225

Mariano Rojas and his “Manual de la lengua nahuatl” (1927)

Willem J. de Reuse
University of North Texas

1. INTRODUCTION∗
I became interested in the figure of Mariano Rojas for four reasons. First, he
had a long life and saw many different leaders and he was an activist through
it all. Second, the pedagogical literature he wrote is more appealing than
others from the late XIXth or early XXth century. Third, he is mentioned in
Whorf’s (1946) sketch of Milpa Alta Nahuatl, an influential description of
modern Nahuatl. And fourth, he was from Tepoztlán, one of the most famous
Mexican villages in American anthropological literature.
Because Rojas is not an extremely well-known figure in the field of Nahuatl,
I will make some remarks on his life in section 2. In section 3, I discuss the
contents, origins, and dialect of Rojas’ Manual, arguably his most interesting
piece of linguistic work. Sections 4 and 5 are more detailed discussions of
the phonology, orthography, and lexicon in Rojas’ Manual. Section 6 deals
with the relationship, and possible influences, between Rojas and the American
linguist Benjamin L. Whorf. Section 7 concludes regarding the influence of
Rojas on Nahuatl studies, and more generally, regarding linguist and native
speaking scholar collaborations in the first half of the twentieth century.

*
In the last eight meetings of the Encuentro, I have been the one and only talking
about Apache, so I thought that for our 10th meeting I would attempt a foray in the
study of Nahuatl, a language that many more of our Mexican colleagues are interested
in. I would like to thank the Nahuatlist colleagues who have given me guidance while
I researched this paper, in particular Frank Karttunen, Karen Dakin, and R. Joe Campbell.
Joe Campbell was kind enough to send me some CD’s of his field recordings, so I have
had the pleasure of listening to Santa Catarina Zacatepec and Tepoztlán Nahuatl. All
errors and misinterpretations are of course my own.
226 Willem J. de Reuse

2. BIOGRAPHICAL SKETCH OF MARIANO JACOBO ROJAS

Mariano Jacobo Rojas (1842-1936) was a Tepoztlán native and native
speaker of Nahuatl, and an early activist for Nahuatl-Spanish bilingual
education. He pleaded the case for Nahuatl language education with
Emperor Maximilian and later with Porfirio Díaz. In Tepoztlán, he attempted
to found Lancasterian schools, which is a system of British origin, where
children teach each other, but with little success. He was involved in
several newspapers, and in 1908 became teacher of Nahuatl at the Museo
Nacional de Arqueología, Historia y Ethnografía in Mexico City. He had a
few famous students, one of whom appears to have been Isabel Ramírez
Castañeda, the first Mexican woman archeologist. Rutsch (2003) surmises
that Isabel Ramírez Castañeda might have learned her Nahuatl from Rojas,
but it is much more likely that she was a native speaker (Karttunen 1991).
By the end of his life, Rojas had become an accomplished teacher, linguist,
writer and student of Classical Nahuatl literature. In 1921, he founded a little
newspaper, for the Tepozteco community in Mexico City, called El Tepozteco.
He also wrote some poetry and a tragic play, called Maquiztli, which,
according to Lastra de Suárez and Horcasitas (1980: 235) is the only
published original Nahuatl play written in the twentieth century.1

3. ROJAS’ LINGUISTIC CONTRIBUTIONS

Rojas’ (1927) Manual was designed to teach Nahuatl at the Museo Nacio-
nal de Arqueología, Historia y Ethnografía, and has not been influential in
Nahuatl linguistics. The reasons for this are easy to find. Rojas’ linguistic
literature displays a sort of Mexican nationalism, leading one to believe that

1
More biographical information about Rojas is available in the piece written by Dávila
Garibi on the occasion of his 90th birthday (1934), in Lastra de Suárez and Horcasitas (1980:
235), and in Ascensión de León-Portilla’s Tepuztlahcuilolli (1988, Vol. I, 162-163.)
Mariano Rojas and his “Manual de la lengua nahuatl” (1927) 227

Nahuatl speakers are in some way better Mexicans than either Spanish
speakers or speakers of other indigenous languages. That is obviously
problematic. He was also a purist in that he avoided Spanish constructions,
and Spanish loans to some extent. He is also a classicist, in that he tended to
favor Classical Nahuatl constructions and sounds over Modern Nahuatl ones.
At first sight, there appears to be little in his Manual that cannot be found in
the original sources on Classical Nahuatl.
Nevertheless, Rojas was aware that one cannot just teach Nahuatl with
the Classical materials. I quote from his foreword below.

