6RXQG6WUXFWXUHDV6RFLDO6WUXFWXUH

$XWKRUV6WHYHQ)HOG
5HYLHZHGZRUNV
6RXUFH(WKQRPXVLFRORJ\9RO1R6HSSS
3XEOLVKHGE\University of Illinois PressRQEHKDOIRISociety for Ethnomusicology
6WDEOH85/http://www.jstor.org/stable/851232 .
$FFHVVHG
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 Illinois Press and Society for Ethnomusicology are collaborating with JSTOR to digitize, preserve
and extend access to Ethnomusicology.

http://www.jstor.org

SYMPOSIUM ON COMPARATIVE SOCIOMUSICOLOGY

SOUND STRUCTURE AS SOCIAL STRUCTURE

Steven Feld
n this paper I address two questions: What are the major ways that the
classless and generally egalitarianfeaturesof one small-scale society reveal
themselves in the structureof organized sounds? What are the majorways that
these same featuresreveal themselves in the social organizationand ideology of
soundmakersand soundmaking?By providingan overview of these areas I hope
to illuminatesome dimensions of a sociology of sound for the Kaluli of Papua
New Guinea, a traditionallynonstratifiedsociety whereegalitarianfeaturesseem
significantto sound structure,and where inequalitiesalso are clearly represented
in the distributionof expressive resources for men and women.
My concernwith these problemsderivesfrom a preoccupationwith merging
ethnomusicological questions (the cultural study of the shared meanings of
musical sounds) with sociomusical ones (the study of musical sounds from
perspectivesof the social structureand social organizationof resources,makers,
and occasions). My work of the last few years (Feld 1981, 1982, 1983) attempts
to understandthe most salient lessons aboutthe structureand meaningof Kaluli
sounds and ways they are inseparablefrom the fabric of Kaluli social life and
thought, where they are takenfor grantedas everydayrealityby membersof this
society. My title alludes to a perspectivethat considers structuredsound as "un
fait social total," in the sense that sociologists like Durkheim, Mauss, G. H.
Mead, and Schutz stress the primacy of symbolic action in an ongoing intersubjectivelifeworld, and the ways engagementin symbolic action continually
builds and shapes actors' perceptions and meanings.
My title also alludes to anotherpaper, Song structureand social structure,
one of Alan Lomax's seminal cantometricreports(Lomax 1962). This reference
is meant to locate this paper, and the Kaluli pattern it reports, in a larger
comparativeframeworkfor the sociomusicalanalysis of classless and egalitarian
societies. In doing so I also want to reconsiderLomax's rationalefor why we
shouldcomparesociomusicalsystems, and what we can comparefrom one to the
next.
For Lomax the "principal message of music concerns a fairly limited and
crude set of patterns"(1962:450); as a form of humanbehaviormusic should be
? 1984 Society for Ethnomusicology

Final version rec'd: 2/1/84

383

384

SEPTEMBER1984
ETHNOMUSICOLOGY,

seen as highly patterned,regular,and redundantin each society, yielding stable

structures.Lomax suggestedthat cantometricsprovideprofiles for each of these
societal musical norms. Moreover, "these stable structurescorrespondto and
representpatternsof interpersonalrelationshipwhich are fundamentalin the
various forms of social organization" (1962:449). Or, as he later put it, the
"salient featuresof songs are symbols for the key institutionsof society such as
the sexual division of laborand the state" (1976:9). Lomax suggestedthat songs
identify, represent, or otherwise reinforce the core structuresof society. By
comparingdistinct patternsof vocalizing, Lomax attemptedto constructworld
stylistic maps, raise questions of evolutionary sequence, and correlate dimensions of performancestyle with basic data on techno-economiccomplexity,
mode of production,and social organizationfor a world ethnographicsample.
In Lomax's conception comparativeresearch is fundamentalto knowing
how propertiesof singing style (musical behavior ratherthan musical content)
significantlyco-vary with social institutionsandotherlevels of culturalbehavior.
The expectation all along was for highly patternedshapes for each culture,
because "singing is viewed as an act of communicative behavior that must
conformto a culture'sstandardof performanceif it is to achieve its social ends"
(Lomax 1976:11).
Comparewhat? Lomax comparedsamples of ten songs from four hundred
cultures and correlated the codings with social structuraldata profiles from
Murdock'scross-culturalsurveys and the HumanRelationsArea Files (Murdock
1967,1969). The small samplesize of songs per society was justified by Lomax's
belief that each society has highly standardizedand highly redundantperformance models. "Cantometricsis a study of these standardizedmodels, which
describe singing ratherthan song. Therefore, it is not primarilyconcernedwith
completecollections anddescriptions,but with locatingprovableregularitiesand
patterns, in the fashion of science" (Lomax 1976:17). Lomax's 37 coding
dimensions attemptto factor all significantuniversalelements of song performance style on gradient scales.
Lomax's majorreportwas greetedwith mixed enthusiasm.Sample size and
time depth, compatibilityof song data with social structuraldata, psychocultural
reductionism,inferentialhistory, readingcorrelationas causation, intracultural
and arealvariability,and the extentto which the coding system normalizedraters
in ways which constrainthe accuracy of patternjudgement were all causes of
critical discussion surroundingthis monumentalwork. Much of the criticism
focused on methodand datainterpretation,and not upon Lomax's basic hypotheses about music as a universal public communication of social identity.
Whateverone's reactionto Folk Song Style and Culture, the publicationof
the cantometricstrainingtapes and coding book (Lomax 1976) must be greeted
as a majorevent in the historyof comparativemusical research.Few researchers
ever make their methods so available to others, and we should be grateful to

FELD:SOUNDSTRUCTURE
AS SOCIALSTRUCTURE

385

Lomax for these materials.Over the past few years I have listened to the tapes,
learned Lomax's core examples, and tried to apply his parametersto a society
and musical system I know well from intensive field research. I began with a
sample of about700 Kaluli songs and reducedthis numberto 500 afterremoving
single performerrenditionsor other reliabilityproblems. Whatcantometricshas
taught me about those 500 songs is that they display so much intracultural
variabilityand subtletythat it is virtuallyimpossibleto code a normalizedprofile
for them. To constructa typical cantometricprofile what I really need is onl'yten
songs, but the problem is, which ten?' How can I maintain the integrity of
patternsdiscoverablein largebodies of datawhen the cantometricssystem seems
to sacrifice so much significant data in order to objectify a "core pattern?"
After wrestling with the trainingtapes (an example of one of my specific
problems, with the social organizationof the vocal group, follows), I still feel
that Lomax is asking many of the right questions about music and social
institutions, but the mechanics of cantometrics crunches them in ways that
cannot satisfy the researcher accustomed to intensive field work, in-depth
analysis, and groundedethnographictheory. So I go back to the initial question:
compare what?
My suggestion is true heresy to many committedcomparativists,but I think
we need to pioneer a qualitative and intensive comparativesociomusicology,
without reified and objectified musical and social structuraltrait lists, without
unsituatedlaminationsof variouslycollected and historicallyungroundedmaterials. Comparativesociomusicology shouldtake the tough questionsand sort them
out with the best materials available for detailed comparison:the thorough,
long-term,historicallyand ethnographicallysituatedcase study. The meaningful
comparisonsare going to be the ones between the most radicallycontextualized
case examples, and not between decontextualizedtrait lists.
The data needed to begin this kind of comparativesociomusicology are
statementsof patternfor single societies, focusing on stylistic integrity, sociomusical coherence, and the role of music in role differentiation.Forthis the best
etic inputwill have to be the most thoroughemic data. By this I meanthatbroad
meaningfulcomparisonswill have to be basedon accurate,detailed, careful local
ethnographicmodels. So to start, the best way to answer Lomax's questions
about the systematicnatureof musical representationin social organizationis to
study them on the ground, in the field, up close, over long periodsof time, where
sound structuresare observably and undeniably socially structured.
A FRAMEWORK

While my firm belief is thatthe basis for comparingthe social life of sounds
must be qualitativeand derived from intensive local research,I also believe that

386

SEPTEMBER1984
ETHNOMUSICOLOGY,

such comparisonscan be framedin generaldomainsthat do not oversimplifythe

culturallyspecific dimensions of every sociomusical reality. I proposethen six
broad areas of inquiry into music as a total social fact, into the social life of
organized sounds. Each area is meant to open up a related set of socially and
musically situated questions that see sound structuresas socially structured,
sound organizations as socially organized, meanings of sounds as socially
meaningful. For each of the six rubricsI elaboratesome rudimentaryquestions;
the lists are not meant to be exhaustive in any sense.
Competence
1. Who can make sounds/music, and who can interpret/usethem?
2. What is the pattern of musical acquisition and learning?
3. Are there stratificationsof skill and knowledge? What types? How are they
sanctioned, recognized, and maintained?
4. Is musical acquisition assumed to be unproblematic?A necessity?
5. Do ideologies of "talent" determine or constrain acquisition and competence?
6. What is the relationshipbetween competence, skill, and desire for music?
7. What are the differences between productionand reception skills, for individuals, across social groups?
Form
1. What are the material musical means and how are they organized into
recognizable codes?
2. How are musical means distributedacross settings and participants?
3. What are the preferredaesthetic orderings?
4. Whatare the boundariesof perceivedforms?Whatdoes it meanto be wrong,
incorrect,or otherwise marginalfrom the standpointof code flexibility and
use?
5. How flexible, arbitrary,elastic, adaptable, open is musical form? How
resistantto changes, internalor externalpressures,or otherhistoricalforces?
Performance
1. What are the relationships between makers and materials?
2. What is the relationshipbetween individualand collective expressive forms
and performancesettings?
3. How are forms coordinatedin performance?How adaptableand elastic is
musicalform when manipulatedby differentperformersat a single momentin
time or over time?

