Tomlinson Ideologies Aztec Song

Ideologies of Aztec Song
Author(s): Gary Tomlinson

Source: Journal of the American Musicological Society, Vol. 48, No. 3, Music Anthropologies
and Music Histories (Autumn, 1995), pp. 343-379
Published by: University of California Press on behalf of the American Musicological Society
Stable URL: http://www.jstor.org/stable/3519831 .
Accessed: 11/07/2013 15:40
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp
.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact [email protected].
University of California Press and American Musicological Society are collaborating with JSTOR to digitize,
preserve and extend access to Journal of the American Musicological Society.
http://www.jstor.org
This content downloaded from 128.97.245.36 on Thu, 11 Jul 2013 15:40:11 PM

All use subject to JSTOR Terms and Conditions
Ideologies of Aztec Song

BY GARY TOMLINSON
have moved far to restore the writing of preSCHOLARS
Columbian America; not so its singing.
Recent developmentsin archaeological,codicological,and anthropological interpretation,abetted by an upsurge in attention to indigenous
America in the years before and after the Columbian quincentenary of
1992, have brought us to a much fuller comprehension of the varieties
of expression and conception encoded in the various American scripts.
Numerous studies have put to work new interpretive strategies
growing from this deepened comprehension. They have reread familiar but insufficiently grasped indigenous texts, and they have begun to
read writings that were rarely before thought by Europeans to be
writing at all. They have shown much-traveled dichotomies of preliterate versus literate or oral versus written societies to be predicated on
simplistic, Eurocentric judgments of what constitutes a sophisticatedly
meaningful script and what does not. They have begun to break down
long-held Western biases that extol the flexibility and semantic
richness of phonetic writing above all other approaches to script. And,
most provocatively, they have begun to uncover orders of meaning
that do not coincide with European notions of time, space, and human
movements through them.
But of the singing that was so often associated with the inscribed
traces of American cultures we still hear relatively little. American
voices have remained largely silent. This essay, in tandem with
another one published elsewhere and discussed further below, aims to
provide prolegomena for a broader restoration of the singing of one
particular group of cultures, the Nahuatl-speaking peoples of central
Mexico, in the years before and after the coming of Cortes.'
SNOW
'
Perhaps the most challenging and synoptic among the new readings of American
scripts is Gordon Brotherston's Bookof the Fourth World:Readingthe Native Americas
through Their Literature (Cambridge: Cambridge University Press, 1992); more
technical, less venturesome, but of consummate interest is Joyce Marcus, Mesoamerican Writing Systems: Propaganda,Myth, and History in Four Ancient Civilizations
(Princeton: Princeton University Press, 1992). The essays collected in Elizabeth Hill
Boone and Walter D. Mignolo, eds., Writing without Words:AlternativeLiteraciesin

344
OFTHEAMERICAN
MUSICOLOGICAL
SOCIETY
JOURNAL
Standing behind the new readings of American scripts, informing

them implicitly or in some cases explicitly, is the set of views and
methods commonly referred to as poststructuralism.2This movement
can be seen to have facilitated these new readings in several general
Mesoamericaand theAndes(Durham, N.C.: Duke University Press, 1994), reached me
late in the writing of this essay; their relevance to my topic will be at least partially
indicated below. For colonial mediations of indigenous writing in the central Mexican
area, see Serge Gruzinski, La Colonisationde l'imaginaire:Sociftesindigeneset occidentilisationdansle Mexiqueespagnol,XVT-XVIIY7
sidcle(Paris:Gallimard, 1988), translated
by Eileen Corrigan as The Conquestof Mexico:The Incorporation
of Indian Societiesinto
the Western World, Sixteenth-EighteenthCenturies[Cambridge: Polity Press,
i993]);
and James Lockhart, The Nahuas after the Conquest:A Socialand CulturalHistory
of the
Indians of CentralMexico, Sixteenth throughEighteenth Centuries(Stanford: Stanford
University Press, 1992). With a few exceptions, most notably two articles in Writing
without Words(Mark B. King's "Hearing the Echoes of Verbal Art in Mixtec Writing"
[pp. I02-36] andJohn Monaghan's "The Text in the Body, the Body in the Text: The
Embodied Sign in Mixtec Writing" [pp. 87-101]), all these studies pay at most
fleeting attention to song. My own partner-essay to this one, discussing from other
vantage points some of the issues taken up here (see the third section below, "The
Cantares Deferred") and advancing views on the connected materialities of Aztec
utterance, writing, and subjectivity, is "Unlearning the Aztec Cantares(Preliminaries
to a Postcolonial History)," in Objectand Subjectin RenaissanceCulture,ed. Margreta
de Grazia, Maureen Quilligan, and Peter Stallybrass (Cambridge: Cambridge University Press, 1995), in press. Both the present essay and "Unlearning the Aztec
Cantares" have profited much from responses to them by Jonathan Goldberg,
Margreta de Grazia, Stephen Greenblatt, Peter Stallybrass,Nancy Vickers, Lilliane
Weissberg, and others.
2For three studies whose reliance on poststructuralist insights is explicit, see
Brotherston, Book of the Fourth World; and Elizabeth Hill Boone, "Introduction:
Writing and Recording Knowledge," and Walter D. Mignolo, "Afterword:Writing
and Recorded Knowledge in Colonial and Postcolonial Situations," both in Writing
without Words, ed. Boone and Mignolo, 3-26 and 293-313 respectively. See also
Brotherston's earlier essay, "Towards a Grammatology of America: Levi-Strauss,
Derrida and the Native New World Text," in Literature,Politicsand Theory:Papers
from the Essex Conference,1976-84, ed. Francis Barker, Peter Hulme, Margaret
Iversen, and Diana Loxley (London: Methuen, 1986), 190-209. Mignolo's poststructuralism is, at the least, ambivalent. He rejects Brotherston's call for a grammatology
of New World writing, ascribing to Brotherston-mistakenly, I think-the advocacy
of Derrida as a "model" for understanding American writing and the view that
Derrida's "thesis ... is ... automatically relevant to account for Mesoamerican and
Andean writing practices before the conquest" (p. 303). In my reading Brotherston
instead advocates-guardedly--Derridean argument as a starting point for critique of
our general, Europe-inflected notions of the relations between writing and speech;
with this use of Derrida Mignolo has no quarrel (see p. 304). Mignolo's idea of
"rereading Derrida's grammatology from the experience of the Americas" (p. 303)
would be more feasible in a situation where the hegemony of Western language
ideologies did not weigh heavily on us. Since in my view (and clearly enough in
Mignolo's as well) it remains burdensome, we need to bring analytic strategies for
exposing its hidden structures together with careful study of native American traces.
It is these parallel paths that I set out on below, with a reconceptualization of song in
mind.

IDEOLOGIESOF AZTECSONG
345
ways. Poststructuralistviews have extended constructivist conceptions

of language developed in the writings of Wittgenstein, Saussure, and
others, in which language is seen as our primary means of conceiving
the reality around us or even the medium in which we make and
remake that reality. At the same time they have deepened Saussure's
insight that the meaningfulness of words arises from their differences
from one another in linguistic networks; they have implicated this flux
of difference more and more radically in linguistic constructions on
and of reality. Poststructuralismhas thus made language seem at once
more fundamental to human cognition and more diverse and malleable in its applications. All these tendencies have promoted a more
ecumenical stance in judging how people different from us, locally or
globally, constitute through words their own worlds. They have also
stimulated useful critique of the cultural limitations-the ethnocentrism, to put it more bluntly-of our most automatic assumptions
about language and its capabilities.
The question of the relations between spoken and written language, which sits at the center of this critique, was raised in clarion
tones in Jacques Derrida's discussion of Claude L6vi-Strauss in De la
grammatologie.3Here Derrida exposed the deep-rooted European
ideology behind "The Writing Lesson," a famous episode in LeviStrauss's Tristestropiques.He showed how L6vi-Strauss's depiction of
the corrupting advent of European writing among the Nambikwara of
Brazil was constructed from a habitual European suspicion of writing,
a nostalgic hankering after pristine "noble savages," and a repressing
of evidence concerning the Nambikwara's indigenous writing. Derrida went on, in Of Grammatology,to uncover the roots of LeviStrauss's distrust of writing and his privileging of speech in a widely
dispersed and age-old Western ideology he named "logocentrism" or
"phonocentrism." This ideology granted a special status to the
phonetic writing that accompanied it; in his critique of logocentrism
Derrida crucially prepared the way for later writers' attempts to
repudiate this bias toward phoneticism.
What scholars who have profited from Derrida's arguments have
tended to ignore, however, is the fact that his central text in analyzing
the structures and conundrums of logocentrism, Jean-Jacques Rousseau's Essai sur l'origine des langues, is almost as much a book about
3Jacques Derrida, De la grammatologie(Paris: Les Elditions de Minuit, 1967),
translated by Gayatri Chakravorty Spivak as Of Grammatology(Baltimore: Johns
Hopkins University Press, 1976); see pt. 2, chap. i. All citations in the text below will
be from Spivak's translation, abbreviated OG.

346
OFTHEAMERICAN
MUSICOLOGICAL
SOCIETY
JOURNAL
singing as it is about speech (and more a book about singing than
about writing). Read in a certain fashion, Derrida's analysis not only

frees writing from its bondage to speech in Western language
ideologies; it also points up the ideological pressures that have led us
to subordinate song to speech. It indicates how song might resist this
subordination and in so doing allies singing with writing, over against
speech and its logocentric privilege. Exploring this Derridean stance
raises suggestive possibilities for hearing song in certain instances of
nonphonetic writing. More specifically, it facilitates a critique of the
forces that have kept European observers of the indigenes of central
Mexico-the Aztecs, as they are now commonly called-from hearing
their singing over the last five centuries. Here I will pursue particularly the second of these possibilities and through it, to a lesser degree,
the first.
ReadingDerrida ReadingRousseau:The Play of Supplements
The most durable lesson of the Grammatologyis that the connections and distinctions the modern West perceives between speech and
writing are not natural and given in human communication but rather
constructed in complicity with a whole metaphysics of presence. This
ancient metaphysics uniquely valued the spoken word as the place
where a divine being and meaning were most directly revealed;it was,
in this way, speech- or sound-centered, logocentric or phonocentric.
With the emergence through the Renaissance (and especially in the
Cartesian moment at its end) of a new subjectivity, the metaphysical
emphasis shifted from a divine presence in logos to an individual,
reflective self-presence there (see OG, pp. 16, 97-98); but still the
spoken word was exalted. As it has continued to be: from Rousseau
through Hegel to Saussure, writing has been devalued as an indirect
representation of a presence found embodied in speech; it has been
seen as a sign of a sign of presence.
In a historical convergence that is not accidental, the period of the
emergence and development of this modern, logocentric subjectivity
is also the period of Europe's growing contacts with and subjugation
of vast stretches of the non-European world. It is the period of
Europe's first sustained encounter with many forms of nonalphabetic
writing, including ideographs, hieroglyphs, and pictographs. In dealing with such writing, Derrida argued, logocentrism operated as
language ethnocentrism and language teleology (OG, pp. 76-81).
in that phonetic writing, while subordinate to
Languageethnocentrism,
speech, could nevertheless be raised over all other writing systems: it

