowel, the opposition between aspirated and
unaspirated stops is realized in a different way. Aspirated stops are pronounced
like simple voiceless stops, whereas unaspirated stops are pronounced like voiced
stops.50 Orthographically, Lagunas writes the voiced allophones of the simple stops
as , <d>, <g>, and notes that these letters do not occur in initial position (cf.
§7); however, allophonic voicing is not indicated in the case of the affricates -
<tz>, <ts>, <ch> - which are also phonetically voiced after nasals. Even today,
authors like Medina Pérez and Alveano Hernandez (2000) continue to distinguish
Copyright © 2005. John Benjamins Publishing Company. All rights reserved.

the voiced stop allophones in their orthography.

6.4 Displacement
In some cases, a phonological contrast was analyzed in terms of a secondary
conditioned feature rather than the primary conditioning feature. That is, the
graphic representation of the opposition is displaced from its primary locus to a

50
This produces a classic case of phonemic overlapping, but without loss of the opposition, of the
type Jakobson, Fant and Halle (1963[1952]:5-6) discuss for Danish. According to these authors, the
contrast [Ch] vs. [ ] / # corresponds to the contrast [ ] vs. [ ] / N V. Thus, | ] has two
different phonological values depending on its position.

secondary one. Such was the case, for example, in the Dominican Juan de
Córdova's Zapotec orthography (Smith Stark 2003). Zapotec has a contrast
between tense and lax consonants which was apparently quite difficult for a
Spanish speaker to hear. Stressed vowels are phonetically short before a tense
consonant, as in the following examples.52
a. quéte "baxo [low]" °geté' "low"
b. níça "agua [water]" °nisa "water"
c. quicha "cabello [hair]" °gica "hair"
In these cases, Cordova usually writes a single stressed vowel and always uses
a non-lax consonantal grapheme for the tense consonant. Compare the following
examples which have a lax consonant following the stressed vowel, an environ
ment where the stressed vowel is also redundantly lengthened.
a. peezo "abeja [bee]" °bizu "bee, wasp"
b. yóogi "arena [sand]" °yuzi "sand"
c. quèeza "picietl [tobacco]" °geza "cigarette"
In these cases, the medial consonants are unambiguously written as lax.
However, the stressed vowel is also written double because of the phonetic
lengthening. This could be considered a case of overdifferentiation. However,
there are also cases such as the following where the tension of the post-tonic
consonant is not marked unambiguously53 and the doubling of the stressed vowel
becomes the sole indication of the fact that the following consonant is lax.
a. piáache "ciruela [hog plum]" "hog plum"
b. xoopa "mayz grano [grain of corn]" °zubá' "corn"
c. quijti "piel [skin]" °gidi "skin"
Copyright © 2005. John Benjamins Publishing Company. All rights reserved.

This is an example of displacement. The nature of the following consonant is

marked by doubling the preceding vowel.

51
This is a well-known phenomenon in the Cyrillic representation of Russian, where special vowel
graphemes are used to mark the presence or absence of palatalization on the preceding consonant.
52
I also give the cognate forms in Juchitan Zapotee (Pickett et al 1979[1959]) to help identify the
probable phonological representation of the forms written by Cordova. I modify Pickett et al.'s
orthography, using standard Americanist symbols for voiced segments to represent lax consonants,
a degree sign at the beginning of the stressed syllable, and an acute accent for high tone. Vowels
without accents have a low tone.
53
The graphemes <p, t, c/qu> are ambiguous since they are used by Córdova for tense as well as
the corresponding lax consonants.

7. Phonotactics
In addition to establishing the inventory of letters (i e., sounds) of a language,
some information is at times provided about the combinations of those letters
which are permitted. For example, the Franciscan Juan Baptista de Lagunas notes,
after presenting the sounds of Purepecha, that "They have all of them at the
beginning and in the middle of words, except for b, d, f, g, i, r, which do not have
words that begin with them, and they are totally lacking ƒ and l." ("Tienen las
todas en principio y medio de diction, sacando B, D, F, G, I, R. que no tienen
vocablos que comiencen en ellas: y totalmente carescen de F, L" 1987[1574],
Arte:l).
It should really come as no surprise that the friars were attentive to possible
sequences of sounds and to syllable structure. Quintilian (c.30-c.l00) recom
mended that after teaching the names and shapes of the letters, the teacher should
then teach the syllables of the language: "As regards syllables, no short cut is
possible: they must all be learnt" ("Syllabis nullum compendium est; perdiscendae
omnes nec", book 1, ch. 1, 30, in Quintilianus 1920[ca. 90], I:35). This viewpoint
seems to be the inspiration of Lagunas' predecessor, Friar Maturini Gilberti, who
presents what he calls the ba be (1987[1558]:10r-11v), that is, a list of 92 CV and
2 CVC syllables in Purepecha, with a word illustrating each one. Unfortunately,
other authors did not take Quintilian's dictum so seriously.

8. Argumentation
The colonial grammars are in general of a practical nature, meant to aid others
in learning and using the Indian languages of New Spain. As a result, the analyses
are generally presented as given facts, with little attempt at argumentation or
Copyright © 2005. John Benjamins Publishing Company. All rights reserved.

justification. There are, however, some interesting exceptions.

Particularly noteworthy is the use of minimal pairs to substantiate the existence
of certain contrasts. For example, the Franciscan friar Maturino Gilberti (1987
[1558]) uses minimal pairs to establish the distinctiveness of stress placement, of
the contrast between <tz> and <ç>, of the contrast between <ts> and <tz>, and of
the distinctiveness of preaspiration, as is illustrated in the following examples.
a. contrastive stress véquanhaxeti "cosa linda [pretty thing]"
vequanhaxeti "cosa triste [sad thing]" (f. 9r)
tzahcamani "blanquear alguna cosa [to whiten some-thing]"
çahcamani "enxugarse el agua [to wipe off the water]" (f. 7r)

tsitsis "abispa [wasp]"

tzitzis "hermoso [attractive]" (f. 7v)
d. VhCV ≠ VCV yehcamani "mentir [to prevaricate]"
[yecamani] "echar algo enel agua
[to throw some-thing into the water]" (f. 109v)
Gilberti's successor, Juan Baptista de Lagunas, is equally attentive to the utility
of minimal pairs, using them to establish the contrasts between aspirated and non-
aspirated bilabial, velar and labiovelar stops and between the two rhotics (1983
[1574]:134-136).
a. [p]≠[ph] piràni "el alcacer en porreta antes del espigar, o echar la
caña: hora sea el del trigo ora el de mayz [the
green shoot of grain which is absolutely naked,
before bearing spikes, or producing the stalk:
whether it be of wheat or of corn]"
phiràni "tomar, o coger a alguno o qualquiera cosa [to
take or grab someone or anything]"
b. [k] ≠ [kh] càni "mucho [a lot]"
kàni "qualquier pierna de manta o sauana, o las hojas
del mayz [any uneven edge of a piece of cloth or a
sheet, or corn leaves]"
c. [kw] ≠ [kwh] cuìni "paxaro [bird]"
kuuini "dormir [to sleep]"
d. [r] ≠ [r] caràmeni "ser corrosiua qualquier cosa, o que con su leche,
o çumo causa entre comezon y escozimiento:
como lo hazen los higos por madurar, y las demas
cosas semejantes [to be corrosive of anything
whatsoever, or that whose sap or juice causes
Copyright © 2005. John Benjamins Publishing Company. All rights reserved.

something like itching and irritation: like figs

about to ripen, and all other similar things]"
carhàmeni "meterse en el agua la propia cosa, id est. yrse
hasta el fondon sumiendo poco a poco [to enter
the water the thing itself, that is, to go to the
bottom sinking little by little]"
Minimal, or near minimal, pairs are also used to good effect by the Jesuit priest
Antonio del Rincón (1595). His Nahuatl grammar includes 41 pairs or triplets
which are distinguished by vowel length or the placement of a glottal stop (ff. 74-
78), such as the following:

Again stands here for an apico-alveolar s.

a. chichi largo en la primera. mamar. [long on the first, "to suck"]

breue en la primera. es perro. [short on the first, it is "dog"]
salto en la primera surzir alguna cosa, y tambien es la saliua
[jump on the first "to darn something", and also it is "saliva"]
b. iecoa largo en la primera, lleuar al cabo vna cosa [long on the first, "to
finish something"]
breue en la primera, probar y si tiene accusatiuo de persona es
verbo inhonesto [short on the first, "to try", and if it has a per
sonal accusative, it is an indecent verb (meaning "to have sex
with")]
c. tepiltzin largo en la .1. hijo de alguno [long on the first, "son of someone"]
breue en la .1. verbum turpe [short on the first, an obscene word55]
Horacio Carochi, Rincón's successor as the Jesuit authority on Nahuatl, also
includes this list in modified form, with 58 sets of contrasting items, including 38
of those given by Rincón and 20 new ones (1983[1645]:126v-129r; cf. Smith Stark
2000 "Rincón").
A second type of argumentation has to do with the identification of segments as
units or sequences of units: the vexing "one phoneme or two" problem (cf.
Martinet 1939). For example, Carochi notes that verbs regularly form their
preterite by eliding the final stem vowel: téma "to bathe" becomes tén in the
preterite.56 However, if the elision of the final vowel would produce a stem which
ends in two consonants, then the final vowel is not lost and a suffix -c is added to
form the preterite: itta "see" becomes itta-c. Carochi (1983[1645]:31v) observes
that the verb nōtza "call" forms its preterite like the first group, not the second:
"You will say that nötza "call" has two consonants, and nonetheless loses the last
a. I respond that it only has one consonant, that in the Spanish alphabet it is
Copyright © 2005. John Benjamins Publishing Company. All rights reserved.