Al lector: (...)2 pues si bien es cierto que tenía a la mano antiguos y modernos
vocabularios, métodos y gramáticas de ilustres autores, su estilo, vocablos arcai-
cos, su antigüedad de expresión y su concisión tanto en unos como en otros,
ninguno de esos vocabularios y métodos pueden ser adaptables a la manera
actual de hablar el idioma mexicano, pues éste ha sufrido una especie de metamor-
fosis durante el curso de más de cuatrocientos años, es decir que en la actualidad
no se habla el idioma mexicano como lo era en la época del Reinado Azteca, sin que
por esto haya perdido su expresión, sonoridad y elocuencia en algunos casos.
(Rojas 1927: II)

As sources for his work, Rojas mentions the classical authors Sahagún,
Molina, and Carochi, as well as the XIXth century authors Chimalpopoca,
author of a short textbook (Chimalpopoca Galicia 1869), and Palma, also
author of a textbook (Palma 1886). He also states that he took into account
“la opinión eficiente de contemporáneos conecedores del idioma” (Rojas
1927: III). It is hard to find specific evidence for influence from the three
classical authors Rojas mentions, since all writing on Classical Nahuatl
grammar is largely based on these authors. I have not managed to get a copy
of Chimalpopoca’s textbook yet, but I have looked at Palma (1886). It is

2
I start in mid-sentence because Rojas writes extremely long sentences.
228 Willem J. de Reuse

more Classical in layout and language than the Manual, and its influence on
the Manual is minor.3
Table 1 below is a synopsis of the contents of the Manual, in which I also
include some preliminary evidence regarding sources and influences.

Table 1. Synopsis of contents of the Manual

Title(s) or Chapter(s) Page numbers; Comments
commentary on
immediate sources
or influences
Dedicatoria p. I Contains one text entitled
Netenamactiliztli and one
piece in verse titled Occe
Tlahtolli, in Nahuatl only
Al lector pp. II-III (cf. quotation from
foreword above)
Introducción pp. 1-2; El Tepozteco About the alphabet and
Oct 1st, 1921 accents.
Capítulo Primero, De pp. 3-42; El Tepozteco Each lesson is about a
las Articulaciones, Oct 1st, 1921, Nov 1st, consonant of the Nahuatl
Lección I through XIV 1921 alphabet. Each lesson
contains an explanation of
how the letter is pronoun-
ced, a vocabulary with
glosses, and three sets of
exercises with short Nahuatl
sentences without
translation, with words from
the vocabulary.

3
I have not had the time to do a complete philological study of the Manual, which
would entail a detailed study of all the sources. Neither have I looked at the non-
pedagogical writings by Rojas, which would include his poetry, his theater, and his
transcription of the Tepozteco myth dialogue, all mentioned in León-Portilla (1988, Vol. II,
341-342).
Mariano Rojas and his “Manual de la lengua nahuatl” (1927) 229

Capítulo Segundo, pp. 43-99 Each lesson is numbered in

Lecciones Prácticas, Nahuatl, and also titled with
Lección I through XL an important Nahuatl word.
Each lesson contains four
short sentences written
vertically, one word per line,
with parallel vertical Spanish
glosses; the four sentences
are then repeated in horizon
tal lines, first in Nahuatl, then
in Spanish translation.
Capítulo Tercero, (starts on p. 100,
Lecciones varias each lesson treated
separately below)
Lección I Del pp. 100-102; El Lists of nouns with
nombre Tepozteco, Feb. absolutive and plural
1st, 1922 suffixes, examples of
compounding of two, three,
four, and five nouns.
Lección II Del pp. 103-107; El Independent pronouns,
pronombre Tepozteco, June 1st, prefixed subject, object,
1922, July 1st, 1922 possessive, and reflexive
pronouns.
Lección III pp. 107-118 Present, past, imperfect, and
Conjugación: future forms of ‘to write’ in
Escribir six subject persons, in a
sentence.
Lección IV El amor pp. 108-110 Sentimental sentences
illustrating the noun ‘love’
and the verb ‘to love’.
Lección V El trabajo pp. 110-111 Sentimental sentences about
work and the human
condition.
Capítulo Cuarto, pp. 112-114; El Moralizing sentences about
Lecciones de diferentes Tepozteco, Oct 1st, God, sin, Montezuma, love
sentidos (actually 1921, Oct. 15, 1922, of country and Mexican
just one Lección) Palma (1886: 106-107) independence.
230 Willem J. de Reuse

Capítulo Quinto. (starts on p. 115, each Question and answer

Lecciones en forma lesson treated sentences with facing
de diálogo separately below); Spanish translation.
there is no evidence
of influence from Pedro
de Arenas’ (1611)
conversation manual
Lección I, two pp. 115-118; there is no
subtitles: Modo de evidence of influence
Saludar and Modo from the Bancroft
de Despedirse dialogues (Karttunen
and Lockhart 1987);
El Tepozteco, Nov 1st,
1921
Lección II, two pp. 118-123
subtitles: Un
caminante and
En la casa
Lección III, one pp. 123-127
subtitle: En un
pueblo pidiendo
que comer
Lección IV pp. 127-130 Conversation about
someone’s health, relatives,
one’s name, and language
knowledge.
Capítulo Sexto, pp. 131-149; Classified Spanish-Nahuatl
Vocabulario, section there is no evidence vocabularies; the entries are
titles are: I País, Nación; that Redfield’s not in alphabetical order.
II Habitación, Morada; (1930: 225-226)
III Recámara, IV kinship term list
Comedor;V Cocina; was influenced by
VI Frutas, VII Partes del section VIII; El
Cuerpo Humano; VIII Tepozteco, March
Afinidad; IX Modo de 15th, 1922
contar
Erratas notables p. 150
Índice p. 151
Mariano Rojas and his “Manual de la lengua nahuatl” (1927) 231