347
aimed to inscribe the sound or phonos of speech, and in this it

represented presence-in-speech more faithfully than nonphonetic
writing systems. Language teleology,in that, as Derrida has it, the
"decentering" that "followed the becoming-legible of non-occidental
scripts"(OG, p. 76) was quickly accommodated in a vision of historical
progress toward phoneticism, a progress then correlated with a
putativesocietalevolution.For Rousseau,pictographsor hieroglyphs,
ideographs, and alphabet "correspondalmost exactly to three different

stages according to which one can consider men gathered into a
nation. The depictingof objects is appropriateto a savage people;
signs of words and of propositions, to a barbaric people, and the

alphabet to civilized peoples."4
Derrida's analysis of Western language ideology gravitates toward
the eighteenth-century, a period, he notes, when questions as to
writing's place first irrupted from within logocentrism: "Neither
Descartes nor Hegel grappledwith the problem of writing. The place
of this combat and crisis is called the eighteenth century"(OG, p. 98).s
Derrida devoted the lion's share of his book to a reading of Rousseau's
Essaybecause he found there "the most energetic eighteenth-century
reactionorganizing the defense of phonologism and of logocentric
metaphysics. What threatens is ... writing" (OG, p. 99). And not only
writing: singing threatens as well, as we may recognize from Rousseau's Essay and from certain tangents in Derrida's presentation.
4Both the ethnocentrism and the teleology have continued to run strong in
general analyses of writing; see, for example, the literature reviewed in Boone,
"Introduction." I read Rousseau's Essai sur l'originedes langues in the translation by
John H. Moran in Two Essayson the Origin of Language:Jean-JacquesRousseauand
Johann Gottfi'edHerder, trans. John H. Moran and Alexander Gode (1966; reprint,
Chicago: University of Chicago Press, 1986), hereafter cited in the text as Essay.For
the original text see Jean-Jacques Rousseau, Essaisur l'originedeslangues,ed. Charles
Porset (Bordeaux:Guy Ducros, 1970). For the quotation here see p. 17 (for Derrida's
analysis of this passage see Of Grammatology,291-302). For another view of Derrida's
analysis of Rousseau, see John Neubauer, The Emancipationof Musicfrom Language:
Aesthetics(New Haven: Yale University
Departurefrom Mimesisin Eighteenth-Century
Press, 1986), chap. 6. Yet another, famous reaction to Derrida's account is Paul de
Man, "The Rhetoric of Blindness: Jacques Derrida's Reading of Rousseau," in
Blindnessand Insight:Essaysin the Rhetoricof*ContemporaryCriticism,2d ed. (Minneapolis: University of Minnesota Press, 1983), 102-41. I agree with Neubauer (pp. 88,
that de Man's reading is marred by a willful ahistoricism of method that is
I02)
imputed to Rousseau as well. See also n. 42 below.
s The fact that this historical localization is questionable does not affect, I think,
the cogency of Derrida's analysis. In forthcoming work I will suggest ways in which
the difficulties in Western language ideology that Derrida sees emerging in the
eighteenth century were broached earlier, in Renaissance humanist conceptions of
language that themselves evolved and shifted under the impact of late-Renaissance
awareness of non-European languages and cultures.

348
SOCIETY
JOURNALOF THE AMERICANMUSICOLOGICAL
Singing and writing converge on speech from different sides, from the
side of its potency as auto-affective phonos and from the side of its
semiological mechanics. By virtue of the characteristic surplus-overspeech that each carries, they resist the preeminence of speech within
the logocentric scheme. They function as speech's dual others, as
traces that condition through the movement of difleranceits meanings
and powers while it conditions theirs.6
Derrida locates a reflection of this problematic alterity of writing
and singing in Rousseau's use of the word suppliment.This word,
Derrida points out (pp. 144-45), can signify both an addition (or
surplus or completion) and a substitution (or replacement or displacement). This dualism, replicating the general grammatological relation
of diffiranceby which the meaning of one sign depends on (is shifted
onto) another, opens a rift in Rousseau's defense of phonologism. For
Derrida the rift is a window on the dispersed operation of the
metaphysics (Rousseau's and ours) that, by suppressingthe dualism
inherent in the supplement,has constructed for the West ostensibly
natural and universal relations among song, speech, and writing. For
us the rift will expose this, and also something more specific: an
ideological apparatusor, if you will, a set of myths within our language
beliefs that has helped determine Western conceptions of the Aztecs
since the sixteenth century. The terms that sustain this apparatuswill
6 To review some basic issues: Derrida's grammatology diverges from Saussurean

linguistics in that it follows through to its conclusion Saussure's view that meaning
arises in systems of signs from the mutual differences that distinguish one sign from
another. In what is one of Derrida's most fundamental insights (it is surely his most
well known), he called attention to the deferralimplied in Saussure'splay of difference
among signs, to the incessant slippage of meaning, from one sign onto another,
entailed in the production of meaning through difference. Derrida coined the term
diffiranceto signal both Saussure's play of difference and the deferral it entails. This
insight carries with it the implication that meaning arises not from Saussure's stable
(if arbitrary) connection between signified things and signifiers (themselves related
through their mutual differences), but from a deferral of difference from one sign to
another that operateswhollywithin the systemof signs; "il n'y a pas de hors-texte," as
Derrida infamously put it. Meaning comes from the jostling and shouldering of one
sign by others, in a manner loosely analogous to Bakhtin's heteroglossia, where any
unilateral significance of a word is constantly challenged and rivaled by numerous
others. Difference, plurality, and otherness inhabit the heart of meaning. To this
alterity, to "the part played by the radically other within the structure of difference
that is the sign" (as Spivak puts it in her introduction to Of Grammatology),Derrida
gave the name trace (p. xvii). Such difference built into the sign creates in any single
instance of it not only meaning but remaindersto that meaning spilling over around
it. These remainders are closely connected in Derrida's thought to the play of
Rousseau's supplement (see below).

IDEOLOGIESOF AZTEC SONG
349
at the same time help us to unmake it by the token of their

supplementarity. They are four: song, writing, poetry, and metaphor.
The dual "logic of the supplement" undoes Rousseau's attempts to
assign song an unequivocal role in his history of languages. Rousseau
wishes to understand song as an addition to speech imitating its
accents by a certain permanence not found in it: "There seems to be
wanting to the sounds which form the discourse, no more than
permanence, to form a real tune," he writes in the Dictionaryof Music.
This permanence seems to be a matter not only of sustained intonation but also of a codified, restricted system of intervals not observed
in speech: "The different inflexions which we give to the voice in
speaking, form intervals which are not at all harmonic, which form no
parts of the system in our music." Because it requires this codification,
it would appear, "the tune does not seem natural to mankind."It is an
artifice added to supplement the speech of civil societies, and in this
spirit it may be provisionally denied to "savages":"Tho' the savages of
America sing, because they speak, yet a true savage never sung."7
At the same time, however, Rousseau envisions song as the point
of origin of speech, as a privileged and primordial imitation of the
passions from which speech first arose: "With the first voices came the
first articulations or sounds formed according to the respective
passions that dictated them.... Thus verse, singing, and speech have
a common origin. ... the first discourses were the first songs" (Essay,
p. 50). In this notion that the first language was sung he extends a
European strain of thought with deep roots in ancient myth and
pastoral ethos and their Renaissance revivals. The speaking voice
reaches back through history to merge into the passionate idyll of
song: "In a language which would be completely harmonious, as was
the Greek at the beginning, the difference of the speaking and singing
voices is null: We should have the same voice for speaking and
singing. Perhaps that may be at present the case of the Chinese,"
Rousseau writes, again in the Dictionary.8 For Rousseau, then, song is
not only a supplementary accretion on speech in civil society; it is also
a supplementary displacementof speech at its beginnings, before civil
7See Jean-Jacques Rousseau, Dictionnaire de musique (Paris, 1768; facs. ed.,

Hildesheim: G. Olms, 1969), s.v. "chant." William Waring's translation of the
Dictionnaire, from which I quote, renders Rousseau's chant as "tune" (A Complete
DictionaryofMusic [London, 1779; facs. ed., New York:AMS Press, 1975]; see p. 45I).
This material is quoted in OG, p. 197; for related material see Rousseau's Essay,
66.
8 See Rousseau, Dictionnaire,s.v. "voix" (Dictionary,trans. Waring, s.v. "voice");
also quoted in OG, p. 198.

350
OFTHEAMERICAN
MUSICOLOGICAL
SOCIETY
JOURNAL
society. It is at once artificial, unnatural to the human voice, unfamiliar
to savages, and the voice's first, natural, most primitive and authentic
expression.
Inevitably, the place of Rousseau's "savages"in this reasoning is as
equivocal as that of song itself. Because song is not natural to
humankind but a civilized artifice, "a true savage never sung." But
because song displaces speech at its origin, aboriginal peoples or those
closer than Europeans to them in Rousseau's estimation remain nearer
the passional origins of speech/song. Their present-day discourse may
even still merge speech with song, as in "the case of the Chinese." The
primal, authentic speech-force of melody survivesin Rousseau's world
only insofar as it is given voice by others inferior to Europeans.
This equivocation in regard to non-Europeans divides song against
itself: civilized peoples systematize melody, in effect creating a phenomenon inaccessible to savages; but at the same time they slide
farther and farther away from a preexistent and aboriginal song
displacing speech. It is not much to my point here that Rousseau's
dilemma arises in part from his advocacy of melody and his polemic
against part writing, harmony, and theorists of harmony like Rameau
(another strain in his musical thought with roots reaching back to the
Renaissance). But it is important to note that the dilemma ramified in
another distinction in Rousseau's thought: the distinction of song, the
enactment of melody, from music,the degraded realm of harmony, far
from authentic imitation of passions. In chapter 19 of the Essay
Rousseau traced a European evolution by which "singing gradually
became an art entirely separate from speech; ...
the harmonics of
sounds resulted in forgetting vocal inflections; and finally ... music,

restricted to purely physical concurrences of vibrations, found itself
deprived of the moral power it had yielded when it was the twofold
voice of nature" (Essay,pp. 71-72). Music is for Rousseau a privileged
depravity limited to only the most advanced societies. Harmony, as
Derrida puts it, is "the evil and the science proper to Europe" (OG, p.
212).
So in counterpoint with Rousseau'sviews on melody, grounded in

the logocentric privilege of speech and its archaeo-teleology involving
song, "music" emerges as an exclusionary principle. It takes shape as
a conception that cannot embrace song fully but instead must overlap
partially and uncomfortably with it, denying some versions of it its
name. This "music"is an exemplaryoutgrowth of Western metaphysics, one anticipated in views earlier than Rousseau's and greatly
enhanced in later ones-and given a positive, not negative, valence in
most of those enhancements. It is an ideological mechanism whose

IDEOLOGIES OF AZTEC SONG
35 I
operation will eventually drown out whole realms of others' singing.

The supplementary equivocation so marked in Rousseau's construction of song is a control for turning down the machine.
In separating song from speech, Rousseau's evolution of music
tends to separatesong from poetry as well. It proposes for the kinds of
song it embraces a system of intervals that, as we have noted,
distinguishes civilized song from savage utterance. By the same token,
it pushes poetry closer to a body of rhetorical theory that historically
had been allied more with speech (oratory) than with song. Of this
civilized poetry, figurative language-metaphor-has always been a
crucial identifying feature: figurativelanguage viewed as a supplementary surplus added onto literal, rational language. For metaphor to
exist, there must first be literal meanings that it may displace, transfer,
and reorder. Within the ideology of music, poetry is no longer
determined by its melody; it is defined by its tropes.
But this poetic phylogeny is undone by a supplementary ambivalence of poetry and metaphor similar to that of song. Following the
pastoral mythos I mentioned above, Rousseau identifies poetry with
song at the origins of language: "The first tales, the first speeches, the
first laws, were in verse. Poetry was devised before prose. That was
bound to be, since feelings speak before reason. And so it was bound
to be the same with music. At first, there was no music but melody and
no other melody than the varied sounds of speech. Accents constituted
singing" (Essay,pp. 50-51; quoted in OG, p. 21 4). Poetry shares with
song an intimacy with and fullness of passion not found in more
rational speech. It cannot be only a refinement on speech but must
simultaneously mark the origin of speech. The early interceding of
reason and philosophy between passions and speech is antithetical to
poetry and song and threatens to destroy them: when Greece became
"full of sophists and philosophers, ... she no longer had any famous
musicians or poets. In cultivating the art of convincing, that of
arousing the emotions was lost" (Essay,p. 69; quoted in OG, p. 201).
Just as the origin of language was song/poetry, so the earliest
language was figurative, not literal. Rousseau develops an elegant
historical fable to depict this first language:
Upon meetingothers, a savageman will initiallybe frightened.Because
of his fearhe sees the othersas biggerandstrongerthanhimself.He calls
them giants.After many experiences,he recognizesthat these so-called
giants are neither bigger nor strongerthan he. Their staturedoes not
approachthe idea he had initiallyattachedto the word giant. So he
invents anothername common to them and to him, such as the name
man,for example,and leavesgiantto the falseobjectthat had impressed