55
This word refers to the female genitals. Pronouncing this pair of words properly was particularly
important since the first, tepiltzin, was used to refer to Jesus when talking about 'the father, the son
and the holy ghost'. Notice the use of Latin to disguise references to certain indelicate topics, a
technique also noted by García-Medall 2004 and first brought to my attention some 30 years ago by
Michael Silverstein (reiterated by correspondence in 2003), who has commented on the "gate
keeping protective quarantine" (and I would add, euphemistic) function of using languages
unintelligible to the uninitiated for translating possibly offensive language. He points out, for
example, that the Loeb Classical Library translates obscene passages in Greek authors into Latin
rather than English; that such passages in Latin authors are translated into Italian; and that Sapir
sometimes reverted to Latin to translate obscene expressions, as in his publication of Takelma texts
in 1909.
56
An /m/ regularly becomes [n] when left in final position in Nahuatl.

represented by two, but the Hebrew language has it, and it is called tsade."
("Diras, que nōtza llamar, tiene dos consonantes, y con todo esto pierde la vltima
a. Respondo que no tiene mas de vna consonante, que en el alphabeto castellano se
suple con dos, pero tienela la lengua hebrea, y se llama tsade."). Even someone as
lucid as Carochi, though, feels the necessity of pointing out that this sound can be
represented by a single letter in Hebrew ( ), in order to prove that it can indeed be
considered to be a single unit.
Carochi derives a similar argument from the fact that the verb icuci "cook" has
the preterite form icucic: "Icuci "to cook" or "to ripen" has the preterite oicucic,
because, even though they write this verb this way, the first c is pronounced after
the u, and so comes to have two consonants in the penultimate syllable." ("Icuci,
cocerse, ô sazonarse algo, hace oicucic, porque aunque lo escriuen assi este verbo,
pero su primera c, se pronuncia mas despues de la u, y assi viene à tener dos
consonantes en la penultima", 1983[1645]:32r). We would currently say that this
verb stem is /ikwsi/, with a consonant cluster before the final vowel. Once again,
Carochi apparenly understands the phonological situation correctly, though he
feels obliged to finagle a consonant cluster rather than clearly recognizing cu as a
consonantal digraph.
A third type of argumentation has to do with whether semivowels should be
analyzed as consonants or as vowels. Olmos, as noted earlier (§6.3), points out that
there is little difference between o and u in Nahuatl. However he recommends
using u when it comes between two vowels, as in nitlaalaua "to slip", for the
following reason: "But it seems to me that when o should happen to be between
two vowels, it should best be pronounced and written with u rather than o, and the
reason that I say this is that in such words, Mexican and Texcocan women
Copyright © 2005. John Benjamins Publishing Company. All rights reserved.

pronounce consonantal v, which they would not do if it were an o." ("Pero

pareceme que quando la o estuuiere entre dos vocales se pronunciara y escriuira
mejor con u que con o, y la razon que a esto me mueue es que, en las tales
dicciones, las mugeres mexicanas y tetzcucanas pronuncian v consonante, lo qual
no harían si fuesse o", 1972[1547]: 198). That is, when the represents the
semivowel /w/ rather than the vowel /o/, it is pronounced like a fricative [ß] by
Mexican and Texcocan women.
The semivowel /w/ is often not written in Nahuatl when it occurs between o and
a (/ o a). However, the Jesuit Antonio del Rincón notes that in the preterite of
verbs like poa "count" and çoa "unfold", the stem final vowel is deleted and uh is
added to the stem, producing pouh and çouh. He suggests that this is "because

some write them with a u before the a." ("porque algunos los escriuen con, u. antes
de la, a.", 1967[1595], f. 18v). That is, some people write poua and çoua,
suggesting that there may be a /w/ which will be left in final position, and
therefore devoiced, when the final vowel is deleted to form the preterite stem.
Currently we would say that such verbs are underlyingly /po:wa/ and /so:wa/, and
that the /w/ that appears in the preterite is part of the stem which is left in final
position when the final vowel is deleted.

9. Phonological processes
The last aspect of phonological analysis in the colonial grammars which I
would like to mention is the notion of phonological process. The colonial
grammarians were familiar with a variety of specialized terms for talking about
phonological processes of various sorts: syncope, aphaeresis, apocope, synaeresis,
synaloepha, prothesis, paragoge, and epenthesis are all mentioned by at least one
author (Smith Stark 2000b). It is also quite common to find discussion of what was
called the addition, loss and mutation or conversion of letters. That is, an item and
process terminology was usual for describing the phonological relationships
among linguistic forms.57
In Nahuatl, for instance, Olmos (1972[1547]: 199-202) recognizes a series of
phonological processes, which we might summarize as follows:
a. n→m / V (usually)
b. t → Ø / 1_1
c. n → Ø / c, tz, u, x, consonantal y
d. ch, x, z → Ø / c, tç, ch, x
The first process refers to cases where an <n> changes to an <m> before a
Copyright © 2005. John Benjamins Publishing Company. All rights reserved.

vowel: "when n is placed before the vowels a, e, i, o, it usually becomes m"

("quando la n se pone antes destas vocales a, e, i, o, por la mayor parte se buelue
en m" [Olmos 1972(1547):201]). We would now say that the process goes in the
other direction: an m is maintained before a vowel but assimilates to a following
consonant in point of articulation and becomes n word finally, since in this form
there are no exceptions. Olmos probably chose the opposite direction of change
57
Hockett (1954:211) notes that IP terminology was used by Boas and his students, and seems to
consider it to be a carry over from processual descriptions of historical change. However, as the
terms cited here illustrate, it is already an integral part of traditional synchronic grammatical des
criptions such as those by New Spain grammarians, well before the advent of diachronic linguis
tics.

because he was influenced by the frequency of occurence of the forms. For

example, the third person plural possessive prefix is /i:n-/ before stems which
begin with any consonant, but it is /i:m-/ before stems which begin with a vowel.
The frequency of /i:n-/ lead Olmos to take it as the basic form of this prefix, and so
he conceived of the change in terms of the /n/ changing to /m/ instead of the other
way around. The formulation of this process which is currently considered to be
correct has the negative consequence, therefore, of claiming that certain very
common grammatical morphemes which usually end in /n/ actually come from a
form with an underlying /m/.
The second process is formulated by Olmos in the following way: "When / is
followed by tl, the t must be lost" ("Quando despues de la / se sigue tl se ha de
perder la t.", p. 201) This statement, which is essentially correct, ignores the fact
that the t forms part of the affricate A/, much as would currently be done in
autosegmental treatments of affricates.
Olmos describes the third process as follows: "when n is followed by c, tz, u, x,
or consonantal y, it is totally lost" ("quando despues de la n se sigue c, tz, u, x, y
consonante se pierde la n del todo", p. 201). This formulation is only partially
correct. For one thing, it does not make it clear that the context consists of
members of two natural classes, the sibilants /s, š, ¢, č/ and the semivowels /w, y/.
And second, it fails to note that the /n/ is not lost altogether when followed by a
sibilant, but rather assimilates to the following segment. Carochi clarifies things
somewhat, claiming that the /n/ assimilates to a following /s/ or /s/, creating a
double fricative: "Before a ç with a cedilla, n is not pronounced. Rather, the ç is
pronounced as if it were double; and for this, some people write a z instead of an n,
Copyright © 2005. John Benjamins Publishing Company. All rights reserved.

[...] Neither is an n pronounced before an x; rather, the x is pronounced with more

force, as if it were double." ("Antes de la ç. cerilla no se pronuncia la n. supiese
con pronunciar la ç. como si fueran dos; y para esto escriuen algunos la z. en lugar
de la n. [...] Tampoco se pronuncia la n. antes de la x. sino que la x. se pronuncia
con mas fuerça como si fueran dos.", 1983[1645]:lv). Andrews (1975:11) agrees
with Carochi in limiting this process to fricative sibilants; however, Olmos also
includes an affricate, /¢/, which leads one to wonder whether the other affricated
sibilant, /c/, should not also be included. Andrews also points out that in the case
of the glides, /w/ and /y/, the nasal simply disappears without compensatory
lengthening, although some nasalization might remain in the case of /w/. This last
observation is apparently based on Carochi, who says that "The same n, placed
before consonantal v, is barely felt, if it is there at all" ("La mesma n. antepuesta à

la v. consonante apenas se siente si la <a>y", 1983[1645]:lv).

Olmos describes the fourth process in the following way: "it is to be noted that
when the letters ch, x, z come before any of the letters c, tç, ch, JC, the one that
precedes is lost" ("es de notar que quando estas letras ch, x, z vienen antes de
qualquiera destas letras c, tç, ch, x se perdera la que precede", p. 202). This too is
only partially correct. The sibilants assimilate totally to a following sibilant,
creating a long or double sibilant (Andrews 1975:10). Olmos' rule in a sense
recognizes the assimilation, but not the consequent doubling. He also leaves out
<tç> as one of the sibilants that can be affected by the rule. A slightly different
description of this process is given by the Jesuit Antonio del Rincón:
Donde quiera que se encuentran vna destas quatro letras [tz, ch, x, ç] con otra, o vna dellas
semejante asimisma no se an de pronunciar dos sino sola vna. [...] donde en todos estos
exemplos se pierden las letras asperas precedentes, y aunque no se pronuncia la letra aspera
gastase, empero en aquella vocal que esta junto a ella doblado tiempo como si la otra se
pronunciara (1967[1595]:67-68)

[Wherever one of these four letters [tz, ch, x, ç] is found with another, or one of them with
another like itself, one should not pronounce two, but rather only one [...] where in all these
examples the preceding asperous letters are lost, and although the asperous letter is not
pronounced, twice the time, however, is spent on that vowel which is with it, as if the other
were pronounced]
This seems to be a clear description of compensatory lengthening of the
preceding vowel, instead of the gemination of the sibilant described by Andrews.
Rincón identifies another phonological process in Nahuatl, which he describes
as follows: "Note about h: whenever a word that begins with m or p follows, it
turns into m or p with the preceding u. For example, maimmochiua, for,
Copyright © 2005. John Benjamins Publishing Company. All rights reserved.

maiuhmochiua; and of the p, for example, nicchippoloa58 instead of nicchiuh-

poloa" ("Nota de la .h. que siempre sigue diction que comiença en .m. o en .p. se
buelue .m. o p. con la .u. precedente, v.g. maimmochiua, pro, maiuhmochiua, de la
p. v.g. nicchipoloa, en lugar de nicchiuhpoloa", 1595:73v). This description is
correct, although we are today bothered by the fact that it is formulated in terms of
<h>, rather than <uh>, and we would incorporate the fact that the /m/ and /p/ form
a natural class; that is, we would currently say that /w/ totally assimilates to a
following bilabial consonant (cf. Andrews 1975:10).

58
Rincón gives this form as nicchipoloa, but it should of course be nicchippoloa. The correct form
is in fact given in Rincón's vocabulary (f. 81r[1967:76]).

