Monolingual fieldwork in and beyond the classroom: the Logooli
experience at UCLA
Betty Mack Twarog, the biologist who discovered serotonin in 1951 as a Harvard
graduate student of John Welsh, later wrote:
“To have worked with Professor John Henry Walsh is to have been apprenticed
to a Prince of Serendip, on a journey of scientific exploration. In science,
‘fortuitous’ discovery depends heavily on the traveler’s choice of itinerary.
Professor Welsh designed wonderful itineraries for his students and encouraged
enthusiasm, eclectic interests, and openmindedness. Many of us have enjoyed the
excitement of discovery because of his guidance” (Twarog 1988: 21).
Serendipitous discoveries are a hallmark of linguistic fieldwork, especially
when it is conducted primarily in the target language. The linguist who listens and
converses in the target language is privy to many unanticipated exchanges. Despite
this, the primary mode of data collection in graduate linguistic field methods
courses is usually translated elicitation. Few field methods instructors train students
in the acquisition of competence in the target language. This means that even if a
novice fieldworker believes in the value of target language competence, s/he likely
has had little guidance in techniques for acquiring this, nor practice in monolingual
It is a pity that most field methods instructors cling to translation from English:
monolingual elicitation is an enjoyable, memorable, and efficient framework for a
graduate field methods course, especially in the early stages. The monolingual
approach gives students a much more realistic introduction to the challenges,
frustrations, and joys of linguistic fieldwork than does a controlled contact language
elicitation model. The first quarter of the 2014-15 UCLA field methods course was
taught using monolingual methods. This paper uses two transcripts from sessions
from this course to show the potential of monolingual methods to gather large
amounts of target language data quickly.
Many thanks to the UCLA 2014-15 Field Methods students and Logooli consultant, Mwabeni
Indire, as well as to Margit Bowler and John Gluckman for connecting us with Mr. Indire. The
opinions expressed here are mine. I am grateful to all of the communities with whom I have been
privileged to work, and to my fieldwork mentors Alexandra Y. Aikhenvald, R. M. W. Dixon, G.
Tucker Childs, and Deborah D. Foster. Thanks also to CLS 51 audience members for questions
Proceedings of CLS 51 (2015), 471-484
Chicago Linguistic Society 2016. All rights reserved. 471
2 Monolingual fieldwork overview
is taken here to denote linguistic fieldwork in which the
researcher communicates with consultant(s) in the target language. Of course, in
fieldwork situations where a contact language is available, this contact language
may be essential for ethical project setup, and on occasion throughout the field
period. As Alexandra Aikhenvald (p.c.) points out, there is no contradiction
inherent in a fieldworker’s both acquiring the target language and occasionally
employing the contact language. The notion of “pure” monolingual fieldwork—
where an available contact language is always eschewed, even in instances in which
it might be useful—is not advocated here.
For nearly a century, linguistic and anthropological fieldworkers have discussed
the utility of learning the target language (Boas 1910; Mead 1939; Lowie 1940;
papers in Newman & Ratliff 2001; Everett 2001; Aikhenvald 2007; Dixon 2007;
Moore 2009; Chelliah & De Reuse 2011; Mosel 2012, among others). The debate
centers around the time burden of language learning and the influence of
intermediary languages on elicited material.
In the first half of the twentieth century, anthropologists sparred over whether
or not language competence enhanced anthropological research. Boas (1911: 61)
wrote: “Much information can be gained by listening to conversations of the natives
and by taking part in their daily life, which, to the observer who has no command
of the language, will remain entirely inaccessible.” Mead (1939: 191) dissented,
claiming that a bit of linguistic knowledge was sufficient for many types of
research. She defended the “time-honored methods of interpreters and lingua
francas,” although still advocating the use of a texts corpus. For her, the acquisition
of language competence by the researcher was an act of “virtuosity” pursued apart
from actual research, akin to learning to paddle a canoe. Lowie (1940: 82)
responded to Mead’s article by challenging the amount of time that interpretation
and lingua francas had been “honored” for, listing earlier North American
ethnologists and linguists: “I believe J. O. Dorsey knew his Omaha, and that several
Russian scholars, as well as Castrén (about a century ago), used the linguistic
approach. In the United States, Frank Hamilton Cushing learnt Zuni during his five
years' stay (1879–1884); Alexander M. Stephen spoke Navaho by 1890 and was
learning Hopi before his death (1894).”
