Learning How to Ask: Native Metacommunicative Competence and the Incompetence... Fieldworkers Author(s): Charles L. Briggs

Learning How to Ask: Native Metacommunicative Competence and the Incompetence of
Author(s): Charles L. Briggs
Source: Language in Society, Vol. 13, No. 1 (Mar., 1984), pp. 1-28
Published by: Cambridge University Press
Stable URL: http://www.jstor.org/stable/4167483
Accessed: 22/02/2009 16:20
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at
http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless
you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you
may use content in the JSTOR archive only for your personal, non-commercial use.
Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at
Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.
JSTOR is a not-for-profit organization founded in 1995 to build trusted digital archives for scholarship. We work with the
scholarly community to preserve their work and the materials they rely upon, and to build a common research platform that
promotes the discovery and use of these resources. For more information about JSTOR, please contact [email protected]
Cambridge University Press is collaborating with JSTOR to digitize, preserve and extend access to Language
in Society.
Lang. Soc. 13, 1-28.
Printed in the United States of America
Learning how to ask: Native metacommunicativecompetence
and the incompetenceof fleldworkers
Department of Anthropology
Vassar College
and the
Committeeon Degrees in Folklore and Mythology
Harvard University
Every speech communityboasts an arrayof devices for characterizingcommunicativeevents. These native metacommunicativerepertoiresare culturally patternedin terms of both use and acquisition. Interviews meet with
varying degrees of success by virtue of their relative (in-)compatibilitywith
the norms underlyingsuch events. An analysis of the way in which Spanish
speakers in rural New Mexico gain metacommunicativecompetence suggests that native metacommunicativeroutines provide a rich source of sociolinguistic and social/culturaldata and that awarenessof these repertoires
can assist fleldworkersin using interviewsmore appropriatelyandeffectively. (Interviewtechniques, metacommunication,acquisitionof sociolinguistic competence, ethnopoetics, New Mexican Spanish)
Speech exerts a powerful role in social interaction,linking the thoughtprocesses
of two or more individuals and coordinatingmany of their bodily movements.'
One of the richest means of orchestratingsuch connections is by the use of
language in describing or evaluating communicative events or processes. This
metacommunicativecapacity is used by speakers to draw the attentionof their
interlocutorsto the ongoing situation(e.g., "are you tryingto tell me that . . ."),
otherspecific exchanges (e.g., "she said, '1reallydidn'tdo it' "), or communicative norms and processes (e.g., "'we don't talk like that at school"). All speech
communitiesprovidetheirmemberswith a varietyof such strategies.These range
between such elementary (and possibly universal) devices as quotation-framing
verbsto complex, culture-specificroutinessuch as discussionsof academicpapers
or Malagasy bride-wealthnegotiations (cf. Keenan 1973).
Fieldworkersrely heavily on metacommunicativeinformation,supplementing
their observations with native exegesis. In obtainingthis material, fieldworkers
utilize a metacommunicativeskill which figures prominentlyin the speech economy of their society: interviews (cf. Strauss & Schatzman 1955; Gnimshaw
0047-4045/84/010001-28 $2.50 ?3 1984 CambridgeUniversity Press
I969, I969-70;
Wolfson, 1976). This paper is one of a pair of essays which
examine the usefulness of interviewingas a fieldworktechnique. My goal is not
to urge practitionersto discardthe practice. I simply wantto show thatinterviews
become highly problematicwhen their metacommunicativepropertiesgo unexamined and when researchersfail to familiarize themselves with the metacommunicative repertoireof the society under study.
The otherpaper(Briggs I983a) examines the difficulties entailed in interviewing members of a speech community which adheres to communicativenorms
which differ substantiallyfrom those presupposedby the interview. These pertain to such factors as communicative channel, social situation, key, genre,
interactionalgoals, type of communicativeevent, social roles, establishingand
maintainingreference, and the like. The data were drawnfrom my own communicative blunders in research with Spanish speakers in northernNew Mexico
This essay focuses on the natureof metacommunicationand the relationship
between interviews and other types of metacommunicativeevents. I wish to
show that metacommunicationis, like communicationin general, culturallypatterned in terms of both acquisitionand use. A sketch of the metacommunicative
repertoireof one society - Mexicano New Mexico - and the mannerin which
individuals gain competence in these routines is used in outlining the sociolinguistic bases of the problemscreatedby relying primarilyon interviewsin
conducting research in this society. A variety of metacommunicativeevents,
including elicitations of repetitionsfrom young children, political rhetoric,and
scripturalallusions, are analyzed in pointing to the researchpotentialof speech
events that natives use in explicating communicative events and processes to
each other. The paperconcludes with a typology of metacommunicativeevents.
This provides a basis for exploring the relationshipbetween interviewsand other
types of metacommunicationin theoreticalterms.
The data were collected in Cordova, a communityof about700 inhabitantsin the
mountains of northernNew Mexico. The residents are Mexicanos, with the
exception of one recent Mexican immigrant;two middle-agedAnglo-Americans
who have marriedCordovans;and occasionally a few transientAnglo-American
youths. Mexicanos are descendants of primarilySpanish and Mexican citizens
who settled in New Mexico and southern Colorado during the seventeenth,
eighteenth, and nineteenth centuries. Their ancestry includes a significant Native-Americanelement, but the Mexicanos consider themselves to be culturally
All native C6rdovansare fluent in New Mexican Spanish;the distinctiveness
of the dialect was recognized through the pioneering work of Aurelio M. Espinosa, Sr. (igi I, 1930) and Juan B. Rael (I937). New Mexican Spanishdiffers
from the standardand from other Southwesterndialects on primarilyphonological, lexical, and suprasegmentalgrounds. Most residentsunderfifty years of age
are bilingual, although to varying degrees. English loanwords are common,
although they are used less frequently in C6rdova than in the largertowns and
Awareness of the fact that language use is culturally patternedis old, and the
rangeof variationhas been documentedwith increasingcare in recentdecades by
sociolinguists. Repertoiresof metacommunicativeroutines vary from society to
society with respect to form and function as well as in the rules of alternationand
cooccurrence (cf. Ervin-TrippI972) which constrain their use. Hymes has argued repeatedly that one of the central tasks which awaits a fieldworkerupon
entering another society is acquiringat least a minimal level of sociolinguistic
competence (see especially I97 I). As Karpand Kendall(I982) rightlyargue, the
task confrontingthe fieldworkeris not identical to that encounteredby a native
child. Nevertheless, the means by which membersof the society gain metacommunicativecompetence is generally the gauge which will be used in interpreting
the efforts of nonnatives. Discovering this process is thus critical for conducting
and for understandingfieldwork.
Native models of socialization
One way in which sociolinguistics has departed from Boasian and structural
traditionshas been to include native conceptionsof languageand languageuse as
an integralpart of the agenda for the ethnographicstudy of communication(cf.
Hymes I972:39). This task is especially importantto researchon the acquisition
of sociolinguistic competence. Ochs (I982) finds that language acquisition, language teaching, and the evaluation of communicative competence are closely
relatedto the views that the members of a society hold of the learningprocess.
My own data similarly suggest that the role of different metacommunicative
routines in language acquisition is best comprehendedagainst the backdropof
Mexicano epistemology.
An agriculturalmetaphoris frequentlyused in explicatingthe mannerin which
a person acquiresa skill or body of knowledge. You must have seeds in orderto
plant. These are equivalent in pedagogical terms to talento 'talent'. Talento
refers to a God-given aptitudefor learninga given thing. Talentowill not develop, however, in isolation - it must be desparramado'spreadout' or 'scattered'.
This refers to the need to observe those who have expertise in the endeavor.
Interes and concentracion are also required;a person must want to learn and be
able to concentrate.The next stage involves wateringand weeding the growing
plants. The allusion here is to taking the knowledge which has been gained
throughobservation and intensifying and extending it throughimitation. In the
case of the acquisition of communicativecompetence, this refers to reiterating
the words of one's seniors. A person who lacks paciencia 'patience', who wants
to absorba topic immediatelyand believes s/he can be an expert shortlythereafter, will not gain competence. Such attemptedprecociousnesswould also demonstratea lack of respeto 'respect'. Respeto, a centralculturalvalue, prescribes
deferringto one's seniors' greatercommand of the pertinentskills.
If properlycultivated, the plants will grow to maturityand can be harvested.
Another means of scattering your 'talent' gains importanceat this point. A
musicianmay be able to play well. If s/he does not get out of the house and play
for a wide range of audiences, however, s/he will not become a competent
performerand will not gain a reputationas a great musician. This correspondsto
the point at which individualsare expected to exhibit their rhetoricalvirtuosity.
Once a speaker has masteredall of the requisite skills, s/he must practicediscussing community affairs, performingproverbs, singing hymns, or the like in
Observationand elicited repetition in early childhood
As Snow (1977:37), Ochs (I982), and others have noted, societies vary greatly
in what is believed about and expected from children. In Mexicano society,
caregivers from six years of age on up are quite voluble with infants. This does
not, however, reflect the assumption that infants are trying, if imperfectly, to
communicatewith them (as has been assertedof Anglo-Americans[cf. Bates et
al. I979; Shotter 1979; Trevarthen'9791). It is ratherbelieved thatchildrenmust
be able to observe a great deal of speech if they are to be expected to "pick up"
the language. Such verbal interactionalso enhances the child's interes and concentracio'n,ensuring that observation will promote learning.
