SIL International Home

The Nature of Field Work in a Monolingual Setting

Those fortunate enough to own or to have seen Kenneth L. Pike's monolingual elicitation lecture as filmed by the Television Studio of the University of Michigan (1979) entitled 'Pike on Language' (in four parts), or who have, by training or by necessity, ever invented or followed a real discovery procedure in the field, will probably agree with me when I opt for ‘all the lexis and zero grammar’ from the imaginary fairy godmother, as offered by Prof. McIntosh.

Why and how does monolingual elicitation work? What theoretical assumptions must we share before we enter into such a situation?

For those who have not yet seen Professor Pike's 30-minute program, or observed him in real life, let me briefly summarize what goes on in the studio room:

Pike enters a well-lit room with a large table in the middle on which there lies a rich assortment of objects, e.g., stones of varying sizes, leaves of varying magnitude; twigs, small and large, as many as there is room for. There will be flowers and small animals on the table, e.g., a tiny stuffed alligator and orchids. (SIL does a lot of its work in tropical rain forests.) There are no tiny motorcycles or imitation Hondas, or imitation Volkswagens; these are irrelevant. The native may very well have seen such objects, but again, he may not have. Rags and cloths of various colors we call white, black, red, yellow, brown or green are also on the table. Beside the table stands a language helper (earlier known as an ‘informant’), a woman or a man, wearing some local shirt, e.g., made from Batik or similar material.

Pike will say 'hello' in Mixtec, an Oto-Manguean language of Central America which he speaks fluently and into which, with several native speakers and SIL colleagues, he translated the New Testament in just over a decade. The native speaker of the language Pike is investigating responds in some appropriate manner. Pike immediately goes to the blackboard and writes down what he heard, or what he thinks he heard. We cannot assume at the outset that the writing he uses is phonemic or phonetic; that sort of analysis will come later.

Pike, with a questioning look, lifts one of the rocks and asks (either in Mixtec, which his language helper has certainly never heard) or by gesticulation something like ‘what's this?’ or ‘what do you call this?’ The helper responds by saying something that Pike perceives as . He then points at two stones and repeats the same questioning gesture or noise. The helper responds with két kõ. (Pike does not know the orthography of the language nor what language it is. He will probably write the first utterance as [kø] and the second one as [ke:t kø].) He will reach for three stones and make the same interrogative facial expression. The speaker says három kõ which Pike tentatively notes as [ha:rom kø]. He has realized by this time that the form for ‘stone’ doesn't seem to change with the ‘numeral’ before it, or what goes as a numeral in this strange tongue. He holds up five fingers and asks, facially, ‘how many?’ to which the helper answers [öt], then he holds up ten fingers, to which the answer is [ti:z].

Pike has barely been in the room for 3 minutes, when he is enabled by this information to make a bold generalization: nouns do not ‘pluralize’ in this language, but are unchanged after the numeral. He is now ready to test this hypothesis and risks his own first construction: [ti:z kø]. He looks at his helper expectantly. The native speaker smiles and nods, and repeats the utterance [ti:z kø] holding up both of his hands. This confirms Pike's hypothesis that whether you say ‘one stone’, ‘two stones’, ‘three stones’, or ‘ten stones’ the form for ‘stone’ will not change in this language.

He now holds up one finger and says questioningly: [kø], uh, uh, uh, [kø]? His helper gets his meaning (Pike wants him to say ‘one stone’) and says [εd' kø]. Pike now has the numerals 1, 2, 3, and 10, and testing his hypothesis starts to count out loud: [εd'], [ke:t], [ha:rom], … [ti:z]? Amused, the helper smiles and shakes his head. He says, helpfully: [εd'], [kεt:ø], [ha:rom], [ne:d'], [öt], [h*t], [he:t], [ñolts], [kilεnts], and [ti:z]. Pike now tries all forms for ‘one’ through ‘ten’ with the word for ‘stone’ only to realize that the word for ‘2’, [ke:t - kεt:ø] is the only one to have a double form, one for counting, one for use before a noun.

He is now ready to test whether this language has no plural form at all, and points at one stone first, then at a larger heap of uncounted stones. The speaker replies with a smile: [kø - kövεk]. This gives Pike the idea of making collections of uncounted objects, and he then holds up one of the objects and points at the whole heap. He starts with ‘leaf’ and gets the following result: [lεve:l lεvεlεk]. He points at a ‘chair’ then at many chairs and gets [se:k se:kεk]. For the pair ‘tree’ vs. ‘trees’ (these may be out of a color picture book) he gets [f* - fa:k].

He now retests the nonpluralization hypothesis by saying the numerals before the singular form of each noun with the exception that he uses the form [ke:t] as a numeral modifier before a noun. His hypothesis is confirmed; all nouns (so far) seem to behave identically.