In Purepecha, Gilberti recognizes that <c> is replaced by <g> when preceded

by a nasal: "Replace the c with a g, because of the n which is found in intermediate
position and which is never followed by c" ("Truecase, la, c por. g. por razon dela,
n. que se offrece en medio, ala qual nunca se sigue, c", 1987[1558]:lllv).59 He
illustrates with examples like the following: aran-guhche xaca thireni, thiren-
guhche xaca, ara-haca-cuhche thireni "we are eating".60 Here it can be seen that
the initial consonant of cuhche "we" is voiced when preceded by a nasal.
Likewise, Lagunas (1983[1574]:44) gives the personal endings of the verb as ca,
re, ti, and in the plural cuhche, tsi, tix. However, in some of the paradigms, he
gives instead ga in the first person and di in the third person. He explains: "And
that g in the first person and the d in the third person which are found in the
optative and in the subjunctive are the result of the conversion of the letter, due to
being preceded by an n." ("Y aquella G. de la primera persona: y la D. dela
tercera, enel Optatiuo y Subjunctiuo. es conuersion de letra, por le auer antecedido
la N.").
Lagunas also identifies another phonologically conditioned process. He notes
that the two suffixes for instrumental nouns, -ti and -ri, are phonologically
conditioned:
Saluo que para mas declaracion delos Nombres mecanicos, la causa desta differencia Ri.
vel. Ti. Enlas terminaciones: no es sino el Accento, vt, Caràri, Escriuano, o pintor. Mítati,
Portero. Porque el vno tiene luenga la. A penultima. Y el otro la I. antepenultima. (Lagunas
1983[1574],p. 69)

[Except that, for further information about the mechanical nouns, the cause of the
difference between the endings -ri and -ti is none other than the accent, as caràri "scribe"
Copyright © 2005. John Benjamins Publishing Company. All rights reserved.

or "painter", mìtati "doorman". Because one has the penultimate a long and the other the
antepenultimate i.]
In other words, the suffix is -ri if immediately after the stressed syllable,61 but

9
As noted above (§6.3), aspirated stops are realized as voiceless unaspirated stops when occurring
after a nasal and before a vowel. Gilberti therefore seems to be mistaken, at least as far as the
modern language is concerned, when he says that <c> can never follow <n>. However, Cristina
Monzon informs me that she has not been able to find counterexamples to Gilberti's claim either in
Gilberti 1558 or in Lagunas 1573.
60
Gilberti does not gloss these forms. Cristina Monzon informs me that the third form is not
marked for the progressive.
61
Probably due to Latin influence, Lagunas wrongly refers to syllable length rather than stress.

-ti if following an unstressed syllable.

The examples in this section, I believe, demonstrate clearly that some colonial
grammarians were able to identify and describe certain phonological processes
coherently.

10. The role of the native speakers

The many accomplishments of the European friars in describing the languages
of New Spain are impressive. And yet it is important to keep in mind that their
successes were due not only to their own skill and dedication but also to the gifted
speakers who served them as informants, consultants, interpreters, translators,
teachers, and models.
At times we can only infer the presence of native speakers indirectly. Consider,
for example, the comments which Olmos (1972[1547]: 200) makes about the h in
Nahuatl.
Pero es de notar que en todos los plurales, que no se diferencian en la boz ni pronunciacion
de sus singulares, pondremos una h, y esto no porque en la pronunciacion se señale la h,
sino solamente para denotar esta diferencia del plural al singular. Y esto se ha de entender
ansi en los verbos como en los nombres donde no ouiere otra cosa por la qual se
diferencian. Exemplo.: tlatlacoani, pecador; plural, tlatlacoanime vel tlatlacoanih; —
tlaqua, aquel come; plural, tlaquah, aquellos comen; — titlaqua, tu comes; titlaquah,
nosotros comemos.

[But it is to be noted that in all plurals that are not different from the singular either in the
word nor in the pronunciation, we will place an h. We don't do this because there is an h in
the pronunciation, but rather solely to denote the difference between singular and plural.
And it should be so understood with verbs as well as with nouns, when there is nothing else
Copyright © 2005. John Benjamins Publishing Company. All rights reserved.

to distinguish them. For example, tlatlacoani "sinner", plural tlatlacoanime or tlatlacoanih;

tlaqua "that one eats", plural tlaquah "those eat"; titlaqua "you eat (2sg)", titlaquah "we
eat".]
Olmos seems to claim that the h is used solely as a diacritic feature in order to
distinguish singular from plural, which otherwise would be identical. The funny
thing is that there is a phonological difference between singular and plural forms

62
This process is somewhat reminiscent of Verner's law, which states that the position of stress in
Proto-Indoeuropean was a crucial factor in determining whether voiceless stops (p, t, k) became
voiced stops (b, d, g) (when neither initial nor immediatley preceded by the accented syllable, nor
second in a voiceless cluster) or became voiceless fricatives (when initial or immediately preceded
by the accented syllable) in Germanic (Anttila 1972:66-67).

and it consists precisely of the presence of a glottal stop (or an /h/) at the end of the
plural forms. If Olmos didn't hear it, then how is it that he wrote it? It seems to me
likely that the native speakers themselves who wrote their language served as his
model. They wrote h because they heard it and pronounced it; Olmos simply
followed suit.
It really should come as no surprise that the Europeans had problems pro
nouncing and hearing the Indian languages properly. After all, they usually came
to New Spain as adults when they began to learn one or more Indian languages. It
seems unlikely that they could learn to speak them with native speaker
proficiency. Rather they probably depended on native speakers with whom they
could check their work and make sure they were using the language properly. At
times the presence of such advisors is overtly acknowledged, though they often
remain anonymous. For example, friars Domingo Guigelmo and Juan de
Villalobos say that, in order to evaluate Cordova's Vocabvlario, "we called
together four Indian interpreters, native Zapotecs, who seemed to us to be the most
able and appropriate that there could be" ("llamamos y juntamos quatro Indios
lenguas, naturales çapotecas los mas abiles que para ello nos parecio y pudieron
ser auidos", Cordova 1578b:iiiv).
Similarly, the Dominican friar Francisco Alvarado, author of a 16th century
Mixtec dictionary, notes in his prologue that he used material that his predecessors
had gathered, but that he himself had to go over it all with care. However, he gives
principal credit to the Mixtees themselves: "the very Indians who are the best
teachers and who, for that task, were and have been the authors" ("los mismos
Indios que son los mejores maestros que para esto eran, y an sido los autores",
Alvarado 1593, "Prologo al lector", 6v).
Copyright © 2005. John Benjamins Publishing Company. All rights reserved.

In a few cases we actually have the names of individuals who helped the friars.
For example, one of the best informants for Andrés de Olmos "fué un indio
principal viejo de Texcuco llamado D. Andrés" ("was an aged Indian lord from
Texcoco named Sir Andrew", Mendieta 1945[1596]: 158). Hernandez de León-
Portilla (1996) has written about Hernando de Ribas (d. [Link].1597), a Texcocan
who helped various Franciscans: Alonso de Molina with his Nahuatl vocabulary
and grammar, Juan de Gaona in his Dialogos de la paz y la tranquilidad del alma
(1582), and Juan Baptista Viseo in various of his works. Ribas was among the first
students of the Royal Holy Cross College in the convent at Santiago Tlatelolco
and, according to Juan Baptista, was "very knowledgeable in Latin, and with much
ease translated anything from Latin and from Romance into the Mexican language,

attending more to the sense than to the letter" ("muy gran latino, y que con mucha
facilidad traduzia qualquiera cosa de Latin, y de Romance en la lengua Mexicana:
atendiendo mas al sentido, que ala letra", Viseo 1606, prologue; cf. Hernandez de
León-Portilla 1996:479). Also trained by the Franciscans at Tlatelolco and in
particular associated with the extraordinary ethnographic and doctrinal work by
the Franciscan Bernardino de Sahagún were other outstanding native Nahuatl
speakers like Antonio Valeriano of Azcapotzalco (governor of the Indians of
Mexico City from 1570-1605 and brother-in-law of the historian Fernando
Alvarado Tezozómoc), Alonso Vegerano and Pedro de San Buenaventura of
Cuauhtitlan, Martín Jacobita and Andrés Leonardo of Tlatelolco, and Juan
Badiano of Xochimilco (cf. Martínez 1981, León-Portilla 1986, de la Cruz
1991 [1552]). About Antonio Valeriano (d. 1605), in particular, Juan Baptista says
that he was:
vno delos mejores latinos, y rethoricos que del [dicho Colegio de Santa Cruz] salieron [...]
fue tan gran latino, que hablaua ex tempore (aun en los vltimos años de su vejez) con tanta
propriedad, y elegancia, que parecia vn Ciceron, o Quintiliano. [...] El qual me ayudo muy
bien, assi en cosas particulares que le consulté, como en la Ethimologia, y significacion de
muchos vocablos. (Viseo 1606, prologue)

[one of those most knowledgeable in Latin and one of the best rhetoricians produced by the
Holy Cross College [...] he was so good at Latin that he spoke it extemporaneously (even in
the last years of his life) with such propriety and elegance, that he resembled a Cicero or a
Quintilian [...] He helped me very much in specific things that I asked him about, as well as
the etymology and meaning of many words]
Juan Baptista also acknowledges the aid he received from Juan Berardo of
Huejotzingo (d. 1594), Diego Adriano and Pedro de Gante of Tlatelolco (d.
Copyright © 2005. John Benjamins Publishing Company. All rights reserved.

XI. 1605), Francisco Baptista de Contreras of Cuernavaca and Estéban Bravo of

San Diego Tlailotlacan (ibid.).
Among the Jesuits, we have the name of one Don Miguel. Karttunen and
Lockhart (1987, p. 168), in their edition of the Bancroft dialogues, mention that at
the bottom of the manuscript appears the following text: "These samples of
excellent Mexican are written by Don Miguel, Father Oracio's teacher." ("Estos
retazos de excelente Megicano son escritos por Don. Miguel, Maestro. del Padre.
Oracio."). Assuming, as seems likely, that the reference here is to Horacio
Carochi, author of what is generally considered to be the most outstanding
grammar of Nahuatl produced until the modern studies by Andrews (1975) and
Launey (1986), then Don Miguel must have been a fine teacher indeed.

Martínez Baracs (1997) notes that Antonio Huitziméngari (c. 1519-1562),

governor of Michoacan from 1545 to 1562, knew Latin, Spanish, Greek, and
Hebrew, read Erasmus and Nebrija, and was a friend and collaborator of the
Franciscan friar Maturino Gilberti.
Two important collaborators of the Franciscans in Yucatan were Juan Cocom
and Gaspar Antonio Chi. Bishop Diego de Landa speaks of the first as follows:
fue hombre de gran reputación y muy sabio en sus cosas y bien sagaz y entendido en las
naturales; y fue muy familiar del autor de este libro, fray Diego de Landa, y le contó
muchas antigüedades y le mostró un libro que fue de su abuelo (Landa 1986[1560]:21)
[he was a man of great reputation and very wise in his matters and well versed and
knowledgeable in natural things; he was well known to the author of this book, Friar Diego
de Landa, and he told him many things of old and he showed him a book that belonged to
his grandfather]
In his edition of Landa's Relación de las cosas de Yucatan (1560), Tozzer
(1941:45, note 219) mentions another of Landa's informants, Gaspar Antonio Chi,
who knew Maya, Latin, Spanish and Nahuatl. He taught Maya to the friars and
worked as a translator and interpreter. He knew something of grammar and wrote
sermons for the friars.
Among the Mixtecs, Diego Osorio from the Achiutla area was an accomplished
orator on Christian doctrine who left many prayers of saints and antiphons
translated into his language. Another gifted Mixtec was Gabriel de Valdivieso
from the valley of Yanhuitlan, praised as an avid reader, and as an erudite and
prolific writer of prayers and spiritual tracts, though Burgoa did not approve of his
explanations of certain religious matters nor of his translation of Apuleius' The
Copyright © 2005. John Benjamins Publishing Company. All rights reserved.