In the second column, I have noted Rojas’ lessons (Lecciones metódi-

cas de lengua mexicana o azteca)4 from his newspaper El Tepozteco
where material was used for his Manual. So El Tepozteco was definitely
the place where Rojas conceived the idea of his Manual. However, he did
not often copy from his newspaper; there are plenty of differences between
the El Tepozteco lessons and the Manual lessons, especially in that the
newspaper lessons contain more Classical texts, and more unambiguously
Classical forms.
It is also interesting to compare the Manual with the third and last
pedagogical and grammatical publication by Rojas: Estudios Gramaticales
del Idioma Mexicano, published in 1935, when Rojas was 93 years old.
León-Portilla states in her bibliography (1988, Vol. II, 342): “En realidad es
una abreviación de esta obra.” (“esta obra” referring to the 1927 Manual).
I do not quite agree with that. It is really a grammatical sketch, with a lot
more grammatical detail from Classical authors, –it even has a discussion of
the six Latin cases– and with no exercises or vocabulary to learn. So it is less
original than the Manual, and is therefore a disappointment. There is more
of a mixture with Classical forms in adverbs and postpositions, as seen in
Table 2 below. Classical plurals are now preferred over Tepoztlán plurals,
whereas that preference was less explicit in the Manual, as seen in Table 3
below.

4
I thank Karen Dakin for these copies of the El Tepozteco lessons, especially
since León-Portilla notes in Tepuztlahcuilolli (1988, Vol. I, 162, and Vol. II, 341) that
she was unable to find copies.
232 Willem J. de Reuse

Table 2. A postposition and an adverb in Classical Nahuatl, Rojas’ grammars,

and the Tepoztlán dialect
Classical form Rojas (1927) Rojas (1935) Tepoztlán form Gloss
-tloc5 -tlac -tlac or -tloc -tlac near (postp.)
iuhqui (often ihqui yuhqui or ihqui so (adv., see
spelled ihqui also Table 5)
yuhqui)

Table 3. Some differences in plural marking between Classical Nahuatl,

Rojas’ grammars, and the Tepoztlán dialect
Classical plural Rojas (1927) Rojas (1935) Tepoztlán form Gloss
tçteoh teôme, tçteo tçteo teômeh gods
mâmazah or mâmazâme mâmaza mazâmeh deer (PL.)
mazâmeh

Rojas (1935: 8) further points out that teôme is also possible, but he now
rejects mâmazâme: “es una redundancia, que debe evitarse”. It is quite
likely, therefore, that mâmazâmeh is a Tepoztlán dialect form as well.
In the next two sections, I focus on comparisons between the Manual
forms with Tepoztlán dialect forms, with Milpa Alta6 dialect forms, and with

5
In Tables 2 through 7, spellings of Classical Nahuatl follow Karttunen (1983); the
spellings of Rojas (1927) or Rojas (1935) are those of Rojas, except for the important
addition of the macron for vowel length. As we will see, Rojas never indicates vowel
length, and vowel length is posited on the basis of Karttunen (1983). For the Milpa
Alta and Tepoztlán dialects, the original spellings have been converted to the system
used in Karttunen (1983), which usually boils down to the addition of the macron for
vowel length, and the addition of h for the “saltillo”.
6
The Milpa Alta dialect of the Distrito Federal is spoken not very far to the north
of Tepoztlán, by people with close connections to Tepoztlán. It is closer to Classical
Nahuatl, and Rojas was familiar with it.
Mariano Rojas and his “Manual de la lengua nahuatl” (1927) 233

Classical forms. I deal with phonology and orthography in section 3, and

with lexicon in section 4. I have no space to discuss morphological, syntactic,
and pragmatic differences, although they are plenty of interesting things to
study.7

4. PHONOLOGY AND ORTHOGRAPHY

In phonology and orthography, the two problematic areas are the “saltillo”,
generally considered to have been a glottal stop in Classical Nahuatl (de
Pury-Toumi 1980), and vowel length.
Rojas writes the “saltillo” consistently as h, which is a good choice, since
it was also the practice of some classical writers, and also because the “saltillo”
sounds like an [h] rather than as a glottal stop in Tepoztlán Nahuatl. For
example, Whorf (1946) emphasizes that the “saltillo” is glottal stop in the
Milpa Alta dialect, but [h] in the Tepoztlán dialect. However, in some phonetic
transcriptions of Tepoztlán Nahuatl (González Casanova 1919-1922, and
Lastra de Suárez 1986), the “saltillo” is often written with the IPA symbol for
the glottal stop. One has to assume, then, that some Tepoztlán speakers
have a glottal stop for the “saltillo”. In the recordings from the Tepoztlán
Municipio (Tepoztlán proper and Santa Catarina Zacatepec) I have listened
to, the etymological “saltillos” are phonetically [h], and the only glottal stops
I hear are phrase-final glottal stops, which are actually good evidence that
the word ends in a vowel and not in an etymological “saltillo”, as pointed out
for other dialects by Canger (1990).

