Linguistic Fieldwork Research Paper

Until recently linguistic fieldwork on unwritten non-European languages did not attract much attention in mainstream linguistics. It is only since the early 1990s that the linguistic departments of universities and the professional societies have become increasingly aware of the need for researching these languages, about 90 percent of which are endangered (Grenoble and Whaley 1998, Krauss 1992).

2. Learning And Analyzing Unresearched Languages

Fieldwork on languages which have not yet been described differs from other kinds of linguistic fieldwork in that the linguist comes to the speech community with no or only marginal knowledge of the language (called field language). The process of analyzing its structure and use is therefore closely tied to the process of language learning. This means that the choice of research topics, the sequence in which they are studied, and the research methods and techniques depend on the fieldworkers’ progress in learning. The better they understand and speak the language, the higher the quality of their data will be, because they will have to rely less on translation. Although the progress of learning a field language depends on the structure of the language and the talent of the researchers and their teachers, experience shows that the methods which have been developed for learning and analyzing a field language work well for very different types of languages and people.

Since languages are never homogenous systems, but show regional, societal, and situational variation, the linguists and the speech communities they work with have to decide which variety should be the object of their first investigation. Only when they have a good command of one variety can they endeavor to study another.

3. Time Plan

Before leaving for fieldwork, the researchers gather as much information as possible about the field language, and the political and sociocultural situation of the speech community they are going to visit. If no information of the structure of the language is available, they should try to access information about related languages in order to get an idea of what to expect. For the work in the field itself, the ideal time plan consists of at least three field trips. The first trip is used for establishing contact and collecting the first samples of data. On the second trip the main body of data is collected, while the last trip will be used for filling gaps and checking what has remained unclear after the analysis and description of the results of the previous trip.

Data collection during the fieldwork is not confined to audio or video recording of spoken language and taking notes during interviews. Rather, all recordings need to be transcribed and translated with the help of native speakers who understand exactly the context of the speech situation and what the people are talking about. In addition, the fieldworker should at least draft a rough analysis of the data, as only analysis and description reveal gaps and unclear items in the data. If for practical and financial reasons it is not possible to visit the fieldwork site several times, the fieldworker has to structure her/his stay into comparable phases of work, so that phases of interviews and recording, translation, and transcription alternate with phases where the researcher steps back from data collection in order to concentrate on analysis and description.

The question of how much time has to be planned for doing fieldwork with the purpose of writing a grammar, compiling a dictionary, or editing a reasonable number of texts cannot be answered as it depends on the situation of the fieldwork, how the fieldworkers and the speech community cooperate, the availability, talent, and the training of the people they work with, the complexity of the language and culture, etc. In other words, the course of fieldwork is influenced by numerous things that cannot be predicted, but which nevertheless have to be planned. In order to cope with all kinds of unexpected difficulties, the research plan has to be structured in such a way that even in the event of a premature abandonment of the project, the data are sufficiently processed to yield at least some useful output. With regard to dictionary work this means, for instance, that you do not start with the letter A and then work through the alphabet letter by letter, because if the project stops for whatever reasons, a dictionary comprising only a limited number of letters is completely useless.

4. The Relationship Between Linguists And Informants

The outcome of fieldwork always depends on how well the researchers are accepted by the community, and on which people they can work with. The ideal fieldwork situation is to live in a family or similar kind of group, and be exposed to the field language most of the day. This, however, presupposes that the researchers can adjust to a different standard of food and hygiene, accept their rules of behavior, and can tolerate less privacy than most academics are used to. After having settled, the most difficult task is to find people who can work with the researcher on a regular basis. Especially at the beginning, these people, traditionally called informants, must be able to communicate with the linguist in a second language; but they should also be good speakers of their native language and enjoy studying it together with the linguist. Which people whom the speech community considers to be suitable informants do not exclusively depend on what kind of skills are required for the work, but may also be determined by the linguist’s age and sex, by family relations, prestige, or political matters (Grinevald Craig 1987, pp. 9–13)

As in Western societies, work relationships can be of very different kinds. But what makes it more difficult is that fieldworkers do not know beforehand how the speech community will define the roles of the linguist and the informants. None of the recently suggested terms for ‘informant,’ e.g., ‘language helper,’ ‘native expert,’ or ‘consultant,’ can cover all these types of relationships. Therefore, instead of inventing seemingly politically correct terms, the traditional term ‘informant’ is given preference in this research paper.