Golden Ass (Terraciano 2001:54; Burgoa 1989[1674], I:349-350, 379).

Clearly, intelligent, highly educated, experienced and thoughtful people such as
these were instrumental in many of the achievements of the friars in the phon
ological description of the languages of New Spain.

11. Conclusion
In this paper, I have tried to present an evaluation of the accomplishments of
the phonological descriptions produced by the New Spain linguists. As a group,
and focusing specifically on their achievements rather than their failures, it is
possible to appreciate many good insights. In particular, they recognized, recorded
and to some degree described numerous types of sounds which were not found in

Spanish. They showed some ability to arrive at phonologically correct appraisals

of the phonetic data. They appreciated the contrastive function of sounds and
justified their analyses with minimal pairs. They engaged in a bit of phonological
argumentation and included phonotactic information. The work of the Franciscans
Olmos on Nahuatl, Gilberti and Lagunas on Purepecha and Carceres on Otomi,
and of the Jesuits Rincón and Carochi on Nahuatl in particular stand out for their
descriptive acumen.
Elsewhere (Smith Stark 2000b), I have argued that the period described in this
paper should be recognized as marking the birth of modern descriptive linguistics.
Do the accomplishments just presented in phonological description justify such an
appraisal? Perhaps the predominantly empirical and practical nature of the work
would suggest a negative answer. To mark the beginning of a scientific period,
some would require work done for the sake of science, work which is integrated
into a scientific community of observational data accumulation, debate, and theory
building (cf. Zimmermann 2004). However, it seems to me that the quantity and
quality of work produced in New Spain does indeed merit recognition as scientific.
It remains to establish to what extent it was accompanied by similar activity in
other parts of America and the world as a whole.

REFERENCES
A. Primary sources
Aluarado, Francisco de, et al, 1962[1593]. Vocabvlario en lengva misteca, hecho por los
padres dela Orden de Predicadores, que residen enella, y vltima mente recopilado, y
acabado por el Padre Fray Francisco de Aluarado, Vicario de Tamaçulapa, de la
misma Orden. Mexico: Pedro Balli. Reedition: Wigberto Jiménez Moreno, editor, Fray
Copyright © 2005. John Benjamins Publishing Company. All rights reserved.

Francisco de Alvarado, Vocabulario en lengua mixteca, facsimilar reproduction with a

study by Wigberto Jiménez Moreno, including an appendix with the vocabulary taken
from the Arte en lengua mixteca by Friar Antonio de los Reyes prepared by Alfonso
Caso, México, D. F.: Instituto Nacional Indigenista e Instituto Nacional de Antropología
e Historia.
Barreda, Nicolas de la. 1960[1730]. Doctrina christiana en lengva chinanteca, añadida la
explicacion de los principales mysterios de la fee. Modo de baptizar en caso de
necessidad, y de ayudar a bien morir, y methodo de administracion de sacramentos.
Mexico: Por los herederos de la viuda de Francisco Rodriguez Lupercio. Reedition:
Howard F. Cline, Nicolas de la Barreda, Doctrina christiana en lengua chinanteca
(1730), facsimilar edition and introduction by Howard F. Cline. México, D. F.: Instituto
Nacional de Antropología e Historia.
Barros, João de. 1540. Grammatica da lingua portuguesa. Lisboa: Luís Rodrigues.

Basalenque, Diego. 1975[1642]. Arte y vocabulario de la lengua matlaltzinga vuelto a la

castellana. Manuscript; the grammar (ff. lv-76v) is followed by a vocabulary (ff. 78r-
383v). Edition: María Elena Bribiesca S., paleography, with a preliminary study by
Leonardo Manrique C, Biblioteca Enciclopedica del Estado de Mexico, vol. 33,
Mexico.
Burgoa, Francisco de. 1989[1674]. Geografica descripcion de la parte septentrional, del
polo artico de la America, y nveva iglesia de las Indias occidentales, y sitio astronomico
de esta Provincia de Predicadores de Anteqvera Valle de Oaxaca: en diez y siete grados
del Tropico de Cancer: debaxo de los aspectos, y radiaciones de Planetas morales, qve
la jvndaron con virtvdes celestes, inflvyendola en Santidad, y doctrina. Mexico: Iuan
Ruyz. Reedited in 2 vols.: Mexico, D. F.: Editorial Porrúa.
Cantares mexicanos. 1994[ca.l611]. Manuscript in the National Library of Mexico.
Facsimilar edition: Jose G. Moreno de Alba y Miguel Leon-Portilla, presenters,
Cantares mexicanos, Mexico, D. F.: Universidad Nacional Autonoma de Mexico.
Capella, Martianus Minneus Felix. ca. 420. De arte grammatica. Third book of a
manuscript in nine books.
Carceres, Pedro de. 1907[ca.l580]. Arte de la lengua othomi. Ms. which survived into the
20th century in two copies, currently of unknown whereabouts. Published as "Arte de la
lengua othomi por Fr. Pedro de Carceres, del orden de San Francisco. (Siglo XVI) Lo
publica por vez primera el Dr. N. Leon", Boletín del Instituto Bibliografico Mexicano,
6:39-155.
Carochi, Horacio. 1983 [1645]. Arte de la lengva mexicana con la declaracion de los
adverbios delia. Mexico: Iuan Ruyz. Edition: Horacio Carochi, Arte de la lengua
mexicana[;] con la declaracion de los adverbios della, ed. by Miguel Leon-Portilla,
facsimilar edition of the one published by Juan Ruyz in Mexico City, 1645, with an
introductory study. Mexico: Universidad Nacional Autonoma de Mexico.
Ciudad Real, Antonio de (?). 1984[end of the 16th century(?)]. [Calepino maya de Motul.]
Ms. in the John Carter Brown Library of Brown University, Providence, Rhode Island.
Edition: Antonio de Ciudad Real, Calepino maya de Motul, 2 vols., ed. by Rene Acuña.
Copyright © 2005. John Benjamins Publishing Company. All rights reserved.

Mexico, D. F.: Universidad Nacional Autonoma de Mexico.

Cordoua, Iuan de. 1987[1578a]. Arte en lengva zapoteca. Mexico: Pedro Balli. Reedited
as Juan de Cordova, Arte del idioma zapoteco, under the direction and care of Nicolas
Leon, Morelia: Imprenta del Gobierno en la Escuela de Artes, 1886. A facsimilar edition
of the 1886 edition was published by Ediciones Toledo, Mexico, D. F.
Cordoua, Iuan de. 1987[1578b]. Vocabvlario en lengva çapoteca, hecho y recopilado por
el mvy reverendo padre Fray Iuan de Cordoua, de la orden de los Predicadores, que
reside en esta nueua España. Mexico: Pedro Charte y Antonio Ricardo. Facsimilar
editions: Wigberto Jimenez Moreno, ed., Juan de Cordova, Vocabulario castellano-
zapoteco, with an introduction and notes, Biblioteca Lingüística Mexicana, 1, Mexico:
Instituto Nacional de Antropología e Historia, 1942; Mexico, D. F.: Ediciones Toledo.
Cruz, Martín de la. 1991 [1552]. Libellus de medicinalibus indorum herbis. Aztec
manuscript from 1552 according to a Latin translation by Juan Badiano. 2 vols. I.

Facsimile; II. Spanish version with studies and commentaries by diverse authors.
Transcription and translation from Latin by angel María Garibay K. 2a ed. Mexico:
Fondo de Cultura Economica e Instituto Mexicano de Seguro Social. (la ed., Mexico:
Instituto Mexicano de Seguro Social, 1964.)
Gilberti, Maturino. 1987[1558]. Arte de la lengua de Michuacan. [Mexico]: Iuan Pablos.
Facsimilar edition: Maturino Gilberti, Arte de la lengua de Michuacan (1558), ed. by J.
Benedict Warren, who prepared an historical introduction, documentary appendix and
was responsible for the photographic preparation of the text, Morelia, Michoacan,
Mexico: Fimax Publicistas Editores.
Guzmán, Pantaleon de. 1984[1704]. Compendio de nombres en lengva cakchiqvel. Ms. in
the John Carter Brown Library. Edition: Pantaleon de Guzman, Compendio de nombres
en lengva cakchiqvel, ed. by Rene Acuna. Mexico, D. F.: Universidad Nacional
Autonoma de Mexico.
Lagunas, Iuan Baptista [Bravo] de. 2002[1574]. Arte y dictionario: con otras obras, en
lengua michuacana, Mexico: Pedro Balli. Facsimilar edition: Arte y dictionario con
otras obras en lengua michuacana, composed by Fray Juan Baptista de Lagunas, ed. by
J. Benedict Warren, who prepared the historical introduction with documentary appendix
and was responsible for the photographic preparation of text, Morelia, Michoacan,
Mexico: Fimax Publicistas Editores, 1983. The grammar was edited separately as Iuan
Baptista de Lagunas, Arte en lengua michuacana, Agustín Jacinto Zavala, transcription,
Benedict Warren, revision, Zamora, Michoacan: El Colegio de Michoacan &
Fideicomiso Teixidor.
Landa, Diego de. 1959[1560]. Relacion de las cosas de Yucatan. Editions: with an
"Introduction" by angel Ma. Garibay K., with an appendix with various important
documents and a letter by the author, Editorial Porrúa, S. A., Mexico, D. F. (13a edicion,
1986); Tozzer 1978[1941] in "secondary sources".
Mendieta, Geronimo de. 1870[1596]. Historia eclesiastica indiana, ed. by Joaquín García
Icazbalceta. Mexico, D. F.: Antigua Librería. Reedited in 4 volumes, Mexico, D. F.:
Editorial Salvador Chavez Hayhoe, 1945.
Copyright © 2005. John Benjamins Publishing Company. All rights reserved.