AS SOCIALSTRUCTURE
FELD:SOUND STRUCTURE

387

4. How do cooperativeand competitivesocial relationsemerge in performance?

What meanings do these have for performersand audience?
5. How do performances achieve pragmatic (evocative, persuasive, manipulative) ends, if at all?
Environment
1. Whatresourcesdoes the environmentprovide?How are they exploited'?What
relationshipsexist between resources, exploitation, and the materialmeans
and social occasions for performance?
2. Are there co-evolutionary patterns, ecological and aesthetic, linking the
environment and sound patterns, materials, situations'?
3. What are the visual-auditory-sensaterelationshipsbetween people and environment, and how is this patternrelated to expressive means and ends'?
4. What myths or models scaffold the perceptionof the environment?Are these
related or complimentary to conceptions of person, society, expressive
resources?
5. What mystical or cosmological associations with the environmentsupport,
contradict,or otherwiserelateto the socioeconomiccontextof musicalbeliefs
and occasions?
Theory
1. What are the sources of authority,wisdom, and legitimacy about sounds and
music? Who can know about sound?
2. Is musical knowledge public, private, ritual, esoteric?
3. What dimensions of musical thought are verbalized?Taught verbally?Nonverbally?
4. Is theory necessary? How detached can theory be from practice? What
varietiesof knowledge and activitycount as musicalor aesthetictheory'?How
is music rationalized?
Value and Equality
1. Who values and evaluates sounds'?Who can be valued and evaluated as a
maker of sounds?
2. How are expressive resources distributed, specifically among men and
women, young and old? How do stratificationsemerge?
3. How do balancesand imbalancesmanifestthemselves in expressive ideology
and performance?
4. Do sounds deceive? Mystify? Who? Why?
5. Are sounds secret? Powerful? For whom? Why?

388

SEPTEMBER
1984
ETHNOMUSICOLOGY,

6. How do musical materials or performancesmark or maintain social differences? How are such differences interpreted?How are they sustained?
Broken or ruptured?Accepted or resisted?
These questions, and the six domains that head them, are intended as an
approachto integratingthe microscopic, ethnographicallydetailed analyses of
musical lives, with an arenaof comparable,general, relevantissues thatwill help
us compare sociomusical realities and practices. After a brief orientationto the
Kaluli, I summarizethe most salient issues in the six namedareas. By providing
key Kaluli metaphorsand concepts for each of these areas, I hope to stay trueto
an emic Kaluli sociology of sound, to the Kaluli articulationand constructionof
a sociomusical coherence system. At the same time, use of this simple comparativegrid will, I hope, make the Kaluli patternmore available to contrast,
comparison, and question in terms of the larger issues surroundingclassless,
small-scale, relatively egalitarian societies.
THE KALULI PEOPLE

Twelve hundredKaluli people live in the tropical rain forest of the Great
PapuanPlateauin the SouthernHighlandsProvinceof PapuaNew Guinea.2On
several hundredsquaremiles of rich land, at an altitudeof about two thousand
feet, they hunt, fish, gather, and tend land-intensiveswidden gardensthat yield
sweet potatoes, taro, pandanus,pumpkin, bananas, and many other fruits and
vegetables. Their staple food is sago, processed from wild palms that grow in
shallow swamps and creeks branching off of larger river arteries that flow
downwardfrom Mt. Bosavi, the collapsed cone of an extinct volcano reaching
eight thousand feet (E. L. Schieffelin 1976).
Kaluli live in about twenty distinct longhouse communities;in each, most
people reside in a single communal house, comprising some fifteen families
(sixty to eighty people). Social life for the village is centeredaroundthe house,
where primaryface to face interactionoccupies most time people are not in their
gardens,on the trails, visiting relativesin othercommunities,or stayingat small
garden homes or sago camps for major food processing activities.
This is a classless society. There are no social or occupationalspecializations, stratifications,or ranks.Thereare no professions, no ascribedor achieved
statuses that form the basis for social differentiation.All Kaluli are assumedby
their fellows to have equal social potential and endowment, which is theirs to
make of as best they can. Adults are responsiblefor getting what they want and
need out of daily affairs; assistance throughfriendshipand networksof social
relations is of primary importancefor all Kaluli.
This is also a generallyegalitariansociety in matterseconomic andpolitical.
There are no appointedor elected leaders, spokesmen, chiefs, bosses, control-

AS SOCIALSTRUCTURE
FELD:SOUNDSTRUCTURE

389

lers, or middlemen. People hunt, gather, garden,and work to producewhat they

need, taking care of themselves and their associates throughextensive cooperation in food sharing and labor assistance. There is little inequality in material
rewardsfor goods or labor because the goods are the only products;sharingin
labor or productioncontinuallypositions people within reciprocitygroups built
upon kin, extendedfamily, or friendshipties. There is also little accumulationof
goods, rewards, or prestige, and no highly valued jobs or roles automatically
rewardedby prestigeor materialobjects. In such a system, egalitarianismrefers
to a general lack of deference to persons, roles, categories, or groups based on
power, position, or material ownership.
Additionally, Kaluli seem less sexually polarized than other classless
societies at the middle range of techno-economiccomplexity; that is, societies
which mix some partialhunter-gatherermode of existence with a majoramount
of herding, fishing, or horticulture.This type of society varies greatly in sexual
equality, from hostility to mutualism(Schlegel 1977:5). While Kaluli men and
women have differentand often exclusive spheresof daily life (E. L. Schieffelin
1976:122-124; B. B. Schieffelin 1979), the degree of sexual antagonismand
animosity widely reportedin the Highlandsof Papua New Guinea (see Brown
and Buchbinder1976, M. Strathern1972) is not presentamongthe Kaluli. Many
domainsof controland productionare cooperative. Bridewealthnegotiationsand
maritalarrangements(usually sister-exchange)are the most obvious arenawhere
men can cultivate authorityand make decisions that expressly control women.
PATTERN
THE KALULISOCIOMUSICAL

Competence
For Kaluli there is no "music," only sounds, arrangedin categoriesshared
to greateror lesser degrees by natural,animal, and humanagents. Knowledgeof
these categories is widespread,tacit backgroundfor everyday life in the forest.
No hierarchies of sound types are imposed, no rationales constructed for
differentiatinghuman-madesounds from those of other sources. It is broadly
assumed that every Kaluli must become a competent maker, recognizer, user,
and interpreterof naturaland cultural sound patterns. Not only does physical
adaptationto the rain forest demand and favor acute auditoryperceptualskills;
Kaluli have developed the kind of ideological and aesthetic scaffolds for these
skills that humanize them and provide a coherent culturalframeworkfor their
acquisition. This is precisely the point of co-evolution: physical imperatives
alone do not explain adaptation;societies invent mutually supportiveadaptive
strategies linking nature and culture.
Acquisition of skill in song, weeping, whooping, cheering, humming,

