Везде
Is contact-induced syncretism possible? A corpus-based study on bilingual verbal morphology of spoken German in Russian Siberia
Table of contents
Annotation Estimate Publication content
References Comments
Share
QR
Metrics
Is contact-induced syncretism possible? A corpus-based study on bilingual verbal morphology of spoken German in Russian Siberia
Annotation
PII
S0373658X0007548-1-1
DOI
10.31857/S0373658X0007548-1
Publication type
Article
Status
Published
Authors
Christiane Andersen 
Affiliation: University of Gothenburg
Address: Sweden, Gothenburg
Edition
Issue 6
Pages
94-112
Abstract

This paper investigates bilingual verbs in a corpus of spoken German by L1 speakers in Russian Siberia. The study analyzes bilingual verb formations of the type hinpostupaje, which are inserted into a position corresponding to a native verb. Different kinds of inflectional and word-formation patterns in bilingual verbs are identified and discussed with reference to the contact-induced morphology of German and Russian. The study demonstrates that the bilingual verbs follow morphophonemic regularities, and that there are several traces of morphophonemic syncretism. It argues that the bilingual speakers make use of German and Russian morphology in an innovative way, which cannot be explained by classical morphology alone. The bilingual formatives identified differ from the form/function equivalents in both German and Russian morphology. Further research should therefore include cognitive aspects of bilingual morphology.

Keywords
bilingualism, corpus linguistics, dialectology, diaspora, German, island dialects, language contact, mixed languages, morphology, Natural Morphology, Russia, Russian, syncretism, verbal inflection
Acknowledgment
I wish to thank Nadezhda Zorikhina Nilsson (professor of Slavic Languages at the University of Stockholm) for her valuable feedback concerning Russian morphology and two anonymous reviewers for valuable advices.
Received
18.12.2019
Date of publication
20.12.2019
Number of purchasers
69
Views
583
Readers community rating
0.0 (0 votes)
Cite   Download pdf
1

1. Goal of the investigation and research questions
2 Morphosyntactic structures from Russian have consistently affected the German varieties spoken in Russia over a period of more than 200 years.1 Syntactic phenomena such as verb-first order in declarative sentences and pro-drop of subject pronouns, which often coincide with morphological borrowings, have recently been the subject of more detailed investigations [Andersen 2016a; 2016b]. But contrastive studies focusing on structural and/or typological phenomena are still rare, which is all the more concerning considering that the time for fruitful applied research is running out: “[…] the onset of the 21st century bears witness to the last potential (semi-)functional speakers of many of these dialect communities, it would seem that the time to conduct meaningful research on these dialects is now” [Putnam (ed.) 2011: 1].2
1. But it should be mentioned that code-mixing with Russian was stigmatized for a long time, see Dinges [1923: 60]: “[…] so geht und fragt die Russen und hört wie sie urteilen. Sie lachen sich aus über die Deutschen, die auf Schritt und Tritt russische Wörter in ihre Sprache hineinflicken und sie noch dabei schlecht aussprechen […].”

2. An exception is the volume “Studies on German-language islands” [Putnam (ed.) 2011] introducing generative and structural studies on German-language islands in Northern America (Wisconsin, Texas, Pennsylvania), Italy (Cimbrian German) and the Netherlands (Plautdietsch). Unfortunately, studies on German-language islands in Russia are missing.
3 Based on the previous investigations, the goal of this study is to investigate bilingual verb units that have hitherto not been included in typologies of German and Russian, although this phenomenon of bilingual speech is well-known in contact linguistics. Muysken investigated bilingual verbs in typologically different languages and noted that code-mixing in verbal systems is often innovative, “leading to structures not present in either of the languages in contact” [Muysken 2000: 184]. This is particularly noticeable in Colonial German from Russian Siberia. Consider the following example3:
3. This example is taken from the transcription of a dialogue between a grandmother and grandchild living in the Russian Altai region [Moskaljuk 2013].
4 (1)
5
Jets henze uns kants naies eivodide4
ADV AUX.3PLඅ OBJ ADV OBJ INF
jetzt haben-sie uns ganz Neues eingeführt
‘Now they have introduced something very new.’
4. This and the following examples from the Siberian German Corpus (see more details in section 2) are in italics, the Russian insertions are in bold. The linguistic description is given as follows: morphosyntactic features for relevant word forms, Standard German lexemes, and English translation of the corpus example.
6 In (1), the verb perfect unit (predicate) henze eivodide has in the position of the German perfect participle a bilingual verb containing the verb stem of the Russian infinitive vodit’ ‘lead’, the German verb formation prefix ei-, and the German infinitive suffix -e. The bilingual verb eivodide is morphologically a German infinitive with a Russian stem. A wide range of several contact-linguistic and cognitive phenomena can activate the underlying triggers of such verb formations. Muysken subsumes these complex phenomena under the term “congruent lexicalization”, when “the grammatical structure is shared by languages A and B, and words from both languages a and b are inserted more or less randomly” [Muysken 2000: 8]. But what does “more or less randomly” mean for morphological units like eivodide and what makes it different from Russian and German morphology? In cognitive approaches to morphology, particularly in connectionist models, morphemes are not represented as discrete entities. “Instead, as the network learns to map from one domain to another (e.g. sound to meaning) it picks up on regularities in the mappings. Morphology arises as a consequence of the correlations between codes” [Seidenberg, Gonnerman 2000: 356]. It is useful to bear the connectionist view in mind when investigating odd contact-induced morphological verb formations like eivodide.
7 While we concentrate on contact-induced phenomena, we will argue in particular that there are specific restrictions for certain language pairs in different bilingual speech situations, i.e. “not anything structural” is possible in contact speech [Myers-Scotton 2002: ix], and that language contact phenomena are constrained by the requirements of the given grammar of the contact-induced speech [Andersen 2016b: 8]. A further approach adopted in the following investigation is the notion of contact-induced syncretism, arising probably in the morphological phenomena in the investigated corpus. We understand simple syncretism as cases in which “two or more cells with different values for a feature are merged” [Baerman et al. 2005: 13].
8 Based on these considerations, the investigation will concentrate on the following questions:
9
  1. What kind of inflectional and/or word-formation morphemes are inserted into the bilingual verb units under investigation?
  2. Can particular word-formation patterns be identified in the bilingual verb units under investigation?
  3. To what extent does the language pair and/or the direction of language transfer matter in this specific case of contact-induced morphology?
10 The background and special conditions of the corpus and several types of investigated data are presented in section 2. Relevant typological contrasts between German and Russian taking into account the grammatical structure of the German varieties in Russia are discussed in section 3. There follows an analysis of two types of verb formation contained in the material in section 4: (1) inserted Russian verb forms and (2) Russian verb stems with German affixes. The analysis seeks to answer research question (i). The particular word-formation patterns are then discussed in section 5 with reference to research question (ii): Finally, in section 6 an attempt is made to answer research question (iii).
11

