THE GENITIVE IN ERZYAN (MORDVINIAN)

AND THE WAY IT IS USED IN THE NOVEL

"ИСЯК ЯКИНЬ НАЙМАНОВ"

BY KUZMA ABRAMOV

ЭРЗЯНЬ КЕЛЕНЬ АЗОРКСЧИНЬ НЕВТИЦЯ ПАДЕЖ,

КОДА ТЕВС НОЛДАВИ

КУЗЬМА АБРАМОВОНЬ СЁРМАДОВТ

«ИСЯК ЯКИНЬ НАЙМАНОВ» РОМАНСОНТЬ

Хельсинкисэ тёжа вейксэсядт вейксэньгемень котоце иень умарьковонь сисемгеце чистэ Эрзянь келень сёрмаксчиденть вельть вишкине сёрмадовксонть анокстызе

Яку (Jack Michael Rueter)

The genitive in Erzyan (Mordvinian) and the way it is used in the Novel "Исяк якинь Найманов" by Kuzma Abramov.

To the folks at home who helped me get started

"Одксцёраксчим сравтния,

Эрьва козонь правтния"

"Тетянь кудо, содамак,

Авань кудо, нолдамак"

(А. Арапов. Цёрань моросто саевкс, Сятко 1996-3)

PREFACE

My first contact with the Erzyan language was in 1987 when I turned down a job opportunity offered in 1987 only to accept a similar one the following year. In April of 1988,1 sat down at the computer, and after a series of mishaps, erased documents etc., I was well under way with the task of doing the data processing for Martti Kahla, chief editor of Heikki Paasonen's dictionary of Mordvinian dialects.

For the first four years, I found myself mainly involved with the computer typesetting and put my efforts into comprehending the German and Russian texts. In fact, it was not until 1992 when I made my first journey to Saransk that I energetically pursued an active knowledge of the Erzyan language. Until then my knowledge had been little more than a series of letters and diacritics. For this reason I say that 1992 is when I actually began my studies of one of the Mordvinian literary languages, Erzyan.

Why was it Erzyan and not Mokshan you may ask. I was invited to Saransk to participate in a dialect expedition with the Erzyan students. It was probably this exposure to youthful speakers that made me believe that this was a language to learn. Perhaps, if the Mokshans had invited me, I would have begun speaking Mokshan at first. Over the years, I have only been aware of Erzyans who have invited people on their expeditions. Once you get to Saransk, however, the table turns and it seems the Mokshans are the ones with more initiative, as far as invitations go.

During my second and third year of studying the Erzyan language, I became more and more familiar with the works of various Western scholars in the Mordvinian field. My readings became more and more interesting as I went because I continually attempted to read modern Erzyan literature. Some of the authors I have read include

people like Abramov, Bryzhinsky, Chesnakov, Doronin, Gorbunov, Martinov to name a few.

In the summer of 1995, I visited the Mordovian Republic, in the European part of Russia just south of Nizhny Novgorod. In the capital, Saransk, I had an opportunity to interview various people from the University. The planned excursion to the Erzyan-speaking countryside had been canceled, so I spent those two weeks in Saransk working on my master's thesis.

Professor Dmitri Tsygankin and Nina Adushkina were very helpful in supplying me with answers to my numerous questions concerning the usage of the genitive in the Erzyan literary language and everyday speech. I also payed frequent visits to the editorial staffs of the Erzyan-language journals, Syatko (a literary journal) and Chilisema (a children's journal).

Later in the autumn of the same year I was in contact with Nina Agafonova, who at that time was a scholarship holder at the University of Turku. And at times it has been quite helpful to discuss translation into English with Niina Aasmäe — an Erzyan resident of Tartu, Estonia — who teaches English and Italian at the University there.

I would like to express my special thanks to my colleagues at work, the Erzyans of Loppi, Saransk, Tartu and Turku who have taken time out of their own leisure to help solve many of the problems a study of the Erzyan genitive has entailed, and to my professor Raija Bartens for her persistence and painstaking efforts in the reading of my rough drafts. My thanks also to John Acher, who read the proofs and helped make my English more accurate and readable. My heart-felt gratitude to Nina Aasmäe, Nina Adushkina, Nina Agafonova, Aleksandr Feoktistov, Mikhail Mosin and Dmitri Tsygankin for their input in the examination of the material and to the grand old man

Kuzma Abramov himself who wrote the books and took the time to explain some of them to an ignorant foreigner.

February 1996, Kyrkslatt, Finland Jack M. Rueter

1. INTRODUCTION

There are two literary languages that represent a historical split in what is often called Mordvinian — Erzyan and Mokshan. Although the published materials in Erzyan and Mokshan might date back to the early part of the nineteenth century — involving translations of parts of the Bible — the active cultivation of the literary language did not come about until nearly a century later.

Today, both languages have developing literary traditions, which for all practical purposes, are based upon writers of the Soviet era. The nineties have seen a call for development in the Erzyan and Mokshan languages, which has included a change in orthography — and in the case of Erzyan — an effort to revive a language without the "Russian" conjunctions.

Other changes observable in the re-Erzyanification of the language have been the tendency of replacing adjective forms borrowed from the Russian with Erzyan equivalents. One of the ways of deriving adjective-attributes is to simply add the indefinite genitive -нь to a noun stem. In this way, the number of glosses might actually be reduced. In practice, however, there seems to exist a variety of forms, for example a direct loan of the Russian дипломатический, дипломатической, has become дипломатиянь "diplomatic", which is an indefinite genitive of дипломатия "diplomacy". On the other hand, the loan word химия has an indefinite genitive form химиянь "chemistry (attribute)" that appears alongside of the Russian loan химической "chemical". In the "Ersalais-suomalainen sanakirja" by Jaana Niemi and Mikhail Mosin, both forms are given the same definitions with no cross reference (see 3.1).

The language revival is still in an infant state. The latest grammar published in Saransk dates back to 1980 — "Грамматика мордовских языков". On the whole it is a treaties of the Mordovian language with examples from both literary languages producing evidence for a Mordovian language. There are, however, other studies of the language which have been printed abroad. These include a study of infinitive forms by Raija Bartens "Mordvan, tšeremissin ja votjakin konjugaation infiniittisten muotojen syntaksi" (Helsinki 1979) and László Keresztes's book "Chrestomathia morduinica" (Budapest 1990).

This material was helpful to me in the understanding of the Erzyan grammar. It was also useful to me in the spring of 1994 when I began doing some morphological coding of a short piece of Erzyan text. The morphological coding of Erzyan was a part of the Euro-typo corpus project lead in Finland by Maria Vilkuna. In anticipation of this work, I had also spent a month in Saransk, the capital of the Mordovian Republic. With me, I had a copy of a chapter by a prolific Erzyan writer, Kuzma Abramov.

Kuzma Abramov — the Mordovian national writer, acclaimed the state laureate of the Mordovian Republic, and holder of the Order Friendship of Nations — has worked with writing from the 1930s until the early 1990s. During this time he has published 63 titles of prose and 52 titles of poetry in the Mordvinian languages — a few are written in Mokshan (KGA 1994). The title I chose was one which I thought would reflect the mood of change in the literary language, i.e. "Исяк якинь Найманов" which is a second edition of Abramov's first novel "Найман". The second edition can most readily be characterized by a reduction of "Russian" particles and a substitution of "Russian" words with Erzyan ones. Later comments by Mordvinian linguists, however, have reduced the credibility of this conjunction hunt (read witch hunt).

After my return to Helsinki in April 1994,1 spent six weeks coding and translating the first chapter of this novel. During this time, I became conscious of certain grammatical phenomena that required more information. Many of my questions on derivation, declination and conjugation were answered by Keresztes's book, and problems involving the understanding of syntax could be answered due to my readings of Raija Bartens's treaties of infinite forms.

Subsequent reading of Erzyan prose, however, led me to further contemplation of the split between so-called infinitives and gerunds in the Mordvinian languages. Was there reason to maintain that a possessive suffix added to a verb form could help differentiate between gerunds and infinitive — not if only intransitives prove the existence of gerunds and transitives that of infinitives.

My involvement with the computer typesetting of the extensive dialect dictionary of the Mordvinian language edited by Martti Kahla — "H. Paasonens Mordwinisches Worterbuch I-IV" (Helsinki 1990, 1992, 1994 and the forthcoming fourth volume) (= MW I-IV) — also gave me an opportunity to question the forms presented in Keresztes's table of possessive suffixes. It seems, however, that his choice of forms is based on part of the presentation made by Heikki Paasonen (1953: 4-5).

In GMYa 1980, the genitive is dealt with separately from the -нь adjective forming ending. There is argumentation, supported by contradictory examples, for the description of the roles played by the genitive. There was dialectical evidence in the Paasonen Dictionary which spoke for and against the claims in the GMYa (1962:191), for instance there exist both a genitive and an absolute form in the literary construction телень чи "a winter('s) day". Of course, the semantics would require the meaning "winter sun; winter day" in "нурькине теле#чись "the winter day is short".

10

Consequently, I returned to the Erzyan author, Abramov. It would be much easier to establish his usage of the language and then expand to other writers. Here I also found evidence that called for a reevaluation of the split between gerunds and infinitives, i.e. it began to seem as though the bulk of gerunds was concentrated on the intransitive verbs and the majority of infinitives was derived from transitive ones (see above).

In the summer of 1995 I revisited Saransk and had an opportunity to meet Abramov himself. During the course of three sessions, I was able to learn a little of his personal background and clarify some of the questions his text had caused.

KUZMA ABRAMOV

Mr. Abramov was born in 1914, in the village of Molnia (now a part of the town of Naiman in the Bolshebereznikovsky rayon). His father, Grigori Stepanovich, a local resident, was a school teacher who had received his education before the October Revolution. His mother came from a neighboring Russian-speaking village two to three kilometers away. In those days, it was customary to find a spouse outside of one's own village.

After their marriage Kuzma's mother began learning her husband's language, and within a year she had become fluent in Erzyan. This phenomenon of a majority-language speaker adapting to a minority language is by no means unknown today in some of the Erzyan-speaking tracts. Even this past summer, 1995, I encountered a couple of examples in the small village of "Viren Tashkina" along what the Erzyans call the Rator River (Russ. Алатыр).

After completing elementary school in Naiman, Kuzma transferred to the middle school in Ichalki — also located in the eastern part of the Mordovian Republic —

11

where his father was director. From there Abramov went to Saransk to attend the Pedagogical Institute. Having completed a year of studies, he took up teaching at his father's school in Kosogory, Bolshebereznikovsky, where he worked from 1935 until 1937. In 1937 he transferred to Semeley, Kochkurova. He then worked briefly as an editor for the Erzyan-language publishers from 1938 to 1939, before joining the army in 1939 where he served until 1945.

Abramov remarried in 1945 — I received no information as to the whereabouts of his first wife. This time he married an Erzyan woman from the Ardatovsky (currently Dubensky) village of Kobayevo. After attending her village school, she had entered the "Mordovian Studiya" and subsequentially graduated. She became an actress in the Erzyan-language theater, where she played roles in both Erzyan and Russian.

Characteristic of the times, the Abramovs spoke both Erzyan and Russian in the home and for all practical purposes brought their children up speaking Russian. This might have also been a way to lighten the load Kuzma's children and family faced because he had been taken prisoner during the war. Nonetheless, the fruit of Abramov's prolificacy as a writer can be enjoyed in both the Erzyan and the Russian languages.

In April of 1942, Abramov was taken prisoner by the Germans and was finally released in 1945 by U.S. forces at Niirenburg. During imprisonment, he had been confronted by personnel from Konigsberg with the question of whether he wanted to transfer to Finland and start fighting for the other side. He did not transfer. When the U.S. forces turned Abramov over to Soviet authorities at the end of the war, he was moved to a filter camp at Nevil near Pskov, and from there he was sent to a detention camp for having been taken prisoner from July 1945 until November of the same year.

12

During these consecutive terms of confinement, Abramov had no contact with speakers of his native language.

The next decade was one of turmoil for Kuzma's career. Back in Saransk in November of 1945, he immediately took a job as musical director of the radio committee, where he worked until 1948. At that time people branded as prisoners-of-war were repressed, and Kuzma was left jobless until 1951. During this interval he and his wife sold rugs at the market. From 1951 until 1954, he worked at the Erzyan and Mokshan Cabinet of the Pedogogical Institute, where he taught Erzyan literature.

Even at that time, Kuzma did some writing in Erzyan, but none of his works were accepted for publication. In fact, it was not until 1957, after he had taught at the Russian-language school in Sivin, Krasnoslobodsky, from 1954 to 1956, that Abramov came back to work for the Mordovian Publishing Company and his first post-war manuscripts were accepted for publication.

When the first printing of the book "Найман", dealt with here, appeared in 1957, Abramov was being allowed back into the public eye, full scale, for the first time since World War II. A second edition bearing the title "Исяк якинь Найманов" appeared in 1987. The second edition was rewritten in places to eliminate many of the so-called Russisms, mainly conjunctions and particles. The other two parts of his trilogy have also undergone such scrutiny but these editions are still awaiting republication.

Although "Исяк якинь Найманов" is a revision of Abramov's first novel, it is still essential in establishing the starting point for the development and refinement of his own literary talents. Perhaps, these two novels should be compared to ascertain the difference of style, Erzyan traits, etc. Kuzma's talents later flourished in books such as "Пургаз" (1988) and "Олячинть кисэ" (1989). By comparing Abramov's use of the

13

genitive with those of his contemporaries and other noteworthy Erzyan authors, we can be provided with some insight into the workings of the Erzyan language. Hence this paper, though only touching the works of one Erzyan author, will help in the evaluation of the entire Erzyan literary tradition.

On the one hand, I have used the modern Cyrillic spelling of Erzyan when quoting Abramov and others. On the other, I have been faithful to Abramov's text in his spelling choice, even though there are new developments in the Erzyan orthography, which would point to a different choice. One exception, however, is that in the case of compound words, which are to be written as single words, I have placed a hash mark # to show which words are connected, for instance цеця#чине "fragrance". The hash mark was a way of making it possible to retain a means of comparison for noun phrases.

In Abramov's texts, I have also found a difference in usage between the absolute stem in цеця "flower" and the indefinite genitive of the same noun цецянь in асодавикс цецянь чине "scent of an unknown flower". This appears to advocate a difference between a noun attribute in the absolute form and a specific indefinite attribute in the genitive. A. S. Yegorova (1976: 86-8) discusses noun attributes in constructions such as кудо#нармунь "domestic fowl" and нармунь#кудо "lit. bird house » chicken coop". A. P. Feoktistov, however, states that in Mordvinian куд пря "house roof can easily be replaced by кудонь пря (Feoktistov 1963: 206).

The novel "Исяк якинь Найманов" also has an example of the use of an -мсто/ -мстэ "gerund" whose possessive suffix refers to the subject of an intransitive verb but it also has an instance where the referent of the possessive suffix denotes the patient. A parallel might be drawn between this and what D. V. Bubrikh states (1947:14), namely the genitive signifies the subject of an intransitive verb and the object of a transitive

14

one. This is shown in noun phrases such as чинь стямо "sunrise, rising of the sun" моронь морамо "singing of a song". GMYa (1962:162) only mentions the role of subject with no reference to transitive or intransitive variation.

Due to various interpretations of the phenomena surrounding the genitive in Erzyan, I have decided to give it an up-dated description, which will attempt to evaluate Abramov's usage and the disparity of descriptions given in the texts above.

All text examples and rules are numbered consecutively (1-208), and examples taken from the novel by Kuzma Abramov have been given a four-digit page number of reference in parentheses without further reference to the author. There are instances where the English translations will consist of run-on sentences, but that has only been done in order to retain the whole of the Erzyan examples. At points in the paper there will word phrases taken out of context, which might have seemingly odd translations. They are, however, contextual translations.

Initially, I will look at the elements involved in the formation of the genitive, i.e. the genitive forms of nouns, pronouns and the possessive suffixes. Usages peculiar to individual forms will also be dealt with at this time. Subsequently, I will move on to the formation of possessive structures and encounter the so-called nonpossessive structures, which in the long run seem to be very few in number.

The habeo-construction and similar locative usages of the genitive will be followed by a look into the objects, which might, in fact, help in the future treatment of the verb forms.

Next come the verb forms where I have observed a three-way split of the verb forms according to the functions "subject of an intransitive verb", "subject of a transitive

15

verb" and "object". It might have been just as easy to limit this split to two ways, but that would have made it more difficult to discern the actual usages. The functions would have been split such that the subject of an intransitive verb would have fallen together with the object. Maybe, this ergative-type split could be dealt with to a greater extent.

Postpositional government looks into the postpositions from a pattern set forth by the absolute vs indefinite genitive vs definite genitive formulas found among the nouns. Finally come some adverbial constructions that are of low frequency in the corpus.

16

2. THE GENITIVE CASE AND MARKERS

In the Erzyan language there are combinations which provide two different ways to show the possessive relation. It can be shown with genitive endings of common nouns (2.1.1), the pronouns in the genitive (2.1.2), the genitive of names of persons and places which also need special explanation (2.1.3), as well as with possessive suffixes (2.2).

2.1 GENITIVE ENDINGS

The genitive endings can be affixed to common nouns, pronouns and proper nouns alike. Although the same types of endings will be brought into question, their paradigmatic usage can vary from class to class. Common nouns have genitives in indefinite and definite forms, whereas pronouns and proper nouns often take indefinite genitives in the singular even when they have definite referents. In the plural, however, they are often open to variation, i.e. usage of the plural definite in Abramov's material had fewer instances and therefore no set pattern could be established.

2.1.1 Genitive endings of common nouns

The genitive in Erzyan has what appears to be three different forms. First, we have the indefinite genitive in -нь which serves both the singular and plural. Second, we have a definite form of the genitive singular in -нть which, like the indefinite genitive ending, can be added to the vowel stems of nouns and adjectives. And third there is a definite genitive plural formed from the allomorphs -тнэ and -тне of the definite nominative plural to which the indefinite genitive -нь is added.

The formation of the definite nominative plural provides for more variety because of the transversion to a newly adopted spelling standard, which renders two coexistent forms in the genitive plural.

17

Certain voiceless consonant-initial morphemes, such as the marker -т/ть of the indefinite nominative plural and definite plural endings, cause the disappearance of the middle stem vowels o and e when they are affixed to the stem in the declension of nouns. Additional instances that cause the ellipsis of middle vowels would include the illative -c; the elative -сто/-стэ; and the inessive -со/-сэ. This is not the case with, for example, the voiced prolative endings -га, -ва, -ка, whereas the -ка form appears affixed to voiceless stems (cf. вирьга "through the forest", пипьгева "along the leg", вакска "past, by").

According to the old orthography, the omission of the middle vowel before voiceless consonant-initial morphemes is standard, but it does not include all words ending in middle vowels, for instance кудо "house; home; building", веле "village". Hence forms such as пильгть "foot" (Indef. Nom. PI) derived from пильге + ть as well as пильгс "foot; leg" (Indef. 111.) derived from пильге + с and similarly back-vowel circumstances such as пангт "mushroom" (Indef. Nom. PI.) < панго + т, пангс "mushroom" (Indef. 111.) < панго + c, conform with the old standard, whereas the new orthography does not call for an ellipsis or the omission of the middle vowels o and e.

Consequently the formation of the genitive plural according to the old orthography is dependent upon this phenomenon of middle vowel omission before voiceless components. See the first variations of the words пандо (3) and пильге (5) below.

Indef. Nom. Sg. Indef. Gen. Def. Gen. Sg. Def. Gen. PI. (1) кал калонь калонть калтнэнь "fish" (2) кедь кедень кеденть кедтнень "hand; arm" (3) пандо пандонь пандонть пандтнэнь "hill" пандотнень (4) кудо кудонь кудонть кудотнень "building; house; home" 18

(5) пильге пильгень пильгенть пильгтнень "foot; leg"

пильгетнень

веле велень веленть велетнень "village"

пря прянь прянть прятнень "head"

2.1.1.1 The indefinite genitive ending -нь

The indefinite genitive suffix is added directly to the word-final vowel of nouns. Nouns ending in consonants, on the other hand, take predictable linking vowels before the indefinite genitive -Hi is added, cf. 2.1.1.2, 2.1.1.3,2.2.1. The linking vowels [o, e] represented in the Cyrillic alphabet by o, e, э are the adhesive necessary for the declension of nouns ending in consonants. They can be predicted according to the rules set forth in (8-11). Here special attention must be paid to the final vowel in the stem as well as the palatal character of the final consonants.