Eliciting Repetitionsfrom YoungChildren. Duringthe first year of a child's
life, caregiversuse interrogativeforms in drawingthe young child into conversation. Caregiverdiscourse often consists of asking a series of questionsover and
over, varying pitch and word stress. Once childrenreach the holophrasicstage,
caregiver emphasis shifts from encouraging observation to eliciting responses
from the child. The point, however, is not to induce the child to come up with
novel utterances. Caregivers assume that comprehensionprecedes production;
the absence of an intelligible response thus is not necessarily interpretedas
reflecting a complete lack of understanding.Observationalone is not a sufficient
developmental basis for gaining the ability to generate the properforms in the
appropriatesocial settings. "Imitation" is seen as a requisiteintermediatestage
between comprehensionwithoutproductionand the acquisitionof truecommand
over linguistic forms. Interestingly,this sequence from comprehensionto imitation to production coincides with a leading theory of linguistic ontogeny (cf.
Menyuk I 977:68-69).
As children approach two years of age, one means of eliciting responses
assumes a centralrole in caregiver-child interactions.Any person who is about
six years of age or older is likely to amuse her/himself by directing the baby to
'say X to so-and-so.' The following tape recordingwas made while Linda (1.9)
was being held by her grandmother,Lupe. Linda's mother, father, grandfather,
and her uncle Ben (Lupe's oldest son) were standingneara large woodpile. I was
standingbetween the two groups, with my tape recorderpointed in the direction
of Lupe, Manra(Lupe's youngest child), and Linda. Maria approachedLinda
frequentlyand gave a gentle yank on one of her legs, to the amusementof all.
The game continued intermittentlyfor eight minutes, while Lupe attendedalternately to the adults' conversation and to Maria and Linda's play. Then Ben
pretendedto charge Linda. She squealed with delight at first, but soon began
crying. Lupe told Ben to desist, using a serious tone of voice, which he did
immediately. When Mariabegan anew to pull on Linda's legs, Lupe triedto end
the game.
i. Eliciting repetitionsfrom Linda (i. 9)
Lupe: iDejala!
(Linda startsto cry)
Lupe: iNo! iDe todos rumbos la jalan!
(everyone laughs, including Linda)
Lupe: Como se rie. Pobrecita.
Leave her alone!
No! They pull at her from all
How she laughs. Poor little
(all laugh)
Lupe: No, Maria, leave her alone.
Maria: Okay.
Lupe: No la jales.
(said in sing-song; Maria and Linda
laugh loudly)
Lupe: 6Quieresque yo te jale los
shongos2 de indio?
Maria: No.
(everyone laughs for 10 seconds)
Ben: Dile, "shongos de perro."
Linda: [anos]
Ben: Yah, "Maria, shongos de
Lupe: iNo, Maria, no la jales!
(previous utterancestated in stem,
serious tone)
Ben: "Shongos de perro," dile.
Linda: [biygal
No, Maria, leave her alone.
Don't pull on her.
Do you want me to pull on your
Indian braids?
Tell her, "dog's braids."
Yah, "Maria, dog's braids."
No, Maria, don't pull on her!
"Dog's braids," tell her.
Ben: No, dile "shongos de perro."
Linda: [sagmiy, samiy]
Lupe: Leave her alone, te digo.
Ben: "Shongos de perro," dile.
Linda: j[miya]?
Ben: "Shongos de perro." That's
No, tell her "dog's braids."
[sagmiy, samiy]
Leave her alone, I'm telling
"Dog's braids," tell her.
"Dog's braids." That's good!
The game then ended; Linda was put down on the ground, and Lupe focused on
conversing with the other adults.
In the interestof brevity, I will confine my analysis of this text to pointingout
(i) the reflection of importantculturalpremises in the discourse, (2) the rangeof
the social roles which are played by the participants,and (3) the role of metacommunication.
Cultural Premises. The interpretationof this conversationinvolves important culturalassumptions.One of these underliesthe mother'sambivalence.She
is obviously enjoying the game, but her admonitionsto Mariaand Ben are not
entirely playful. Infantsand young childrenare inocentes 'innocentones'. They
cannot be blamed for their own actions, since they are too young to know the
difference between right and wrong. But they are also highly susceptible to
supernaturalharm, which can manifest itself as a potentially fatal physical illness, such as mal de ojo 'evil eye' or susto 'fright' (cf. Trotter & Chavira
Playing with babies is enjoyable, but it is also potentially dangerous. If an infantor young child laughs loudly for an extendedperiodof time,
s/he becomes a potential target for supernaturalillness. (Anglo-Americansare
notoriouslyindiscreetin this regard.Since my daughterspentthe first year and a
half of her life in the field, I was continuallybeing lecturedon the inadvisability
of eliciting too much laughter.) Lupe was becoming more and more concerned
about this possibility throughoutthe interaction.This element seems to simply
add to the tension, hence the humor, in the course of the game. Lupe's fear is
beginning to outweigh her enjoymentduringthe second partof the text, and she
brings the game to a halt shortly thereafter.
It should be noted that Ben's feigned attacks on Linda elicit quite different
reactions from Lupe. Ben is immediately admonished No me gusta que la
agarres, Ben 'I don't like you grabbingher, Ben'. The children's reactionsare
similarlycontrastive- Ben ceases this form of play at once. He then picks up on
Lupe's semantic lead by pointing out Maria's long and beautiful braids, thus
becoming a second ally for Linda. This contrastin the children'sbehaviorcannot
simply be attributedto age or gender (Mariawas six and Ben was fourteen).Two
boys span this age gap in the family, and they persist longer in the face of such
admonitionsthan Maria. It is rathernecessary to understandthe special role that
accrues to the eldest male child, who is often referredto as the papacito 'little
father', in Mexicano families. Being accordeda quasiparentalstatus, his responsibilities are much greater, the range of acceptable behavior is much narrower,
and his rights are correspondinglygreaterthan those of his siblings.
A thirdculturalpremise involves the importanceof fighting one's own battles
and defending one's dignity. From the time that they are ambulatory,children
are not picked up when they injurethemselves slightly or are roughedup by other
children. (Anglo-Americans are considered horrendouslyindulgent in this respect; I was constantly being told that I was spoiling my daughter.) Constant
interventionsare said to preventchildrenfrom learninghow to stand up for their
own rights and to resist assaults on their person or reputation.Maintainingone's
dignidad de la persona 'personaldignity' is the key to preservingself-respectas
well as a good reputationin the community, and actions which are construedas
attacking an individual's dignidad de la persona are treated with the utmost
seriousness. Linda is still too young to fight her own battles. But Lupe is
encouraging her to defend herself (rather than to rely solely on caregivers'
efforts) by framing some of her admonitionsas Linda's own words throughthe
use of the dile X "Y" formula.
These data strongly supportOchs's contention that the mannerin which individuals gain communicative competence reflects and is based on "a particular
set of cultural values and beliefs" (i982:88). Elucidating the meaning of this
shortconversationprovides the analyst with a good point of departurefor investigating Mexicano conceptions of the person and the importanceof this cultural
constructfor social interaction.This is, of course, hardly surprising.As CookGumperzand Gumperz(1976) and Ochs (1979, I982) have argued, caregivers'
speech often focuses on basic cultural knowledge as well as features of the
communicativesituation itself that are often assumed in adult-adult discourse.
Alternationof Social Roles. Interpretingthis interactionalso presupposesthe
ability to discern the manner in which the participantsare moving between a
numberof different social roles. Lupe's participation,for example, is structured
by her movements between two different role sets. On the one hand, she alternates between a mother role, playing with and supervising the children, and
interacting with the other adults. Lupe plays two roles within the children's
game, altemately "playing along" with and enjoying the game and trying to
extricate Linda from a potentiallydangeroussituation. Manrasimilarlyplays the
partof the mischievous child, pretendingto attackher niece, as well as thatof the
obedient daughter,respondingseriously to her mother's admonitions("okay").
She also distances herself from the child's role entirely earlier in the interaction
by assuming the role of a dictatorialteacher (sie'ntensey mirense 'sit down and
watch'). Note that this places both Linda and Lupe in the role of naughty
students;Lupe apparentlyappreciatesthe humorof this suddenreversal, because
she laughs heartily.
Ben enters the interaction, having just finished carrying a truckloadof firewood to the woodpile, by displacing Maria in the role of mischievous attacker.
