This paper explores patterns in the integration of Hungarian and Romanian nouns as well as adjectives in the German dialect of the speech community of Palota, a German Sprachinsel in North-West-Romania. The main focus of the study is on both inflectional and derivational noun and adjective morphologies and on how they behave in the case of some more or less distantly related contact languages. Based on a select number of examples from first hand data and following standard code-mixing models such as that of Muysken (2000) and Myers-Scotton’s (1993, 2002) mlf model, it establishes a typology of code-mixing morphology ranging from more matrix language-like, i.e. German-like to more embedded language-like, i.e. Hungarian- and Romanian-like patterns and bare forms, suggesting an ongoing shifting process in the local German dialect of Palota towards a fused lect (Auer 1998). In terms of linguistic complexity, the present paper argues that this language shift process favour simplification of morphology in some domains, but also complexification in some other domains, supporting the idea that languages in long-term intensive contact settings become linguistically more complex (Trudgill, 2010, 2011; Fenyvesi, 2005; de Groot, 2005, 2008).
In studies of contact linguistics, bound morphemes, measuring them on a “borrowing scale” (Thomason and Kaufman, 1988 for the term), are generally described as being more difficult to borrow than free morphemes, and only under specific social circumstances, i.e. intense contact situations (Matras, 2010: 77–82; Chamoreau, 2012: 72). On this scale, content and function words figure higher, i.e. they are more “borrowable” than derivational suffixes, and derivational suffixes are more prone to borrowing than inflectional morphology. It was also suggested by Field (2002) that agglutinating affixes enable borrowing at a higher position in the hierarchy than fusional affixes (Matras, 2010: 78). Likewise, hierarchies of content-lexical word classes predict that nouns are more borrowable than non-nouns, and adjectives score higher than verbs (Wilkins, 1996: 114). One explanation for this noun-dominated asymmetry lies in the referential function of these words (Kovács, 2001: 130; Haspelmath, 2008: 49). Moreover, their structural autonomy facilitates their integration into another language (Matras, 2010: 78). Myers-Scotton (2002: 240) suggests that “nouns are borrowed more frequently than any other category because they receive, not assign, thematic roles”.
The aim of the present paper is to examine both inflectional and derivational noun and adjective morphologies and on how they behave in a multiple German-Hungarian-Romanian language contact situation in which the languages in contact do not share a typologically similar morphological system, Hungarian being an agglutinative language as opposed to the mainly fusional languages German and Romanian. The main focus of the study is given to the German dialect of the speech community of Palota, a German Sprachinsel in North-West-Romania, which, due to several historical-political and macrosociological changes during the last centuries has become the subject of two language shift processes that now run parallel, resulting in the convergence of linguistic structures of three genetically and typologically more or less related languages.
The main goal of the study is to present a typology of the major patterns used by bi-/trilinguals to structurally integrate Hungarian and Romanian intrasentential, phonologically unintegrated, accidentally borrowed single nouns and adjectives as well as whole noun phrases into the German matrix language. Therefore, the term “code-mixing”, as it has been used by Muysken (2000), seems to be appropriate to refer to all cases of the juxtaposition of two or more languages. According to Muysken’s classification of code-mixing into insertion, alternation and congruent lexicalization, content words such as nouns and adjectives are likely to be insertions (Muysken, 2000: 97). In Myers-Scotton’s (1993, 2002) Matrix Language Frame (mlf) model that will be taken as a theoretical background for the grammatical analysis, insertions correspond to mixed embedded language (el) and matrix language (ml) constituents, alternations to el islands combined with ml islands, and congruent lexicalization is akin to the notion composite ml (Muysken, 2000: 17).
In a more recent line of research at the intersection between linguistic typology and sociolinguistics, it has been suggested that in contact situations extralinguistic factors interfere as well in the behaviour of bound morphology in that they can lead to both structural simplification and complexification of a language/variety, respectively (Adamou, 2012: 143). Trudgill (2009, 2010, 2011) argues that different types of contact settings have different effects on morphology, high-contact settings involving adult language learning leading to simplification of morphology as opposed to stable long-term contacts involving child bilingualism that favour complexification via borrowing of additional morphology. He argues further that small community size and isolation may promote complexification and also the growth of morphological irregularities, redundancy and low transparency (Trudgill, 2009: 109). Thus, in the case of Palota, a small speech community with long-term extensive, but non-stable language contact, isolated from the other German speaking communities, the expectations for the German dialect are ambiguous. It is also the aim of the present study to at least tentatively answer this question.
The paper is structured as follows: Section 2 provides a brief outline of some relevant structural properties of the contact languages in order to better understand what the source languages have on offer and into what kind of a system the el elements have to be integrated in the recipient language. In sections 3 and 4, the types of structural integration of Hungarian and Romanian nouns and adjectives in the German dialect of Palota (in what follows, referred to as gp) will be presented. Finally, section 5 offers some concluding remarks.