The relationship between linguists and speech communities should be based on a collaborative approach in which the linguists not only research indigenous languages, but also become involved in the training of native linguists and in language maintenance projects (Grinevald Craig 1997, Wilkins 1992). Apart from their work relations, it is desirable that fieldworkers cultivate friendly relationships within the speech community; in an atmosphere of mutual trust, native speakers will be more ready to advise them on culturally appropriate behavior, and allow them to gain deeper insights into the culture and language.

5. Types Of Method

The methods used in linguistic fieldwork depend on the research topic, the progress in the linguist’s ability to communicate in the then field language, and the kind of work that informants regard as being worthwhile. The best way to ensure the interest and cooperation of informants is to explain to them what one is doing and why. The more the informants understand the linguist’s work, the more they can contribute. Thus linguistic fieldwork should be understood as a process of mutual learning in which the linguist learns something about the structure and the use of the field language, and the informants learn something about linguistic methods and the structure of their language.

In particular, linguists employ three kinds of methods: participant observation, interviews, and the recording of narratives, conversations, and other kinds of verbal interaction. Participant observation means that linguists take part in the activities of the speech community so that they can observe how the people communicate and, if they are allowed to play an active role, can also learn to communicate themselves. In the early stages of fieldwork, for instance, games are a good occasion for active participant observation, as the vocabulary used in games is restricted and repetitive, and the grammatical structure of what is said is usually simple.

Participant observation is an excellent means of learning the rules of verbal conduct and the appropriate speech formulas in various situations. It also helps to extend one’s vocabulary, as it is easy to memorize new words and their contextually bound meaning when they can be associated with a particular Activity. But for the analysis of grammatical structure it is less useful; in order to capture various types of grammatical constructions, one needs to write down hundreds of examples, which is hardly possible while being engaged in activities.

Inexperienced fieldworkers might think that interviews are the easiest and fastest way to obtain all the information you need for grammatical analysis and the compilation of a dictionary. This is an error. Before even thinking of an interview, the linguists need to study ‘the local ecology of questioning,’ i.e., ‘to find out who is allowed to question who, when and how’ (Duranti 1997, p. 104). If it is all right to ask questions, the researcher must be careful only to ask those questions the informants can easily answer; not being able to answer a question about their mother tongue can be so embarrassing that they will try to avoid further interviews, or invent a false answer in order to save face. Questions must be worded in such a way that informants can spontaneously answer without being influenced by what they think the linguist wants to hear.

Two types of interview can be distinguished: the metalinguistic interview and the direct elicitation of language data such as words and sentences. The main purpose of the metalinguistic interview is to help with the transcription and translation of recorded texts, and to provide the necessary background information for understanding their content. Before recording the first texts, however, the linguists should learn the basic vocabulary and grammatical structure of the language. Usually they will do this by elicitation, although there are reportedly cases where the speech community appoints one or two people to teach the linguist in their own way ( Wilkins 1992).

All grammatical descriptions must be based on texts which have been recorded, transcribed, and translated, because the function of many grammatical constructions only becomes clear when the wider context of their use is known. In interviews the informants only produce isolated sentences; how these would fit into the natural flow of narratives or other genres of speaking will remain unclear. Nevertheless, elicitation remains necessary in all phases of fieldwork, because even a very large corpus of texts does not contain all word-forms and grammatical constructions which belong to a systematic grammatical description. Similarly, the lexicography of a field language relies on elicitation and the recording of texts.

6. Fieldwork For Grammatical Descriptions

If the field language is an unwritten language, the very first thing that needs to be done is to investigate the sound system and develop an orthography. A good method which has been practiced by many linguists starts with a minimum dictionary (Foley 1991, Kibrik 1977, pp. 35–9, Nida 1949, pp. 175–91). This minimum dictionary contains 200–500 words of a maximum length of two syllables which denote different kinds of people, bodyparts, animals, concrete objects, every day activities, states of being, and properties (e.g., words for ‘child,’ ‘dog,’ ‘house,’ ‘eat,’ ‘walk,’ ‘sleep,’ ‘sick,’ ‘big,’ ‘small’). As these words can be combined to form meaningful utterances, they do not only serve for the investigation of the sound system, but will also, later on, be used to form the first phrases and sentences (e.g., ‘the small dog,’ ‘the child is sick,’ ‘the child is sleeping,’ etc.).