390

SEPTEMBER1984
ETHNOMUSICOLOGY,

drumming,bird call and animal identification,as well as environmentalsound

recognition, are all fundamentallyrelated. Kaluli assume that the acquisitionof
skill in symbolic modes for expression of sound, as well as the natural
recognitionof sound, is unproblematic,naturallyrequiredof all social beings. In
fact these are very much considered like the acquisition of verbal or gestural
competence, requiringsimilaradult input, engagement, interaction,and explicit
"showing" instruction(B. B. Schieffelin 1979). While we in the West assume
some sort of necessity for symbolic competence in the verbal and gestural
modes, we assume no such thing for othervarietiesof symboliccompetence, and
thereforeutilize culturallyinvested notions (like "talent") to explain or rationalize stratificationsin awareness,expressive production,and interpretation.Nothing parallel exists among the Kaluli.
A basic Kaluli metaphorembodies the constructionof a culturalideal for
competence;this is the notion of "hard." Myth has it that the world was once
mushy and soft; alin the goura pigeon and ode:n the scrub turkey together
stampedon the ground, "making it hard," so life could flourish. The model for
social life recapitulatesthe model for the world. Infantsare "soft"; their bodies
and bones must "harden." A "hard" personis one who is strong, assertive, and
not a witch (witches have soft yellow hearts). Childrendo not eat soft, mushy
foods until theirbodies and language"harden." "Hardtalk" is grammaticaland
appropriatespeech, the talk that gets one what is needed in social interaction
(Feld and B. B. Schieffelin 1982). When a song "hardens," coalescing poetic
and performativestructures,people are moved to tears (Feld 1982:211-212).
"Hard" (halaido) then means social competence, physical maturity,verbal
competence, sound competence; control of energies, ability to perform, influence outcomes, take controlover one's life, invoke propersocial strategies.It
is assumedthatthe acquisitionanddevelopmentof social and intellectualskills is
the natural "hardening" process that all Kaluli go through. This central
metaphorlinks land, body, maturity,control, vitality, language, aesthetics, and
social action. For Kaluli, "hardness" is a core element of personhood.
Clearly not all Kaluli desire, develop, or attain uniform or even vaguely
identicalcompetenciesin the compositionof songs and poetic texts, the performance of songs or drumming, or funerary weeping. Some men and women
compose hundreds of songs in a lifetime, perform often, provide insightful
exegesis of texts, and discourse on musical matters in ways that extend far
beyond the assumed or recognized abilities of others. Two factors seem importantto interpretcorrectlythe meaningof this variation.First, the actualspreadof
competencies in productionskill is greatly downplayedby Kaluli in relationto
the much less stratified competencies in meaningful interpretation.Second,
Kaluli assume that differentiationin song performanceand composition are
explained by interest and desire, and not by special biological endowment or
talent. My remarksto the effect thatcertainpeople seemed to compose, perform,

AS SOCIALSTRUCTURE
FELD:SOUNDSTRUCTURE

391

or drum more than others were usually met by Kaluli with nonchalantreplies
citing the obvious: some garden more than others, some often make netbags,
some cook well, othersknow aboutbuildinghouses, and so forth. In sum, Kaluli
seem to have no investment in rationalizingdifferences in competence; they
simply assume that skills for interpretingand making sounds are naturally
acquiredand required,and that with instructionand encouragement,all children
will learn to sing and compose as part of their general socialization.
Form
Most all Kaluli sound expression is vocal, dependenton a strong interplay
of poetic and melodic elements. Five types of song exist, one of uniqueinvention
(gisalo) and four borrowedfrom neighbors(heyalo, ko:luba, sabio, iwo:). Each
of these is organizedby pentatonicprinciples;reductionsto three- and four-tone
variantsare common, and there is evidence of melodic convergence paralleling
the historyand diffusion of these four styles in the Bosavi area(Feld i.p.). Kaluli
compose in three of these five forms (gisalo, heyalo, ko.:luba);the other two
styles consist of closed sets (sabio, iwo:). Gisalo is limited to ceremonial and
seance occasions and is composed only by men. Heyalo and ko:luba are
composed for ceremonies but widely sung during everyday work and leisure
activities. Only in the case of heyalo do men and women participateequally as
composers.
Cutting across these song forms are other varieties and means of sound
expression:women's sung-texted-weeping(sa-ye:.lab),and cheering (uwo:lab),
men's whooping (ulab), and instrumentaldrumming(ilib) and bamboo jaw's
harp playing (uluna). Three other rattle instruments(sologa, seed-pod rattle;
degegado, crayfish claw rattle;sob, mussel shell rattle)are used for ceremonial
accompanimentto songs. Only the sologa is used more casually for accompaniment; additionally it is the only one of these instrumentsused by women.
Beyond a taxonomyof expressive forms, and the distributionof expressive
means in song, poetics, instruments, and performance, one very significant
generalizationcan be made about all Kaluli sound forms: no Kaluli sounds are
performedunison. Kaluli know quite well what unison is because missionaries
have tried to get them to sing this way for twelve years and churchleadershave
been taughtto count "one-two-three" before each song. Aside from church
activities (where only the most committed Christianscan actually manage this
new form of vocal organization)it is rare to hear anythingapproachingunison
sung by Kaluli or emitted from any sound sources in their environment.
Kaluli sound preferences,modeled consciously on bird sounds and the rest
of the forest environment,involve extensive overlappingand alternation,layering parts and sounds in coordinatednondiscretetextures. In Western musical
terms there is much canon and hocket. In cantometricterms, from Lomax's

392

SEPTEMBER
1984
ETHNOMUSICOLOGY,

characterizationof the social organizationof the vocal group (1976:86,177180), threepatternsare found:interlock,overlap, andalternation.Interlockis the
term Lomax uses for the social organizationof musical groups with an equality
of parts, common among acephalousbands of hunter-gatherersor noncomplex
forest societies. Lomax finds thatoverlap is most typical of largersocieties with
herd animals and a somewhat more complex production system. He cites
alternationof partsas morecharacteristicof societies with clear divisions of parts
and productive systems (Lomax 1976:86).
Listening to Kaluli songs makes it clear that no one of these three
characterizationspredominates;there seems to be an equal mix of all three. At
the same time, the Kaluli techno-economicpicture does not follow any of the
threeLomax cites as typically matchingthese sound types. The maingeneralization for Kaluli social organizationof the vocal group is that interlock, overlap,
and alternationare equally salient, precisely to the exclusion of unison forms.
Lomax (1976:86) cites unison as "the simplest technique of coordinating
effort"; he finds it "resorted to everywhere but most prominent in the performancesof small tribalsocieties, especially among planterswithoutlarge herd
animals." While this techno-economic characterizationis closer to Kaluli
society than any of the others cited above, Kaluli sound organizationsimply
lacks a unison principle altogether.
Like the use of "hardness" to characterize a Kaluli ideal of social
competence, there is also a metaphorthat draws together the dimensions of
interlock,overlap, and alternationwhich are so importantin Kalulisoundmaking
style. This is duluguganalan, "lift-up-oversound." Parts, sounds, whetherfew
or many, must constantly "lift-up-over" one another; one cannot speak of
sounds "leading" or "following" or "starting" or "finishing." Human
soundmakingmust stagger in layers, like bird calls, or arch up and over, like
waterfalls.The idea is more spatio-acousticthanvisual; Kaluli like all sounds to
be dense, compacted,withoutbreaks,pauses, or silences (Feld 1983). Whentwo
people sing together, the subtletiesof the shifting lengths of overlap (or, in the
case of a leader and a group, the nuances in the alternation)are the locus of
aesthetic play and tension. In the forest, sounds constantly shift figure and
ground;examples of continually staggered alternationsand overlaps, at times
sounding completely interlockedand seamless, are abundant.For Kaluli this is
the naturally coherent organizing model for soundmaking, whether human,
animal, or environmental:a constant textural densification constructed from
"lift-up-over sounds."
Performance
The performanceof all Kaluli sound expression focuses upon collective
texture and coordinationof layered parts. No competitive agendas play out

