SUSA/JSFOu 94, 2013

Merja Salo (Helsinki)

Deverbal reflexive and passive in Chuvash1

1. Introduction

The question about reflexives and passives in Chuvash is complicated. Usually, the
grammatical elements of the Turkic languages are quite similar in many ways; for
example, reflexive and passive verbs are kept separate, each having its own suffixes.
Some sources, however, claim that the passive category in Chuvash is formed differ‑
ently than it is in the other Turkic languages, or even that it does not exist in the first
place (Ašmarin 1898: 258–259, Ramstedt 1952: 165, 1957: 149, Serebrennikov 1976:
29, Serebrennikov-Gadžieva 1986: 200). The latest Chuvash grammar does not even
mention the word passive when describing the verbal system (V. Sergeev 2002). Yet,
the reflexive forms are quite uniform throughout the Turkic languages. This study
tries to prove that both of these old Turkic categories do exist in Chuvash, although
the line between them can be blurred and their meanings might overlap. The material
in this study has been taken from grammars, dictionaries, folklore and translated fic‑
tion. Furthermore, the results have been compared to some extent with some of the
languages surrounding it (Tatar, Bashkir, Mordvin and Mari) and also with the most
studied language, Turkish. The analysis presented here is based on sentences, which
has seldom been done in the reference literature used.
Few of the native reseachers of Turkic languages explain the passive. However,
their concept of it can be assumed to be similar to the construction in Russian, German
or English, where the active sentence, e.g. a) Mary slapped John can have two passive
variants: b) John was slapped or c) John was slapped by Mary (Keenan 1985: 243).
In b), the primary actant is demoted from the subject position and replaced by the
secondary actant. The primary actant can be present as an agent, as it is in c). The
foregrounding feature of passive exists in Chuvash, but the backgrounding does not,
which has confused Turkologists for a long time. In Chuvash, even agentless passive
sentences are rare and the passive and reflexive are expressed by two suffixes, which
can be synonymous.

1.  I am indebted to the reviewers and editors of this volume for many helpful comments on the form
and content of the paper, and to Kimberli Mäkäräinen for revising it.
224 Salo

1.1. Historical background

The grammatical tradition in Russia and in adjacent areas deviates from the Western
tradition in that many verbal suffixes are dealt with as voice formatting elements,
instead of being dealt with as parts exclusive to the greater derivational system. It has
been supposed that all of the Turkic languages have five voices – basic, passive, recip‑
rocal, reflexive and causative (Serebrennikov & Gadžieva 1986: 200, Levitskaja 1988:
269). According to Róna-Tas (1998: 75) Proto-Turkic had four voices: a cooperative
(or reciprocal) e.g. kör-üš- ‘see another’, a middle (or reflexive), e.g. kör-ün- ‘become
visible’, a passive e.g. kör-ül- ‘be seen’, and a causative, e.g. kör-t-kür- ‘show’ (which
might consist of two separate causative suffixes).

1.2. Common Turkic passive suffix ‑l-

According to Serebrennikov and Gadžieva (1986: 200–201) the passive suffix ‑l/‑ıl/
‑il already existed in the ancient Turkic languages known to us and these forms can
be found on various literary monuments. The passive suffix can be added to transitive
stems. Levitskaja (1988: 303) points out, however, that an exception does exist: Yakut.
The lateral suffix is very productive in Chuvash, yet it is hardly ever used as a passive.
The original meaning of the suffix ‑l- is to form denominal (probably intransi‑
tive) positional verbs characteristically so that the derived verb expresses the acqui‑
sition of the property or state indicated by the root noun (Levitskaja 1976: 178–179,
1988: 303–304). It is not difficult to notice that the subject is rather passive with these
verbs. Some examples of denominal verbs usually derived from adjectives include
čĕrĕl ‘revive, recover’ ← čĕrĕ ‘living; lively; fresh, new’, and śĕnel ‘be renewed’ ←
śĕne, śĕn ‘new; fresh’ (cf. Appendix for more examples).
There are a remarkable number of works dedicated to finding the prototype of
the suffix ‑l/‑il/‑ul. For instance, Serebrennikov (1976: 29–32) places this marker for
the passive voice into a group of verbs with similar denominal derivational suffixes.
According to Fedotov, the voice suffix ‑ăl/‑ĕl in Chuvash also serves as a derivational
suffix for forming verbs from nouns (Fedotov 1986: 47). Interestingly enough, ‑l- can
also take on a frequentative2 meaning in the Turkic languages, although this is not
very productive (Serebrennikov-Gadžieva 1986: 200). Some other turkologists have
presumed that the suffix originates from the copula verb bol or ol ‘to be’; this ety‑
mology being first proposed by Kazem-Bek in 1846 (cf. also Ščerbak 1981: 107–108,
150, Fedotov 1986: 48). On the other hand, it has been proposed that the passive is
a rather late phenomenon, particularly because Chuvash does not have this feature
(Ramstedt 1952: 165, 1957: 149). Serebrennikov (1976: 30, 33) points out that many
Turkic languages have a lateral frequentative suffix, and that in the Permic languages,

2.  The ‑l- is now considered to be a derivational suffix (slovoobrazovatelʹnyj suffiks).

 Deverbal reflexive and passive in Chuvash 225

frequentative suffixes (Komi ‑ś-, Udmurt ‑śk-) have later developed into reflexive
suffixes. The line between the passive and reflexive is to a certain extent blurred, due
to the fact that stems ending with ‑l currently take a passive suffix ‑n, which can be
the result of dissimilation. Older texts show that even lateral stems were followed by
passives with ‑l (Ščerbak 1981: 1083).

1.3. Common Turkic reflexive suffix ‑n-

The reflexive suffix ‑n/‑ın/‑in can be found in all the Turkic languages existing
today and even in the already extinct Turkic languages too, which can be seen on
the aforementioned monuments. This suffix probably already existed in Proto-Turkic
(Serebrennikov & Gadžieva 1986: 200). The reflexive voice expresses that the action
is directed towards its performer itself as a direct object. The reflexive suffix is mostly
added to transitive verbal stems. It is assumed that, at first, it had been both a deverbal
and a denominal suffix at the same time (Ramstedt 1952: 169, 1957: 152).
In some cases, the reflexive verbs can be derived directly from adjectives, e.g.
pušan ‘be freed, released from’ ← pušă ‘empty, free’, ăšăn ‘warm oneself, warm up,
heat up’ ← ăšă ‘warm, hot; heat; warmly’, sivĕn ‘get cold[er], get cool[er]’ ← sivĕ,
siv ‘cold, frost’ (cf. Appendix). Surprisingly, this development has not been accepted
by all scholars. Levitskaja (1976: 166–167) has even invented transitional verb forms
similar to the adjectives above: (*puša-, *ăšă-, *sivĕ-) based on findings in other
Turkic languages, as has been pointed out by Fedotov (1986: 48).
The origin of this suffix has been traced back to an incorporated 3rd person
pronoun that was then connected to the possessive form an of the pronoun ol. It is
generally known that the Turkic languages have preserved traces of the 3rd person
pronoun in the form in ~ ın, which had not only the pointing function of ol, but also
the distincting. The situation is quite the same as it is in the Slavic languages and
other languages, too. In Russian, the reflexive suffix has two phonologically condi‑
tioned allomorphs ‑sja and ‑sʹ that originate with the accusative form of the reflexive
pronoun sebja (Lehmann 1995: 44–49). It is only quite recently that this pronoun has
been firmly attached to the verbal stem. The various Slavic languages are in different
stages of development on this matter and in Russian, it occurs exclusively as a verbal
suffix. It cannot be ruled out that the examples from the Turkic languages are a result
of contact with their foreign neighbours (Ščerbak 1981: 111–112, with references).

3.  Referring to Osttürkische Grammatik … by C. Brockelmann, which is not used in this study.
226 Salo

1.4. Previous work on the Chuvash reflexive and passive

1.4.1. 19th century grammars

The first scientific Chuvash grammar (and third one ever printed) De lingua tschuwa­
schorum was written in the beginning of the 1840s by Wilhelm Schott (1802–1889), a
German orientalist, sinologist and professor at the University of Berlin. The booklet
was tiny, only 32 pages long, but it gave a short description of Chuvash phonetics
and morphology and compared this to the other Turkic languages based on regular
sound correspondences. As text material, he used the four Gospels translated under
the leadership of the bishop of Kazan in 1820 (Benzing 1959: 699). An example of
verbal derivation as given by Schott is as follows:

(1) onda oz-il-tschi-s wul-sam-yn koz-sam (Schott 1841: 31)

then open-ref-pret-pl3 (s)he-pl-gen eye-pl
‘tunc aperti sunt eorum oculi’
‘then their eyes were opened’ (Luke 24:31)

In this context, he revealed that the medium (= reflexive) and passive are equal in
form in Chuvash and mentioned that the deverbal suffix ‑tar was used to express tran‑
sitivity. Schott did not need to write his grammar from scratch; he actually had two
predecessors4. His main source was the second printed Chuvash grammar Načertanïe
pravilʺ čuvašskago jazyka i slovarʹ, sostavlennye dlja duhovnyhʺ učilišč Kazanskoj
eparhii 1836 goda, published anonymously, although this is commonly attributed to
Viktor Petrovič Višnevskij (1804–1885), dean and teacher of the Religious Academy
in Kazan. Despite its many mistakes and shortcomings, Višnevskij’s book served as a
textbook for students and scholars for decades afterwards (Alekseev 1970: 207–220,
Alekseev & Sergeev 1988).
Another of the earliest mentions of the passive in Chuvash is found in one of
the classics of general linguistics Über das Passivum. Eine sprachvergleichende
Abhandlung, written in 1861 by Hans Conon von der Gabelentz (1807–1874). The
author was able to create a typological classification of the passive construction by