2. Research corpus and types of data
12 The corpus data for this study are taken from the digitalized Siberian German Corpus (SGC)5 at the University of Gothenburg.6 It has been specifically created as a sample of the German language which has been spoken in Russia for two hundred years by ethnic Germans who immigrated to Russia as colonists. The speakers7 are between 70 and 80 years old, and are probably the last generation speaking German as their heritage language [Andersen 2016c].
5. 5 The corpus data consist of audio recordings from the Krasnoyarsk region between 1988 and 1998, collected by Valentina Djatlova (V. P. Astafyev Russian State Pedagogical University in the city of Krasnoyarsk) and video recordings from the Krasnoyarsk region of 2010 collected by the author with the help of Russian colleagues at the Astafyev University. The transcription and annotation of the two linguistic corpora was a part of the research project “Syntax in Contact. Word Order in a Variety of German Spoken in East Siberia” at the University of Gothenburg in collaboration with the Astafyev University between 2008 and 2016. The transcription follows the broad outlines of the transcription system of Spoken German GAT [Selting et al. 1998].

6. In collaboration with the Centre of Language Technology, University of Gothenburg, two corpora (Siberian German and Siberian German women) are available at Korp, the concordance search tool of Språkbanken (The Swedish Language Bank); see the language resources at Språkbanken, Korp [Borin et al. 2012]; SGC is in open access at >>>>

7. Special thanks to my wonderful informants Maria, Emma, Linda and Minna, who told me about their lives as ethnic Germans in the former Soviet Union and in Russia today.
13 The SGC consists of transcriptions of spoken narratives and dialogues of the German variety spoken by about 36,000 people in the Siberian region of Krasnoyarsk (Russia). The SGC contains a total of 50,413 tokens, see the interface of the SGC in the figure below:
14

Figure. The interface of the SGC showing examples of the token war ‘was’
15 The SGC is partly annotated: Russian words and German-Russian bilingual word units are given in square brackets; verb units have the attribute FINIT (finite verb) or INFINIT (perfect participle or infinitive). A quantitative breakdown of the data in the SGC is as follows:
16

Table 1. Total amount of annotated tokens in the SGC
17 About 3 % of the total number of tokens is borrowed from Russian. These borrowings include sequential code switches and different types of lexical borrowings. However, the proportion of lexical borrowings is relatively low. Typical borrowings are nouns, adjectives, adverbs and various discourse markers as illustrated in (2):
18 (2)
19
di lait leve družno
ARTඍ SBJ.PL PRS.1PL ADV
die Leute leben einträchtig
‘People live in harmony.’
20 In (2) the adverb družno, derived from the adjective družnyj, is used in the phrase die Leute leben einträchtig. The sample is a direct transfer of the Russian phrase žit’družno ‘to live in harmony’, which can be translated into Standard German as in Eintracht leben. As we can see, the transfer of a lexical item mostly triggers further borrowing of lexical and structural constructions. Like the adverb družno, which appears only once in the SGC, most of the lexical borrowings are used spontaneously. Only Russian discourse markers are more integrated in the German speech. The Russian discourse marker nu ‘well’ appears most frequently in the SGC.8
8. The discourse marker nun (34 tokens) is used less often than its Russian counterpart nu (59 tokens) in the SGC. Compared with other Russian borrowings, nu is used relatively frequently and it shows a higher heterogeneity concerning the word order type in the German contact variety. Most of the examples with nu surprisingly contain a verb-first order and not the typical German verb-second order [Andersen 2016a: 280ff]. Russian discourse markers seem to be a crucial linguistic unit also in other typologically different languages spoken in Russia as in an Udmurt-Russian mixed code [Kaysina 2014], which shows similar phenomena to German in Siberia. But apart from that, it is not the subject of the present study.
21 High frequency of finite auxiliaries is another typical property of the SGC, see table 1. As a German contact variety, German in Siberia is still a sample of spoken German. However, it cannot come as a surprise that the verbs are mostly used in the present and perfect tenses. Verb forms of the German past tense do not occur at all (exceptions are the auxiliaries war, hatte etc.). Another interesting quantitative result of the SGC is the absence of finite bilingual verbs. In other words, the German finite verbs are not linked with Russian morphemes in the sample under investigation. But about 1.5 % of the total number of borrowings in the corpus are nonfinite bilingual verbs with a Russian stem and German affixes as in example (1) above. On the other hand, there is not a single case in the SGC where the nonfinite verb consists of a German stem and Russian inflectional morphemes, which is certainly remarkable.
22 But studies which compare the speech of Russian and Russian-German immigrants in Germany show opposite patterns. Russian immigrant speakers produce bilingual verbs in Russian speech with a German verb stem and Russian inflectional morphemes, as the following examples show: behandlevat’ (behandeln ‘treat’, Russian infinitive suffix -vat’), vermissevaem (vermissen ‘miss’, Russian inflectional suffix -vaem, 1PLඅ, PLPRS) [Berend 2014: 232]. The same phenomenon is documented in the speech of Russian immigrants in Sweden. The Russian immigrants use Swedish verb stems with Russian infinitive suffixes as in bukovat’(boka ‘book’) and sjuklat’ (cykla ‘bike’) [Lisik 2013: 63].
23 The following section discusses the crucial typological differences between German and Russian morphosyntax.
24