2 Noun and Adjective Morphology in German, Romanian and Hungarian
2.1 German and the German Dialect of Palota
German is an inflectional language with a grammatical three-gender system, i.e. each noun is marked on the determiner as a masculine, feminine or neuter. Gender assignment in German follows semantic, morphological and phonological rules, thus grammatical gender does not necessarily reflect the natural sex of its referent. Besides gender, German nouns are also marked for number and case. The plural marking system of Standard German (StG) involves eight different allomorphs for plural formation, the use of the one or the other depending on several factors like gender and number of syllables (Boas, 2009: 226). The four-case system of StG consists of the nominative, genitive, dative and accusative cases. They encode syntactic and semantic information in the noun phrase. The assignment of a particular case to a noun phrase is mostly determined by a particular verb or preposition.
Attributive adjectives normally precede the head noun and are marked together with the determiner according to the gender, number and case of the head noun they modify. There are three paradigms of endings, depending on the art of the determiner the adjective is following after in the noun phrase: the weak paradigm after the definite article, the strong paradigm after the zero-article and the mixed paradigm, a combination of weak and strong adjective endings used after the indefinite article ein and its negated form kein. Adjectives in a predicative position are not declined. The degrees of comparison are mostly realised via morpho-syntactic means, the comparative taking the inflection ending -er and the superlative both the inflection ending -sten and the preposition am; suppletion occurs only in a few cases.
German dialects differ from StG, not only what phonology and lexicon concerns, but also in morpho-syntactic respect. Moreover, there are also differences between different dialects. The German settlers of Palota came from various parts of Germany, thus they brought along different donor dialects which over the time developed through mixing and leveling to a new dialect. According to Gehl (2003: 39), the gp has its origins in Palatinate and Mosel Franconian dialects. However, German-speaking inhabitants of the village use to call it Swabian or simply unser Sproch ‘our language’ and they treat it as a separate language, distinguishing it clearly from StG, which is seen as a foreign language and which is not understood by mostly old people.
Regional varieties are in general distinguished by their less complex morpho-syntactic structure as compared to the standard variety. In nominal morphology, there was particularly drawn attention to variation in gender marking, plural formation and case marking. In the gp, a reduction of the plural allomorphy from eight to three morphemes could be attested1: -e (e.g. Finster-e ‘windows’ instead of StG Fenster-Ø), -er (e.g. Bett-er ‘beds’ instead of StG Bett-en), -r after the diminutive suffix -che (e.g. Mädche-r ‘girls’ instead of StG Mädchen-Ø). The case system also shows simplification by the loss of the genitive case and the reduction of the number of case endings, thus a noun having one single form for all cases in the singular and in the plural, respectively (e.g. Haus – Haiser ‘house’ instead of StG Haus, Hauses, Häuser, Häusern). A similar simplification took place in the inflection paradigm of the attributive adjective, too, the ending -e being largely overgeneralized for both the strong and the weak paradigm. However, the strong endings -er for the masculine and -es for the neuter can also occur. No doubt, such changes have likely to do with the influence of the contact languages, too.2
Romanian is considered a highly inflectional language with a rich inflectional system particularly in the nominal paradigms (Mallinson, 1986: 205). Both nouns and adjectives are marked for number and case by means of suffixes. Romanian distinguishes five cases, there are, however, no inflectional differences between nominative/accusative and genitive/dative. The only explicit distinctions are made by the use of the preposition pe, marking the syntactic status of the accusative as the case of the direct object, and by the presence of possession articles in the genitive case, respectively. Prepositions play an important role in marking syntactic-semantic functions. Whereas the accusative, the genitive and the dative are prepositional cases, the nominative and the vocative are non-prepositional cases (Guțu Romalo, 2005: 72). Romanian nouns are categorized into three genders, however the neuter is morphologically identical with the masculine in the singular and the feminine in the plural. The term gender is meant grammatically, but has some relationship to natural sex. In addition to number and case marking, definiteness represents an important morphological feature of Romanian nouns. In contrast to German, the definite article in Romanian is an enclitic article, rendered by a definite suffix.
Attributively used adjectives are in most of the cases found after the head noun they modify, the position in front of the head noun representing a topicalized pattern. Both attributive and predicative adjectives agree with their head nouns in number and gender. The derivation of non-positive forms takes place with preposed, unbound markers.