In order to avoid translations from the second language into the field language, the informants are asked to answer questions like ‘Could you tell me the names of the most common animals here? What are the basic food items?’ and so on. If the word is not too long, the linguist writes it down in a phonetic transcription and asks the informant to translate it or describe its meaning. After at least 200 words have been recorded, the informant’s pronunciation of these words are recorded on a tape or disk. First the linguist says the translation equivalent, or the paraphrase, of each word, then the informant repeats the word in question two or three times in the field language. Afterwards the linguist makes a more refined phonetic transcription which is used for a phonological analysis and the design of a practical orthography, both of which will probably have to be revised later on. If there are people in the community who know to read and write in the regional lingua franca, they should help with the development of the orthography to make sure that it is also acceptable and practicable from a technical point of view.

Once the orthography has been designed, the fieldworker can start with the elicitation of short utterances. Again translations should be avoided, as they could result in interferences from the second language and, for example, show an unnatural word order. One strategy to overcome this problem is to give the informant only the idea of the meaning and present it in a way that it cannot be directly translated. For instance, instead of asking to translate the sentence ‘Eat the rice!,’ one can ask: ‘Tell me, what a mother would say to her little son if she wants him to eat up his rice?’ In order to elicit utterances of the most natural wording, the linguist should help the informant to imagine a realistic scenario. The linguist writes down the sentences given by the informant and later, preferably not on the same day, reads the sentences to informants and asks them to correct the pronunciation and translate the sentences into the second language. These short sentences should also be recorded; while the recording helps the informant to get used to the tape recorder, the linguist can use the recordings for learning the language.

As for the various components of grammar, it is most difficult to elicit the inflectional paradigms of verbs, since paradigms like ‘I ate, you ate, he ate, …’ do not make sense to informants who are not trained in linguistics. But later on, after they have become more familiar with language work, at least some of them will understand what such paradigms are. Since it is impossible to identify all morphemes of complex verb forms in the very beginning, the elicitation should start with plain imperatives as they are used by adults when addressing little children. Probably these are, as in many languages, the least complex forms.

After having captured the basic sentence structure, one should start to record, transcribe, and translate short texts, and use them as the basis for further elicitation by asking the informant to paraphrase certain expressions or to transform sentences. If, for example, the text is about a third person, one can ask the informant to repeat it in the first person: ‘How would you say this sentence if it was you who did this?’

7. The Collection Of Texts

Which kinds of texts fieldworkers collect depends not only, of course, on the focus of their research, but also on what the speech community regards as both worthwhile recording and suitable for publication. The language a speech community uses for their various communicative interactions is not a homogeneous system, but consists of many styles differing in their phonology, grammar and lexicon. Since it is impossible to handle extensive variation when investigating a language from scratch, the fieldworker should start with one well-defined style and at first, if possible, stick to speakers of the same generation and background. As for transcription, translation, and grammatical analysis, the easiest types of text are narratives of personal experiences. Narratives are usually told at a moderate speed, and contain fewer elliptical sentences and a larger variety of grammatical constructions than casual conversations. Furthermore, the recording of conversations is often impaired by several people talking at the same time, which is very difficult to transcribe. Legends and folk tales, on the other hand, are less suitable than personal narratives early in research, as they may contain archaic words and proverbs which require a deeper knowledge of the language and culture. If, however, members of the speech community ask for a particular text to be recorded and written down, the linguist should never refuse.

In order to guarantee that the informants speak as naturally as possible, and do not simplify their language in reaction to the linguist’s low degree of communicative competence, the linguist should ask them to tell the stories in the presence of other native speakers. Ideally, the topic should get the informants so emotionally involved that they forget the tape recorder and the linguist. Since our ethics forbid recording people without their knowledge, emotional involvement can help to overcome the ‘observer’s paradox,’ formulated by Labov (1972, p. 113) ‘To obtain the data most important for linguistic theory, we have to observe how people speak when they are not being observed.’