3. Typological differences between German and Russian taking into account the morphosyntactic structure of spoken German in Siberia
25 In the spirit of Hawkins [1986], the contrastive approach will be applied as a complement of typological studies. Hawkins, who has suggested a contrastive typology of German and English, argued that it “adopts a methodology which is in many ways the exact inverse of the comparative-universal approach. Whereas this latter examines a small number of variant linguistic properties in a large number of languages, the present approach looks at a large number of variant linguistic properties in a small number of languages” [Ibid.: 3]. Moreover, the contrastive approach is motivated by the goal of this investigation. It is undeniable that the sample under investigation in the SGC is a German contact variety. Russian is present throughout as an embedded language in the vocabulary as well as in the grammatical structure of this German variety.
26 Even though Russian and German are not closely related languages, as Indo-European languages, they both have verbal systems that function as the core of the sentence, determining syntactic and semantic role assignments. In addition, German and Russian verbs have inflection-bearing morphemes, marked for person, number and tense. That means code-mixing in the verbal system is to be expected but is unpredictable in detail. To begin with, let’s have a look at an example [Andersen 2016a: 273] from the SGC:
27 (3)
hat nich gefunde weg
AUX.3SG NEG PTCP OBJ
hat nicht gefunden Weg
‘He did not find the way.’
28 (3′)
Ne našël put’
NEG PST.SG.M OBJ
not found way
‘He didn’t find / has not found / had not found the way.’
29 While (3) consists of exclusively lexical items from German, we assume that its grammatical structure exhibits traces of structural transfer from Russian. If we contrast the utterance with Standard German, it would correspond to Er hat den Weg nicht gefunden. But striking differences are the absence of the subjective pronoun, the exclusion of the object from the sentence bracket and the missing definite article of the object. In typology, German and Russian are classified as languages with SVO9 basic word order with relatively free constituent orders. However, the contrasts in usage-based word order are rather intricate. Table 2 shows the relevant typological features of German and Russian which have a crucial impact on the present investigation.
9. See more typological features of German and Russian in the World Atlas of Language Structures (WALS): [Dryer, Haspelmath (eds.) 2013].
30

Table 2. Typological contrasts between Russian and German
31 German has the characteristic feature of verb-final order (SOV) in subordinate clauses. Some linguists, like König and Gast, consider the SOV order in subordinate clauses as the basic word order for German. Among the arguments they give is the observation that German verbs with separable prefixes like Ich will, dass Karl das Licht ausschaltet – Karl schaltet das Licht aus leave their prefix behind in final position when they occur as finite verbs in a main clause [König, Gast 2009: 165]. This order of constituents is not possible in Russian, because it does not separate verbal prefixes. Moreover, German has analytical forms of the perfect and past perfect, whereas Russian has a synthetic preterite form, which semantically corresponds to these tenses. If we translate (3) into Russian, as in (3′), we derive the perfective (singular masculine) past tense form našël from the Russian infinitive pair najti / naxodit’ ‘find’. The German translation offers the past, perfect and past perfect tenses. It should be noted that the purity of only three Russian tenses (present, past and future) is compensated for by the grammatical category of aspect, including a morphologically different perfect and imperfect form for almost every verbal infinitive.
32 Since the Russian past tense našël also denotes the masculine gender, the masculine pronoun can be dropped without losing this grammatical information. In (3), the temporal auxiliary hat does not mark the gender morphologically, but the pronoun is dropped anyway. In (3′), moreover, the negation precedes the finite verb – ne našël.
33 Another conspicuous difference is that the Siberian German perfect hat gefunde evokes the German “sentence bracket” (Germ. Satzklammer) including the negation nich in the middle field. In German, the finite verb invariably occupies the second position in main clauses and exactly one constituent has to be placed in the position before the verb, in the so-called “forefield” (Germ. Vorfeld). Thus, in Standard German the pro-drop in (3) is considered to be salient.
34 The non-finite verb, here gefunde, typically occurs in final position. The position behind the non-finite verb (“extraposition”) is called “postfield” (Germ. Nachfeld). It usually contains subordinate clauses (cf. [König, Gast 2009: 167]). As we can see, there are several contrasting grammatical features, which can trigger the morphosyntax of the contact variety.
35

The German sentence bracket is perhaps the crucial typological feature in the German variety which has resisted any structural change. A noticeable finding is the exclusive usage of perfect tense when talking about the past10. The structure of perfect tense causes the German sentence bracket. As mentioned above, there are no tokens of German past tense forms (except auxiliaries) in the entire SGC. This finding does not seem to be a coincidence. Investigations on Colonial German in Eastern Europe that are almost a hundred years old show the same empirical results in parts of Ukraine, Russia and the South Caucasus (Transcaucasia), cf. [Ström 1926/1927; Schirmunski [1926–1931/1992]. The past tense was already replaced by the perfect a hundred years ago. Table 3 illustrates the frequency of auxiliary usage in the SGC.

10. In written Standard German, the past tense has the meaning of a narrative tense used primarily for describing past narratives (cf. [Duden 2016: 523]).
36

Table 3. Auxiliary usage in the SGC, the most frequent finite forms
37

It is obvious that about half of the finite verbs occurring in the SGC are variants of haben and sein followed by a lower frequency of the finite forms of können and müssen. In the majority of cases, haben is used in the perfect and past perfect, which build the sentence bracket. This strong structural contrast to Russian morphosyntax has not changed despite long-term language contact. Another interesting quantitative result is the substantial frequency of sein. Inflectional forms of sein are frequently used as copulas in nominal predicates, as in examples (4)–(6). In table 4, we can see the variation of the finite forms (types)11 of auxiliaries in the SGC.

11.The table shows the occurring word forms (types) of the most frequent auxiliaries in SGC. The auxiliary sein for instance appears in more than 19 different conjugational forms; every variant occurs in a number of tokens like bin (108), bist (9), war (460) etc., for more frequencies see table 3. The variety of conjugational forms can indicate the status of Siberian German: Spoken German, mixture of dialects, absence of Standard German etc.
38