4.  The first Chuvash grammar Sočinenïja, prinadležaščïja kʺ grammatikĕ čuvašskago jazyka was
printed at the printing house of the St Petersburg’s Academy of Sciences in 1769. This has been reliably
attested with many quite recently found documents by Dimitriev (1967). In the literature, it is often
mentioned erroneously as having been published in 1775 and Egorov (1951: 86) incorrectly assumes
that Sočinenïja was reprinted at that time. In all likelihood, it was created by a group of missionaries
under the command of Veniamin (1706–1782) (born Vasilij Grigorʹevič Puček-Grigorovič, Ukrainian
of noble birth), the bishop of Kazan and Svijaž, who had in 1769 spent 27 years near the Chuvashes
(Dmitriev 1967: 159). Supposedly, this booklet of 68 pages served as the model for the grammars of
Mari and Udmurt printed in 1775. (Šamraj 1955, Benzing 1964: 842, Alekseev 1970: 203–207, Fedotov
1987: 10–12). It was reviewed in Germany in 1770. Although the grammar was based on grammars of
classical languages, it still had a great impact on the field by convincing many scholars of Chuvash’s
relationship to the Turkic languages. It is especially valuable due to its rich vocabulary, some of which
ended up disappearing later. Višnevskij’s grammar is mainly based on the first one (Egorov 1949:
117–118, 1951: 90–92).
 Deverbal reflexive and passive in Chuvash 227

successfully using materials from different languages. It is generally known that this
polyglot was able to read texts in more than 80 different languages. In his linguistic
studies, he dealt with more than 200 languages from many corners of the world. His
masterpiece Über das Passivum deals with 200 languages that have a passive, its
usage and its synchronic connections to other categories. He discussed the Chuvash
passive in the chapter entitled “Passivum durch eine Reflexivform”, where he notes
that, unlike the other Turkic languages, Chuvash has two suffixes for the passive: l
and n. He writes:

“In den türkisch-tatarischen Sprachen scheinen ebenfalls die Formen für

Reflexivum und Passivum ursprünglich identisch zu sein, wenigstens ist die
Bildung durch n beiden gemeinschaftlich, während die Bildung durch l meis‑
tens nur dem Passivum zukommt. Eine Ausnahme macht hier vielleicht nur
das Tschuwassische, das beide Formen für das Passivum hat, dabei aber, wie
Schott (De Lingua Tschuwaschorum p. 31) ausdrücklich bemerkt, Media und
Passiva der Form nach nicht unterscheidet, also beide Formen ebenso gut für das
Medium oder Reflexivum wie für das Passivum gebraucht” (von der Gabelentz
1861: 523; he had no examples of the voices in Chuvash).

Soon after that, József Budenz (1836–1892) published a grammatical study of

Chuvash based on his own fieldtrips in 1845–1848, Antal Reguly’s (1819–1858) notes
from 1843, and various grammars and dictionaries (Fedotov 1987: 23–25). He uses
four pages to discuss the derivation of words, only one of which he dedicates to a
brief mention about the derivation of verbs and that the derivational suffixes n, en, in
are used to form passive verbs, such as oldala-n-as ‘be dissappointed, be deceived’
← oldala-s ‘deceive, cheat, lie’ [probably a dialectal form, according to the modern
orthography of the literary language: ultalanas ← ultalas] (Budenz 1863: 66–67).
One of the leading turkologists and a specialist in Chuvash at the turn of the
century, Nikolaj Ivanovič Ašmarin (1870–1933), wrote in his grammar from 1898 that
Chuvash has no passive voice, only a reflexive voice. All the same, he found some
traces of the passive in special expressions, such as:

(2) Ут окҫ-и тат-ăл-ман-ха5 (Ašmarin 1898: 259)

horse money-px3sg break-refl-neg.ptc-emph.part
‘Denʹgi za lošad’ ešče ne zaplačeny (= ne otorvany).’
‘The money for the horse has not yet been paid.’

He states that some reflexive verbs are not derivations as they do not have a root word,
such as tapran ‘move (itr. from its place); begin, rise’, which has the parallel form
taprat ‘move (tr. from its place); start up, begin’; pătran ‘become muddy’, with the

5.  The numbered examples have been kept as much as possible near the original form, but in the run‑
ning text they have been unified and written in the Latin alphabet, the derived and stem verbs presented
under the examples are presented in the same way.
228 Salo

parallel form of pătrat ‘make muddy, surge; stir; dig, root, turn over, muddle’; and
pălxan ‘get worried; be worried; rebel; become muddy or misty; move (itr.), sway’,
which does not have a parallel form. In Ašmarin’s grammar, the nature of deverbal
verbs with the suffix ‑n can be best seen when the root verbs are transitive, e.g. parăn
‘submit, surrender’ ← par ‘give, present; grant’, tıtăn ‘begin (itr.), start; stay, stand;
stammer’ ← tıt- ‘hold; take; catch; reach; have, own; use, control’, and perĕn ‘hit
oneself; touch’ ← per ‘hit; throw, shoot’ (Ašmarin 1898: 258–259, cf. Appendix for
more examples).
The author reveals that the reflexive nature of the suffix ‑l can be seen in some
verbs that have roots ending in the consonants r, s, ś or t, e.g. xuśăl ‘break (itr.), get
broken, be cut off, fold up’ ← xuś ‘break, cut, fold’ and tatăl ‘break (itr.), tear; end’
← tat ‘break (tr.), tear (to pieces), cut, saw; pick, collect’ (Ašmarin 1898: 260–261).

1.4.2. 20th century grammars

The next remarkable postwar grammar was printed in 1957 in Cheboksary. In the
foreword, the book is said to be the result of collective work, but in fact, only three
men participated in writing it, each one having written from two to six chapters. The
most troublesome part in describing Chuvash, as with so many other languages as
well, fell to Vasilij Georgievič Egorov (1880–1974). He had studied in the Chuvash
Pedagogical College and Religious Seminary in Simbirsk, at the Religious Academy
in Kazan, and finally, at the University of St Petersburg under the tutelage of many
famous scholars (Fedotov 1987: 66–67). Unlike his fellow researchers, Egorov sur‑
prisingly sees in the reflexive suffix ‑n relics of the former 2nd person pronoun and
divides the suffix into several groups based on its meanings (similar to the way this
has been done in Russian grammars, too):

Proper reflexives, where the subject is also the object of action:

śapăn ‘take a sauna bath, slap o.s. with a sauna whisk’ ← śap ‘hit’,
parăn ‘surrender, give up’ ← par ‘give’,
muxtan ‘boast’ ← muxta ‘praise’,
saltăn ‘undress’ ← salt ‘open, take off, let go’.

Reflexives expressing changes in subject’s physical or mental state:

savăn ‘be happy, be pleased’ ← sav ‘love, like’,
śěklen ‘rise, get up’ ← śěkle ‘raise, lift up’,
puśtarăn ‘gather (itr.)’ ← puśtar ‘collect, gather’,
tavrăn ‘return (itr.), go back, turn back’ ← tavăr ‘return (tr.), bring back’.

Reflexives, where the subject experiences something it has no effect over:

avăn ‘bend (itr.)’ ← av ‘bend (tr.)’,
kurăn ‘be seen’ ← kur ‘see’,
üsen ‘breed, increase (itr.)’ ← üs ‘grow (itr.)’ (Egorov 1957: 167–168).
 Deverbal reflexive and passive in Chuvash 229

The passive verbs have been divided into two groups based on their meaning:

Reflexive-passive, which is said to be weakly developed:

tıtăn ‘get arrested’ ← tıt ‘hold; take; have’,
vitěn ‘get covered’ ← vit ‘cover’,
šumlan ‘get counted’ ← šumla ‘count’.

Reflexive-medial verbs including all ‑l cases, stems ending with ‑r, ‑s, ‑ś or ‑t:
xuśăl ‘be broken’ ← xuś ‘break’,
sarăl ‘broaden, widen (itr.)’ ← sar ‘broaden, widen (tr.)’,
věrěl ‘catch [a] cold after being ill’ ← věr ‘blow’ (Egorov 1957: 169–170).

In the next decade, Ivan Andreevič Andreev (1928–), a pupil of Egorov, presents four
voices: 1) basic kăškăr- ‘shout’, xăvala- ‘follow, chase’, pulăš- ‘help’; 2) reflexive-
passive, e.g. xuśăl- ‘break (itr.)’, păsăl- ‘go bad, be spolt’, ěślen- ‘go well’; 3) recipro‑
cal śapaś- ‘fight (each other)’, pallaš- ‘get acquainted with (a person/thing)’, těkěš-
‘butt, gore, toss’, perkeleš- ‘shoot at each other’; and 4) causative (Russ. ponuditelʹnyj)
kultar- ‘make laugh, entertain’, čakar- ‘back, push away’, tart- ‘make run’. The last
one can be added to the other voices, as can be seen in vit-ěn-ter- ‘make cover one‑
self’, which has both a reflexive-passive and a causative suffix. Slightly confusing is
the fact that the author adds that all of the voice-formatting suffixes can act as word-
formatting suffixes: tap-ăn- ‘attack’ ← tap- ‘kick, push, hit’, ut-tar- ‘walk’ ← ut-
‘walk’ (-tar seems to be semantically empty with this stem) (Andreev 1966: 54–55).
In 1975, Gennadij Emelʹjanovič Kornilov’s (1936–) expansive article on the
reflexive and passive voices in Chuvash was published. He writes that the passive
constructions can have two or three participants and that they do not correspond
to passive structures in other languages such as Russian. The closest are reflexive
structures, which have two participants: śurt tăvănat ́ cf. dom stroitsja ‘the house is
being build’, pulă tıtănat ́ cf. ryba lovitsja ‘the fish is being caught’. These now exist
in the literary language and have been taken from the lower dialects; they seldom
occur in the upper dialects. Reflexive verbs can be formed from transitive or intransi‑
tive stems: üs- ‘grow’ → üsěn- ‘go off, pass off’, śit- ‘come, reach’ → śitĕn- ‘grow
up, ripen’ (Kornilov 1975: 46–47). He provides extremely detailed information on
how reflexive and passive suffixes, with their numerous allomorphs, are used to form
verbs from denominal and deverbal stems. According to him, the Turkish type of
constructions with three participants has developed under Arab influence; this can be
seen in (3) in Tatar:

(3) komanda oficer tarafınnan6 bir-el-de (Kornilov 1975: 72)

order[s] officer pop‘from the side of’ give-pass-past.3sg
‘komanda otdannaja oficerom’
‘order[s] was given by the officer’

6.  Even the word itself is a loanword from the Arabic taraf ‘side’ (Kornilov 1975: 72).
230 Salo

There is also a way in Chuvash to express passive structures with three participants:
the comitative or instrumental suffix ‑pa/‑pe ~ ‑pala/‑pele ~ ‑palan/‑pelen, which
corresponds to the Ancient Turkic postposition bilä, bilän ‘with’, as seen in Chuvash
(4), and Tatar (5):

(4) śil-pe(len) uś-ăl-na (Kornilov 1975: 72)

wind-instr(pop‘with’) open-pass-perf.ptc
(5) žil belän ač-ıl-gan išek (Kornilov 1975: 72)
wind pop‘with’ open-pass-IIpast door
‘the door was opened by the wind’

He has noted that sometimes the reflexive suffix can have a passive meaning, which
can be seen from the following example:

(6) ală supăń-pe lajăx śăv-ăn-at ́ (Kornilov 1975: 73)

hand soap-instr well wash-pass-pres.3sg
‘the hand(s) can be cleaned well with soap’7

Sometimes the passive and reflexive suffixes in Chuvash are identical in meaning:
pěkěr-ěl- or pěkěr-ěn- ‘bend, crook (itr.)’, which has also been borrowed by Mari as
pügirn-aš (Kornilov 1975: 73–74).
In a concise account of Chuvash in a relatively new reference book, the reflexive
voice is said to be formed with ‑(Ă)n, e.g. śăvăn- ‘wash oneself’ ← śu- ‘wash (tr.)’,
although transitive verbs ending in ‑t, ‑s, ‑ś, ‑r form the reflexive with ‑(Ă)l, e.g. uśăl-
‘open (itr.)’ ← uś- ‘open (tr.)’. Verbs with ‑(Ă)n sometimes function as passives, e.g.
śırăn- ‘be written’ ← śır- ‘write’. The active voice lacks the suffix, cooperative-recip‑
rocals are formed with ‑(Ă)ś, and causatives with ‑(t)tAr (Clark 1998: 443). The same
distribution can be found in many Chuvash textbooks and minor works on grammar
(V. Sergeev 2000, L. Sergeev 2004).