7
Regarding morphology, these would include reciprocal prefixes, second person
plural prefixes and independent pronouns, honorific forms (Dakin 1978), plurals
(Karttunen 1978), verb compounding with continuative -tica, and diminutives by
reduplication (Karttunen, p.c.). Regarding syntax and pragmatics, one could study
word order, the function of the particle in (Rojas 1935: 16), relative clauses, and polite
formulae in conversation (Barlow and Newman [n.d.], Karttunen and Lockhart 1987).
234 Willem J. de Reuse

As far as vowel length is concerned, it is phonemic in the Tepoztlán dialect

but hard to hear. Karen Dakin (p.c.) pointed out to me it took her a while to
hear it, and this explains why her text publications (Dakin 1972, 1977) do
not mark it. Recordings by Joe Campbell confirm that it is possible to elicit
minimal pairs for vowel length. Karttunen’s (1991) observation for Milpa
Alta that vowel length is phonemic but that no one writes it consistently,
seems to be true for Tepoztlán as well. Indeed, in the Tepoztlán Municipio
sources that attempt to write vowel length (González Casanova 1919-1922,
Guzmán Betancourt 1979, Lastra de Suárez 1986, Whorf 1946, Whorf,
Campbell and Karttunen 1993), it is written rather inconsistently.8 Karttunen
in her discussion of the Tepozteco, a mythical text (Karttunen and Céspedes
1982: 138) states, however, that “vowel length as such does not seem
distinctive”. This might be because the people reciting the text might not have
been very fluent speakers, and might have lost the vowel length distinction
altogether.
We can assume that Rojas had the same problems modern writers have,
and that he had a difficult time writing vowel length, assuming he had not lost
vowel length altogether. But unlike modern writers, and like modern linguists,
he knew about the Classical literature marking vowel length. In the
introduction to the Manual (1927: 1) he briefly mentions a system of accents,
either from Rincón’s (1595) or from Carochi’s (1645) grammars of Classical
Nahuatl (Whorf, Campbell and Karttunen 1993: 207), but apparently without
realizing that these accents refer to vowel length. Then, on the next page of
the introduction, he discusses vowel length, as follows:

Como el mexicano da gran importancia al valor de las vocales, teniendo al igual de

las lenguas clásicas largo y breve el siguiente, (—) colocado sobre una vocal es
largo, como si la letra que carga fuese duplicada. Ejem.; [sic] Amoxtli, (amooxtli),
libro; calli, (caalli) casa.” (Rojas 1927: 2)

8
A comparative statistical study of vowel length as recorded by these four authors
would be an interesting project.
Mariano Rojas and his “Manual de la lengua nahuatl” (1927) 235

It appears that Rojas knows exactly what vowel length is, but is quite
unable to apply his knowledge to Nahuatl. The two words he gives as
examples do not have vowel length where he doubles the vowel; if one
indicated vowel length on them with a macron, they would be written âmoxtli
and câlli. What Rojas appears to indicate by vowel doubling on these words
must be stress, which is normally penultimate in Nahuatl.
In the rest of the Manual, Rojas gives up on the whole business of Classical
style accents or macrons. He does not mark vowel length at all, and the only
mark in vowels is penultimate stress, written with the acute accent, following
the orthographic rules of Spanish stress marking.
Other interesting phonological correspondences are in Tables 4 and 5.

Table 4. Final Classical and Milpa Alta -uh [-w] to Tepoztlán –n [-n, -N]
correspondences
Rojas (1927) Classical Milpa Alta Tepoztlán Gloss
ôquichîuh ôquichîuh ôquichîuh ôquichîn s/he did it
*noconçuh9 noconçuh noconçuh noconçn10 my son
mocihuâuh mocihuâuh mocihuâuh mozohuan your wife
-tîn -tîuh -tîuh -tîn centrifugal
verbal
compounding
element

Table 4 exemplifies a well-documented correspondence, the final -n being

typical of most Morelos dialects (Whorf 1946, Dakin 1974, Campbell 1976,
Campbell and Karttunen 1994). Rojas was aware of this correspondence,
and chose the Classical or Milpa Alta -uh for his Manual. However, he did
not carry this choice through with complete consistency, as shown in the last
row of Table 4.

9
The asterisk means: does not occur in these data, but this is what is expected.
10
But Lastra de Suárez (1986: 605) has nocone and mozohua, showing loss of final
n, a characteristic of many Nahuatl varieties (Karttunen and Lockhart 1976: 9).
236 Willem J. de Reuse

Table 5. Preconsonantal Classical -uh- [-w-] to Milpa Alta and Tepoztlán -h- [-h-]
correspondences
Rojas (1927) Classical Milpa Alta Tepoztlán Gloss
âtlahtli âtlauhtli *âtlahtli âtlahtli canyon
cuah- cuauh- cuah- cuah- tree
(compounding
form, as in the
below)
cuahtla cuauhtlah cuahtla11 cuahtlah mountain,
wilderness
ihqui iuhqui ihqui ihqui so
icniuhtli icniuhtli *icnihtli icnihtli 12 brother

In Table 5, there is a correspondence between Classical -uh- and Milpa

Alta and Tepoztlán -h-. It seems that Rojas chose the Milpa Alta and Tepoztlán
forms here, with some exceptions, as in the last row of Table 5.

5. LEXICON
There was a habit, maybe started by Franz Boas (Boas 1920), of putting
Spanish loanwords in Nahuatl texts in italics, so they stand out. Rojas subverts
this practice in an interesting way: all Spanish loans in Nahuatl are given in
italics, but all Nahuatl loans into Spanish are also given in italics, as one can
see in Table (6) below. So what we have here is an ideology of reciprocal
borrowing, implying the full equality of Nahuatl and Spanish.