AS SOCIALSTRUCTURE
FELD:SOUNDSTRUCTURE

393

throughsong performance;the value of layering,juxtaposing, arching, "liftingup-over," and densifying are conceived as social activity. Even when the
situationinvolves a single voice, the sound is coordinatedwith the surrounding
acoustic featuresof the environment;this is particularlyso when Kaluli sing at
work.
At Kaluli ceremonies composed songs are sung all night long by members
of a visiting communityat the longhouse of their hosts. The hosts find the song
texts in particularto be sad and evocative because they concentrateon maps and
images of the places in the immediate, surroundingforest, places to which the
hosts have a sentimentalattachment.The performanceusuallyprovokesthe hosts
to tears, and to a sung-melodic-weeping performed in polyphony with the
ongoing song to which they are responding. The intense grief and sadness
experiencedby the hosts resultsfrom theirbeing remindedof the dead who have
lived, worked, and sharedmany experienceswith them at the places sequentially
mentioned in the song (E. L. Schieffelin 1976, 1979; Feld 1982).
It would appearthat this sort of evocation in performancemight create a
tremendousfocus on the composer and performerof the song as an individual
creator. Indeed, direct social manipulationis involved in composing a song that
might move a specific individualto tears. At the same time, these are not the
featuresof a song that Kaluli stress;they preferto cast the whole activity into an
explicit social and community-widemessage framework,largely throughimagery of land, the central Kaluli metaphorfor the accumulationand meaning of
social experience relating, sharing, and being with others in the forest.
When Kaluli men sat down with me aftera ceremonyto talk aboutthe songs
and their meanings, they always stressed the social rather than individual
motivationof song poetics. There is always collective sorrow it seems, because
of the underlying assumption that audience members will empathize with
whomever is moved to tears. This derives from the feeling of common experience among those people who sharea longhouseand surroundingforest community. Over and beyond ties of direct close kin, Kaluli feel deeply connected by
the places they have lived, gardened,worked,exchanged, and travelledtogether.
Soundmakingprovidesno formatfor the assertionof power, dominance,or
personalexcellence at the cost of others. The recognitionof skill in composition
and performanceis clear, and its pragmaticoutcome is the weeping of the hosts.
Despite all this, competition is not a major agenda in the ceremonies, and the
provocation involved is not a manipulationfor the sake of power. The Kaluli
constructionof a performanceis really activity in concert ratherthan activity
involving sequentialsolo performances.Individualityis not played up the way it
might be; costumes in fact conceal identities, so only the voice singles out the
singer.
If "lift-up-over sound" and "hardness" are the central metaphorsfor the
form of and competence in soundmaking,then its centralmetaphorfor perform-

394

SEPTEMBER1984
ETHNOMUSICOLOGY,

ance is "flow" (e:be:lan). Like a waterwaysituatednext to you in the forest,

whose water "flows" beyondperceptualimmediacyalthoughyou know it is still
moving at otherlocations, sound must have a physical presenceat its momentof
performanceand a staying power that carries it beyond the moment, so that it
flows in your mind, and stays with you beyond the boundariesof the immediate
event. Performance, like the substance of sound, must be dense, layered,
collective, thick in connotationand resonance;this is what makes it "flow."
Environment
For the Kaluli, the Bosavi tropicalrain forest environmenttakes on several
levels of meaningand abstraction.In the most basic sense the environmentis like
a tuningfork, providingwell-known signals that markand coordinatedaily life.
Space, time, and seasons are marked and interpretedaccording to sounds.
Sounds give indexical informationaboutforest height, depth, and distance. The
time it takes a sound to travelthroughvariouskinds of bush;the echoes through
landformations,waterfalls,and rivers;the layersof birdsoundin the canopyand
at forest openings;-all these provide clock and spatial information to the
accustomedinhabitantof the rainforest.Daily cycles of birdpresence, migration
cycles throughoutthe year, as well as cycles of cicadas and insects are taken by
Kaluli as indicatorsof location, season, and time of day. These signals have
differentauditoryappearancesfrom the village longhouse, from the forest edge,
from the gardens, from the trails, or from forest depths. Ratherthan counting
monthsor moons, Kaluliconceive seasons and cycles largely in termsof changes
in vegetation, changes in birdpresence, soundsof high and low wateraccumulation, or white water runoff in relation to rainfall. Numerous sounds then are
continuallyavailable and interpretedby Kaluli as the clocks of quotidienreality
in the forest.
At another,clearly relatedlevel, the forest is a model of balance, of plenty,
of resources,of deathand life and cyclical regeneration,of birds, animals, plants
and watersin mutualdependence. In effect, a model of what is natural,given as
normal, predictable;a model for humanlife and a challenge to it, since human
relationshipsand society depend on co-existence and managementof the forest,
throughmaintenanceand exploitationof its resources. Additionally, it is on this
basis that Kaluli explicitly rationalizethe form and performanceof song and
human sound with metaphoricideals about the layering of bird calls and insect
swells, the flowing of waters, arching of waterfalls, and general relationship
between perceptual immediacy and what lies beyond.
Anotherlevel of forest-Kalulirelationshipis more mystical. Kaluli believe
that two co-extensive realities, one visible, the other a reflection, comprisethe
world. In the reflection or mirroredrealm men may appear as wild pigs and
women as wild cassowarieshigh on the hills of Bosavi. The dead returnas "gone

FELD: SOUND STRUCTURE AS SOCIAL STRUCTURE

395

reflections" (ane mama) to the visible, usually in the form of birds in the forest
treetops. Thus the immediate village area is surroundedwith the presence,
throughvoices and sounds, of friendsand relatives. Because birds sing, whistle,
say their names, make a lot of noise, weep, or speak, they provide a simultaneous index of the environmentas well as a deepersymbolic understandingabout
self, place, and time.
Beyond these notions-tuning fork, model, mystery-there is also a deeply
pleasurable aspect to the way Kaluli approach the forest, which couples a
sentimentalitybased on land as mediatorof identity(E. L. Schieffelin 1976:2945) and an outrightenjoymentof the soundscape. Kaluli find the forest good to
listen to, and good to sing with as well. Improvisedhuman duets with birds,
cicadas, or other forest sounds are not uncommoneveryday events. Sometimes
people will find themselves a waterfalljust for the pleasure of singing with a
shimmeringaccompaniment.Again, the co-evolutionarytendencies for ecology
and aesthetics: Kaluli not only take inspiration from, listen to, and enjoy
the forest, but become part of it, which ultimately intensifies their sentiments
about it.
In sum, the Kaluli relationshipwith the forest is neither antagonisticnor
destructive,patternsthatare typical among some swiddenhorticulturalists.Little
ecological pressure, extremely low population density, no competition for
resources, and constantlyavailablefood (fish, fowl, vegetables) all contributeto
easy materialextractionand exploitationof the environment.At the same time,
the mystical, pleasurable, and tuning-fork dimensions of the forest-Kaluli
relationship reinforce this materialist basis. In all, the forest is a mama, a
"reflection," or mirrorfor social relationships,particularlyas mediatedthrough
the poetic imagery of songs that concern maps, lands and identities (Feld
1982:150-156), as well as through formal structureand singing style.
Theory
For Kaluli, the theory and concepts of where sounds come from and how
they can be organized, and, particularly,what they mean, is not contained in
esoteric knowledge or in a body of private lore controlled or circulated by
specialists. Myths about human-birdtransformationsexplain the origins of
categories of sounds that humans share with each other and with the natural
world, namely, weeping, song, poetics, whistling, talking, noise, mimicry.
These myths frame the meanings of sounds in terms of the range of social
sentiments associated with categories of bird-spirit"reflections." No special
occasions are requiredfor their telling, and no constraintsexist on who may tell
or hear them. All in fact are quite short and even Kaluli who don't volunteerto
narratethem can certainlyrecountthe punchlineor generalpoint to each. These
myths are central to the meaning and theory of sounds for Kaluli; the facts of

396

SEPTEMBER1984
ETHNOMUSICOLOGY,

theircodificationas mythsand theirqualityas chartersfor social realityspeaksto

their collective importance.
While the generalsocial outlines linkingsoundsand sentimentsarerevealed
in myths, Kaluli theory about sound organizationgoes quite a bit further(Feld
1981). Kaluli are energeticallyverbal about song, composition, and mattersof
musical form; to discuss sound they principallyrely on lexical and discourse
metaphors.All Kaluli terms for water motion, waterwayparts, and particularly
waterfallsare polysemous with the semanticfield of sound. Terms for intervals,
contours, and other structuralaspects of song form are inventedthis way. Song
parts are metaphorizedas bends of waters, or branchingstructuresof trees.
A numberof other theoreticalnotions are even less specialized or marked;
while myths and musical terminologycode theoreticalideas about what sounds
mean and how they are organized, not all Kaluli have much occasion to recite
stories or discourse about compositional matters. But all Kaluli know what
sounds can mean, what they can bring forth in social occasions, and how they
must be interpreted.Kaluli share a cultural logic for organizing their listening
experiences, and the extent to which that logic is systematic is the extent to
which it must be treated as musical theory, a foundation for Kaluli musical
epistemology, or a Kaluli theory of musical interpretation.
Like the duality of the cosmos alreadymentioned, Kaluli assume dualities
to all expressive forms; the idea is conventionallyexpressed by the notions of
"inside" (sa), and "underneath"(hega). Intentionallysymbolic behaviorsare
not transparent;they must be interpreted,and the act of interpretingthem is what
the Kaluli call finding the "inside" or "underneath." Sometimes these meanings are quite conventional.Even so, the interpretiveturnKalulitake is one that
assumes concentration, attention, and active listening.
Sounds have "insides" and "underneaths"because they often are heardas
the reflected images of birds; birds in turn are spirits, and the particularset of
emotional and personalassociations with spirits, spirit places, and the like are
deeply forceful for Kaluli. Additionally,poetic song texts are cast in a language
of "bird sound words" (o:be.:
to), and these have various conventional
go.'no:
and deeply ambiguous "underneaths"which involve a variety of obfuscatory,
mystifying, and evocative devices (analyzed in Feld 1982:138-144). In short,
Kaluli listeners are not passive participantsin symbolic affairs; they performa
kind of active social work based on the interpretiveassumption that there is
always a reflected "inside" or "underneath"meaningto sounds. Interpretation
is a necessary, active component in the social life of Kaluli sounds, bridging
individual competence with realms of theory and epistemology.
Kaluli exhibit markeddivergencesin the propensityto verbalizeaboutsong
or poetics. Many Kaluli, including both men and women who have composed
many songs, simply do not go in for extensive verbalizationabout song. Elder
men tend to be the most talkative and intellectually invested in this kind of