Molina, Alonso de. 1998[1571a]. Arte de la lengua mexicana y castellana. Mexico: Pedro
Ocharte. A second edition, "corregida, enmendada y añadida, mas copiosa y clara que la
primera", was published in 1576, Mexico: Pedro Balli. Facsimile of the edition of 1571
on CD: Hernandez de Leon-Portilla 1998 in "secondary sources".
Molina, Alonso de. 1970[1571b]. Vocabvlario en lengva castellana y mexicana[;]
Vocabvlario en lengva mexicana y castellana (augmented and corrected version of the
first edition of 1555). Mexico: Antonio de Spinosa. Reedition "facsimilaria" by Julio
Platzmann, Leipzig, 1880. A facsimile of this edition, with a preliminary study by
Miguel Leon-Portilla, was published as Vocabulario en lengua castellana y mexicana y
mexicana y castellana, Editorial Porrua, Mexico, 1970 (2a ed., 1977).
Nebrija, Antonio de. 1481. Introductiones latinae. Salamanca. The second edition, Zamo
ra, c. 1492, includes a Spanish translation of the Latin. The last edition supervised by the
author: Introductiones in latinam gramaticem, Alcala: Arnao Guillén de Brocar, 1523.

Nebrija, Antonio de. 1984[1492]. ...Comiença la gramatica que nueva mente hizo el
maestro Antonio de lebrixa sobre la lengua castellana ..., Salamanca. Reeditions:
Antonio de Nebrija, Gramatica castellana ed. by Pascual Galindo Romeo & Luis Ortiz
Muñoz, 2 vols., Madrid: Edicion de la Junta del Centenario, 1946; Antonio de Nebrija,
Gramatica de la lengua castellana, ed. by Antonio Quilis, Madrid: Editorial Nacional.
Nebrissensi, Aelio Antonio. 2001[1495(?)]. Dictionarium ex hispaniensi in latinum
sermonem. Salamanca. A facsimile is included in Real Academia Española, Nuevo
tesoro lexicografico de la lengua española, DVD, Madrid, España: Espasa Calpe, S. A.
Nebrija, Elio Antonio. 1977[1517]. Reglas de orthographia en la lengua castellana.
Alcala de Henares. Reedition: Antonio de Nebrija, Reglas de orthographia en lengua
castellana, ed. by Antonio Quilis, Bogota, Colombia: Instituto Caro y Cuervo.
Neve, y Molina, D. Luis de. 1975[1767]. Reglas de orthographia, diccionario, y arte del
idioma othomi, breve instruccion para los principiantes. Mexico: Imprenta de la
Bibliotheca Mexicana. Facsimile: Luis de Neve y Molina, Reglas de ortografía,
diccionario y arte del idioma otomi, prepared by Mario Colín, Mexico, D.F.: Biblioteca
Enciclopedica del Estado de Mexico.
Oliveyra, Fernão d'. 1536. Grammatica da lingoagem portuguesa. Lisboa: Germão
Gallardo.
Olmos, Andres de. 1972[1547]. Arte de la lengua mexicana. There are six surviving
manuscript copies. The first published version appeared as Grammaire de la langue
nahuatl ou mexicaine, composee, en 1547, par le franciscain Andre de Olmos, et publiee
avec notes, eclaircissements, etc. par Remi Simeon, Paris, France: Imprimerie Nationale,
1875. A facsimilar reedition of this one appeared in 1972 as Andres de Olmos, Arte para
aprender la lengva mexicana, with prologue and Spanish translation of the introduction
for this edition by Miguel Leon-Portilla. Guadalajara, Jalisco, Mexico: Edmundo Aviña
Levy, editor.
Quintilianus, Marcus Fabius. 1920[ca. 90]. Institutio oratoria. Edition: The Institutio
oratoria of Quintilian. English translation by H. E. Butler, 4 vols., London: William
Heinemann; New York: G. P. Putnam's Sons.
Copyright © 2005. John Benjamins Publishing Company. All rights reserved.

Reyes, Antonio de los. 1976[1593]. Arte en lengva mixteca. Mexico: Pedro Balli.
Reeditions: by H. de Charencey in Actes de la Societe Philologique, 18:1-96, 1889 and
as a book, Paris: Klincksieck, 1890; facsimilar edition of that of 1890, Nashville,
Tennessee: Vanderbilt Publications in Anthropology.
Rhodes, Alexandre de. 1651. Dictionarium annamiticum lusitanum, et latinum ope sacrae
congregationis de Propaganda Fide in lucem editum ab Alexandro de Rhodes. Romae.
Congregado de Propaganda Fide.
Rincon, Antonio del. 1967[1595]. Arte mexicana. Mexico: Pedro Balli. Reprinted under
the care of Antonio Peñafiel, Mexico: Oficina Tipografica de la Secretaría de Fomento,
1885. Facsimilar reedition of that of Peñafiel, Guadalajara, Jalisco, Mexico: Edmundo
Aviña Levy.
Romero (?), Eugenio. 1990[lst third of 17th century (?)]. Arte de la lengua totonaca.
Published as Arte de la lengua totonaca by unknown author, introduction, transliteration

corrected and normalized, and indices of lexical and grammatical content by Norman A.
McQuown, with a facsimile of the original manuscript and an opinion as to a possible
author by Miguel Leon-Portilla. Mexico, D. F.: Universidad Nacional Autonoma de
Mexico.
Ruggieri, Michele & Matteo Ricci. 2000[ca.l580]. Diccionário portugues-chines. Por
tuguese-Chinese dictionary. John W. Witek, S. J., editor. San Francisco, California:
Ricci Institute for Chinese-Western Cultural History, University of San Francisco;
Instituto Portugues do Oriente; National Library of Portugal.
San Buenaventura, Gabriel de. 1996[1684]. Arte de la lengva maya, compuesto por el R.
P. Fr. Gabriel de San Buenaventura Predicador, y difinidor habitual de la Provincial de
San Joseph de Yucathan del Orden de N. P. S. Francisco. Mexico: Viuda de Bernardo
Calderon. Reedition: Arte de la lengua maya de Gabriel de San Buenaventura, ed. by
Rene Acuna Sandoval, Mexico: Universidad Nacional Autonoma de Mexico.
Sousa, Gabriel Soares de. 1938[1587]. Tratado descriptivo do Brasil em 1587. 3rd edition.
Sao Paulo: Nacional/EDUSP. (1st complete edition, Lisboa: Academia Real de Ciências,
1825.)
Tapia Zenteno, Carlos de. 1985[1767]. Noticia de la lengua huasteca, [...] con
cathecismo, y doctrina Christiana. Mexico: Imprenta de la Bibliotheca Mexicana.
Facsimilar edition: Carlos de Tapia Zenteno, Paradigma apologetico y noticia de la
lengua huasteca, con vocabulario, catecismo y administracion de sacramentos, ed. by
Rene Acuña, bibliographical study and notes by Rafael Montejano y Aguinaga. Mexico,
D. F.: Universidad Nacional Autonoma de Mexico.
Urbano, Alonso. 1990[1605]. [Arte breve de la lengua otomí y Vocabulario trilingüe
español-nahuatl-otomí.] Ms in the National Library of París. Edition: Alonso Urbano,
Arte breve de la lengua otomí y vocabulario trilingüe español-nahuatl-otomí, ed. by
Rene Acuña, Mexico, D. F.: Universidad Nacional Autonoma de Mexico.
Vazquez, F. 1937-1944[1714-1717]. Cronica de la provincia del Santísimo Nombre de
Jesús de Guatemala de la Orden de Nuestro Serafico Padre san Francisco (1714-1717).
Biblioteca "Goathemala", vols. 14-17. Guatemala.
Copyright © 2005. John Benjamins Publishing Company. All rights reserved.

Varo, Francisco. 2000[1703]. Arte de la lengua mandarina. Reprinted: Francisco Varo's

grammar of the Mandarin language (1703): an English translation of 'Arte de la lengua
mandarina', translated by Joseph A. Levi & W. South Coblin; introduction by Sandra
Breitenbach, Amsterdam & Philadelphia: John Benjamins.
[Viseo,] Ioan Baptista. 1998[1606]. Sermonario en lengua mexicana. Mexico: Diego
Lopez Daualos. Reproduction: Hernandez de Leon-Portilla 1998 in "secondary sources".

B. Secondary sources
Acuña, Rene. 1990. "Introduccion". Urbano 1990[1605]:XIX-LXX in "primary sources".
Altman, Cristina. 2003. "Revisiting Linguistic Relativity". Paper delivered at the First
International Conference on Missionary Linguistics, Oslo.
Anderson, Arthur J. O., Francis Berdan & James Lockhart, eds. and translators. 1976.
Beyond the Codices!;] the Nahua View of Colonial Mexico. With a linguistic essay by

Ronald W. Langacker. Berkeley, etc.: University of California Press.

Andrews, J. Richard. 1975. Introduction to Classical Nahuatl. Austin & London:
University of Texas Press.
Anttila, Raimo. 1972. An Introduction to Historical and Comparative Linguistics. New
York: The Macmillan Company / London: Collier-Macmillan Limited.
Arias, Beatriz. 2001. "La figura de las voces: la relacion entre ortografía y sonidos en el
español de los siglos XVI y XVII". Paper to appear in Smith Stark (2001, forthcoming).
Bierhorst, John. 1985. A Nahuatl-English Dictionary and Concordance to the 'Cantares
mexicanos/ with an Analytical Transcription and Grammatical Notes. Stanford,
California: Stanford University Press.
Boas, Franz. 1917. "Introduction". International Journal of American Linguistics 1:1.1-8.
Repr. in Franz Boas Race, Language and Culture. New York: The MacMllan Company,
1940.199-210.)
Butler, H. E. 1920. Cf. Quintilianus, Marcus Fabius, in primary sources.
Canfield, Delos Lincoln. 1934. Spanish Literature in Mexican Languages as a Source for
the Study of Spanish Pronunciation. New York: Instituto de las Españas en los Estados
Unidos.
Carmack, Robert M. 1973. Quichean Civilization[;] the Ethnohistoric, Ethnographic, and
Archaeological Sources. Berkeley, etc.: University of California Press.
Carmack, Robert M. & James L. Mondloch, eds. 1983. El título de Totonicapan[;] Texto,
traduccion y comentario. Mexico, D. F.: Universidad Nacional Autonoma de Mexico.
Cline, Howard F. 1960. Cf. Barreda, Nicolas de la, 1730, in primary sources.
Cruz Perez, Víctor de la. 2001. El pensamiento de los Binnigula'sa'. Cosmovisión y
calendario, con especial referencia a los binniza. Ph.D. Thesis, Mexico: Universidad
Nacional Autonoma de Mexico (thesis defended in 2002.)
Esparza Torres, Miguel angel & Hans-Josef Niederehe. 1999. Bibliografía nebrisense[;]
las obras completas del humanista Antonio de Nebrija desde 1481 hasta nuestros días.
Amsterdam & Philadelphia: John Benjamins.
Fox, James Allan. 1978. Proto-Mayan Accent, Morpheme Structure Conditions, and Velar
Copyright © 2005. John Benjamins Publishing Company. All rights reserved.