352
JOURNAL OF THE AMERICAN MUSICOLOGICAL SOCIETY
him duringhis illusion.That is how the figurativewordis bornbeforethe

literalword. (Essay,p. 13; quotedin OG, p. 276)
Metaphor is thus imbued with supplementary ambivalence: it is at
once a de-rationalizing ornament added as surplus to rational speech
and a primal link between language and passion beforerational speech.
Rousseau writes: "As man's first motives for speaking were of the
passions, his first expressions were tropes. Figurative language was the
first to be born. Proper meaning was discovered last. One calls things
by their true name only when one sees them in their true form. At first
only poetry was spoken; there was no hint of reasoning until much
later" (Essay,p. 12; quoted in OG, p. 27I).
Here again, as in the case of the linkage of melody and speech,
Rousseau turns to the "barbarians"of his own world for an example:
"The genius of oriental languages, the oldest known, absolutely
refutes the assumption of a didactic progression in their development.
These languages are not at all systematic or rational. They are vital
and figurative" (Essay,p. i i; quoted in OG, p. 273). Imprecision and
irrationalityof style are linked with less-than-fully-civilized languages.
(And not only by Rousseau; Condillac made the same association in
his Essaisur l'originedes connaissances
humaines,as Derrida noted, and
offered an example we will recall in considering Mexican song: "Even
now, in the southern parts of Asia, pleonasms are considered as an
elegance of speech." See OG, p. 273.) But at the same time imprecise,
irrational metaphor forms a truer expressive language than any
rational speech.
In this the poetry of primitive language would seem to be exalted
above modern, theorized poetry, just as primitive song could seem
more authentic than civilized song (or music). Except that by an
imperious sleight of hand Rousseau allows himself an escape: "The
illusory image presented by passion is the first to appear, and the
language that corresponded to it was also the first invented. It
subsequently became metaphorical when the enlightened spirit, recognizing its first error, used the expressions only with those passions
that had produced them" (Essay, p. I3; quoted in OG, p. 276).
Enlightened Europeans created metaphor, even though savages had
first naturally deployed it, by joining it only with the passions
appropriate to it. Derrida remarks of this quotation: "The Essaythus
describes at the same time the advent of the metaphor and its 'cold'
recapture within rhetoric" (OG, p. 276). There is more at stake here
than a chronological narrative, however. In a movement exemplary of
the whole European construction of others over the last half millen-

353
nium, Rousseau attributes metaphor to savages even as he withdraws

it from them, recovering it for control by Europeans. He allows
savages to construct their linguistic world, but with toolsdelimitedand
defined by Europeans. The strategy sits at the heart also of
poetic/aesthetic readings of Mexican song, as I will relate. Meanwhile
it is important to remember that Rousseau's movement is a sleight of
hand, not a substantive resolution of the dilemma of the supplement.
He recovers tropes for civilized peoples by means of their ability to use
them "only with those passions that had produced them." But where
in his conception of primitive language use was there the possibility of
savages misusingtropes, that is, aligning them with passions inappropriate to them?
Rousseau's views of writing, finally, are caught up in much the
same dilemmas of supplementarity as his views of song, poetry, and
metaphor. Here it is the pictograph that stands in place of cherished
melody, the alphabet in place of despised harmony. This devaluing of
alphabetism is unexpected. In Rousseau's logocentrism, writing as a
whole is an addition to speech that more or less distantly reflects the
presence embodied in it. Alphabetic writing is the calculus that comes
closest to retrieving the spoken accent. It is phonography,an inscription
of speech sounds themselves and hence closer to the presence of
speech than other writing systems. For this reason, as we have seen, it
is for Rousseau the most advanced writing, appropriate to "civilized
people."
But what of a writing that elides the signifying distance between
speech and things? Pictography, if it claims no ability to inscribe
sounds, inscribes things themselves. It writes gestures, which for
Rousseau constituted a preverballanguage, universal and independent
of convention, in which primitive people expressed their needs and,
less so, their passions. The original speech/song of these peoples was,
in the historical scheme of the Essay,a supplement to earlier communicative gesture that arose in order to express passions more fully and
effectively than gesture could (see the Ersay,chaps. 1-2). If, as Derrida
writes (p. 301), both pictography and phonography tend to efface
themselves as signifiers in the presence of the things they signify, those
things are nevertheless different in status for Rousseau. The modern
spoken word signified by the conventions of alphabetic writing is
farther from the passional origins of language than the gesture
signified by a pictograph. As Rousseau puts it in the Essay, "The
primitive way of writing was not to represent sounds, but objects
themselves whether directly, as with the Mexicans, or by allegorical
imagery, or as the Egyptians did in still other ways. This stage

354
SOCIETY
corresponds to passionate language" (Essay,p. 17; quoted in OG, p.

237).
So while phonography is the writing closest to modern speech, it
has alreadytaken a step with that speech away from the articulation of
passion at the source of languages. And this is a step larger than the
step taken by pictography/gesture. Or, in the language of supplementarity: writing can be both a surplus brought to speech, recording its
sound, and the record of a pre-phonic substitute for speech at or
before its origin, namely gesture. Pictography can displace speech,
substituting for it at the point of its emergence from gesture-at the
very point, that is, where speech was still song. An alliance of song and
pictography emerges from the supplementary play of Rousseau's
historical schema. The writing that touches this preverbal origin, that
approaches more closely than phonography the primordial presence
of passion, is not civilized writing but the writing of savages. It is the
very savagery of pictography that ensures its priority over the purest,
most authentic speech. The play of the supplement again renders
equivocal Europe's judgments of savage others.
In the play of the supplement the demand of metaphysics for an
unequivocal, unilateral, and universal positioning of song, writing,
metaphor, and poetry is undone. The working of logocentrism, which
would make those things familiar by setting them in fixed, immobile
relation to speech-a relation, usually, of devalued accretion-is
replaced by a different operation of fluid interrelations between speech
and its others. In their supplementarymobility they refuse to be tamed
by metaphysics, standing outside the meanings found for them within
it in an ineffable flux of dififrance.
Song's uncanny otherness from speech-an otherness lurking (if
quietly) near the heart of many musicological endeavors-thus partakes of the alterity and dififrancehardwired in the grammatological
structure of meaning. Poetry, too, is understood now neither as a
canonized "literature"nor as "primitive"song itself, object of nostalgic European attempts at capture and literary domestication, but
rather as any number of heightened utterances that will always slip, in
some measure, out of reach of either of these Western projects. This
slippage means that the alterity of song and poetry cannot be assuaged
from within metaphysics, by appeal to code words like "aesthetics,"
"the sublime," or even "music."
Neither will it be assuaged by the development, parallel to and as
it were from within alphabetic writing, of separate writing systems for
song (i.e., music notations). As westerners have understood and
deployed these, they work within the ideology of music to separate

355
song and poetry and fix the relations between them. By the same
motion they assimilate poetry to speech, to speech's own, limited
writing (alphabetism), and to its "literatures."In the encounter of
non-European others, music notations arise-we will see two examples later-as a function of an absence within alphabetism, an
outgrowth of the egregious insufficiency of phonetic writing to
encompass the otherness of song/poetry. Music notations work in
intimate complicity with the logocentrism that has determined our
orderings of speech, song, and writing. They record and represent
from within these Western orders, in the process excluding other
musical writings that do not. Which is finally only to say that music
notations, like the other phenomena we have treated here, must find
their own supplementary flux. They instance once more the basic
lesson ethnohistorians might draw from Of Grammatology:to create
and define savages it is enough to rely on Western metaphysics; to
converse with others we need to embrace the logic of the supplement.
Mexican Song Captured/TheSupplementRepressed
Since the sixteenth century the ideological patterns of logocentrism have shaped European views of indigenous Americans. They
have worked steadily to impose on Americans regimes of communication that have no necessary indigenous resonance. Controlled
within these regimes, native cultures could be evaluated according to
teleological evolutionary schemes more and more firmly associated
with the regimes across the last four centuries. This is what is at stake
in the European domestication of Mexican utterance: it is a broadly
dispersed discursive adjunct-the most fundamental discursive adjunct, I believe-to the European conquest, subjugation, colonization,
and extermination of native Americans. The logocentric capturing of
Mexican utterance began to take shape in sixteenth-century Spanish
responses to Mexican speech, song, and writing. It coalesced into a
full-fledged (if unselfconscious) offensive on behalf of Western ideologies by the mid-eighteenth century, not coincidentally the era of
Rousseau. And it has controlled the range of Western conceptions of
indigenous Mexicans ever since.
Here I will sketch only these constraints on Mexican utterance,
within the categories of song, poetry, metaphor, and writing that have
emerged from my reading of Derrida. I am interested not so much in
the substance of individual debates over Mexican poetry, writing, and
the like as in the more basic assumptions reflected in them. These
assumptions unite most if not all the parties to the debates, but they

356
OFTHEAMERICAN
MUSICOLOGICAL
SOCIETY
JOURNAL
also foreshorten their views. For the Western capturing of Aztec

utterance reflects the West's fundamental failure to comprehend the
categories by which it has extended its control. In this way it portrays
our repression of supplementary logic.9
The logocentric regime is most evident in the case of writing.
Almost from the first moment of contact the Europeans worried over
Mexican picture writing. The most common way of familiarizing this
threateningly unknown script was by assimilating it to a cryptic
writing nearer to home: Egyptian hieroglyphs. Peter Martyr pointed
up the similarities of the two systems as early as 1530 (DecadeIV; see
Keen, p. 64), only nine years after the fall of Tenochtitlan, and they
were raised again by numerous writers over the next two centuries (by
Andr6 Thevet, for example, in his Cosmographie
universelleof 1575; see
At
least
the
middle
of
the
seventeenth century,
Keen, p. 152).
by
however, the attempt to identify the two systems was challenged. The
Jesuit AthanasiusKircher, alwaysready to express (at length) a view on
any arcane topic, denied that Mexican pictographs carried the kind of
secret meanings Egyptian hieroglyphs did (OedipusAegyptiacus,I65254; see Keen, p. 208). But the gulf opened here between Mexican and
Egyptian writing did not undermine the familiarizing intent behind
their earlier identification. Instead it marked the Western consolidation of an evaluative scale of writing systems in which Mexican
pictography could be at once familiarizedand devalued. The scale was
accepted as the basis for discussion both by writers who adopted
Kircher's dismissive attitude toward Mexican writing, like Cornelius
de Pauw (Recherches
sur lesAmericains,1768; see Keen, p.
philosophiques
and
those
like
others,
261),
Juan Jos6 de Eguiara y Eguren
by
(Bibliotecamexicana,1755; Keen, p. 224), Gian Riccardo Carli (Lettere
americane,1780; Keen, p. 27 I), or Francisco Javier Clavigero (Storia
antica del Messico, 1780-81; Keen, p. 298), who insisted that the
Mexican pictographs were not literal renditions of things but carried
symbolic meanings as subtle as those of Egyptian writing.
By the late eighteenth century the full logocentric teleology of
writing systems we have seen in Rousseau, from pictography through
hieroglyphs and ideographs to alphabetism, was brought to bear on
the Western understanding of Mexican writing. William Robertson,
9 This preliminary sketch will reflect my own reading of numerous reports on the
Aztecs dating from the sixteenth to the twentieth century; but for many other reports
I rely for now on information culled especially from Benjamin Keen's valuable survey
TheAztecImagein WesternThought(New Brunswick:Rutgers University Press, 1971).
Further references to Keen will be given in the text.