By the end of the twentieth century, Boas and Lowie’s approach had won out,
at least in theory, among anthropologists. Indeed, Moore (2009) and Newman
(2013) identified what Newman (2013) called “the great anthropological myth”:
the notion that anthropologists are all competent in the languages spoken by their
subjects. Among linguists, however, target language competence is often seen as a
special feat of the particularly-gifted, following Mead (1939).
Called “blind fieldwork” by Vaux & Cooper (1999: 10).
472 HAN NAH SA RVASY
2.1 The positives: language comprehension for participant-
A partial list of the benefits to a linguistic fieldworker of target language
competence follows as B1-10, given in no particular order.
B1. Minimize interference. Working solely through the target language
helps limit effects from a contact language.
B2. Guide research. Learning the target language can help the researcher to
build pseudo-intuitions about the language that may indicate profitable
directions for investigation.
B3. Access the best speakers. If the researcher need not rely on interpreters,
s/he may have direct access to members of the community who do not know
the lingua franca, such as the elderly or women: these are likely the most
proficient and knowledgeable speakers of the target language.
B4. Obtain reliable negative proofs. Others’ corrections of the researcher’s
mistakes are negative proofs of language structure in real-world contexts.
Davis, Gillon, & Matthewson (2014) and many others have discussed how
essential negative evidence is to establishing a grammatical analysis. One
excellent way to obtain negative evidence is to produce an utterance
yourself in the course of elicitation, which the consultant corrects
automatically, as if correcting a child. This depends, of course, on the
consultant’s willingness to correct the researcher, which must be cultivated.
B5. Please the community. For the (probable) majority of community
members who conceive of the linguistic research project as ‘learning
(about) our language,’ the researcher’s progress is gauged in terms of
language fluency. If a researcher claims to be investigating language, yet
makes no progress in using it, some might interpret the researcher’s claims
as a cover for some other type of investigation.
B6. Endear the researcher to the community. When a prestigious outsider
partially drops his/her expert status to become a student of the community’s
language—inevitably making embarrassing mistakes in it—this humbles
the researcher and shows his/her commitment to the community.
B7. Translate texts and examples more precisely. Eventually, language
competence allows for maximally precise translation of texts and examples.
Note that this discussion takes for granted that the local community sees no political, spiritual or
other problem with the researcher’s acquisition of the target language. If the community were to
object to the researcher’s learning their language, this must be respected.
MON OLI NG UAL FI ELD WOR K IN A ND BE YO ND TH E CL ASS ROO M:T HE LOG OO LI E XPE RI ENC E AT UCLA 473
B8. Discover the unexpected through participant-observation. Target
language competence allows the researcher to notice grammatical structures
in casual speech outside the elicitation setting
B9. Leave No (Linguistic) Trace (to coopt the US Forest Service mantra).
The researcher should be acutely conscious of any potentially negative
impact of his/her conduct on a language’s social status. A researcher
working on an endangered language may claim to value this language but
converse only in a contact language. This may negatively impact the
endangered language’s potential for survival: even the linguist apparently
avoids using it! Such a researcher could be leaving a heavy “trace” in the
community, cementing the language’s low status. If the researcher instead
honors the language by favoring it over the contact language in
conversation, this may help elevate the language’s status.
B10. Maximize efficient data collection. If the researcher converses
primarily in the target language, this means that all interactions become
potential data sources, or at least potential catalysts for novel elicitation
2.2 What opponents say
Some linguistic fieldworkers who did not acquire competence in their target
language claim, along with Mead (1939), that such competence is unnecessary for
describing a language’s grammar. Indeed, a sizable corpus of transcribed natural
speech and a decent passive understanding of the language’s structure may suffice
for a basic grammatical description. These fieldworkers may also add that acquiring
competence would have taken too much time. But it is noteworthy that of
fieldworkers who learned to speak the target language, no one seems to declare: “I
wasted too much time learning the language.” Potential drawbacks to target
language acquisition are listed below as D1-4.