1.4.3. 21th century grammar

In spite of its modest layout, Vitalij Ivanovič Sergeev’s (1942–) grammar from 2002 is
quite comprehensive, almost 300 pages in all. The contents are detailed, the verb sec‑
tion consisting of approximately 65 pages. The book is clearly meant for native speak‑
ers of Chuvash, since most of the examples have not been translated into Russian. Not
all of the voices can be formed with all of the verbal stems; there are remarkable lexi‑
cal restrictions. Morphologically, the voice in the Turkic languages comes between
the concepts slovoizmenitelʹnyj and slovo- or glagoloobrazovatelʹnyj, and a new

7.  According to others this is not proper Chuvash. Nonetheless, even though it might not be the best
example, the two main meanings for śăvăn- in dictionaries are passive and reflexive.
 Deverbal reflexive and passive in Chuvash 231

term osnovoobrazovatelʹnyj has even been created for it (Grunina 1987: 14). In some
textbooks of higher education, the term reljacionnyj has been used. Traditionally,
Chuvash lexicography has treated voice forms as word forming and included them in
dictionaries (V. Sergeev 2002: 234).
Nowadays, the passives described in some earlier grammars are no longer
accepted. Structures such as the ones in (7) have developed under Russian influence
and they are not considered to be proper Chuvash (V. Sergeev 2002: 236–237):

(7) Kěneke xitre vula-n-at ́ (V. Sergeev 2002: 236–237)

book nicely read-pass-pres.3sg
‘Kniga xorošo čitaetsja.’
‘The book is nice to read.’

1.4.4. Compound derivatives connected to reflexives

According to Andreev, the denominal verbal suffix ‑lAn is a compound derivative

suffix from ‑lA and ‑ăn, ‑ěn, ‑n, (the latter forming mainly intransitive verbs, e.g.
jěpe-n- ‘to get wet, damp’). The suffix ‑lAn forms reflexive verbs, such as av-lan- ‘get
married (of a man)’ (← *av ‘house’) (Levitskaja 1976: 166), cf. Turkish ev-len ‘id.’
(← ev ‘home, house’), ikkě-len ‘hesitate’ (← ikkě ‘two, twice’) (Andreev 1966: 54,
Fedotov 1986: 57). The suffix ‑lAn is traceable back to all of the Turkic languages. At
least one minor study has been dedicated to these verbs in Chuvash (Orlova 1976).
The verbs formed with the suffix ‑lAn are quite common and are one and a half
times more common than verbs with the suffix ‑lA. The ‑lAn derivatives have been
divided into six groups, all having nouns as their stem word, mainly substantives and
adjectives expressing the emergence of a property or the changeover to a new state.
However, in most cases, the ‑lA derivatives are formed from words denoting instru‑
ments and objects (Orlova 1976: 157–158).
The suffix ‑lAn has been borrowed into Mari8 from Chuvash or Tatar in many
denominal verbs, e.g. sěmsěr-len- → sümsır-lan-aš or xajar-lan- → ajar-lan-aš ‘get
provoked/angry/worried’. Hesselbäck (2005: 113) writes that it is certain that this has
been copied from the Turkic languages, since it mainly occurs with adjective and
substantive stems of Turkic origin (Alhoniemi 1985: 162, 1993: 151, without mention‑
ing the suffix’s etymology; Fedotov 1986: 57–58). As the analyzed part of this study
concentraites on deverbal verbs, ‑lA and ‑lAn verbs have more or less been left out.

8.  In Hill Mari, the suffix ‑lAn is (more) productive.

232 Salo

2. Materials and methods

2.1. The data

The first textual sources used was Gebräuche und Volksdichtung der Tschuwassen
(= Gebr), collected in 1900 by Heikki Paasonen (1865–1919), but published only
1949 by other peoples9. This material comes mainly from Novoye Yakushkino in
the Buguruslan district, Samara government. The second source used was John R.
Krueger’s Chuvash Manual from the year 1961 (= ChM). These have been very con‑
venient due to the German or English translations. A comprehensive Chuvash–Russian
dictionary Čuvašsko-russkij slovarʹ (= ČRS), a smaller Chuvash–Finnish dictionary
Moisio–Fomin–Luutonen: Tšuvassilais-suomalainen sanakirja (2007) (= TšSS), and
a small reference book by Ašmarin: Sbornikʺ čuvašskihʺ pĕsenʺ10 (1900) (= Sbor),
with material from 11 villages in 3 governments, have also been used. On the referee’s
recommendation, the electronic corpus Pavlik Morozov (= PaMo) has been studied.
This originally Russian story by Vitali Georgievič Gubarev of a young pioneer and
his hard fate in the 1930s has been translated into many Finno-Ugrian languages and
Chuvash, too. The electronic corpus consists of 1,608 sentences on 50 pages. These
parallel corpora are made by and kept in the Research Unit for Volgaic Languages at
the University of Turku.

2.2. Methodology

For the sentences in this study, I have applied valence roles from case grammar and
examined the relations between deverbal verbs having ‑n or ‑l suffix in order to illus‑
trate the syntactic changes caused by the suffix. For the analysis, I have used the same
roles as distinguished in Geniušienė (1987: 39–41). In the passive sentences, I have
distinguished between three roles for the semantic subjects or first actants: agentive,
experiencer, and causer (or force).
At the top of the subject hierarchy is the agentive – an animate, who is con‑
sciously controlling his/her actions. The experiencer is an animate, who participates
in an action, but does not influence it, the action influencing its consciousness instead.
The causer is a moving meteorological force, not materia. The fourth role, neutral,
occurs only in automative expressions, it is a participant in a space or process, but
the action does not affect it considerably (applied to the Ob-Ugrian languages, see
Kulonen 1989: 11). The term neutral is approximately the same as Anderson’s (1971:
37) nominative. The fifth role, actor, occurs only in reflexive utterances, when the
subject and object conflate.

9.  Some parts of the collection have been sent some years later to Finland by Paasonen’s (only?)
informant, more details in (Salo 2010b: 70, 88–89). Paasonen’s influence on other turkologists is dis‑
cussed by Fedotov (1987: 44–46).
10.  In the examples, the original orthography has been slightly altered.
 Deverbal reflexive and passive in Chuvash 233

For semantic objects or second actants, I have two roles: goal/patient and con-
tent. One role is sufficient for the semantic dative: benefactive, a living being, who
benefits from the action. Sometimes other roles are necessary, too: locative, which
expresses motion into/towards, or the location of the referent or time, source, which
expresses the starting point of a motion or an indicator of origin that can express
time, too. Finally, the role instrument, the inanimate participant of action, has also
been included. These roles have been used in describing basic sentences in Finnish
(Hakulinen & Karlsson 1979: 102–104). I have already used these roles to analyze
verbs in Mordvin (Salo 2006b, 2010a) and Mari (Salo 2006a).
In order to better analyze the Chuvash verbs that have an ‑l or ‑n suffix, I have
applied the tripartite division of derivational suffixes created by Kangasmaa-Minn
(1982: 43–44). She has started to use the terms changer (Finn. muuttaja), transformer
(muuntaja) and modifier (modifioija). changer moves words from one main category
to the other: nouns into verbs, and vice versa. It might be the same as the Russian term
slovoizmenitelʹnyj. transformer operates in a single category causing changes to the
argument structure, e.g. changing intransitive verbal stems into transitives and transi‑
tives into intransitives. The number of (obligatory) arguments or actants decreases
or increases. It appears to be the same as the Russian term slovoobrazovatelʹnyj.
The third group, modifiers do not usually cause any great changes in the behaviour
of verbs; they just add some nuances. This is the same division I have used in my
Master’s Thesis on Erzya Mordvin derivatives (Salo 1988).

3. Data analysis

3.1. Passives

All Chuvash grammars agree that passives occur very infrequently. The passive is
only used if an agent is not present (Benzing 1943: 87). Often the passive meaning
is combined together with the reflexive in describing derivational suffixes. Benzing
has searched through all 17 volumes of Ašmarin’s dictionary for passive examples
and based on his findings, he went on to state that the passive is rare: kas-ăn- ‘rubit'
sam soboju; cut oneself’, cf. kas-ăl- ‘byt' razrezyvaemym; be cut’. Furthermore, he
states that the latter often has more of a reflexive meaning than a passive meaning
(Benzing 1959: 720). In an active sentence, the first or primary actant appears in the
subject position. When the primary actant is demoted from the subject position, its
place is occupied by a secondary actant. In Chuvash, the agentive is always hidden in
passive sentences, although it can be inferred that the actant is animate. In Mari, the
situation is similar to this. In Mordvin, however, the agent in the dative is possible,
yet not very common (Salo 2006a: 333, 2006b: 172–176). The surface subject has the
semantic role of goal/patient, as in (8) to (11). With the passives, the derivative suffix
is a transformer.
234 Salo

(8) Sămax-sem ujrăm sasă-sem śine pajlan-aććĕ. (ChM 200, 249)

word-pl separate sound-pl pop(into, on) be.divided-pres.3pl
‘Words are divided into separate sounds.’

pajlan ‘divide (itr.), be distributed’

← pajla ‘divide (tr.), share, distribute’
← paj ‘part, share; branch, section; detail’

(9) tărn-i külěn-ně, te-t, śerźi (Gebr 185)

crane-px3sg get.harness-perf.ptc say-pres.3sg sparrow
akapuź-ě tıt-nă, te-t.
plough-px3sg take-perf.ptc say-pres.3sg
‘Der Kranich war vorgespannt, der
Sperling faßte an der Pflugsterze.’
‘The crane was harnessed, says, the sparrow took
ahold of the plough handles, says.’

külěn ‘get harnessed‚ go (enter) into’

← kül ‘harness (verb)’

Sometimes the dictionary says that the ‑an verb has a passive meaning, as in (10):

(10) ҫăка йывăҫ лайăх чавăн-ать (ČRS 576)

lime.tree wood good be.hollowed-pres.3sg
‘the wood of a lime tree is good for hollowing out’

čavăn ‘be dug out, be hollowed’

← čav ‘dig, scratch, paw, hollow; claw, scrape’

(11) šămi-zem pir śir-e puśtar-ăn-nă. (Gebr 192)

bone-pl one place-dat/acc collect.-pass-perf.ptc
‘Die Knochen waren an einer Stelle zusammengekommen.’
‘The bones were collected into one place.’

puśtarăn ‘collect, gather (itr.), be collected, be gathered’

← puśtar ‘collect, gather (tr.)’