For the representative documentation of a language, the editing of narratives is not sufficient. In order to give a comprehensive picture of how the language is used across a large variety of speech events, it must also contain casual conversations, ceremonial events, political speeches, jokes, poems, and whatever else might count as a distinct genre in this culture (Duranti 1997, pp. 96f ). The recording of such speech events is, however, only useful if the linguist and her or his team of native speakers are able to transcribe and translate them. Sometimes it is difficult to get the informants to speak on the recorder because they cannot find a topic which they think would be suitable for recording and publication. In this case it can be helpful to show them a film or a picture story and ask them to describe what they have seen. A short film which was especially designed for linguistic purposes is the so-called PearStory (Chafe 1980). Employing the same film for the recording of a number of people can, amongst other things, reveal recurrent patterns of discourse structure which are significant for the respective language and culture. Linguists interested in cognitive anthropology have designed fieldwork kits for highly specialized cross-linguistic investigations such as the conceptualization and expression of spatial relations and the structure of events. Besides questionnaires and video clips, these fieldwork kits also contain materials for language-related games in which native speakers have to fulfill certain tasks while being recorded.

8. Lexicographic Fieldwork

The dictionaries of field languages are usually bilingual dictionaries in which the meaning of the words of the field language are translated into an international language. Some dictionaries are trilingual and have, in addition, translations into the regional lingua franca. Both the fieldworker and the informants must be aware that writing a good comprehensive dictionary takes years, if not decades. If their time is limited, it is more reasonable to plan a glossary to accompany the edition of texts and smaller specialized dictionaries on selected topics such as the terms for food, parts of the house, animals, or speech-act verbs.

The work on a dictionary starts with compiling lists of those words which become the headwords of dictionary entries. There are two complementary methods of compiling word lists: the extraction of words from texts, and elicitation. The extraction of words from texts has the advantage that the words are embedded in sentences which can be used as natural examples in the dictionary. The disadvantage is that even the most common words may be missing. Therefore this method has to be complemented by elicitation. Elicitation begins with the identification of subject areas which should be covered by the dictionary and which are most suitable to start with because the meaning of the words is easy to understand and explain.

The biggest problem of lexicographic fieldwork is that informants are usually not fluent speakers of the second language, whereas the linguist only has a limited knowledge of the field language and culture. Furthermore, many words cannot be translated by single words of a European language because they denote very different concepts. In such cases a translation should be accompanied by a definition. If space does not matter, the definition can be given both in the indigenous and the second language so that the semantics of the headword as understood by native speakers is clearly shown.

A good example for the usefulness of definitions are animal and plant names. A dictionary which only gives their translations would not show the taxonomy of animal and plant names in the indigenous language and culture, as the semantic relations between generic and specific terms would not be explained. In Samoan, for instance, all fishes, dolphins, whales, and turtles belong to a single category of animals which is called i’a. A bilingual dictionary would translate the word laumei as ‘turtle,’ but not explain that it belongs to the class of i’a, which is mostly misleadingly translated as ‘fish.’

Bibliography:

1. Chafe W L 1980 The Pear Stories: Cognitive, Cultural and Linguistic Aspects of Narrative Production. Ablex, Norwood, NJ
2. Chambers J K, Trudgill P 1980 Dialectology. Cambridge University Press, Cambridge, UK
3. Craig C 1987 Jacaltec: Fieldwork in Guatemala. In: Shopen T (ed.) Languages and Their Speakers. University of Pennsylvania Press, Philadelphia, PA
4. Duranti A 1997 Linguistic Anthropology. Cambridge University Press, Cambridge, UK
5. Foley W A 1991 Fieldmethods. In: Malmkjaer K (ed.) The Linguistic Encyclopedia. Routledge, London
6. Grenoble A, Whaley L J 1998 Endangered Languages. Current Issues and Future Prospects. Cambridge University Press, Cambridge, UK
7. Grinevald Craig C 1997 Language contact and language degeneration. In: Coulmas F (ed.) The Handbook of Sociolinguistics. Blackwell, Oxford, UK
8. Kibrik A E 1977 The Methodology of Field Investigations in Linguistics. Mouton, The Hague
9. Krauss M 1992 The world’s languages in crisis. Language 68: 4–10
10. Labov W 1972 Some principles of linguistics methodology. Language in Society 1: 97–120
11. Nida E 1949 Morphology: The Descriptive Analysis of Words. University of Michigan Press, Ann Arbor, MI
12. Samarin W 1967 Fieldlinguistics. Holt, Rinehart and Winston, New York
13. Wilkins D 1992 Linguistic research under Aboriginal control. A personal account of fieldwork in Central Australia. Australian Journal of Linguistics 12: 171–200