Table 4. Conjugational patterns of auxiliaries (types) in the SGC
39 The greatest amount of variation appears in the temporal auxiliaries haben and sein in the first and third person plural followed by the first and third person singular. The most frequent modal auxiliaries are müssen and können in the finite forms of the first and third person plural. It is also obvious that the second person plural is not used in SGC (except habt, 13 tokens). In contrast, the occurrence and variation of the temporal auxiliary haben is very high, particularly in the first person plural (seven variants) and third person plural (six variants), for the frequency of the tokens see table 3. It means that the Siberian German speakers only use a few conjugational patterns. In these patterns, the variation of forms is evidently high.
40 Furthermore, the occurrence and variation of the modal auxiliaries12 sollen and möchten are very limited. The conjugational form soll appears 16 times, the form möchte only seven times and möchten eight times. In conclusion, it can be said that the usage of conjugational patterns of auxiliaries in SGC is limited to a few conjugational patterns of the auxiliaries haben and sein and the modal auxiliaries können and müssen. But the above discussed corpus data are difficult to explain. On the one hand the results show that Siberian German is still vital spoken language, but, on the other hand, the distribution of the conjugational patterns shows a limited field of verbal discourse.
12. The modal auxiliaries dürfen and mögen do not occur in SGC.
41 (4)
42
Wir war ganz allein.
SBJ.1PL COP.PST.3SG PRED
wir waren ganz allein
‘We were quite alone.’
43 (5)
Es war voennye.
3SG.N COP.PST.3SG PRED.PL
es war Soldaten
‘They were soldiers.’
44 (6)
Das sind ja auch billiger.
DEM.SG COP.PLPRS.3PL ADV ADV PRED
das sind ja auch billiger
‘They are cheaper, too.’
45 The number of the subject does not correspond with the inflection of the copula war, as illustrated in (4). In (6), the construction das sind does not correspond to the predicative billiger; in Standard German: Das ist ja auch billiger. Borrowing is an additional component in (5). In addition, in this case the number of the copula war does not match the number of the nominal predicative voennye, as is required in Standard German. The Russian voennye ‘militaries’ is inserted as a nominal predicative; in Standard German it would be Es waren Soldaten. We argue that the bilingual speaker retrieves both Russian and German morphological rules: the Russian predicative (plural) and the German (singular) copula war. We do not know how this cognitive process works, nor is it the topic of the present investigation. But we can make the argument by comparing the existing structural parameters. In particular, both grammars can form a nominal predicate using a copula — German sein or Russian byt’— followed by a predicative, but the Russian copula is not realized in present tense. This means that the irregular personal forms of the German copula sein (bin, bist, ist, sind, seid) correspond to a zero form in Russian. The German past tense forms of sein are war, warst, waren, wart, signalizing the categories of person and number, whereas the Russian past tense forms of byt’ are byl, byla, bylo, byli, signalizing the categories of number and gender, but not the category of person. Retrieving the morphologies of both languages, the speaker of the contact variety makes use of different rules from the two languages. The output is a contact-induced sentence construction constrained by both languages. To illustrate the underlying morphological diversity of the bilingual speech, the regular inflectional paradigms of the verbs occurring in the SGC can be compared to German and Russian as follows:
46

Table 5. Regular synthetic verb paradigms in German and Russian: spielen / igrat’ ‘play’
47 German and Russian have two synthetic verb paradigms, non-past and past. In table 5, the paradigms of the regular verbs spielen and igrat’ ‘play’ are compared in person, number and gender. The non-past tense forms in German and Russian mark person and number, and it is obvious that the German inflected forms show block syncretism, i.e. blocks of inflectional morphemes (cf. [Stump 2001: 217]); see also Synkretismusfelder, [Eisenberg 1998: 164]) in several cases: 3SG and 2PL, 1PL and 3PL in German non-past; 1SG and 3SG, 1PL and 3PL in German past tense. This is not the case for the Russian non-past forms. The inflectional suffixes in the Russian verb differ for every person and number, while the Russian past tense forms are uninflected for person, but inflected for number and also gender. But the grammatical category of gender appears neither in the Russian non-past paradigm nor in the German inflectional system.
48 We argue that the tension between partial uninflectedness (past) and form diversity (nonpast) in Russian on one hand and block syncretism in the German variety on the other hand are responsible for several morphological contact phenomena, as will be discussed in the following analysis of bilingual verb units.
49

4. Types of inserted Russian verb morphemes in Siberian German
50 In his study of bilingual speech, Muysken [2000] investigated a large number of bilingual verb formations in several language pairs in order to explore different ways of borrowing for a specific word class.13 He established four main types of bilingual verb constructions in typologically different languages: (i) the new verb is inserted into a position corresponding to a native verb, in an adapted form or not; (ii) the new verb is adjoined to a “helping verb”; (iii) the new verb is a nominalized complement to a causative “helping verb” in a compound; (iv) the new verb is an infinitive and the complement of a native auxiliary [Ibid.: 184]. Muysken’s contact-linguistic approach is clearly typological and general in nature, whereas the following analysis focuses on the particular case of inserted Russian verb morphemes in a German variety. The embedded Russian morphemes in the SGC are in all cases inserted into the position of a German nonfinite verb. That means the Russian morpheme occupies the position of a German perfect participle or an infinitive.
13. In my view, this study is still the most extensive overview of morphosyntactic borrowings in verb formations, but Russian borrowings are not mentioned.
51 There are two options in German grammar to create non-finite verb units: (1) the regular formation of the perfect participle is the prefix ge-, added to the verb stem, and the suffix -t; (2) the formation of the regular infinitive is the verb stem plus the suffix -en. Additionally, German irregular verbs often have different stem paradigms as in schreib (present stem) and schrieb (past tense stem); the perfect participle of schreiben is geschrieben. But in the corpus under investigation, bilingual verb units with the affixes geand -en do not occur. Therefore, it is not necessary to discuss more verb inflections in German than just the regular one. In the SGC data, Russian verbs basically come as one of two main insertions:
  1. The stem of a Russian verb is inserted into the position of the German perfect participle or infinitive and accompanied by German word-formation affixes.
  2. The inflected form of a Russian verb is inserted into the position of the German perfect participle or infinitive without German affixes.
52 Following the classification by Muysken [2000: 184], only the insertion of a borrowed verb in an adapted or unadapted inflected form occurs in the SGC. The types (ii–iv) above were not found in the corpus under investigation. One reason is probably that German and Russian are both fusional languages building the inflectional systems by suffixes and changing their semantic and syntactic functions by prefixes etc. But alongside the similarities, there are specific morphosyntactic contrasts, as is illustrated above; see specifically examples (3), (3′) and tables 2 and 5.
53 It is remarkable that in most of the examples in the SGC the inserted Russian verb units remain unadjusted to the morphosyntactic sentence structure in some way or other. The next section contains a detailed analysis of the bilingual verbs discovered.
54

4.1. Russian verb stems with German word-formation affixes
55 In (7)–(10), Russian verb stems are integrated in the rest of the German sentence. There is no other lexical borrowing from Russian.
56 (7)
das vas mr net vo ze vil hinpostupaje
DEM.SG PRS.1PL 1PL NEG Q 3SG AUX.3SG INF
das wissen wir nicht wo sie will eintreten
‘We don’t know where she wants to apply.’
57 (8)
un des nae johr vstretschaje um zvelf uhr
ART.N ADJ ACC INF
und das neue Jahr treffen um zwölf Uhr
‘And we meet New Year at 12 o’clock.’
58 (9)
militajres hebn sich gsobirajet
SBJ.PL AUX.3PL REFL PTCP
Soldaten haben sich gesammelt
‘The soldiers have met.’
59 (10)
mir huns Gapridilait
SBJ.1PL AUX.1PL PTCP
wir haben-es erkannt
‘We have realized it.’
60

The stems of the Russian verbs postupat’‘enter’, vstrečat’‘meet’, sobirat’‘collect’, opredeljat’ ‘identify’ seem to be rather seamlessly integrated into the German verb morphology. In (7)–(8), the Russian stem takes the German suffix -e14, and in (9)–(10), the Russian stem is integrated into the German perfect participle with the prefix gand the suffix -t, as summarized in table 615.