After being scolded by his mother, he assumes the role of an older sibling who is
admonishinghis junior. He does not, however, discardhis playful maliciousness
entirely, but simply converts it into a means of playfully sanctioningthe misbehaver. His allusion to 'dog's braids' is derogatory;it may be slightly off-color
as well, equatingMaria's braidswith a dog's tail and thus its posterior,but I did
not pursuethis possibility with my consultants. Ben also distanceshimself from
the child's role, assisting his mother in admonishingMaria. This movement is
completed within the game when he switches into English and changes voice to
play the role of parentwith "that's good!" Finally, both Lupe and Ben alternate
between playing themselves and assuming Linda's part. The latter is accomplished by creating lines for Linda and then reciting them themselves.
Withinthis shortdialogue, Lindahas been exposed to a broadrangeof familial
roles. Each is similarly presented vis-a-vis complementaryroles, providinginsights into the norms for interactionbetween elder and junior siblings, parents
and children, teachers and students, and so forth. Exchanges between persons
playing these roles also provide Linda with insights into the rights and obligations which accrue to individuals who stand in these relationships.Lupe's frequent reversion to a serious, parentalposition enables her to provide a running
metacommunicative commentary on the appropriatenessof these roles. Obviously, close observation and analysis of such interactionscan provide fieldworkers with insights into these areas, which are of central socioculturaland
sociolinguistic importance.
The Role of Metacommunication. The precedingmaterialraises the question
as to how both Linda and the fieldworkercan perceive such alternationsin the
discourse. Cook-Gumperzand Gumperz (1976) argue that communicationis
punctuatedwith "contextualizationcues" which mark relevant featuresof the
social and linguistic setting, thus providinginterpretiveframeworksfor deciphering the meaning of other participants'signals and for shapingone's own contributions. The central contextualizationcues in the precedingtext are changes in
prosodic features and the elicitation of repetitionsfrom Linda. Lupe, Ben, and
Manraalter the pitch, quantity, rhythm and stress patterns, and the speed of
utteranceof their speech. In some cases, these are used in markingan utterance
as mothereseor baby talk. But they are also used by Lupe in framingthe relative
ludic versus admonitorycharacterof her statements.
The best example of such variations is providedby the three ways that Lupe
articulates no. No's which are framed as feigned prohibitionsfeature a slight
increase in pitch and quantity over that of the adjacentlexemes, with a slight
prolongation of the vowel. Lupe marks another admonition (no la jales) as
facetious throughthe use of a "sing-song" intonationwith exaggeratedcontrasts
in pitch. Lupe also uses no in an elicitation device earlier in the game. Here the
no in "'no,' dile" ' "no," tell her' is distinguishedby a greaterrise in pitch on
the word itself; the utteranceas a whole is markedby a diminutionin the quantity
and by laryngeal constriction - the illusion of a "small" voice. All of Lupe's
no's following her initial "leave her alone" conformedto a thirdpattern.Here,
stridencywas complementedwith greatly increasedpitch and quantityto create a
sharp, stem tone. The no's which are followed by a laugh are perhapsthe most
interesting. They not only constitute a strikingcontextualclue to Lupe's simultaneous enjoyment in and fear of this situation, but also suggest that she is at
times playing this tension for dramaticeffect.
A second type of contextualizationcue is providedby the repetitionelicitation
formulae. These provide the child with a host of pragmaticinformationon the
speaker'sperceptionof the ongoing social situation.They are often used to draw
the child into the interactionand to keep her/his attention.They provide a clear
index of the caregiver's desires, telling the languagelearnerwhom to addressand
what to say. The formula specifies the types of assistance that the baby can
reasonably expect to obtain from the third party, thus illuminating the norms
which govern such relationships.The utterancealso inforrnsthe child as to which
speech acts are appropriatefor obtaining the desired effect. Lupe and Ben are
using these formulae to teach Linda to tease, scold, and to defend herself verbally in this case. Such formulae are also used in teaching children how to
shame, insult, console, plead, demand, and the like. Repetition elicitation formulae thus provide a link between a particularconversational setting and the
relevant social and sociolinguistic norms.
The questionremains, however, as to the natureof Linda's involvementin the
interaction. In this text, it is apparentthat Linda has acquired the ability to
identify the first partof adjacencypairsand to providesecond parts,a skill which
commonly emerges at about this age (Ervin-TrippI979:412). She repeats segments of the embeddedutterances;her success in this pursuitat 1.9 years of age
varies between being accurate,even if partial,and unrecognizable.Note thatBen
evaluatesher approximationsand providesfeedback. Having left the field shortly
afterthis interactionwas recorded, my next observationswere taken when Linda
was 2.3. At that time she had mastered this interactionalpattern, providing
accurate and complete repetitions. The forms she repeated then were not only
multiclausal, but many contained English phrases embedded in Spanish sentences (e.g., dile a tu grandma ['tell your grandma'] "I love you eight-nineten").
Interestingly, the distribution of this strategy for the acquisition of sociolinguistic competence appearsto be quite widespread. Eisenberg (I982) presents an extensive description of the use of dile among recent Mexican immigrants in northernCalifornia. The similarity of the two sets of data is striking:
"The childrenwere told what to say to initiate interaction,to get theirown way,
to tease andjoke, to greet, to apologize, etc. Ratherthan intercedefor the young
child in an interaction,the adult speakerspushed them to interacton their own"
Ochs (1982) and Schieffelin (1979) have also reportedclosely analogous data from WesternSamoa and PapuaNew Guinea. How do we accountfor
these parallels?I have tried to show that caregivers' speech containscontextualizationcues which foregrounda wide rangeof aspects of the presentinteraction.
Childrenthus "'read"these subtle cues, along with the widercontext, in discerning conventionalizedexpectationsfor these types of settings (cf. Cook-Gumperz
& Corsaro 1976). Caregivers construct their utterancesin such a way that the
formalfeaturessimultaneouslyindex (in Peirce's 11932:2.305] sense of the term)
what they take to be the crucial dimensions of the present interactionand its
socioculturalbackground(cf. Corsaro 1979; Ochs 1979; Cicourel 1970, 1974).
Repetitionelicitation formulaeare particularlyuseful in this regard,becausethey
convertconventionalizedexpectationsinto one of the parametersof the context the child's speech. They constitute the clearest (although intentionallynot the
most direct) means of telling the child "this is how the normativebackground
intersects with this particularinteractionat this particularmomentfor you."
Bearing messages in later childhood
Beginning at about six years of age, children assume an importantrole in adult
communication. The dile X "Y" formula is used in asking a child to relay a
short message to a member of a nearby household. A child will be told, for
example, dile al Santiago que me preste una pala 'tell Santiago to loan me a
shovel'. The child will then go immediatelyto Santiago'shouse and tell Santiago
(or, in his absence, anothermemberof the household) in a very high-pitchedand
excited voice, dijo me daddy si tienes una pala pa' prestarle 'my daddyasked if
you have a shovel to loan him'. Note thatthe requestitself foregroundsthe nature
of the utterance(as reportedor indirect speech), the identity of the sender, the
purposeof the visit, and the mode of transaction(a loan). Childrenof this age are
thus expected to be able to interpretand generatea numberof speech act types.
Since the child provides a report of the senior's speech (ratherthan a direct
quotation), s/he must modify pronouns, verbal inflection, and syntacticform as
well as transformone type of directive (an explicit request to convey information) into another(an implicit request by a third party for a materialitem).
What sort of communicative competence, then, do six throughtwelve year
olds possess, and how are these skills regardedby the community?As Goffman
notes, children possess incomplete selves. They can enter other
householdsmore freely thanadults, and theirvisits do not requirethe residentsto
shift from intimate, mundane to formal and ceremonial presentations. As
Hotchkiss (1967) suggests, this renders them perfect messengers for errands
which would cause adultsto lose face (e.g., requestinga petty loan) or would be
nearly impossible (e.g., finding out what your neighborsare up to). Cordovan
children are thus sometimes told by adults or older siblings to observe what is
said and done at a neighboringhousehold or in the street and to reportback to
their seniors. This advantageousposition is not, however, without its price tag.
Childrenare not yet consideredfull-fledged actorson the householdor community stage, and they accordingly are given few lines in the substantivedialogues.
The youths' strictly linguistic competence is not at issue here. They have already
mastereda broadrange of speech acts and, as demonstratedby the verbalgames
they enjoy with peers, they can be quite creative. They are, however, expected to
exhibit respeto for theirseniors at all times, that is, to honorthe latters'considerably greateraccess to the center stage of interaction.In short, they are expected
to have mastered a set of complex metacommunicativeskills relating to the
of speech in a mannerwhich will be deemed approtransference/transformation
priatein a broadrange of social settings. I noted above that repetitionelicitation
formulaeprovideyoungerchildrenwith crucial informationregardingtheirrights
vis-a-vis other children and adults. Carryingmessages provides older children
with models regardingadult privileges and responsibilitiesvis-a-vis otheradults.
The metacommunicativepersuasiveness of adults
Once individualshave begun earning a living, married,startedraising a family,
and establisheda household, they are presentedwith opportunitiesto make their
voices heard, so to speak, in interactions with members of their and their
spouse's extended families. As they progress from thirty to sixty years of age,
many assume roles of importancein religious voluntaryassociations, irrigation
ditch associations, parishaffairs, domestic waterand land grantassociations, and
other intracommunitygroups. If they prove themselves to be thoughtful and
persuasivespeakers, their statementswith regardto communityaffairscan come
to be taken quite seriously by persons of all ages. These are the years in which
men and women who possess talento for public speakingare expected to develop
and exhibit their rhetorical facility. A great deal of prestige accrues to the
community member who can sway an audience in the course of a meeting or
other public gathering.