357
for example, noted in his History of America of 1777 the Mexicans'

inability to progress from pictographs to hieroglyphs and the alphabet; he ascribed it to the brief duration of their empire (see Keen, pp.
281-82). (Robertson subscribed also to a Rousseau-like division of
societies into savages, barbarians, and civilized peoples, though he
located the Aztecs somewhat higher on this scale than Rousseau did.)
It remained only for a rearguard action to defend the Mexicans,
emphasizing not only ideographic symbolism in their glyphs but also
an incipient phoneticism that had long been recognized in some of the
codices. This deciphering of Mexican rebus writing was the work
especially of J.-M.-A. Aubin (Mimoire, 1849; see Keen, p. 340). It
brought the conceptualization of Mexican writing more or less to the
point where it remains today in accounts like those of Miguel
Le6n-Portilla, with their untroubled distinctions of pictographs,
ideographs, and phonetic glyphs."'
It is perhaps needless to emphasize that views like Le6n-Portilla's
have the power only to shift the Aztecs slightly from one position to
another on Rousseau's scale of linguistic evolution. They have no
power to undermine the authority itself of the scale. To do this we
need to think again of the supplementary play in Derrida's notion of
writing, which loosens the bond of writing to speech in the Western
relation of alphabetic phonography. Writing in this conception can
come into meaning through a diffirancethat engages orders of reality
other than speech differently and more directly than alphabetism.
Pictography exemplifies this other dififrance.In a broadened conception of writing it does not take its place in an evolutionary queue,
patiently awaiting the forces of enlightenment that will nudge it
toward phoneticism. Instead it finds its own set of valences between
utterance and the world. We will return below, in an analysis of
speech and song glyphs, to these valences as they may have formed in
the Mexican world."
"oLe6n-Portilla discusses Aztec writing in several works; for a recent overview see
his study The Aztec Image of Self and Society:An httroductionto Nahua Culture, ed. J.
Jorge Klor de Alva (Salt Lake City: University of Utah Press, 1992), 52-55.
" The history of the development of Mesoamerican writing systems may be seen
to reverse the logocentric teleology, that is, it moves in the broadest view from the
phoneticism of Maya glyphs to the generally iconic usages of Mixtec and Aztec
writing. For the outlines of large cultural shifts that may have militated for this
history, see John M. D. Pohl, "Mexican Codices, Maps, and Lienzos as Social
Contracts," in Writing without Words,ed. Boone and Mignolo, 137-60, esp. 155-56.
From such outlines scholarly views of the cultural efficacy (rather than inadequacy) of
post-Classic Mesoamerican writing are beginning to emerge.

358
SOCIETY
Mexican song and poetry too were brought under the sovereignty
of logocentric thought. More precisely: Mexican song was incorporated into this reign by means of the European invention of Aztec
poetry and music. The logocentric repression of song's ambivalent
supplementarity was the driving force of this invention. It came more
and more to enforce, across the seventeenth, eighteenth, and nineteenth centuries, a Europeanized view of Mexican song as a compound
of speech and a separate, more or less straightforwardaddition to it. It
distributed Mexican song across two familiar Western discursive
practices. In the process it obscured less familiar ties to other forms of
utterance and to the world that might have made Mexican song
meaningful in its indigenous contexts. It ushered what is most
accessible to the West in Mexican song into the realm of poetry,
aligned in the early-modern period especially with oratory and in later
centuries more and more with a new category, Literature. (It is a
poignant irony that the voices of a civilization so often dismissed for
its lack of letters should ultimately be coopted under the imprimatur
of Literature.) In so doing, the repression of supplementarityset apart
the elements of Mexican song "added" to speech-its songish elements, so to speak-as a substandard, primitive, and in any case
essentially unknowable form of another Western category, Music.
This Western dissection of Mexican song set in slowly after the
first Euro-American contacts. Some of the earliest Europeans in
Mexico, men like the clergyman Vasco de Quiroga, could hear in
native song echoes of the pastoral Golden Age that would later
unsettle Rousseau's evolutionary schemes. The most substantial accounts from the sixteenth century, some of them almost ethnographic
in their on-the-scene observation and detailed reportage on society
and culture, adopt a related stance. Toribio de Benavente alias
Motolinfa, Bernardino de Sahagvin,Diego Durhin,and others of this
period depict Mexican song as an activity integrated in a broader
setting of (idolatrous) festival, ceremony, and religious observance.
They tend not to use words like "poetry" and "music," preferring to
call the songs cantaresand to refer to the usual ensemble of dance,
song, and drumming by means of the Taino-derived word areito."
" For Quiroga's references to Mexican song, from his legal brief of 1535 entitled
"Informacion en derecho," see Vasco de Quiroga, La Utopia en Amirica, ed. Paz
Serrano Gassent (Madrid: Historia 16, 1992), 218; also Keen, The Aztec Image, Io8.
For Motolinia's account see Memorialesde Fray Toribiode Motolinia, ed. Luis Garcia
Pimentel (Mexico: published by the editor, 1903), bk. z, chap. 26; although it
remained unpublished until Pimentel's edition, this account was widely dispersed,
being echoed in Francisco L6pez de G6mara's Istoriade las Indiasy conquistade Mexico

359
Already in the waning sixteenth century, however, perhaps as the

shock of unfamiliarity of Mexican society wore off, European observers began to graft their own categorical distinctions of poetry and
music onto autochthonous practices. Diego Mufioz Camargo, writing
his Historia de Tlaxcalain the last quarter of the century, assimilated
the Nahuatl language to European poetic traditions, extolling it in
part because "one can easily compose verses in [it] according to the
rules of meter and scansion" (quoted in Keen, p. 129). Michele
Zappullo, writing in 1603 in the vein of condemnatory writers like
Francisco L6pez de G6mara, asserted that the Mexicans possessed
neither music nor letters, in one motion dismissing the Mexicans and
imposing on them the European division of the two (see his Historiedi
By the
quattroprincipalicitt? del mondo,cited in Keen, pp. 140-41).
middle of the next century the Europeanized view was prevalent.
Eguiara y Eguren celebrated "the love of the Mexicans for poetry and
oratory" (quoted in Keen, p. 224). And Joseph Joaquin Granados y
Gailvez,in his dialogues of 1778 entitled Tardesamericanas,discussed
separately and at some length ancient Mexican music and poetry. On
the one side he praised the consonance and melodiousness of the
music, its ensemble of voices and instruments ably concerted so as not
to overwhelm the words, and its varied and delightful melodic figures
capable of inducing ecstasy in its listeners. He compared the Mexican
musicians to celebrated musicians of ancient Greece, and he distinguished three styles of Mexican "compositions,"warlike, pathetic, and
grave, that clearly enough harken back to Platonic descriptions of
modal ethos. Although he could not find in Mexican poetry all the
genres practiced in the ancient European world, he succeeded in
locating at least the "mathematical" (i.e., astronomical) style of
of I552 (see Robert Stevenson, Music in Aztec and InzcaTerritory[Berkeley and Los
Angeles: University of California Press, 1968], 105) and taken over almost verbatim
in Juan de Torquemada's massive Monarchiaindianaof 6i 5 (see the facsimile reprint
of the 1723 edition [Mexico, 1943], bk. 14, chap. ii). Sahaguin'sremarks on song are
scattered through his huge Nahuatl compilation known as the FlorentineCodex(see
Fray Bernardino de Sahaguin,FlorentineCodex:GeneralHistoryof the Things of New
Spain, ed. and trans. ArthurJ. O. Anderson and Charles E. Dibble, 12 vols. [SantaFe,
N.M.: School of American Research and University of Utah Press, 1950-82]) and his
Spanish Historia general de las cosasde Nueva Espafa, ed. Angel Maria Garibay K.
(Mexico: Editorial Porrfia, 1956), abstractedfrom the FlorentineCodex;for particularly
rich descriptions of song and dance in situ see the accounts of merchants' banquets in
either work, book 9, chaps. 8-io. For Durain'saccount, from his Historiade las Indias
de Nueva-Espaia e islas de la tierra firme, see Stevenson, Music in Aztec and Inca
Territory,109-io. Stevenson's book gives a useful overview of many early accounts on
pp. 85-120.

360
SOCIETY
Manilius, the tragic and comic modes of Seneca and Euripides, and
the heroic style of Silius Italicus.13
By the end of the nineteenth century, finally, in the wake of
intensive new source studies led by Eduard Seler and Francisco del
Paso y Troncoso and with the recovery, of paramount significance, of
the manuscript of alphabetized Nahuatl song texts entitled Cantares
mexicanos,Mexican utterance was well on the road to consolidation
under the aegis of Literature. The capstone of this whole logocentric
co-option and domestication would wait another half century. Father
Angel Maria Garibay K.'s monumental Historiade la literaturandhuatl,
published in 1953-54, divided and classified prehispanic Mexican
speech/song according to the Western genres that had defined
Literature in the first place: religious poetry, dramatic poetry, epic
poetry, lyric poetry, didactic prose, historical prose, and fiction
("imaginativeprose"). Meanwhile pre-Hispanic "music"was left in its
own Western camp, the province of conservatory-oriented orchestral
composers with a nationalist and archaeological bent like Carlos
Chaivez.
The European invention of Aztec music and poetry depended on
the denial of song's supplementary dualism, on the preeminence of its
supplementing-by-addition to speech and the submergence of its
supplementing-by-substitution for speech. But this second diffiranceof
song was not entirely repressed;its propensity to surface unexpectedly
is signaled, after all, in the possibility of views like Vasco de Quiroga's.
Another such surfacing occurs in the Idea de una nueva historia
general de la America septentrionalof Lorenzo Boturini Benaducci
(1746), a precis for a larger work on ancient Mexico never completed.
The extraordinaryimportance for Mexican studies of Boturini's work
was signaled at some length by Benjamin Keen in The Aztec Image in
Western Thought; it remains to be fully explored. Moreover, the
significance of the Idea extends well beyond its subject matter: it is a
bellwether of European conceptions of the speaking, singing, and
writing of others. Boturini's small book marks the advent in European
studies of the Americas of a full-fledged Rousseau-like aporia concerning song.
Boturini did not entangle himself in song's supplementary ambivalence by direct contact with Rousseau's thought but ratherby reading
an earlier work important in forming Rousseau's ideas as well: the
Scienza nuova of Giambattista Vico (I725). Boturini borrowed from
'3 See Joseph Joaquin Granados y Gilvez, Tardesamericanas:Gobiernogentil y
catolico:Brevey particularnoticiade toda la historiaindiana (Mexico, 1778), 88-90.

361
Vico a whole historical philosophy that divided societal evolution into

three ages, the divine, the heroic, and the human or rational. He took
over the idea that all societies pass, independently of one another,
through these ages in their evolutions and thus that all societies share
a common developmental history. Armed with these notions Boturini
proclaimed the autonomous achievements of Mexican culture and
combatted diffusionist views, current in his day, that linked them all
to some time-shrouded dependence of Americans on Europeans.
Most importantly, Boturini borrowed Vico's conception of knowledge in the divine and heroic ages as distinct stages in a pre-rational
"poetic wisdom." This wisdom, in Vico's view, shaped every aspect
of society and thought-logic,
morals, economy, politics, and
to
of
some
Vico's
physics,
give
categories. It took the forms of
and
poetry, fables,
hieroglyphic writing, and was superseded by
natural reason only in the third, human age. Following the lead of
Vico's poetic wisdom, Boturini in his Idea gave an emphasis to
songs and pictographic writings that was probably unprecedented
in earlier Mexican studies.'4
But Boturini did not accept Vico's estimation of the historical age
that American societies had attained at the time of the European
invasions; from this his difficultiessprang. Whereas Vico regarded the
Americans as still in the heroic age, Boturini wished to portray them
(or at least the Mexicans) as more advanced. He located all three of
Vico's ages in the history of Mexico, dating the last of them from the
establishment of Toltec society in 660 C.E.In the process he extolled
the sophistication of their writing, the truths garbed as fables in their
pantheon of gods and their teoamoxtli or divine book, and the
refinement and subtlety of their poetry. On this last point he was
unequivocal; once he completed his history, he claimed, European
"poets would drink, in the cantares,the nectar of the Indian Parnassus." Boturini devoted most of a lengthy chapter to the Mexican
cantares.He ascribed them to poets of the heroic age, distinguished
their different genres according to European models, and described
'4 Keen, pp. 225-37, discusses Boturini's work and his dependence on Vico. For
Boturini I use the facsimile edition of the Ideade una nuevahistoriageneralde la America
septentrional(Paris: Genet, 1933). For Vico's work see The New Scienceof Giambattista
Vico,trans. Thomas Goddard Bergin and Max Harold Fisch (1968; reprint, Ithaca:
Cornell University Press, 1988). Poetic wisdom is the primary subject of the Scienza
nuova;for an axiomatic sketch of it, see The New Science,bk. i, sec. 2; for a filling-in
of this sketch, see bk. 2. For Rousseau's dependence in his Essay on Vico, see The
Autobiographyof GiambattistaVico, trans. Max Harold Fisch and Thomas Goddard
Bergin (Ithaca: Cornell University Press, 1944), 72-73.