D1. It (may) take more time. It has been claimed that acquiring target
language competence makes the field research last at least an extra six
months (Everett 2001). This claim must be examined against the increased
depth and other positive aspects to language acquisition in §2.1. Further, the
quality of some of the UCLA graduate student grammatical sketches
produced after an academic quarter of part-time monolingual elicitation
may prove that monolingual fieldwork does not necessarily entail longer
D2. Training is lacking. A willing fieldworker may have little training in
language acquisition. Moore (2009) and Mosel (2012) note that very few
guides to fieldwork, and few field methods classes, teach in any detail how
to learn the target language. Thus, the novice fieldworker who wishes to
eventually work monolingually may not have any guidance in this. Even
474 HAN NAH SA RVASY
veteran fieldworkers who themselves prefer monolingual elicitation may
not train students to do this in field methods courses (Dixon 2007).
D3. Pressure to learn quickly. Not only are the stakes high for the researcher
to prove s/he is making progress in learning the language, but the
researcher’s language mistakes may also be leveled at his/her language
teachers (as in McLaughlin & Seydou Sall 2001). In 2012, in the midst of
my dissertation research on the Papuan language Nungon, I gave what I
thought was a rousing speech in Nungon before a primary school assembly.
I meant to say that in the USA, if students do not do homework, they are
punished, but did not yet understand that the verb ‘punish’ has suppletive
forms depending on the object argument person/number. Instead of
nisopnangkang ‘they will punish us,’ I said niinangkang ‘they will bite us.’
This provoked spontaneous roars of laughter from the assembled students.
When I returned to the hearth of my adopted clan in the evening,
conversation around the fire was hushed and grave. My public mistake had
been leveled against my adopted family: they had been accused of failing
to teach me well.
D4. Negative influence on recording content. Consultants may simplify
their speech or over-explain during recording sessions if they are directing
their speech at the researcher and perceive him/her to have sub-optimal
language comprehension. This danger may be minimized by using native
speakers as interviewers in recording sessions.
For some, the notion of seriously trying to learn to speak the target language
may clash with an idea of “scientific research” in which the expert researcher goes
to the field with a pre-established hypothesis, not to pursue side serendipities (Davis
et al. 2014). In fact, influential scientific advancements have come about through
unexpected discoveries, as with Twarog’s Prince of Serendip (1988). Although
Charles Darwin touted hypothesis-driven research (Darwin & Seward 1903: 195),
he was of course greatly influenced by field observations.
2.3 What no one discusses: efficiency!
Benefit B10 of §2.1, “Maximize efficient data collection,” is a major boon of target
language competence, but it has been discussed little.
Elicitation through a contact language means that both consultant and elicitor
are conscious of building their working relationship primarily through the contact
language, and only secondarily around the target language. If either elicitor or
consultant is not a near-native speaker of the contact language, that person may feel
compelled to prove competence in the contact language through expert
commentary, witty parentheticals, jokes, or other meta-discussion. All this may be
welcomed by the other party to lessen the intellectual strain of grappling with the
less-familiar target language, but it takes up valuable time in elicitation sessions.
MON OLI NG UAL FI ELD WOR K IN A ND BE YO ND TH E CL ASS ROO M:T HE LOG OO LI E XPE RI ENC E AT UCLA 475
Further, it may be difficult to ask the other person to curtail such commentary in a
respectful manner. The need to build a relationship through the contact language
can thus mean much less actual target language data, and much more elicitation
session time taken up with social niceties, jokes, or extraneous information.
these occurred in the target language, of course, this would be valuable additional
data, possibly yielding serendipitous discoveries!
Rebecca Paterson (p.c.) normally elicits through contact languages (English and
Hausa) at her field site in Nigeria. When checking previously-obtained data with a
new consultant on a recent field trip, however, she worked only in the target Kainji
̱t-Ma’in. She reports that this approach made for more streamlined data
collection and much greater efficiency.