Innovations. Ph. D. Thesis, Chicago, Illinois: University of Chicago.

García-Medall, Joaquín. 2004. "Notas de lexicografía hispano-filipina: el Bocabulario de
lengua bisaya, hiligueyna y haraya de la isla de Panay y Sugbu y para las demas islas, de
fray Alonso Mentrida, OSA (ca. 1637)". Zwartjes & Hovdhaugen 2004.201-232.
Guzmán Betancourt, Ignacio, ed. 1996. Estudios de filología y lingüística nahuatl 1995-
1996 (= monographic number of the Revista latina de pensamiento y lenguaje 2(2B):
377-540). Guadalajara, Jalisco, Mexico: Sociedad Iberoamericana de Pensamiento y
Lenguaje.
Herbruger, Jr., Alfredo & Eduardo Díaz Barrios. 1956. Metodo para aprender a hablar,
leery escribir la lengua cakchiquel. Volume 1. Guatemala, C. A.: Tipografía Nacional.
Hernandez de Leon-Portilla, Ascension. 1996. "Hernando de Ribas, interprete de dos
mundos". Guzman Betancourt, ed., 1996.477-493.
Hernandez de Leon-Portilla, Ascension, ed. 1998. Obras clasicas sobre la lengua nahuatl.

Coleccion Clasicos Tavera, no. 16; Serie IX, Fuentes Lingüísticas Indígenas, vol. 8.
Compact disc. Madrid, España: Fundacion Historica Tavera/ Mapfre Mutualidad /
Digibis Publicaciones Digitales.
Hernandez Lopez, Pedro. 2000. Actitudes, redes de comunicacion y dialectología:
propuestas para la planeacion lingüística en la Chinantla de los dzämo'. M.A. Thesis,
Mexico, D. F.: Centro de Investigaciones y Estudios Superiores en Antropología Social
& Instituto Nacional Indigenista.
Hockett, Charles F. 1954. "Two Models of Grammatical Description". Word 10:2-3.
210-231. (Repr. in Readings in Linguistics I. The Development of Descriptive
Linguistics in America 1925-1956, Martin Joos, ed., Chicago & London: University of
Chicago Press, second printing of the fourth edition, 1967:386-399, originally published
in 1957 by the American Council of Learned Societies).
Jakobson, Roman, C. Gunnar M. Fant & Morris Halle. 1952. Preliminaries to Speech
Analysis. The Distinctive Features and Their Correlates, 5th printing: Cambridge,
Massachusetts: The M.I.T. Press, 1963.
Karttunen, Frances. 1983. An Analytical Dictionary of Nahuatl Austin, Texas: University
of Texas Press. (Reedition: Norman & London: University of Oklahoma Press, 1992.)
Karttunen, Frances & James Lockhart. 1987. The Art of Nahuatl Speech!;] the Bancroft
Dialogues. Los Angeles, California: University of California.
Lastra, Yolanda. 1992. "El vocabulario trilingüe de Fray Alonso Urbano". Scripta
Philologica in Honorem Juan M. Lope Blanch, coordinated by Elizabeth Luna Traill,
3.39-46, Mexico, D. F.: Universidad Nacional Autonoma de Mexico.
Launey, Michel. 1979. Introduction à la langue et à la litterature aztèques. Ed.
L'Harmattan. Spanish translation by Cristina Kraft. 1992. Introduccion a la lengua y a
la literatura nahuatl. Mexico, D. F.: Universidad Nacional Autonoma de Mexico.
Launey, Michel. 1986. Categories et operations dans la grammaire nahuatl. Ph.D. Thesis,
Universite de Paris-TV.
Leon-Portilla, Miguel, ed. 1986. Coloquios y doctrina cristiana con que los doce frailes
de San Francisco, enviados por el papa Adriano VI y por el emperador Carlos V,
Copyright © 2005. John Benjamins Publishing Company. All rights reserved.

convirtieron a los indios de la Nueva España. Mexico D. F.: Universidad Nacional

Autonoma de Mexico y Fundacion de Investigaciones Sociales, A. C. (based on a
manuscript from c. 1569)
Leon-Portilla, Miguel, ed. 1995. Diccionario Porrúa de historia, biografía y geografía de
Mexico. 6th edition, corrected and augmented. 4 volumes. Mexico, D. F.: Editorial
Porrúa, S. A.
Lucas Gonzalez, Rosa. 2002. Translations of the Latin texts in Lagunas 2002[1574], in
"primary sources".
Martinet, Andre. 1939. "Un ou deux phonèmes?" Acta linguistica 1:94-103. (Repr. in
Readings in Linguistics II, ed. by Eric P. Hamp, Fred W. Householder & Robert
Austerlitz. Chicago & London: The University of Chicago Press, 1966.116-123.
Martínez, Jose Luis. 1981. "Fray Bernardino de Sahagún y sus informantes indígenas.
Vida y obra". Bernardino de Sahagún. El Mexico antiguo (seleccion y reordenacion de

la Historia general de las cosas de Nueva España de fray Bernardino de Sahagún y de

los informantes indígenas) ed. by Jose Luis Martínez, ix-cii, Caracas: Biblioteca
Ayacucho.
Martínez Baracs, Rodrigo. 1997. "El Vocabulario de la lengua de Mechuacan (1559) de
fray Maturino Gilberti como fuente de informacion historica". Historia y sociedad[;]
ensayos del seminario de historia colonial de Michoacan ed. By Carlos Paredes
Martínez, 67-162. Morelia, Michoacan: Instituto de Investigaciones Historicas-
Universidad Michoacana de San Nicolas de Hidalgo.
McQuown, Norman A., ed. 1967. Linguistics. Vol. 5 of the Handbook of Middle
American Indians, Robert Wauchope, general editor. Austin, Texas: University of Texas
Press.
Medina Perez, Alberto & Jesús Alveano Hernandez. 2000. Vocabulario p'urhepecha-
español y español-p'urhepecha. Mexico, D. F.: Plaza y Valdes, S. A. de C. V.
Monzon, Cristina. 1997. Introduccion a la lengua y cultura tarascas. Valencia:
Universitat de València.
Monzon, Cristina. 2001. "LaREPRESENTACIÓNortografica de las consonantes aspiradas del
tarasco en la obra gramatical de los frailes: una reconstruccion historica". Paper to
appear in Smith Stark (2001, forthcoming).
Monzon, Cristina. 2005. "La pronunciacion tarasca de las grafías <ts, tz, ç, s> en el siglo
XVI". Paper delivered at the First International Conference on Missionary Linguistics,
Oslo, published in this volume.
Newman, Stanley. 1967. "Classical Nahuatl". Norman A. McQuown, ed. 1967.179-199.
Norton, Peter B., ed. 1995[1974]. The New Encyclopœdia Britannica, Micropœdia Ready
Reference. 15th ed., Chicago, etc.: Encyclopaedia Britannica, Inc.
Orie, Olanike Ola & Victoria R. Bricker. 2000. "Placeless and historical laryngeals in
Yucatec Maya". International Journal of American Linguistics 66:3.287-317.
Ostler, Nicholas. 2004. "The Social Roots of Missionary Linguistics". Zwartjes &
Hovdhaugen, eds. 2004.33-43.
Pellicer Ugalde, Alejandra. 1997. IM reconstruccion de la escritura por parte de niños
Copyright © 2005. John Benjamins Publishing Company. All rights reserved.

maya-hablantes: el caso específico de laREPRESENTACIÓNde las consonantes

glotalizadas. Ph. D. Thesis. Mexico, D. F.: Instituto Politecnico Nacional.
Pickett, Velma et al, compilers. 1959. Castellano-zapoteco y zapoteco-castellano[;]
dialecto del zapoteco del Istmo. Mexico, D. F.: Instituto Lingüístico de Verano en
cooperacion con la Direccion General de Asuntos Indígenas de la Secretaría de
Educacion Pública. (2nd edition, corrected and augmented, with the title Vocabulario
zapoteco del Istmo[;] castellano-zapoteco y zapoteco-castellano, 1965. 8th printing,
Mexico, D. F.: Instituto Lingüístico de Verano, 1979.)
Pike, Kenneth L. 1946. "Phonemic Pitch in Maya". International Journal of American
Linguistics 12:2.82-88.
Quilis, Antonio. 1984. "Estudio". In Nebrija 1984[1492]:7-92, in "primary sources".
Radin, Paul. 1943-1944. "Cuentos y leyendas de los zapotecos". Tlalocan, 1:1.3-30
(1943), 1:2.134-154 (1943), 1:3.194-226 (1944). Pages 3-10 were translated into Spa-

nish by Pilar Mascare Sacristan and published as "La gramatica y el diccionario

zapoteco de Juan de Cordoba", Guchachi' reza[;] Iguana rajada (Oaxaca), 13:3-9
(1982).
Rensch, Calvin R. 1989. An Etymological Dictionary of the Chinantec Languages. Dallas,
Texas: The Summer Institute of Linguistics.
Robins, Robert Henry. 1967. A Short History of Linguistics. Bloomington, Indiana:
Indiana University Press. (Second edition, 1969, London: Longman.
Rowe, John Howland. 1974. "Sixteenth and Seventeenth Century Grammars". In Dell
Hymes, ed., Studies in the History of Linguistics[;] Traditions and Paradigms, 361-379.
Bloomington & London: Indiana University Press.
Ruz, Mario Humberto, ed. (with the collaboration of Diana Birrichaga and with
appendices by Søren Wichmann and Thomas Smith). 1997. Las lenguas de Chiapas
colonial[;] manuscritos en la Biblioteca Nacional de Paris. Vol. 2, Lengua zoque.
Mexico, D. F.: Universidad Nacional Autonoma de Mexico.
Sanchez García, Daniel. 1919. "Bibliografía cachiquel". In the introduction to the edition
of Gramatica del idioma cakchiquel, written in 1748 by the Franciscan friar Carlos
Joseph Rosales, which was prepared for publication by Daniel Sanchez García, friar of
the same order, with an introduction, a Cakchiquel-Quiche-Tzutujil bibliography,
corrections, notes, a "Paralelo del cachiquel" and a "Compendio de la Doctrina cristiana
en cachiquel y castellano". Guatemala: Tipografía "San Antonio".
Smith Stark, Thomas C. 1992a. "Apuntes sobre la lexicografía novohispana". Paper
presented at Jornadas Lingüísticas, Instituto de Investigaciones Filologicas, Universidad
Nacional Autonoma de Mexico, Mexico, D. F. (Most recent version, 1995.)
Smith Stark, Thomas C. 1992b. Unpublished review of Alonso Urbano, Arte breve de la
lengua otomí y vocabulario trilingüe español-nahuatl-otomí ed. by Rene Acuña.
Mexico, D. F.: Universidad Nacional Autonoma de Mexico, 1990.
Smith Stark, Thomas C. 1995. "Apuntes sobre la evolucion de la ortografía del nahuatl".
Paper presented at Jornadas Lingüísticas, Escuela Nacional de Antropología e Historia,
Mexico, D. F.
Copyright © 2005. John Benjamins Publishing Company. All rights reserved.