AS SOCIALSTRUCTURE
FELD:SOUNDSTRUCTURE

397

discourse and interpretation,and their enjoyment of this has no doubt been

prompted by several years of dialogue with E. L. Schieffelin and myself,
althoughI greatlydoubtthatthe kinds of discussions we've had with Kaluliwere
absent before our arrivalin Bosavi. Again I would point out that differences in
predispositionto musicaldiscussionand exegesis derivesostensibly from interest
and personal desire; no traditionalpositions, ascribed social roles, honorific
status, or rewards are involved.
In short, for Kaluli, theory is not an activity that takes place at a remove
from compositional, performative, or interpretive practice. All are deeply
associated, and theoretical ideas about musical form are part of the realm of
acquired competencies assumed to be essential for Kaluli.
Value and Equality
Soundmaking is highly valued and considered necessary for survival,
expression, and social interactionfor all Kaluli men and women. At the same
time thereare obvious inequalitiesin the pragmaticends these resourcesachieve.
Kaluli maintainthat two kinds of structuredsound forms, weeping and song,
emerged together in myth (Feld 1982:20-43). These are separate but complementarystructuringsof sound for social evocation. Women's funeraryweeping, which turnsinto wept song, is consideredby Kaluli to be the closest sound
form to being or becoming a bird, as it expresses immediatesorrows over loss
and abandonment. Men's ceremonial gisalo song, which ultimately moves
listeners to weeping, is composed and performedas a deliberateprovocationof
the fears and emotions of its listeners. Both forms derive from the same source
(birdsound)and reachthe same ends, bringingforthdeep emotion aboutloss and
abandonmentin the context of village-wide public rituals. Women are highly
valued and evaluated as funerary weepers, and men are highly valued and
evaluatedfor composing and performingpersuasivegisalo songs. In both cases a
reputationmay develop for individual performers, at least in the immediate
circumstancessurroundingan event; this sort of reputationdoes not seem to lead
to any marked consequences in terms of social prestige.
No such male-female complementarityexists for other Kaluli expressive
forms, that is, forms not invented by Kaluli but borrowed by them from
neighboring groups. The oldest addition to the ceremonial song repertoireis
heyalo; here in fact there is a greatdeal of female compositionalactivity, but the
ceremony is the domain of men. Young men often compose quite a numberof
songs in the hevalo formatbefore tackling the more poetically and melodically
demandinggisalo form. Young men performbut do not compose sabio, a song
form from the Lake Kutubuarea to the East, introducedby carriersworking on
governmentpatrolsto Bosavi in the 1950s. Ko:lubawas importedinto the Kaluli
area in the mid-1960s: men performand compose in this style but there are very

398

ETHNOMUSICOLOGY,SEPTEMBER 1984

few compositionsby women, althoughsome women with family ties to the area
from which ko:lubaderives know quite a numberof the songs and sing them for
work. Iwo.:is a fixed cycle of songs sung the night before killing pigs; it has an
explicit women's counterpartcalled kelekeliyoba, sung the morning after the
ceremony. These songs also use fixed formulaewith only new placenamesand
pig names added by the singers.
In terms of instruments,handdrumsand bamboojaw's harpsare relatively
recent introductionsto the Bosavi area, althoughthe latter is said to replace a
monochordicmouthbow, whose history is uncertain.Both instrumentsare only
availableto men;drummingtakes place beforeor duringceremonialactivity, and
the jaw's harpis associatedwith the personalrecreationof young men. The drum
involves magical constructionsecrets which are kept from women; even though
it is not hiddenfrom their sight, women do not touchthe instrument(Feld 1983).
Even so, the dimensions of drum secrets seem to be nothing like the secret
character of flutes and garamuts reported elsewhere in Papua New Guinea
(Gourlay 1975). Of the three rattleinstrumentsused for ceremonialaccompaniment, only one is sharedwith women (sologa), while the one used for gisalo, the
sob, is clearly consideredsecret, and should not be touched by women. Women
are not aware of the mystical events through which the cane handles (olo:
se:se:lo:) of these instrumentsare passed on to Kaluli men from spirit mediums
(E. L. Schieffelin 1976:214).
One other major soundmakingresource does return us to a very basic
male-femalecomplementarity;these are the matchedpair of demonstrativeand
assertivecollective soundscalled ulab, ("says u") and uwo:lab, ("says uwo.:").
For men ulab is a loud group whoop, celebratingthe call of the eagle usulage, a
booming UUU! For women, uwo.lab is a raucousgroup cheer, celebratingthe
call of the SuperbBird of Paradiseuwo:lo, a screeching U-WO:O:!These two
birds are prominentspirit representationsof Kaluli men and women, and the
group soundmakingusually takes place when men collectively work or prepare
for a ceremony, or when women cheer during a ceremony.
The complementarydistributionof gisalo and weeping, whooping and
cheering speak for the way Kaluli men and women have coordinatedseparate
expressive spheres that are relatedand mutuallysignificant in modes of appeal
and assertion. At the same time the distributionof instrumentaland ceremonial
song resources clearly points to the way Kaluli men have appropriatedand
control new expressive resources.
Finally, it is importantto note the clear inequalitiesin what it is that these
expressive resources allow men and women to achieve, symbolically and
pragmatically.For men, the compositionand performanceof ceremonialsongs
creates a grand social focus around them and their powers of evocation.
Ceremonial action of this sort is the height of Kaluli stagecraft, drama, and
collective celebration (E. L. Schieffelin 1976:172-196, Feld 1982:163-216).

AS SOCIALSTRUCTURE
FELD:SOUNDSTRUCTURE

399

Gisalo is the most powerful of these, but hevalo and ko:luba achieve the same
effect for men. In all threethe performancemay be so moving thata woman will
lose her heartto the dancer, wish to elope, and follow him home. Drummingcan
contribute to the same social power (Feld 1983).
What weeping achieves for women is far less sweeping. It certainlycreates
a social focus on women as performers,and thatfocus is significantin validating
the role weepers play in articulatingcommunity sentiments. But no persuasive
social ends and no long-term changes in social life are effected by weeping;
largely it is an intensely aesthetic public display of personal grief.
If there is a key Kaluli metaphor that sums up male-female relations
generally, and male-female expressive means and ends specifically, it is the
rather ambiguous term ko.li, "different." While this was about the most
common term I heard men use when speaking of women, or women use when
speaking of men, it certainly means different things from the relative vantage
points, and can have positive ("different" in the sense of new, exciting,
valuable), flat ("different" in the sense of bland, not quite right) or neutral
connotations.Kaluli recognize real social difference, but there is often little that
is negative or antagonisticin that recognition. At the same time, male ideology
clearly casts women's differences in a dangerous light: aversion to menstrual
blood and belief that women drain male energy are rationalesfor taboos and
other social practices which reproduce beliefs in and actions against female
contamination.The tensions and contradictionsfound here parallel the duality
of expressive domains, which have both a complementarydimension (songweeping, whooping-cheering) and an overlay of male-appropriatednew resources.
Discussion
To roundout the shape of the Kaluli pattern,I organize a summaryaround
three questions:
(1) How are the obvious egalitarianfeaturesof Kalulilife markedwithin
the organizing schemes of Kaluli sound structuresand soundmaking?
(2) How is the most obvious form of Kaluli inequality (men over
women) markedand resolved within the organizingschemes of Kaluli
sound structuresand soundmaking?
(3) How does the Kaluli pattern fit into the broader comparative
framework of Papua New Guinea societies?
. in egalitarian societies the division of labor by sex has led to
complementarity and not female subservience;

. . .