Monolingual elicitation is especially suited for the beginning stages of linguistic
fieldwork on an undescribed language, when the goals are amassing a sizable
vocabulary and filling in morphological paradigms, and there is little need to set up
complicated scenarios involving nuanced modalities. These beginning stages of
fieldwork are akin to the situation simulated by linguistic field methods courses.
3 Field methods courses in linguistics
Field methods courses are included in the curricula of many graduate and some
undergraduate linguistics programs. Although most graduate students in the U.S.
will never pursue fieldwork in a faraway place, all linguists have to interact
throughout their careers with language descriptions that stem from fieldwork. A
field methods course should, at least, give students a window into how the data for
these descriptions are gathered and analyzed. Without such experience, a linguist
has no basis on which to evaluate such descriptions as reliable. This is like
evaluating a statistical claim with no understanding of statistical methods.
Newman (2009: 124) concurs with these generalizations about the import of
these courses even for linguists who never plan to do fieldwork: “One has to
approach each description with a certain degree of skepticism. Theoreticians who
come up with sophisticated models and explanations naively depending on data
from one description by one person of one language (about which they know
nothing) do so at their own peril.”
For many, their field methods course is also the one chance they get to
investigate a new language as a complete system, grappling with its grammar as an
integrated whole. A field methods course generates the ultimate messy, real-world
massive problem set for students to work out as best they can. In this problem set,
phonology is not clearly separated from morphology, and syntax is tied to
pragmatics. It should be a challenging synthesis of everything students have learned
about language thus far.
Once the researcher and consultant have begun working through a contact language, it may be
difficult to later transition to only speaking the target language together.
476 HAN NAH SA RVASY
Finally, a field methods course should give students a taste of serendipity. It
should give them glimpses of intriguing facets of language they could not have
anticipated finding. At first, students in Martin Walkow’s 2014 undergraduate field
methods class could only imagine exploring components of grammar that they
knew from general theoretical literature: “Three out of four students who wanted
to work on syntax, wanted to work on wh-questions. The fourth wanted to work on
quantification” (Martin Walkow, p.c.). After exposure to idiosyncracies of the
target language, Shanghainese, however, none of the students ended up focusing on
these areas, instead honing in on new discoveries they had “just stumbled into.”
Newman (2009) reports an early-1990s informal survey of US and Canadian
graduate linguistics programs; of 42 that responded, 34 offered a field methods
course. In 5 the field methods course was required for the MA degree, and in 14 it
was required for the PhD. In many of the departments, field methods courses were
not offered every year. The survey did not probe teaching methods for the field
methods courses. Anecdotal evidence today suggests that few, if any, university
graduate field methods courses use monolingual methods for any sizeable portion
of the course.
Why not use monolingual methods? The major drawback to monolingual
methodology for fieldwork in general is the investment of time and energy in
learning the language. Faculty members undertaking fieldwork may have
high-stakes, time-sensitive deliverables that necessitate a rush job using contact
language translation. This is not the case with field methods courses: the stakes are
low, and the goals are learning-related, not publications. A field methods course
would seem to be the ideal venue for students to try monolingual elicitation
4 The 2014-2015 Logooli field methods course outline
All UCLA PhD students in linguistics must take the graduate field methods course.
As with many US graduate programs in linguistics, the UCLA field methods course
is two academic quarters long. I taught the 2014-15 UCLA field methods sequence
working with Logooli (Eastern Bantu) consultant Mwabeni Indire. All class
sessions in the Fall quarter were monolingual, and this quarter culminated in each
student’s production of a grammatical sketch of Logooli. Following the UCLA
tradition established by Pam Munro, Bantu and Logooli-related literature was off-
limits for the first five weeks of the Fall quarter, so students wrote mid-term draft
grammatical sketches without reference to outside literature. In the Winter quarter,
students were free to use English in class sessions, and each pursued a specialized
A class database in FLEx included class session transcripts, as well as
transcribed narratives from Mr. Indire, interview clips from the film Maragoli
(Nichols & Ssenyonga 1976), and transcribed narratives from Michael Diercks’s
Logooli texts corpus (Diercks 2014).