For the verb in (11), only the reflexive reading is indicated (TšSS: 141): pěr śěre
puśtarăn ‘gather (itr.) together in one place’ and śula tuxma puśtarăn ‘get ready for
a journey’. Another source gives more readings including several passives: tırpul
xăvărt puśtarănat ́ ‘the corn shall be reaped quickly’ and větě śırla čas puśtarănmast ́
‘small berries are not picked quickly’ (ČRS: 319).
 Deverbal reflexive and passive in Chuvash 235

3.2. Automatives

In automative sentences, the agentive is never present. The occurrences arise sponta‑
neously. The majority of Chuvash verbs with an ‑n or ‑l suffix represent a specific type
of agentless passive. For this group, Kulonen (1985: 290) has started to use the term
automative for the active-passive axis alongside the reflexive, as reflexive and passive
are often loosely used to refer to several different semantic categories. In automative
events, the agentive is absent from both the deep and surface structure. The only act‑
ant of this type of verb, the subject, does not control the event or state. Many automa‑
tive verbs can be derived from adjectives, e.g. măkal- ‘become blunt, dull; become
stupid’ ← măka ‘blunt, dull; stupid’, śutal ‘twinkle, shine; lighten; dawn, brighten up’
← śută, śut ‘light; bright, enlightenment, education’, tasal ‘become clean, be purified;
dissappear, be lost’ ← tasa ‘clean, tidy’ (Levitskaja 1988: 303–304). (Depending on
the arguments, the reading can also be passive.) With nominal stems the derivative
suffix is a changer, with verbal stems being transformers. There are two kinds of
subjects for the types of verbs: experiencers (12, etc.) or neutrals, as in śĕśĕ măkalnă
‘the knife became blunt’ (TšSS 97).

(12) Хӗр-не хора ан кал-ăр, (Sbor 42)

girl-dat/acc black neg.imp.part say-imp.2pl
Ар-па вырт-сан сарал-ать.
husband-instr lie-crd.ger 2 turn.yellow-pres.3sg
‘Don’t say that the girl is black, when she sleeps
with (her) husband, she turns yellow.’

saral, sarăx ‘turn yellow’

~ sarat ‘make yellow’
← sară, sar ‘yellow’

Cognitive and mental verbs form an important group of automatives. Although the
first actants are experiencers, they are not visible in the surface structure and are hid‑
den somewhere in the background. In these cases, the second actant is the content,
which does not participate in any way in the situation expressed by the verb. Often
an argument in a local case is also present and the source in the ablative case (13) or
locative (14).

(13) Ката-ран ҫутă корăн-ăть.  (Sbor 41)

far-abl light be.seen-pres.3sg
‘Far away, a light is visible.’

korăn, kurăn ‘be seen, become visible’

← kor, kur ‘see, look; experience, endure, hold, bear’
236 Salo

(14) Śül-te, pĕlĕt aj-ĕnče // (ChM 212, 262)

high-loc cloud under-loc.px3sg
tărri jur-ri iltĕn-et,
lark song-px3sg be.heard-pres.3sg
‘In the sky under the clouds // the song of the lark can be heard;’

iltĕn ‘be heard’

← ilt ‘hear’, metathetic itle ‘listen; obey’

The following examples have an external participant in the instrumental case ‑pa/‑pe,
which can only be considered a reason, not an agent. In some examples from Mari in
an earlier study, I considered the wind to be the only representative of the force agent
(Salo 2006a: 335–336). This could also be the case in Chuvash, too, as can be seen
in examples (15) and (16). Nonetheless, they are borderline cases. More examples
of instruments: ujsem jurpa vitĕnčĕś ‘the fields were covered with snow’ (ČRS 82),
xĕvelpe xĕrtĕn ‘tan in the sun’ (ČRS 555). In example (17), the eyes in the instrumen‑
tal case, are an instrument.

(15) epĕ sasă-pa văran-t-ăm (Gebr 120)

I voice-instr waken-pret-1sg
‘Ich erwachte, als sie rief.’
‘I woke up, when she shouted.’

văran ‘wake up (itr.)’

~ vărat ‘wake up, revive; stay awake’

(16) Ҫил-пе пӗрле ҫӗмӗрт ҫыхлан-ать (Sbor 5)

wind-instr together bird.cherry get.tangled-pres.3sg
‘The leaves of the bird cherry are tangling up in the wind.’

śıxlan ‘get mixed, get tangled’

? ← śıx ‘knit, twist; tie, knot’

(17) Kuś-ĕ-sem-pe păx-nă čux //  (ChM 214, 263)

eye-px3sg-pl-instr look-perf.ptc pop(when)
kaččă-n čĕr-i śĕklen-et.
young.man-gen heart-px3sg surge-pres.3sg
‘When she looked with her eyes // men’s hearts surged.’

śĕklen ‘rise, ascend; grow; start, begin, become inspired’

← śĕkle ‘raise; carry, bear; stimulate, encourage’

In (18), the ‑n and ‑l suffixes, here used with the same verbal stem, are even mutually
exchangeable in the given context.
 Deverbal reflexive and passive in Chuvash 237

(18) йывăҫ-ҫем ҫил-пе тайăл-аҫҫӗ / тайăн-аҫҫӗ (ČRS 437–438)

tree-pl wind-instr bend-pres.3pl / bend-pres.3pl
‘The trees are bending in the wind’

tajăl, tajăn ‘bend, lean (itr.), sway, sink down’

← taj ‘weigh; bend, lean (tr.), press down’

In the following two examples, the door opens and closes, probably by itself. Actually,
the involvement of a hidden agent cannot be excluded. But in (21), the tree fell down
all by itself.

(19) śapla kala-san-ax alăk-ĕ te uśăl-nă… (Gebr 108)

so speak-crd.ger 2-emph.part door-px3sg and open-perf.ptc
‘Kaum hatte er diese Worte gesprochen, so öffnete sich das Tor, …’
‘No sooner had he spoken these words than the door opened, …’

uśăl ‘open (itr.); start, be established; be cleaned, air; freshen up’

← uś ‘open; start (to do something), establish’

(20) Прийомăн алăк-ӗ – йӗс алăк, … (Sbor 11)

consulting door-px3sg brass door
Епĕр кӗр-се кай-сассăн хопăн-ать.
we enter-crd.ger go-crd.ger 2 close-pres.3sg
‘The consulting door – the brass door, … when we enter, it closes.’

xopăn‘close (itr.)’
← xop ‘close (tr.)’

(21) juman-ĕ xăjĕssĕnex kasăl-sa  (Gebr 109)

oak-px3sg by.itself be.cut-crd.ger1
vaklan-sa vırt-nă.
break-crd.ger1 lay-perf.ptc
‘Die Eiche fiel von selbst um und lag in Stücken auf der Erde.’
‘The oak fell down by itself and lay chopped up into pieces on the ground.’

kasăl ‘be cut, be chopped; be spoilt’

← kas ‘cut, chop, tear’
vaklan ‘get smaller, get chopped, crumble (itr.)’
← vakla ‘make smaller, chop, crumble’
← vak, vakă ‘small, little, minor’

In (22), the subject is inanimate, but with an animate subject, the verb would be
reflexive:
238 Salo

(22) Шерте карта-на пыр-са ҫапăн-чĕ (PaMo 958)

pole fence-dat/acc go-crd.ger1 bump-past.3sg
‘A pole bumped into the fence.’

śapăn ‘bump [into], crash into’

← śap ‘hit, slam, slap’

In (23), the reason is expressed with the ablative case. In (24), the tree as a subject is
situated between experiencers and neutrals; it lives and dies and reacts to climate
and seasons. In (25), the source is in the ablative case; in (26), the locative is in the
dative/accusative case.

(23) kurăk-sem jĕpe-ren pičče (Gebr 116)

grass-pl moisture-abl elder.brother
jĕpen-se śit-r-ĕ.
become.wet-crd.ger1 perf.vb-pret-3sg
‘Der Bruder wurde von dem feuchten Gras ganz naß.‘
‘The brother was quite wet from the damp grass.’

jĕpen ‘get damp, get wet, grow moist’

~ jĕpet ‘water, wet, moisten’
← jĕpe ‘wet, moist, damp; dampness, humidity’

(24) хăва йăвăҫ авăн-ать; (Sbor 50–51)

willow tree bend-pres.3sg
‘The willow tree bends;’

avăn ‘bend, break, crook (itr.)’

← av ‘bend, break, crook, twist (tr.)’

(25) Атте-пе анне-рен уйăрăл-т-ăм;  (Sbor 48–49)

father-instr mother-abl leave-pret-1sg
‘I said farewell to my father and to my mother;’

ujărăl, ujrăl ‘be separated, fall apart; come loose/off/out’

← ujăr ‘divide, separate; determine, nominate; identify, recognize’

(26) Шыв-а йар-сан кĕмĕл ирĕл-мĕ. (Sbor 22)

water-dat/acc put-crd.ger 2 silver melt-neg
‘Silver doesn’t melt if you put it in the water.’

irĕl ‘melt, dissolve (itr.)’

?← ir ‘mix, crush, knead’
 Deverbal reflexive and passive in Chuvash 239

3.3. Reflexives

In the reflexive utterances, the subject (agentive) and object (patient) have been con‑
flated, and the only obligatory actant is actor, which fills both roles. Benzing’s (1959:
720) reflexive examples are e.g. śăvăn- ‘wash oneself’ ← śu- ‘wash’, vitĕn- ‘cover
oneself’ ← vit- ‘cover’. The reflexive derivational suffixes are transformers. A typi‑
cal example is tumlan- ‘dress’ (27). Often, when the subject changes its location, a
locative in the dative(/accusative) is present, as in (28).