The prediction of linking vowels requires a division of front and back vowels, in addition to the concept of palatal for consonants. Since the phenomenon of "vowel" harmony in Erzyan is of a progressive nature, the quality of vowels is ascertained by the actual quality of each vowel. Hence the front vowels consist of the set /i, e I {и, ы, e, э}, and the back vowels consist of the set/м, o,a/ {y, ю, о, ё, а,я).

Rule (8) renders the linking vowel / o I in a back-vowel surrounding with no palatalization of the consonants following the vowel of the stem.

(8) -o-

whereas the internal stem vowel Vi {+back}

and whereas the word final consonants C are {-palatalized}

Vi(C(C))C_

19

(8a) сув сувонь "fog" (8b) юр юронь "root" (8c) тол толонь "fire" (8d) ён ёнонь "side" (8e) ваз вазонь "calf (8f) мияв миявонь "beaver" (8g) ярсавкс ярсавксонь "food" (8h) ялт ялтонь "puff of wind" In rules (9-11) the linking vowel rendered is the front middle vowel lei represented in the Cyrillic alphabet by {e, э}. Since the phonetics of the Erzyan language is well adapted to the Cyrillics, a familiarity with the separation between palatalizable consonants and nonpalatalizable consonants is helpful in the establishment of these rules.

In rule (9) we are dealing with word stems that have a stem vowel I i, e / depicted by the Cyrillic {и, ы, е, э}, and a set of subsequent consonants that are not subject to palatalization, e.g. the labials / b, v, p, f, m I, the sibilants / zh, sh, ch, shch I and the palatals I g,kl. The resulting linking vowel is written {e} and, in fact, there is a palatal glide present when it is preceded by an nonpalatalizable consonant.

(9) -e-

whereas the internal stem vowel Vi {+front}

and whereas the word final consonant C3 {nonpalatalizable}

C3 = {6, в, г, ж, K, M, п, ф, 4, ш, щ)

Vi(C,(C2))C3_

(9a) серб сербень "Serb" (9b) чев чевень "shake (of wood)" (9c) кеж кежень "anger" 20

(9d) чек чекень check (monetary) (9e) Рим Римень "Rome" (90 верф верфень "wharf" In (10-1) the linking vowel lei can be triggered by a front stem vowel, or a nonfinal palatalized dental. Rule (10) differs from rule (11) in that the former requires a word-final palatalized dental or palatal glide and the latter requires a word-final nonpalatalized dental.

(10) -e-

whereas the final consonant C2 = {дь, зь, й, ль, нь, рь, сь, ть, ць} ЩС1)С2+_

(Юа) кедь кедень "hand; arm" (ЮЬ) ривезь ривезень "fox" (Юс) пай паень "lot" (10d) пуль пулень "dust" (10е) ломань ломанень "person" (10f) умарь умарень "apple" (10g) сёксь сёксень "autumn, fall" (lOh) пежеть пежетень "sin" 810i) кузнець кузнецень "smith" In (10), I was unable to find many nouns ending in ць in the corpus. In examples (lOc-i) it will be observed that the stem vowel is a back vowel. Therefore the presence of a front middle vowel as linking vowel is dependent upon the word-final palatalized dental or palatal glide as seen in (10c).

Rule (11) gives an example in (lid) in which the front middle linking vowel has been determined by the presence of a palatalized dental following a back vowel. This

21

palatalized consonant, however, has been followed by a nonpalatalized, word-final dental, these are circumstances observed in (1 la, 1 le), as well.

In the case of stems ending in vowels, the indefinite genitive ending is merely added to the word-final vowel as stated above, (see definite genitive plural and singular.)

whereas the internal stem vowel Vj {+front} or whereas a subsequent nonfinal Ci {+palatal} and whereas the word final consonant C2 {-palatal} Сг= {д, з, л, н, р, с, т, ц)

(Па) гольф гольфень "golf' (ПЬ) ПИКС пиксэнь "Une" (lie) ТЫЛ тылэнь "(military) rear" (Hd) карькс карьксэнь "thong, lace (for binding on bast shoes)" (He) кирькс кирьксэнь "loop" (llf) сэкс сэксэнь "grime, filth" (Hg) верьгиз верьгизэнь "wolf 2.1.1.2 The definite genitive singular ending -нть

The definite genitive singular ending can be added to word stems after the same fashion as the indefinite genitive ending is added, i.e. the same rules for linking vowel prediction are applicable (2.1.1.1), and this ending is likewise added to a word final vowel where such exist.

22

2.1.1.3 The definite genitive plural ending -тнень/-тнэнь or merely an indefinite genitive added to the definite nominative plural -тне/-тнэ + -нь

In the new orthography, where there is no rule for omission of certain middle vowels, the rules for the distribution of the two allomorphs can be shown as follows.

Rule (12) is applied to words ending in nonpalatalized dentals, regardless of the vowel found in the stem, in other word, the vowel / e /is always a part of the definite nominative morpheme.

(12) -тнэ- + -нь

whereas the final consonant (д, з, л, н, р, с, т, ц}

(12а) лад ладтнэнь "way" (12Ь) паз пазтнэнь "god" (12с) кал калтнэнь "fish" (Ш) сан сантнэнь "vein; sinew (12е} мар мартнэнь "knoll" (12f) пизёлкс пизёлкстнэнь "rowan" (12g) брат браттнэнь "brother" In Erzyan the word for brother (12g) could actually be replaced by one of the following: леля "older brother; uncle", ялакс "younger brother" or a combination of the two, which would render the meaning "brothers" лелят-ялакст. My choice of that word is due to the shortage of Erzyan word ending in -r. Could this shortage of -r final nouns be accounted for in its homonymy with the plural phoneme -т/-ть as observed above in rule (10). In like manner, it has also been difficult to find words ending in -ц (see -ць above).

In rule (13) the ending can be preceded by any vowel whatsoever, or any word ending in a consonant other than a nonpalatalized dental.

23

(13) -тне- + -нь

whereas word ends in V

or whereas word ends in Q

or whereas word ends in C2

Ci = {дь, зь, й, ль, нь, рь, сь, ть, ць)

С2={б, В, Г, Ж, К, М, П, ф, Ч, Ш, Щ}

(13а) ялга. ялгатнень "friend" (13Ь) пильге пилгетнень "leg; foot" (13с) веле велетнень "village" (13d) пандо пандотнень "hill, mountain" (13е) кудо кудотнень "house; home; building (13f) пизэ пизэтнень "nest" (13g) меню менютнень "menu" (13h) пря прятнень "head" (13i) кедь кедтнень "hand; arm" (13j) ривезь ривезтнень "fox" (13k) пай пайтнень "lot" (131) пуль пултнень "dust" (13m) умарь умартнень "apple" (13n) сёксь сёкстнень "autumn, fall" (13o) клуб клубтнень "club" (13p) тов товтнень "flour" (13q) круг кругтнень "circle" (13r) кеж кежтнень "anger" (13s) ёрок ёроктнень "skill" (13t) лём лёмтнень "bird cherry" (13u) шеф шефтнень "boss; chief (13v) палач палачтнень "exicutioner" (13w) ош оштнень "town, city" (13x) лещ лещтнень "bream" The examples in (13a-x) show a number of phonemes that are newcomers to the Erzyan language, for instance the Russian loanword in (13x) has an Erzyan synonym чалган (ERV: 737), which could readily be substituted. The same could probably be

24

said of the French loan in Russian found in (13v). So as you can see, this group has been maximized to illustrate a liberal usage of the language.

The old orthography would require a reshuffle for certain vowel-final words ending in middle vowels. There are a number of words of this type which lose their final vowel before the definite genitive plural ending, e.g. пеке "stomach", пандо "hill", пильге and "leg; foot". Since there is no immediate way of identifying words which are subject to middle vowel ellipsis, it would be easiest to mark these words in the vocabulary separately. To my knowledge, the Erzyan language has no nouns ending in two consonants, whereas the second consonant is voiced. This makes it simpler to examine the words пандо and пильге, whereas here the final vowel is predictable according to the rules set forth for the prediction of linking verbs. On the other hand, the word валдо "light" does not lose its final vowel. Thus we are forced to return to the marking of the vocabulary.

2.1.2 The genitive of pronouns

Formation of the genitive in pronouns requires a familiarity with genitive formation in nouns. The personal pronouns seem to call for knowledge of the possessive suffixes (2.2), and other pronouns call for knowledge of the three varieties found with the common nouns (2.1.1).

2.1.2.1 Personal pronouns

Personal pronouns can be separated from a group of other pronouns and proper nouns, which take their normal genitive in the indefinte form, (see 2.1.2.2) and (2.1.3). This feature, however, is limited to the personal pronouns, and yet not all of them are subject to it. The second person singular, for instance, deviates from the others, i.e.

25

whereas the other personal pronouns appear to have a possessive suffix included in the basic form of the word, it has none. The first person singular might also be considered seperately, since the palatalized [И] could be interpreted as either the indefinite genitive or possibly the special possessive for members of kinship. In the dialects, however, there are instances of possessive suffix like endings for the second person singular, as well (see below).

MOH + нь » монь

TOH + нь » тонь

COH + нзэ » сонзэ

минь + нек » минек

тынь + нк >> тынк

сынь + ст » сынст

my; mine "(sg.) your, yours" "his, her, its; his, hers, its" "our; ours" "(pi.) your; yours" "their; theirs"

The second person singular resembles the first person singular in that the ending is the same, i.e. a palatalized -нь, which actually could be compared to the indefinite genitive ending, as well. In his minimal grammar, Bubrikh (1947:17) gives the two forms — тонть as well as тонь — for the genitive of the second person singular. The modern written language has preferred the latter form, which, as we have seen, does not fit into the set pattern of the other personal pronouns, i.e. a combination which would look like a pronoun stem and a subsequent possessive suffix.

The construction of the personal pronouns in the oblique cases is reflected in my reconstruction of the genitive, which could have been reconstructed with the vowel stems {MO-, TO-, CO-, МИ-, ТЫ-, СЫ-}. Such a solution would partially reflect the system found in Northern Sa*mi, where the genitive of personal pronouns singular are vowel stems, cf. mu, du, su. There is a difference however, in Northern Sami it is the genitive vowel stem that reoccures in the oblique cases, cf. mus, dus, sus (inessive-elative), muinna, duinna, suinna (comitative) or the questionable form munnje, dutnje,

26

sutnje (illative). The case ending can essentially be separated from the vowel stem — in other words the genitive.

The oblique stem in Erzyan would point to a palatalized consonant, which in the case of the first and second persons singular would be the genitive form, see the ablative forms of the personal pronouns I—Ш Sg. and I—1П PL in succession: мондень, тондеть, сондензэ, минденек, тынденк and сындест. In the plural the palatalization is expected, (cf. 17-9). In section (3.5) below there are the forms of the illative тоньзэть (II Sg.) and минзэнек (I PL), perhaps the latter is merely a typographical error, but the former makes reference to a bare genitive form. This type of formulation would, in fact, follow that of the Mokshan language in which the nominative forms of the personal pronouns singular are the only ones with the nonpalatalized final -н. Other forms in the personal pronouns singular except the dative, which requires a different root word all together, are based on what appears to be a genitive form, (see Polyakov 1993: 19-20).

Reconstruction of a vowel stem in Erzyan for the genitive would cause no problems, but perhaps then the simple indefinite genitive ending -нь would be the element present in the oblique cases of the third person singular. The indefinite genitive element would then be followed by a case ending and consequently a possessive suffix. Evidence in support of this kind of hypothesis is found in the relative pronoun forms конаньстэ "out of which, where" and конаньсэ "with which" cf. конасонть "where; in which".

2.1.2.2 Other pronouns

There are other classes of pronouns, i.e. the demonstrative, relative etc., which unlike the personal pronouns (2.1.2.1) with their possessive suffix-like constructions, take their genitive in either the indefinite or definite genitive. Hence we have the

27

demonstrative pronouns (14-9) and the relative pronoun (20a-b), with two different forms.

(14a) re тень "this" (14b) теке текень "this very" (15a) ce сень "that" (15b) секе секень "that very" (16) тона тонань "the other" (17a) неть нетнень "these" (17b) неть нень "these" (18) сеть сетнень "those" (19) нонат нонатнень "the others" (20а) кона конань "whose, of which, that... of (20b) кояа кинь "whose, of which, that... of (21а) мезе мезень "what, which" (21b) мезе мезенть "which" The genitive form in (17a) is used in constructions where it serves as a determiner, and the form in (17b) is the object. Perhaps, more extensive corpus material would show whether analogy could be found in (19), which had no genitive representation in "Исяк якинь Найманов". Forms (15) and (18), with their simple genitive paradigm, show a contrast to the [г-: н-] paradigmatical change found in (14a : 17) and (16 : 19). In examples (20a-b) the use of the forms is problematic. On the one hand, the genitive form in (20b) appears to be only used in reference to +humans, (see 22-3). And on the other hand, the genitive in (20a) can also be used for reference to humans, (see 24).

In examples (21a-b) there is an overlapping of functions in the nominative form. Hence the pronoun мезе can be used as an interrogative and as a relative pronoun, just to mention two functions. In the genitive, however, the functions are divided, i.e. the simple indefinite genitive is used interrogatively and the definite genitive is used as a relative pronoun.

28

In examples (22—4) we are presented with evidence of varied relative pronoun usage for +human referents. There was not sufficient material, however, to establish whether indefiniteness of the referent could call for the use of the genitive form конань in example (24).

(0247) Сынь a васенцеде истя кортнить чокшне ланга ды лецтнить се ломаненть лемензэ, кинь стака сэредемась валызе масторонть покш мелявкссо... "This wasn't the first time that they talked in the evenings mentioning the name of the man whose serious illness had filled the land with great anxiety."

(0023) Надя, кинь лемс пурназель те праздникесь, аштесь бабанзо элес нежедезь ды ваннось инжетнень. "Nadya, in whose honor the party had been organized, was leaning into her grandmother's lap and studying the guests."

(0045) Ванстыцякс сиведекшнесь ало ульцянь атя, конань усад песэ ашти чёскась. "He had hired a man from the low street, who owned the garden at the end of which the woolen mill was located, to stand guard."

There is another instance where a differentiation is made between the genitive chosen on the basis of the referents involved. First, the word весеме and a shorter form весе can take the indefinite genitive when a +human referent is involved. Both forms весень and весемень are used as objects and as determiner of adverbials, but there might be a difference in usage, whereas весень can be used as a determiner for the subject, (see 25), and весемень appears in the predicate, (see 26). This question, however, should be addressed when more corpora are available. Second, the pronoun весементь is restricted to nonhumans, and in this corpus it has only been used as either the head of postpostional construction (27) or as an object (28).

(0095) ...ды сестэ весень сельмтне кармильть ваномо Канаевень пелев. "... and then everybody's eyes would start looking toward Kanayev."

(0170) ...лавкась ули весемень эли кияк вейке ары азорокс?... "...does the shop belong to everyone, or will some one person become its owner?..."

29

(0262) Авазо сонськак зярдо-бути ульнесь тейтерензэ таркасо, умок сивсь авань каштанчизэ ды ней весементь лангс ваны мирдензэ вановтсо, арси сонзэ превсэ. "Even her mother had once been in the girl's predicament, she had lost her womanly pride and looked at everything through her husband's eyes and thought with his mind."

(0125) Сювордык кедем, ды весеиенть стувтсынек! "Squeeze my hand, and let's forget everything."

2.1.3 The genitive of personal names and place names

In Erzyan names of people and places the formation of the genitive is similar to that of the pronouns, i.e. in the possessive construction the indefinite genitive is used, (see 3.1.1.1, 3.1.1.2, 3.1.1.3, 3.1.2.1.1). For the position of object and postpositional valence with a definite article there is a difference. First, personal names retain the use of the indefinite genitive thoughout. Second, place names, at least in "Исяк якинь Найманов", take the definite genitive to mark the roles of definite object and definite valency for a majority of the postpositions, (see 3.3.1,3.5). I have not spoken of proper nouns because that would include the names of holidays, days of the week, etc. — material that is of a minimal quantity in this corpus.

2.2 POSSESSIVE SUFFIXES

When dealing with the genitive in the Erzyan language, it is essential that one be familiar with possessive suffixes. The Erzyan possessive suffixes hold a position equal to that held by pronouns in many other languages. In context, it can be observed that the possessive suffixes express nearly all of the same notions as the personal pronouns do when they are in the genitive. In fact, only the direct object of the finite verb cannot be shown by the possessive suffix. Thus possessive suffixes will be used thoughout the paper as evidence for genitive usage.

30

2.2.1 The possessive suffixes and their variation in the paradigms

The strategy of linking vowel prediction set forth in the derivation of indefinite genitives (2.1.1) is applicable in the formation of the nouns modified by possessive suffixes. Hence no possessive suffix can be added to a noun stem without the presence of a linking vowel. This, it will be observed, is one means of distinguishing between the plural marker -т/-ть and -т/-ть of the second person singular, for instance тевть "things; tasks" versus теветь "your task(s); your thing(s)".

In his subdivision of the Mordvinian language, László Keresztes (1990: 58) presents the Erzyan possessive suffixes according to the following subgrouping (29-34):

Sg (Nom) PI (Norn) Oblique cases (29) Sgl -M -H/-Hb -я (30) Sg2 -т/-ть (-нть) -т (31) Sg3 -30/-ЗЭ -Н30/-НЗЭ -Н30/-НЗЭ (32) Pll -нок/-нек -нок/-нек (33) P12 -нк -нк (34) P13 -ст (-ск) -ст (-ск) The variation I have actually encountered in the Erzyan literary language is more limited than the dialectical possibilities would have allowed for, i.e. only the third person singular has a rigid alternation between two forms, (see 31, 37) and (34, 40), (2.2.2). Examples (29-30) show us instances of what might be a result of haste, i.e. the oblique cases — according to the literary standard — take possessive suffixes that alternate between palatalized and non-palatalized. In example (29) the oblique cases are often seen to take an alternate -м form as well; this is where it varies from the plural. According to Bubrikh (1947:10-2) there are accusative forms for all of the possessive suffixes — except that of the third person singular — which are equal to

31

those of the nominative. In the third person singular the accusative form is the same as that of the genitive. Indeed, Bubrikh claims the third person singular is the only person that has possessive suffixes for the genitive and dative, whereas N. Agafonova (1981:6-8) presents a paradigm with both genitive and dative forms for the first and second persons singular. Here only the paradigms of the personal pronouns plural are left incomplete, i.e. they have no genitive or dative forms, whereas the object shares the form of the nominative possessed. A check of the corpus material revealed no instances of a possessor marked by the possessive suffix of the first and second persons plural. They did, however, mark objects and nouns taking postpositions.

In example (30) the literary language calls for forms identical to those of the nominative singular in -т/-ть for the plural and oblique forms (cf. (29) above). In the northwestern dialect, which calls for an -нт/-нть form of the possessive suffix in the nominative plural, as shown above, there is also an identical -нт/-нть form for the oblique cases (G. Yermushkin: 1984: 18, 119). This form resembles the -нть of the definite article.

Unlike example (30), example (32) is not given a dialectical form. If we were to revisit the northwestern dialect, however, we would find that once again there was a distinction between nominative singular and the rest of the paradigm, i.e. the nominative singular -мок/-мек. This same dialect, which contains the -нт/-нть forms for the second person singular, also makes a difference in forms for the third person plural, which follows the pattern nominative singular possessed versus other. Hence the northwestern dialect has the variation -ст versus -нст, e.g. аваст сась "their mother has arrived" versus аванст тесэть "their mothers are here" (Yermushkin: 1984: 123).

32

In (35) the -M form is even shown for words in the nominative plural as Abramov occasionally uses it. Other native speakers, for instance D. Tsygankin, M. Mosin, N. Adushkina, it seems, accept only -н/-нь forms for the nominative plural in standard usage, (personal discussions 1995-96).