362
SOCIETY
the poetic means, including their "elevated metaphors and allegories,"

that made them "una optima Poesfa." And he singled out for special
praise the poems of the Texcocan king Nezahualcoyotl.'5
Boturini's dilemma is all too clear: How can the Mexicans have
traversed some eight centuries of their Viconian human and rational
age and still embody the truths of their culture in the pre-rational
form of poetic wisdom? How can primordial poetic wisdom display all
the aestheticized refinement and evolutionary unfolding of "una
optima Poesia"? How can a voice like Nezahualcoyotl's represent so
manifestly the poetic, heroic age almost a millennium after the end of
that age? Vico portrays song unilaterally, as the collective voice of the
peoples of divine and heroic societies. Boturini, in wishing both to see
it as such and to assert the rational attainments of the Mexicans,
instead broaches the supplementary dualism of song that would soon
plague Rousseau. The cantares, for him, bear the dual burden of
aboriginal utterance and post-rational refinement.
Metaphor is the final Western construct that has captured and held
Aztec utterance. The earliest European notices of Mexican songs
already called attention to their frequent and difficult metaphors.
Diego Durninwrote that "all their cantaresare made up of such obscure
metaphors that there is hardly anyone who understands them."
Sahagfin also reported on the obscure meanings of songs he knew,
though for him this amounted to Satan's attempt to conceal their
idolatry.He reservedthe wordmetaphoras-andthe Nahuatlcoinage

(approximately"sign-speech")that he used intermachiotlahtolli
it-for his admiringobservationson the huehuetlahwith
changeably
tolli,the "elders'talk"or ritualisticspeechesdeliveredon variouscivic
and domesticoccasions.The mestizochroniclerJuanBautistaPomar,
writing his Relacionde Tezcocoin 1585 and probably himself the
delossen"ores
dela
compilerof a collectionof cantaresentitledRomances
NuevaEspafia,closely relatedto the Cantaresmexicanos,
admiredthe
many truths to be gleaned from such songs but admitted that it
required a "great linguist" to understandtheir images. Another
mestizo writer, Fernandode Alva Ixtlilxochitl,a descendantof the
kingsof Texcoco andthe firstgreatpropagandistfor theirwisdomand
learning, absorbedEuropeanconceptions of poetry and figurative
language from the Franciscansat their College of Tlatelolco. He
'5sFor Vico's views on Americans, see for example The New Science, 116-17,
142-43, 414. Boturini's "IndianParnassus"remark is on p. I6o of the Idea;the cantares
are discussed in chapter 15. For one of the earliest accounts of Nezahualcoyotl's
songwriting abilities, see Torquemada, Monarchiaindiana, bk. 2, chap. 45-

363
described the songs as "very obscure, being allegorical in form and

adorned with metaphors and other figures of speech."'6
Later writers continued to call attention to the metaphors of
Nahuatl songs, whether admiringly'7 or with the view that they
reflected the roundabout expression of an impoverished language.'8
Particularly in the latter, negative assessment we are reminded of
Condillac on the imprecisions and pleonasms of less-than-fullycivilized languages. In the coexistence of both views we are brought
again to face the duality of metaphor-as-supplement. Figurative
language is a pre-rational confusion of savage and barbarous tongues
and, at the same time, a civilized transcendence of reason in poetry.
(Or again, in Rousseau's imperious attempt to resolve the dilemma, it
is a natural impulse not meriting the name "metaphor"until brought
under European control.)
It was left for Garibay to exalt the positive interpretation of
Mexican metaphor and dispel the last suspicions of Nahuatl expressive
inadequacy. He did so, not unexpectedly, by bringing Nahuatl metaphors under the umbrella of a Western aesthetic made universal (in
language that looks all the way back to the seventeenth-century
theorist of metaphor Emanuele Tesauro): "The metaphor is the
mother of all beauty. In essence it comes to be the nucleus of all
poetry. ... Nahuatl poems teem with [metaphors]."'9But in effect this
only reiterates Rousseau's evasion of metaphor's supplementarity
through the arrogation of all metaphor to Europe. Metaphor, for
Garibay, is not a speech-act that might be rethought at or beyond the
boundaries of Europe's "poetry."It is, instead, the very engine of the
poetic ideology.
And it is a powerful engine. Garibay's conception of Nahuatl
metaphors has been dispersed by now through many accounts,
x6 Durain'squotation comes from his Historiade las bidias de
Nueva-Espania;I take
it from Angel Maria Garibay K., Historia de la literatura ndhuatl, 2 vols. (Mexico:
Editorial Porrua, 1953-54), i:74; see also the translation of this passage in its larger
context in Stevenson, Music in Aztec and hIca Territory, o09-o0. For
see his
FlorentineCodex,ed. Anderson and Dibble, i:8i. Pomar's remark is Sahagtin
quoted by John
Bierhorst in Cantaresmexicanos:Songs of the Aztecs (Stanford: Stanford University
Press, 1985), 17; Ixtlilxochitl's Relacioneshistdricasare quoted by Keen, p. 199.
'7 See for instance Boturini, Idea, 2, 6, 88, 96, 162; Granados y Gailvez, Tardes
americanas,94; and Thomas F. Gordon, Histoiy of Ancient Mexico (1832), cited in
Keen, The Aztec Image, 350.
'8 For example Francisco G. Cosmes, La dominacidnespaiiolay la patria mexicana
(Mexico, 1896); Eugene-Emmanuel Viollet-le-Duc, in Desire Charnay, Citeset ruines
americaines(Paris, 1863); and EdwardJ. Payne, Historyof theNew WorldCalledAmerica
(London, 1904); all cited in Keen, The Aztec Image, 435, 438, and 445 respectively.
'9 Garibay, Historia 1:76; my translation.

364
scholarly and popular, of Aztec culture. To judge from these, its

attraction is the result especially of a specific technique Garibay
identified as the essence of Nahuatl figurative language. He called this
technique difrasismoor diphrasis."'It is the joining of two concepts in
grammatical conjunction to signify a third: for example, in atl in tepetl,
literally meaning something close to "water and hill" but metaphorically "town" or "settlement." The most famous of the difrasismos
Garibay identified is an expression that, for his followers, came to
signify poetry itself in indigenous culture: in xochitlin cuicatl,approximately "flower and song."
In this technique of diphrasis Garibay banished any negative
valuation of Nahuatl pleonasms. He upheld the literary status and the
figurative subtlety of Mexican "poetry" and even found for it an
indigenous name. Some of his followers have gone farther. Writers
like Birgitta Leander, Luis Alvelais Pozos, and, most importantly,
Miguel Le6n-Portilla have seen in the metaphorical naming of poetry
the reflection of a whole indigenous humanism and aestheticism.
They have linked this view of life built around xochicuicatlto the great
university supposedly sponsored at Texcoco by Nezahualcoyotl and
his son Nezahualpilli, the story of which is traceable back through
Boturini's Idea and other sources to Ixtlilxochitl. And they have seen in
the figures of these Texcocan rulers poet-philosophers embodying
universal wisdom in their songs and opposed to the militarism and
sacrifice of neighboring Tenochtitlan. In David Carrasco's summary
of this interpretation, the poet-philosophers "used language, instead
of blood, to communicate and make offerings to the gods.... They
preserved honored traditions, produced and read the painted manuscripts, and developed refined metaphors and poems to probe the true
foundations of human existence." They devised "techniques to open
the depths of the human personality to the illusive world of truth. The
main technique was the creation of in xochitl,in cuicatl,or flowers and
"
songs.
2o Ibid., 1:18-19.
For an influential statement of Le6n-Portilla's views, see his Aztec Thoughtand

Culture,trans.Jack Emory Davis (Norman: University of Oklahoma Press, 1963), esp.
74-79; also Birgitta Leander, In xochitl in cuicatl:Flor y canto,la poesiade los Aztecas
(Mexico: Instituto Nacional Indigenista, 1972), and Luis AlvelaiisPozos, Los cantosde
Nezahualcdyotl(Toluca, Mexico: Instituto Mexiquense de Cultura, 1989). For Carrasco's words see his Religionsof Mesoamerica:Cosmovisionand CeremonialCenters(San
Francisco: Harper, 1990), 79-80. For Boturini's description of the university at
Texcoco, see his Idea, 142; for Ixtlilxochitl's, his Historia de la nacidnchichimeca,ed.
Germin Visquez (Madrid:Historia 16, 1985), 136 (chap. 36); and for Torquemada's,
his Monarchiaindiana, bk. 2, chap. 41.
2"

365
The linchpin of this construction of Aztec poetic humanism has

been the Cantaresmexicanos,the most important of the few collections
of song texts in alphabetized Nahuatl surviving from the sixteenth
century. This is true not least because the headings of some of the
cantareshave encouraged some scholars to consider them the work of
the Texcocan kings and other like-minded rulers. It is true also
because the Cantaresmexicanosis a chief source, along with the other,
closely related collection of song texts mentioned above, the Romances,
of the diphrasis and xochicuicatlimagery singled out by Garibay and
basic to Le6n-Portilla's poetic humanism. So even those scholarswriters like David Damrosch-who do not put much stock in the
ascriptions of particular cantaresto specific pre-Hispanic rulers nevertheless tend to ground their readings of the texts in Le6n-Portilla's
and Leander's humanism, aestheticism, and belles-lettrism.22
Among post-Garibay writers on the cantares,only John Bierhorst
has seriously challenged the ideology of metaphor and poetry imposed
on them. In his transcription, translation,and analysisof the song texts
he advanced an interpretation of them as "ghost-songs" intended, in
the manner of certain more recent native North American practices,
to revive spirits of dead heroes to combat the Spaniards. Bierhorst
dismissed the pre-Hispanic authorship of the cantares,dating most of
them from well into the colonial period. He questioned the tradition
of ascribing poems to Nezahualcoyotl and other pre-Cortesian luminaries. He even doubted whether the use of in xochitlin cuicatlto mean
"poetry" reaches any farther back than Garibay himself.23
While Bierhorst's transcriptions of the cantareshave been recognized as authoritative, his translations and ghost-song interpretation
were immediately controversial. James Lockhart and Le6n-Portilla
both have condemned them, and they have found little general
acceptance.24 Largely forgotten in the hubbub of interpretive particulars, however, has been the value of Bierhorst's project as an
ideological corrective to earlier views. Here at last is an attempt to
understand the cantaresas an indigenous colonial discourse, a discourse of resistance in which the Mexica tried to construct an
22
See David Damrosch, "The Aesthetics of Conquest: Aztec Poetry before and
after Cortes," Representations
33 (1991): 101-20.
23
Bierhorst, "General Introduction," Cantares mexicanos,esp. 17, 97-99, and
o13-924 See James Lockhart, "Care,
Ingenuity, and Irresponsibility: The Bierhorst
Edition of the Cantares Mexicanos," Reviewsin Anthropology16 (1991): 119-32; and
Miguel Le6n-Portilla, Fifteen Poets of the Aztec World (Norman: University of
Oklahoma Press, 1992), 41-44.