He reaches for a green rag and for a green leaf and holds the two objects beside each other. He then takes a white rag and holds that before the green leaf. The helper understands his intention and pointing to the green objects says [zöld]. When Pike says the word for green and holds up the object he wishes so characterized, the speaker responds: [zöld lεve:l] for 'green leaf', and [zöld roñd'] for the green rag. Pike now gets more objects of the same color and gestures as before. The native answers, surprisingly: [zöld lεvεlεk] and [zöld roñd'ok] for ‘green leaves’ and ‘green rags’, respectively.

Pike tentatively concludes that this language has no adjective-noun-agreement as the adjective (if that’s what it turns out to be) stays in the ‘singular’ form; even though the noun was pluralized. He suspects some kind of V[k] structure is at work in pluralization in this language and tries the form [zöldεk lεvεlεk] on the helper. The helper laughs, and corrects him by saying: [zöldεk * lεvεlεk] and by also volunteering [* lεvεlvεk zöldek] with a rising intonation and an interrogative look on his face. Then he answers his own question: [igεn * lεvεlεk zöldεk]. From this Pike hypothesizes that adjectives will co-pluralize with the noun they refer to not as Head-Noun modifiers with the noun phrase, but as predicating adjectives. He correctly guesses that the helper said ‘the leaves are green’ and that he also asked ‘are the leaves green?’ The helper's intonation had a lot to do with guessing that the one was a question and the other was an assertion or a statement.

Pike now moves in for the first verb. (He has realized that in order to say ‘the leaves are green’ or ‘are the leaves green?’ this language uses no verbs. He sits down in a chair, gets up, and motions to the man to sit down. The helper says: [ülök, üls, ül, ülünk, ültök, ülnεk] and gestures first at himself, then at Pike, with the third utterance he obviously pretends that a third person is present; with the fourth he makes Pike sit down while he, too, sits down, etc. This makes him realize that the language has morpheme endings on the end of what seem to be ‘verbs’ and that these ‘verbs’ seem somehow to be ‘inflected’ in the six persons corresponding to ‘I’, ‘thou’ ‘she/he/it’, ‘we’, ‘you’, and ‘they’.

There are some toy dogs on the table. This enables Pike to create a situation in which some dogs sit (these are the ones nearer him), and the other ones (farther from him) stand. He gets [εzεk * kut'a:k ülnεk, *zok * kutya:k a:l:n k.]. He correctly concludes this means ‘these dogs are sitting, those dogs are standing’.

Barely fifteen minutes into the elicitation, he has the rudimentary alphabet, presumably a phonemic script of some sort, figured out; he has vowels, consonants, word initial stress; and due to the strange behavior of the plural suffix k, he begins to guess that this is a vowel-harmony language. Moreover, he has taken steps to discover and to describe the morphology of the language, which—lacking gender and number agreement—seems to be an agglutinating Asian tongue. The absence of the copula in stative sentences of the sort ‘these leaves (are) green’ reminds him of Russian, but he rules out Slavic, because of the lack of gender and the probable presence of vowel-harmony. He has begun to build a theory of Noun Phrase and Verb Phrase in the sentence and at the end of the 20th minute can ask questions of the type ‘what is this?’, ‘what are these?’ ‘what are those?, since his helper has inadvertently slipped him the right lexis to do so.

I can attest to the exactitude of this procedure, since I was the informant. Pike, I have found out, has never done Hungarian, and I thought it would be fun to surprise him. I dressed up as an old Hungarian peasant, repleat with moustache, beard and white wig; torn, old clothes, and a shepherd's drinking flask on my shoulders and a walking cane. MY make-up helpers assured me that I was unrecognizable and that I looked 98 years old. Pike knew I was in Dallas that week, but he just saw me a few minutes before wearing a business suit with a white shirt and tie and was assured that an illiterate old peasant from Asia was to be interviewed that evening. In 38 minutes (I looked at my watch stealthily when he turned to the board) he had the main sketch of Hungarian morphosyntax worked out. When he got to the point he could ask questions, he knew it was a Finno-Ugric language, and laughter from the audience tipped him off. I lifted my garb and the show, for that night, was over. But on the University of Michigan tape he interviews a man who speaks the rural Javanese dialect, similar to Bahasa Indonesia, but sufficiently different, and he does this year in and year out in Africa, Papua New Guinea, and other distant points on the planet somewhat like a tournament chess-master who plays 40 games simultaneously. These ‘games’ are, of course, the workshops Ken and Evelyn Pike hold both in the States and abroad, when they go on one of their numerous around-the-world trips to teach beginners in language description.


From Adam Makkai, "The lexo-centric approach to descriptive linguistics," pp. 54–57. In Benjamin F. Elson, ed. 1986, 47–61, Language in global perspective: papers in honor of the 50th anniversary of the Summer Institute of Linguistics 1935-1985. Dallas: Summer Institute of Linguistics.