(27) xaj kaxal tumlan-sa tux-nă ta… (Gebr 108)

that lazybones dress-crd.ger1 go.out-perf.ptc and
‘Der Faulenzer zog sich an, ging hinaus, …’
‘The lazybones got dressed, went out,…’

tumlan ‘get dressed in (festival dress, uniform)’

←← tum ‘dress, clothing’ (obviously the form *tumla has disappeared)

(28) Irxi zarjadka tu-r-ĕś te šıv-a  (ChM 198, 247)

morning exercise do-pret-3pl and water-dat/acc
kĕr-se čüxen-čĕ-ś.
go.in-crd.ger wash-past-3pl
‘They did morning exercises and bathed (washed
themselves) by going in the water.’

čüxen ‘[s]plash, dabble; wash o.s.; pour over o.s., have bath; swing, sway’
← čüxe ‘rinse, wash out; disinfect (the seeds)’

In my earlier studies on Mordvin (Salo 2006b: 181–182, 2010a: 80), I have separated
a small group of reflexives that have the special meaning of moving in some direction
and named them intentionals. The second role is very frequently lative or locative loc-
ative. This kind of meaning is also quite common among the Chuvash verbs presented
in (29) to (31). In this function, even a verbal structure kur-ma ‘to see’ is possible (30).

(29) “tură man-a xa-m-a pıtan-ma xuš-at, (Gebr 181)

God I-dat/acc self-px1sg-dat/acc hide-inf order-pres.3sg
ăś-ta pıtan-as?” te-nĕ.
where-loc/lat hide-fut.ptc say-perf.ptc
‘“Gott befiehlt mir, mich zu verstecken, wo habe ich mich zu verstecken?”’
‘“God told me to hide myself, where shall I hide myself?”’
[he] said. [The boy is talking to his horse.]

pıtan ‘hide oneself, go into hiding; be hidden’

~ pıtar ‘hide [away], conceal, mask; bury, cover’
240 Salo

(30) pĕtĕm śın tuj kur-ma puxăn-nă. (Gebr 231)

all people wedding see-inf gather-perf.ptc
‘alle Leute versammelten sich, um die Hochzeit zu sehen.’
‘all the people gathered to see the wedding.’

← puxăn ‘assemble, gather (together) (itr.)’

← pux ‘gather, collect’

(31) kil-ne tavrăn-san man-a xa-măr (Gebr 121)

home-px3sg.dat return-crd.ger 2 I-dat/acc self-px1pl
kil-e jeś-r-ĕ.
home-dat/acc drive-pret-3sg
‘Nachdem wir heimgekehrt waren, führte er mich in unser Haus.’
‘After we came home, he drove me to our own house.’

tavrăn ‘come back, return, arrive’

← tavăr ‘turn around, wrap; return, give back’

A source in the ablative is present in (32). Obviously, the temporal meaning can also
be in the second role, as in (33).

(32) Вăл амăшӗ-нчен ҫăмăллăн-ах вӗҫерӗн-чӗ (PaMo 713)

(s)he mother-px3sg.ablat easily-emph.part withdraw-past.3sg
‘He drew apart easily from his mother.’

vĕśerĕn ‘withdraw [from], free’

← vĕśer ‘loosen, release’

(33) ҫулла кашкăр тапăн-ма-ст те-ҫҫӗ, … (PaMo 5)

in.summer wolf attack-neg-aor.3sg say-past.3sg
‘it was said that wolves do not attack in the summer,…’

tapăn ‘attack’
← tap ‘kick, push, hit’

Sometimes the local element can be included in the meaning of the verb:

(34) Чăххи-сем шавлăн вӗҫ-се салан-чӗҫ, (PaMo 1240)

hen-pl loudly fly-crd.ger1 scatter-past.3pl
‘Hens flew off loudly in different directions,…’

salan ‘undress; scatter; loosen’

~ salat ‘ spread (tr.); take to pieces’
 Deverbal reflexive and passive in Chuvash 241

The next verb tipĕn- ‘dry up’ is marked as a reflexive verb in the dictionary, meaning
that it is impossible for it to have an object.

(35) типӗн-ме пӳрт-е кӗр (ČRS 483)

dry-inf house-dat/acc enter
‘come inside to dry (one’s clothes)’

tipĕn ‘dry up (one’s clothes, but itr.)’

← tip ‘dry (itr.), become dry’ ~ tip, tipĕ ‘dry; dry place, dryness, hot weather season’

In my material, (36) is especially interesting due to the fact that it proves that the
suffix ‑l- in Chuvash can also be read as a reflexive. Examples (37) and (38) present
another case where the suffixes ‑l- and ‑n- can be used equally. With a neutral sub‑
ject, the reading would be automative, e.g. păr xuskančĕ ‘the ice moved’ (ČRS 567).

(36) Алăк-ран тух-р-ăм, тайăл-т-ăм – (Sbor 48–49)

door-abl go.out-pret-1sg bend-pret-1sg
‘I went out the door and bent (down) –’

tajăl ‘bend, lean (itr), sink down’

← taj ‘weigh; bend, lean (tr), press down’

(37) хресчен-сем кӗрешӗв-е хускал-нă (ČRS 567)

peasant-pl battle-dat/acc move-perf.ptc
‘the peasants started to move to the battle’

(38) ан хускан! (ČRS 567)

neg.imp.part move.imp
‘don’t move!’

xuskal, xuskan ‘move, start to move (itr.)’

~ xuskat ‘move, start to move (tr.)’

3.4. Zero meaning

Since Ašmarin, it has been recognized that some Chuvash verbs can take the reflex‑
ive suffix ‑n without this changing the original meaning of the verbs, e.g. xĕpĕrte or
xĕpĕrten ‘be happy, rejoice’ (Ašmarin 1898: 260). In this case, the derivative suffix
can be considered to be semantically empty and is then classified as a modifier; there‑
fore the term zero meaning is suitable. Such pairs are easiest to find in dictionaries,
e.g. xĕr and xĕrĕn ‘become hot’ (ČRS 553–554), or šıś and šıśăn ‘swell, puffy, dis‑
tended; get greasy, get thick[er]’ (ČRS 627, TšSS 272). The first argument can be (43)
242 Salo

experiencer, (40) quasi-experiencer, or (39), (41)11, (42), (44), neutral. Often a reason
or instrument in the instrumental case can be present, as can be seen in (42), (43), and
(44). In the following three pairs of sentences, the suffix has zero meaning:

(39) ура-на пылчăк ҫыпăҫ-нă (TšSS 186)

foot-dat/acc mud stick-perf.ptc
‘the mud stuck to [his/her] feet’

(40) хуҫăл-нă шăмă ҫыпҫăн-нă (ČRS 433)

be.broken-perf.ptc bone stick-perf.ptc
‘the broken bone(s) grew back together’

śıpăs, śıpśăn ‘stick, fasten (itr.), adhere’

(41) улмуҫҫи хăр-нă (ČRS 547)

apple.tree dry-perf.ptc
‘the apple tree dried up’

(42) уяр-па уй-хир хăрăн-ать (ČRS 548)

dry.weather-instr field-meadow dry-pres.3sg
‘the dry weather is parching the agricultural land’

xăr, xărăn ‘dry, get dry, get parched’ ~ xăr ‘dry (adj.)’

(43) сивӗ-пе чӗтре (TšSS 257–258)

cold-instr shake
‘shake with cold’

(44) йывăр машинă-сем пыни-пе кӗпер чӗтрен-ет (ČRS 593)

heavy car-pl going-instr bridge vibrate-pres.3sg
‘the bridge vibrates when heavy cars go across [it]’

čĕtre, čĕtren ‘shake, tremble, shiver, rattle, vibrate’

~ čĕtre ‘shaking, tremble, (substantive or adjective)’

3.5. One more potential meaning

The next example shows that texts are sometimes incompatible with the meaning
derived from dictionaries. Here, in this context, I would like to propose a modal
nuance for the suffix: ‘be able to find’. The argument structure does not change, so the
derivative suffix type is modifier. This case is very similar to Mordvin v-derivatives,

11.  Cf. previous ex. (18)

 Deverbal reflexive and passive in Chuvash 243

where dynamic modality ‘be able to do something, can’ is one of the many mean‑
ings of this ambiguous suffix (Salo 2006b: 182–183). In Mordvin, this meaning is
so important that the word vozmožnyj or vozmožnostʹ ‘possible/possibility’ is always
mentioned in the relatively new comprehensive Erzya and Moksha dictionaries beside
the term reflexive (vozvratnyj), including the passive and automative.

(45) ku pět-ně kěrü jeple tup-ăn-čě? (Gebr 255)

this die-perf.ptc son-in-law how find-mod-past.3sg
‘Wie hat dieser verstorbene Schwiegersohn hergefunden?’
‘How has the dead son-in-law been able to find [his way] here?’

tupăn ‘be found, be able to find’

← tup ‘find’

4. Other Turkic languages

4.1. Reflexive and passive in Tatar and Bashkir

The closely related Tatar and Bashkir languages have quite similar derivational sys‑
tems and there is no clear boundary between the passive and reflexive. Verb stems
ending with ‑l or ‑lA take the passive suffix ‑n: Tatar and Bashkir al- ‘take’ → alın- ‘be
taken’, bašla- ‘begin’ → bašlan- ‘be started’, alda- ‘deceive’ → aldan- ‘be deceived’,
Tatar sukala- ‘plough’ → sukalan- ‘be ploughed’, ešlä- ‘make’ → ešlän- be made’.
Other verb stems take the passive suffix ‑l: Tatar ju- ‘wash’ → juıl- ‘be washed’
sat- ‘sell’ → satıl- ‘be sold’, Bashkir bor- ‘turn’ → borol- ‘be turned’, Tatar yasa- ~
Bashkir yaha- ‘make’ → yasal- ~ yahal- ‘be made’ (Poppe 1964: 69, 1968: 97–98,
Berta 1998: 291).
Reflexive verbs are formed with ‑n and may thus coincide with the passive: Tatar
tot- ‘hold’ → totın- ‘hold on’, ju- ‘wash’ → juın- ‘wash oneself’, Tatar ki- ~ Bashkir
kěj- ‘dress’ → kijăn- ~ kějen- ‘dress oneself’, Tatar bize- ~ Bashkir biδe- ‘adorn’ →
bizen- ~ biδen- ‘adorn oneself’ (Poppe 1964: 69, 1968: 97–98, Berta 1998: 291).
It has been reported that (cooperative-)reciprocals with ‑š have displaced reflex‑
ives in Bashkir dialects in particular. This has been attested in some eastern and
southern dialects. In one eastern subdialect (govor), Aysk, this is even more wide‑
spread, as the reflexive forms have displaced some basic forms: ašan ‘eat’ cf. aša ‘id.’
in the literary language, respectively tegen ‘sew’ cf. tek ‘id.’ (Maksjutova 1976: 59,
58, 198). In the southern subdialect of Ik-Sakmarsk, reciprocals fill basic, reflexive
and passive functions (Miržanova 1979: 66, 211–212).
It is rare to be able to study the conditions under which passives and reflexives
can be used. A study on 1,950 verbs in Tatar revealed that more than half of the
verbs studied (1,013) could not take an ‑l or ‑n suffix; most of them were intransitives
(Zinnatullina 1965: 189). Another study says that the suffix ‑n with intransitive stems
expresses intensiveness and frequency (Tumaševa 1969: 211–212).
244 Salo

In the Turkic languages, unambiguously reflexive verbs are mostly expressed

by reflexive pronouns: e.g. Turkish kendini öldür ‘kill oneself’ is a causative of öl-
‘die’. Sometimes reflexive verbs have special lexical meanings, e.g. the Turkish sev-
‘love’ → sevin- ‘rejoice’ has a parallel counterpart in Chuvash: sav- ‘love’ → savăn-
‘rejoice’ (Johanson 1998: 55). In Turkish, the passives can be derived from intransitive
verbs (46) and the first actant of the corresponding active sentence is not necessarily
expressed, which makes them close to Finnish impersonals. Obviously, this structure
does not occur in Chuvash.