MON OLI NG UAL FI ELD WOR K IN A ND BE YO ND TH E CL ASS ROO M:T HE LOG OO LI E XPE RI ENC E AT UCLA 477
How can a classroom setting with a single speaker accommodate monolingual
field methods? My initial monolingual demonstration generated some lexical items,
which students then used to prepare potential Logooli sentences. Mr. Indire
corrected the students’ sentences in the next session and provided further
elaboration of them, generating more lexicon. Props, mime, drawings, etc. were all
used to elicit vocabulary and inflectional paradigms. Beginning very early, Mr.
Indire provided short Logooli narrative sequences that served as models for
students’ own statements in Logooli (see Moore 2009 on memorization of folklore
and other texts as a tool for acquiring target language competence).
Students were forced to pay close attention to all monolingual sessions. In the
UCLA program students are encouraged to choose programmatic affiliations
early—aligning themselves with the “P-side” (phonetics, phonology) or “S-side”
(syntax, semantics)—the monolingual approach and the demand on each of them
to produce a Logooli grammatical sketch meant that P-siders could not disregard
syntax, and S-siders had to grapple with phonetics and phonology.
A further demand on students—that they submit an abstract to a special session
of the Annual Conference on African Linguistics on the Luyia Bantu languages—
put pressure on them to quickly master Logooli grammar: a tall task. When this
abstract was due, at the end of the first quarter, some students balked at the
monolingual approach. But in the second quarter, these same students reported that,
having now worked using English, they understood the benefits of working only in
Logooli in the first quarter. They felt they grasped the overall structure of Logooli
better than they would have if they had used English translation from the beginning.
5 Analysis of sample monolingual and bilingual sessions
The following two sections introduce excerpts from a sample monolingual session
and a sample bilingual session. The initial 12:19 of a 1:40:00 monolingual session
was transcribed and analyzed, while the last 10:51 of a 42:00 bilingual session was
also transcribed and analyzed. The entire 12:19 and 10:51 transcripts are not given
here, but all material from these excerpts contributes to the counts in Table 1 in
The two sessions have necessarily different characters: the monolingual one
was part of the period of general information-gathering in the Fall quarter, while
the bilingual session represents one student’s investigation (in front of the class) of
a topic of special interest. This means that if the bilingual session sampled here had
been conducted monolingually, it would still have had a somewhat different
character than the monolingual session sampled here.
Regardless of this difference in content and style, the emphases here are on the
serendipitous discoveries of the monolingual session and the excessive
parenthetical statements and social niceties of the bilingual session. These do not
relate to the sessions’ content; rather, they relate to the choice of primary language
478 HAN NAH SA RVASY
In the monolingual session, the consultant performs self-corrections,
elaborations, explanations, and other meta-linguistic commentary in the target
language. Given the number of these explanations in the bilingual sample, an equal
or greater number of serendipitous discoveries could be expected from such a
session if conducted monolingually.
5.1 Monolingual session
The excerpt in (1) below represents a 1:28 segment of a class session in October
2014. The segment belongs to the 12:19 excerpt that was transcribed for analysis;
for length considerations, only this 1:28 excerpt is given here. Participants are: M
(the consultant), J, H, NA, NL, and S. The session began with two students, NA
and J, reading to Mr. Indire the self-introductions they have prepared in Logooli,
modeled on his own; this excerpt begins from the start of the second student’s self-
introduction. The students pause after each sentence for any corrections from Mr.
In this session, most interactions were in Logooli, with occasional English
asides whispered between students to clarify points among themselves, and an
English introduction to the session by the instructor.
(1) Monolingual session transcript excerpt
I was born
I walked [sic]
M: I walked,
H: I walked,
Not ‘to go’,
Not ‘to go’.
M: I walked
NA: [whispers to J]
J, to NA: I walk to school.