14. In (8), the German Suffix -e has the grammatical function of an infinitive suffix because most of the verbal infinitives of SGC have the suffix -e. This is a common feature in various German dialects and not only typical for Siberian German. In contrast, the regular infinitive suffix in written German Standard is the suffix –(e)n (cf. [Duden 2016: 446]). But in several Spoken Regional Standard varieties one can also find the pronunciation [ə], i.e. the suffix -e. Moreover, the suffix -e (schwa) is one of the most syncretic suffixes in the German nominal and verbal inflectional system including the plural of masculine and neutral nouns and the first person singular, present tense. In several Spoken German varieties, including SGC, the first person plural is often pronounced as schwa, transcribed as -e, like in example (2). However, discussing the various options here would go beyond the scope of the paper. 15. The Standard German affixes appear in brackets.
61

Table 6. Examples for integrated Russian verb stems with German affixes in the SGC
62 In example (7), the Russian stem takes the derivational prefix hinin addition to the infinitive suffix -e, while the borrowed Russian infinitive postupat’ does not have a prefix. Furthermore, it is striking that the (bilingual) infinitive vstretschaje in (8) is used instead of a finite verb and without the subject pronoun wir (‘we’).
63 Another crucial contrastive aspect affecting the morphology of bilingual verb forms is the existence of the verbal aspect in Russian. The selection of Russian verb stems by the bilingual speaker is due to the fact that Russian verbal infinitives are systematically represented by an infinitive pair, the perfective and the imperfective, see table 6. In the examples (7)–(10), the Russian verb stem is possibly derived from the imperfective infinitive. The vowel -ain the stem of the hybrid forms in (7)–(10) could be an indication of the imperfective form. But this is only a cautious assumption. In this study, it is not possible and not intended to consider the functional-semantic category of aspectuality in the Russian embedded verb forms, which does not mean, however, that this functional opposition of the two aspectual infinitives is not relevant. But the present study considers only the morphological form and not the function of this category.
64 However, this does not completely explain the existence of the formative -ain the bilingual verbs. The choice of the imperfective Russian verb stem might be motivated by phonetic analogy: if we separate the suffixes as -aje, -ajet, -ait, we arrive at potentially well-formed phonetic clusters in both languages. The phonetic value /ait/ is a part of several German stems such as arbeit ‘work’16, and /(a)je/ is a frequently occurring value in the SGC as in fraje ‘woman’ etc. But /aj/ is also the imperative singular in Russian verbs, and /ajet/ is the third person present tense in Russian verbs ending in -at’. It should be mentioned here that the Russian verb stems in (7)– (10) belong to the five most productive verb classes in Modern Russian (cf. [Sidorov 1945]). These factors — phonetic similarity and morphological frequency — could be evidence for contact-induced syncretism. At this point we argue with [Baerman et al. 2005] that a sort of nested syncretism is compounded across different environments as in the example of the West Slavonic language Upper Sorbian. In Upper Sorbian, “a-stem nouns have syncretism of the dative, locative in the singular, while all nominals have syncretism of the dative, locative and instrumental in the dual. Thus the syncretic pattern of the singular can be said to be nested within the larger syncretic pattern of the dual” [Baerman et al. 2005: 14].
16. Arbeit appears in SGC 77 times, the perfect participle gearbeit 15 times, but the phonetic value /ait/ is with a total frequency of 307 tokens strikingly often used in the SGC; for example, in the dialect verb forms geit and keit ‘goes’, steit ‘stands’ and in zeit ‘time’, krankheit ‘illness’ etc.
65 This is also supported by Muysken’s [2000: 192] observation that in the case of verb insertion, no extra structure is created, but the morphophonemic requirements of the language pair evokes the bilingual infinitive construction with an adapted stem, as in Dutch with French verb stems. French verbs can easily be integrated into Dutch when the stem is affixed with -er. However, the addition of -er to incorporate alien stems seems to be limited to French stems in standard Dutch, e.g. blesseren ‘hurt’ (< Fr. blesser). Muysken [Ibid.: 191] also found that Dutch bilingual verbs which derive their stems from French -ir verbs also take the suffixes -er and -en as in offreren ‘offer’ (< Fr. offrir).
66 As in Dutch, the German suffixes -er and -en are very productive word-formation morphemes. Basically, the German infinitive suffix -en can be seamlessly affixed to most English verb stems like save — saven, start — starten, move — moven, while in the German variety of the SGC the Russian verb stems need an additional interfix, as can be seen in (7)–(10), and also in table 6. Here it is undeniable that the language pair plays an important role in language contact.
67 In Standard Russian, verbs of foreign origin often need to be affixed with -ova-/-evafollowed by the Russian infinitive suffix -t’as in klassificirovat’, stilizovat’, kollektivizirovat’or remontirovat’. This morphological process also matters for the formation of bilingual verbs in the SGC as can be seen in (11):
68 (11)
69
war noch remontiert jez
AUX.PST.3SG ADV PTCP ADV
war noch renoviert jetzt
‘It was renovated then.’
70 The form remontiert is a Russian stem integrated in German verbal morphology. This is not necessarily obvious, because there is a verb remontieren in German — in fact, it is a seldom-used botanical term meaning ‘to blossom after the main flowering time’ (a French loan from remonter ‘wind up’). But the Russian loan verb remontirovat’‘renovate’ is a derivative of the noun remont and belongs to the class of ova/eva-stems, which often have interfixes such as -iror -iz(-ir-) as in stilizovat’, kollektivizirovat’ etc. [Mulisch (ed.) 1975: 68].
71 The context of remontiert in the SGC definitely supports the meaning of repair. However, the fact that the noun remont is also a loan in Russian makes it to an appropriate candidate for bilingual word-formation because the morphophonemic similarity of the Russian interfix -irin -irovat’ and the German -iermakes it easy to integrate the Russian stem into German morphology. The formation of the perfect participle with -ieren infinitives is frequently used in foreign loans such as diskutiert, pariert, fabriziert etc. Several verb tokens of passiert, palviert, interessiert etc. also appear in the GSC.
72