In contrast to German and Romanian, Hungarian is an agglutinative language with a rich morphology. The syntactic and semantic functions of noun phrases are expressed primarily by case endings, the number of cases in Hungarian ranging from 17 to 27 (Kenesei et al., 1998: 191). Besides the four cases existing also in German and Romanian, Hungarian possesses a series of locative cases for the use of nouns as adverbials.3 In German and Romanian, these syntactic and semantic functions are expressed by using prepositions. The nominative case is unmarked. Like in German and Romanian, nouns can have plural markers (-k/ok/ak/ek/ök) that appear in front of the case suffix. Hungarian lacks gender distinction.
In Hungarian, attributive adjectives occur to the left of the head noun (Szabolcsi, 1994: 184). They do not agree with the overtly expressed head noun either in number or in case. They can be inflected when their head nouns are not expressed overtly, i.e. they constitute pro-forms (Kenesei et al., 1998: 331). Predicative adjectives are marked for number. The comparative of adjectives is formed with the -bb/abb/ebb suffix, the superlative by attaching the prefix leg- to the comparative form. Since Hungarian is an agglutinative language, derivation has an important role in the word formation process. There are a number of suffixes that derive adjectives from nouns, verbs, adjectives, adverbs and numerals.
Although nouns are the most commonly insertions in the gp, Hungarian elements are not at all restricted to the category of nouns and adjectives that will be analysed in sections 3 and 4. There are also a notable number of verbs and particles in a broader sense, including adverbs, conjunctions, interjections and discourse markers that entered the language of the German speaking community.
3 A Typology of Noun Morphologies in Contact
Especially the first three options seem to be similar to the strategies suggested by Muysken (2000) and Myers-Scotton (1993, 2002), the first type corresponding to insertional code-mixing, the second to the so-called “bare forms”, and the third to alternational code-mixing.
(1) To treat borrowed nouns just like native nouns, and integrate them into native inflection patterns. (2) To avoid integration and maintain just a simplified representation of borrowed nouns. (3) To integrate nouns along with their original inflection in the source language. (4) To apply a special integration strategy that marks out borrowed nouns as loans.
In this section, various strategies developed by bi-/trilingual speakers in the speech community of Palota in order to morphologically integrate Hungarian and Romanian nouns will be discussed. Phonologically integrated, fully established loan words which emerge in the gp from the Romanian majority language via Hungarian, the target language of the first shift that had become a minority language itself in Romania, were excluded from the following analysis (see for example the noun kondzselátor ‘freezer’ in example 9). The examples4 were drawn from audio-recorded interviews conducted by the author with 33 speakers all of German descent, both men and women, who were born in Palota and who had lived there for their whole lives. The speakers, who range in age from 3 to 88, are all German-dominant and most of them are fluent in gp, Hungarian and Romanian. Only some elderly German-Hungarian bilinguals are not familiar with Romanian.
3.1 Bare Forms
Myers-Scotton (1993: 112; 2002: 21) asserts that using bare forms, i.e. el content morphemes without system morphemes from either language that would make them well-formed in a ml frame, is a common strategy if there is not sufficient congruence between the two languages. However, she calls attention to the fact that these bare forms must be distinguished from those which would occur with no obligatory system morphemes even in a monolingual el utterance.
In the case of the Hungarian noun gyerektartás ‘child support’ in example (1), the accusative case ending -t is missing, according to the rules of the el. Moreover, the el noun is neither preceded by a German definite article that would explicitly mark it as a direct object. The use of the ml zero-article is in this case motivated both by the ml and el: in the corresponding German and Hungarian sentences, the noun also has to be marked with the zero-article. Thus, although a conflict between the ml and the embedded item’s grammar was avoided, the example presented fits only partially the definition suggested by Myers-Scotton. It rather seems to be a form between a bare and an unmarked form.
3.2 Mixed German-Hungarian and German-Romanian Constituents
Quite frequently, Hungarian and Romanian nouns are treated like native nouns being embedded in the gp inflection pattern, using the unmarked form in the nominative case of the el noun and ml system morphemes. These can be unbound morphemes such as articles and prepositions leading to mixed ml + el noun or prepositional phrases, but also bound inflection morphemes such as plural markers.
As to case marking, unmarked nouns fit well into the German ml as subjects and predicative nouns as can be seen in the following German-Hungarian examples:
Similar to the Hungarian nouns, Romanian noun insertions in the unmarked nominative form satisfy the German ml as well. Consider example (4), where the Romanian noun capră ‘goat’ (here ‘blockhead’) is used as a predicative:
The following two examples show cases of switched Hungarian (5) and Romanian (6) objects:
In both examples, the switched singular unmarked nouns are provided with the masculine German definite article in the accusative case, according to the matrix morphology. Regarding gender, the Hungarian noun zaj ‘noise’ becomes masculine, probably by analogy with German Lärm; in the case of the neuter Romanian dicționar ‘dictionary’, there is, however, no agreement. A possible explanation for this disagreement lies probably in the ambigeneric character of the Romanian neuter.