366
OFTHEAMERICAN
SOCIETY
MUSICOLOGICAL
JOURNAL
efficacious vision of the strangers among them. In this interpretation

Bierhorst seeks to understand the cantaresas embodiments of otherthan-European ideologies. Admittedly his effort imposes on the
Mexica more recent discourses of resistance native to North America;
in moving against the overlaying of European ideology on the cantares
he substitutes another ideology not necessarily so much closer to the
Mexica, with little evidence for doing so other than his willful reading
of the texts themselves. But the importance of his gesture should not
be underestimated.
The supplementary analysis of metaphor opens a space where we
can see other alternativesthan Bierhorst's to the European captivity of
the cantares.To come up against Rousseau's aporia is to begin to sense
the westernizing power of metaphor and to question the preeminence
it has had in Western conceptions of Aztec figurative language. It is to
scrutinize the relations of language to the world imposed by the
discovery of metaphors in the speech of others. It is to wonder
whether the dichotomy of literal and figurative all told is as transparent and universal as the preeminence of metaphor makes it seem. All
of which issues a question of some importance to postcolonial
colonialism,with its
historiography: Has the historyof European-American
been
a
also
Western
enforcementof
metaphysics,
history
of the impositionof
views
and
the
world
on
metaphorical
of language,psyche,
diferent indigenous
views?25
Let us retrace another route toward the same question. The
poststructuralistview of language as infinite sign-chain, with meaning
produced in deferral from one sign to another, might also challenge
the preeminence of metaphor in Western conceivings of others. In its
tropological operation metaphor would seem to match better the leap
from signifier to signified of Saussurean semiology than the slippage
from signifier to signifier of grammatology. The very proximity of the
signs between which meaning arises in grammatology suggests that
another trope will be more broadly helpful in comprehending different language uses, a trope of displacement from one entity to other
contiguous ones: metonymy. The pervasive metaphoricity of modern
Western culture might well be a measure of our inability to perceive
2s One recent writer who would answer this question with a resounding "yes" is
Eric Cheyfitz. In The Poeticsof Imperialism:Translationand Colonizationfrom "The
Tempest"to "Tarzan"(New York: Oxford University Press, I991) he opposes views
broached in recent years that metaphor works as a figure of intercultural communication, and he traces some of the ways European perceptions of and actions toward
indigenous Americans were driven by metaphorical conceptions of language. See
chap. 6, esp. pp. 0o4-9.

367
propinquities of the world apparent in many other cultural settings.

Perceiving these contacts between things might be a matter of
traversing the world linguistically in a different way than we customarily do, of linking signs with signs rather than separating them,
indeed finally of collapsing together the realm of signs and the realm
of things. A shift of emphasis from metaphor to metonymy might
begin to enable some of these other shifts.
The Cantares Deferred
To pursue this shift to metonymy we might start from the Cantares
mexicanos and summarize some of the effects of the most basic
European technology at work in them: the technology of the alphabet.'6 The transformation of spoken or sung Nahuatl into alphabetized words-performed utterance into fixed inscription-enforces
various regimes of Western writing on the Nahuatl songs. At the local
level these include a solidification of the distinction and individuality
of single words that is alien to the plasticity and flexibility brought
about in spoken Nahuatl by complex procedures of agglutination and
compounding; a tendency to arrange words according to the dictates
of Latinate syntax; a crystallization of the inscribed space so as to
create a desired "poetic form"; and a coercive and distorting distinction of semantic elements from "meaningless"vocables.
Let me briefly pursue the last of these points, by way of exemplifying how a new kind of engagement with the cantarestexts might
challenge Western assumptions that have heretofore been unquestioningly applied to them. Table i gives the beginning of one of the
cantaresin two formats:in a transcriptionadapted from Bierhorst's and
in Bierhorst's translation. In the transcription I have inserted underscoring borrowed from an "analytical transcription" of the song
Bierhorst published in a companion volume to the Cantaresmexicanos.27These underlines indicate what Bierhorst, in the introductory
notes to his analytical transcription, calls "meaningless song syllables,
or vocables"; as a glance through that transcription shows, they are
found frequently in almost all the cantares.But they are omitted from
his translation, as indeed they were omitted from Daniel Brinton's
26
Here this account will be brief. In forthcoming work I will enlarge on it through
a closer reading of the cantares.
27 See Bierhorst, A Nahuatl-English
Dictionaiy and Concordanceto the Cantares
mexicanoswith an Analytical Transcriptionand GrammaticalNotes (Stanford: Stanford
University Press, 1985), 501-2.

TABLE I
Beginning of Song 44 from John Bierhorst, Cantaresm

Transcriptionwith Underscored"SongSyllables"
I
2
Nican ompehua Teponazcuicatl.

Tico, tico, toco toto, auh ic ontlantiuh cuicatl Tiquiti
4
5
6
Tollanaya huapalcalli manca noqan in mahmani coatlaquetzalli ya

quiyacauhtehuac Nacxitl topiltzin on quiquiztica ye choquililo in
toplThuanahuay ye yauh in polihuitiuh nechcan tlapallan ho ay.
7
8
Nechcayan cholollan oncan tonquigaya poyauhtecatitlan in quiyapanahuiya-yyacallanon quiquiztica ye cEo-quililoneta
12
Nonohualco ye nihuitz ye nihuiquecholi nimamali teuctla nicnotlamatia

oyahquin noteuc ye ihuitimali nechyaicnocauhyan i mi'tlacxochitl ayao ayao o ayya yyao ay.
In tepetl huitomi ca niyaychocaya axalihqueuhcanicnotlamatiya
I3
[marginal:]yehuayan
oyaquin noteuc etta
to
II
Tico tico to
titito titi.

In Tollan stood a hous

Nacxitl Topiltzin lef
bewailed with conch
Tlapallan.
Yonder you are passing
traverses, and Acalla
Now he goes to his des
I come from Nonoalco,
Gone is my lord Ihu
That the mountain coll
Gone is my lord Ihu
369
pioneering attempt to translate some of the cantaresa century ago.S8

In fact, the nonsemantic status of such syllables in the cantareshas been
accepted in all the commentaries on the songs I have seen.99
But what does it mean to deny meaning to these syllables? How
can we draw the line so confidently between significance and insignificance in Mexican song? The complications of doing so are not very
apparentin the case of syllables added to the end of a "verse"(see, e.g.,
line i i). But they come into full view when the syllables are prefixed,
infixed, or suffixed to sentence-words themselves. Such altered
Nahuatl sentence-words inhabit a liminal space between Nahuatl
sense and European nonsense that should undermine the confidence
with which we demarcate either, even in ostensibly simpler cases.
Look at five instances in Table i: in line 12 niyaychocayais derived
from nichoca,"I weep," by the infixing of yay and the suffixing of ya; in
line i o nechyaicnocauhyan
comes from teicnocahua,"he/she leaves
someone in sorrow," with the addition of ya and yan; in line 5
builds on tecauhtehua,"he/she goes away or dies and
quiyacauhtehuac
leaves someone"; in lines 7-8 tonquifayaand quiyapanahuiaare both
derived through the addition of ya from simpler sentence-words
meaning something like "he/she passes through/across" (quiza and
panahuia respectively). In all five cases the added syllables seem to
touch the sense of departure, passing, and bereavement that any
attempt at translation, however problematic, would recognize in this
song text. Far from being Bierhorst's "meaningless song syllables," in
other words, they sit at or near the semantic heart of the song. They
are extragrammatical, in this sense nonsyntactic, but profoundly
meaningful. Moreover other syllables to which Bierhorst assigns
meaning carry a much less distinct semantic charge than they do (the
om- of ompehuain the heading of the song, for example, is a variant of
the ubiquitous Nahuatl prefix on-, which carries a subtle suggestion
of rhetorical or directional emphasis). Distinct Western categories of
significance and insignificance are breached by such examples.
At the broadest level the alphabetizationof the cantaresworks to fix
these songs on a Western grid of relations between language and the
28
Daniel G. Brinton, trans. and ed., AncientNahuatl Poetry,Containingthe Nahuatl

Text of XXVII Ancient Mexican Songs (Philadelphia, 1890; reprint, New York: AMS
Press, 1969); for the present cantar see 104-529 See for example Leander, In xochitl hincuicatl, 53-54; Pozos, Los cantos, 32;
Frances Karttunen and James Lockhart, "La estructura de la poesia nihuatl vista por
sus variantes," Estudiosde cultura ndhuatl 14 (i98o): 15-64, esp. 23-28; and Miguel
Le6n-Portilla, "Cuicatl y tlahtolli: Las formas de expresi6n en nihuatl," Estudiosde
culturandhuatl 16 (1983): 13-Io8, esp. 36-37.

370
world. It connects the songs to a network of our assumptions about

reading including the dual nature of language, spoken (or sung) and
written, and the primacy, within the written, of alphabetic technology.
It encourages the comprehension of these songs as literature-written
poetry the loss of whose sung medium affects its meaning in no
essential way. And it embeds the songs in a system of graphemes that,
despite their own materiality, give themselves over as representing
reality rather than participating materially in it.
But, as I have stressed, the relations of other writing systems to the
world might well be different, and this difference might involve altered
degrees of material participation in the world. In the indigenous
Mexican context, while there clearly existed distinctions of speech
(tlahtolli) and song (cuicatl),of speech and painting (tlahcuilolli),it is
not clear that anything like our distinction of speech and writing
existed. Or that it could have existed, given the differing valences of
pictographs and alphabet.30 The relations between spoken or sung
enunciation and the pictographic codices might be conceived better in
other terms: as relations of speech/song and pictures of material
things, or speech/song and color, or speech/song and counting, or
speech/song and embodied memory-of-the-people. Moreover, while
our dichotomy of speech and alphabetic writing tends to divorce
language from the material world, in all these other dichotomies
pictographic inscription pulls vocal enunciation toward a material
immersion in the world: toward pictures of things sacred and palpable,
toward the differently colored realms and associations of Aztec
cosmogony, toward counting objects as opposed to manipulating
abstract numbers (Nahuatl numeration builds object-names into the
words for numbers in order to suggest the kind of object being
counted in a given circumstance), and toward the geographical history
of the migration and self-definition of the Mexica. In the most general
terms, asJohn Monaghan has recently emphasized in regard to Mixtec
codices, the relations of utterance and painting might need to be
rethought as connections of bodily gesture, of a whole system of
linguistic, pictographic, and danced choreography with an elaborate
semantics of its own.3'
30
On this topic see, in general, Gruzinski, La Colonisationde l'imaginaire,chap. I

and, for persistent cultural differences in the later colonial period, chap. 7.
3' See Monaghan, "The Text in the Body," esp. 87-91. Monaghan argues from
linguistic evidence that dance, in Mixtec understanding, was a kind of embodied
singing related to prayer; in thus merging dance with song Monaghan suggests that
our customary notion of dance is a Western category in need of a critique similar to
mine above of poetry, song, and so on. He does not, however, allow for any separation

371
In the light of these other affiliations we might reevaluate the

westernization of Mexican song by means of metaphor. Where there
is less space between language and the world than we habitually
presume, there might be all told less room for a figurative distance
between words and things. Language might function through its
material participation in the world instead of across the abyss between
it and the world.32 In this situation metaphor, a trope of overleapt
distance, might find little congenial soil and give way, once more, to
metonymy, a trope of contact and participation:not a metonymy that
figuratively betokens a world that is in reality different, but one that is
the cipher of the propinquities of the world and language, spoken,
sung, or painted-a metonymy that reveals the volumetric presence by
which all these forms of language participate in the world.
Such a metonymic conception might help us look anew at another
mode of inscribingMexican song, one closer to indigenous technologies
than the alphabetof the cantares:the song glyph. Figure I reproducesa
famous exampleof a song glyph from the Codex Borbonicus,paintedin
the valley of Mexico around the time of the Spanishinvasion;the glyph
in questionis the decoratedvolute or scroll extendingfrom the mouth of
the smaller deity pictured. Such elaborate volutes can be found in
Mesoamerican pictography at least as far back as the murals of Teotihuacan,dating from the seventh or eighth centuryC.E.33They are not all
specificallysong glyphs;in their degree of elaborationthey seem, in some
uses at least, to convey a gradation of utterance from plain speech
through heightened speech to song.
of song and speech within Derrida's analysis of logocentrism: "But if we are now
coming to the realization that these [Mixtec] texts cannot be understood apart from
their performance, the focus on song and chant betrays our logocentricity. Were these
texts only put into words? Is that the way they were 'read'?"(p. 89). In this he misses
the supplementary role of song in regard to speech I have endorsed above. (For
further ramifications of such failure to pry song apart from speech in Derrida's
analysis, see n. 42 below.) A rethinking along Monaghan's lines for the Nahua
culture-areas would start from Alfredo L6pez Austin's magisterial The Human Body
and Ideology:Conceptsof the Ancient Nahuas, trans. Thelma Ortiz de Montellano and
BernardOrtiz de Montellano, 2 vols. (Salt Lake City: University of Utah Press, 1988);
for a study of similar issues in Inca culture with suggestive remarkson sound and song,
see Constance Classen, hIca Cosmology
and theHuman Body(Salt Lake City: University
of Utah Press, 1993), esp. 68-7332 The relation between these
thoughts and my earlier work on Renaissance occult
traditions is clear. In Music in RenaissanceMagic: Towarda Historiographyof Others
(Chicago: University of Chicago Press, 1993) I discuss views of language that set it
clearly apart from the world it describes as emergent early-modern alternatives to
linguistic views basic to magical thought.
33See for examples Kathleen Berrin and Esther Pazstory, eds., Teotihuacan:Art
firomthe City of the Gods(New York: Thames and Hudson, 1993), 198-99.