4.2. Bare inflected Russian verbs
73 A further group of inserted verb forms are inflected Russian verbs, as can be seen in examples (12)–(16). Most of the tokens found in the SGC are finite forms or in few cases passive participles, but there are no borrowed infinitives.
74 As is well-known, infinitives are basic morphological forms in traditional grammar. Bare infinitives also appear in several interim structures in first and second language acquisition. It is widely accepted that verbal infinitives occur in the two-word-phase of one year and 10 month old children as in sentences like Ball haben ‘ball have’; the morphological assignment is a later process [Müller 2013: 52].
75 Studies in second language acquisition with German as the second foreign language after English in Swedish schools show similar patterns. The learners choose the uninflected verb forms in an early phase using the suffixes -en and -t, but master the details of the different inflectional categories much later and often incompletely [Fredriksson 2006: 150–160].
76 However, in the present case of bilingualism, the speakers seem to avoid bare infinitives of the embedded language.
77 (12)
mir huns apridilili
SBJ.1PL AUX.1PL PST.PL
wir haben-es erkannt
‘We have defined it.’
78 (13)
die kvartira ist alles arendovan
ART.F SBJ.F AUX.3SG PTCP.M
die Wohnung ist alles gemietet
‘The flat is rented.’
79 (14)
Junge sind gange tantsuju
SBJ.PL AUX.1PL PTCP PRS.1SG
Jungen sind gegangen tanzen
‘The youngsters went dancing.’
80 (15)
so viel äpfel und die emer marinujut
PL DEM.PL AUX.1PL PRS.3PL
so viele Äpfel und die haben-wir eingeweckt
‘So many apples and we have them marinated.’
81 (16)
des alte johr provozhaj und des nae johr vstretschaje um zvelf uhr
ART ADJ ACC IMP.SG ART ADJ ACC INF
das alte Jahr begleite und das neue Jahr treffen um zwölf Uhr
‘Let us say goodbye to the old year and meet the New Year at 12 o’clock.’
82 In example (12), the inserted Russian apridilili is the past tense plural of the infinitive opredelit’. It correctly corresponds to the Siberian German subjective pronoun (nominative) mir ‘we’. This borrowed verb is also used in the SGC as an adapted form with German affixes — gapridilait’, ‘realize’ — as is illustrated in (10).
83 A different case is example (15). In this case, the inserted form marinujut is the third person plural of the present tense of the corresponding infinitive marinovat’‘marinate’; but the subject emer ‘we have’ is the first-person plural!
84 In example (13), the Russian participle arendovan ‘rented’ is incorporated in place of the predicative complement with the German copula ist. The German sein-passive ist gemietet ‘is rented’ is replaced by the Russian equivalent, actually the short form of the past passive participle, masculine arendovan from the infinitive arendovat’. Standard Russian has four participles (present active and passive, past active and passive), but only the short form of the past passive participle occurs in the SGC.
85 (17)
86
kvartira arendovana
SBJ.F PTCP.SG.F
‘The flat is rented.’
87 (17′) Die Wohnung ist gemietet. ‘Idem.’
88 As illustrated in (17), Russian perfective passive sentences are built synthetically from the short form of the passive participle, in this case with the inflectional suffix -a (feminine) in agreement with the feminine noun kvartira ‘flat’. The copula byt’ ‘be’ is not realized in the present tense. Unlike Russian, the translated German construction (17′) realizes the copula ist in the present tense. This is also the case in the contact variety, as is illustrated in example (13). The insertion of the Russian noun kvartira in (13) is particularly apparent because of the use of the corresponding German definite article die in the feminine form and the Russian participle arendovan. As can be seen in (17), gender agreement between subjective pronoun and predicative participle is obligatory in Russian. But in the German variety, see (13), the masculine form arendovan is used.
89 In example (14), the inflected verb tantsuju (1SG, PLPRS; from the infinitive tancevat’ ‘dance’) is inserted in the verbal predicate sind gange tantsuju. The word order is FINITE-NONFINITE-FINITE (borrowing).
90 In example (16), the imperative provozhaj (IMP, SG from infinitive provozhat’ ‘see off ’) is inserted together with the bilingual infinitive vstretschaje, discussed in (8), literally translated as “Begleite das alte Jahr und treffen das neue Jahr um 12 Uhr”. Still, the morphological non-alignment of the borrowed verbs is in a way pragmatically motivated. The grammatical function ‘invitation’ is an inherent feature of the Russian imperative.
91

5. Summarizing discussion
92 Based on assumptions of language typology and related principles of Natural Morphology, inflectional forms (a) “are (in general) more often and more systematically subject to analogical leveling than derivationally related words […]”, (b) inflections are “(prototypical) category-constant”, and (c) “inflectional formatives usually have a peripheral position in the word form” [Dressler (ed.) 1987: 5–6].
93 In the analysis of tokens of inserted Russian verbs occurring in the SGC, a striking amount of variation in morpheme combination and types of insertion has been discovered, but it has become obvious that the borrowed inflectional formatives basically show the features named in (a–c) above.
94 The variation of borrowed morphemes has to be related to the quantitative patterns of the finite and nonfinite verbs occurring in the corpus. In contrast to the large number of monolingual finite verbs, the occurrence of nonfinite verbs is much smaller (see table 1), and the number of bilingual nonfinite verbs is even lower. However, the bilingual verbs investigated show promising results. Here I want to turn around Myers-Scotton’s [2002: ix] thesis that “not anything structural is possible in contact speech,” and ask: what is possible in bilingual speech when German is contrasted with Russian in this contact variety?
95 First, it has been shown that the system of inflectional suffixes which are added to German verb stems do not show any morphological borrowing from the Russian verbal inflection over a period of 200 years. On the other hand, the variation of monolingual German verb suffixes is high compared to Standard German inflection (see table 4). It is especially significant for 1PL, PLPRS/PST: , -en, -n, -e, -t, -mr, -te and 3PL, PRS/PST: , -en, -n, -e, -t, -te, which correspond to -en (PRS.1/3PL) and -ten (PST.1/3PL) in Standard German.
96 Second, the inserted Russian verb forms in the SGC have the shape of an inflected form with or without German inflectional affixes, and these verbs are sometimes loans from other European languages. The total number of formatives in bilingual verbs is shown systematically in table 7.
97