In GMYa (1962:94) Kolyadenkov shows a division of the possessive suffixes for the first person singular in which м and -н/-нь are used in the oblique cases, whereas -м is used for singulars and -н/-нь for plurals. Thus far the evidence for this kind of representation has been confusing and appears to be highly theoretical. The dialectical evidence, found in the writing of other contemporary authors, usually points toward a split between the nominative singular and nominative plural, e.g. -м versus -н/-нь. The possessive suffix in the oblique cases, however, appears to be either -м , -н/-нь or in places -м/-нь irregardless of its actual number, e.g. седейсэм/ седейсэнь. (see also Bubrikh 1947:10, which gives the -м form for the accusative as well.)

Nom.Sg. Nom.Pl. ОЫ. (35) lP.Sg. M н/нь, (м) н/нь, м (36)2P.Sg. т/ть т/ть т/ть (37) 3P.Sg. зо/зэ нзо/нзэ нзо/нзэ (38) 1P.P1. нок/нек нок/нек нок/нек (39) 2P.P1. нк нк нк (40) 3P.P1. ст ст ст Here, we can see that a distinction is made between nominative singular and plural with regards to the first and third singular possessive suffixes. The only clear-cut division is that of the third person singular which marks the nominative singular in one

way -зо/-зэ and retains the form with -н нзо/-нзэ — in all other instances. This

division, nominative singular possessed versus other, is slightly obscured in the first person singular where there is variation even in the literary language. The second

33

person singular and all three plurals use one set of forms for all three contexts, nominative singular, oblique, and nominative plural.

When it comes to the use of the first person singular possessive suffixes in oblique cases, there is a controversy. A survey of modern Erzyan fiction and poetry would show the use of both the -si and -н/-нь variants. In Abramov's native dialect, the -м form is used as a marker of the nominative singular and the oblique cases (Yermushkin: 1984: 117). It seems, however, that in his writing Abramov uses both the -м and -н/-нь forms interchangeably for all forms except that of the nominative singular where the -м form is the only possible choice.

According to grammars such as GMYa (1980:198) , the first person singular form -M is only used when the word in question is in the nominative singular. Other sources such as GMYa (1962:94) and general fiction would provide a more liberal interpretation allowing -м to be used in the singular and -н/-нь alone in the plural.

2.2.2 Special possessive suffixes limited to the genitive and dative

According to GMYa (1980:200) the genitive forms -нь, -ть and -нстэнь of the first, second and third person singular respectively are applied to words of both the palatal and nonpalatal category when the referent in question is one of a relation of kinship rather than a simple possessed referent. Hence авать "your mother's", тетянь "my father's" and цёранстэнь "his/her son's". Agafonova (1981:6-8), on the other hand, gives what must be a regularly alternating form of the possessive suffix in цёранзо "his/her son's", additionally she gives a dative form цёранстонзо "to his/her son", which also shows a back vowel. The form -нстэнь is by no means a genitive one, but rather the third person possessive suffix - dative used for both the singular and plural, according to a discussion with N. Adushkina (summer 1995). It has, however,

34

been maintained that the form as seen in цёранстэнь "to his/her/their son" is the type of construction that would also only occur with kinship relation possessees. Bubrikh, on the contrary (1947:10-2), offers the dative form моданстэнь "to his/her land", which is not a kinship relation. Agafonova (1981:6-8) shows additional dative forms for the first and second persons singular -нень, -тень respectively.

In Abramov's writings the use of the third person possessive suffix - dative can be seen attached to referents other than kin. These referents are, however, similar to terms of kinship in that they denote inalienables, for instance атя#ялганстэнь "to his fellow elder" (0208) and шабрансто "to his neighbor" (0242). The example шабрансто resembles that offered by Agafonova above except that a possessive suffix is missing from the end. Thus there is reason to assume that it is a dialectical form which would be replaced by шабранстэнь in the literary language.

There seem to be variations in Abramov's writings with disparity of usage found with the genitive of the first and second persons singular versus the dative of the third person singular, i.e. use of the special genitive form of the second person singular is not found with the construction "with your mother-in-law" ававтот марто (0048), whereas he extends the third person possessive suffix to nonkinship, as well.

3. USAGE OF THE GENITIVE

According to GMYa (1980: 161-3) the functions of the genitive can be split into the categories (1-8). I have edited the quote below to include only the Erzyan portions of the examples, which I have then translated into English, for convenience. At times it would seem that the presence of the Mokshan examples might make for better illumination of the categories in question. The genitive is used to mark:

35

1) a person possessing something:

Чарькодьсы ли сон авань седеенть? (В. Коломасов) "Does he understand a mother's heart?"

Ансяк вана Микита тусь, ды свадьбанть ланга атянть весе кемеманзо ёмасть (А. Лукьянов) "Mikita had hardly left, and all the hopes the old man had of a wedding were lost."

2) possessors in relation to the possessed as whole in relation to a part:

Ho сёстэ плотинань основаниянть сави озавтомс витьстэ грунтонть лангс (А. Лукьянов) "Well then the base of the dam should be placed directly on the bottom."

There is a misprint here in the second word of the example that should be corrected to сестэ "then".

3) an object upon which the action of a verb is directed:

Агрономось икелеяк тесэ ламоксть ульнесь, но сестэ васняяк решасть войнадо мейлень задачатнень (А. Лукьянов). "The agronomist had been here several times before, but then first of all they solved the postwar problems."

4) an item which possesses some kind of characteristic or property:

Ней уш паксясто вармась канды тикшень ды васень цецянь чине (К.

Абрамов). "Now the wind already bore a smell of hay and the first flowers from the

field."

Ней сон ваны лангозонзо мельспаросо, кецязь никси кенерезь сюронь чиненть

(К. Абрамов). "Now he looks at it with joy and happily sniffs at the smell of ripe

grain."

5) a subject of an action:

Пилесэнь те шкас гайги эйдень авардемась (К. Абрамов). "The cry of a baby is still ringing in my ears."

Марят ансяк атякшонь морамо ды школанть вакссо, переменань шкасто, эйкакшонь шалнома (А. Лукьянов). "You only hear the call of a rooster and next to the school, during recess, the noise of children."

6) a person in some state:

Bece велесь марясь сех вадря пчеловодонть сэредемадо (А. Лукьянов) The entire village heard of the best beekeeper's illness.

Марюша седейсэ чарькодизе цёранзо ёжонть, мольсь эйзэнзэ, вадяшизе прява (К. Абрамов). "In her heart Maryusha understood the feelings of her son, she went up to him and patted him on the head."

36

In the first example of instance 6) the name of the source was incomplete and only the last two syllables of the Russian translation were present. Hence the translation is directly from the Erzyan.

7) a person or thing and his / its relationship to other things or persons:

Сон аламос чатьмонсь, мейле велявтсь станциянь начальникенть

помощникензэ ёнов (А. Лукьянов). "He was quiet for a little while, then he turned

toward the station master's assistant."

Сон нурькинестэ ёвтась, мезень кис пурнызь те открытой партийной

промксонть вейсэ колхозонь активенть марто... (А. Лукьянов). "He briefly

stated why they had called an open party meeting with the active members of the

kolkhoz."

8) division or distribution:

Удалце кудосонть эзть эря, голландкась пельскалозь, штукатуркась таркань-таркань певерсь (К. Абрамов). "They weren't living in the back room, the tiled stove was half collapsed and the plastering had fallen in places."

It can, of course, be debated whether all the categories given above are necessary for the description of the Erzyan genitive. For instance, is category 6) necessary at all? In considering the rule set forth in 6) we can, in fact, perceive that any reference to state or condition is held by the words сэредемадо "being sick (deverbal noun, Indef. Abl.)" and ёжонть "feeling(s) (Def. Gen.)". The former type is dealt with under subject of the intransitive verb in (3.4.1) or rule 3) above, and the latter in determiner of the object (3.1.1.2).

From the above categories I have been able to compile a list deviant in some basic features. This has entailed the combination of certain groups, as well as, the addition of new ones. The genitive can, for example, also be used in marking the valency for various postpositions, (see 3.5). It has bearing on the interpretation of infinite structures (3.4), and the use of definiteness can also show different frequencies with regard to grammatical function etc. The rules 1), 2), 4) and 7), which all are concerned with aspects of

37

possession, will be dealt with in 3.1 under possession and 3.2 under habitive. Rule 3) will be dealt with in (3.3), accompanied by information concerning ablative complements of the verb. Rule 5), which has only mentioned the subject of verbs here, will be expanded to greater dimensions involving nonfinite deverbals in (3.4). There it will be compared with the object and a division will be shown for transitive versus intransitive verbs. Finally, there will be a brief discussion of the problems involved in the workings of postpositions (3.5), followed by a slight expansion of rule 8) in (3.6).

Additional information concerning the usage of the indefinite genitive could also be of interest to the learner of the Erzyan language. The indefinite genitive, as seen above (2.1), can be affixed to a number of nouns and pronouns both definite and indefinite. In fact, if the word is definite by nature, the indefinite genitive is usally the best choice for expressing notions of possession and habitive, alike. One thing particular to the semantics, however, is that the presence of an indefinite genitive presupposes a specific person, place, or thing, (discussion with Feoktistov, Jan. 1996), (see 3.1).

3.1 THE GENITIVE AS A DETERMINER OF A NOUN PHRASE

Since the genitive can be used to indicate the possessive relation for different noun phrases, it is important that we differentiate between various parts of speech when looking for possible alternation between markings for possession. This will include establishing the existence of an alternation between definite and indefinite genitive formulas.

Second, it is important that a comparison between proper nouns and common nouns be made, so that the discussion in 3.1.2.1 below can more readily be given evidence concerning the actual usage of the genitive forms with regard to definiteness and specificness.

38

The corpus shows instances of the indefinite genitive followed by a possessee with no marking for possession or definiteness, and this occurs regardless of whether the determiner is animate or inanimate, see (41a-b). The possessee can, however, be modified by a definite article, see (42a-b). Here, as well, there is no difference between the properties of animate versus inanimate. In examples (43 a) we can find that a noun with temporal properties can appear in the definite genitive and have a possessive relation similar to that of an animate noun, cf 3.1.2.1 below.

(41) [A: indefinite genitive] + [B: neither definite article nor possessive suffix] (41a) (0152) а#содавикс цецянь чине "smell of an unknown flower"

(41b) (0165) Тона пельде марявсть тейтерень пейдема ды цёрань вайгель. "From the other side came a girl's laughter and a young man's voice."

[A: indefinite genitive] + [B: definite article] (42a) (0137) кардазонь тевтне "the work in the yard" (42b) (0237) тейтерень мазычись "the girl's beauty"

[A: definite genitive] + [B: possessive suffix] (43a) (0274) теленть якшамозо "the chill of winter" (43b) (0238) тейтеренть кедензэ "the girl's hand"

In examples (41-3) I have selected examples which show the combination formulas for possessees in the core functions, object and subject. This is the same criterion that will be used in the formulations of examples (53-5). One might ask whether the formulas are different for possession of the adjuncts. Formula (44) reveals something about the use of the indefinite genitive in relation to definite referents, such as those indicated by names of persons (44a-b) as well as places (44d) and interesting enough the inalienable "godmother of the groom", сваха (44c). For this reason formula (44) can be compared with formula (46), which has a definite referent indicated by a definite article (46a). In formula (45) we can observe the use of the indefinite genitive on the possessor, for both definite and indefinite referents, and the definite article marking the possessed. Surprising, however, is the use in formula (47), which marks both elements with the definite article. Could this be

39

a question concerning the semantics of the referent of the possessed — perhaps countable versus noncountable. Feoktistov said that there is reason to differentiate between the singular and the plural, that is something that would not be observable in the indefinite. In the entire novel "Йсяк якинь Найманов" I have found only one word marked for definiteness whose possessor was also marked for definiteness. Hence no conclusions can be drawn concerning this feature of the Erzyan language.

(44) [A: indefinite genitive] + [B: indefinite oblique case] (44a) (0198) Салдинэнь пирестэ "from Saldin's garden" (44b) (0093) Захгронь кирьгас "onto Zakhar's neck"

(44c) (0223) Свахань кирьгас "onto the neck of the groom's godmother" (44d) (0202) Явлеень вирьс "into the Yavlei forest"

(45) [A: indefinite genitive] + [B: definite oblique case]

(45a) (0005) асатыксэнь эрямостонть "from this life of shortages"

(45b) (0102) Явлеень паксястонть "in the Yavlei field"

(45c) (0152) Танянь кедьстэнть ды черьбулотнестэ мольсь а содавикс

цецянь чине. "There was the scent of an unknown flower coming from Tanya's

arms and braids."

[A: definite genitive] + [B: indefinite oblique case] (46a) (0223) айгоронть кирьгас "onto the stallion's neck"

[A: definite genitive] + [B: definite oblique case] (47a) (0008) Лелянть валтнэстэ "in his brother's words"

In example (47a) the possessor is being stressed, hence the underlining in the translation. Formulas (48-50) show part of the characteristics of possessive constructions in which the possessive relation is shown by a possessive suffix as opposed to a genitive ending. As discussed above in 2.2.1 there are discrepensies with regards to the functions a possessive ending can hold. In fact, in the corpus material only the possessive suffix of the third person singular, as suggested by Bubrikh above, is used for showing possession, see (48a-b, 49a, 50a).

(48) [A: possessive suffix (genitive)] + [B: indefinite oblique] (48a) (0261) тетянзо-аванзо кудосо "in her parents' home"

40

(48b) (0243) аванзо самодо "about her mother's arrival"

The use of the indefinite with the oblique cases in (48a-b) is interesting. The former possessee has a definite referent, whereas the latter might be the result of the verb form involved, see 3.4.1.5 below. On the other hand it might be reflective of definite referents as possessors in (44a-d) above. Examples (49a, 50a) seem to follow the pattern for definite referents as possessors, whereas the stress of the nominal phrase falls on the possessed.

[A: possessive suffix (genitive)] + [B: definite article] (49a) (0268) Аванзо валтнэ "his mother's words"

[A: possessive suffix (genitive)] + [B: definite oblique] (50a) (0182) Аванзо вайгельстэнть "in his mother's voice"

3.1.1 The genitive as a marker for possession

One of the main questions in Erzyan possession appears to be that brought forth in the three-way division of nominal phrases found in the language. Words can be formed by simply placing one after another without any case markings, i.e. the first element is in the absolute form. At the same time, noun phrases and adverbial phrases can be formed with, for instance, the genitive as an adhesive. Therefore the relationships of these various word and phrase constructions must be closely examined.

First we have the compound words that would appear to show a notion of possession. In fact, GMYa (1980:179-81) speaks of the first element as being in the function of the possessor: письмар#кудо "birdhouse", whereas письмар "starling" and кудо "house". Upon further examination, however, the compound words such as вить пелька#кенже "thumbnail" in the nominal phrase "right thumbnail" (0017), цеця#чине "fragrance (lit. flower scent)" (0070), and кардо#кенкш "barndoor" (0014) offer meanings with the

41

stress falling upon the latter element. The former element merely characterizes the latter, and at no time does it identify an existent referent, i.e. пелькакенже does not attest to the existence of a thumb, rather a nail. In the case of a "birdhouse", there is no presupposition of a bird, which would be obligatory if we were to speak of possession.

Second, there are the word groups containing a true sense of possession but at the same time a lack of definiteness in form. Here the first element must be considered existent, and the second element then becomes the possessed in much the same way as a part is possessed by a whole. This means that the constructions асодавикс цецянь чиие "scent of an unknown flower" (0152), скалонь одарстонть "from the (lit. a) cow's udder" (0037), and that of каршо кудонь вальматне "the windows of [the] house opposite" (0211) are deviant from those in the previous paragraph in that both elements are concrete. It will be noticed that in the case of possessors body parts and elements equivalent to them are not possessed, as it were, by a definite genitive. The word "definite" here is actually a misnomer. In truth the article involved is more of the deictic nature (cf. Mokshan: Feoktistov 1963: 205 Стлпа Ваня... versus Стяпать Ваняць...). Hence possession, although specific, can be shown by an indefinite genitive, cf. цецянь чине, скалонь одарстонть and кудонь вальматне above.

The long string of qualifiers found in example (41) shows how Abramov uses the specific, indefinite genitive. The discussion above would cause us to inquire about the genitive qualifiers ловсонь "of milk" and вельксэнь "of cream". Both the qualifiers ловсонь, вельксэнь and the possessed object тансть "taste, flavor" are subordinated to the possessor скал "cow", which in turn is further subordinated to the nominal phrase "of every home". If we were to propose the formula скалонть ловсо# ды велькс#танстезэ, we would be speaking of the "milk-like flavor of the cow". Hence in order to insure a sense "the taste of the milk of the cow", the indefinite genitive must be used, whereas the word flavor is first subordinated to the words milk and cream. It should be noted,

42

however, that the possessive suffix designates the cow as the possessor of flavor, whereas the immediate possessors of flavor, milk and cream, are overridden in their claim as possessor. The milk and cream, it will be noted, are also possessed by the cow. Nina Aasmäe would have preferred a construction with possessive suffixes instead of the indefinite genitive скалонть ловсонзо ды вельксэнзэ танстест. This also entails a switch of the possessive suffix to refer to the taste of the milk and cream and not that of the cow.

(51) (0039) Кемготово иесэ содылизе эрьва кудонь скалонть ловсонь ды вельксэнь танстензэ, лиякс меремс, ушодсь совсеме мацтка. "At the age of sixteen he knew the milk and cream taste of the cow in every home, in other words, he had started going into cellars."

If, on the other hand, we consider human, specific possessors of something concrete and yet being formed with the indefinite genitive, we will most likely be dealing with an indefinite possessor цёрань виев кедтне "the man's strong arms (lit. the strong arms of a man)" (0169), and ломанень нешкепире "a stranger's abiary" (0037). This however is not always the case. In fact, what we might actually be witnessing is a variation in the emphasis of possessor and possessed, see Feoktistov above.

With the names of people, i.e. in the case of given names, surnames as well as appellations of kinship preceded immediately by names, usage requires that the indefinite genitive be used. Unlike indefinite nouns, however, these definite referents can cause possessive suffixes to be affixed to the possessed, for instance Кандрань вайгелесь "Kandra's voice" (0303, 0304), where the word "voice" is modified by a definite article, but also Кандрань вайгелезэ "Kandra's voice" (0024), where the same word has been modified by a possessive suffix. Here it might be worth while looking at Feoktistov (1963: 205) where the operation of Mokshan might offer us a solution in the evaluation of the possessive ending.

43

It appears that Kandra's voice where the possessive suffix is used is new information, whereas Kandra himself is presupposed. This would also reflect the fact that no instances of unmarked possessees occur with personal names in the indefinite genitive, i.e. the possessive suffix is used to mark indefinite possessions, (oral information Feoktistov, Jan. 1996). There were no instances of personal names in the genitive followed by a possessee in the indefinite, nominative singular. On the other hand, the place name Naiman can take indefinite, nominative singular possessees, see (52, 53b). In the corpus, definite forms and forms affixed with a possessive suffix were used systematically when the possessed was in the role of subject. The only instances of absolute forms, as it were, are with a place name determiner, whereas the nominal phrase occurs either in the predicate (53b) or as an appositive (52). How would similar nominal phrases work with personal names, or is it that place names cannot have possessees affixed with possessive suffixes.

(52) (0022) Тон симат-ярсат, иайманонь эрзя, ... "You eat and drink, Erzyan of Naiman..."

Since proper nouns seem to be split in their nominal-phrase structures, such that personal names can take the formulas (54-5), and place names take the formulas (54-5) (see also (41) p 39), the question concerning common nouns arises.

(53) [A: indefinite genitive] + [B: definite article]

(53a) (0104) Захаронь самось сезизе Стяпанонь ёвтнеманть: ... "Zakhar's

arrival interrupted what Styapan was saying:..." (53b) (0019) Ларькань омбоце ёно озадо Найманонь попось, Гавря. "On the

other side of Larka sat the Naiman priest, Gavrya."

(54) [A: indefinite genitive] + [B: possessive suffix]

(54a) (0171) Захаронь ялганзо аволь стада ваныцят. "Zakhar's friends aren't herders."

(55) [A: indefinite genitive] + [B: no article]

(55a) (0291) Васянь a велявтсак, сон ней аволь Найманонь эриця. "You can't bring Vasya back, he isn't a resident of Naiman anymore."

44

Common nouns, on the one hand, can be used in the composition of compound words, whereas the first element, a determiner in the absolute form, can be directly followed by the head. On the other, they can be modified by the genitive endings both indefinite and definite. A question then arises as to which genitive forms are used with possessive suffixes and which ones are used with definite articles. Are there instances of possessees lacking both?