MON OLI NG UAL FI ELD WOR K IN A ND BE YO ND TH E CL ASS ROO M:T HE LOG OO LI E XPE RI ENC E AT UCLA 479
The excerpt in (1), although brief, includes several noteworthy discoveries not
likely to have emerged in routine translated elicitation. Some of these concern
discourse functions of known expressions, such as ndijɔ ‘like that, thus,’ which is
used here for encouragement. Pivotal here is Mr. Indire’s self-correction after he
supplies an infelicitously-worded sequence. He uses an expression pɔlɛ ‘sorry’ to
apologize for supplying misinformation, then negates the infelicitous term kuzja ‘to
go’ using a secondary negator, mba. Whenever Mr. Indire self-corrected in Logooli,
he used the negator mba rather than the primary negator he had taught the group,
daave. If the session here had been conducted through English, none of the three
expressions would have been used in these ways. Here, the richness and flexibility
of Logooli as a living language is brought to the fore.
5.2 Bilingual session
The transcript in (2) is a 4:30 excerpt from the 10:51 transcribed segment of the
bilingual class session on 27 January 2015. In this session, a single student, N,
interacted with the consultant, M, to explore the differences between three copular
forms. Other members of the class also asked questions at various points during the
session, but the excerpt here shows only N and M. In this excerpt, Mr. Indire’s
speech is in brackets throughout.
(2) Bilingual session transcript excerpt
N: It’s something like, the translation is, is that accurate to say that when you say
mwivi ne mwahi, [M: m-hm] it means ‘the thief is the doctor’ in English? [M: No.]
N: And in the other case— No? [M: Sorry, say it again.]
N: Mwivi ne mwahi is something like, ‘the thief is the doctor,’ talking about a
[M: Yes, you can get that too, uh-huh.]
N: Whereas when you say mwivi ave mwahi [M: M-hm] it’s more like, ‘the thief
is a doctor.’ [M: A doctor, exactly.]
N: But it’s a possible sentence. [M: That—] It’s correct— under this reading, is it
[M: The thief is a doctor, mwivi ave mwahi?]
N: Uh-huh. [M: Absolutely, yeah. It’s very correct.]
480 HAN NAH SA RVASY
[M: Now, for the other one, because you said it’s more specific, I should add, you
can say mwivi ne mwahi ora.]
N: M-hm. [M: The ora just makes it more specific, like it’s that doctor.] N: The
specific. [M: M-hm.]
[M: Oh no, no, mwivi ne mwahi, is that what I said, right?] N: M-hm.
[M: Yeah, okay. Cool. If I get confusing or something, just let me know.]
Like (1), this excerpt also shows self-correction and elaboration by the
consultant. Here, however, these reflections serve only to help clarify judgments
and reported connotations, and to grease the social wheels between consultant and
elicitor. Because English is used as a metalanguage, no unexpected discoveries
arise; rather, the elicitor is given nuanced English interpretations to weigh against
previous ideas of the differences between the copula forms.
The turns in (2) are longer than in (1). Further, both elicitor and consultant in
(2) use parenthetical statements such as it’s more like, the translation is, and I
should add. These arguably contribute nothing of linguistic interest and serve only
to take up valuable time. Likewise, the last statement by the consultant serves no
It appears that when English is used, elicitor and consultant are necessarily
conscious of their relationship-building in English, mediating statements with
niceties and parentheticals. Unfortunately, these niceties fill time in the session.
When the elicitor(s) and consultant use only Logooli, these niceties are set aside;
the elicitor(s) are exempt from using them, as language-learners, and the consultant
strives to teach the basics first. Thus, monolingual elicitation techniques spare
everyone the empty social graces and parenthetical comments of English or another
5.3 Comparison of the two sessions
The two transcribed session segments from which (1) and (2) were taken are
analyzed in Table 1. Number of turns, total words spoken, number of words spoken
in Logooli, percentage of words spoken in Logooli, and the number of unique
Logooli words used are given for both consultant and elicitor(s).
