Grammar

 
As for the Yami grammar, I mainly relied on information resulting from my conversations and correspondence with Shigeru Tsuchida of Tokyo University and on the research of Tsunekazu Moriguchi of Kumamoto University (1980).
 
Moriguchi calls (o) a topic marker. I decided to call (o) a nominative (NOM) marker. Choosing not "focus" but "subjectivisation" as a basis, I justify my NOM in the following comparative manner: In both English and in Yami some nominal argument must take the nominative case. In English, the subject takes NOM and this is the case in Yami (o). In both languages there are case markers for non-subjects:
 

English

Yami

subjects

NOM

subjects

o

agents

by

agents

no

instr

with/by

instr

no

objects

ACC

objects

so

 
In English, objects and indirect objects can be moved to subject position (passive) and marked with nominative case. In Yami, objects, agents, and instruments can be grammatical subjects and will be marked with nominative (o), but cannot move as they do in English. Thus, I believe, we can call the particle o nominative (Woolford 1986).
 
As for the particle (no), I shall not call it OBL, as Moriguchi does, but genitive (GEN). I also prefer to make a difference between GEN and agentive (AGT) (Mirikitani 1972).
 
The problem of the "mystery particle" (to) remains unresolved. Tsuchida has not published his findings yet, and Moriguchi has not treated (to). I have observed it in a multitude of its occurrences in my texts, but it seems to be a problem which requires great linguistic expertise and I do not feel qualified enough to tackle it. To get around the problem, I simply translated the particle (to) as required by the sentence in which it occurred. In most cases it expresses either suddenness or the continuity of an action. Another problem is the particle (a) when it is "fused" to the end of an adjective as in rakwa baka. My first intention was to separate the two, as in rako a baka, but that would have been somewhat unfair to the recordings, in which they hardly ever occur separated. Finally, I decided to transcribe them in the text as they were pronounced on the tape: in one unit. I intended, however, to break them up in the final form of the dictionary at a later date. The grammars of Itbayaten and Ivatanen are very similar to that of the Yami language. After having learned Yami, I could work quite easily with these languages. When I could not be assisted by my informants, the source which I used most frequently while working on the texts recorded in Itbayat was Yamada's description of Itbayaten grammar and his excellent dictionary of the language.
 
 
Transcription
 
Most of the errors that occur in my transcription of the material presented here are due to an occasional lack of consistency in transcribing and to insufficient research on the phonetic system prior to transcribing the texts. The final [o] has been inconsistently marked by [aw], occasionally pre-, inter-, or post-vocalic [i] was marked /y/, and there are many questionable cases of marking the velar fricative /h/.
 
The greatest difficulty I encountered in transcribing the Yami texts was due to the fact that the Yami men or women were simply not used to telling or listening to stories without chewing their betelnut. Because of the betelnut, the lime and vine bark on it, and the green leaf that wraps it all, the speech of some elder informants occasionally was reduced to something between a mumble and a slurp in which many sounds got fused into untranscribable noise. The informants' talking with a full mouth sometimes reduced the clarity of the recordings to a point where almost no phonologic structure could be made out from the speech signal and even the informants themselves could not understand the replays. In transcription the phoneme most affected by the severe articulatory impediments caused by the nut-chewing addiction of the natives was the velar fricative /h/. This phoneme occurs not only in fixed positions in a given word, but may also be employed as needed to express the meaning of "also," "again," "too," and "frequently." Depending on the position in which it occurs, the pronunciation of /h/ may vary from a strong velar sound to a slight glottal stop. In poor recording conditions, /h/ is very hard to detect even by the native ear.
 
The transcription of Yami names requires a few comments as well. The Yami practice technonomy, which means that the parents are called according to the name of their child. In the case of the Yami, if the first child is called Mazat, the parents will be called by others "father of Mazat" and "mother of Mazat." The name change applies to the grandparents as well. Fourth-generation grandparents, however, are not called by the name of their first grandchild, but are called siapen kotan or fourth-generation grandparents. Since this term is used as a name, in the transcriptions I capitalize it. In order to identify the persons better, I add in parenthesis the name by which the grandparents were known before they became kotan.
 
Because of technonomy there is always a particle that precedes a person's name. This particle, which indicates the social status of the individual, is separated by a hyphen from the name itself, as in: "Si-Mogaz" or "Sinan-Mogaz." If there are grammatical particles that precede the social status marker, they will not be separated from the social status marker, as in: "Ninan-Mogaz." If the recordings show, however, that the informant separated the particles in his speech, then in transcription the grammatical particle will be separated as well: "ni Inan-Mogaz." The names of legendary characters may be preceded by Simina, which is translated as "the late."
 
 
Translation
 
The texts presented in this study appear in four versions. For each text there is a transcription, a word-by-word translation, and a more readable line translation. At the end of the text is a glossary with a numbered occurrence index for each word in the text. After the glossary I have added an interpretative translation. Both the linear and the interpretative translations may not read smoothly at times for the sake of fidelity to the original recorded versions.
 
The difficulties encountered in translating the Yami texts were so many that there is neither space nor time to deal with all the problems in detail. The greatest difficulties in rendering the texts into English arose when the natives themselves did not know what the words meant in their own language. Strange as it may seem, this happened very frequently, and it did not involve young people who, due to acculturation, did not know how to speak their own tongue. In all cases the texts were ancient chants of the rawod, or anohod category. The translation of such texts was very time-consuming because a great number of informants had to be consulted for each individual word, both in and out of context, in several villages. The translations obtained were frequently not only different but sometimes contradictory. After having consulted the vocabulary of all Bashiic languages, I decided on the most likely variants, but I am unsure of the correctness of some of the choices.
 
Another problem was the translation of certain verbal expressions that simply do not have a corresponding equivalent in English. For example, the word sawon in all three languages, Ivatanen, Itbayaten and Yami, expresses emphasis, but in English there is no word which could express emphasis in similar fashion. In most cases I translated sawon as "so," "indeed," or "surely," but I am not entirely satisfied with this translation.
 
I should also mention that all Bashiic languages have typical narrative discourse markers. The Itbayat narrators use the word kono, which means "reportedly." It may be used to mark a "general past tense," but it also can inform the listener that whatever he or she is listening to is not something that the narrator just made up or knows for a fact, but "it was said," by someone, somewhere, once upon a time. The frequency of the the word kono can reach a point where all the verbs, adverbs, and sometimes even adjectives in a phrase may be followed by it. In the transcriptions kono is present as it occurred in the recordings, but in the interlinear translations it has not been included because of its high frequency.
 
The Yami narrators, depending on the composition of their audience, use the expressions sira kahakey (friends), ripos (relatives), or manganak ko, which means "my children." In the case of my recordings, it was the last one that was used most often because, no matter how big the audience, the stories were primarily told to me. In each case, given the age of the narrator, a father-son or grandfather-grandson relationship could have been considered to exist between the narrator and my Yami companions and me, (who were about my age). The use of the expression manganak ko occasionally became so frequent that it obscured the meaning of the sentence. In all cases of what appeared to be a verbal abuse of the expression, it was evident that the narrator repeated it only to gain time to find the proper word for what he was going to say next, and not to reinforce the awareness of the existing age difference between narrator and audience.
 
In the translation of the narratives the following abbreviations are used: NOM nominative; AGT agentive; ACC accusative; BDN bird name; EXP expression; FSN fish name; GEN genitive; ITJ interjection; LI link; LOC locative; PLN place name; PSN person name; PTN plant name; SSN season name; TAG tag question.