women lost their equal

status when they lost control over the products of their work" (Leacock

400

SEPTEMBER1984
ETHNOMUSICOLOGY,

1978a:79, see also Leacock 1978b, 1977). Leacock's examinationof huntergathererand horticulturaleconomics shows thatwomen foragedand furnishedas
much if not more basic resourcesto hunter-gatherer
life thanmen did. Moreover
they distributedthe food to networksof kin as well as nuclearfamilies. Societies
which typically mix hunting and gatheringsubstratawith a dominanthorticulturalpracticepartiallysharein the patternLeacock describes, and presenta wide
range of egalitarian tendencies as well as unequal or contradictoryones.
Schlegel's (1977) notion that theories of sexual stratificationmust account for
person-to-personrelations and ideology as well as person-to-goodsrelations is
clearly essential if we are to understandsocieties like the Kaluli that are more
politically and economically egalitarianthan sexually so. This is also important
because Kaluli have the economic resources to produce a surplus far beyond
subsistence needs. The theory that stratificationsystematically co-varies with
surplushas been recentlydismantledby both Marxistsand non-Marxists,largely
in favor of theories that see stratificationmore closely related to population
dynamics (Cancian 1976).
Given all this, one thing we can clearly predict is that a society like the
Kaluli with such a combinationof demographiccharacteristicsas low density,
easy subsistence, ecological diversity, and economic resources like combined
hunting,fishing, sago production,and gardening,will not neatly matchidealized
social-economic-technologicalcharacterizationslike "hunter-gatherer,""horticultural," "peasant," and so forth. Our expectation therefore should be for
complex, and ambiguousor contradictorytendenciesin sociomusicalpracticeas
well as social organization.This seems to be the case in several regards.On the
one hand, the image of deep mutualismand co-aesthetic relationshipwith the
forest ecology is reminiscentof hunter-gathererslike the pygmies of the Ituri
(summarized,from Turnbull, in Lomax 1962). The general situationfor competence, musical form and performance, environment and theory depicts an
emphasison cooperation,complementarity,autonomy,valuingself in relationto
othersand to ecology. It is easy to imaginehow these musicalbeliefs, structures,
and actions can be coherent within an egalitariansocietal framework,and how
they can be locally perceived and sustained within that framework.
At the same time there are imbalances in the distributionof expressive
resourcesbetween men and women; more significantly, there are differences in
the pragmatic ends served by these resources. Men can mystify, impress,
persuade, and even win women, but what are women getting from weeping at
funerals and singing at work besides personal enjoyment, fulfillment, and
general social solidarity?Male-femaleinequalitiesthen are definitely markedin
the sphere of musical organization,but to push the qualitativepoint, it also is
importantto realize that the form in which they are markeddoes not involve
secret instruments stolen from women in mythic times, does not involve
deception maintenance, does not involve daily antagonisms, and does not

FELD:SOUNDSTRUCTURE
AS SOCIALSTRUCTURE

401

involve sanctionsof personalviolence. All of these variants,found elsewhere in

Papua New Guinea, are not manifest in the Kaluli situation.
A final way to scrutinize the patternof male-female inequalities in the
Kaluli patternis to look at the largerPapuaNew Guinea ethnographicsituation
througha widespreadsocial institutionwhere inequalitiesare ideologically and
practicallyenacted:male initiationcults. In an introductionto a recentcollection
of essays on this topic Roger Keesing (1982:7-11) summarizes some of the
prominentthemes involved in PapuaNew Guineacult activity. These includethe
idea that men and women are physically and psychologically different, that the
fluids and essences of women are potentiallyif not actuallyharmfulto men, that
boys must be subjectedto rigorousinitiationsso thatthey will not be weak, that
male homosexualactivity is essential to the ideology of sexual separationand the
creation of differentiated men, and that residential separation of males is
necessary in order to avoid female contamination.
A central feature of many of these cults is secret instruments,particularly
flutes. K. A. Gourlayhas devoted a monographto the distribution,symbolism,
and meaning of these instruments,showing that the male secrecy surrounding
their use is a central component of the deception of women, and a central
revelation during the rites.
of secrecyis linkedcloselywith the men'sdeceptionof the womenand
Maintenance
sounds.... Forthose
uninitiated
youthsthrough
givenaboutthemysterious
explanations
to whomthe secretshave been revealed,emphasisis on maintaining
the deception
(1975:102).

The Kaluli seem to have practiced (until 1964) an extremely moderate

variantof this pattern,without traumatic,violent rituals, or secret flutes; it was
called the baiua. In an essay on this institution,E. L. Schieffelin (1982) argues
that this was not an initiation cult because there was no formal status change
involved and because novices assumed no submissive and dependent role.
Nevertheless,
The bauat exhibitedmanyfeaturestypicalof a maleinitiationprogram,includingthe
seclusionof themembers,ritualactivityaimedat promoting
male
growthandenhancing
qualities,and the teachingof secretlore (E. L. Schieffelin1982:156).

The bau a stressed hunting activities, promotion of strength including

homosexual intercourseto promotegrowth, long expeditions, and development
of knowledge of the forest geography, flora and fauna. The ideological stance
toward women was clear: ". . . Kaluli believe that women have a debilitating

influence on men. A man who has had too much to do with women is likely to
lose his stamina, become fatigued on the trail" (E. L. Schieffelin 1982:178).
Moreover,

402

ETHNOMUSICOLOGY,SEPTEMBER 1984
To non-participants,and in particularwomen, the bau a was presentedas a mystically
powerful and dangerous institution (ibid:163).

The bau a was in earshotof the village longhouse, so women heardsounds

of the men; this was supposedto be tantalizing.While women pretendednot to
know any of the secret aspects of what was going on, some in fact knew quite a
bit about it (ibid:163). But the bau a "... expressedwhat men liked best about
themselves, what they stood for and wanted to be...
"(ibid:166).
Forthe youngmen, the baua removesthemfromthe marginsof socialattentionand
thruststhemto its center;theyemergefromseclusionin theirbestandmostappealing

cultural image (ibid:194).

In other words, they emerge as controlled, energetic, good hunters, knowledgable about the forest, including the spirits who reside there. They also emerge
supplying a large amount of cooked meat to members of their own and the
surroundingcommunities, thereby placing them all in their debt.
The differences between the bau a and some of the other traumatic
separationrites practiced in the Papua New Guinea Highlands is clearly underscoredby societal differences.Complex initiationswith grades, alliances, and
the like require large populations. Although the bau a was symbolically and
ritually elaborate and complex, it had none of the stratified complexities of
graded initiations common for many of the other societies summarized by
Keesing (1982) and discussed in the literature.
Ecological pressures, root crop intensification, and increasing population
density are common in many Papua New Guinea Highlandssocieties. A major
shift towardintensivecultivationincreasesthe burdenson women, and seems to
be accompaniedin many societies by increasedmale control throughappropriation of the products of women's labor. The Kaluli, by contrast, are low in
production,lackingin ecological pressuresand crop intensification.They are not
competing for food or territory, and their population size is stable. Men's
hunting, clearing and gardening activities seem complementaryto women's
gardening and sago-making work. This produces none of the competition
necessary for the organizationand maintenanceof a "big man" social system
exchanging women and their products.
While the bau a shares many surfacefeatureswith initiationcults in Papua
New Guinea, it appears to have been a more subdued variantof most of the
institutionsdiscussed in the literaturesurroundingissues of male-femalerelations
(Keesing 1982, Allen 1967, Gourlay 1975, Langness 1974, Murphy 1959, A.
Strathern1970). Kaluli men seem far more interestedin impressingwomen than
in maintaininghostilities with them. Ceremonialsong, costuming and performance, hunting and providing meat for exchange are the activities males most
cultivate and use for impressingwomen. These activities are not filled with the

AS SOCIALSTRUCTURE
FELD:SOUNDSTRUCTURE

403

preoccupationsreportedelsewhere:maintainingseparateness,antagonism,keeping ritualsecrets, using secret instrumentswhich are forbiddento women under

penalty of rape or murder. Kaluli seem to promote male mystique without
promotingextensive female deception. The lack of secret instruments,so pivotal
to ritualdeception, was replacedby whooping and singing at the bau a; women
could hear this from the distance, but its purposewas to excite and impress, not
so much to deceive. Here, as in ceremonies, Kaluli men assume an evocative
stance toward both women and other men. Toward women that evocation
translatesas power throughdifferencewhile towardmen thatevocationtranslates
as solidarity and nostalgia.

CONCLUSION

Keil argues that:

The greatforce shapingmusicand its meaningis social inequalityspecificallyas it
manifestsitselfin fourdirections
atonce:Thedominance
of somemenoverothermen,of
menovernature,of menoverwomen,andof somesocietiesoverothersocieties.Among
those remainingprimitivepeoples who have managedto maintainthe social, natural,and

sexualequalities,I assumethat"music"is a vitalpartof thatmaintenance.