Smith Stark, Thomas C. 1996. "La grafía del nahuatl de Antonio del Rincon". Guzman
Betancourt 1996.407-433.
Smith Stark, Thomas C. 1997. "La ortografía colonial y la fonología de la lengua zoque".
Appendix to Ruz 1997.345-354.
Smith Stark, Thomas C. 2000a. "Rincon y Carochi: la tradicion jesuítica de descripcion
del nahuatl". Las gramaticas misioneras de tradicion hispanica (siglos XVI-XVII), ed. by
Otto Zwartjes, 29-72. Amsterdam & Atlanta: Rodopi.
Smith Stark, Thomas C. 2000b. "La gramatica descriptiva en la Nueva España: 1521-
1645". Paper to appear in Smith Stark (2001, forthcoming).
Smith Stark, Thomas C, ed. 2001, forthcoming. Los gramaticos de Dios: las raíces de la
lingüística descriptiva en la Nueva España. Mexico, D. F.: El Colegio de Mexico.
Smith Stark, Thomas C. 2003. "La ortografía del zapoteco en el Vocabvlario de fray Juan
de Cordova". Escritura zapoteca: 2,500 años de historia, coordinated by María de los

angeles Romero Frizzi, 173-239. Mexico, D. F.: Instituto Nacional de Antropología e

Historia, y Miguel angel Porrúa.
Stark Campbell, et al. 1986. Diccionario mixteco de San Juan Colorado. Mexico, D. F.:
Instituto Lingüístico de Verano.
Suarez, Jorge A. 1983. The Mesoamerican Indian Languages. Cambridge, etc.:
Cambridge University Press.
Suarez Roca, Jose Luis. 1992. Lingüística misionera española. Oviedo: Pentalfa.
Tellez Nieto, Herendira. 2004. Acercamiento filologico al vocabulario trilingüe, atribuido
a fray Bernardino de Sahagún. Undergraduate Thesis. Mexico, D.F.: Universidad
Nacional Autonoma de Mexico.
Terraciano, Kevin. 2001. The Mixtecs of Colonial Oaxaca: Ñudzahui History, Sixteenth
Through Eighteenth Centuries. Stanford, California: Stanford University Press.
Thompson, Laurence C. 1987. A Vietnamese Reference Grammar. Mon-Khmer Studies
XIII-XIV, Stephen O'Harrow, ed., Honolulu, Hawaii: University of Hawaii Press.
Tozzer, Alfred M. 1921. A Maya Grammar[;] With Bibliography and Appraisement of the
Works Noted. Cambridge, Massachusetts: The Peabody Museum. Repr. New York:
Kraus Reprint Corporation, 1967.
Tozzer, Alfred M., ed. 1941. Landa's Relacion de las cosas de Yucatan. A Translation.
Cambridge, Massachusetts: The Peabody Museum. Repr. Millwood, N. Y.: Kraus
Reprint Co., 1978.
Whitecotton, Joseph W. & Judith Bradley Whitecotton. 1993. Vocabulario zapoteco-
castellano. Nashville, Tennessee: Vanderbilt University.
Zimmermann, Klaus. 2004. "La construccion del objeto de la historiografía de la
lingüística misionera". Zwartjes & Hovdhaugen, eds. 2004.7-32.
Zwartjes, Otto & Even Hovdhaugen, eds. 2004. Missionary Linguistics / Lingüística
misionera[;] Selected papers from the First International Conference on Missionary
Linguistics, Oslo, 13-16 March 2003. Amsterdam & Philadelphia: John Benjamins.
Copyright © 2005. John Benjamins Publishing Company. All rights reserved.

Missionary Linguistics II: Orthography & Phonology
No ratings yet
Missionary Linguistics II: Orthography & Phonology
300 pages
A Concise History of The Language Science - Koener - 1995 - Book
100% (6)
A Concise History of The Language Science - Koener - 1995 - Book
435 pages
Prescriptive vs. Descriptive Linguistics
No ratings yet
Prescriptive vs. Descriptive Linguistics
5 pages
History of Linguistics Overview
No ratings yet
History of Linguistics Overview
3 pages
4b2ac53a3ea2a005ddd7117a8a39eca4
100% (1)
4b2ac53a3ea2a005ddd7117a8a39eca4
411 pages
19th Century Historical Linguistics Overview
No ratings yet
19th Century Historical Linguistics Overview
3 pages
Traditional vs Modern Grammar Overview
No ratings yet
Traditional vs Modern Grammar Overview
10 pages
The Syllable-De Gruyter (1999)
No ratings yet
The Syllable-De Gruyter (1999)
809 pages
Overview of Modern Celtic Languages
No ratings yet
Overview of Modern Celtic Languages
30 pages
Pre-modern Linguistics: Indian & Western Insights
No ratings yet
Pre-modern Linguistics: Indian & Western Insights
2 pages
Tarascan Orthography in the 16th Century
No ratings yet
Tarascan Orthography in the 16th Century
24 pages
Saussure's General Linguistics Overview
No ratings yet
Saussure's General Linguistics Overview
17 pages
Saussure Ferdinand de Course in General Linguistics 1959
No ratings yet
Saussure Ferdinand de Course in General Linguistics 1959
9 pages
Modern Linguistics: A Historical Overview
100% (2)
Modern Linguistics: A Historical Overview
218 pages
Linguistic Traditions Through History
No ratings yet
Linguistic Traditions Through History
6 pages
Proto-Indo-European Language Insights
No ratings yet
Proto-Indo-European Language Insights
26 pages
Overview of Linguistic Anthropology
No ratings yet
Overview of Linguistic Anthropology
20 pages
Greek Verbal Network: A Probabilistic Analysis
No ratings yet
Greek Verbal Network: A Probabilistic Analysis
39 pages
The Greek Verbal Network Viewed From A Probabilistic Standpoint (Stanley E. Porter)
No ratings yet
The Greek Verbal Network Viewed From A Probabilistic Standpoint (Stanley E. Porter)
39 pages
Historical Overview of Linguistics
No ratings yet
Historical Overview of Linguistics
24 pages
Overview of Linguistic Analysis Concepts
No ratings yet
Overview of Linguistic Analysis Concepts
4 pages
Historical Overview of Comparative Linguistics
No ratings yet
Historical Overview of Comparative Linguistics
10 pages
Ten Lectures on Linguistics by Stambolieva
No ratings yet
Ten Lectures on Linguistics by Stambolieva
160 pages
Modern Linguistics vs Traditional Grammar
No ratings yet
Modern Linguistics vs Traditional Grammar
11 pages
Traditional Language Studies Overview
No ratings yet
Traditional Language Studies Overview
13 pages
(Studies in Language Companion Series 138) Gabriele Diewald, Leena Kahlas-Tarkka, Ilse Wischer-Comparative Studies in Early Germanic Languages_ With a Focus on Verbal Categories-John Benjamins Publish
100% (1)
(Studies in Language Companion Series 138) Gabriele Diewald, Leena Kahlas-Tarkka, Ilse Wischer-Comparative Studies in Early Germanic Languages_ With a Focus on Verbal Categories-John Benjamins Publish
325 pages
Grammar Evolution in the Middle Ages
No ratings yet
Grammar Evolution in the Middle Ages
6 pages
Early New English: Historical Overview
No ratings yet
Early New English: Historical Overview
27 pages
History of Linguistics
No ratings yet
History of Linguistics
13 pages
Essentials of Comparative Linguistics
No ratings yet
Essentials of Comparative Linguistics
24 pages
History of Modern Linguistics Overview
No ratings yet
History of Modern Linguistics Overview
11 pages
Type of Explanation in History by Lightfoot
No ratings yet
Type of Explanation in History by Lightfoot
21 pages
Overview of Historical Linguistics
No ratings yet
Overview of Historical Linguistics
18 pages
Evolution of European Word Classes
No ratings yet
Evolution of European Word Classes
18 pages
Introduction to Linguistic Studies
No ratings yet
Introduction to Linguistic Studies
8 pages
Introduction to Historical Linguistics
No ratings yet
Introduction to Historical Linguistics
14 pages
Gramatica Comparata A Limbilor Germanice - ID
No ratings yet
Gramatica Comparata A Limbilor Germanice - ID
140 pages
Overview of Global Linguistic Traditions
No ratings yet
Overview of Global Linguistic Traditions
21 pages
New French: Emerging Linguistic Crisis
No ratings yet
New French: Emerging Linguistic Crisis
7 pages
Historical Linguistics and Grammar Types
No ratings yet
Historical Linguistics and Grammar Types
4 pages
History of Linguistics: Greeks to Indians
No ratings yet
History of Linguistics: Greeks to Indians
3 pages
History of Language: Greeks, Romans, Indians
100% (1)
History of Language: Greeks, Romans, Indians
3 pages
Historical Syntax and Language Change
No ratings yet
Historical Syntax and Language Change
21 pages
Linguistics: Key Concepts and History
No ratings yet
Linguistics: Key Concepts and History
20 pages
Evolution of Linguistics Through Ages
No ratings yet
Evolution of Linguistics Through Ages
10 pages
Grammar Evolution During the Renaissance
No ratings yet
Grammar Evolution During the Renaissance
3 pages
Noam Chomsky - Morphophonemics of Modern Hebrew
100% (1)
Noam Chomsky - Morphophonemics of Modern Hebrew
82 pages
(Current Issues in Linguistic Theory 223) Teresa Fanego (Ed.), Javier Pérez-Guerra (Ed.), María José López-Couso (Ed.) - English Historical Syntax and Morphology_ Selected Papers from 11 ICEHL, Santia
100% (1)
(Current Issues in Linguistic Theory 223) Teresa Fanego (Ed.), Javier Pérez-Guerra (Ed.), María José López-Couso (Ed.) - English Historical Syntax and Morphology_ Selected Papers from 11 ICEHL, Santia
317 pages
Evolution of Linguistics (18th-19th Century)
No ratings yet
Evolution of Linguistics (18th-19th Century)
16 pages
A Grammar of Proto Germanic PDF
100% (2)
A Grammar of Proto Germanic PDF
118 pages
Diacronía en la Gramática Inglesa
No ratings yet
Diacronía en la Gramática Inglesa
7 pages
Overview of Descriptive Linguistics
No ratings yet
Overview of Descriptive Linguistics
9 pages
Discovery of Language Linguistic Science in The Nineteenth Century
No ratings yet
Discovery of Language Linguistic Science in The Nineteenth Century
393 pages
Overview of Greek Linguistics History
No ratings yet
Overview of Greek Linguistics History
4 pages
Formation of Colonial Yucatec Maya Language
No ratings yet
Formation of Colonial Yucatec Maya Language
23 pages
Sahagun's Research Methodology Explained
No ratings yet
Sahagun's Research Methodology Explained
24 pages
Nahua-Christian Moral Dialogue in Mexico
No ratings yet
Nahua-Christian Moral Dialogue in Mexico
286 pages
Tagalog Responses to Colonial Conversion
No ratings yet
Tagalog Responses to Colonial Conversion
27 pages
Nahua Drama: Holy Wednesday Analysis
No ratings yet
Nahua Drama: Holy Wednesday Analysis
91 pages
Primeros Memoriales: Nahuatl Paleography
No ratings yet
Primeros Memoriales: Nahuatl Paleography
26 pages
Aztec Theater: First Christian Play
No ratings yet
Aztec Theater: First Christian Play
15 pages
Nahua Gold Casting in Florentine Codex
No ratings yet
Nahua Gold Casting in Florentine Codex
17 pages
Jones Owen - Wills 2015
No ratings yet
Jones Owen - Wills 2015
16 pages
Missionary Linguistics in Mesoamerica
No ratings yet
Missionary Linguistics in Mesoamerica
23 pages
Bricker - Kaua
No ratings yet
Bricker - Kaua
136 pages
Yucatan's 17th Century Cabildos Analysis
No ratings yet
Yucatan's 17th Century Cabildos Analysis
3 pages
Kellogg - Philology Plus. New Studies in Mesoamerican Ethnohistory
No ratings yet
Kellogg - Philology Plus. New Studies in Mesoamerican Ethnohistory
7 pages
Hanks - Sanctification, Structure, and Experience in A Yucatec Ritual Event
No ratings yet
Hanks - Sanctification, Structure, and Experience in A Yucatec Ritual Event
37 pages
Interdisciplinary Sound and Listening Insights
No ratings yet
Interdisciplinary Sound and Listening Insights
16 pages
Stages of the Speaking Process
100% (1)
Stages of the Speaking Process
2 pages
Understanding Phonetics and Phonology
75% (4)
Understanding Phonetics and Phonology
68 pages
Kate Dictionary: Pacific Linguistics Se Ie C - No. 41
No ratings yet
Kate Dictionary: Pacific Linguistics Se Ie C - No. 41
552 pages
Understanding Phonetics and Sounds
No ratings yet
Understanding Phonetics and Sounds
30 pages
Understanding Phonetics and Phonology
No ratings yet
Understanding Phonetics and Phonology
33 pages
English Phonetics and Phonology Quiz
No ratings yet
English Phonetics and Phonology Quiz
10 pages
Communication for Various Purposes Module
No ratings yet
Communication for Various Purposes Module
10 pages
Voicing and Consonant Articulation
100% (1)
Voicing and Consonant Articulation
52 pages
Understanding English Pronunciation
50% (2)
Understanding English Pronunciation
58 pages
Overview of English Phonetics
No ratings yet
Overview of English Phonetics
39 pages
Organs of Speech in Phonetics
No ratings yet
Organs of Speech in Phonetics
5 pages
Introduction to Academic English Course
No ratings yet
Introduction to Academic English Course
10 pages
Speech Sound Production Basics
No ratings yet
Speech Sound Production Basics
34 pages
Idiolect Pronunciation Analysis Factors
No ratings yet
Idiolect Pronunciation Analysis Factors
5 pages
Perspectives: An Open Introduction To Cultural Anthropology: Second Edition
No ratings yet
Perspectives: An Open Introduction To Cultural Anthropology: Second Edition
28 pages
EDU410 Mid-Term Literacy Quiz Guide
No ratings yet
EDU410 Mid-Term Literacy Quiz Guide
4 pages
Standard Austrian German Phonetics Analysis
No ratings yet
Standard Austrian German Phonetics Analysis
10 pages
BBM 203: Business Communication II Guide
No ratings yet
BBM 203: Business Communication II Guide
2 pages
Understanding Places of Articulation
No ratings yet
Understanding Places of Articulation
44 pages
Phonetics I Practice Activities - Vowels - Diphthons - Consonants
No ratings yet
Phonetics I Practice Activities - Vowels - Diphthons - Consonants
6 pages
IPA Chart for English Phonetics
No ratings yet
IPA Chart for English Phonetics
4 pages
Mẫu Đề Thi Ngữ Âm Đại Học HCM
No ratings yet
Mẫu Đề Thi Ngữ Âm Đại Học HCM
4 pages
Analyzing Natural Classes in Phonology
No ratings yet
Analyzing Natural Classes in Phonology
22 pages
Understanding Speech Organs and Vowels
No ratings yet
Understanding Speech Organs and Vowels
16 pages
Understanding Phonemes in Linguistics
No ratings yet
Understanding Phonemes in Linguistics
26 pages
Phonetics Review Homework Assignment
No ratings yet
Phonetics Review Homework Assignment
7 pages
Mouth Positions for English Consonants
No ratings yet
Mouth Positions for English Consonants
4 pages
Diagram of Speech Organs in Phonetics
No ratings yet
Diagram of Speech Organs in Phonetics
3 pages
Phonology Quiz: Vowels & Consonants
No ratings yet
Phonology Quiz: Vowels & Consonants
5 pages