MON OLI NG UAL FI ELD WOR K IN A ND BE YO ND TH E CL ASS ROO M:T HE LOG OO LI E XPE RI ENC E AT UCLA 481
Number of turns by consultant
Number of turns by elicitor(s)
Total words spoken by consultant
Logooli words spoken by consultant
Total words spoken by elicitor(s)
Logooli words spoken by elicitor(s)
Percentage of words spoken by consultant
Percentage of words spoken by elicitor(s)
Discrete Logooli words used by consultant
Discrete Logooli words used by elicitor(s)
Table 1: Comparison of bilingual and monolingual sessions.
Table 1 shows that although the bilingual session excerpt is shorter and
comprises fewer turns, both consultant and elicitor spoke many more words in those
fewer turns than in the monolingual session. This confirms the impression from
§5.2 that the substantive turns in (2) are generally longer than the turns in (1). In
terms of different Logooli words, however, in this wordy session fewer than 20
unique Logooli words were used, in contrast to the monolingual session.
In the bilingual excerpt, only 105 of 369 words uttered by consultant were in
the target language. The other 264 functioned to translate, explain usage, or
elaborate on previous explanations—or were simply niceties, like the utterance:
Cool. If I get confusing or something, just let me know, seen in (2). Only 13% of
words uttered by elicitor in this excerpt were in the target language.
In the bilingual excerpt, the consultant used a total of 16 discrete words in
Logooli throughout the session. The elicitor used a total of 17 discrete words. One
word used by the consultant was not used by the elicitor, and two words used by
the elicitor were never spoken by the consultant. This is because Logooli was only
used in a small number of example sentences, repeated by both elicitor and
consultant, with English used to discuss the nuanced differences between the
sentences. In contrast, in the monolingual session, the consultant used 27 more
forms than did the elicitors while explaining terms in the target language. This may
mean that the session generated extra data, some of which the elicitors were unable
to parse at the time, but could lead to insights later.
Because the two sessions occurred at different points in the trajectory of the
course, the bilingual session was productive in its own way. But if the initial
data-gathering component of the course had proceeded with such a heavy
English-to-Logooli balance, course participants’ understanding of Logooli as a
whole language, and their collection of data, would have been greatly slowed.
482 HAN NAH SA RVASY
Monolingual elicitation is not only useful for long-haul stints in bush villages; it
can also be a stimulating and memorable format for an advanced field methods
course. Attempting to speak the target language from the beginning of such a course
forces students to think on their feet, drop their armor of theories and analyses, and
to boldly make mistakes in public—as fieldworkers must, regardless of
methodology. This alone makes monolingual methods better preparation for actual
fieldwork than translation-based methods.
But the differences between the monolingual and bilingual class session
excerpts examined here also show a new side to the debate: the relative efficiencies
of monolingual and contact language-based field methods for data collection. The
bilingual excerpt here is weighed down by excessive use of English parentheticals
and niceties by both elicitor and consultant; it also lacks any new lexical or usage-
related discoveries in the target language. While turns in the monolingual excerpt
are shorter, the range of target language vocabulary is much broader. The
monolingual excerpt also reveals several usage-related quirks related to self-
correction that would not turn up in a bilingual session, since there all correction is
done through English.
Since monolingual methods curtail excessive meta-discussion in a contact
language, they can facilitate highly-efficient gathering of data at the beginning of a
field methods course. In a graduate field methods course run using conventional
English translation, the consultant may feel s/he needs to prove intellectual mettle
or expert status to the PhD students in the class, thus increasing English
commentary or niceties.
Further, because students in field methods courses are under no time pressure
to publish their results, drawback D1—the potential increased time burden of the
monolingual method—is irrelevant. If every linguist could follow a Prince of
Serendip at one point during graduate training, what wondrous discoveries the field
of linguistics would behold.
Aikhenvald, A. Y. 2007. Linguistic fieldwork: setting the scene. In Focus on: Linguistic Fieldwork.
Special issue of Sprachtypologie und Universalienforschung, ed. by A. Y. Aikhenvald. 60.3–
Boas, F. 1911. Introduction. In Handbook of American Indian Languages, Part 1, ed. by F. Boas.