The following transcriptis taken from a tape recordingof a meeting of the
boardof directorsand the membershipof an associationof the users of the local
water system. As is the case in any small community, an issue occasionally
arises which creates a persistent dispute. Problems with the community water
system had resultedin frequentshortagesover a periodof years. This serious and
at least temporarilyirreconcilablesituationhad produceda numberof lasting rifts
that surfaced during the meeting. One involved two geographic parts of the
community, while the other divided the association's boardand its membership.
The presidentof the board used his opening remarksin directing the discussion
toward the technical and legal dimensions of the situation. Everyone knew,
however, that the true agenda was the members' feeling that the boarditself was
behind many of the difficulties. A room in the local school was accordingly
packed. One member, whom I refer to as the Disputant, had been particularly
vocal in denouncing the actions of the board. He interruptedthe president's
remarkswith a series of questions, but he had not been able to induce the board
members to address the central issue.
2. Dispute at board meeting3
What is needed . . .
I. Lo que se necesita . . .
2. Yo soy de eso opini6n.
A mi no me importa
quien es usted,
de comision o como
quiera que sea.
Para ml, todos son
Pero si digo una cosa.
IO. Lo que se necesita
en este negocio
12. [es que] se arregle
13. y camine todo
I4. como sopone caminar.
I am of this opinion.
It doesn't matterto me
who you are,
a board memberor whatever
you want it to be.
For me, all are good.
(meaning 'I am impartial')
But I will say one thing.
What is needed
in this business
[is that] it work itself out
and that everythingproceeds
the way it is supposed to proceed.
15. Y no que . . .
And not that . . .
i6. Y no camine bajo de,
17. bajo de politica
I8. o bajo de envidia
I9. o bajo de esto y el otro.
And that it doesn't proceed under,
under [the influence of] politics
or under [the influence of] envy
or under this and that.
20. Es hacer . . .
It is to do . . .
I am revealing the fact
that there is a lot of envy here,
as much from one direction
as from the other,
so we are all wrong.
I like to make [myself] clear
and I like to say
what I am going to say
in public.
Yo estoy cantando a conocer
que aqui hay mucha envidia,
tanto por un rumbo
como por el otro,
ya estamos todos mal.
A mi me gusta hacer claro
y me gusta decir
lo que voy a decir
a un puiblico.
I don't . . .
30. Yo no . . .
Yo no le rodeo a nadie.
32. Se necesita que todos nos
trabajemoscomo hombres
34. y que todos digamos la verdad
35. y no la mentira.
Other board member:
36. La verdad es/
I don't walk aroundanybody.
What is needed is for all of us
to work like men
and for all of us to tell the truth
and not lies.
The truthis/
37. /Porque yo se ciertamente
38. que hay muchas cosas
escondidas en esa linea.
/Because I know with certitude
that there are many things
hidden in that line.
The other boardmemberthen responded, referringdirectly to the chargethatthe
board had been getting more than its fair share of water.
The Disputant's (hereafterD) words constituteda focal event in the meeting,
and they were highly effective in bringing out the topic he wished to discuss.
Wherein lies their efficacy? Theform of the discourse is crucial. In comparison
with the preceding statements, D's words were marked from the start by an
increase in volume and a decrease in the speed of utterance.The pitch, loudness,
and stress evince a much greater range of variation than the surroundingdiscourse. The individual utterances are separatedinto clear tone groups, which
come close to being poetic lines. The nucleus rests in the final word in all
utterances save i6, 20, and 30, which are false starts, and in the last line
(38-39). Each is followed by a longer than normal pause. The speech is also
segmented prosodically into quasi-stanzalikeunits (I-8, 9-19, 20-25, 26-3I,
and 32-39). The end of each "stanza" is markedby a sharpdropin pitch and an
even longer pause.
D's words were preceded by a period of short turns with frequent interruptions. Once D secured the floor in i, all side conversationsceased, and all eyes
turned to him. Even when the other board member tried to prevent D from
coming to the point in 36, D was able to retain the floor. His slow, clear,
measured, rhythmic, forceful speech marked his statement as climactic. This
formal elaboration highlighted his abilities as rhetorician, thus increasing the
stakes, so to speak, of the success or failure of his efforts. When D shifted
register, he signalled the audience that the meaning of what he was aboutto say
must be sought less in its overt referentialmeaning than in the way in which the
poetic form provided clues to underlying, implicit meanings.
This strategywas crucial. D and his fellows had promptedthe boardto call the
special meeting in orderto bringtheircharges againstthe latterout into the open.
This had not taken place thus far, and the discussion of technical problemswas
becoming so tedious that many of the participantswere losing interest and
beginning side conversations. D had to bringthe key issues to the fore soon if the
momentum produced by the special circumstances of the meeting was to be
preserved. Nevertheless, to accuse the board directly of taking more than their
shareof the scarce water and of covering up their actions would have constituted
a real affront. Such public accusations controvertthe accused's dignidad de la
persona 'personal dignity', and are likely to spark a real confrontationand to
engender lasting enmity. D thus sought to force the issue withoutdirectly referring to it himself.
D accomplished this feat by juxtaposing allusions to three fundamentalMex13
icano values with implicit allusions to the basic conflicts which underlaythe
meeting. Lines 3-8 express a basic ideal of interpersonalrelations- tratarigual
a todos 'to treateveryone equally'. One of the highest complimentswhich can be
paid a person is that s/he 'treatseveryone well' (bien) or 'the same' (igual). This
value is seen as following from Christ's love for all humans. Christianvalues
dictate the avoidance of pettiness, spitefulness, and conceit. D thus implicitly
contrastshis compliance with these normswith the members'assertionthatsome
board members feel that their position entitles them to a greater share of the
water. In otherwords, D is saying to the president, "you may thinkthatyou, as a
board member, are better than I am, but I treat everyone equally well."
Lines io through 25 juxtapose a second value, corporatism, with another
dimension of the conflict, that between the two sections of the community.La
gente 'the people' should ideally be united, sharinglabor and resourcesin times
of need. This, too, has a Biblical precedent; it is often said that 'we are all
brothersand sisters, because we are all children of God.' This is the way that
community affairs 'are supposed to go.' Envidia 'envy' and politica 'politics'
are the majorobstacles to realizing corporatism.D goes on in 21-25 to suggest
that all partieshave given way to envidia. The use of rumboin this context is a
double entendre. Todos rumbosis the common expression for 'all over'. But the
otherside of the communityis generallyreferredto locally as el otro rumbo 'the
other side'. D's allusion to the internecine conflict thus similarly combines a
veiled but clear reference to a basic aspect of the dispute with a value-laden
A final juxtapositionof value and conflict involves D more directly. Priorto
the meeting, D had personallyqueriedmost of the memberswith respectto their
feelings aboutthe board's actions. Since he had not yet spokendirectlyto anyone
on the board, some of the commissionersaccused him in privateof being afraid
to confrontthem directly. This constituteda serious charge. The inabilityto face
one's opponent in person constitutes a loss of face and a diminution in one's
sense of personal dignity. D asserts, on the contrary,that he likes to speak not
only clearly but publicly, and that he 'won't walk around'(i.e., isn't afraidof)
Now that he had adumbratedboth the conflicts and the perceived strengthof
his own position, D goes on to state his charge publicly - if still implicitly.
Having declared himself to be forthrightand truthful,he exhortsall participants
to behave similarly and to tell the truth.A boardmembercorrectlyperceivesthat
D is just aboutto drive the last nail into the board'swould-becoffin, so to speak,
and tries to interrupthim in 36. Not lacking in rhetoricalstrengthat the moment,
however, D concludes 'because I know with certitudethatthereare many things
hidden in that line'. This may seem cryptic at first glance. Several of the key
issues in the dispute revolved, however, arounda water line that passed next to
the houses of several of the board members, and this line had been discussed at
length in the precedingdialogue. Linea also denotes a line of reasoning, and the
ambiguityis very much a partof D's strategy. In essence, D is telling the board
directly "we all know that your statements about the real cause of the water
shortage are lies, and it is time to tell the truth."
D's plan worked. Withoutmakingany directaccusations,he managedto bring
the chargesout into the open. His allusions were so clear thatit was necessaryfor
the board members to refer to them directly in responding to D. The second
commissioner proceeded to summarizethe people's feelings with respect to the
board's actions and then to detail her own involvement. My analysis suggests
that D's success rested both on his use of sub rosa argumentationand on his
juxtaposition of commonly accepted, traditional values with specific issues.
These values provide an indisputablesource of legitimacy for his points. They
also provide him with a means of interpretingthe events in line with his position.