372
i
?...
ii~~i~i!iiiiiiiiiiiii!iiii~iiiiiiii!i!iiiil''
:~iiiiiiiSiiii!
:'i
iiiIn
iiiiiiiii
, ........
N
~iiliiiiiiil~ii
..................:iiiii',iiiiiiiii~iii~
............a',i4
::::::::::::::::::::::
::::::::::::
:::::::::::::::::::::
li'~e
4s
iiiiiiiiiii
iiiiiiiiiiii-iiiii~
il
4;?::i'iii~il';ii:iiji:liiiiii
iiiiiiiiiii.....
....
::=:::i:ili:!iiii
.....
:::
EMMI::i::
iiii
ciiii"
i iiijij
i:i:-i:i~i:-:::::::
!iiii ii:ii
iiiiiai:!iiii~i!ii
::iiiiii::iii::i:.iiil
i::ii:i
i
i~i4ii!!iiiZiii!ii!iiiiiiiii:i::~~~:::::i:iiijii-~
~~iiiiii~~:;i-:iii~~
.._:::::::::::Iw::~:~:-i
::::::::::r::::a::::
::
:::::::::::-li:iiiiiii
o"::::
~:::::::::-::::.j::?Ag
:8i:i~i:::
.-:-_-:::::::.::
:_..:.-:
:l::::-:ini":i
::*i:~i..._
......................i
::WIN
Figure I. Song glyph, Codex Borbonicus (Bibliotheque de l'Assemblee Nationale

Franqaise, Paris, Y120), p. 4; printed by kind permission of Akademische Druck- u.
Verlagsanstalt, Graz, Austria
But I am increasingly convinced that they convey something else

more distant than this from our notions of speech writing or music
writing. They depict the palpable, worldly volume of indigenous
utterance. Their effectiveness as music writing-let us for the moment
think of the glyph in Figure i as "music notation"--does not reside in
their answering to Western requirements of such writing. They do
not enable us-and were not intended, I think, to enable those who
first viewed them-to re-create more or less precisely some specific

373
speech-act (or, better, song-act). But to demand that they do so is only

to constrain once again the Mexican inscription of song in logocentric
bonds; it is to demand that inscription aspire teleologically to the form
of a representation of speech from which speech might be exactly
reconstructed. I think the glyph in the Borbonicus answers to different
affiliations than these and embodies other powers of writing. It
conveys through the tangible medium of paint, itself laden with
cultural values and meanings we are only beginning to understand, a
meeting of sung language and the world in their coextensive material
substance. It is not properly musicnotationbut musicsubstantiation.34
In the manuscript of the Cantaresmexicanosthere appears a music
writing closer than this to European expectations. It takes the form of
percussion cadences, indicated by the four syllables ti, to, qui, and co,
between the headings of some of the songs and the song texts
themselves (as for instance in Table i above). We do not know much
about the origins of this syllabic notation. It was clearly dispersed
more widely than in this and the related manuscript, the Romances,
where it appears-widely enough so that by the middle of the
seventeenth century a whole genre of colonial Mexican song, the
tocotin,could be named after it.3s It may well reflect pre-Hispanic oral
mnemonic practices for teaching, disseminating, and preserving percussion cadences.
But even if it does look back on these nonalphabetic origins, the
syllabic system is subtly transformed by its alphabetization. In its
alphabetized form in the cantaresmanuscript it comes to be perceived
as the product of an absence created by the alphabetization of the song
texts themselves. Only their inscription, after all, pries their words
apart from their sung medium, which comes then to require a writing
of its own. Their writing presses the demand for a music notation
answering, however imperfectly, to the European requirement of
song-act reconstructibility. In their phonetic writing they prescribe a
role for music writing. Music notation in the West is fundamentally
defined by this bond to and alienation from alphabetic "word notation"; it is the roduct of a gap (and hence a felt lacuna) introduced by
the alphabet.3 On the other hand, in the music notation of the Codex
develop this point at greater length in "Unlearning the Aztec Cantares."
For seventeenth- and eighteenth-century accounts of tocotines, see Bierhorst,
Cantaresmexicanos,88-90; and Stevenson, Music in Aztec and Inca Territory,i65-66.
36 Leo Treitler adumbrates this view in stressing that the earliest music notation
of Gregorian chant aimed not to define a sequence of pitches but instead to write the
melodic inflections of individual syllables of sung words; see "The Early History of
Music Writing in the West," this JOURNAL35 (1982): 237-79, esp. 244. These
34I
35

374
Borbonicus, the song-volute, there is no such absence. There song is

rendered complete, in its full indigenous materiality, but without
response to any Western demand for precise reconstitution.
Writing down the syllabic percussive cadences, however, brought
them within the gravitational control of European inscriptive technology. In doing so it pervaded them with a European verdict of their
inadequacy. The Western expectation that word- and music-writing
would aspire to the telos of more-or-less exact speech-act reconstructibility was cleanly mapped onto a system responsive instead to
indigenous contexts of ritual preservation and oral/pictographic dissemination. This mapping resulted in the transformationof a Mexican
music enunciation (the orally transmitted syllabic cadences) into a
European musicnotation,and carried with it the corollary judgmentalmost inevitable, it would seem-that the one is really only a failed
version of the other. The sense of this failure suffuses the many
scholarly attempts that have been made over the years to "decipher"
the syllabic notation of the Cantaresmexicanos,to find out "what it
really means" so as to reconstruct "how they really played."37
The European control exercised by the writing of the percussive
syllables in the Cantaresmexicanosmight be dramatized by comparison
with an even more explicit irruption of European music notation
concerning another American locale, this one also linked to an
attempt to alphabetize indigenous song. This is the account of Tupi
singing from Jean de Ldry'sHistoired'un voyagefait en la terredu Bresil,
first published in 1578 and reprinted frequently thereafter. Recently
Lery's book has been a cartede visite of sorts for fascinating meditations on European encounters of Americans by Michel de Certeau,
Frank Lestringant, Stephen Greenblatt, and others. De Certeau in
particular places the passages recording Ldry's ravishment by Tupi
song at the heart of his essay "Ethno-Graphy." He argues that these
moments in L6ry's experience of the Tupi betray in paradigmatic
fashion a presence of others' voices that transgressesethnographic and
historiographic attempts to capture and discipline them through
writing. As de Certeau puts it, we cannot fully "replace with a text
what only a voice that is other could reveal." The bodied voices of
inflections were absent, of course, from the alphabetic writing of the chanted words
and syllables-an absence recuperated, then, by music notation.
37 See for examples Garibay, Historia i:8o-8i; Samuel Marti, Canto,danzay mzisica
(Mexico: Fondo de Cultura Econ6mica, 196i), 140-48; Stevenson, Music
precortesianos
in Aztec and Inca Territory,46-53; and Bierhorst, Cantaresmexicanos,72-78.

375
others will always escape full domestication in our writing.38

De Certeau's conception intersects in numerous ways (if at times
obliquely) with Derrida's critique in Of Grammatology.He shares with
Derrida the broad notion that a set of relations between writing and
speech has formed Western categories for construing and controlling
others. Moreover, the escape he describes of others' voices from our
writing, the surplus by which they always exceed writing, might be
understood as the neat converse of writing's supplementary trumping
of speech in Derrida's account-a grammatology of speech, if you will.
But unlike Derrida, in making his dichotomy of speech and writing de
Certeau leaves little room for a dichotomy of speech and song. This
leads him to elide the two and effectively to deny any supplementary
status of song in relation to speech. While he subtly examines the
problems of the West's attempt to form its relations with others
through writing, he sees no qualitative difference in the difficulties
speech (on the one hand) and song (on the other) pose to alphabetism
as it makes this attempt to bring "primitive reality" into Western
discourse.39
L6ry himself, instead, registers the supplementarity of song, the
excess by which it transgresses speech, in the form of his writing: not
so much in the earliest editions of his travelogue, where the brief Tupi
song texts are set apart only in being italicized, but in later editions,
from 1585 on, where Lery introduced music examples in European
notation for the five Tupi songs. Here we see, as it were, a step-bystep realization of the lacuna opened up by phonetic writing, a move
from a distinction of song within alphabetization (by means of italics
for the sung words) to an attempt to close (with music notation) the
absence made palpable by alphabetization. But if, in the Cantares
mexicanos, an indigenous system is constructed as inadequate by
alphabetization,here, as de Certeau makes clear, it is European
notation that fails in the face of the other. L6ry's readers gain little
hint even from his extraordinarynotational recourse of how the Tupi
38 See Michel de Certeau, "Ethno-Graphy: Speech, or the Space of the Other:

Jean de L6ry," in The Writing of History,trans. Tom Conley (New York: Columbia
University Press, 1988), 208-43 (for the quotation, p. 212); Stephen Greenblatt,
The Wonderof the New World(Chicago:
"Introduction," in his MarvelousPossessions:
Frank
of
and
Press,
1991);
University
Chicago
Lestringant, "The Philosopher's
33 (1991): 200-2 i1. For
Breviary:Jean de Ldry in the Enlightenment," Representations
Lery's Histoire I have used the facsimile of the 1580 edition, edited by Jean-Claude
Morisot (Geneva: Droz, 1975), and the translation of it by Janet Whatley, Historyof
a Voyageto the Land of Brazil, OtherwiseCalledAmerica (Berkeley and Los Angeles:
University of California Press, 199o).
39 De Certeau,"Ethno-Graphy,"
223.