In the corpus material, as seen above, there are instances of absolute forms of nouns nonspecific followed by inflectible nouns specific, for instance пелькакенже, цецячине. The use of the indefinite genitive presupposes a notion of specificness and therefore it is first here that I shall begin treating the connective material required for notions of possession. It might be contested that there is no difference in the usage of indefinite genitive and absolute forms in the position of attribute. To be sure, MW (I: 186) gives two different forms for the concept "man's voice, male voice": [цёра-вайгель [E:?Mar]~ цёрань вайгель [E:?Kazl]], and conversation with Agafonova and Feoktistov has revealed that the two constructions are synonymous. Perhaps, the usage involves variation between the dialects. Agafonova did not accept the first formula цёравайгель as representative of Erzyan, but that could be a sign of modern usage.

3.1.1.1 The genitive can show possession of the subject

In examples (56—8) we can see how both personal and place names take the indefinite genitive when indicating the possessive relation (see 2.1.3, 3.1.2.1.1). In examples (56-63) the possessor can be marked by either the indefinite or definite genitive and the possessed subject can be marked by either a possessive suffix (see 56, 59, 60) or a definite article (see 57, 58, 62, 63), or it can be left unmarked (see 61). The a possessive suffix can also affixed to the possessor as shown in (64), i.e. here we have an instance of two

45

determiner elements in succession, see also (54a, 55a) above. It will further be noted that the position of determiner can be held by both animates and inanimates.

(0156) Захаронь весе чокшнензэ ютнесть - ликбезсэ ды ячейкасо. Zakhar spent all of his evenings - in the reading center and at the cell, (lit. All of Zakhar's evenings passed...)

(0063) Меельце валтнэде Захаронь чамась вельтявсь якстере тюссо, ды сон нолдынзе сельмензэ, мерят, кундызь саламсто, чатьмонезевсь. "At these last words Zakhar's face flushed red, and he lowered his eyes, as though he had been caught stealing, he fell silent."

(0019) Ларькань омбоце ёно озадо Найманонь попось, Гавря. "On the other side of Larka sat the Naiman priest, Gavrya."

(0274) Теленть якшамозо сивсь. "The chill of (the) winter was broken"

(0004) Кудонть потмозо свал сундерькс чопода. "The inside of the house [was] always dusk dark."

(0165) Тона пельде марявсть тейтерень пейдема ды цёрань вайгель. "On the other side she could hear a girl's laughter and a man's voice."

(0169) Таня эзь кенере фатямояк, кода цёрань виев кедтне кепедизь верев. "Tanya didn't realise how she had been picked up by the strong arms of a man."

(0003) 1922 иень тундось Найманс сась шкадонзо икеле. "The spring of 1922 came to Naiman ahead of time."

(0059) Еленань аванзо кудось аштесь сэтьме ульцясо. "Yelena's mother's house was on a quiet street."

The evidence for the possessive construction of nouns with spatio-temporal reference in (58-60, 63) can be applied in the argumentation against the claim of separate spatio-temporal adjective derivation, cf. 3.1.2.1 below.

In examples (65-6) we have instances of personal names indicating the possessee. Here the possessive relation can be shown by a possessive suffix as in example (65) or a genitive marking the possessor and a definite article affixed to the possessee as in example (66) can be used. Tsygankin maintains that all use of the definite article with personal

46

names is strictly a dialectical feature, related to contact with the Mokshan language, and that it is not part of the literary norm (oral information summer 1995). Nina Aasmäe on the other hand claimed that both the definite article and second person singular possessive suffix are used to a certain extent throughout the Erzyan dialects to express closeness of kinship, endearment etc. (discussion Feb., 1996).

(0268) — Зняро эзинь энялдо, Еленат эзь ушто. "-- No matter how many times I pleaded, your Yelena didn't heat [it]"

(0268) Банявтомо, келя, седе курок кулат. Вана мезть корты тонь Еленась!.. "They say you die quicker if you don't visit the bathhouse. That's what that Yelena of yours says."

3.1.1.2 The genitive can show possession of the object

As is the case of the subject, the possessor - determiner of the object can be marked by the indefinite or definite genitive or a possessive suffix. Likewise the possessed object can be marked by either a possessive suffix (67-8), a definite article, (69-70, 72-3), or it can be left unmarked (71). In (72-3) it can be seen that the indefinite genitive is used with place names when the possessed is marked by a definite article.

(67) (0247) Сынь a васенцеде истя кортнить чокшне ланга ды лецтнить се ломаненть лемензэ, кинь стака сэредемась валызе масторонть покш мелявкссо... "It wasn't the first time they spoke like this in the evening and mentioned the name of the man whose serious illness had filled the land with anxiety."

(68) (0005) Моданть пелешканзо миилизе кодамояк сюпав азорнэнь, кадовиксэнть сокавтылизе ды видевтилизе чачк. "About half of the land he had sold to some rich man, and the remainder he had had plowed and seeded for another half."

(69) (0174) Веть пракшнось кельме ды ульцянь рудазонть теизе калгодо каворекс. "During the night it had frozen and made the mud in the street into hard clods."

47

(0220) Лиза маринзе ялганзо вайгельтнень ды лисексэль ушов. "Liza heard her friends' voices and wanted to go outside."

(0276) — Велень ярмак симить, -- чова вайгельсэ цийнесь каршонзо Ларька. "— They're drinking the village's money, whined Larka's high voice in reply."

(0096) Улиндеряй кавтолдома монь ули-паронть коряс кевкстинк Найманонь эрицятнень, сынь ёвтасызь. "Should there be any doubt about my wealth, ask the people of Naiman, they'll tell [you], (lit. they'll tell it)."

(0126) Захаронь седеенть, мерят сезнесть кискат. "It was as though Zakhar's heart was being torn apart by hounds."

The referent of велень in example (71) could — in context — actually be synonymous with Найман. Hence the construction would be one of an emphasis on the possessor. There are, of course, other instances of the word велень that have indefinite referents but that is more a matter of marking definiteness.

Examples (74-5) show that it is possible to separate the possessed object from its qualifier by both additional qualifiers, as seen in example (74), and the predicate itself, see (75). In this latter example, change in word order is associated with the pragmatics of topicalization. Hence the object пирензэ "her garden" has undergone topicalization, (discussion with Feoktistov, Jan. 1996).

(0039) Кемготово иесэ содылизе эрьва кудонь скалонть ловсонь ды вельксэнь таистензэ, лиякс меремс, ушодсь совсеме мацтка. "At the age of sixteen he knew the milk and cream taste of the cow in every home, in other words, he had started going into cellars."

(0046) — Мезе теньсэ, аванть сокия пирензэ? "What of it, it's the lady's garden I plowed."

48

3.1.1.3 The genitive can show possession of an adverbial phrase

The possessive construction used with adverbials differs from that of the one used with the subject and object. This difference can be seen in the use of the definite article or possessive suffix with subjects and objects and the redistribution of their use with oblique cases, (see 41-50). There even seems to be a difference is usage compared with the government of postpositions, namely the genitive used to show the government for postpositions in place names is definite, see 3.5 below.

Example (77a) gives us further reason to contemplate (3.1.2.1), is the fact that a quant or for that matter any attribute present sufficient evidence to overrule the notion that the form туидонь is an adjective. It might not overrule this conception, but it would, at least, show the existence of homonymy with an indefinite genitive. In examples (81-2) we can see how personal and place names take the indefinite genitive as possessor of adverbials (cf. 50b-c), whereas all noun types can have this pattern, indefinite genitive possessor plus definite possessee, see also (50a). In distinction from place names the personal name, as found in example (81), will retain the same formula with the possessor marked by an indefinite genitive in the government of postpositions, as well (see 3.5). Place names take definite genitives for the government of postpositions.

(76a) (0014) Пахом чатьмонезевсь, вансь братонть чамас ды турвасо чевилясь эчке цигарка пенть. "Pakhom fell silent, he looked into his brother's face and chawed away on the butt of a thick cigar."

(76b) (0091) Захар саиксэлизе тейтеренть кедьстэ кеченть, сон эзизе максо. ёртызе ды каявсь Захаронь кутморямо. "Zakhar wanted to take the dipper out of the girl's hand, she didn't give it [to him], she threw it and overwhelmed Zakhar with embraces."

49

(77a) (0045) Ваннызе ниле тундонь ведьсэ салсезь сёлтонть, ловизе зяро ломань сави сиведемс сёлтамганзо. "He examined the pond which had been eroded by the water of four spring floods. He counted how many people he would have to hire in order to dam it."

(77b) (0113) — Зярдо ульнесь: тейтерень кудосто казнень рамсемс ярмак вешевельть? "-- When have you ever heard of money for buying presents being asked for in the girls home? (lit. when was it that money for buying presents would be asked for in a bride's home?)"

(0274) Наста эзь тонадо нееме мирдензэ чамасто арсема мелензэ ды потмо ёжонзо, секс нейгак мезеяк эзизе тюнгольдсте. "Nasta wasn't used to reading (lit. seeing) her husband's thoughts and feelings on his face, that's why nothing surprised [her] now either."

(0008) Лелянть валтнэстэ сонензэ марявсь пейтькшнема ды згилямо, мезеяк эзь мере тень каршо, згилямонзо кадызе апак пандо. "In his brother's words he could sense mirth and irony, he said nothing in reply and left his brother's joke without retort."

(0182) Зярс удось a помнясы, сыргойсь аванзо вайгельденть. "How long he had been sleeping he didn't remember, he was woken by his mother's voice."

(0154) Захаронь тонавтнемантень топодсть кавто ковт. "Zakhar had been studying for two months, (lit. Zachar's studying had filled two months)."

(0205) Захар a курок вейсэндявсь Найманонь эрямонтень. "Zakhar joined the life of Naiman after a while."

3.1.2 Nonpossessive use of the genitive

With regard to the discussion begun in 3.1 as well as GMYa (1980) rules 2), 4) and 7), there is reason to consider relations other than possessive ones when dealing with the genitive marked attribute of a noun. GMYa (1980:179) introduces the idea of parts of an organism or body мако#лопакс "like a poppy leaf' or part of an item тумо#лопа "oak leaf', to be sure пей#сывепь "gums (lit. tooth meat)" has little to do with parts of a whole. These could better be characterized as specific elements qualified by a nonspecific element.

50

To refresh our memory let us look at some examples from Abramov's text: яксяргонь пуло "a duck's tail" (0021) can be contrasted with мацей#толгат "goose feathers" (0078). In examples (83—4) we can see instances of compound words with attributes, both seem to be attributes of the compound word (cf. 3.5.1). Example (84) supplies us with evidence which might be used in further contemplation of word structures, i.e. (0042) сэрей каль прятненень "into the high willow tops" cf. сэрей калень пряс "into the top of a tall willow". In example (84), it would be tempting for the English speaker to assume the word турва as the head were it not for the possessive suffix on the second element вельксэзэ, but that is a characteristic of English and not the language of study.

(0298) Потмура варштавкссо варштызе кельме кудо потионть ды таго авардезевсь. "She gloomily looked at the cold room and burst into tears again, (lit. With a gloomy look she looked...)"

(0007) Верце турва вельксэзэ кармась чополгадомо, сон те шкас а соды бритвань ёжо. "His upper lip was starting to turn dark, it still hadn't seen the edge of a razor (lit. The top side of his upper lip was starting to turn dark, to this time it hadn't known a razor's edge.)"

3.1.2.1 Special instances of the so-called genitive adjective

According to GMYa (1962:189-92) relative adjectives indicate qualities of items with relation to place велень эриця "villager, village dweller", material of origin чувтонь кудо "wooden house", and time тундонь чи "spring day". One of the elements for derivation is -нь. At first glance, one might consider this a homonym of the indefinite genitive dealt with earlier in 2.1.1.1 and 3.1. In fact, the functions of this ending are extremely close to those of the English genitive. In English for instance the genitive can be used both temporally a long winter's night and with reference to material of origin a house (made) of wood.

51

The grammar, however, leads one to assume that certain classes of nouns can be used to derive nonpossessive determiners, which in many cases can be viewed as equal to adjective attributes. At least, they are readily translated into Russian as relative adjectives. These classes of nouns include ones with meanings of spatial (3.1.2.1.1) and temporal (3.1.2.1.2) magnitudes as well as ones denoting material of origin (3.1.2.1.3), (see GMYa 1980:112-3, GMYa 1962:189-96). GMYa (1962:195-6) speaks of a transition from relative adjective to adjective of quality, as well. I have addressed derivations with -нь, denoting "characteristic of under possession (3.1), further consideration will be given below.

According to GMYa (1962:192-3) a small number of relative adjectives can be derived from spatio-temporal adverbs with the ending -нь. The examples, given to better illustrate the situation, include other adverbs, as well. For instance алкуксонь шофер "real chauffeur" has no actual relation to temporal or spatial but maybe the Russian настоящий does. On the other hand, this does show us that the number of derivates might actually be much greater.

In addition, GMYa (1962:193-4) shows that a past participle form in -нь is also present, for instance чапавтонь ловсо "sour milk (lit. soured milk)", сиыень тейтерь "betrothed girl (lit. girl who has been drunk off/away)". In addition to this infrequent past participle, which appears to be limited to the transitive verbs, there is the negative gerund with апак. It can be used for deriving adjective attributes with the same -нь ending, which still could be called the indefinite genitive. In Erzyan there are other instances of two case endings being placed one after the other, see конаньсэ in 2.1.2.1 above. According to Riho Griinthal this tendency of stringing case endings together is found in other languages of the Volga region, as well (private conversation Jan. 1996).

52

In Abromov's language the -нь derivations should be considered different from adjectives in that they do not take signs of the plural when they are part of the predicate. This is true of the Erzyan written language, such that the when the adjective is the predicative, it agrees with its subject for number, for instance кудотне пижеть "the houses are green" but кудотне чувтонь "the houses are wooden".

3.1.2.1.1 Is the indefinite genitive a derivational suffix for spatial attributes

Places denoted by ало ульця "low street", веле "village", ош "town" etc. can, by taking an indefinite genitive, provide a gloss giving the spatial origin of the "possessed". The confusion — if any — that results may stem from the fact that an indefinite genitive can by nature refer to both Singulars and Plurals. Hence whereas ало ульцянь in (85a) is, by context, translated "from the low street", it could actually be given a different translation in a different context, for instance "from a low street", whereas we would be dealing with general or for that matter plural concepts. Hence (85a-86b) could be compared with a place name such as "Naiman", whereas (85a-b) could be placed together with examples from (3.1.1.1,3.1.1.2), determiners of the subject (see 58) or object (71).

(85a) (0045) Ванстыцякс сиведекшнесь ало ульцянь атя, конань усад песэ ашти чёскась. "As a guard he had hired an old man from the low street, at the end of whose garden the carding mill was standing." (85b) (0039) Bece кудонь тевтне ды паксянсетнеяк кирдевсть нинзэ, Палай бабань, вийсэ. "All of the chores around the house and even those in the field were managed by his wife, granny Palai."

On the other hand, (86a-b) are examples of indefinite-genitive attributes with a possessive relation. Example (86a) should be dealt with in a treaties of definiteness versus indefiniteness, because of the uniqueness of велень "village" and its synonymy with Naiman, the name of the village. The phrase велень совет should be compared with

53

вельсовет (Zhuravlev, 1993:77). This comparison would require us to reclassify example (86a), it should be an example of possession (3.1.1). On the other hand the form вельсовет might be a reflection of the Russian, сельсовет.

(86a) (0298) Кaлмосъ чувозель велень Советэнь кудонть икелев. "The grave was dug in front of the village's Soviet building."

(86b) (0235) Ошонь конёвонть марто Миколь чийнесь веленть келес ды шнась прянзо тонавтнеме молемасо. "With the paper from town in hand Mikol ran about the village and bragged of going to study."

Although example (86a) can be compared with the possessive construction, which it most likely represents, example (86b) shows an instance of indefinite but specific reference. Furthermore, I have been unable to find instances in the corpus, other than the вельсовет above, that advocate a spatial attribute without a genitive ending. In MW (111:1472) there are forms with the absolute form of ош "town" [ошбазар "city market", ошэри "urbanite"] as well as ошсо эриця "someone living in the city".

3.1.2.1.2 The indefinite genitive a derivational suffix for temporal attributes

As was the case with spatial nouns temporal ones can, with the indefinite genitive, provide a derivation denoting the temporal origin of the "possessed". Indeed, by temporal and spatial origin we know that the possessed can — but does not necessarily have to — be located in its point of origin. The inessive in present spatio-temporal usage, for instance Римсэ папась "The Pope in Rome" and ниленьгемень иесэ цёра "a forty-year-old man" versus Римень цифрат"Roman numerals" and 1922 иень тундось "the spring of 1922", shows the temporary situation involved in the usage of the inessive as opposed to that of the genitive (discussion with M. Mosin Dec. 1995).

54

It would be interesting from the aspect of possession to see whether example (88) could be rewritten on the basis of what we have learned, i.e. could it be written апрель ковонь чизэ or апрель ковонть чизэ. From a pragmatic point of view these hypothetical formulations, with the possessive suffix, would not be possible, i.e. the emphasis falls upon the spacial adverb "outside", and this leaves no reason to expect an additional — simultaneous — emphasis on the temporal attribute. On the other hand, GMYa (1980:185) maintains that the III Sg. Px is used with natural phenomena and expressions of time, so one might ask whether it would be possible to use the possessive suffix on the word for "sun" when there are attributes present. In examples (87a-b, 89) the use of a possessive suffix would be impossible because of the indefiniteness of the referent. Example (88), on the other hand, is open to pragmatic alternation that could involve the use of a possessive suffix, whereas the compound word апрель ковонь "April (Absolute) + month (Indef. Gen.)" represents the same [proper noun] + [appellation] formula found in Олдай баба "Oldai (Absolute) + grandmother (Indef. Nom.)", cf. GMYa (1980:181). the lack of further determiners leaves the door open for thematic alternation.

(87a) (0200) Тундонь васень читнестэ кармасть лембе мастор ёндо ливтямо нармунтне, Захаронь седееськак кармась тердеме Найманов. "In the first days of spring the birds began flying from the south, Zakhar's heart, also, began calling for Naiman."

(87b) (0248) Телень нурька читне ютасть вейке омбоце мельга. "The short days of winter passed one after the other."

(0010) Ульцясо апрель ковонь чись кумболдсь ведь#пурнавкстнэсэ. "In the street the April sun (lit. the sun of April) sparkled in the puddles."

(0070) Валскень экше коштось ванькс, эйсэнзэ качады цеця#чине. "The cool morning air (lit. the cool air of morning) is clean, it bears a fragrance."

55

3.1.2.1.3 The indefinite genitive indicating material source

Unlike the preceding usages of the genitive, the indefinite genitive derivation of material nouns does not have an inessive counterpart showing present state. Instead, there seems to be vacillation between indefinite nominative singular and indefinite genitive, whereas the latter sometimes appears to emphasize the material of origin.

According to GMYa (1962:194) absolute forms can be used as relative adjectives, for instance сывель#ям "stew (lit. meat soup)". Of the examples given лов#шушмо-... "snowdrift" draws my attention to examples of this very construction in which the first element of the compound word is an uncountable and the latter an expression of shape/form or measurement, cf. GMYa (1980:180-1). Abramov's text has plenty of examples for this formula, whereas the first element is an uncountable and the second one describes an actual form or limitation/measure, e.g. конёв#панкске "slip of paper", сал#поколь "piece of salt", менель#панкс "patch of sky", but also measurement/purpose медь#парь "tub of honey; honey tub", фасоль#мешок "sack of beans; beansack", etc.

In examples (90a-b) the use of the word сёвонь "clay" might indicate a difference, whereas in (90a) we would be speaking of the form knoll, and in (90b) we would be emphasizing the composition of the mound. Example (91) demonstrates that the genitive origin can be used in an attribute of character which, in turn, allows for a transformation such as "silk-like", see also GMYa (1962:196).

(90a) (0063) Церькованть эйстэ a васоло тюжа сёвонь марокс неявсь Осип Ортянь кирьпецень кудось. "Not far from the church was Osip Orta's brick house which looked like a ruddy clay knoll."

56

(90b) (0298) Модась вана уш пачкодсь калмо чирентень. шкань ютазь калмонть вельксс кепететсь тюжа сёвонень бугорькске... "The earth had already reached the edges of the grave, with time a little mound [made] of ruddy clay rose over the grave."