D's use of a formal, quasipoeticform for his words highlightedthe importance
of his statement as well as his own rhetoricalskills. By relating these current,
specific issues to basic values, he enhancedtheirconsequence. Gaining recognition for one's verbal abilities is an importantpart of the process of moving
throughthe status of muchacho (lit. 'boy', meaning 'young man') and into fullfledged adult status. Speaking out on the affairs of the community is the most
importantmeans of establishing one's reputationat this point in life.
Just as such individuals gain access to new sociolinguistic skills during this
period, however, importantconstraintsare operative as well. Muchachos draw
on traditionalvalues in coping with contemporaryaffairs, but they lack the right
to speak authoritativelyon the locus of these principles, 'the elders of bygone
days.' These values are most directly embodied in 'the talk of the elders of
bygone days,' which consists primarilyof such folkloric forms as proverbs,oral
historicalvignettes, and scripturalallusions. Although persons underthe age of
sixty may refer to such forms, they are not generally accorded the right to
performthem, particularlyin front of their elders. Nevertheless, public rhetoric
provides them with opportunitiesfor practicingjust the skills that will later be
necessary for mastering 'the talk of the elders of bygone days.'
Geriatric metacommunicativepower: 'The talk of the elders of bygone days'
Los ancianos or los viejitos 'the old folks', females and males of about seventy
years of age and older, are the legitimate bearers of traditionalknowledge in
Mexicano society. The centralprincipleof respeto 'respect'constrainsboth what
and when individuals can speak. The oldest members of the community are
obligated to preserve and transmit 'the talk of the elders of bygone days'; they
are the only speakers who are accorded the right to use performancesin specialized genres in doing so. Just as the more esteemed speech types and the
knowledge they presuppose belong to the elders, theirs is the floor as well. A
'respectful' younger speaker will always yield the floor to an older person, will
not contradicther/his senior, and will not grow angryif contradictedby a viejito.
The elders thus do much of the talking, especially when two or moreof them are
'The talk of the elders of bygone days' emerges in three main types of social
situations. First, elders who are highly versed in these traditions enter into
exchanges of oral historical vignettes, proverbs, and the like. These are true
virtuoso performances. The participantsare experts in presentingthese speech
forms;the young and inexperiencedmay watch, but they keep theirdistance.The
texts are presented in an abbreviatoryfashion, and they are not explicated.
Anothercontext consists of discussions of the past by personsof roughlythirty
to sixty years of age. These take place in informalsituations, such as conversations between good friends and trips to the mountains.The materialis generally
drawn from the more recent past, and it is presentedas the personalrecollection
of the speaker (e.g., yo no tenia mas que como diez anos cuando . . . 'I was
only about ten years old when . . .'). Such exchanges of stories are seen as
recreational,ratherthan competitive or pedagogical.
The thirdand most common arenafor presentationof this 'talk' is pedagogical
discourse. Here, an elder or elders engage one or more younger persons in a
dialogue about the past. The explicit object is to inculcatethe basic moralvalues
that exemplify the actions of 'the elders of bygone days' to succeeding generations to such a degree that they come to be reflected in the thoughtand action of
the latter group.
Common settings for pedagogical discussions are informal gatherings in
livingroomsor kitchens;elders also frequentlylaunchinto pedagogicaldiscourse
upon passing a local landmarkon a walk or a trip into town. These conversations
are generally initiatedby the elder, althoughyoungerpersons' questionsregarding some facet of the past can serve as an entree into a lengthyexposition. Elders
have much more control over what will be discussed and how the conversation
will proceed; nevertheless, pedagogical discourse is dialogic in nature. The
elders often query theirjunior(s) as to whethers/he is comprehendingwhat has
been said by ending a statement with such questions as Gno ves? 'you see?'
isabes como te digo? 'do you know what I'm telling you?' or 6verdad? 'right?'
Similarly, the "student" is free to ask for reiterationsor furtherelaborations.
Younger persons use questions (e.g., iverdad? or e;si? 'really?') or such exclamations as si 'yes,' "um hum," and the like in signaling the elder that they
are comprehendingthe "lesson."
Scriptural Allusions. Having presented analyses of other genres elsewhere
(0983b, in press), I will draw on performancesof scripturalallusions in illustrating 'the talk of the elders of bygone days.' Scripturalallusions are a fairly
flexible genre, so to speak, since they can be used in a varietyof situationsand
can incorporatetexts of various degrees of length and complexity. They consist
of the insertion of a text bearing on moral principles which is identified as
scripturalinto ongoing discourse. The following scripturalallusion was used by
Aurelio Trujillo; he and his wife, Costancia, informally adopted me when I
enteredC6rdova in 1972. Mr. Trujillo was highly religious, and he placed great
emphasis on teaching me the importanceof placing total faith in God. He was
commentingon a growing tendency on the partof Christiansto lose the sense of
the divine presence and to maintaintheir devotional practices.
3. Scripturalallusion
Aurelio Trujillo:
Pero al cabo que Dios los sabe
lo mismo que premia al (pecador),
porque dice
"perdonaral inocente,"
"porque no sabe lo que hace."
Y todos semos [sic] hijos,
todos semos brothers,
todos semos hermanos.
Aurelio Trujillo:
Y muchos no,
porquetiene un nickel more
que el otro;
es orgullo.
la vanidad
se acaba,
no tiene fin.
El dinero se acaba,
no tiene fin.
De modo que hay tres cosas
que no tienen fin.
Y la amistad reina en la vida.
But in the end God knows how to
them just as he rewardsa (sinner),
because He says
"pardon the innocent,"
He says,
"because he knows not what he
does. "
And we are all children,
we are all brothers,
we are all siblings.
But not for many,
because s/he has a nickel more
than the other;
it's pride.
comes to an end,
it is pointless.
Money comes to an end
it is pointless.
And so there are three things
that are pointless.
And friendshipreigns in life.
This scripturalallusion is metacommunicativein two majorrespects. The allusion is to Christ's invocation on behalf of his crucifiers(Luke 23:34). The term
inocente refers to those who lack the knowledge of good and evil, and it is
generally applied to young childrenand to imbeciles. The term is extendedto all
humans in the text, however, since we lack divine omniscience and our actions
are fallible. God will accordingly forgive those who forget Him just as He
forgives sinners. Mr. Trujillo extends this logic in this allusion and in two
subsequent allusions to suggest that humans must love their fellow inocentes,
regardlessof what wrongs they may commit against them. He also lists what he
sees as the three greatest obstacles to the expression of such brotherly/sisterly
affection - pride, vanity, and avariciousness.
The second majorthrustof this allusion - its bearingon the situationat handis especially clear. 1 had inauguratedmy fieldworkin the communitythreeweeks
previously, and I met Mr. and Mrs. Trujillo four days before the presentmeeting. During this initial session, my position was quite ambiguous. I was a
stranger, an Anglo-American, who came with a tape recorder, and wished to
conduct research. As an educated representativeof the superordinatesociety, I
possessed a high degree of status. I was, however, also a nineteen-year-oldwho
obviously sought the Trujillos friendship and wanted to learn from them. Mr.
Trujillo's speech reflected this ambiguity. He used formal/deferential,usted
forms while occupying the preeminentposition in an asymmetricmode of interaction - pedagogical discourse.
But Mr. Trujillo had an obvious interest in playing down the deference-tohearer and emphasizing the solidarity-with-hearerdimension (cf. Silverstein
His speech was accordingly pronouncedly performative in that he
sought to create a close friendshipand to be accordedthe deferencethathe, being
fifty years my senior, deserved. Like Mr. Trujillo's utterancesin this conversation as a whole, a numberof features of the scripturalallusion were carefully
suited to this goal. His condemnationof pride, vanity, and avariciousnessis not
simply reflective of basic values - Anglo-Americansare commonly stereotyped
as proud, vain, and avaricious. Moreover, these characteristicsare believed to
engender in Anglo-Americansa sense of superiorityto Mexicanos and a reluctance to develop close friendshipswith them. Mr. Trujillogoes on to arguethat
friendship (amistad) conquers these obstacles. He indexes the fact that his remarks are directly applicable to Anglo-Americans (presentcompany included)
by inserting three English lexical items into the discourse. Thirty-five seconds
after this segment, Mr. Trujillo argues that the obligation to extend fraternal/sororallove to everyone admits no racialdistinctions:No importaquien sea,
no importaque sea negro o que sea lo quefuere o austriago o lo quefuere 'it
doesn't matterwho it is, it doesn't matter if s/he's black or s/he's whateverit
may be or a foreigner or whatever it may be.'
Friendshipdid come to reign in the situation.The Trujillosbecame two of my
best friends, central consultants, and major sponsors within the community.
They shifted to tti forms before long, and we adoptedkin termsfor use in address
a year later (I addressed them as papd 'father' and mamd 'mother', and they
addressed me as hijo 'son').