376
SOCIETY
singing could have awakened in him the strong emotions he describes:

fear, wonder, transporting delight. His notational technology cannot
keep Tupi song from escaping European control.
So, also in the 1585 edition, another European technology intrudes, this time a conceptual one. Hard on the heels of his description
of one of the songs, where he had in earlier editions suspected that the
singers were possessed by Satan, L6ry now breaks off to introduce
learned material on witchcraft and the witches' sabbath from Jean
Bodin's De la dimonomaniedessorciers.40Where the Tupi before had
been merely enraglesthey now become Dimoniaques,as if the deployment of a technical, even clinical, vocabulary (note the capital D)
could redress the limitations of a notational technique. The performative presence of song that escaped constraint by notation is
disciplined instead through an abominated European diabolism. The
analytic categories of this evil, more and more fully developed in
Europe through the sixteenth century, are made to compensate for the
inability of alphabetic writing and its attendant music writing to
capture performed utterance and justify affective response.
L6ry's gesture is the most common conceptual mechanism by
which Europeans sought to bring an indigenous American imaginary
under the aegis of a Christian supernatural. It has been analyzed by
several writers, most notably Serge Gruzinski for central Mexico and
Sabine MacCormack for Inca Peru.4' Its connection to song has
seldom been remarked, but I believe it runs deep; the automatic and
repeated European association of American song with things diabolical may tap the unsettling and finally uncanny transgression inherent
in others' singing.4' Here I wish only to point out that the whole
40 On L6ry's use of Bodin and for an interpretation along other lines of its
significance, see Greenblatt, MarvelousPossessions,i5-19.
See Gruzinski, La Colonisationde l'imaginaire;and Sabine MacCormack, Religion
4'
in the Andes: Vision and Imagination in Early Colonial Peru (Princeton: Princeton
University Press, i99I).
42 One writer who has
begun to analyze this association for Inca territories is Juan
Carlos Estenssoro Fuchs, "Los bailes de los indios y el proyecto colonial," Revista
andina 10 (1992): 353-89; I thank Nancy Farriss for bringing this important essay to
my attention. The transgression of others' songs is felt not only across wide
intercultural distances; it occurs in more narrowly European contexts as well. Its
intrusion in familiar repertories has, for instance, repeatedly animated Carolyn
Abbate's brilliant musings on nineteenth-century opera. There the transgression has
the force to thwart the monologic aspirations of a unilateral compositional authority,
to breach conventional narrative strategies, to "destroy language" and with it the
presence of enacted character, and even to undo the structures of power associated
with visually identified gender; see her Unsung Voices:Operaand MusicalNarrative in
the NineteenthCentury(Princeton: Princeton University Press, 1991) and "Opera;Or,

377
complex is related also to European regimes of writing. The excess of

song could be brought home especially forcefully by an absence at the
heart of alphabetism. At the moment of generating a musical notation-moments like those in the Middle Ages, of capital importance
for European music history, or like the one marked in successive
editions of Lery's Histoire-alphabetic writing declares this absence.
By this generation it moves to capture in its own inscriptive terms an
aspect of sung utterance that will always escape it. The escape is
uncanny, then, in part because of what it tells us about our speech and
our writing: it gives the lie, perhaps more forcefully than any other
kind of enunciation, to the illusion by which phonetic writing could
align itself with a Western metaphysics of presence in spoken speech.
Song contains a dimension of phonos, therefore an intimacy with
speech/logos, but this phonos remains always farther alienated from
alphabetic writing than the phonos of speech. Because of this distance
singing, chanting, and intonation in general mark experiences of
Europeans in America that cannot be reduced and domesticated, but
only exposed and exacerbated, by the exegetical technology the
Europeans brought. They are traces that undermine the place of
phoneticism in the metaphysics of Western language and thereby
open out a space for other disruptions and other arrangements of our
connections of language and the world.
the Envoicing of Women," in Musicologyand Difference:Genderand Sexualityin Music

Scholarship,ed. Ruth A. Solie (Berkeley and Los Angeles: University of California
Press, 1993), 225-58. Abbate's "unsung voices," which she analyzes from a Barthian
and narratological perspective and which repeatedly escape the cultural forms that
seem to give rise to and constrain them, could also be richly viewed in the light of
Derridean categories such as supplement and remainder, as I have suggested here in
rereading Derrida's treatment of Rousseau. Abbate's own reservations concerning the
usefulness of poststructuralist thought for her project (Unsung Voices,pp. 12-19)
spring from her sense that such thought (i) "evaporates"subjectivity and (2) insists on
analogies between music and language, thus submerging music's own, unique
phenomenological attributes in the play of literary theory; her instance of the second
tendency is Paul de Man's analysis of Derrida's reading of Rousseau (cited in n. 4
above). But poststructuralism exists in many guises. Its most rewarding versions do
not dissolve subjectivity so much as they posit its continual reformulation in changing
intersubjective and dialogic situations (and, as Foucault taught, in shifting metasubjective environments). And I hope to have described above how Derrida at least, if
not de Man, builds the implication of a productive slippage between speech and song
into his analysis of the Essayon the Origin of Language.Thus he allows for (if he does
not develop at length) a surplus or remainder of song over speech that is not
dependent on the nineteenth-century musical formalism and transcendentalism that
Abbate rightly senses in de Man. He allows, in other words, a distinction of singing
and speaking that might intersect in provocative ways with Abbate's phenomenology
of voice.

378
The Powersof Mexican Song

So where would we be left, were we to pursue along the lines
sketched here an understanding of the cantaresand, more broadly, of
all the diverse traces of Mexican song that have come to us? Not,
certainly, with an alternative view of the Aztecs "the way they really
were" that could make the same claims to truth as the views it
displaced. Not with a reconstituting of "authentic"Mexican voices.
We need to acknowledge, again with de Certeau, that our texts do not
replace what only others' voices might reveal; indeed, let us complete
de Certeau's thought: our texts do not replace what only others' voices
can reveal about theplace wherewe write. In Of GrammatologyDerrida
insists (an insistence not always heeded by his acolytes) that deconstructive reading does not provide us with a vantage point outside the
metaphysics it undoes; it only provisionally erases effects of that
metaphysics, in the process providing glimpses, but no more, of others
pressing at its horizons. It is not in our ken ever to know fully what
others might tell us about where we are, in the same way that we
cannot ever see undimmed our situation from theirs. Instead it can be
our privilege to reach tentatively toward the horizons of our situation,
sensing the otherness that lives there in its own situations. Instead
History and Ethnography-spell them finally with capital letters to
betoken the particular visions of storytelling about self and other,
present and past, that must be seen as two great Western projectscan be reshaped to bring us to other sorts of claims than the claim to
know the truth about others. In revised form, after they have
unlearned enough, they can shed their capital letters and offer ways of
reading the traces of others that have been more or less systematically
excluded till now by ideologies like logocentrism. In the process they
can make availablehypotheses that have been inaccessible, bringing us
to the juncture where both the hypotheses themselves and the
conditions of their inaccessibility emerge. But the hypotheses remain
hypothetical. They are sustained by a flux of dialogue that never
hardens into unilateral conclusion; they are produced by a thinking
from within our own situation that will be provisionally erased by, but
that will not allow the absolute construction of, other situations.
The Western (perhaps peculiarly Western) self-denying materiality of the alphabetic graphemes in which the Cantaresmexicanosare
written packs an extraordinarypower, a power to deploy an ideological regime where specific relations of speech, writing, and song to one
another and the world are fixed. The hypotheses that begin to emerge
from a reconsideration of these regimes, stimulated by both poststruc-

379
turalist critique and the new perspectives on Mexican traces it has

abetted, bring into view other possibilities and different relations.
They suggest-this much I can hazard now, by way of a conclusion
promising further study--that the powers of Mexican song unfolded
as its uttered materiality came into contact with other materials.These
probably included the paint of the codices, the materials of sacral and
mythic reality they encoded, the medicinal and talismanic substances
of therapeutic practice, the landscape of a migratory past, nexus of
Mexican historicity, and the "waterand hill" itself of civic space, locus
of societal order. In imagining such material contiguities we might
discover unsuspected ways of approaching ancient Mexican expression, subjectivity, and cosmogony.
Universityof Pennsylvania
ABSTRACT
Attempts to understandthe place of song in indigenousNew World

societieshavebeenstymiedby powerfulEuropeanideologiesof musicandits
uses. This is particularlyevidentin the case of the Mexica or, as they are
of European
commonlyknown,Aztecsof centralMexico.The crystallization
be
followed
five
centuries
of
accounts
of Aztec
can
conceptions
through
of
contradictions
The
and
these
limitations,
nature,
conceptionscan
singing.
be analyzedthrough a re-readingof paradigmaticEuropeantexts such as
Jean-JacquesRousseau'sEssaisurl'originedeslangues,a re-readingthat takes
us to the heartof earlypoststructuralist
thought.From all this emergesnot
a
our
of Aztecsong itself
for
broadened
only
conceptualrange
understanding
but also a defamiliarizing
approachto the singingof othersin anynumberof
contexts.


Tomlinson Ideologies Aztec Song

Uploaded by

Copyright:

Available Formats

Tomlinson Ideologies Aztec Song

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Tomlinson Ideologies Aztec Song

Uploaded by

Copyright:

Available Formats

What is the main topic of the document?

What is the main topic of the document?

What are some of the challenges in understanding indigenous music traditions?

What are some of the challenges in understanding indigenous music traditions?

Ideologies of Aztec Song

Author(s): Gary Tomlinson

This content downloaded from 128.97.245.36 on Thu, 11 Jul 2013 15:40:11 PM

Ideologies of Aztec Song

This content downloaded from 128.97.245.36 on Thu, 11 Jul 2013 15:40:11 PM

Standing behind the new readings of American scripts, informing

This content downloaded from 128.97.245.36 on Thu, 11 Jul 2013 15:40:11 PM

ways. Poststructuralistviews have extended constructivist conceptions

This content downloaded from 128.97.245.36 on Thu, 11 Jul 2013 15:40:11 PM

singing as it is about speech (and more a book about singing than

about writing). Read in a certain fashion, Derrida's analysis not only

This content downloaded from 128.97.245.36 on Thu, 11 Jul 2013 15:40:11 PM

aimed to inscribe the sound or phonos of speech, and in this it

putativesocietalevolution.For Rousseau,pictographsor hieroglyphs,

ideographs, and alphabet "correspondalmost exactly to three different

nation. The depictingof objects is appropriateto a savage people;

signs of words and of propositions, to a barbaric people, and the

This content downloaded from 128.97.245.36 on Thu, 11 Jul 2013 15:40:11 PM

6 To review some basic issues: Derrida's grammatology diverges from Saussurean

This content downloaded from 128.97.245.36 on Thu, 11 Jul 2013 15:40:11 PM

IDEOLOGIESOF AZTEC SONG

at the same time help us to unmake it by the token of their

7See Jean-Jacques Rousseau, Dictionnaire de musique (Paris, 1768; facs. ed.,

This content downloaded from 128.97.245.36 on Thu, 11 Jul 2013 15:40:11 PM

society. It is at once artificial, unnatural to the human voice, unfamiliar

sounds resulted in forgetting vocal inflections; and finally ... music,

So in counterpoint with Rousseau'sviews on melody, grounded in

This content downloaded from 128.97.245.36 on Thu, 11 Jul 2013 15:40:11 PM

IDEOLOGIES OF AZTEC SONG

operation will eventually drown out whole realms of others' singing.

This content downloaded from 128.97.245.36 on Thu, 11 Jul 2013 15:40:11 PM

JOURNAL OF THE AMERICAN MUSICOLOGICAL SOCIETY

him duringhis illusion.That is how the figurativewordis bornbeforethe

This content downloaded from 128.97.245.36 on Thu, 11 Jul 2013 15:40:11 PM

IDEOLOGIES OF AZTEC SONG

nium, Rousseau attributes metaphor to savages even as he withdraws

This content downloaded from 128.97.245.36 on Thu, 11 Jul 2013 15:40:11 PM

corresponds to passionate language" (Essay,p. 17; quoted in OG, p.

This content downloaded from 128.97.245.36 on Thu, 11 Jul 2013 15:40:11 PM

IDEOLOGIES OF AZTEC SONG

This content downloaded from 128.97.245.36 on Thu, 11 Jul 2013 15:40:11 PM

also foreshorten their views. For the Western capturing of Aztec

This content downloaded from 128.97.245.36 on Thu, 11 Jul 2013 15:40:11 PM

for example, noted in his History of America of 1777 the Mexicans'

This content downloaded from 128.97.245.36 on Thu, 11 Jul 2013 15:40:11 PM

This content downloaded from 128.97.245.36 on Thu, 11 Jul 2013 15:40:11 PM

IDEOLOGIES OF AZTEC SONG

Already in the waning sixteenth century, however, perhaps as the

This content downloaded from 128.97.245.36 on Thu, 11 Jul 2013 15:40:11 PM

This content downloaded from 128.97.245.36 on Thu, 11 Jul 2013 15:40:11 PM

Vico a whole historical philosophy that divided societal evolution into

This content downloaded from 128.97.245.36 on Thu, 11 Jul 2013 15:40:11 PM

the poetic means, including their "elevated metaphors and allegories,"

idolatry.He reservedthe wordmetaphoras-andthe Nahuatlcoinage

142-43, 414. Boturini's "IndianParnassus"remark is on p. I6o of the Idea;the cantares