(91) (0212) Мерят, a умок панштнесть умаринатне, вешксть письмартнэ, Вишкалей чирева пижелдсь парсеень нарось. "It was as though not long ago the apple trees had bloomed, the starlings had whistled, and the silky glade along the edge of Vishkalei had glistened green."

In Abramov's texts we find one of two variables used in modern Erzyan, as exemplified by the term скалонь стада "cattle herd" (92). This term is in fact a concept, whereas it can possibly include other elements in its referent, i.e. скалонь стада can include the herders and dogs etc. On the other hand, a compound word like скал#стада "herd of cows" can only mean "a herd [the measurement] + cows [consistency]". My information is based upon discussion with N. Adushkina in Saransk (summer 1995), cf. солдат#полк "regiment of soldiers" and солдатонь полк "regiment [including officers etc.]" information contained in (MW IV). Perhaps the first component here, a countable has a different effect on the interpretation of the genitive.

(92) (0274) — Тарказо скалонь стада удало... "— His place is behind a cattle herd..." (i.e. not as head of the village Soviet).

3.1.2.1.4 Genitive for constructions expressing quantity

Close to the concept of material of origin for noncountables is the one of quantity shared by countable referents. In examples (93—4) the genitive attribute preceding the head expresses quantity. Hence вете#ашонь лопа "five-rouble note" entails a note whose value is that of five roubles. Likewise, example (94) speaks of their land whose extent is that of seven yedoks. Apparently the first element must be a quant, and therefore

57

the resulting construction is equal to that found in example (82) ало ульцянь атя "a man from the low street".

(MW II: 1055) [мейле таргась вете#ашонь лопа.] Darauf zog er einen Fiinfrubelschein hervor.

(0003) Арасть видьместкак, арась кемемаяк, сокаволь ды видевель сисем#едаконь модаст. "They have neither seed nor even faith, yet their seven 'yedoks' of land ought to be plowed and sowed."

3.1.2.2 Adjectives formed from adverbs

The property of derivation found in the adverbs васолдо "from afar" > васолдонь "from afar [attr.]" and течи "today" > течинь "today's", GMYa (1980:112-3), can be applied to adverbial phrases, as well. During my discussions with N. Adushkina in the summer of 1995, we arrived at a pattern that illustrates the application of the -нь affix to adverbials, (see 95-7). In (95) we have the simple existential clause stating the location of a tree. Instances (96-7) appear to be almost identical in form, whereas the meaning of (96) is one defined as present - temporary. On the other hand, (97), a hypothetical form, appears to indicate the tree that belongs next to the house is a birch.

Кудо вакссо чувто. "Next to the house is a tree."

Кудо вакссо чувтось килей. "The tree next to the house is a birch."

Кудо вакссонь чувтось килей. "The tree for keepings next to the house is a birch."

In examples (98-9) two different forms of the word шкава, which is in the prolative, illustrate how Abramov varies in his usage. Had a genitive form been used in example (99), an emphasis would have been placed on the attribute монь "my", cf. (98). Example (100) shows a determiner derived from the adverb чияк "recently, within a day's time", which in the written language appears to have a varient in чияконь. Going further

58

(101) provides us with an example of derivation from the ablative, and (102) has a interrogative pronoun, in the elative, for its base, see also GMYa (1962:192-3).

(0115) Тесэ ульнесть Озяз Иван, Сергей Андреевичень тейтерезэ Лиза ды знярыя истят жо, сынст шкавань, од ломанть. "Here were Ivan Ozyaz, Sergei Andreyevich's daughter Liza and a few such young people, of their age."

(0050) Монь шкава авантень налксемс ды пейтькшнемс... "At my age a woman needs to laugh and play." (Actually, (play and laugh) Perhaps, word order follows the same principle as that in English, i.e. the shorter comes first.)

(0075) Лия мезеяк эзь са прязонзо, ды сон таго лецтизе чиень кулянть: "Nothing else came into his head, so he rementioned the recent news..."

(0267) Мартонь васень читне кандсть лембе ёндонь летьке кошт. "The first days of March brought in moist southern air."

(0164) Валоськак кие соды костонь, паро ладсо а ёвтавияк. "And the word itself, who knows where it comes from, it's hard to even say it properly."

3.1.2.3. Adjectives formed from the negative gerund апак

In my labeling of gerund versus infinitive, I have chosen to employ the unorthodox usage name for function. This entails differentiating between gerunds and infinitives, whereas the major functions of the gerund are qualifier of the entire sentence and qualifier of the verb. The presence of the indefinite genitive in deriving attributes from adverbials can be seen to show a difference of emphasis in instances where the adverb is preceded by a genitive attribute of its own, see above. This usage has its parallels in the formation of nominal phrases including nouns of spatio-temporal reference. Likewise, we can expect the usage of the genitive in the derivation of attributes from verb forms.

Examples (103-4) are deverbal forms that are derived from the negative participle/gerund forms апак + [conneg] : апак тонадо "uncustomarily" and апак учо "unexpectedly". In GMYa (1962:354) the two examples given for Erzyan апак вадне кемтне "the ungreased boots" and апак учонь инжеденть "because of the unexpected guest", show that not only is the plain participle/gerund form апак вадне possible but

59

presumably the compound апак учо, as well. On the other hand, Abramov sets the forms апак учо "unexpectedly" and the dirived апак учонь "unexpected" apart.

(0242) Тевесь ульнесь вельть од, апак тонадонь. "The concept was quite new, uncostomary."

(0203) Захаронь апак учонь самось абунгавтынзе весемень. "Zakhar's unexpected arrival startled everyone."

In examples (105-6) we are presented with the use of the pair anак содань versus апак сода. One might ask whether my translation of (106) is the only possible one, could the form апак сода have a participle use, as well. Here, however, I will ask whether there is actually a negative participle of the апак сода formula in Erzyan. The text corpus would speak against such a hypothesis.

(0006) Покшнек-вишкиненек сельмень цитнезь ванкшность ге апак содань ярсавксонть лангс. "Both big and small looked at this unfamiliar foodstuff with eyes aglow."

(0243) Тонсь арсик, кода молян апак сода ломаннень. "Now think yourself, how am I to visit strangers."

3.1.2.4 The -нь participle, is it an old genitive of the verb stem?

There is a passive past participle -нь in Erzyan, which is synonymous with the -зь participle. The same form, it seems, can also appear in Erzyan folklore in contexts that lack the notion of past participle, for instance вечкень ава "beloved mother", сравтонь черть "with hair in tangles". (GMYa 1980:353).

According to Ledyaikina (1980:85), this participle can only stand in the position of qualifier preceding the head. Bubrikh (1947:43), on the other hand, gives the constructions сынь кундазь and сынь кундань "они схваченные" as equivalents. With regards to Ledyaikina's claim, the low frequency of the -нь participle can be more readily

60

understood when compared to the more common -зь perfect participle, which can play the role of predicative, as well.

In MW (П: 1528) three examples are given for панжонь versus панжозь all three with the definition "open(ed)". Here, there is a difference observable between the use of the -нь form and that of the -зь.

(107a) [кучумак — пекстань ортань панжицякс, панжонь ортань пекстыцякс] E:Mar "Send me — to be the opener of closed gates and the closer of opened gates."

(107b) [ортазо панжозь келейстэ] E:Mar "Her gates [were] wide open." (107c) [панжозь вальма#чирьксынем ] E:VVr "My window [was] slightly open."

In example (107a) the -нь form is used as though it were a genitive attribute, whereas examples (107b-c) show the -зь form as the predicative. It should be noted that in (107c) the same word order could be used with the -зь form as an attribute. This differentiation can be observed in the pair: пурнань versus пурназь "gathered, set" (MW III: 1644), but not for example in the pair мадень versus мадезь "?lying" (MW II: 1156). Could this be a difference between dialects. Abramov uses the -нь in the position of attribute (108—9), whereas the -зь form is used as both an attribute and a predicative.

(0007) — Ярмаком эзь кадово синдень трешникак, — мерсь каршонзо Пахом. "--1 don't even have a broken copeck, said Pakhom in reply."

(0235) Ноября ковонь меельце читнестэ губерниянь центрасто сась кувать учонь конёвось. "In the last days of November, the long awaited 'paper' arrived from the provential center."

According to Nina Aasmäe there is a differentiation between the use of the -нь and -зь forms. She sees the -нь form as one which expresses something generic, whereas the -зь form is very concrete. Hence the difference could be seen on much the same lines as those found between adverbials and their genitive counterparts in 3.1.2.2. For example, the use

61

in сон ярсы чапавтонь капстадо "she eats sauerkraut (generalization)" is different from that found in сон ярсы чапавтозь капстадо "she is eating soured cabbage (cabbage which has been soured, specific)."

3.2 THE GENITIVE IN THE HABEO AND LOCATIVE CONSTRUCTIONS

In Erzyan there areinstances of the genitive, which can readily be interpreted to indicate the habeo or locative construction. In many of the neighboring, and related languages of Ezyan there are no habeo-verbs. In Erzyan the habeo-construction is represented by a usage of the genitive that can overlap with that of the possessive construction, see (56, 59, 60, 67-8, 74-5) above. Likewise, there are locative constructions — especially ones formed with collectives which apply the genitive to a locative context, see 3.2.2 below.

3.2.1 The genitive in the habeo-construction

The possessive suffix plays an important role in the habeo-construction. In fact, the presence of the possessive suffix appears to be mandatory for this construction. On the other hand, its mere presence is not sufficient to render the habitive interpretation, i.e. a copula is often needed to reduce the number of interpretations to the habitive. In examples (110-2) we encounter two different copulas, the positive preterite in (110) and the negative present in (111-2). Furthermore, alternation of word order, in the habeo construction, is acceptable as well as the addition of further adverbials, for instance locatives, (see 112).

(110) (0004) ... хоть Стяпанонь ульнесть низэ ды эйкакшонзо, Пахомт-Захарт — ськамостольть. "although Stapan had a wife and children, Pakhom and Zakhar were single."

62

(0033) Кинь арасть эсензэ кудонзо-чинзэ, лишиезэ, се синтри уцянзо ломань кардайсэ. "He, who has no home or horse of his own, breaks his back in someone else's yard."

(0017) Малавикс роднянзо Наймансо Кандрань арасть. Низэ, Елена, саевть ошсто. "Kandra doesn't have any close relatives in Naiman. As for his wife, Yelena, she had been taken from the city." (For more information concerning the peculiar usage of the agentive participle in (112) see 3.4.2.1 below.)

Example (113) shows that the possessive suffix is even used with indefinite habeo-subjects — something which does not seem to occur in instances of simple possession.

(113) (0265) Мезеяк теть эзь тееве, авань ды каткань сисем ойиензэ. "Nothing happened to you, a woman and a cat have seven lives."

3.2.2 Use of the genitive in locative constructions

In deviance from the habeo-construction above the possessive suffix does not appear to be a required element. Something more characteristic of the locative construction might be the use of collectives, such as the plural pronouns, and surnames. The genitive form минек "our" is used in example (114) to express the habeo agentive or relative location. In a similar manner the surnames Канаев (115) and Гаруз (116) are used, with the notion of collective, to denote a relative location. The sentences have verbs in the third person plural, which would be similar to a Finnish usage where a locative is combined with a passive verb, cf. (115) Kanajeveilla tiedettiin...

(0006) Мон потмозонзо кунцевлинь таракант, каштом удало минек вельть ламо. "I'd catch cockroaches in it, we have a lot [of them] behind the stove."

(0265) Нама, весенень ёвтнемс теде а мейс, Канаевнень снартсь ёвтамо. Цы эзь кенере — Марюша икельдизе. Канаевень содасть, мезе теевсь Лабырень кудосо. "Of course, there's no reason to tell everyone about it;

63

she intended to tell Kanayev. But she didn't manage it in time ... Maryusha beat her to it. At Kanayev's they knew what had happened at Labyr's house." (116) (0004) Гарузонь тезэньвесе промкшныть ансяк телень якшамотнестэ, ды сестэ Бишка кудыненть потсо эрси а кува ютамс, а кува велявтомс. "At the Garuzes everyone gathers here only during the cold of winter, and then there is nowhere to walk or turn around in the little house."

3.3 THE GENITIVE AS THE SIGN OF THE OBJECT

The genitive is one of the many signs of the object in Erzyan (Alhoniemi, 1994: 21), see also 3.4.3.5 below. This sign of the object is used with the definite and indefinite conjugation alike, whereas use of the indefinite conjugation shows an incompleted action or action with no change of state in the patient, see (123—4) below. On the other hand, there are instances of genitive objects which do not refer to the patient, such as the source - indirect object of the English "ask", see the Erzyan кевкстемс in example (122). Another interesting feature revealed in the objects is the use of the definite genitive with place names versus the indefinite genitive with personal names, cf. (122-3).

Certain verbs take not only an object but can take an additional adverbial to specify the point of contact in kicking, which in Erzyan is expressed with the illative. Another case used for this type of construction is the ablative, which is used to show the part of the object by which it is led, held, etc. This use of adverbials should be considered when dealing with infinitives, as well.

The definite genitive, as seen in (117-22), can mark a definite object be it concrete, abstract, animate (121-2) or inanimate (117-120). Example (117) shows a part of the body as an object, whereas example (119) shows the deverbal noun in a similar construction.

64

(117) (0144) Захар мезеяк эзь мере, лавтовонь лыкавтозь айгизе Миколень кеденть ды а мельсэ варштась оев сельмезэнзэ. "Zakhar said nothing, with the shrug of a shoulder he moved Mikol's hand and looked into his (Mikol's) oily eyes with disgust."

(118) (0021) Теде мейле сон апак корта каподизе икелензэ аштиця стопканть ды, лиянь апак учо, комавтызе кургозонзо. "After this, without saying a thing, he snatched the glass standing in front of him and, without waiting for the others, he tipped it into his mouth."

(0023) Попось маризе Кандрань тошкамонть ды серьгедсь: ... "The preacher heard what Kandra was whispering and cried out:..."

(0276) — Нуртик кадовиксэнть, кум, симсынек ды тутано, — пшкадсь те шкас чатьмониця Кандра. "— Pour out what's left, my friend (lit. (god parent of my child) but used thoughout the text as a word of address.), we'll drink it and go, — said Kandra who until now had been quiet."

(0278) Миколь кежень сазь чавсь лишменть, ялатеке те эзь лезда. "Mikol furious hit the horse, but that didn't help"

(122) (0278) Елена саинзе ождятнень, пульзясь ды ушодсь лишменть панеме. "Yelena took the reins, knelt down and began urging the horse on."

The conjugation form of the verb with which the genitive object is combined, in many instances, gives a sense of aspect. For instance, in (121) the horse is being hit but no change of state is taking place, and in example (122) Yelena is beginning to urge the horse on. Definite objects in combination with indefinite conjugations, however, cannot always be interpreted solely on the basis of incompleted action or action without result. In (123) we have an instance of information being related for the second time in the text, (cf. 124). It would be interesting to know whether the difference in word order or pragmatic difference in these sentences could be used to account for the variation of the definite and indefinite conjugation.

65

In (125) Zakhar is pensive, and perhaps emphasis falls on the verb, i.e. there is a stress place upon the means as opposed to the object and result. In contrast, example (126) shows a simple introduction of facts with a result clearly shown.

(123) (0010) Пахомгак вана: комсь вайгельбеть уцянзо лангсо кандсь фасоль мешоконть, бажась андомс вачодо эль кулыця семиянть. "Why look at Pakhom: it was twenty versts that he carried that bag of beans on his back, he wanted to feed the nearly starving family."

(124) (0006) Фасолесь колмошка пондоль, ды неть колмо пондтнэнь Пахом кандынзе уцянзо лангсо Явлеень станциясто, комсь вайгельбе. "There were about three poods of beans, and those three poods Pakhom carried on his back from the Yavlei station - twenty versts"

|

In examples (125-6) we can see the difference between personal and place names. The place name in (126) shows the usage of the definite genitive for the role of object. This type of usage was not present in the possessive constructions, but maybe there were simply no instances of emphasis placed on the toponym, cf. 3.1.1.

(125) (0212) Таня кармась кевкстнеме Захаронь, косо ульнесь, мезе роботась. "Tanya started asking Zakhar where he had been and what he had done."

(126) (0238) -- Сёрмадт, Захар, иляк стувтне Найманонть, — пшкадсь Таня. "- Write, Zakhar, don't forget Naiman, said Tanya."

Examples (127-9) show that in addition to simple definite objects there are constructions containing complement adverbials, which specify the contact surface or point at which the patient is affected by the action of the verb. In examples (127-8) a point of contact with a concrete entity is being expressed, whereas example (129) can be seen as an application of the pattern found in example (127).

66

The -мадо, -меде, -модо form represent by чувномодо "a-digging" in example (129) is presented in GMYa (1980:271) as an infinitive, see also Bartens (1979:78-9) where the infinitive is shown as a complement. Hence (127) and (129) are similar in their ablative forms. I have shown three different forms for the ablative infinitive above "-мадо, -меде, -модо" instead of the two usually considered "-меде, -модо", because Abramov appears to make no difference in their usage. If one of usages of the ablative deverbal noun is an adverbial expressing state, then perhaps there are even fewer reasons to consider the -мадо, -меде, -модо forms separately in a grammatical presentation of the Erzyan infinites in Abramov's texts (see 3.4.3.7).

In example (128), on the other hand, we find the surface where the object is affected by the action of the verb. There might be parallels drawn between this illative usage and that found in -Mc infinitive, i.e. the illative can be used both temporally and for the expression of goal.

(0035) Тейтересь ноцковтызе Миколень ожадо ды таго энялдозевсь:... "The girl tugged Mikol by the sleeve and pleaded again..."

(0045) Марюша кутмордызе цёрынензэ ды палызе коняс... "Marusha gave her son a hug and kissed him on the forehead..."

(129) (0313) Удалце пирявтонть трокс ютамсто умарь#пирестэ неизе чувномодо Оксянь. "When he came over the back fence, he saw Oksya digging in the garden."

3.4 THE GENITIVE AS DETERMINER OF NONFINITE VERB FORMS

The genitive and consequently the possessive suffixes can be used to show the agent and patient roles with various verb forms including gerunds, participles and deverbal nouns. According to Bartens (1979:45-6) it is typical of Mordvinian infinitives that possessive suffixes affixed to them give reference to the patient, object. She also states that the ability to take a possessive suffix denoting agent is proof enough to show that a

67

verb form is not an infintive. The genitive and possessive suffixes can be used with gerunds, such as the temporal elatives -мсто/-мстэ to denote the agent.

In the corpus, however, we are confronted with different evidence for Abramov's language. Sometimes, especially in the case of deverbal nouns, including those referred to as -модо/-меде infinitives above, only the context can tell what role the genitive — and consequently the possessive suffix denote. In these cases the genitive can, in fact, show the roles "subject of the intransitive verb", "subject of the transitive verb" and "object of the transitive verb".

As for the role "subject of the transitive verb" there might be a sense of pseudo-intransitive verb present, but I have been unable to find an instance where both the patient and the agentive are shown simultaneously (see 130). Instead, there are other formulas available, which reduce the components in (130) by one (cf. 131-2).

*Тейтеренть моронь морамсто столь экшсэтне киштезевсть. "While the girl was singing a song, the people at table began to dance."

Тейтеренть моршсто столь экшсэтне киштезевсть. "While the girl was singing, the people at table began to dance."

Моронь морамсто столь экшсэтне киштезевсть. "While the song was being sung, the people at table began to dance."

Furthermore, a primary transitive verb such as печкемс "butcher" will not accept a genitive marker for the agent, i.e. "атянть печкемазо "the butchering of the man" could only refer to the man as a patient and not as the agentive (see 3.4.1.5,3.4.2.3,3.4.3.6).

68

3.4.1 Genitive playing the role of subject of an intransitive verb

According to Bubrikh (1947:13-14) the genitive can be used to show the role of the subject of an intransitive verb чинь стямо "sunrise". In the corpus the genitive and sometimes possessive suffix can be used to denote the subject of the intransitive verb with the following verb forms: -зь gerund, -мсто/-мстэ temporal gerund, -мс temporally limiting gerund, and the апак negative gerund. I have chosen to speak of gerunds here because only verb forms of this function can have genitive attributes. The deverbal noun -мо/-ма, like the other nonfinite forms, can be observed to follow the pattern, genitive attribute refers to the roles "subject of an intransitive verb" or "object" in Abramov's texts.