11
This h is occasionally written as c: cuactla, due to writer reinterpretation
(Karttunen 1991, pace Launey 1979: 349).
12
According to Guzmán Betancourt (1979: 49), the Santa Catarina Zacatepec form
is icnitl.
Mariano Rojas and his “Manual de la lengua nahuatl” (1927) 237

Table 6. Spanish words in italics or Nahuatl words in italics

(italics as in the original)
Rojas (1927) Spanish glosses, Rojas (1927) Nahuatl words
and my English translation
mi caballo ‘my horse’ nocahuayo §13
clacloyo ‘tortilla wrapped around beans’ tlahtlayôlotl §
col ‘cabbage’ collex
frijol gordo, ayocote ‘type of large bean’ ayocohyetl §
haba ‘broad bean’ huçyetl, ahuax
higo ‘fig’ pepetolxocotl §, icox
jitomate ‘tomato’ xîtomatl
mole de guajolote ‘turkey sauce’ huexôlômôlli
pan ‘bread’ pantzin
plátano ‘banana’ tzapalotl §, polantanex
tamal ‘tamale’ tamalli
vela ‘candle’ candela (!)
zacate ‘grass, hay’ zacatl
durazno ‘peach’ tohmixocotl, tolaxno (italics forgotten)
cincuenta centavos ‘fifty centavos’ nâhui tomin (italics forgotten?)
cuatro pesos ‘four pesos’ nâhui pçzzo (italics forgotten?)

For tolaxno, from Spanish durazno ‘peach’, Rojas seems to have

forgotten the italics. In the last two rows of Table 6 there are Spanish loans
for monetary units, tomin and pçzzo, and it is not clear why these should not
be in italics either.
As first noted in Redfield (1930), a pioneering ethnography of the village
of Tepoztlán, Rojas was a purist in the sense of Hill and Hill (1986: 122-
140). I quote Redfield here:

Mariano Rojas, a native of Tepoztlán, in his Manual de la lengua Nahuatl (Mexico,

1927) lists Nahuatl names for many articles of European origin, but most of these,
while intelligible, are the artificial constructs of as linguist, and few are in general
use. (Redfield 1930: 50)

13
The symbol § marks words for which vowel length marking is uncertain.
238 Willem J. de Reuse

In view of the preceding discussion of Spanish loanwords, it seems that

Rojas was a reasonable purist. Redfield’s comments would seem to be
applicable only to the Manual’s section on household items and furniture
(pp. 133-135), a small portion of the vocabulary, and even there Rojas
avoids the ridiculously long sorts of compounds reported by van Zantwijk
(1965) in the writings of more fanatical purists.
I finish this discussion of the lexicon with Table 7, a list of items that often
vary in Nahuatl dialects.

Table 7. Diagnostic lexical items

Rojas (1927) Classical Milpa Alta Tepoztlán Gloss
âxcân âxcân âxcân, âxâ(n) âxân, âxcân now
yetl, etzintli etl yetl, yetzintli yetl beans
yçxpa çxpa *yçxpa *yçxpa three times
yeztli, eztli eztli yeztli yeztli blood
nochi mochi nochi nochi all
nonâna nonân nonâna nonâna14 my mother
notlac notloc notlac, notlacuh notlac with me
cihuâtl, zoatl cihuâtl15 cihuâtl zohuatl woman
(once)

What one can conclude from Table 7 is similar to what the data in Tables
4 and 5 already suggested. When the Milpa Alta and Tepoztlán forms are
the same, Rojas will prefer these over the Classical forms for his Manual;
but if the Milpa Alta forms line up with the Classical forms, Rojas will prefer
the Milpa Alta or Classical forms over his native Tepoztlán forms for the

14
But according to Guzmán Betancourt (1979: 49), the Santa Catarina Zacatepec
form is nonân. In any event, this form is not a very good diagnostic one, since nâna
type forms might be motherese.
15
The form zohuatl, widespread in Morelos and in other Central dialects, also
does occur in Molina’s Classical dictionary (Karttunen 1983: 348).
Mariano Rojas and his “Manual de la lengua nahuatl” (1927) 239

Manual. So, for Rojas both Classical and Milpa Alta seem to have more
prestige than Tepoztlán, probably because Milpa Alta is on average closer
to the Classical language. But if Milpa Alta lines up with Tepoztlán, he will be
“Milpa Altaicizing” rather than classicizing.