Scripturalallusions thus provide metacommunicativecommentaryon the ongoing conversation in two primaryways. First, scripturalallusions drawupon a
textual traditionwhich is sharedby the speaker, her/his interlocutor(s),and their
audience. Such phrases as Dios dice 'God says' or dice mi Sefior Jesucristo en
sus evangelios 'my Lord Jesus Christ says in His gospels' tell the hearer "these
are not my words, these are the words of Jesus Christ." Use of a scriptural
allusion thus places the highest source of legitimacy in the society behind the
speaker's words. Since the scripturaltext is irrefutable, the legitimacy of the
speaker's point of view can only be disputed by challenging the bearing of the
text on the issue at hand. This is no simple task, and such challenges rarelyoccur
outside of rhetoricalduels between elders.
The rhetoricalforce of scripturalallusions cannot be attributedentirely, however, to the connection of the discourse of this shared textual background.The
effective use of a scripturalallusion uses this source of legitimacy in advancing
the speaker's view of an issue of presentconcern. Text 3 adumbratesthe values
which Mr. Trujillodeems central. It also affords us insight into his perceptionof
Mexicano-Anglo-American relations, our friendship, and of what was taking
place on an interactionallevel that afternoon. The scripturalallusion thus functions as a blueprintor, in Peircean terms, a diagramof the speaker's view of a
particularsituation. This diagrammaticfunction is two-sided, so to speak, since
it also includes an iconic condensationof one aspect of 'the talk of the elders of
bygone days.'
In anthropologicalterms, scripturalallusions connect an ideal model with a
real, temporallybounded event, specifying how the formerapplies to the latter.
Such speech events thus show great promise for ethnographers.Some fieldworkerscollect extensive dataon structuralprinciplesor decontextualizedideals.
Others are less concerned with norms and meanings than with the details of
"objectively" observable behavior. The ideal and the real are, however, never
equivalent, and ethnographersencounter difficulties in relating the two. The
study of these sorts of speech events thus provides the ethnographerwith a large
set of examples of how the people themselves envision the relationshipbetween
culturalnorms and concrete situations.
Summary:The evaluation of metacommunicativecompetence
Taken as a whole, these materialspoint to the centralityof threetypes of tasks to
the developmentalprocess. Lacking basic linguistic competence or commandof
a given type of speech event, individuals are expected to learn initially through
observation. Once the learnerhas gained some grasp of the requisiteskills, this
mechanismis complementedby a second type of acquisitiondevice - repetition
of the words of one's seniors. Once this reiterativecapacity has been acquired,
the focus turns to practicing the production of original utterancesin dialogue.
These skills roughlycorrelatewith basic stages of the acquisitionprocess, with
childrenobserving and reiteratingthe wordsof theirseniors, adolescentsgenerating original utterances in interactions with peers, and adults generating increasingly sophisticatedspeech acts in public. Nevertheless, all threedevices are
utilized at each stage as well, albeit in differentsocial contexts. Young children,
for example, are not encouragedto offer originalcontributionsto adults' conversations. But they possess competence in and a rightto performverbalgames that
are hardly sharedby adults. Similarly, while most adultsof thirtyto sixty years
of age can speak freely on issues which pertainto wage labor employmentand
communityaffairs, they are still primarilyrestrictedto observationandrepetition
of the elders' presentationsof 'the talk of the elders of bygone days.'
This progressioncontinues throughthe lifespan, and a personnormallyreaches her/his point of maximumcompetence in these areasjust priorto joining 'the
elders of bygone days,' barring senility. Rhetorical facility and the right to
perform the most esteemed genres are not, however, simply guaranteedby
growing older. The degree of rhetoricalcompetencewhich is requisiteto discussing such "'advanced"topics as oral traditions,ethnotheology, and moralvalues
is thus gained in the course of a gradualand lengthy process of masteringsocial
and linguistic skills. The combined effects of talento, interes, concentracion,
andpaciencia renderit possible for some persons to gain rhetoricalcompetence
much earlier than others.
Two factors play a major role in assessing rhetoricalability. First, ceteris
paribus, the greater the numberof folkloric genres in which a person can perform, the greater her/his status as a rhetorician.The second factor entails the
development of a heightenedpragmaticsense. The best speakersselect an utterance type, such as a joke, a story, or 'just words,' which best fits the social
context of the conversationand the predominanttopic. Likewise, the utterance
does not falter (through slurring of words, a lapse in memory, etc.) and is
accompaniedwith appropriategesticulation, body movement, facial expressions,
and prosodic features. A good speaker will maintain the cohesion of the discourse, making her/his hearer(s)aware, whetherexplicitly or implicitly, of the
relevance of the utteranceto the present situation. Such individualsthus subtly
manipulate both linguistic forms and aspects of the social situation to lend
compelling force to their utterances.
Question-answer sequences play several distinctand quite importantpartsin the
acquisition process. Young children repeat adults' questions to other adults and
then convey the latter's answers to the former. Childrenare also accordedthe
rightto pose permissionrequeststo the appropriateadults, as in our society (e.g.,
"'mama,may I go to the store with Mary?"). This act presupposesa knowledge
of appropriateage and genderroles and of authorityrelations,since unreasonable
requests or queries directed to the wrong person would be ineffectual or even
At a much higher level in the rhetoricalhierarchy,individualsare permittedto
query their seniors with regard to "traditional" Mexicano knowledge - eth20
notheology, oral historical traditions, spoken and musical genres, and so fortn.
Posing questions in this way does not, however, even resemble an interview.
Persons who wish to acquire such knowledge frame a repetition of part of the
elder's preceding utterancewith an interrogativemarker(Briggs I983a). Common questions thus assume the form of Zque no dijo usted que . . . ? 'didn't you
say that . . .?' or ique quieredecirX? 'what does X mean?' It is importantto
note that I) the referentialframe for such questions is provided by the elder's
speech, 2) the latter's pedagogical discourse provides the dominant conversational structure for the interaction, and 3) the utterance of such a question
presupposes both basic knowledge of this type of information(gained through
observation) and the mannerin which it is disseminated.
Standardanthropologicalinterviewing technique inverts these conversational
norms in three ways. First, nonnative ethnographersenter the society lacking
acquaintance with norms for comportment and speech. Instead of acquiring
communicativecompetence by ascending throughthe established succession of
developmental tasks, however, interviewers skip the stages of observationand
repetition of their seniors' words to move immediately to the generation of
original utterances- questions which emerge from their own interests. Furthermore, such questions frequently pertain to the most esoteric topics - those in
which only the most advanced speakers are competent.
Second, control over the interactionlies in the hands of the interviewer.It is
s/he who exerts the most control over the process of turn-taking,it is s/he who
introducestopics, by and large, and it is s/he who decides when to move onto the
next topic. This constitutes an inversion of the normativestructurefor the conveyance of informationon such topics between a senior and a junior.
Finally, the interviewer'sinitial lack of familiaritywith the relevantreferential
frames and accepted mode of dissemination for the informationfrequentlyrenders her/his questions disruptiveto the cohesion of the discourseor inappropriate
for the social situation. These problems are compoundedwhen the interviewer
attemptsto shift the discourse to a subjectof greaterrelevanceto her/his research
interests or moves from topic to topic. Problems emanating from insufficient
competence in the language or dialect or in basic culturalpremises also undermine the effectiveness of the interviewer's speech. As I argued in the paperon
interviewing(I983a), questions which contain such stumbling blocks are difficult to answer. The repeated emergence of proceduralproblems, to use Churchill's (1978) term, prompts the consultant to reassess the former's ability to
effectively engage in this level of discourse.
The pragmatic effectiveness of a person's speech - providing the proper
referentialframe, avoiding significant formal or semantic flaws, making one's
remarksappropriateto the social situation, and maintainingthe cohesion of the
discourse - is one of the two primarymeans of judging rhetoricalcompetence.
The emergence of such proceduralproblems thus suggests to the hearersthat
their interlocutorlacks sufficient competence to enable the latter to generate
original utterances.One's consultantsare accordinglynot obligatedto repairthe
proceduralproblem, answer the question, and give the turn back to the questioner, as is prescribed by Anglo-American, middle-class speech norms (cf.
Churchill 1978:89). It is indeed far more appropriatein the case of the fieldworker to signal her/him of the failure of her/his efforts to assume a sophisticated communicative role.
Inverting norms for the acquisition of metacommunicativecompetence, nevertheless, is par for the course. Although some fieldworkers, such as Geertz
(1973), provide notable exceptions, most ethnographersseek to impose their
own metacommunicativenorns on theirconsultants.Ratherthanadoptingnative
metacommunicativenorms, consultantsare taught a subset of the fieldworker's
own metacommunicativedevices - those pertainingto interviewtechnique.This
is the practice that I have referredto as communicativehegemony (i983a).