3.4.1.1 The -зь gerund

According to Bartens (1979:70-2) the gerund function of this form can be seen as a verb complement, a free qualifier of the verb, and a qualifier of the entire sentence. Which of these functions can be accompanied by a genitive attribute? When affixed to intransitive verbs, the -зь gerund can take a genitive marked determiner which denotes the subject. It cannot, it seems, be affixed with a possessive suffix. Would it be conceivable to expect a determiner marked with a definite article or possessive suffix; examples (133-4) are both marked with an indefinite genitive.

(133) (0040) Шкаиь ютазь сон лиссь мекев, оршнезь. "After a while she came back out, dressed."

(134) (0015) Сон кежень сазь комась тевезэнзэ ды эзь бажа ваномо пейдицянть ёнов. "Infuriated he bent over his work and didn't want to look in the direction of the one laughing."

Both (133-4) are representative of the function, qualifier of the entire sentence. Since the "subject" of the gerund would necessarily concur with the subject of the finite verb in

69

the instance of "a gerund as a free qualifier of the verb", it would be impossible to have a genitive attribute with an intransitive verb form, cf. 3.4.3.2.

3.4.1.2 The elative gerund -мсто/-мстэ

As a form of intransitive verbs, the -мсто/-мстэ gerund can take a genitive marked determiner which denotes the subject, but in Abramov's texts there seems to be no instance of this, cf. 3.4.3.3. But then again, this might have to do with the fact that when this gerund has an agentive determiner, it can also be marked with a possessive suffix (135-6), Bartens (1979:46). The elative -мсто/-мстэ form works as a qualifier for the entire sentence.

(0265) — Чачомстонзо эшкемапь кавто пильгеде, — мерсь Лабырь сынст пелев ды таго лиссь кудостонть. "When he was born, he should have been slammed by [his] two feet, — said Labyr in their direction and then left the house again." (i.e. [he should have been slammed against something by his two feet and killed...])

(0193) Туемстэнзэ Пахом удыль, эзизе нее, кода лелясь оршизе пиджаконзо. "When he [Styapan] went out Pakhom was sleeping, he [Pakhom] didn't see how his brother had donned his suit jacket."

3.4.1.3 The illative temporal gerund -uc

As a form of intransitive verbs, the -мс gerund can take a genitive marked determiner which denotes the subject (137-8). In examples (139-40) a possessive suffix is used to show this same role. Typical of this form and many others of intransitive verbs, the gerund can have a "subject" other than that of the finite verb, cf. 3.4.3.4. When the subjects are separate, it would seem that the gerund could only be used as a qualifier of the entire sentence. In Abramov's texts only separate subjects have been used.

70

(0086) Патянть самс Аганень самоварось, нама, эзь лакавтово. "Aga hadn't been able to bring the samovar to a boil before her big sister came."

(0072) Банянть явамс кудось таго пешкедсь сыця ломанде. "By the time the bathhouse had gotten warm the house was once again filled with arriving people."

(0009) — Пасиба, Пахом, топодемезэнек андымизь,— мерсь Стяпан ды сормсевезь конястонзо панар ожасо нардызе ливезенть. "Thank you, Pakhom, you've filled us up (lit. "we've gotten full" the Erzyan verb is intransitive.), said Stapan and wiped the sweat off his wrinkled brow with a shirt sleeve."

(140) (0025) Лиясто симиль a помнямозонзо ды ёмавтнилизе весе амунициянзо — "Sometimes he drank until he didn't remember, and spent all of his amunition..."

Agafonova has assured me that there is no reason to set example (141) apart from (137— 8), there is only a temporal sense present. Is the temporal sense the only one available for intransitive verb forms, the transitive verbs offer a broader field, cf. 3.4.3.4. Examples (140-1), for sure, can also be used as transitive verbs, furthermore the -мозонзо in (140) could essentially be formed from both the -мо+с and -мс. Here, however, the context speaks for the interpretation intransitive.

(141) (0054) Сон таргась келей эзементь алдо гигерень покш парго, варчась понкс-панарт ды Захароиь оршнемс ладизе фонаренть. "He pulled a big birch basket out from under the broad bench, he tried out some clothes, and before Zakhar had dressed he had the lantern working."

3.4.1.4 The gerund of negation апак

The апак gerund modifying an intransitive verb only appeared once with a determiner bearing the role "subject of the intransitive verb". It can be seen in (142), that the type of gerund that takes a genitive attribute refering to the "subject" is the same as that found in 3.4.1.1 above, i.e. it qualifies the entire sentence.

71

(142) (0031) Кинень лиянень, азоронь апак мере, двигателенть Петя, нама, аволизе нолда. "Of course, Petya wouldn't have turned on the machine for anyone else without his master's order."

One of the questions which arises in Erzyan is how can an intransitive verb be set apart from a transitive one. Can the verb form апак мере here be set separately from one in a different context which would take an object?

3.4.1.5 The deverbal noun -мо/-иа

The -Mо/-ма deverbal noun can take a genitive-marked determiner which has the role "subject intransitive verb", (see 143). Likewise a possessive suffix affixed to this noun, (see 144-5) also has the same role. Example (146) shows that, at least, a look-alike of the -модо/-меде gerund is governed in the same manner as the nominative in (143) and genitive in (144). Example (145) could be compared with the construction 3. учизе лисемеде "Z. waited for her to be leaving." Are we waiting for "her in the process of leaving" or simply "her leaving".

(0068) Седеень кецямось эзь максо марямо ды чарькодеме валтнэнь. "The rejoicing of the heart didn't let [her] hear or understand the words."

(0244) Сонзэ кецявтсь самозо. "She [Tanya] was pleased by her [her mother's] arrival"

(0085) Захар учизе лисеманзо ды совась тей. "Zakhar waited for her to leave and then he went in."

(0268) Петук Ёгань туемадо мейле весемесь савкшны теемс эстензэ. "After Yoga Petuk left he had to do everything himself."

In the corpus there were no instances of the -модо/-меде infinitive with either genitive attributes of possessive suffixes. Perhaps, Abramov simply does not use this infinitive with attributes when it is derived from an intransitive verb. Martynov, on the other hand, seems to practice a more extensive use of possessive suffixes at least, cf. (1992:130)

72

...лоткак лавгамодот! "...stop blabbering!". Here the possessive suffix affixed to the ablative infitive refers to the second person singular as does the verb. This use of the possessive suffix is possible because the infinitive in question has been formed from an intransitive verb. It seems that the need of concurrence of subjects with the ablative infinitive makes instances of genitive attributes rather infrequent, see 3.4.3.7 below.

3.4.2 Genitive playing the role of subject of a transitive verb

The genitive and sometimes possessive suffix can be used to denote the agentive with the following deverbals: -зь gerund, апак gerund, and the deverbal noun in -мо/-ма. In examples (135-6) I briefly mentioned the problem faced with the elative gerund -мсто/-мстэ and their relation to the agentive. Much the same can be said of the -мс gerund, as well. This one, however, exhibits an ability to take either a genitive or an indefinite nominative singular determiner in the role of "subject of an intransitive verb". In Kuldyrkayev (1994: 14/16) Ансяк илямак атякш#морамс кирдть "Only, don't keep me till the cock crows." According to N. Adushkina, both атякш#морамс and атякшонь морамс would be possible, but this use of the absolute form is archaic (summer 1995). Compare the formations стадань ваныця and стада#ваныця. Feoktistov said that the formation атякш#морамс in question is actually an allegro form of атякш#морамос, (oral information Jan. 1996). Thus the temporal interpretation might readily be associated with a compound noun construction, whereas no notion of specific indefinite for "rooster" is present. Hence a comparison with the construction стада#ваныця "herder" even козяйка#вешнеме "in search of a wife" is merited, cf. 3.4.3.5,3.4.3.8 below.

73

3.4.2.1 Is there a -зь agentive participle?

In contrast to the gerund -зь, Abramov uses a participle which, in the case of transitive verbs, can be determined by an agentive role in the genitive (147-8). In standard Erzyan this -зь participle alternates with a preferred -вт/-вть agentive participle. According to Bubrikh (1947:14) the -вт/-вть participle is used with the agent, for instance тейтерень сокавт пакся "a field plowed by a girl". On the other hand, GMYa (1980:353-4) also presents this participle with the stipulation that it requires the presence of the genitive. In addition to an example of an agentive nature Kapn атя стясь, саинзе содавикс таркастонть одирьванть ёравт карсемапелензэ ды варштась вальмава (Т. Raptanov) "Old man Karp got up, took his shoes, which his daughter-in-law had prepared, from the usual place, and looked out the window", they also give one of a temporal nature Чись яла шашты верев ... ды нолси вановтсонзо валскень правт росанть (Т. Raptanov) "The sun keeps drudging upward ... and licks at the dew, which fell in the morning, with its gaze." On the other hand there is one example that brings about a different element, namely that of the instrument and surface Игаев сельмень сявадовт вановтсо ильтясь айгоронть мельга кепедевть пуленть (Т. Raptanov) "Igayev followed the dust, which had risen behind the horse, with an envious gaze." Since all of the examples come from the same quill, as it were, one might question the author's background, does his dialect have a -зь form.

Abramov does have examples of the -вт/-вть form, but they are limited (in distribution) to his quotations of Erzyan folksongs. On the other hand, he occasionally used a -вт/-вть affixed to transitive verbs in the predicate, (see 149).

(0250) Сонгак маризе Ларькань кандозь кулянть. He too heard the news Larka had brought.

(0252) Зярдояк a кадсынек Ленинэнь ушодозь тевенть!.. "We shall never abandon the work begun by Lenin."

74

(149) (0017) Малавикс роднянзо Наймансо Кандрань арасть. Низэ, Елена, сгевть ошсто. "In Naiman Kandra has no close relatives. His wife, Yelena, comes from town." (lit. ...is taken from town.)

In my discussion of the examples (147—49) with Feoktistov and Agafonova (Jan. 1996), it was concluded that the examples (147-8) were ambiguous, i.e. in the case of transitive verbs, the genitive qualifier is expected to represent the patient, whereas an agentive is only representative of a Russian construction. The fact that it is not typical of Erzyan might be anticipated from Bubrikh's presentation of -зь gerunds (1947:14) where he only gives the examples чинь стязь "the sun risen"and нумолонь кундазь "with a rabit caught". Example (147) might have sounded a little more Erzyan if the gerund кандозь had been left out. On the other hand, Abramov's use of the -вт/ -ВТЬ agent participle in the predicate might reflect his dialectical background, but V. A. Ledyaikina (1980:85) states that the -вт/-вть participle requires an accompanying qualifier in -нь.

3.4.2.2 The gerund of negation апак

The апак gerund when used with the transitive verbs can take a genitive determiner which refers to the agentive. Examples (150-1) have verbs that could be interpreted as pseudo-intransitive, i.e. they are not "real transitive verbs". As is the case with intrasitive verbs, cf. 3.4.1.4 above, the gerund here is used as a qualifier of the entire sentence, cf. 3.4.3.9 below.

(0176) Эрявсь саемс кодамояк конёв, саемс Канаевень апак сода секретаресь сёрмадызе тензэ Канаевтеме. "[He] needed to take some kind of paper, to take [it] without Kanayev knowing. The secretary wrote it for him without Kanayev."

(0207) Течи, несак, Дрига ды волостной прявтось якасть паксява, ванность, илязо кадово киньгак апак виде. "Today, you see, Driga and the administration district head walked through the fields and checked that nothing be left unseeded." (lit. ... and checked lest someone be left without sowing seed.)

75

3.4.2.3 The deverbal noun -%ло/-ма

The deverbal noun in -мо/-ма can be modified by a possessive suffix denoting the agentive role (see 152-3), whereas this role can also be denoted by a genitive, (see 155). Example (154) shows an instance of where context and empirical experience would show that Petya was not going to pinch the bees, i.e. the possessive suffix here could, indeed, refer to either the agentive or the patient. Since the use of the ablative in example (154) is one of cause, the rections of the verb, this form should not be considered an infinitive of its own right, cf. 3.3.2,3.4.3.7.

(0092) Ней ойми Маря, учомазо учовсь. "Now Marya can rest, her waiting is over, (lit. her waiting has been waited.)"

(0302) Елена маризе ававтонть ёвтнеманзо ды капшазевсь лисеме удалце кудов. "Yelena heard what her mother-in-law said and hurried out into the back room."

(0037) Пирев чиемстэ Петянь коняс эшксесть ливтниця промот, сон тонадсь сынст бизнэмантень ды тунь а пели пупамодост. "As he runs into the yard Petya is hit in the forehead by flying bees, he's become used to their buzzing and isn't afraid at all of their sting."

(0104) Захаронь самось сезизе Стяпанонь ёвтнеманть... "Zakhar's arrival interrupted what Styapan was saying..."

3.4.3 Genitive playing the role of patient

According to Bubrikh (1947:13-14) the genitive can be used to show the role of the direct object моронь морамо "singing of a song". In Abramov's texts the genitive determiner of infinitives and other deverbals is used most frequently to mark the object of a transitive verb. His use of the deverbal noun and -мадо form tend to unify them as a set, although he does make at least one individual set of minimal pairs (cf. 170, 174). The genitive and sometimes possessive suffix can be used to denote the patient with the

76

following deverbals: -мга gerund, -зь gerund, -мсто/-мстэ gerund, -мс gerund, -мо/-ме infinitive, deverbal noun -мо/-ма, -мадо form, -ця participle, and the апак gerund.

3.4.3.1 The prolative gerund -мга

In my conversation with Yermushkin (spring 1994) I learned it was possible to use the possessive suffix to denote the subject of an intransitive verb and the object of a transitive. Likewise the genitive attributes can, at least, be used for showing the object. The -мга gerund in Abramov's texts can have a genitive determiner as in (156), or it can be affixed with a possessive suffix (157), both of which refer to the role "object of the transitive verb". Both instances are one in which the gerund is used to qualify the entire sentence.

(0284) Дурнов Иван эрьва скалояь ветямга саи пель пондо сюро. "For every cow serviced, Durnov Ivan takes half a pood of grain."

(0045) Ваннызе ниле тундонь ведьсэ салсезь сёлтонть, ловизе, зяро ломань сави сиведемс сёлтамганзо. "He examined the pond which had been eroded by the water of four spring floods. He counted how many people he would have to hire in order to dam it."

3.4.3.2 The gerund -зь

According to Bartens (1979:70-2) the gerund function of this form can be seen as a verb complement, a free qualifier of the verb, and a qualifier of the entire sentence. Which of these functions can be accompanied by a genitive attribute. The -зь gerund of transitive verbs takes a genitive determiner which refers to the object, (see 158-9). As was the case with the intransitive verb, no instances of possessive suffixes have been observed.

77

(0144) Захар мезеяк эзь мере, лавтовонь лыкавтозь айгизе Миколень кеденть ды а мельсэ варштась оев сельмезэнзэ. "Zakhar said nothing, with the shrug of a shoulder he moved Mikol's hand and looked into his (Mikol's) oily eyes with disgust."

(0237) Таня кадовсь аштеме прянь нолдазь. "Tanya was left sitting with her head hanging, (lit. ...with head lowered.)"

In contrast to the gerund usage of the intransitive verb in 3.4.1.1, the transitive verb provides a gerund which has the function of a free qualifier of the verb. Could there be other functions than the one shown in (158-9).

3.4.3.3 The genitive attribute with elative gerund -исто/-мстэ

The -МСТ0/-МСТЭ gerund in (160-3) shows a large variety of determiners - all referring to the patient, i.e. genitive (160, 163), pronoun (161) and possessive suffix (162). The

usage of this gerund construction is also as a qualifier of the entire sentence.

(0109) Эрьва калонь таргамсто Надя сеересь ды цяпась кедьсэ. "At the pulling-in of each fish, Nadya cried out and clapped her hands."

(0024) Минек эрзянь моротне лужанят, сынст морамсто паро ансяк авардемс. "Our Erzyan songs are sad, when you sing them it's only worth crying."

(0065) -- Эно кода, каземстэнзэ мерить, илясан ёмавто. "What did you expect, when [you] gave it [to me] you said that I not lose it."

(0032) Салдин Кандра, мерть, весе мельсэ,— кортась сон улавтнень ильтямсто. "Say, Kandra Saldin would be very pleased to, - he said as he saw the wagon out."

3.4.3.4 The illative infinitive -ыс

According to Bartens (1979:31-43) the -мс form can be used as a qualifier of the verb, an attribute of a noun, an adverbial of an adjective, and as a qualifier of the entire sentence

78

as well as the infinitive of an "in order to" phrase. With transitive verb forms preceded by a genitive attribute Abramov tends to use his -мс infinitives as attributes of the noun. In the Erzyan language as a whole, one would also expect the gerund usage with transitive verbs.

In the "Erzyan pravda" (Oct. 19, 1995) there was an example of an -мс infinitive in the position of attribute of a noun: Истямо покш комиссиясь вансы, прядовсть-эзть хозяйстванть паксянь тевензэ, витнезь-арась техникась, кода анокстазь телентень ракшань кирдемс кардтнэ. "This kind of large comission checks to see if the work in the field has been completed or not, whether the machinery has been fixed or not, and how the barns have been prepared for holding animals over the winter." This can be compared with the example Bartens gives from folklore (1979:85) rauzho bis'or surguton' paks'an' sokams-modazo "mustan helmen kaltainen on Surgutin alueen peltomaa". Examples of the -мс infinitive used as an attribute of a noun can also be found in Abramov's texts, where they are limited to forms of transitive verbs.

Adushkina stated that the difference between the use of -мс and -мо/-ма as attributes of nouns is one of modality (summer 1995 oral information). Hence the structures ракшань кирдемс кардтнэ "barns where animals can be kept" and ракшань кирдема кардтнэ "barns where animals are kept" can help us to interpret the example from folklore in a slightly different way. First, there is the full phrase Сургутонь паксянь сокамс модазо in which the elements can be split into attribute паксянь сокамс "arable, suitable for plowing fields" and the qualified noun модазо "land+(in Sg Px)".

One might, of course, hypothesize an "in order to" clause. This hypothesis would best be supported by example (166), which gives the phrase лишмень рамамс "for buying a horse" as the adverbial. The preceding dative is, at least, the required valency for the verb

79

to give. Do there exist instances of the dative designating the agentive of the -мс infinitive with a genitive qualifier for patient.

(0235) Захар лаксесь коник эземсэ, тейсь ули-паронзо путомс чова лазонь горобкине. "Zakhar was carving on the bench, he was making a box of thin boards for putting his belongings in."

(0233) Тетясь мольсь икеле Миколь мельганзо, кедьсэнзэ сювань кандомс плетюшка. "The father went in front and Mikol after him, in his hands a wicker basket for carrying chaff in."

(0137) Кудов сыця красноармеецтнэнень лишмень рамамс макснесть ссуда, кудонь теемс нолтнесть вирь. "The Red Gaurd returning home was give a loan to buy a horse with, for building a house they allotted forest"

Examples (164-5) provide a contrast to (166), i.e. the -мс phrase acts as an attribute of a noun in (164-5), whereas example (166) shows a qualifier of the verb — an "in order to"

clause.

3.4.3.5 The (nominative/lative) infinitive -uo/-ue

The -мо/-ме infinitive can only take a genitive attribute or be affixed with a possessive suffix with a object referent if it is derived from a transitive verb. According to Alhoniemi (1994: 24) the MORMULA corpus in Turku shows no instances of indefinite genitives following the infinitive -мо/-ме. On the other hand his example 8), with an absolute form in the position of object preceding the infinitive, appears to exhibit a parallel construction to those found in стада#ваньщя "herder" (3.4.3.8) and атякш#морамс "till the cock crows" (3.4.2), cf. [тусь козайка#вешнеме] "he went to look for a wife".

According to Battens (1979:28-31) this infinitive can be used as a complement of the verb, as a subject, and she quotes a claim from Uchkina's discertation that it can also

80

represent the predicate. In example (167) the object is denoted by a possessive suffix affixed to the infinitive, and example (168) show its counterpart with a genitive attribute.

(0311) Танянь туемадо мейле Марюшанень сась Палай баба паксяв кучомонзо. "After Tanya left, granny Palai came to Marusha's to send her out in the field."