6. ROJAS AND BENJAMIN WHORF

Rojas is briefly mentioned in Whorf’s well-known and influential 1946 ske-
tch of Milpa Alta Nahuatl (1946: 368), as follows:

I obtained further valuable insights from Professor D. Mariano Rojas of the National
Museum, himself a native of Tepoztlán, but well acquainted with Milpa Alta and a
most learned and scholarly exponent of the classical speech. (Whorf 1946: 368)

But the above seems to be the only place in the published literature where
Whorf acknowledges him. Unfortunately, in another very important manuscript
by Whorf, recently published and edited by Campbell and Karttunen (1993)
which has a lot of interesting data on Tepoztlán, no consultants are mentioned.16
So we do not know whether Rojas was the consultant for the Tepoztlán data
and texts, or whether it was another Tepoztlán speaker that Rojas had directed
Whorf to. If Rojas was the source of the Tepoztlán data in this Whorf
manuscript, then we would have evidence for what sort of Nahuatl Rojas
really spoke when he let his hair down. This looks like nonpurist (or in Hill
and Hill’s (1986) terminology, syncretic) Tepoztlán Nahuatl, with Spanish
prepositions like de ‘of’, and Spanish loanwords such as carga ‘load’, and
cuenta ‘account’ (Whorf, Campbell and Karttunen 1993: 192-193), the
sort of thing Rojas generally avoided in his pedagogical works.
Another potential influence of Rojas on Whorf might have to do with
Rojas’ confusion of stress with vowel length, discussed in section 4. We

16
Footnote 6, p. 170, probably by editor Karttunen, mentions Doña Luz Jiménez as
one of Whorf’s Milpa Alta consultants.
240 Willem J. de Reuse

might wonder if Rojas’ confusing statements regarding accents, stress and

length might have given Whorf the idea that something else, such as pitch,
was also involved. This, of course, is pure speculation. It is striking, however,
that Rojas spelled cá ‘it is’ with an acute accent, probably in imitation of
Spanish está, with the same meaning, and that Whorf wrote cá ‘it is’, where
the assumption is now that the acute accent marks high pitch (Whorf, Campbell
and Karttunen 1993: 188, 202).

7. CONCLUSIONS
I would like to conclude with some general observations regarding
collaborations between Nahuatl speakers and linguists in the first half of the
XXth century, and how Rojas fits into that picture. Then I will conclude
regarding Rojas’ influence on Mexican and American linguists, and his
significance in Nahuatl studies.
It appears that in the first half of the XXth century there were quite a few
talented native speakers both in the Milpa Alta and in the Tepoztlán areas
who were collaborating with linguists who either helped them write or
transcribed for them. Such speakers were Doña Luz Jiménez, the celebrated
writer in the Milpa Alta dialect of Nahuatl who collaborated first with Benjamin
Whorf, then with Robert Barlow, and then with Fernando Horcasitas
(Karttunen 1991, 2000); the Tepoztlán speaker Apolonio Escalada17 who
collaborated with Byron McAfee on an extensive collection of unpublished
Tepoztlán lessons, and the Hueyapán (Morelos) speaker Miguel Barrios
Espinosa who collaborated with Robert H. Barlow on writing the Nahuatl
Mexihkatl Itonalama newspaper, published in 1950 (Karttunen 2000).18

17
But see also Escalada (1937), for a sample of Escalada’s writing, similar to Rojas,
more Tepoztlán oriented, but not quite as consistent as Rojas.
18
Another less often mentioned case of collaboration between speaker and linguist,
which took place in the fifties was that of Ezequiel Linares Moctezuma and Federico
Mariano Rojas and his “Manual de la lengua nahuatl” (1927) 241

It is interesting that Barrios did not write pure Hueyapán dialect, but was
taught by Barlow to write a more general central sort of Nahuatl (Lastra de
Suárez and Horcasitas 1980: 284), the sort of thing that Rojas was doing by
himself.
Rojas must have known most or all of those these people personally, but
he does not mention them, nor does he appear to have collaborated with
them, except with Whorf, as already mentioned. Part of this must have had
to do with the age difference, but maybe Rojas’ classicizing and purist
tendencies might have intimidated other native speakers and/or linguists who
were not nearly as self-confident in writing Nahuatl. So Rojas seemed to
prefer older sources rather than collaboration with linguists.
What was the influence of Rojas, then? He was as a pioneer in Modern
Nahuatl literature, and as such is dutifully mentioned in recent surveys of
Nahuatl literature (León-Portilla 1992, Montemayor 2001, León-Portilla
and Shorris 2001, Figueroa Saavedra 2005). But what was his influence
from a linguistic point of view? Among Mexican scholars, he was a definite
influence on Dávila Garibi, who mentions him in many of his publications. He
is also mentioned in Miguel León-Portilla’s foreword to Swadesh and San-
cho (1966). Among American scholars, he is mentioned once by Whorf,
once by Dow Robinson,19 and a few times by the American anthropologists
Redfield and Lewis.20 That does not amount to very much, but nevertheless

Wagner, which resulted in a pedagogical work (1st edition 1953, 2nd edition 1961, but
now recently republished, 2002) somewhat comparable to Rojas’ Manual, but with
the difference that they vary inconsistently between Classical forms such as ôquichîuh
and Morelos forms such as ôquichîn (See Table 4). Linares Moctezuma was a speaker
from Morelos, but maybe not from Tepoztlán. Karen Dakin (p.c.) suggested to me he
might have been from Hueyapán.
19
The SIL linguist Dow Robinson, in the literature survey beginning his monograph
Sierra Nahuat Word Structure (1970: 2), not quite accurately, considered Rojas (1927)
to contain a popular description of the phonology of a Nahuatl dialect.
20
Even though Robert Redfield and Oscar Lewis never learned Nahuatl (as pointed
out by Lastra de Suárez and Horcasitas (1980: 236) it is interesting that Redfield
242 Willem J. de Reuse