Table 7. German and Russian formatives of bilingual verbs occurring in the SGC
98 In one case, the derivational prefix hinwas found, see example (7). But German prefixes compounded with Russian stems have been found in other corpora as well. For this reason, we believe that the single case in the SGC is not an incidental finding, see also example (1). The prefix hinis otherwise a regularly used word-formation morpheme in the SGC, as can be seen in table 7.
99 The prefix of the perfect participle g(e)was found in several cases in combination with the corresponding suffix -t. Since ge+ ඌඍൾM + -t is the regular formation of German perfect participles, the prefix g(e)has the highest number of tokens in the SGC, see table 7. It is assumed that the bilingual formation of the German perfect participle with a Russian stem is the most regular bilingual formation in this German variety. This is confirmed by Berend, who gives many bilingual examples, including some without the prefix but with the suffix -t, as in uznajt ‘recognized’, machajt ‘waved’ etc. [Berend 2013: 89]. The formative -tseems to be a very frequently used morphophonemic element in the entire inflectional verb paradigm of the SGC. It therefore seems that the formative -tis a contact-induced syncretic formative of spoken German in Siberia. It is a noteworthy formative in bilingual verb formation, as well as in the German conjugational paradigm, see table 4. Moreover, it also appears in the Russian non-past conjugation, see table 5. However, arguing for contact-induced inflectional syncretism is a bit risky. Several assumptions are required that these instances of “inflectional homophony” might be seen as systematic and “somehow represented in morphological structure” [Baerman et al. 2005: 9]. But of course this question seldom has a clear-cut answer. The most rigorous analysis of distinct pattern identity is given by Pike [1965]. He isolated the phoneme sequences of the German irregular verb sein ‘be’ and found identity in the elements /b/, /z/, /ai/, /ist/, /in/, /n/, and /t/, as in table 8.
100

Table 8. Identical phoneme sequences in German sein (adapted from [Pike 1965: 195])
101 In spite of the fact that this proposal is not compatible with most morphological models,17 the resulting parts are suitable for the present analysis of contact-induced verbal morphology. It is argued that the formatives -ai-, -t, -n, and also -( j)e form very productive patterns in the bilingual verb morphology, see table 8. The special aspect here is that these formatives occur in Russian, Standard German and also in spoken German in Siberia. The formatives are part of either the German or the Russian inflectional paradigm and merge in the bilingual morphological context. They do not have exactly the same function, but they are part of the system of verbal morphology. We propose here that they belong to the contact-induced syncretic inflectional system.
17. Baerman et al. [2005: 9] argue that these patterns of identity may be suitable for a specific analysis, but are not desirable in the context of a large-scale typological investigation. But for a contrastive analysis such as ours, this proposal represents a promising approach.
102 Finally, we should discuss the Russian inflectional suffixes -ili, -uju, and -ujut. In contrast to the formatives discussed above, they do not have analogous morphophonemic forms in German, but occur relatively frequently as Russian inflectional suffixes in the SGC, see table 7. As we saw in examples (12), (14), and (15), the inflectional morphemes -uju, -ujut have the functional values of grammatical person and number, and -ili only marks number, see example (12) and table 5. In the Russian inflectional system, they may be finite verbs, but in the contact-induced context they appear in the position of the German perfect participle.
103

6. Conclusion
104 This investigation has demonstrated that in a mixed language (the variety of Siberian German) the bilingual verb formations (Russian as the contact language) follow morphophonemic regularities to a certain extent:
  1. The borrowed Russian inflectional morphemes are exclusively units of Russian verbal morphology, i.e. the diversity of borrowed morphemes is restricted.
  2. The borrowed morphemes under discussion are derived from Russian verb stems belonging to the five most productive verb classes in Modern Russian.
  3. There are obviously several traces of syncretism in the entire verbal inflectional system including cross-linguistic verb constructions.
105 However, in spite of the regularities we uncovered, we can neither explain precisely why just those morphemes were chosen, nor can we say with certainty if these formatives are regular morphemes. If we finally have a look at the internal structure of the bilingual verb gapridilait ‘identified’ (example 10), components such as gor ga-, apridilaior pridil-, -ait or -aiand -t etc. could be possible morphological units of this bilingual token, if we assume the classical morphological approach18 of word-formation, i.e. that words consist of morphemes that are minimal meaningful units in language. However, the formatives investigated in the SGC do not have exactly the same form/function equivalents in both the German and Russian morphological systems. It is argued here that the bilingual speakers make use of German and Russian morphology in an innovative way and that this becomes apparent through the production of new (bilingual) types of verb formation. The high degree of variation in the verbal inflectional system of this German variety could be another indication of innovative morphological rules induced by language contact.
18. Classical morphology has its roots in structural linguistics, as demonstrated in [Hockett 1958].
106 For further research, it may be productive to focus specifically on cognitive aspects in bilingual morphology, taking into account different forms of language acquisition and other types of bilingualism.
107 ABBREVIATIONS
108 ACC — accusative ADJ — adjective ADV — adverb ART — article AUX — auxiliary verb COP — copula DEM — demonstrative pronoun F — feminine MPL — imperative INF — infinitive M — masculine N — neuter NEG — negative particle OBJ — object PL — plural PRED — predicate PRS — present tense PST — past tense PTCP — participle Q — interrogative pronoun REFL — reflexive SBJ — subject SG — singular

References

1. Andersen 2016a — Andersen Chr. Syntax in contact. Word order in a contact variety of German spoken in Eastern Siberia. Journal of Language Contact, 2016, 9(2): 264–292.

2. Andersen 2016b — Andersen Chr. Nachfeld im Kontakt. Eine Korpusuntersuchung am Russlanddeutschen in Sibirien. Göteborger Arbeitspapiere zur Sprachwissenschaft, 2016, 6. URL: https://gupea.ub.gu.se/ bitstream/2077/52526/1/gupea_2077_52526_1.pdf (accessed 31 July 2019).

3. Andersen 2016c — Andersen Chr. The status of Russian German in Siberia. A case study of four women living in the region of Krasnoyarsk (Russia). Call: Irish Journal for Culture, Art, Literature and Language, 2016, 1(1). Article 4. URL: http://arrow.dit.ie/priamls/vol1/iss1/4 (accessed 31 July 2019).