(0316) Кандра эзь тонадо тевень тееме апак арсе. "Kandra wasn't used to doing things without planning."

3.4.3.6 The genitive as an attribute of the deverbal noun -мо/-ма

The deverbal noun -мо/-ма can take a genitive attribute (169) or a possessive suffix (170) refering to the patient. In examples (169, 171) we can see the deverbal noun in the position of noun attribute, whereas example (170) shows that it can play the role of subject, as well. Example (172) might seem problematic, but according to Bartens (1979:33) the object conjugation can never appear with the -мо/-ме infinitive. Hence the form шлямонзо must be regarded as a noun. It is, however, uncertain whether the possessive suffix should be interpreted as refering to the agent or the patient.

(0252) Ведьгевень теемадо кулясь пачкодсь Салдин Кандранень. "Word of making a mill reached Kandra Saldin."

(0008) Te, келя, кие соды мезе, паряк, грех ярсамозояк, — ёвтнесь Матя ды элестэнзэ мекев мешоконтень чукадизе фасоленть. "Who knows what this is, maybe it's even a sin to eat it, said Matya and shook the beans out of her lap and back into the sack." (Literally: ...maybe, even the eating of it is a sin...)

(0274) Ларька стясь валдонь a токамо таркас, аштесь апак лыка ды кунсолось кудостонть кайсетиця вайгельтнень. "Larka stepped into a shadow, he waited without moving and listened to the voices coming from the building."

81

(172) (0028) Захар прядызе шлямонзо ды сыргась вишка кудынентень, панжадо кенкшентень поводезельть панарозо ды нардамо#нулазо. "Zakhar finished his washing and headed for the little shed, his shirt and towel were hung on the open door."

The examples (171, 173) show the use of the -мо/-ма form as attributes of a noun. This usage in combination with the genitive attribute denoting the object is evidence of a problem faced in the new orthography. There are people, such as Mosin and Feoktistov who would generally write the -мо/-ма form together with the following noun, as though it were a question of two nouns forming a compound word. Is the genitive attribute strong enough evidence to show that the -мо/-ма form could also be a participle. Would this fact have any effect upon the development of the new Erzyan orthography.

(173) (0109) Te таркасонть сон пурнавсь келей ды сэрей сёлтокс, тесэ зярдо- бути ульнесь керень лопавтнема эрьке. "At this point it expanded into a wide and deep pond, at one time there had been a lake here for soaking bark."

3.4.3.7 Is there a -модо/-меде infinitive in Abramov's writing?

Abramov actually uses the -мадо/-модо/-меде forms for the ablative infinitive. These might be compared to the ablative construction used with concrete nouns, for instance Тейтересь кундызе цёранть ожадо "The girl grabbed the boy by the hand", see also 3.3.1. A second, and perhaps more likely comparison is that found in the adverbials of position — озадо "in sitting position". On the other hand, there are no instances in Abramov's texts of the ablative infinitive with a genitive attribute.

The -модо/-меде construction presuposes a concurrence of agentive of the ablative infinitive with either the object or subject of the finite verb. Any possessive suffixes affixed to the infinitive or genitive attributes would denote the patient or agentive of the infinitive according to whether it was formed from a transitive verb or an intransitive verb.

82

Hence the short dialogue — Кие печкизе буканть? — Шабра атянть неия печкедемедензэ. "Who butchered the bull? — I saw the neighbor man butchering it" — shows the cohesion between the object of the verb "to see" and the agentive of the verb to butcher (Agafonova, Jan. 1996).

As seen in examples (124, 126) the object phrase can take an adverbial in the ablative. However, there appear to be no instances of such an adverbial being qualified by either a genitive or a possessive suffix.

3.4.3.8 The active participle -ы(ця)/-и(ця)

According to Bartens (1979:58-62) the -ы(ця)/-и(ця) participle is an active participle which denotes continuous/incompleted action. Its main function is that of a participle attribute, but it might also be seen as predicative, complement of the verb, and even transformed to agent noun. Bubrikh (1947:40) gives further information, which states that the genitive indicates the direct object in the participle construction калонь кунды "angler". In the corpus, the -ы(ця)/-и(ця) "participle" can take a genitive determiner that, by and large, would refer to the patient (174-5). The sense of this genitive attribute, however, must be a temporary one — something not found in the term стада#ваныця "cowherd" of (176). Furthermore, we can ask whether лездыця "helper, assistant" of (177) can be governed by a genitive which is not comparable to the expected dative valency of the finite verb лездамс "help". The fact that the -ця form can take a genitive attribute seems to demonstrate that the name of an agent is in question (174-5, 177). Examples (176-7) also show an ability to function as complements of the verb, but the question arises as to whether these should not be considered noun.

83

(0097) Зярс сон ваннось военной билетэнзэ, Канаев кевкстизе, кода сонензэ муемс волостьсэ партиянь тевень ветицянть. "While he examined the military card, Kanayev asked him how he could find the director of party work in the administrative district."

(0051) — Кинть эйстэ a васоло эрясь вирень ванстыця, кудозо палсь юртонтень яла, паряк, мезеяк кадовсь. "-- Not far from the road there lived a forest warden, his house burnt down to the foundation, but maybe something is left of it."

(0171) — Ней, Пахом, сы кизна а улят стада#ваныцякс? "-- Now, Pakhom, you won't be a cowherd next summer."

(0039) Be кизэ аравтнизь стада#ваныцянь лездыцякс. "One summer they appointed him a cowherd's assistant."

3.4.3.9 The negative gerund апак

The апак gerund when used with the transitive verbs can take a genitive determiner which refers to the patient (178-9), cf. 3.4.2.2 and 3.4.1.4 above. As seen in the government of the -зь gerund 3.4.3.2 above, a gerund derived from a transitive verb with a genitive determiner refering to the patient can be used as a free qualifier of the verb (178). Does example (179), on the other hand, show an instance of the an an gerund in the function of qualifier of the entire sentence.

(0133) Сюпавсе, кона чинек-венек роботы виень апак жаля! "A rich man is one who works day and night using all his strength!" (Literally: (... works day and night without holding back on strength.)

(0147) — a мельсэ пшкадсь Кандра ды, Ёгань пельде валонь апак учо, тусь эсь тевезэнзэ. "— Kandra said in disgust, and without waiting for a reply from Yoga, he went back to his work."

84

3.5 THE GENITIVE FOR GOVERNMENT OF POSTPOSITIONS

GMYa (1980) does not deal with the government of postpositions in any clear manner. On the other hand, Martti Kahla (1974:38) give the following characterization of the government of postpositions whereby the absolute form of the noun is used when it is considered to express something new, not mentioned previously, or something in general. Other declinations (i.e. the definite and possessive declinations) are used with the genitive when the matter in question is known or specified. He goes on to mention two further groups of exceptions that are found only for Mokshan — the Erzyan dictionaries did not mention rection. First, there is the rather extensive group whose heads are always in the definite genitive. Here, I have taken the liberty to translate the Mokshan postpositions Kahla has given for this group into Erzyan — something which might be helpful when examining Abramov's texts.

ЭЙС "into" ЭЙСЭ "in; with" ЭЙСТЭ "out of эземс "in place of; to the level of кедьс "to (receiver, residence)" кедьсэ "in the possession of; at" пачк "through" перть "during, throughout" пингстэ "during the time of; in the presence of удало "behind" удалдо "from behind" удалга "(passing) behind" In the corpus, however, this grouping set for the Mokshan language does not seem relevent. The эй-/эз- stem is the only one whose head was always in the definite genitive. Otherwise, the group given above seemed to be distributed among absolute heads as well as those affixed with either a definite or indefinite genitive ending.

85

The second group of exceptions, which Kahla presented, includes postpositions whose heads are affixed with either the indefinite genitive or the definite genitive:

келес "throughout, across (spatial)"

кондямо "like (something)"

коряс "in relation to"

кувалма "(speaking) of, about"

This division provided for the Mokshan language is useful for the Erzyan language, as well. There is, however, one discrepency of definition. Since postpositions generally have adverbial roles, it seems rather peculiar to call the word кондямо "resembling" a postposition — it is used as either an adjective or a noun. Furthermore, my discussion with Nina Aasmäe (Feb. 1996) revealed that there might be even more intricate rules for the usage of marking for the head in the Erzyan language. Thus the information given below will be based upon evidence from the corpus.

The postpositions in the corpus can be divided into two groups, one requires the genitive for its government, and the other one takes the oblique cases such as the ablative or elative. The first group is by far the most extensive and can be divided into three subgroups — much like the Mokshan above.

All three subgroups share the definite genitive for their government in contexts where definiteness is required. Consideration of the indefinite counterparts, however, immediately divides this group into three subgroups. First there are those postpositions which take the absolute form in the indefinite context. These postpositions are mainly ones of spatial expression, and in many circumstances they could be compared with compound words. Unlike the compound words where the first element is in the absolute form, this absolute form has a specific referent or referents. The use of postpositions with

86

pronouns also presents an interesting feature of emphasis depending on whether the pronoun or possessive suffix is used to denote the head.

According to Tsygankin, an alternation between use of the pronoun in the genitive for the government of a postposition and a postposition with a possessive suffix can be discerned (discussion, summer 1995). The emphasis falls on the pronoun when it is used together with a bare postposition. A postposition with a possessive suffix affixed to it has no emphasis on the person. Hence сынст марто is a formula that emphasizes the third person plural and мартост is the neutral form, both could be roughly translated to "with them". Hungarian has a tendency somewhat like this, although the possessive suffix their is retained throughout: ennekem versus nekem "to me".

In example (180) we have the absolute or indefinite nominative singular, which in this position can denote both singular and plural. Examples (181-2), on the other hand, show that definiteness requires the knowledge of number.

цёра марто "with a man/men"

цёранть марто "with the man"

цёратнень марто "with the men"

The second group of postpositions, is one in which the genitive is always used regardless of whether the word is definite (184—5) or indefinite (183). Like the indefinite of the previous subgroup the word тейтерень can have both singular and plural referents.

тейтерень кисэ "for a girl/girls"

тейтеренть кисэ "for the girl"

тейтертнень кисэ "for the girls".

87

Group three is apparently limited to the postposition stem эз-/эй- which in Abramovs text alternates with the illative, inessive and elative cases. In the Erzyan language, however, use of the illative эйс form or эйзэ- stem is relatively infrequent, and its role is played by other case forms. This stem, it will be recalled, was present in the Mokshan group as well. In fact, there might be reason to include the эйс paradigm among the pronoun paradigms which seldom use their own illative, elative and inessive forms. No matter how you look at it, the эз-/эй- stem is used with high frequency. It cannot necessarily be dealt with as full-fledged postposition. Although it shares many of the traits of the postposition, one feature that links it firmly to the pronouns, is the fact that it does not appear together with indefinite attributes.

Further examination of the эз-/эй- stem might be carried out with regard to the pronouns сонзэс (3P Sg. Gen. + 111.) (0226), тоньзэть (2P Sg. Gen. + 111 + Px (2P Sg.)) (0049), минзэнекак (IP PI. + 111 + Px (IP PI.) + Inclusive) (0301). It appears that these forms alternate with the forms эйзэнзэ, эйзеть and эйзэнек respectively. Many of the Erzyan speakers I have confronted, however, did not recognize the ones formed with the pronoun stem. Could this alternation be used in the same way as the formula of emphasis seen in сынст марто and мартост above. There are no instances of the bare эйс in the entire text. Its functions are usually played by the simple -c form attached to the noun, so the pronouns are required to provide some kind of alternation.

3.5.1 The postpositions that have the absolute form for government in indefinite circumstances

Here I shall deal with the first two groups only. In both groups, the definite forms are identical. If a possessive suffix is added to the postposition, a preceding genitive determiner is not obligatory. With the indefinite circumstances in example (187), there is

a tendency nowadays, among grammarians, to consider the resulting word a compound word although it is homonymous with the noun plus postposition of example (186).

ош куншкасо "in the middle of town"

ошкуншкасо "in a city center"

ошонть куншкасо "in the middle of this city"

ошкуншкасонть "in the city center"

Examples (186-9) show how closely the postpostional structure can be related to that of the compound word. In this group of postpositions, nearly all of them have a homonymous noun. Hence we must rely upon the existence of a meaning not inherent in the nominative forms of those nouns to attest to the postposition status — indeed a very difficult distinction.

As a rule of thumb, the postpositions of the first group, with only a few exceptions, express spatial relations. In the presentation below, I have used brackets surrounding x to indicate state of rest and additional arrows > to indicate direction. Thus we can have the approach (lative, illative) [>x], the state of rest (locative, inessive) [x], the departure (elative) [x>] and the path (prolative) [>x>]. Following the illative forms there is a second form in parantheses that shows the illative stem to which possessive suffixes (Px) are added. The function of the lative, where possessive suffixes are added, is taken by the form of the locative with a possessive suffix.

stem lative locative ablative prolative (lative) [>х] [х] (+Рх) [х>] (+Рх) [>Х>] (+РХ) ал- алов ало алдо алга "under" ён- ёнов ёно ёндо ёнга "toward" икель- икелев икеле икельде икельга "in front of удал- удалов удало удалдо удалга "behind" 89

In this first series a peculiarity arises — the opposite of the an- stem верь- does not appear in the corpus as a postposition. Instead, it is used as an adverb only.

stem illative

lative inessive elative [>x] [x](+Px)

ланг- лангс (лангозо-) лангов лангсо лангсто

мала- малас (малазо-) малав маласо маласто

пе- пес (пезэ-) пев песе пестэ

ПОТ- ПОТС (П0ТМ030-) ПОТМОВ ПОТСО ПОТСТО

экш- экшс (экшезэ-) экшев экшсэ экшстэ

ютк- юткс (юткозо-) ютков ютксо ютксто

prolative

ланга

малава

пева

потмова

экшка

юткова

"onto, on top of; off of

"up to, near, vicinity of

"at the end"

"into; in"

"behind"

"between, among, through"

The postpositions can be split into four groups according to the variety of case endings used. The paradigms show that in many instances these so-called postpositions could be easily classified as nouns. Hence the postpositions missing include ones such as бока-"side", пря- "top" etc. (see (190-1) below). In fact, the text only used the lative form ютков as part of a compound word уяьця#ютков "into the street". On the other hand, there was no instance of the word куншкасто. Finally,

stem

illative

inessive elative prolative

[x] (+Px) [x>] (+Px) [>x>] (+Px)

вакс- ваксс (ваксозо-) вакссо вакссто вакска "next to"

велькс- вельксс (вельксэзэ-) велькссэ вельксстэ велысска "over"

куншка- куншкас (кушказо-) куншкасо куншкава "in the middle"

Finally, two words which can readily be acknowledged as postpositions are "with" and "after (procurement)". The use of the term postposition in the Erzyan grammar is very liberal and comparable to that used in the Komi language grammar. Of course, these paradigms are sometimes lacking in nominative forms, for instance икелькс is the form given in MW as nominative of the икель- series.

90

марто мельга

"with" "after"

The "Исяк якинь Найманов" text seems to alternate in its preference of (pronoun in the genitive + postposition) versus (postposition + possessive suffix). Perhaps, the best solution would be to treat all postpositions separately. The division between the two variants of эйсэ might be attributed to a split in grammatical function, whereas эйсэ + possessive suffix is the (core) object and poss. pron. + эйсэ is the oblique. Another possibility would be to posit the "[pronoun in the genitive] + [postposition]" formula as emphatic and the "postposition + possessive suffix" formula as neutral. A hint of such a division is to be found in the use of марто "with" and the use of border-line words like мельс "mind; feeling". Of course, it would take a much larger corpus to reach any conclusions with regards to Abramov's language.

Although this division of the first two sets of postpositions holds true for common nouns, I have noted variation which I have been unable to explain in the valencies with regard to personal names, i.e. Захар марто "with Zakhar" but Захар лангс and Захаронь лангс "at Zakhar (look at; be mad at)". Although I would have expected the absolute form Захар лангс through out the text, I was surprised to find that the other personal names only took the government of the indefinite genitive.

Nina Aasmäe suggested that there was actually a phenomenon of three different possible combinations with personal names and postpositions. We were unable to establish a semantical basis for distinguishing between the indefinite genitive and absolute forms when adding the postposition лангс. Apparently, there is an actual difference. (Discussion, Feb., 1996.)

91

Finally the definition of postposition should be established for according to (190-1).

(A) words lacking complete paradigms and used with a determiner, for instance word stems like мала- expressing vicinity which has no nominative or genitive forms.

(B) words with a specific meaning found only in a part of the paradigm, for instance in the case of кедьсэ "in the possession of; at someone's place" there is a semantical break from the noun кедь "hand".

Below are a few of the problems presented in the valencies of postpositions in the assumed groupings, where a new viewpoint must be taken. Section (3.5.1) represents that group of postpositions which only has one clearly defined, nonambiguous valency for postpositions in this corpus, i.e. the definite genitive. Group (3.5.2), on the other hand, has two regularly used forms, one for the indefinite and the other for the definite. Futhermore, it has no nominative singular element it can be affixed to.

In instances of this pattern, it may prove difficult to establish the role of the nominative indefinite element. In the instance of compound words like кудо#прясо "in the attic; on the roof and what would be its counterpart кудонть прясо "on top of the house" we can observe the flaws of misinterpretation, i.e. in the case of the "postposition" прясо the indefinite pair is by no means the nominative. Instead, an indefinite genitive form кудонь прясо "on top of a/the house" would be used (oral information, Adushkina, summer 1995). In examples (192-4) the word ёнов "in the direction, toward etc." the confusion of three different instances can be seen, whereas (192, 194) seem, at first, to contradict each other.

Upon closer examination one will note that the two instances of the word эрямо "life" in (194) give us reason to re-examine (192). It appears that the words рунгозо "his body" and шанжавонь "of a spider" cannot represent the same level in the text, i.e. the word

92

рунго has been ellipted where one might have predicted *шанжавонъ рунго ёнов. Proof of this assumption can be found in (193) where instead of a common noun we once again find an ellipsis; in fact there were no instances of nouns in the definite genitive with the word ёнов.

(192) (0018) Нурькине ды эчке рунгозо вельть моли шанжавонь ёнов. "His short and stocky body quite resembled that of a spider."

(193) (0053) Конязо келей, сакалонзо нурькинеть, вельть моли стенасо портретсэнть ёнов. "His forehead [was] broad, his beard [was] short, and he quite resembled the one in the portrait on the wall."

(0177) Салдин Кандрань эрямозо ней вельть мольсь кажварчонь эряыо ёнов. "Kandra Saldin's life now quite resembled the life of a mole."

(0104) Миша эзь понго аванзо кедь#малас, седе курок эцесь каштом удалов. "Misha wasn't caught in arm's reach of his mother, he made a quick retreat behind the stove." (The noun phrase here would consist of the elements (hand + vicinity) » arm's reach.)

Could it be that the only true analysis of postpositions would be one performed on a much larger corpus? It would at least require a greater deal of time to establish criteria for use of the term postposition. Example (195) shows how close the root мала is to being a noun, (MWII: 1166) also testifies to this.

Example (196) shows one of the problems with relation of place names. The so-called postpositions куншкас "to the middle of and чирес "to the edge of could just as easily be part of a compound-word appellation following the place name. In the place name Вишкалей, for example we have the two elements Вишка "little" and лей "river". Example (197) shows that no appellation is necessary, and that the word чиресэ can follow its determiner directly.

93

(196) (0037) Петя вечки арсеме — ды истят кудотнень сон тейнесь а весть, эрьва кува — Пиже луга куншкас, Вишкалей чирес. "Petya likes to plan — this wasn't the only time he had made houses like this, [he made them] everywhere: in the middle of 'Pizhe luga', and on the shore of 'Vishkalei'."

(197) (0033) Гаруз Василь весе пингензэ ютавтызе Рав чиресэ, кудов сакшныль ансяк нинзэ ды эйкакшонзо содавтомо. "The shore of the Volga is where Vasil Garuz had spent his entire life; he had only come home to pay his wife and children a visit."

3.5.2 Postpositions which take the genitive in both the indefinite and definite

Unlike the postpositions of the previous group, the ones in (3.5.2) tend to bear more abstract meanings, even though there would appear to be concrete notions in postpositions such as кедьсэ "in the possession of; in the abode of and the stem томбал- "on the other side of. In fact, it could very well be the concreteness found in the referents of the stems of the first group of postpositions that calls for the absolute form of the governing head. Conversely the abstractness found in the referents of the stems of postpositions in the second group might account for the use of the indefinite genitive (see 198), i.e. in parts of the human body, (3.1.2).