I would argue he is an interesting figure in Nahuatl and Mexican linguistics

for three reasons.
First, to use a word put back in fashion by US Republican presidential
candidates in their 2008 campaign, Rojas appears to be a bit of a “maverick”:
he liked to do things his own way. While other Nahuatl speakers of central
Mexico tended to collaborate, in the first half of the XXth century, with linguists
in various ways, he did not, and he was his own linguist. It is clear that he did
talk with Whorf, but how much is really not clear. It is also good to remember
that Rojas was already 88 years old when Whorf met him in 1930.
Second, Rojas also distinguished himself through his relatively consistent
Classical inspired spelling system. Better known linguists, such as González
Casanova (1946) were much better phoneticians, but they were less consistent
than Rojas when they had to write in a Classical Nahuatl inspired orthography.
Third, and more importantly, there is a need for pedagogical materials in
Modern Nahuatl. Horcasitas’ (1996) short manual of Nahuatl for beginners,
is the only recent exemplar of such materials. It is written in a supradialectal
Nahuatl covering the Distrito Federal, the states of Mexico, Morelos, Pue-
bla and Tlaxcala (1996: 11), but mostly based on Barlow’s unpublished
class notes for Milpa Alta and Hueyapán. Rojas’ Manual, with its relatively
consistent choices as to some middle ground between “too classical” and
“too Tepoztlán”, and his implicit view that Milpa Alta is a model for a
supradialectal middle ground might well be the earliest prefiguration of the
supradialectal ideal explicit in Horcasitas’ work. Therefore, both Rojas’
Manual and Horcasitas (1996) should be of interest for the study of ideologies
of language engineering or dialect leveling.

spelled Tepoztlán Nahuatl in a consistent system reminiscent of that used by Rojas.

Maybe that was directly Rojas’s influence, or the influence of his teacher Edward
Sapir. Lewis (1963), on the other hand, wrote the few Nahuatl words and sentences in
his book in a much less consistent system.
Mariano Rojas and his “Manual de la lengua nahuatl” (1927) 243

REFERENCES
Arenas, P. de. 1611. Vocabulario manual de las lenguas castellano y
mexicana. [Facsimile 1982]. México, D.F.: Universidad Nacional Au-
tónoma de México.
Barlow, R. H., and S. Newman. [n.d.] El náhuatl moderno de Milpa Alta.
Unpublished manuscript in the Benson Latin American Collection,
University of Texas Library, Austin.
Boas, F. 1920. “Cuentos en mexicano de Milpa Alta, D.F.” Journal of
American Folklore 33: 1-24.
Campbell, R. J. 1976. “Underlying /Nw/ in Hueyapan Nahuatl”. International
Journal of American linguistics 42: 46-50.
Campbell, R. J., and F. Karttunen. 1994. Foundation course in Nahuatl
grammar. 2 Vols. Austin: University of Texas, Institute of Latin American
Studies.
Canger, U. 1990. “Philology in America: Nahuatl: what loan words and the
early descriptions of Nahuatl show about stress, vowel length, and glottal
stop in sixteenth century Nahuatl and Spanish”. In J. Fisiak (ed.).
Historical Linguistics and Philology. Berlin and New York: Mouton de
Gruyter. 107-118.
Carochi, H. 1645. Arte de la lengua mexicana con la declaración de los
adverbios della. México: Juan Ruiz.
Chimalpopoca Galicia, F. 1869. Epítome o modo fácil de aprender el
idioma náhuatl o lengua mexicana. México: Tipografía de la Viuda de
Murguía.
Dakin, K. 1972. “El carbonero, un cuento náhuatl”. Estudios de cultura
náhuatl 10: 329-335.
______. 1974. “Dialectología náhuatl de Morelos: un estudio preliminar”.
Estudios de cultura náhuatl 11: 227-234.
______. 1977. “Pedro Cuaresma and other Nahuatl stories”. Tlalocan 7:
46-66.
244 Willem J. de Reuse

______. 1978. “Respect and indirect reference in Santa Catarina (Morelos)

Nahuatl”. Actes du 42e Congrès International des Américanistes, Paris
1976. Paris: Société des Américanistes. 485-493.
Dávila Garibi, J.I. 1934. “Conferencia biográfica-anecdótica acerca del
distinguido mexicanista C. Profesor Mariano Jacobo Rojas, epilogada
con felicitación breve en lengua náhuatl, con motivo del nonagésimo ani-
versario de su natalicio”. Boletín de la Sociedad Mexicana de Geogra-
fía y Estadística 43: 431-467.
Escalada, A. H. 1937. “Tepoztecatl. El Tepozteco”. Investigaciones
Lingüísticas 4: 271-301.
Figueroa Saavedra, M. 2005. “Palabras olvidadas, letras borradas. La lite-
ratura de los pueblos indígenas de México”. Cuadernos del Minotauro
1: 67-78.
González Casanova, P. 1919-1922. “Un cuento en mexicano”. El México
Antiguo 1: 291-307.
______. 1946. Cuentos indígenas. Biblioteca de Filología y Lingüística
Indígenas 1. México, D.F.: Universidad Nacional Autonoma de México.
Guzmán Betancourt, I. 1979. Gramática del náhuatl de Santa Catarina,
Morelos. México, D.F.: Instituto Nacional de Antropología e Historia.
Hill, J. H., and K. C. Hill. 1986. Speaking Mexicano. Dynamics of a
syncretic language in central Mexico. Tucson: University of Arizona
Press.
Horcasitas, F. 1996. Nahuatl práctico. Lecciones y ejercicios para el
principiante. México, D.F.: Instituto de Investigaciones Antropológicas,
Universidad Nacional Autónoma de México.
Karttunen, F. 1978. “The development of inanimate plural marking in
postconquest Nahuatl”. Texas Linguistic Forum 10: 21-29.
______. 1983. An analytical dictionary of Nahuatl. Austin: University of
Texas Press.
______. 1991. “A hundred years of Milpa Alta Nahuatl”. Estudios de
cultura náhuatl 21: 271-308.
Mariano Rojas and his “Manual de la lengua nahuatl” (1927) 245