(46) Burada güzel yasa-n-ıyor. (Johanson 1998: 55)

here nice live-pass-aor.3sg
‘One lives well here / Life is fine here.’

4.2. Some remarks about modality in Turkic languages

In Chuvash there are also forms of possibility and impossibility. In the literary lan‑
guage possibility is formed with the suffix ‑aj/‑ej following the stem: kur-aj-r-ăm ‘I
could see’, sometimes with the final suffix vowel dropped. It is rarely found and the
Viryal “upper” dialect does not have this feature at all. The corresponding negative
form in the literary language is kur-aj-ma-r-ăm ‘I could not see’, which can be found
in the Viryal dialect as kur-i-ma-r-ăm ‘id.’. In some Anatri ‘lower’ subdialects, the
suffix is ‑ajr/‑ejr: śit-ejr-ĕn ‘I can go’ (Andreev 1966: 55). According to Clark (1998:
443), the suffix order is: voice + ability + negation. Ramstedt (1952: 190–191, 1957:
170) sees the origin of this modality in the incorporated verb *u‑ ‘können; can’.
In Tatar and Bashkir, possibility and impossibility can be expressed by various
converb constructions (Berta 1998: 291). Dmitriev has proposed that Bashkir and
other Turkic languages have four parallel conjugation or aspect paradigms: positive,
negative, possible and impossible. The possible aspect is formed using the gerund
‑a and the auxiliary verb al- ‘take’, e.g. jaȥa aldı ‘(s)he could write’ (Dmitriev 1948:
135–137).

4.3. Different suffix combinations

Recently, a substantial amount of time has been devoted to investigating combina‑

tions of verbal suffixes with each other in Turkic linguistics. The passive morpheme
has two phonetically conditioned allomorphs: ‑(I)n- after vowel-final stems and lat‑
erals, and ‑Il-12 elsewhere. The n-variant is identical with the medium voice, and

12.  The capital “I” indicates that the vowel is subject to a four-way alteration, rendered orthographi‑
cally by <u, ü, ı, i>, and determined by the rules of vowel harmony (Haig 2000: 219).
 Deverbal reflexive and passive in Chuvash 245

in some cases, it is difficult to distinguish the medium from a passive (Haig 2000:
219). In Yakut, there can even be suffixes of three voices incorporated into a single
verb (Ščerbak 1981: 101). It has become apparent from large text corpora that in con‑
temporary Turkish materials monoclausal double passives do occur:

(47) Bu şato-da boğ-ul-un-ur. (Özkaragöz 1986: 77)

This chateau-loc strangle-pass-pass-aor.3sg
‘One is strangled (by one) in this chateau.’

A human agent in Turkish passives can be expressed in the clause via the quasi-
postposition tarafından ‘from the side of’, and a non-human agent with the ablative
case, e.g. rüzgar-dan ‘by the wind’ (Haig 2000: 225). Even one word sentences are
possible: Öl-ün-ür or Öl-ün-ül-ür ‘It is [has] died.’ There are also complex causative-
passive, causative-causative, and double causative-passive forms, such as öl-dür-ül-
‘be killed’, öl-dür-t- ‘cause to kill’, öl-dür-t-ül- ‘be caused to kill’, all derived from the
stem öl- ‘die’ (Johanson 1998: 56). Sometimes a sentence is ambiguous as to whether
or not it is reflexive or passive:

(48) Mehmet yıka-n-dı. (Özkaragöz 1986: 78)

Mehmet wash-refl/pass-past.3sg
‘Mehmet washed himself. / Mehmet was washed.’

Once a second passive morpheme has been added, the sentence can only be read as
passive: Mehmet yıka-n-ıl-dı. This type of double passive is called a passive intensi‑
fier by Özkaragöz, who separates also a second type of double passive connected to
modality, one with the abilitative modal auxiliary suffix ‑Ebil:

(49) Burada çalış-ıl-abil-in-ir. (Özkaragöz 1986: 79)

here work-pass-mod-pass-aor.3sg
‘Here it can be worked.’

A diligent search has exposed cases where the suffixes are in a different order:

(50) Bu durum-a tabii ki üz-ül-ün-ebil-ir. (Heß 2011: 263)

such situation-dat of.course conj hurt-pass-pass-mod -aor.3sg
‘One can be hurt by such words.’

Andreev has examples of Chuvash causative markers that can be added twice: śi-ter-
ter ‘make to feed’ or can be added to other voices, as kala-ś-tar ‘make to speak/dis‑
cuss’ or vit-ěn-ter ‘make to cover oneself’ (Andreev 1966: 54–55). The last one can
also have another meaning:
246 Salo

(51) vărlăx-a tăpra-na lajăx vitěnter-ně  (ČRS 82)

seeds-dat/acc earth-dat/acc well earth.over-perf.ptc
‘the seeds are well embedded in the soil’

Even cases with three causative suffixes have been reported: xăp-ar-t13-tar14 ‘make
to get up’ ←←← xăp ‘come loose’. These suffixes can be regarded as allomorphs
for the sake of convenience (V. Sergeev 2002: 234). A reflexive suffix can be added
after a frequentative suffix ‑kala/‑kele: śap-kala-n ‘push, poke several times; rush;
gesticulate; ramble’ ←← śap ‘hit’ or vět-kele-n ‘hurry; try’ ←← vět ‘scorch, burn’
(V. Sergeev 2002: 235–236).
Sometimes verbs having a reciprocal ‑š or other suffixes can be synonymous
with ‑l or ‑n verbs as in (52) and (53), e.g. kutănlan and kutănlaš and ‘be stubborn,
recalcitrate, be capricious’ (TšSS 85), ača kutănlašat ́ ‘the child is capricious’, laša
kutănlašat ́ ‘the horse is agitated’ (ČRS: 198). It is also possible to say tumtir tipšerně
or tipšenčě ‘the cloths got dry’, even though the stem tip has the meaning ‘get dry’
(ČRS: 483)

(52) санталăк кашт майлан-чӗ (ČRS 223)

weather a.little get.better-past.3sg
‘the weather cleared up a little’

(53) санталăк майлаш-са кай-рӗ (ČRS 223)

weather get.better-crd.ger1 go-past.3sg
‘the weather cleared up’

majlan, majlaš ‘turn out well, to be fixed, calm down’

← majla ‘fix, repair’

The sentence in (54) shows a combined suffix with a more specified meaning than its
semantically related verbs from the same adjectival stem.

13.  Almost no attention has been paid to causative suffix ‑at/‑et/‑t due to its rarity (V. Sergeev 2002:
234).
14.  It sometimes has the allomorph ‑ttar/‑tter, but not in this case.
 Deverbal reflexive and passive in Chuvash 247

(54) Sivĕ, xajar xĕl irt-er // (ChM 212, 261)

cold bad winter pass-pret.3sg
kaj-at jĕr-se, xurlan-sa;
go-pres.3sg weep-crd.ger1 grieve-crd.ger1
‘[The] cold unpleasant winter passes //
goes a-weeping, grieving15;’

xurlan ‘mourn, grieve’

← xurla ‘defame, abuse, tell tales’
cf. xural ‘turn black, blacken; darken; get dirty’
xuran ‘be seen as black, dark’
← xura ‘black; dark, gloomy: dirty, dirt; bitter, acrid’

5. Conclusions

Whereas the Chuvash passive and reflexive suffixes are systematically used for form‑
ing lexemes, in many other languages passivization and reflexivization are purely
syntactic processes. The Chuvash state of affairs is not relevant to the grammar – lexi‑
con dichotomy, because the causative and passive morphemes are both used within
the grammar and within the lexicon. They can make nouns into verbs, especially the
reflexive suffix, which has a wider range of use, but verbs with the passive suffix can
also have a reflexive reading. The picture is not clear, however, because some stems
accept only ‑n suffixes and do not take ‑l suffixes at all. In addition, reflexive suffixes
are the same as the suffixes deriving nouns into verbs. Rezjukov (1959: 136) tries
to avoid the problem by combining both suffixes under the passive-reflexive voice.
Dictionaries have provided simple solutions throughout the decades: Egorov (1954:
317–318) lists derivational suffixes, the ‑l and ‑n suffixes are divided into two catego‑
ries, deverbal and denominal, and the ‑n is also considered to have a passive meaning
(ČRS 662). The newest grammar does not say anything about the passive, but states
that the suffixes of the reflexive and reciprocal voices are derivational, while suffixes
of the causative voice are multifunctional and can be added very widely to verbal
stems (V. Sergeev 2002: 237).
In Turkish, the difference between voices is sometimes clear: ög-il ‘be praised’
cf. ög-in ‘boast’ (Gusev 1986: 6–7). Serebrennikov suggests, that in the Turkic lan‑
guages, there are two passive suffixes ‑l- and ‑n-, which have developed indepen‑
dently, although in some languages these have conflated. He thinks that the passive
meaning has evolved from the reflexive meaning (Serebrennikov 1976: 33).