In example (200), however, the presence of an indefinite genitive is coincidental, whereas the referent of кудыкелькс плетень is one unique to the household, all other instances of томбале etc. are restricted to definite circumstances. Example (198) shows the reclassification of the adjective сюпав "rich", it has become a noun; furthermore, you can only be in the abode or household of someone existent. Example (199) simply shows a different meaning of the postposition каршо "before" not found in MW or ERV.

(198) (0010) Захар вельть лездаволь тензэ, ансяк кода сон максы лескс, сонськак эри сюпавонь кедьсэ сиведезь. "Zakhar would have really liked to help him but how could he be of assistance, he himself employed and lived in the home of a rich man."

94

(0253) — Рамакшныя пильгень ёзамс, вельть лезды. Пиземень каршо яла чулксети. "I bought it to rub [my] legs with, it really works. They always start aching before the rain comes." (Literally: ...before a rain.)

(0096) Кудыкелькс плетенень томбале атякшось апак учо вачкодсь селмосонзо ды моразевсь. "The rooster, on the other side of the porch wicket, unexpectedly flapped its wings and crowed."

The corpus included postpositions of the second group, which had more abstract meanings. Hence there are fewer instances of paradigmatical series, and those present were incomplete.

[>х] (+Рх) [x] (+Px) [x>] (+Px) [>x>] (+Px) (illative) (inessive) (elative) (prolative) кедь- кедьс(кедезэ) кедьсэ кедьстэ кедьга "under jurisdiction of, in possession of тарка таркас (тарказо) таркасо "instead, in lieu of [>x] (+Px) [X] (+PX) [X>] (+PX) (lative) (locative) (ablative) пель- пелев пеле пельде "toward; from" томбаль- томбалев томбале томбальде "beyond, on other side" каршо (+Px) "against; preceding" also possessive form каршозонзо. кисэ (+Px) (inessive) "for, after, on behalf of..." no instances of кис. койсэ (+Px) (inessive) "in oppion of коряс (корязо) (+Рх) (illative) "in relation to" кувалт "because of, along" перька (+Px) (prolative) [>x>] "around" перть "during" пингстэ (+Px) (elative) "in presence of (spatio-temp)" трокс "through, over" туртов "to, for" Perhaps, statistics based on a much larger corpus would help produce a better view of the paradigms available in Erzyan postpositions. A three way division of the Erzyan

95

postpositions with regard to the genitive used for government is still unsufficiently founded. Further work, pending a detailed study of the definite/demonstrative article, is still required.

3.6 DIVISION AND DISTRIBUTION

According to Bubrikh (1947:46) adverbs ending in -нь answer the question "no-скольку", i.e. чочконь-чочконь "a log at a time", кавонь-кавонь "two at a time". GMYa (1980:163) gives the example: Удалце кудосонть эзть эря, голландкась пельскаладозь, штукатуркась таркань-таркань певерсь (К. Abramov). "They didn't live in the back room, the tile stove was dilapidated, the plaster had pealed in places." The chief editor of the Erzyan-language journal "Syatko", Aleksandr Doronin, would have preferred to see the forms таркань-тарка or таркань-таркат.

Abramov uses the -нь suffix in two different formulas common to Erzyan. Both consist of a doubled element, whereas formula "(201)" affixes an indefinite genitive -нь to the end of both segments and formula "(202)" affixes the -Hi only to the first one.

Rules (201-2) such that

X = {noun, quant} such that a notion of measurement is apparent

and Х-нь such that X has an affixed indefinite genitive

(201) Х-нь-Х-нь

(201a) (0253) Минь аволь ламо, ансяк стопкань-стопкань. "We don't [want]

much, just a glass or so." (201b) (0018) Истямо ладсо, аламонь-аламонь, Салдин Кандра теевсь

велесэнть сюпавдояк сюпавокс. "In this way, Kandra Saldin little by little

became a very rich man in the village." (201c) (0009) Лиятне ярсасть кургонь-кургонь ды шнасть: те ярсамопеленть

марто мик кшияк а эряви. "The others ate with gusto (lit. by the mouthful)

and boasted: with this food you don't even need bread."

96

(20Id) (0150) Минь тенстемеяк аламоненек. карминдерятано сёвномо эсь юткова, кадовтано ськамонь-ськамонь... "There are few of us without that, if we start quarreling amongst ourselves we'll be left all alone"

As seen in examples (201a-d) the genitive-genitive adverb can be formed from different word types concrete measurement (201a, 201c), abstract measurement (201b, 201d). The concrete reference of (201c) has been transformed to an abstract, and the pronoun usage in (20Id) is a rarity (Feoktistov Jan. 1996).

(202) Х-нь-Х

(202a) (0080) Кувшинэсь сяворсь, кеверсь масторов ды тусь пельксэнь-пелькс. "The scoop fell, it rolled onto the floor and went to pieces."

With the examples (201a-202a) N. Adushkina and I were able to discern a split in definition. In examples (201a-d) the notion expressed is one of measurement. In examples (202a), on the other hand, there might be an aspectual notion present. To illustrate this we constructed the examples (203-^).

Кармасть симеме пенчень-пенчень. "They started to drink a little at a time (lit. a spoon at a time)." An ongoing process (Adushkina, summer 1995)

Явизь весе ули-паронть пенчень-пенч. "They divided up the wealth spoon for spoon." All of the spoons have been distributed (Adushikina, summer 1995)

In example (204) the wealth has been divided up to the very last spoon — a completed action. In example (203) there is an ongoing process, and no mention is made of finishing the beverage off. As seen above, however, there is a difference of opinion present amongst the speakers of Erzyan — perhaps even three different formulas are present. Abramov's text on the other hand, gives very few examples of this adverbial construction. Maybe, a larger and diverse text sampling could show the usages better. Feoktistov maintained that the variation present probably dialectical (Jan. 1996).

97

3.7 IS THE -т/-ть ADVERBIAL ENDING A FORM OF THE GENITIVE

In a discussion with M. Mosin (spring 1995) I learned that a dialectical form of the adverbial кепе пильгть "bare-foot" is actually кепе пильгсть. According to Feoktistov (MW I: XXXVII) the northwest dialect uses the demonstrative flexion -sY for the oblique cases instead of the -«/'found in the central dialect and literary language.

Abramov uses adverbials of this type in what has been the basis of this corpus. Certain peculiarities can be observed, such as the construction чавонь кедть "empty-handed" in example (205). In ERV this construction is headed чаво кедть but the only example includes the indefinite genitive on the adjective чаво, which in Agafonova's words might refer to a transversion from adjective to noun.

(205) (0086) Тона куроксто велявтсь чавонь кедть. "The other soon returned empty-handed."

In examples (206-7) we are confronted with initial adverbials that have been incorporated into the the marked noun пандо "hill". Example (208) is a suggestion of my own, whereas the second element ланк could be explained by the pronunciation of the consonants [ngt ]. Finally, the back-vowel words would indicate that if this was originally a genitive form such as the one found in Mokshan, then it is no longer, i.e. the palatalization is missing from the [t ].

(0238) Сэденть томбале кись тусь веревбандт. "Beyond the bridge, the road went up-hill."

(0228) Аловбандт тусь седеяк пек. "Down-hill he went all the faster"

(208) (0190) Ялга марто кортнезь снартнесь витеме мелензэ, теевсь мекевланк. "He tried to pull himself together by talking to a friend, but the reverse happened"

98

4. CONCLUSION

The application of the genitive in Erzyan involves both a knowledge of linking vowels and the proper choice of genitive forms — definite versus indefinite. In general, the Erzyan language shows a tendency of avoiding redundant usage of definiteness when the genitive is affixed to nouns, such as proper nouns and pronouns. Additional information is required, such as function — possession versus direct object — when deciding whether the definite or indefinite genitive is required.

The use of the definite and indefinite genitive forms on the possessor have an effect upon the choice of marking for the possessed. There are three different possibilities for marking the possessed — (no marking), definite article, and possessive suffix. On the other hand, not all of the combinations are possible, for example a personal name with an indefinite genitive in the role of possessor can only be followed by a possessee with either a definite article or possessive suffix if the possessee is a subject or object of the finite verb. If, however, this possessee should be in one of the oblique cases, the possessive suffix is removed leaving a pattern "definite article" versus { } for the oblique cases.

The use of the genitive is extensive, and there is little reason to separate the use of the genitive attribute from that of the other genitive constructions. There is, however, reason to distinguish between uses of the absolute and genitive form of nouns in the formation of noun phrases. Hence the use of the genitive in the novel "Исяк якинь Найманов" can be established in accordance with the following presuppositions and rules:

Whereas only a specific referent can be denoted by a noun marked with the genitive, and whereas there is an alternation between definite and indefinite morphologically which is further subjected to cultural notions of definiteness.

I It can be used to express possession of both animates and inanimates regardless of their posistion in the sentence.

99

П It can be used to form genitive attributes from certain verb forms and adverbials. П1 It can be used to show the definite object of a finite verb, whereas the verb can be conjugated in both the definite and indefinite paradigms.

It can be used as a determiner of nonfinite verb forms, whereas it, generally, has the function of the subject of the intransitive verb or the object depending on whether the verb is transitive or intransitive.

It can be used as a determiner of a few nonfinite verb forms, whereas it has the function of the agent.

It can be used in the government of postpositions.

It can be used in the formation of adverbials consisting of a doubled element, and perhaps there are further historical adverbial formations to be found.

Rules IV and V have been distinguished, in order to readily adapt Abramov's writing to the main rules of Erzyan, namely rule IV is prevalent and rule V is quite limited in use.

There are varied opinions on the use of the genitive with regard to its role in possession and its position as an attribute. The genitive can be used to mark the possessor -determiner of a noun phrase. First, we have a very similar pattern when the genitive is used to mark the determiner of a subject or object.

The determiner can be marked with an indefinite genitive (la) or a definite one (lb). If the determiner is marked by an indefinite genitive (la), the possessee subject or object is open to alternation between definite and indefinite marking (2b, 2a). If, on the other hand, the determiner is marked by a definite genitive (lb), the possessee subject or object will be marked by a possessive suffix (2c). If these determiner markers, however, are obscured by the use of a possessive suffix (lc), the possessee subject or object will be marked by a definite article (2b).

In the case of a personal name the indefinite genitive (la) is used, and the subject or object will alternate in its marking with the variations (2b, 2c). A topographical name as a determiner would take the marking (la) with the possessee subject or object alternating in

100

its marking with (2a, 2b). There were no instances of a personal or topographical name in the position of genitive attribute, marked with the definite genitive in the corpus material.

(la) [possessor + -нь]

(lb) [possessor + -нть: -тнень/-тнэнь]

(lc) [possessor + Px]

(2a) [possessed + 0]

(2b) [possessed + сь: -тне/-тнэ]

(2c) [possessed + Px]

Hence example (4) is open to analogy with all nouns, as a marker of the neutral definite phrase, whereas (5) is only possible with a personal names, with which more emphasis falls on the possessor, and such that the possessive suffix can be used on indefinite possessees. Formula (3) is open to all nouns but personal names for the expression of indefinite possessee. Once again, no evidence for emphasized place names was available.

(la) # (2a) [цёрань вайгель]

(la) # (2b) {1922 иень тундось)

(la) # (2c) {Захаронь весе чокшнензэ}

Examples (6-7) are formulas not acceptable in Abramov's texts, whereas formula (8) would be a counterpart to (3-4). This in turn would give us evidence opposed to a formulation for special genitive-looking suffixes used for deriving spatio-temporal adjectives. Instead, these indefinite genitives are merely the neutral marker for the genitive attribute in a noun phrase.

(lb) # (2a) {* цёранть вайгель}

(lb) # (2b) {* цёранть вайгелесь)

(lb)#(2c) {телентьякшамозо], {эрзятненьвиест)

Example (9) is impossible because of the definiteness of "Yelena's mother". One might actually expect a variant like (11) to exist, but perhaps the presence of a previous genitive determiner, i.e. Yelena, renders the formula unfunctionable. As we recall certain

101

grammarians only allowed for the third person singular possessive suffix to be used in a possessive function.

(9) (lc) # (2a) {* Еленань аванзо кудо]

(lc) # (2b) {Еленань аванзо кудось}

(lc) # (2с) {*Еленань аванзо кудозо)

Determiners of adverbial phrases can be marked in the same manner as they are when they are the determiners of subject and object phrases. On the other hand, marking of the possessee adverbials is limited to a few instances with definite articles, but no possessive suffix usage was observed (14). Otherwise the possessee adverbial is bare (12) when it follows a definite genitive determiner. There was a lone situation in which a definite article modifies both the possessor and possessed. Hence the pattern found in (13) is allowable for adverbials whereas its analogous form in (7) is not possible for nuclear elements of the sentence. Could this be more a question of the semantic barrier between "word" and "speech" (see (47a) p 40).

(lb) # (2a) { цёранть кедьстэ)

(lb) # (2b) {лелянть валтнэстэ)

(lb) # (2c) {*теленть якшамостонзо)

The indefinite genitive can be used for relations other than ones of possession, but where should the line be drawn. Common nouns of temporal and spatial reference can be determiners in both the definite and indefinite genitive. Subjects and objects determined by them can thus take markers for definiteness according to the pattern stated above.

There is a difference observed in countable versus noncountable regarding the alternation between the indefinite genitive and the absolute form. The problem of delimiting compound words versus determiner possessed noun phrases is one that appears in the complexities of postposition government, as well, and deserves closer examination.

102

Attributes can be formed from adverbs and adverbial — also gerund — phrases, as well. The difference between the adverbial usage without -нь and the attribute with it is one of present temporary as opposed to general or permanent. The -нь participle might have originally been part of a complementary distribution, in which it was the attribute and the -зь participle appeared in the position of predicative, as well as gerund.

The habeo construction in Abramov's texts is usually shown with a definite determiner and possessed element. Although the usual word order for possession is possessor followed by possessed, the order can be altered for pragmatic purposes. There can also be instance of possessor separated from possessee, for instance the possessed can precede the possessor or various sentence elements can be placed between the possessor and possessee (see (75) p 48).

Group words, such as surnames, in the genitive can be used in much the same way as personal pronouns plural, which are adverbially used for the expression of locative. These genitives are used with verbs in the third person plural and they could be rendered in much the same way as the Finnish construction locative plus passive. Kanayeveilla jo tiedettiin "They already knew at Kanayev's" it could be an instance of ellipsis such as that found in English.

The genitive object is the same as the definite object. It is here that personal and topographical names diverge. Place names are affixed with definite genitive endings, which is a property they exhibit with postpositional government, as well. Furthermore, the genitive object can be used with both the definite and indefinite conjugation of verbs, whereas aspect - topicalization appear to be decisive.

In Abramov's texts there is a difference made in the functions "subject of an intransitive verb", "subject of a transitive verb" and "object of a transitive verb". The gerund апак

103

and the deverbal noun -мо/-ма are the only instances where all three functions can be held by a genitive determiner.

There are generally problems with the definition of the function "subject of a transitive verb", whereas there were no instances of infinitive forms with genitive attributes or possessive suffixes showing both a patient and agentive function. And finally, it is doubtful whether the agentive of a real transitive verb could be shown by genitive determiner, excepting the "dialectical" use of the agentive participle -зь, this of course could be an explanation for everything.

The gerunds -мсто/-мстэ and -зь can take genitive attributes — and with the former possessive suffixes — that denote the functions "subject of an intransitive verb" or "object of a transitive verb". The -мс form makes a different split by acting as a gerund when the genitive determiner preceding it has the function "subject of an intransitive verb". In the corpus, -Me forms of transitive verbs that are preceded by genitive determiners act as infinitive attributes with a quality of modality.

In the corpus text, there is no reason for the terms gerund and infinitive to be associated with the use of the possessive suffixes. The Erzyan language here appears to have the genitive analogy of an ergative system, such that use of the genitive in infinitive constructions marks either the object, or the subject of the intransitive verb. Since possessive suffixes can be used in the same manner as genitive attributes for denoting the functions "object" and "subject of the intransitive verb", their usage can be placed under the same category. Therefore, when instances of the gerunds -мс, -зь, -мсто, -мга take genitive attributes or possessive suffixes, the functions denoted would — according to the previous definition for gerund versus infinitive — show that gerunds with determiner could only be formed from intransitive verbs and infinitives with determiners could only

104

be formed from transitive verbs. This, as seen above, is not the case. Hence the results of R. Bartens (1979:88) must be corrected.

Gerunds appear to favor certain functions according to whether they are formed from intransitive verbs or transitive verbs. Since the intransitive form with a genitive attribute show a subject separate from that of the finite verb, most instances of the gerund will be qualifiers of the entire sentence. The exception to this will be found in such verbs such as "be envious", which show a body part as subject.

The active participle -ця provides further evidence for the argumentation "indefinite genitive determiners for countables" versus "indefinite nominative singular determiners for concepts, noncountables". At the same time, however, it seems to show a genitive attribute only with nouns.

The postpositions with genitive government have been dealt with in Mokshan but there is still plenty of studying to be done concerning the division between two visible groups. This will require a better knowledge of the use of the definite/determinite articles in Erzyan.

Finally, there are certain adverbials which can be formed with the genitive. These can most readily be divided into three groups, two of which would consist of doubled elements. The first two groups could be further divided according to function, whereby the first expresses process and the second completed action. The third, or final, group of genitive adverbials is questionable with regard to its genitive origin, but it is a productive one.

105

Bibliography:

Abramov, Kuzma: Исяк якинь Найманов, Saransk, 1987.

Agafonova, Nina: Fenno-ugristica 8: 3-12. Tartu, 1981.

Alhoniemi, Alho: 06 объекте т.н. номинативного инфинитивав мордовском

языке, 21-30 Turku, 1994. Bartens, Raija: Mordvan, tseremissin ja votjakin konjugaation infiniittisten muotojen

syntaksi. SUST 170: Helsinki, 1979.

Bubrikh, D. V: Эрзя-мордовская грамматика минимум. Saransk, 1947. Chetvergov, Yevgeni Vladimirovich: Сырнень човалят, Saransk, 1995. Ersäläis-suomalainen sanakirja. Jaana Niemi- Mihail Mosin. Turku, 1995. ERV = Эрзянь-рузонь валкс, Moscow, 1993. Feoktistov, Aleksandr Pavlovich: O латентной форме выражения поссессив-

ных отношений в мордовских языках. Congressus Internationalis Fenno-

Ugristarum. Budapest, 1963. GMYa 1962 = Грамматика мордовских языков, Ed. Kolyadenkov, Saransk,

1980. GMYa 1980 = Грамматика мордовских языков, Ed. Dmitri V. Tsygankin,

Saransk, 1980. Kahla, Martti: Mordvan kielen postpositioiden syntaktiset funktiot. Filosofian

lisensiaatin tutkintoa varten, Helsingin yliopisto 1974. Keresztes, László: Chrestomathia morduinica. Budapest, 1990. Kolyadenkov: Грамматика мордовских языков, Saransk, 1962. Kuldurkayev: Кезэрень пингеде, эрзянь раськеде, Saransk, 1994. KGA 1994 = Кузьма Григорьевич Абрамов, биобиблиографический указа-

тель. Saransk, 1994. Ledyaikina, V. A: Синтаксические функции причастия законченного

действия с суффиксом -зь в эрзя-мордовском языке. Финно-угристика

84-95. Saransk, 1980.

106

Martynov, Aleksandr: Даволдо икеле. Saransk, 1992.

MW = H. Paasonens Mordwinisches Worterbuch I-IV. herausgegeben von Martti

Kahla Helsinki 1990,1992,1994 and (the forthcoming fourth volume 1996). Polyakov, Osip Ye: Русско-мокшанский разговорник. Saransk, 1993. Ваймонь лихтибрят, «Мокшеть» и «Якстерь тяштенять» книгасна. Saransk,

1994. Yegorova, A. S: O принципах выделения категории определения в эрзян-

ском языке, Советское финно-угроведение 2: 81-9. Tartu, 1976. Yermushkin, Grigori: Ареальные исследования no восточным финно-

угорским языкам, 1984. Zhuravlyov, Chislav: Овто латко, Saransk, 1993.

107