Washington: Smithsonian Institution.
Chelliah, S. & W. De Reuse. 2011. Handbook of Descriptive Linguistic Fieldwork. New York:
Darwin, F. & A. C. Seward, eds. 1903. More Letters of Charles Darwin: A Record of His Work in
a Series of Hitherto Unpublished Letters, Volume 1. London: John Murray.
Davis, H., C. Gillon, & L. Matthewson. 2014. How to investigate linguistic diversity: lessons from
the Pacific Northwest. Language 90.e180–e226.
MON OLI NG UAL FI ELD WOR K IN A ND BE YO ND TH E CL ASS ROO M:T HE LOG OO LI E XPE RI ENC E AT UCLA 483
Diercks, M., ed. 2014. Maragoli texts. Unpublished collection of sound recordings, with some
transcriptions. Pomona: Pomona College.
Dixon, R. M. W. 2007. Field linguistics: a minor manual. In Focus on: Linguistic Fieldwork. Special
issue of Sprachtypologie und Universalienforschung, ed. by A. Y. Aikhenvald, 60.12–31.
Everett, D. 2001. Monolingual field research. In Linguistic Fieldwork, ed. by P. Newman & M.
Ratliff, 166–188. Cambridge: Cambridge University Press.
Gil, D. 2001. Escaping Eurocentrism: fieldwork as a process of unlearning. In Linguistic Fieldwork,
ed. by P. Newman & M. Ratliff, 102–132. Cambridge: Cambridge University Press.
Heath, J. 1985. Clause structure in Ngandi. In Grammar Inside and Outside the Clause: Some
Approaches to Theory From the Field, ed. by J. Nichols & A. C. Woodbury, 89–110.
Cambridge: Cambridge University Press.
Hyman, L. M. 2001. Fieldwork as a state of mind. In Linguistic Fieldwork, ed. by P. Newman & M.
Ratliff, 15–33. Cambridge: Cambridge University Press.
Lowie, R. H. 1940. Native languages as ethnographic tools. American Anthropologist 42.81–89.
Macaulay, M. 2004. Preparing linguistics students for the realities of fieldwork. Anthropological
McLaughlin, F. & T. Seydou Sall. 2001. The give and take of fieldwork: noun classes and other
concerns in Fatick, Senegal. In Linguistic Fieldwork, ed. by P. Newman & M. Ratliff, 189–210.
Cambridge: Cambridge University Press.
Matthewson, L. 2004. On the methodology of semantic fieldwork. International Journal of
American Linguistics 70.369–415.
Mithun, M. 2001. Who shapes the record: the speaker and the linguist. In Linguistic Fieldwork, ed.
by P. Newman & M. Ratliff, 34–54. Cambridge: Cambridge University Press.
Moore, L. 2009. On communicative competence in the field. Language and Communication
Mosel, U. 2012. Morphosyntactic analysis in the field: a guide to the guides. In The Oxford
Handbook of Linguistic Fieldwork, ed. by Nicholas Thieberger, 72–89. Oxford: Oxford
Newman, P. 2009. Fieldwork and field methods in linguistics. Language Documentation and
Newman, P. 2013. The law of unintended consequences: how the endangered languages movement
undermines field linguistics as a scientific enterprise. Linguistics Department Seminar, School
of Oriental and African Languages, University of London.
http://www.hrelp.org/events/seminars/paulnewman. Retrieved 1 November 2014.
Newman, P. & M. Ratliff. 2001. Introduction. In Linguistic Fieldwork, ed. by P. Newman & M.
Ratliff, 1–14. Cambridge: Cambridge University Press.
Nichols, S. & J. Ssenyonga. 1976. Maragoli. Berkeley: University of California Extension Media
Twarog, B. M. 1988. Serotonin: history of a discovery. Comparative Biochemistry and Physiology
Vaux, B. & J. Cooper. 1999. Introduction to Linguistic Field Methods. Münich: Lincom Europa.
484 HAN NAH SA RVASY