15.  The original translation by Krueger is ‘growing black’, but this might be incorrect, as the diction‑
aries do not include the meaning ‘grow black, blacken’, the usual word for ‘grow black’ is xural (ČRS
565).
248 Salo

It seems to me that in the turkological work published in the Soviet Union or

Russia the term reflexive is often used loosely and sometimes might even be close
to the term intransitive. Usually, the term passive is defined in relation to the active.
It has been most disadvantageous to Soviet turkology that there tends to only be
a superficial knowledge of the Finno-Ugrian languages, which are also agglutina‑
tive in nature. Anyone who has studied Finnish grammar in depth can clearly see
that the passive voice formed with the ‑t‑ affix in Finnish is obviously something
other than a derivational suffix. Both the reflexive and passive categories are clearly
derivational in the Turkic languages, as many of the examples have shown. At least
from the morphological point of view there is no reason to maintain a separate voice
category. Ultimately, the language decides what category to place its verbs in: in
Russian vernutʹsja is a reflexive verb, but the English return (itr.) and the Finnish
palata are not. In Russia, not much in-depth research on the derivation of verbs has
been carried out, due to the fact that lists of verbs cannot reveal much at all; context
is what greatly influences results. Some researches consider this situation unpleas‑
ant and propose more studies on verbal categories in their native languages, such as
Zinnatullina (1965: 185), who considered the five voices of Tatar dating back to 1895
to be insufficient.
So far, the Chuvash meanings presented are very similar to the meanings of ‑Alt
verbs in Mari. Verbs in Mari using the am-forms have a passive, automative, reflexive
or zero-meaning, too. The Mari Alt-verbs seem to be indisputedly deverbal (as are
the Mordvin v-verbs), so their derivational suffixes are transformers and modifiers,
while in Chuvash, derivational suffixes from all three categories are used. There are,
however, other meanings16 for these Alt-suffix verbs, especially if they take em-forms
(cf. Salo 2006a: 337). Although the Pavlik Morozov corpus has not turned up any
new meanings, the result might be different if more extensive corpora of modern lan‑
guages were to be examined. The Mordvin v-verbs have more meanings and occur in
texts more frequently than the corresponding Chuvash and Mari verbs do.
Except for in the case of zero meaning, ‑l and ‑n suffixes are used to reduce the
verbs’ valences. This is particularly valid with deverbal derivatives. In very old cases
of nomen-verbum, the few stem verbs are intransitive and seem to be more common
than their derivatives, which have more restricted uses. As is often the case with nom‑
inal stems, the derived intransitives and transitives seem to be equal. On the following
pages in Appendix, the term intransitive refers to the reflexive, automative and pas‑
sive, with context often being the determining factor. For the sake of completeness,
the causatives are also represented, as they are needed to describe the alternation
between the transitive and intransitive meanings.

16.  The frequentative meaning of am-verbs could also be referred to the l-frequentative on the Turkic
side.
 Deverbal reflexive and passive in Chuvash 249

Appendix

Chuvash derivational patterns for verbs are mainly based on materials gathered from
ČRS and TšSS. The case inventory of Chuvash makes finding an object rather dif‑
ficult due to the fact that the object is marked with the dative/accusative case, which is
also used to mark local utterances. So it is more than likely that not every intransitive
or transitive meaning of the treated verbs has been included.

Deverbal patterns

Transitive stems
tr. itr. tr.
taj bend → tajăl bend → tajăltar make to bend
→ tajăn bend → tajăntar make to bend
kas cut → kasăl be cut, cut oneself → kasăltar make to be cut,
→ kasăn be cut, cut oneself cut oneself
vit cover → vitĕn be covered → vitĕnter make to cover,
make to cover oneself
iš break → išel be broken → išelter make to break
→ išter make to break
śıx tie, bind → śıxăn be bound up → śıxăntar connect
→ śıxlan get involved → śıxlantar join
śır write → śırăn be written → śırăntar make to be written
kur see → kurăn be seen
ilt hear → iltĕn be heard → iltĕnter make to be heard
→ iltter make to hear
salt open, take off → saltăn undress → saltăntar make to undress
→ salttar make to open, take off
tavăr return → tav[ă]răn return
→ tavărttar make to return
muxta praise → muxtan boast → muxtantar make to boast
→ muxtattar make to praise
śĕkle raise, lift up → śĕklen rise, get up → śĕklenter make to rise, get up
→ śĕkletter make to raise, lift up
250 Salo

Intransitive stem
xĕr warm up → xĕrel redden
→ xĕrĕn warm up
→ xĕret paint red (tr.) → xĕretter make to paint red
→ xĕrt warm (tr.) → xĕrtĕn warm to up (itr.) →
xĕrtĕnter make to warm up

No clear precedence
tapran move (itr.) ↔ taprat move (tr.)
↓ ↓
taprantar make to move (itr.) taprattar make to move (tr.)
pıtan hide (itr.) ↔ pıtar hide (tr.) → pıtaran hide (itr.)
↓ ↓
pıtantar make to hide pıtarttar make to hide

Nomen-verbum patterns
itr. + noun itr. tr.
tip dry → tipĕn get dry → tipĕnter make to dry
→ tipĕt dry (tr.) → tipĕtter make to dry
→ → tipšer get dry
→ → tipšen get dry
xăr dry → xărăn dry up → xărăntar make to dry up

Denominal patterns
noun itr. tr.
tum dress → *tumla? → tumlan dress → tumlantar dress
śilĕ anger → śillen get angry → śillenter make angry
üsĕr drunken → üsĕrĕl get drunken → üsĕrĕlter make to get drunken
jĕpe wet → jĕpen get wet
→ jĕpet wet (tr.) → jĕpetter make wet
noun tr. itr.
paj part → pajla divide into parts → pajlan be divided into parts
pajlattar make divide into parts
vak little → vakla make smaller → vaklan get smaller
→ vaklattar make to make smaller (tr.)
 Deverbal reflexive and passive in Chuvash 251

noun tr. tr.

tăxa? buckle → tăxan dress → tăxantar dress somebody →
tăxantart make dress comebody
noun several derivatives
pĕkĕ harness bow → pĕkĕn bend (itr.)
→ → pĕkĕrĕl bend (itr.) → pĕkĕrĕlter bend (tr.)
→ → pĕkĕren bend (itr.)
→ → pĕkĕrt bend (tr.)
tasa clean → tasal become clean (itr.)
→ tasalan become clean[er] (itr.)
→ tasat clean (tr.) → tasattar make clean (tr.)
puś head → puśăn begin (itr.) → puśăntar make to begin (itr)
→ puśar start
→ puśla begin (tr.) → puślattar make to begin (tr.)
↓
puślan begin (tr.)

Glosses and abbreviations

abl ablative mod modal(ity)
adj. adjective neg negation
aor aorist neg.imp.part negative impera‑
carit caritive tive particle
conj conjunction neg.ptc negative participle
crd.ger1 first coordinative p. person
gerund ‑sa, ‑se part particle
crd.ger 2 second coordinative past past
gerund ‑san, ‑sen, perf.ptc perfect participle
‑sassăn, ‑sessĕn pl plural
dat/acc dative/accusative pop postposition
emph.part emphatic particle pres present
Finn. Finnish pres.ger present gerund
gen genitive pret preterite
id. idem purp purposive
imp imperative px possessive suffix
inf infinitive ref reflexive
instr instrumental Russ. Russian
itr. intransitive sg singular
lat lative tr. transitive
loc locative
252 Salo

References

Data sources
ChM = J. R. Krueger 1961: Chuvash Manual. Introduction, Grammar, Reader, and Vocabu-
lary. Indiana University Publications, Uralic and Altaic Series, 7. The Hague, Nether‑
lands: Mouton & Co.
ČRS = М. И. Скворцов (ed.) 1985: Чувашско–русский словарь./Чăвашла вырăсла
словарь. Москва: «Русский язык».
Gebr = Heikki Paasonen 1949: Gebräuche und Volksdichtung der Tschuwassen. Gesammelt
von Heikki Paasonen, heraugegeben von Eino Karahka und Martti Räsänen. Mémoires
de la Sociéte Finno-Ougrienne 94. Helsinki: Suomalais-Ugrilainen Seura.
PaMo = Pavlik Morozov in Chuvash. One of parallel corpuses made by and kept in the
Research Unit for Volgaic Languages, University of Turku. The Chuvash translation
made by Eduard Fomin in 2007 is based on the Russian version published on pages
63–111 in the Mansi translation: В. Губарев 1953: Павлик Морозов. На мансийский
перевёл Н. М. Садомин. Москва & Ленинград: Гос. уч.-пед. изд. Мин. просв.
Sbor = Н. И. Ашмаринъ 1900: Сборникъ чувашскихъ пѣсенъ, записанныхъ въ губерніяхъ
Казанской, Симбиркой и Уфимской. Казань: Типо-литография Императорскаго
Университета.
TšSS = Arto Moisio & Eduard Fomin & Jorma Luutonen 2007: Tšuvassilais-suomalainen
sanakirja. Turun yliopiston suomalaisen ja yleisen kielitieteen julkaisuja 76. Turku:
Turun yliopiston suomalaisen ja yleisen kielitieteen laitos.

Other sources
Alekseev = А. А. Алексеев 1970: Заметки о первых грамматических трудах по чуваш­
скому языку. – Учёные Записки НИИ при Совете Мин. Чув. АССР 49: 203–220.
Alekseev & Sergeev = А. А. Алексеев & Л. П. Сергеев 1988: Памятник чувашского
языка 1836 года. – Н. И. Егоров, И. П. Павлов & М. Ф. Чернов (eds), Исследования
по этимологии и грамматике чувашского языка. Чебоксары: НИИЯЛИЭ при
Совете Мин. Чув. АССР. 78–103.
Alhoniemi, Alho 1985: Marin kielioppi. – Hilfsmittel für das Studium der finnisch-ugrischen
Sprachen X. Helsinki: Suomalais-Ugrilainen Seura.
Alhoniemi, Alho 1993: Grammatik des Tscheremissischen (Mari). Hamburg: Helmut Buske
Verlag.
Anderson, J. M. 1971: The Grammar of case. Towards a localistic theory. Cambridge Studies
in linguistics 4. Cambridge: Cambridge University Press.
Andreev = И. А. Андреев 1966: Чувашский язык. – В. В. Виноградов (ed-in-chief), Языки
народов СССР 2. Москва: Наука. 43–65.
Ašmarin = Н. И. Ашмаринъ 1898: Матеріалы для исслѣдованія чувашскаго языка.
Казань: Типо-литография Императорскаго Университета.
Benzing, Johannes 1943: Kleine Einführung in die Tschuwaschische Sprache. Berlin: Otto
Stollberg.
Benzing, Johannes 1959: Das Tschuwaschische. – Jean Deny & al. (eds), Philologiae Turci-
cae Fundamenta I. Wiesbaden: Franz Steiner. 695–751.
 Deverbal reflexive and passive in Chuvash 253