The legitimacy of this approachis enhancedby its enshrinementin manualson
ethnographicfieldwork. One of the most commonly used discussions of anthropologicalfieldwork, Pelto and Pelto's Anthropologicalresearch: The structureof inquiry,suggests that ". . . humansdiffer in their willingness as well as
their capabilitiesfor verballyexpressing culturalinformation.Consequently,the
anthropologist usually finds that only a small number of individuals in any
communityare good key informants"(1978:72). Note, of course, thatthe basis
of the selection is the communicativenorms of the fleldworker,not those of the
natives. Then, having selected the "right" person(s), "some of the capabilities
of key informantsare systematicallydeveloped by the fieldworkers,as they train
the informantsto conceptualizeculturaldata in the frameof referenceemployed
by the anthropologists"(ibid.).4 We are thus left with two fundamentaland, it
seems, commonly accepted assumptions regardingthe epistemology of fieldwork. First, obtaining large quantitiesof ethnographicdata entails (or is at least
aided by) the trainingof one or several natives in the metacommunicativenorms
presupposedby the interview. Second, this process is facilitatedby the selection
of "key informants" from among the ranks of those who appearto have the
greatest facility for operating within this mode of discourse.
I found the second premise to be quite accuratein my own fieldwork. George
and SilvianitaLopez served as my primaryconsultantswhen I beganmy research
in Cordova.They had also assumedquasi-parentalroles, being recognizedby the
communityas my adopted family. After a numberof preliminaryvisits in 1971
and I972, I proposed writing a book on the local wood carving industrywhich
would focus on the couple (cf. Briggs 1980). They respondedquite favorably. I
inauguratedthe fieldwork with the traditionalopen-endedinterview,concentrating on Mr. L6pez's fatherand on Silvianita and George Lopez's participationin
the industry.Mrs. Lopez respondedto nearlyall of my questionswith iooo, pos,
quien sabe! 'ooo, well, who knows!' Mr. L6pez alternatedbetween this reply
and the provision of highly abbreviatedresponses. The interviews got nowhere
for over two weeks.
On the otherhand, I proposeda set of interviewsto FedericoCordova,another
communityelder. He agreed both to the interviews and to their tape recording.
When I returnedfrom my car, Mr. Cordovaasked me 'Now, what is it that you
wanted to know?' I provided him with one of the questions that had fared so
poorly with the Lopezes. He then proceededto producea long, flowing narrative
history of the local carving industry.
How are we to account for such differences? Some researchersmight have
worked almost entirely with the more "cooperative" of the three. But the ability
and desire to communicate were not the problems here. After the Lopezes restructuredour interactionsby inducing me to carve wood along with them, I
found that simply sitting in the couple's kitchen, carving as they carved, produced a fair amountof exegesis aboutthe art. I laterdiscoveredthat I could elicit
additionalinformationthroughthe use of a repetition-elicitation formula - repeating one of their statements with a rising intonationor with a pre- or postpositioned tag question. The conversationswhich were structuredby the couple
turnedout to be extremely fertile sources of data, since the form of the discourse
provided crucial information on how the carvers perceived the history of the
industryand their own participationin it.
I ratherthink that an awarenessof the vastly differenteducationaland linguistic experiences among the three can bring us closer to understandingthis situation. Mr. C6rdova was fluent in English, and he was literatein both languages,
while the L6pezes are virtuallymonolingual. Mr. Cordovagained his first extensive contact with English speakersin WorldWar1, and he masteredthe language
while attendingteacher-trainingcourses at the regional normal school. He then
taught for many years in the local school. He later worked for the U.S. Forest
Service, a job which involved interactingwith English speakers, asking questions, and completing writtenreports. Mr. Cordovaoften readEnglish-language
newspapers. Interestingly,Mr. C6rdova's wife, who is monolingual, responded
to my questions in the same fashion as the Lopezes.
Mr. Cordova had acquirednot only the phonological, syntactic, and semantic
systems of AmericanEnglish, but he had masteredits conversationalstructureas
well. He had such a sophisticatedidea of the interviewingprocess that he even
thoughtto make sure that the tape recorderhad been turnedon before beginning
his account. (The one stipulationthat he imposed in helping me was that I was to
give him a copy of the final publication.) Why, then, was the latterinterviewso
painless and so "successful"? Even though Mr. Cordovaand I spoke in Spanish,
the interview was bilingual, since the frame of referenceand the conversational
structurethat we used were derived mainly from AmericanEnglish. Fortunately,
this is not true of the sections of the interactionsin which Mr. C6rdova "wandered off the point," that is, gained control of topical selection. Here the tape
recordingsreveal the same richness of metacommunicativeroutinesand rhetorical structurethat characterizethe pedagogical dialogues with the L6pezes.
My research methods thus dictated the imposition of my own conversational
norms on my consultants. The gap between American English discourse structure (especially interview techniques)and that of New Mexican Spanishis sufficiently wide that my initial position of communicativehegemony was successful
only with a bilingual consultant.
Having outlined some of the problems which are posed by the use of interviews
in fieldwork in a previous paper (i983a), I proceed to analyze native metacommunicativeroutines here. I argue for the richness of such routinesas sources of
ethnographicand sociolinguistic data. A sketch of the mannerin which metacommunicativecompetence is acquired in one society suggests that the use of
interviews, particularlyearly in the fieldworkperiod, may be highly disruptive.I
argue that a better proceduremight be to learn the native metacommunicative
repertoire. Interviews can then be utilized where they prove to be compatible
with these patterns.
The limitationson the usefulness of interviewingemanate from the natureof
the interview qua speech event. Shifting into this mode of interactionsends two
crucial signals to the participants.First, it selects for certaintypes of messages.
These are, in Silverstein's (1979, 198I) terms, surface segmentable,referential,
and relatively presupposing.The meaningof surface segmentableforms accrues
to units such as affixes, lexemes, and phraseswhich are susceptibleto traditional
syntactic and semantic analysis. A focus on referentialityentails a concern with
the generationof propositionsabout events which exist outside the actual utterance itself (althoughthe referentsmay be speech acts, too). The meaningfulness
of relatively presuppositional(as opposed to creative) statementsrests on the
prior existence of specific aspects of the present context. Use of the token
"she," for example, presupposes a shared understandingas to the fact that a
nonpresentfemale is the subjectof discourse. Interviewsthus rest on the assumption that the respondentwill supply "answers" which are segmentable,referentially rich, and for which the necessary presuppositionalinformationhas been
provided. The interview frame also alerts the participantsthat messages will be
decoded (and should be decodable) via the contributionof segmentableunits to
propositionsunder specifiable (presuppositional)conditions.
Second, interviews negatively select for metacommunicativeevents which are
less surface segmentable, more creative, and whose meaning hinges less on
reference. This description closely fits nearly all of the Mexicano metacommunicativeroutinesthat I have presentedin this paper. Stylistic features,such as
tonality, pitch register, stress, phrasing, laryngealconstriction,Spanish/English
code-switching, and so on, convey crucial information regarding meanings
which are not decodable vis-a-vis business-as-usualreferentiality.The less segmentableand referentialmessages provide the juxtapositionof clarity with indirectness that is the sine qua non of more advanced rhetorical strategies. The
emphasis on creativity is obvious; speakers use their words to transformthe
parametersof the interactionitself.
The complementarydistributionof interview situationsand a broaderrangeof
metacommunicativeevents should come as no surprise.Silverstein (1979, I98I)
has suggested that segmentable, referential, and presuppositionalelements are
the type of signals that universally lie within the conscious control of speakers.
Interview questions prompt natives to constructexplicit, conscious statements.
They accordingly provide a useful means of discovering the range of social/culturalknowledge and metacommunicativeforms which lie within the conscious control of speakers and which jibe with the sociolinguistic norms of the
interview situation. The Mexicano data suggest, however, that a vast arrayof
metacommunicativeroutines lie outside of these narrowlimits. The exclusion of
native metacommunicativeevents by fieldworkersis unfortunate,because they
provide crucial informationon the interactionbetween language use and social
behavior as a whole and between social/culturalnorms and observed patternsof
These materialsthus point to the critical need to achieve a deeper understanding of the interview process. It is perhapsthe most central mode of data collection in the social sciences and linguistics, but it is one of the least understood.
Only occasionally have interviews formed the focus of theoretical analyses or
ethnographicdescriptions. Scholarshave begun to compile descriptionsof native
metacommunicativeroutines and to demonstratetheir usefulness as sources of
sociolinguistic and ethnographicdata. Unfortunately,few practitionershave related specific routines to the nature of metacommunicativecompetence as a
whole, and theoretical and methodological studies are badly needed. Grasping
the natureof interviews and their relation to other types of metacommunicative
events thus forms an importantitem on our collective researchagenda.
Financialsupportfor the various fieldwork periods was providedby the IntemationalFolk Art
Foundation, the National Institute of Mental Health, the National Science Foundation, and the
Research Committee of Vassar College. Dell Hymes kindly invited me to write the paper;James
Femandezencouragedmy pursuitof this line of research. Ben Blount providedvaluablesuggestions
on the section on the use of repetitionelicitation routineswith older children, and JulianJosue Vigil
submittedthe paperto a meticulous reading. The threeelderly C6rdovancouples who are mentioned
in this paper have not been pseudonymed, as is the case with the other consultants. The C6rdovas,
L6pezes, and Trujillos have requested that their names be used in characterizingtheir pedagogical
discourse. Given the many kindnesses they have extended to me, I am happy to do so.
Interestingly, C6rdovansreduce 161to IgI. This featureenables Mexicanos from surrounding
communities to identify a speaker as a resident of C6rdova.