Phonological Insights in Colonial Mexico

Uploaded by

Phonological Insights in Colonial Mexico

Uploaded by

PHONOLOGICAL DESCRIPTION IN NEW SPAIN *

"no abasta escreuir ni pronunciar las tales diciones como

Tradicionalmente se ha afirmado que las gramáticas coloniales de lenguas ame­

merosos tipos de sonidos novedosos no existentes en el español, que lograron

of course continues to be, religious proselytization. Missionaries driven by the

2. The linguistic context: Nebrija

As partially indicated by this quote, Nebrija distinguishes clearly between the

include l, m, n, r, s, z, and perhaps consonantal i -. This distinction seems to be

3. The graphic representation of sounds in the languages of New Spain

the entry for <monarchia>.

ph = /ph/ th = /th/ tzh = /ch/ ch/qhu = /kh/ h = /h/

That is, an <h> represents a glottal fricative, the aspiration of a preceding

obstruent, or the preaspiration of a following nasal.

the stop intervocalically, which corresponds to the phonetic realization of

3.1.2 Glottalization. An important category of new sounds encountered by the

pp = [p'] tt=[t'] ttz=[ts'] cc/cqu = [k']

Doubling, doubling of the vertical stroke, barring or doubling together with

In addition to the glottalized ejectives, Otomi had a preglottalized /?d/. This

3.1.8 Interdentals. In Mixtec, Reyes (1976[1593]) and Alvarado (1962[1593]) use

representation of the allophone [tn] found in this context. A /k/ followed by a

3.3 Prosodic elements

d. yo+sána÷ndi "atrauesarseme el bocado en el garguero [for a bite of food

que entra, y por metaphora dizen de multitud de gente [for

widespread tendency to ignore tone when writing descriptions of Mexican

4. The articulatory description of sounds

4.1 Descriptions by analogy with other languages

maravilla" ("en la pronunciacion mas parece allegarse al latin que no a nuestro

4.2 Explicit articulatory descriptions

L. Lingua palatoque dulcescit.

S. Sibilum facit dentibus verberatis.

Here, by forceful pronunciation, Gilberti is describing the difference between a

4.2.2 An interdental fricative. Spanish sibilants underwent several important

4.2.3 Glottal stop. Another good example of a successful phonetic description is

Nasalization of vowels is also clearly described in Otomi, where the Franciscan

4.2.8 Retroflexion. As a further example of the relative success of some of the

articulatory descriptions achieved by the colonial linguists, consider the

4.2.10 Concluding remarks aboutarticulatorydescription. I do not mean to over­

5. The sound system, the alphabet and Quintilian's method

6. Grapheme and phoneme

6.1 Phonemes and allophones

with u, no general rule can be given.]

practice appears to respond to a phonetic difference in the pitch of long vowels

Rincón (1595) Carochi (1645)

the voiced stop allophones in their orthography.

This is an example of displacement. The nature of the following consonant is

justification. There are, however, some interesting exceptions.

tsitsis "abispa [wasp]"

something like itching and irritation: like figs

Again stands here for an apico-alveolar s.

a. chichi largo en la primera. mamar. [long on the first, "to suck"]

pronounce consonantal v, which they would not do if it were an o." ("Pero

vowel: "when n is placed before the vowels a, e, i, o, it usually becomes m"

because he was influenced by the frequency of occurence of the forms. For

[...] Neither is an n pronounced before an x; rather, the x is pronounced with more

la v. consonante apenas se siente si la <a>y", 1983[1645]:lv).

maiuhmochiua; and of the p, for example, nicchippoloa58 instead of nicchiuh-

In Purepecha, Gilberti recognizes that <c> is replaced by <g> when preceded

-ti if following an unstressed syllable.

10. The role of the native speakers

to distinguish them. For example, tlatlacoani "sinner", plural tlatlacoanime or tlatlacoanih;

XI. 1605), Francisco Baptista de Contreras of Cuernavaca and Estéban Bravo of

Martínez Baracs (1997) notes that Antonio Huitziméngari (c. 1519-1562),

Golden Ass (Terraciano 2001:54; Burgoa 1989[1674], I:349-350, 379).

Spanish. They showed some ability to arrive at phonologically correct appraisals

Francisco de Alvarado, Vocabulario en lengua mixteca, facsimilar reproduction with a

Basalenque, Diego. 1975[1642]. Arte y vocabulario de la lengua matlaltzinga vuelto a la

Mexico, D. F.: Universidad Nacional Autonoma de Mexico.

Varo, Francisco. 2000[1703]. Arte de la lengua mandarina. Reprinted: Francisco Varo's

Ronald W. Langacker. Berkeley, etc.: University of California Press.

Innovations. Ph. D. Thesis, Chicago, Illinois: University of Chicago.

convirtieron a los indios de la Nueva España. Mexico D. F.: Universidad Nacional

la Historia general de las cosas de Nueva España de fray Bernardino de Sahagún y de

maya-hablantes: el caso específico de laREPRESENTACIÓNde las consonantes

nish by Pilar Mascare Sacristan and published as "La gramatica y el diccionario

angeles Romero Frizzi, 173-239. Mexico, D. F.: Instituto Nacional de Antropología e

You might also like

Tradicionalmente se ha afirmado que las gramáticas coloniales de lenguas ame

4.2.10 Concluding remarks aboutarticulatorydescription. I do not mean to over