Benzing, Johannes 1964: Die tschuwaschische Literatur. – Louis Bazin & al. (eds), Philolo-
giae Turcicae Fundamenta II. Wiesbaden: Franz Steiner. 841–861.
Berta, Árpád 1998: Tatar and Bashkir. – Lars Johanson & Éva Á. Csató (eds), The Turkic
Languages. London & New York: Routledge. 283–300.
Budenz, József 1863: Csuvas közlések és tanulmányok III. Nyelvtudományi közlemények II.
Pesten: A Magyar Tudományos Akadémia. 14–68.
Clark, Larry 1998: Chuvash. – Lars Johanson & Éva Á. Csató (eds), The Turkic Languages.
London and New York: Routledge. 434–452.
Dimitriev = В. Д. Димитриев 1967: Документы об издании книги «Сочинения, принад­
лежащие к грамматике чувашского языка». – Учёные Записки НИИ при Совете
Мин. Чув. АССР XXXIV: 153–162.
Dmitriev = Н. К. Дмитриев 1948: Грамматика башкирского языка. Москва & Ленин­
град: Изд. АНСССР.
Egorov = В. Г. Егоров 1949: Чувашские словари XVII века. – Записки НИИЯЛИ при
Совете Мин. Чув. АССР II: 111–142.
Egorov = В. Г. Егоров 1951: Первая печатная грамматика чувашского языка 1769 г.
(К 180-летию со дня вычода в свет.) – Тюркологический сборник I. Москва &
Ленинград: Изд. АН СССР. 85–92.
Egorov = В. Г. Егоров 1954: Краткий очерк грамматики чувашского языка. – В. Г.
Егоров, Чӑвашла–вырӑсла словарь. 2-мӗш изд. Шупашкар: Чӑваш АССР
Государство издателъстви. 299–320.
Egorov = В. Г. Егоров 1957: Глагол. – Материалы по грамматике современного
чувашского языка. Часть первая. Морфология. Чебоксары: Чув. кн. изд. 151–239.
Fedotov = М. Р. Федотов 1986: Чувашский язык в семье алтайских языков III. Под
редакцией Н. А. Баскакова. Чебоксары: Чув. кн. изд.
Fedotov = М. Р. Федотов 1987: Исследователи чувашского языка. Чебоксары: Чув. кн.
изд.
Gabelentz, H. C., von der. 1861: “Über das Passivum. Eine sprachvergleichende Abhand­
lung”. – Abhandlungen der Philologisch-historischen Classe der Königlich sächsi­
schen Gesellschaft der Wissenschaften 3 (= Abhandlungen der Königlich sächsischen
Gesellschaft der Wissenschaften 8): 449–546.
Geniušienė, Emma 1987: The typology of reflexives. Empirical approaches to language typol‑
ogy 2. Berlin: Mouton der Gruyter.
Grunina = Грунина, Э. А. 1987: К теории тюркского залога. – Советская тюркология
1987 (2): 3–15.
Gusev = В. Г. Гусев 1986 Значения форм страдателъного и возвратного залогов.
Состав категории залога. (На материале староанатолийско-тюркского языка.) –
Советская тюркология 1986 (4): 5–12.
Haig, Geoffrey 2000: Towards a unified account of passive in Turkish. – Turkic Languages 4.
Wiesbaden: Harrassowitz Verlag. 215–234.
Hakulinen, Auli & Fred Karlsson 1979: Nykysuomen lauseoppia. Suomalaisen Kirjallisuu­
den Seuran Toimituksia 350. Helsinki: Suomalaisen Kirjallisuuden Seura.
Heß, Michael Reinhard 2011: Towards explaining double passive marking in Turkish pos‑
sibility verbs. – Acta Orientalia Academiae Scientiarum Hungarica 64 (3): 257–286.
Hesselbäck, André 2005: Tatar and Chuvash Code-copies in Mari. Acta Universitatis Upsa‑
liensis. Studia Uralica Upsaliensia 35. Uppsala: AUU.
254 Salo

Johanson, Lars 1998: The Structure of Turkic. – Lars Johanson & Éva Á. Csató (eds), The
Turkic Languages. London & New York: Routledge. 30–66.
Kangasmaa-Minn, Eeva 1982: Derivaatiokielioppia I: verbijohdokset. – Sananjalka 24:
43–63.
Keenan, Edward I. 1985: Passive in the world’s languages. – Timothy Shopen (ed.), Lan-
guage typology and syntactic description 1. Clause structure. Cambridge: Cambridge
University Press. 243–281.
Kornilov = Г. Е. Корнилов 1975: К вопросу о залогах в чувашском языке. Пассив и
рефлексив. – Труды НИИ при Совете Министров Чувашской АССР 53: 44–61.
Kulonen, Ulla-Maija 1989: The passive in Ob-Ugrian. Mémoires de la Société Finno-Ougri‑
enne 203. Helsinki: Suomalais-Ugrilainen Seura.
Lehmann, Christian 1995: Thoughts on Grammaticalization. LINCOM Studies in Theoreti‑
cal Linguistics 01. München–Newcastle: LINCOM EUROPA.
Levitskaja = Л. С. Левитская 1976: Историческая морфология чувашского языка.
Москва: Наука.
Levitskaja = Л. С. Левитская 1988: Категория залога. – Э. Р. Тенишев (ed-in-chief),
Сравнительно-историческая грамматика тюркских языков. Морфология.
Москва: Наука. 269–324.
Maksjutova = Н. Х. Максютова 1976: Восточный диалект башкирского языка в
сравнительно-историческом освещении. Москва: Наука.
Miržanova = С. Ф. Миржанова 1979: Южный диалект башкирского языка. Москва:
Наука.
Orlova = А. П. Орлова 1976: Глаголы с аффиксом в чувашском языке. – Труды НИИ при
Совете Министров Чувашской АССР 66: 149–158.
Özkaragöz, İnci 1986: Monoclausal double passives in Turkish. – Dan Isaac Slobin & Karl
Zimmer (eds), Studies in Turkish Linguistics. Amsterdam & Philadelphia: John Benja‑
mins. 77–91.
Poppe, Nicholas 1964: Bashkir manual. Indiana University Publications, Uralic and Altaic
Series 36. Bloomington: Indiana University.
Poppe, Nicholas 1968: Tatar manual. Indiana University Publications, Uralic and Altaic
Series 25. 2nd, revised ed-n. Bloomington: Indiana University.
Ramstedt, G. J. 1952: Einführung in die altaische Sprachwissenschaft II. Formenlehre.
Bearb. und hrsg. von Pentti Aalto. Mémoires de la Société Finno-Ougrienne 104:2.
Helsinki: Suomalais-Ugrilainen Seura.
Ramstedt = Г. И. Рамстедт 1957: Введение в алтайское языкознание. Морфология.
Москва: Изд. иностранной литературы.
Rezjukov = Н. А. Резюков 1959: Сопоставительная грамматика русского и чувашского
языков. Чебоксары: Гос. изд. Чув. АССР.
Róna-Tas, András 1998: The Reconstruction of Proto-Turkic and the Genetic Question. –
Lars Johanson & Éva Á. Csató (eds), The Turkic Languages. London & New York:
Routledge. 67–80.
Salo, Merja 1988: Ersämordvan v-johtimiset verbit. (Manuscript, kept in the Kaisa-library in
Helsinki.)
Salo, Merja 2006a: The Derivational Passive and Reflexive in Mari Grammars. – Juhani
Nuorluoto (ed.), The Slavicization of the Russian North. Slavica Helsingiensia 27:
328–340.
 Deverbal reflexive and passive in Chuvash 255

Salo, Merja 2006b: The passive in Erzya-Mordvin folklore. – Werner Abraham & Larisa
Leisiö (eds), Passivization and typology: Form and function. Typological Studies in
Language 68. Amsterdam/Philadelphia: John Benjamins. 165–190.
Salo, Merja 2010a: Mordvin t derivatives – semantic equivalent for impersonal. – Ural-Altaic
Studies / Uralo-altajskie issledovanija 2010, 2 (3): 72–87.
Salo, Merja 2010b: Heikki Paasonen – mittavan aineiston kerääjä ja keräyttäjä sekä jälki­
polvien työllistäjä. – Paula Kokkonen & Anna Kurvinen (eds), Kenttäretkistä tutkimus-
tiedoksi. Uralica Helsingiensia 4. Helsinki: Suomalais-Ugrilainen Seura. 57–106.
Schott, Guilelmus 1841 (or 1840 or 1842?): De lingua tschuwaschorum. Dissertatio. Berolini:
Veitii et socii sumptibus.
Serebrennikov = Б. А. Серебренников 1976: К проблеме происхождения аффиксов
страдательного залога ‑l-, ‑yl-, ‑il-, ‑n-, ‑yn-, ‑in- в тюркских языках. – Советская
Тюркология 1976 (2): 29–34.
Serebrennikov & Gadžieva = Б. А. Серебренников & Н. З. Гаджиева 1986: Сравнительно-
историческая грамматика тюркских языков. Изд. 2-е, испр. и доп. Москва:
Наука.
Sergeev = В. И. Сергеев 2000: В. Г. Егоров профессор залог категорийě пирки каланă
шухăшсем. – Чăваш чěлхи: Аваллăхран малашлăха. (Василий Георгиевич Егоров
çуралнăранпа 120 çул çитнě май.) Конференци материалěсем. Шупашкар. 35–36.
Sergeev = В. И. Сергеев 2002: Морфологические категориальные и некатегориальные
формы в современном чувашском языке. Чебоксары: Изд-во Чуваш. Ун-та.
Sergeev = Л. П. Сергеев 2004: Залог. – Чăваш чěлхи. Энциклопеди словарě. Шупашкар:
Чăваш кěнеке издательстви. 44.
Šamraj = Д. Д. Шамрай 1955: К истории появлeния первой грамматики чувашского
языка. – Учёные Записки НИИЯЛИ при Совете Мин. Чув. АССР XII: 226–230.
Ščerbak = А. М. Щербак 1981: Очерки по сравнительной морфологии тюркских языков.
(Глагол). Ленинград: Наука.
Šluinskij = А. Б. Шлуинский 2006: Акциональмость и аспектуальный показатель:
конструкция со вспомогательным глаголом il- в чувашском языке. – Вестник
Московского Университета. Сер. 9. Филология 2006 (1): 32–47.
Tumaševa = Д. Г. Тумашева 1969: Глагол. – Х. Р. Курбатов & al. (eds), Современный
татарский литературный язык. Лексикология, фонетика, морфология. Москва:
Наука. 210–295.
Zinnatullina = К. З. Зиннатуллина 1965: О возвратном и страдательном залогах в
татарском языке. – М. З. Закиев (ed-in-chief), Вопросы татарского языкознания
II. Казан: Изд. Каз. Ун-та. 185–211.