Indeed,in
such societieswhatwe call music-dance-ritual-religion-ecology
seem to be fusedinto
nearly one homeostatic system, symbolizing nothing or everthing (1979:1).

The Kaluli seem both to uphold and to contradictthese three statements.

First, social inequalityis not the main force shaping music and its meaning for
Kaluli. The dominanceof some men over othermen, of men over nature,and of
society over other societies is not a majortheme in the content of Kaluli songs,
not related to the occasions of their performance,and not related to the social
organizationor performancestyles of song makers. At the same time we have
seen that social inequalityof men over women is clearly manifestamong Kaluli
despite other complementaryand cooperative tendencies. This inequality is
sociomusically markedin two ways: differentialexpressive resourcesof men and
women, and differentialsocial-pragmaticends those resourcesmay serve. In all,
social inequalityis markedin some dimensionsof Kaluli musical life but not in
others; where the markingoccurs, it is neithersystematicnor withoutcontradiction, particularlygiven the prominenceKaluli invest in the complementarityof
song/weeping and whooping/cheering.
Keil's next statementalso has validity for Kaluli, but again, the dictum is
less than wholly applicable. Kaluli traditionallydid maintainmany social and
naturalequalities, and in their notions of competence, of form, of performance,
of theory, and of the environment,many indicationsof equalityand balancecan
be found. Indeed, it seems that "music" is a vital part of maintainingthose
concepts and those relations with each other and the environment.

404

SEPTEMBER1984
ETHNOMUSICOLOGY,

Keil's final statement also makes a lot of sense for the Kaluli: the
interpenetrationof the environment with all Kaluli sound recognition and
expression really does symbolize everythingin some way: competence ("hardness"), form ("lift-up-oversound"), performance("flow") are all tied together
by theory (in myths of human/birdtransformation).The success of Kaluli
soundmakingas a deeply affective and emotional mediumof communicationis
grounded in this invented coherence.
Moving to the generalfrom the specific, my concernswith comparisonhave
in great part been stimulatedby reading Alan Lomax's work and by trying to
apply his training tape examples and coding procedures to Kaluli data. I
obviously differ from Lomax at the outset in a basic philosophicalmatter:I do
not equateexplanationwith normativestatisticalcorrelationsor causal analysis. I
am more concernedwith explaining the situatedmeanings of sound patternsin
the intersubjectivelycreated world of actors and actions, and I am concerned
with the role local ideologies play in constitutingand maintainingthose local,
specific sociological models of and for musical realities. I am thereforebiased
towardwhat Lomax calls the "narrowfeatures" of both the streamand content
of music, while his own focus is on grosser, objectifiableand redundantfeatures
of the behavioralstream. At the same time, I have arguedthat it is possible to
conceive a set of comparativeresearchquestions that make use of local models
and metaphors while at the same time identifying issues that have broader
comparativevalue, concerning both the conceptual and materialdimensions of
musicality. While these six domains may look more like artifacts of the
cultural-ideationalsphere rather than the sociological, I have indicated that
culturalconstructsare essential precisely because of the way they lead to and
provide local models for social structure (social roles, division of labor,
stratification, and differentiation).
I have also stressed that there is not a yes/no issue about correlatingsong
structureand social structure.Obviously such correlationsare possible, whether
undertakenin world sample quantitativetermsor small-scale ethnographicones.
The issue is: how do we interpretsuch relationships,and how do we argue for
whatthey mean?Lomax's tendencyhas been towardcausalconnectionsbetween
social evolution and singing style. He arguesfor causal covariancefrom societal
complexityto melodic intervalcomplexity, text precisionand rhythmicfreedom.
He argues for causal covariancefrom societal sexual restrictivenessto singing
voice nasality, and social cohesion to choral cohesion.
The most complete and convincing quantitativereanalysis of cantometric
data thus far, by one of Lomax's early statisticalco-workers, uses multivariate
techniques to argue for historical diffusionist rather than evolutionary interpretationsof cantometricdata. Manysong featuresinterpretedby Lomaxto be
correlates of evolutionary process can be explained almost exclusively by
regional location(Erickson 1976). If culturehistoryis enough to explain the kind
of taxonomic variance Lomax originally found, Erickson insists that:

FELD:SOUNDSTRUCTURE
AS SOCIALSTRUCTURE

405

canbe shown
Fornearlyeverymajorstylein thecantometric
taxonomy,a widevariation
correlates.By contrast,such styles do not often
on the positedcultural/institutional
homes.Whentheydo, it appears
to be inthe
of theirethnohistoric
oversteptheboundaries
contextof an important
eventin history.The attemptto explainsongin termsof some
universalunilinearprocessof social evolution-for all thatsuch a dimensioncan be
shown to exist-is to over-simplify our understandingof this most humanof behaviors
(1976:307).

Erickson argues that song style is less a causative reflection of social

institutionsthan an emblem of social identity. Lomax's early papersalso stress
the functional importanceof song style as an indicatorof social identity, thus:
S.

effectof musicis to givethe

. fromthepointof viewof itssocialfunction,theprimary

listener a feeling of security, for it symbolizes the place where he was born, his earliest
childhood satisfactions, his religious experience, his pleasurein communitydoings, his
courtshipand his work-any or all of these personalityshapingexperiences (1959:929).

The notion thatthe most general relationshipbetween sound structuresand

social structuresconcerns identity has strong supportfrom qualitativeresearch,
and certainly is a conclusion that could be derived from the Kaluli material,
where musical structuresframe the message: this is what we are, where we are,
who we are. At the same time these musical structuresmore specifically move
the listenerto an interpretiveplane beyond general messages about identityand
boundariesto more explicit Kaluli semanticsof sound. Here is wherethe listener
"turns over" the sound organization of singing and poetics to find "'underneaths"aboutbirdand spiritmanifestations,and throughthese, "'insides"of
social emotion and world-sense coherence.
The Kaluli case makes it clear thatit is difficult and confusingonly to argue
from objectified social structuresto musical structures;the result is a net of
reifications. Everything that is socially significant and institutionallyreal for
Kaluli is not necessarily representedin musical order, occasions, or resources.
One cannotdirectlypredictthe shapeof the Kalulimusicalsystem, singing style,
or performanceorganizationfromtheirmode of productionand techno-economic
complexity. The Kaluli sociomusical system is variedin its resourcesand forms,
and includes manifestationsof musicalfeaturesand style thatcan also be fully or
partially found among societies of hunter-gatherers,horticulturalists,herders,
planters,and even peasants(the most unegalitariansocieties known to us; Fallers
1977).
At the same time, it is equally unappealing to reverse the argument,
predictingsocial structuresand modes of productionfrom musical forms. Many
similar surface musical forms are found in societies of widely varying social
complexity, although the musical forms have greatly divergent meanings and
identitiesin these separatehistoricaland geographicsettings. (Witnessthe use of
hocket in the Renaissance.) In short, for all societies with similar techno-

406

SEPTEMBER1984
ETHNOMUSICOLOGY,

economic-social featureswe might expect more musical varietythan Lomax has

indicated. Conversely, for all similar sound patterns and modes of musical
productionwe might expect more social patternsto which they are significantly
related. For any given society, everything that is socially salient will not
necessarily be musically marked. But for all societies, everything that is
musically salient will undoubtedlybe socially marked,albeit in a greatvarietyof
ways, some more superfluous than others.
In a series of recent papers, Judithand Alton Becker have elaboratedsome
of the meanings of "coherence" in symbolic systems (A. Becker 1979, J.
Becker 1979, J. and A. Becker 1981). They follow KennethBurke, arguingthat
the naturallyfelt sense of experience in differentsymbolic modes is deeply tied
to world-sense, to the constructionof culturalboundaries,and to whole logical
orientationsto living and feeling. A central feature in this coherence building/
maintainingprocess that sustainssymbol makingand interpretation,and thereby
human cultures as we know them, is metaphor. "Metaphorsgain power-and
even cease being taken as metaphors-as they gain iconicity or 'naturalness'"
(Becker and Becker 1981:203). In this paper I have also approached the
coherence of Kaluli sound and social order through Kaluli metaphors(which
indeed, are not metaphorsto Kaluli, but simply what is real) while also tryingto
indicate that, as local sociocultural models, metaphors can and should be
comparedfor what they tell us aboutpossible coherences, possible sociomusical
realities. It is my hope that a comparativesociomusicology will develop along
these lines, elaboratingnot correlationsof song structuresand social structures,
but coherences of sound structuresas social structures.