______. 2000. “The Linguistic career of Doña Luz Jiménez”. Estudios de

cultura náhuatl 30: 267-274.
Karttunen, F., and G.W. Céspedes. 1982. “The dialogue of El Tepozteco
and his rivals, September, 1977”. Tlalocan 9: 115-141.
Karttunen, F., and J. Lockhart. 1976. Nahuatl in the middle years.
Language contact phenomena in texts of the Colonial Period.
Berkeley, Los Angeles, and London: University of California Press.
______. 1987. The art of nahuatl speech: The Bancroft dialogues.
(Nahuatl Studies Series 2). Los Angeles: UCLA Latin American Center.
Lastra de Suárez, Y. 1986. Las áreas dialectales del náhuatl moderno.
México, D.F.: Universidad Nacional Autónoma de México.
Lastra de Suárez, Y., and F. Horcasitas. 1980. “El náhuatl en el estado de
Morelos.” Anales de Antropología (II Etnología y Linguística) 17: 233-
298.
Launey, M. 1979. Introduction à la langue et à la literature aztèques.
Tome 1: Grammaire. Paris: L’Harmattan.
León-Portilla, A. H. de. 1988. Tepuztlahcuilolli, impresos en náhuatl.
Historia y bibliografía. 2 Vols. México, D.F.: Universidad Nacional
Autónoma de México.
León-Portilla, M. 1992. Literaturas indígenas de México. Madrid: Edito-
rial MAPFRE.
León-Portilla, M., and E. Shorris (eds.). 2001. In the language of kings.
An anthology of Mesoamerican Literature-Pre-Columbian to the
present. New York and London: W. W. Norton and Co.
Lewis, O. 1963. Life in a Mexican village: Tepoztlán restudied. Urbana:
University of Illinois Press.
Linares Moctezuma, E., and F. Wagner. 1961. Método autodidáctico es-
pañol-nahuatl nahuatl-español. Edición revisada y simplificada.
México, D.F.: Antigue Librería Robredo y Editorial Aurora. [First edition
of 1953 is republished: 2002, México, D.F.: Ce-Acatl, A.C.]
Montemayor, C. 2001. La literatura actual en las lenguas indígenas de
México. México, D.F.: Universidad Iberamericana.
246 Willem J. de Reuse

Palma, M. T. 1886. Gramática de la lengua azteca o mejicana escrita

con arreglo al programa official para que sirva de texto en las es-
cuelas normales del estado. Puebla: Imprenta de M. Corona.
Pury-Toumi, S. de. 1980. “Le “saltillo” en nahuatl”. Amerindia 5: 31-45.
Redfield, R. 1930. Tepoztlan, a mexican village. A study of folk life.
Chicago and London: The University of Chicago Press.
Rincón, A. del. 1595. Arte mexicana. Mexico City: Pedro Balli.
Robinson, D. F. 1966. Aztec studies II: sierra nahuat word structure.
Norman, Oklahoma: Summer Institute of Linguistics.
Rojas, M. J. 1921-1923. “Lecciones metódicas de lengua mexicana o azte-
ca”. El Tepozteco. México, D.F.
______. 1927. Manual de la lengua nahuatl. Método práctico para ha-
blar, leer y escribir la lengua mexicana. México, D.F.: José Donaciano
Rojas.
______. 1935. Estudios gramaticales del idioma mexicano. México, D.F.:
José Donaciano Rojas.
Rutsch, M. 2003. “Isabel Ramírez Castañeda (1881-1943): una antihistoria
de los inicios de la antropología mexicana”. Cuicuilco 10: 1-18.
Swadesh, M., and M. Sancho. 1966. Los mil elementos del mexicano
clásico. Base analítica de la lengua nahua. México, D.F.: Universidad
Nacional Autónoma de México.
Whorf, B. L. 1946. “The Milpa Alta dialect of Aztec with notes on the classical
and the Tepoztlán dialects”. In C. Osgood (ed.). Linguistic structures
of native America. (Viking Fund Publications in Anthropology, No. 6.)
New York: Viking Fund. 367-397.
________
. 1993. “Pitch Tone and the “Saltillo” in modern and ancient Nahuatl”.
In L. Campbell and F. Karttunen (eds.). International Journal of
American Linguistics 59: 165-223.
Zantwijk, R. van. 1965. “La tendencia purista en el náhuatl del centro de
México”. Estudios de cultura náhuatl 5: 129-141.