4. Baerman et al. 2005 — Baerman M., Brown D., Corbett G. G. The syntax-morphology interface. A study of syncretism. Cambridge: Cambridge Univ. Press, 2005.

5. Berend 2013 — Berend N. Über verschollene Dialekte und flexible Sprecher. Am Beispiel eines MutterTochter-Gesprächs im russlanddeutschen Migrationskontext. Zeichen und Stil. Der Mehrwert der Variation. Festschrift für Beate Henn-Memmesheimer. Albert G., Franz J. (eds.). Frankfurt am Main: Peter Lang, 2013, 81–93.

6. Berend 2014 — Berend N. Im Spannungsfeld zwischen Herkunftssprache, Dialekt und Standardsprache: Migration und Remigration am Beispiel russlanddeutscher und russischsprachiger Zuwanderer aus der ehemaligen Sowjetunion. Deutsche Sprache, 2014, 3: 218–235.

7. Borin et al. 2012 — Borin L., Forsberg M., Roxendal J. Korp – the corpus infrastructure of Språkbanken. Proc. of LREC 2012. Istanbul: ELRA, 2012, 474–478.

8. Dinges 1923 — Dinges G. Über unsere Mundarten. Beiträge zur Heimatkunde des deutschen Wolgagebietes. URL: http://wolgadeutsche.net/history/Ueber_unsere_Mundarten.htm (accessed 30 November 2017).

9. Dressler (ed.) 1987 — Dressler W. U. (ed.). Leitmotifs in Natural Morphology. Amsterdam: John Benjamins, 1987.

10. Dryer, Haspelmath (eds.) 2013 — Dryer M. S., Haspelmath M. (eds.) The World Atlas of Language Structures Online. Leipzig: Max Planck Institute for Evolutionary Anthropology, 2013. URL: http://wals. info (accessed 31 July 2019).

11. Duden 2016 — Duden. Die Grammatik. Duden Band 4. Berlin: Dudenverlag, 2016.

12. Eisenberg 1998 — Eisenberg P. Grundriss der deutschen Grammatik. Das Wort. Stuttgart: Verlag J. B. Metzler, 1998.

13. Fredriksson 2006 — Fredriksson Ch. Erwerbsphasen, Entwicklungssequenzen und Erwerbsreihenfolge. Zum Erwerb der deutschen Verbalmorphologie durch schwedische Schülerinnen und Schüler. (Studia Germanistica Upsaliensia, 50.) Uppsala, 2006.

14. Hawkins 1986 — Hawkins A. A comparative typology of English and German. Unifying the contrasts. London: Croom Helm, 1986.

15. Hockett 1958 — Hockett Ch. F. A course in modern linguistics. New York: Macmillan, 1958.

16. Kaysina 2014 — Kaysina I. The emergence of an Udmurt-Russian mixed code: Evidence from discourse markers. Eesti ja soome-ugri keeleteaduse ajakiri. Journal of Estonian and Finno-Ugric Linguistics, 2014, 5(2): 9–27.

17. König, Gast 2009 — König E., Gast V. Understanding English-German contrasts. Berlin: Erich Schmidt, 2009.

18. Lisik 2013 — Lisik L. Idiosyncrasies of the speech of Russian-speaking immigrants in Sweden. Russkii yazyk za rubezhom, 2013, 1: 58–65.

19. Myers-Scotton 2002 — Myers-Scotton C. Contact linguistics. Bilingual encounters and grammatical outcomes. Oxford: Oxford Univ. Press, 2002.

20. Moskaljuk 2013 — Moskaljuk L. Verschiedene Arten von Entlehnungen und Interferenzen in russlanddeutschen Inselmundarten. Göteborger Arbeitspapiere zur Sprachwissenschaft, 2013, 5. URL: https://gupea. ub.gu.se/bitstream/2077/52525/1/gupea_2077_52525_1.pdf (accessed 31 July 2019).

21. Mulisch (ed.) 1975 — Mulisch H. Morphologie. Die russische Sprache der Gegenwart. Band 2. Leipzig: Verlag Enzyklopädie, 1975.

22. Muysken 2000 — Muysken P. Bilingual speech. A typology of code-mixing. Cambridge: Cambridge Univ. Press, 2000.

23. Müller 2013 — Müller H. M. Psycholinguistik – Neurolinguistik. Die Verarbeitung von Sprache im Gehirn. Paderborn: Wilhelm Fink, 2013.

24. Pike 1965 — Pike K. L. Non-linear order and anti-redundancy in German morphological matrices. Zeitschrift für Mundartforschung, 1965, XXXII. Jahrgang: 193–221.

25. Putnam (ed.) 2011 — Putnam M. (ed.). Studies on German-language islands. Amsterdam: John Benjamins, 2011.

26. Seidenberg, Gonnerman 2000 — Seidenberg M. S., Gonnerman L. M. Explaining derivational morphology as the convergence of codes. Trends in Cognitive Scienes, 2000, 4(9): 353–361.

27. Selting et al. 1998 — Selting M., Auer P., Barden B., Bergmann J., Couper-Kuhlen E., Günthner S., Meier Ch., Quastoff U., Schlobinski P., Uhmann S. Gesprächsanalytisches Transkriptionssystem (GAT). Linguistische Berichte, 1998, 173: 91–122. URL: https://www.mediensprache.net/de/medienanalyse/transcription/gat/gat.pdf (accessed 31 July 2019).

28. Schirmunski 1926–1931/1992 — Schirmunski V. Linguistische und ethnographische Studien 1926–1931. Ed. by Hutterer C. J. München: Verlag Südostdeutsches Kulturwerk, 1992.

29. Ström 1926/27 — Ström A. Deutsche Mundarten an der Newa. I. Teuthonista. Zeitschrift für deutsche Dialektforschung und Sprachgeschichte, 1926/27, 3: 39–62.

30. Sidorov 1945 — Sidorov V. N. Verb classification. Ocherk grammatiki russkogo literaturnogo yazyka. Part I. Fonetika i morfologiya. Avanesov R. I., Sidorov V. N. Moscow: Uchpedgiz, 1945.

31. Stump 2001 — Stump G. Inflectional morphology. Cambridge: Cambridge Univ. Press, 2000.

Comments

No posts found

Write a review
Translate