ACKNOWLEDGMENTS
Revision of a paper read at the Society for Ethnomusicologyannual meetings, Tallahassee,
October 1983 at a session co-organizedby CharlesKeil and myself on ComparativeSociomusicology
of Classless and EgalitarianSocieties. Many of the issues discussed here have emerged through
conversation with Charlie and been enhanced by our ongoing materialist-idealistdialogue. Comments at the SEM panel from MarinaRoseman, John Blacking, and Tony Seeger were helpful in
reformulatingmy remarks,as were commentsfromSimon Frith,GeorginaBorn, andJody Berlandat
the Sociology of Music Conferenceat Trent University in August 1983. 1 am also gratefulto John
Shepherdfor inviting me to presenta version of this materialthereamongstsociologists and popular
music researchers. For research support 1976-1977 and 1982 1 am grateful to the National
Endowmentfor the Arts, the Universityof PennsylvaniaResearchFoundation,the Instituteof Papua
New Guinea Studies, the Archives of TraditionalMusic, and the AnthropologyFilm Center. In
Kaluli words, the typed symbols /e:/ and /o:/ equal c and 3. Otherletterscarrytheirphoneticvalues;
for orthographysee Feld 1982:17-19.
NOTES
1. Lomax (1976:16) maintainsthat "Generally, the more songs per culture we analyzed, the
clearerthe core style became." This is a difficult statementto evaluate. While it should be obvious

FELD: SOUND STRUCTURE AS SOCIAL STRUCTURE

407

that largersamples indicateclearerpatterns,the real issue is whetherthe cantometricscales factorout

or blur too many significant sociomusical dimensions to begin with, thus insuring that the "core
pattern" will be overly simple, and thereby making additional sample data superfluous.
2. I write in an ethnographicpresentof the late 1960s to mid-1970s, based on my work and that
of two co-workers, ethnographerE. L. Schieffelin and linguist-ethnographerB. B. Schieffelin.
Massive changes took place in Bosavi in the 1960s and 1970s; some of these are assessed in termsof
culturaldestructionin E. L. Schieffelin 1978. When I returnedto the Kaluli in 1982 I found that
evangelical Christianity,so strong by the mid-1970s, had lost much of its appeal. Kaluli tired of
waiting for the second coming of Jesus Christ after so many years of being told to preparefor an
imminent arrival. Some ceremonialism was reinvigorated, but there were new disruptions and
changes relatedto outside contact, circulationof cash, and desire (largely throughmission influence)
for materialWesterngoods. The changes that have taken place have differentlyaffected the various
longhouse communities, and differentialswithin communitiesare also clearly apparent.New forms
of exploitationand confusion have spread,andculturalanxietyhas heightened.While this essay does
not touch on the implicationsof all this for Kaluli social organization,it is importantto note thatnew
forms of stratificationare evident: Christianityhas given status and advantage to a small few,
cultivated high rewardsfor pastors, introducedcash and social differentiation,and greatly debased
traditional personal autonomy for Kaluli.

REFERENCESCITED
Allen, M. R.
1967 Male Cults and Secret Initiations in Melanesia. Melbourne:MelbourneUniversityPress.
Becker, Alton
1979 "CommunicationAcross Diversity," in A. L. Becker and Aram Yengoyan, eds., The
Imagination of Reality (Norwood: Ablex), pp. 1-5.

Becker, Judith
1979 "Time andTune in Java," in A. L. Becker and Aram Yengoyan, eds., TheImaginationof
Reality (Norwood: Ablex), pp. 197-210.
Becker, Judith and Alton
1981 "A Musical Icon: Power and Meaning in Javanese Gamelan Music," in Wendy Steiner,
ed., The Sign in Music and Literature (Austin: University of Texas Press), pp. 203-215.

Brown, Paula and Georgeda Buchbinder, eds.,

1976

Man and Woman in the New Guinea Highlands. Washington: American Anthropological

Association.

Cancian, Frank
1976

"Social Stratification,"

Annual Review of Anthropology 5:227-248.

Erickson, Edwin
1976 "Traditionand Evolution in Song Style: A Reanalysis of CantometricData," Behavior
Science Research

11(1):227-308.

Fallers, Lloyd
1977 "Equality and Inequalityin HumanSocieties," in S. Tax, ed., Horizons in Anthropology
(Chicago: Aldine, 2nd. edition), pp. 257-268.
Steven
Feld,
1981 " 'Flow like a Waterfall': The Metaphors of Kaluli Musical Theory," Yearbookfor
1982

Traditional Music 13:22-47.

Sound and Sentiment: Birds, Weeping, Poetics,

and Song in Kaluli Expression. Phil-

adelphia: University of Pennsylvania Press.

1983 "Sound as a Symbolic System: The Kaluli Drum," Bikmaus 4(3):78-89.
Kaluli Weeping and Song. 12" stereo disc with notes in English and German. Kassel:
i.p.
Bdirenreiter-Musicaphon.

408

ETHNOMUSICOLOGY,SEPTEMBER 1984

Feld, Steven and Bambi B. Schieffelin

1982 "Hard Talk: The FunctionalBasis for Kaluli Discourse," in D. Tannen, ed., Analyzing
Discourse: Text and Talk (Proceedings of the Georgetown University Roundtable in
Linguistics and Languages, 1981). Washington:Georgetown University Press, pp. 351371.
Gourlay, K. A.
1975 Sound-producingInstrumentsin TraditionalSociety: A Study of Esoteric Instrumentsand
their Role in Male-female Relations. Port Moresby and Canberra:AustralianNational
University-New Guinea Research Unit.
Keesing, Roger
1982 "Introduction,"in GilbertHerdt,ed., Rituals of Manhood:Male Initiationin Papua New
Guinea (Berkeley: University of California Press) pp. 2-43.
Keil, Charles
1979 "Marxism as the Context for a Symbolic Analysis of Music," Paper presented at the
Society for Ethnomusicology annual meeting, Montreal.
Leacock, Eleanor
1977 "Women in Egalitarian Societies," in Renate Bridenthal and Claudia Koonz, eds.,
Becoming Visible: Women in European History. Boston: Houghton-Mifflin.
1978a "Society and Gender," in Ethel Tobach and B. Rosoff, eds., Genes and Gender (New
York: Gordian Press), pp. 75-85.
1978b "Women's Status in EgalitarianSociety: Implications for Social Evolution," Current
Anthropology 19(2):247-275.
Langness, L. L.
1974 "Ritual Power and Male Dominancein the New GuineaHighlands," Ethos 2(3):189-212.
Lomax,
1959
1962
1976

Alan
"Folk Song Style," American Anthropoologist61:927-954.
"Song Structureand Social Structure," Ethnology 1(1):425-451.
Cantometrics: an Approach to the Anthrolpologyof Music. Berkeley: University of
California, Extension Media Center.

Lomax, Alan, et. al.

1968 Folk Song Style and Culture. Washington:AmericanAssociation for the Advancementof
Science.
Murdock, G. P.
1967 "EthnographicAtlas: A Summary," Ethnology 6:109-236.
1969 EthnograplhicAtlas. Pittsburgh:University of PittsburghPress.
Murphy, Robert
1959 "Social Structureand Sex Antagonism,'' SouthwesternJournal of Anthropology 15(2):
89-98.
Schieffelin, Bambi B.
1979 How Kaluli ChildrenLearnWhatTo Say, WhatTo Do, and How To Feel: An Ethnographic
Colmpetence.Ph.D. diss., Columbia
Approach to the Development of Commnunicative
University.
Schieffelin, Edward L.
1976 The Sorrow of the Lonel/'and the Burningof the Dancers. New York:St. MartinsPress.
1978 "The End of TraditionalMusic, Dance, and Body Decoration in Bosavi, Papua New
Guinea." (Discussion paper 30-32.) Boroko: Instituteof Papua New Guinea Studies.
1979 "Mediatorsas Metaphors:Moving a Manto Tearsin PapuaNew Guinea," in A. L. Becker
and Aram Yengoyan, eds., The imaginationof reality (Norwood: Ablex), pp. 127-144.
1982 "The bau a CeremonialHuntingLodge: An Alternativeto Initiation,in GilbertHerdt,ed.,
Rituals of manhood: male initiation in Papua New Guinea (Berkeley: University of
California Press), pp. 155-200.

FELD: SOUND STRUCTURE AS SOCIAL STRUCTURE

409

Schlegel, Alice
1977 "Toward a Theory of Sexual Stratification,"in Alice Schlegel, ed., Sexual Stratification
(New York: Columbia University Press), pp. 1-40.
Strathern, Andrew
1970 "Male Initiation in New Guinea Highlands Societies," Ethnology 9(4):373-379.
Strathern, Marilyn
1972 Women In Between. London: Seminar Press.