The discourse transcribedas texts 2 and 3 is formally elaborated.Such featuresas tone groups,
pauses, and stress parallel the semantic and pragmatic content of the words, creating carefully
constructedrhetoricalstructures. In keeping with the ethnopoetic principles elucidated by Hymes
(1981), McLendon(I982), and Tedlock (1971, 1972), I have used these formalfeaturesin attempting to preserve a sense of the rhetoricalstructurein the transcriptionsand translations.
It should be noted that Pelto and Pelto do express some reservationswith regardto the use of
key informantsand do allude to the fact that ". . . the interactionbetween fieldworkerand informants is a complex social process" (1978:74). Nevertheless, like nearly all writers who offer
critiquesof interview techniques, they simply point to the sources of potentialbias ratherthanurging
fieldworkersto be constantly aware of their patternsof interactionwith their consultants,especially
in comparison with native norms of interaction. (See Karp and Kendall [1982] for a penetrating
critique of the book.)
Bates, E. (1976). Language and context: The acquisitionofpragmatics. New York:AcademicPress.
I Camaioni, L., & Volterra,V. (1979). The acquisitionof performativespriorto speech. In
E. Ochs & B. B. Schieffelin (eds.), Developmental pragmatics. New York: Academic Press.
Bloch, M. (I975). Introduction.In M. Bloch (ed.), Political language and oratory in traditional
society. London: Academic Press, 1-28.
Briggs, C. L. (1980). The wood carvers of C6rdova. New Mexico: Social dimensionsof an artistic
'revival." Knoxville: University of Tennessee Press.
(i983a). Questions for the ethnographer:A criticalexaminationof the role of the interview
in fieldwork. Semiotica.
(I983b). Treasuretales and pedagogical discourse in Hispanic New Mexico. Unpublished
(in press). The pragmaticsof proverbperformancesin New Mexican Spanish. American
Churchill, L. (0978). Questioning strategies in sociolinguistics. Rowley, Mass.: Newbury House.
Cicourel, A. V. (1970). The acquisition of social structure:Towards a developmentaltheory of
language and meaning. In J. Douglas (ed.), Understandingeveryday life. Chicago: Aldine.
Some basic theoretical issues in the assessment of the child's performancein
testing and classroom settings. In A. V. Cicourel et al., Language use and school performance.
New York: Academic Press.
Cook-Gumperz,J., & Corsaro, W. A. (1976). Social-ecological constraintson children's communicative strategies. In J. Cook-Gumperz& J. J. Gumperz, Papers on language and context.
Working Paper #46, Language Behavior Research Laboratory, University of Califomia,
Cook-Gumperz,J., & Gumperz,J. J. ( 976). Papers on language and context. WorkingPaper#46,
Language Behavior Research Laboratory,University of Califomia, Berkeley.
Corsaro, W. A. (1979). Sociolinguistic patterns in adult-child interaction. In E. Ochs & B. B.
Schieffelin (eds.), Developmentalpragmatics. New York: Academic Press. 373-89.
Eisenberg, A. R. (1982). Language acquisition in cultural perspective: Talk in three Mexicano
homes. Dissertation, University of Califomia, Berkeley.
Ervin-Tripp,S. (1972). On sociolinguistic rules: Alternationand co-occurrence.In 1. J. Gumperz&
D. Hymes (eds.), Directions in sociolinguistics: The ethnographyof communication.New York:
Holt, Rinehart, & Winston. 213-50.
(1979). Children'sverbal tum-taking.In E. Ochs & B. B. Schieffelin (eds.), Developmental pragmatics. New York: Academic Press. 391-414.
Espinosa, A. M. (1911 ). The Spanish language in New Mexico and southernColorado. Historical
Society of New Mexico, PublicationNo. i6. Santa Fe.
Estudios sobre el espaiiol de Nuevo Mejico. 2 vols. Translatedand edited by A.
Alonso & A. Rosenblat. Buenos Aires: Imprentade ia Universidad.
Geertz, C. (1973). Deep play: Notes on the Balinese cockfight. In The interpretationof cultures.
New York: Basic Books.
Goffman, E. (1959). The presentation of self in evervday life. New York: Anchor.
Grimshaw,A. D. (I969). Languageas data and as obstacle in sociological research.Items 23:I7-21.
Some problematic aspects of communication in crossracial research in the
United States. Sociological Focus 3(2):67-85.
Hotchkiss, J. C. (I967). Childrenand conduct in a Ladino communityof Chiapas, Mexico. American Anthropologist69(4):71 i - i 8.
Hymes, D. H. (1970). Sociolinguistics and the ethnographyof speaking. In E. Ardener(ed.), Social
anthropology and language. (ASA Monographs io.) London: Tavistock. 47-93. Reprintedand
revised in B. G. Blount (ed.), Language, culture, and society: A book of readings. Cambridge,
Mass.: Winthrop, I974. I89-223.
Models of the interactionof language and social life. In J. J. Gumperz& D. H.
Hymes (eds.), Directions in sociolinguistics: The ethnography of communication. New York:
Holt, Rinehart, & Winston. 35-71.
(1 98 I). "In vain I tried to tell you": Essays in Native Americanethnopoetics.Philadelphia:
University of Pennsylvania Press.
Karp, I., & Kendall, M. B. (I982). Reflexivity in field work. In P. F. Secord (ed.), Explaining
social behavior: Consciousness, humanaction, and social structure. Beverly Hills, Calif.: Sage.
Keenan, E. (1973). A sliding sense of obligatoriness:The polystructureof Malagasy oratory.Language in Society. 2:225-43.
McLendon, S. (1982). Meaning, rhetorical structure, and discourse organization in myth. In D.
Tannen (ed.), Analyzing discourse: Text and talk. Georgetown University Round Table on Languages and Linguistics, 1981. Washington:Georgetown University Press. 284-305.
Menyuk, P. (1977). Language and maturation. Cambridge, Mass.: MIT Press.
Ochs, E. (1979). Introduction:Whatchild languagecan contributeto pragmatics.In E. Ochs & B. B.
Schieffelin (eds.), Developmentalpragmatics. New York: Academic Press. 1-17.
Talking to children in Western Samoa. Language in Society I (0:77-I04.
Peirce, C. S. (1932). Collected papers of Charles Sanders Peirce. Vol. II, Elements of logic, C.
Hartshorne& P. Weiss (eds.). Cambridge, Mass.: HarvardUniversity Press.
Pelto, P. J., & Pelto, G. H. (1978). Anthropological research: The structure of inquiry. 2nd ed.
CambridgeUniversity Press.
Rael, JuanB. (1937). A study of the phonology and morphologyof New Mexican Spanish, based on
a collection of 4I0 folktales. Dissertation, StanfordUniversity.
Schieffelin, B. B. (I979). Getting it together:An ethnographicapproachto the study of the development of communicative competence. In E. Ochs & B. B. Schieffelin (eds.), Developmental
pragmatics. New York: Academic Press. 73- io8.
Shotter, H. (1979). The culturalcontext of communicationstudies:Theoreticaland methodological
issues. In A. Lock (ed.), Action, gesture and symbol: The emergence of language. New York:
Academic Press. 43-79.
Silverstein, M. (I976). Shifters, linguistic categories, and culturaldescription. In K. Basso & H. A.
Selby (eds.), Meaning in anthropology. Albuquerque:University of New Mexico Press. II-55.
(1979). Language structureand linguistic ideology. In P. R. Clyne, W. Hanks, & C. L.
Hofbauer (eds.), The elements: A parasession on linguistic units and levels. Chicago: Chicago
Linguistic Society. 193-247.
The limits of awareness. Sociolinguistic WorkingPaper84. Austin, Texas: South(I981).
west EducationalDevelopment Laboratory.
Snow, C. (1977). Mother's speech research: An overview. In C. Snow & C. Ferguson (eds.),
Talking to children: Language input and acquisition. CambridgeUniversity Press.
Strauss, A., & Schatzman, L. (0955). Crossclass interviewing: An analysis of interaction and
communicationstyles. Human organization 14(2):28-31.
Tedlock, D. (I971). On the translationof style in oral narrative.Journal of AmericanFolklore 84
("Toward new perspectives in folklore," A. Paredes & R. Bauman, eds.): 14-33.
Finding the center: Narrative poetry of the Zuni Indians. Lincoln: University of
Nebraska Press.
Trevarthen,C. (1979). Communicationand cooperation in early infancy: A descriptionof primary
intersubjectivity.In M. Bullowa (ed.), Before speech. CambridgeUniversity Press.
Trotter, R. T., II, & Chavira, J. A. (198 1). Curanderismo:Mexican Americanfolk healing. Athens:
University of Georgia Press.
Whorf, B. L. (1956). Language, thought, and reality: Selected writingsof BenjaminLee Whorf.J.
B. Carroll, ed. Cambridge, Mass.: MIT Press.
Wolfson, N. (I976). Speech events and naturalspeech: Some implicationsfor sociolinguistic methodology. Language in Society 5:i89-209.