Embedded nouns as adverbials usually occur as prepositional phrases in which the noun of the source language is unmarked and the German ml provides the preposition and in some cases the article. The Hungarian (7–9) and Romanian examples (10–11) illustrate cases for locatives where the German preposition in ‘in’ corresponds both with the Hungarian inessive/illative case and the Romanian preposition în ‘in’ which would have been expected, according to the el morphology:
It should be noted that the two-way German preposition in was in all of the cases used according to the matrix morphology. It governs the accusative case in examples (9) and (11) in order to express direction/destination and the dative case in (7), (8) and (10) in order to express stable position/situation, thus corresponding, at least in the Hungarian examples, with the opposition illative/inessive, respectively.
A similar course is also found for embedded Hungarian and Romanian unmarked nouns following e.g. the ml prepositions of (StG auf) ‘on’, zu ‘to’, fier (StG für) ‘for’. Consider the following German-Hungarian (12–13) and German-Romanian examples (14):
Example (14) reveals the insertion of the Romanian noun tanti ‘Mrs.’ into the German ml as a prepositional phrase with the gp preposition fier and the German definite article die where, according to the German ml, the dative case would be appropriate. The outcome of the prepositional form is certainly interpretable in two ways: On the one hand, taking into account that the speaker is dominant in gp and Romanian, it could be the result of an analogy to a prepositional use in Romanian (la tanti ‘to the Mrs.’). However, the Romanian preposition la would rather correspond with the German preposition zu. On the other hand, because the speaker has also a good command of Hungarian, the preposition fier could also correspond with the Hungarian dative case ending -nak.
In section 2.1, it was noted that in the gp there is a decrease of morphological plural markers. The data show the use of the plural allomorphs -e and -er in the case of Hungarian (15–17) and the use of -s in the case of Romanian nouns (18). They occur in structural contexts such as the German plural determiner die and quantifiers like viel ‘much’ or zwei ‘two’:
The determining factors of the use of a certain allomorph are yet to be examined in greater detail. Based on the examples, the phonological shape of the word could be such a factor. So, Hungarian singular nouns ending in a consonant will get a plural morpheme beginning with a vowel. To a Romanian noun ending in a vowel that is perhaps felt a more foreign word as a Hungarian noun, the plural morpheme -s will be added. The vowel -a in mutr-a-s stands for the Romanian feminine enclitic definite article. Thus, the outcome is a mixed construction containing an el island and a ml system morpheme.
3.3 Paradigm Transfer along with el Nouns
A third type of noun morphologies in contact represents the insertion of Hungarian nouns along with their original inflection in the el. This is the case in example (19) where, in contrast to example (1), the Hungarian noun gyerektartás ‘child support’ as a direct object is provided with the Hungarian accusative case ending, thus the sentence being well-formed according both to the el and ml morphology.
Similarly, the Hungarian noun internátus ‘boarding school’ in example (20) acquires the ending of the Hungarian illative case and is integrated in the German ml as an adverbial:
It should however be noted that in this example the inessive case (-ban ‘in(side)’) would be appropriate for the local adverbial instead of the illative case (-ba ‘into’). This is, however, a quite frequently used practice in spoken Hungarian in which the inessive case merges with the illative case.
In this case, a transfer of inflectional morphemes takes place leading to a juxtaposition of a ml and el island and thus fulfilling the syntactic functions required by the ml. Heine and Kuteva (2005) term this type of morphological integration “paradigm transfer”. Since there can be observed a shifting away from the analytical prepositional forms discussed in section 3.2. to a more synthetic, agglutinative morphology, this can mean an increase of system complexity.
3.4 Double Marking
Double marking applies to cases where el content morphemes get “affixes (system morphemes) from both the el and the ml for the same function” (Myers-Scotton, 1993: 132), thus agreeing both with the matrix and the embedded language. Cases of double marking have been quite frequently observed by Földes (2005: 155–159) in the German dialect of Hajós in present-day Hungary. However, Kovács (2001: 196) points out, based on her study of Australian Finnish and Hungarian, that double marking can be more or less common in some language contacts.
Matras (2009: 174) notes that especially plural markers and definite articles are frequently doubled through native morphology. Both cases could be attested in the data. Consider the following excerpts from the German-Romanian (21–22) and German-Hungarian (23) set of data:
In example (21), the analytic German masculine definite article is doubled through the Romanian enclitic definite article -ul. However, the case of the Romanian definite suffix (due to case syncretism, it is unclear whether it is in the nominative or accusative case) does not coincide with the dative case of the German definite article. The same applies to the Romanian feminine definite article -a and the German definite article die in the dative case in example (22). In example (23), the Hungarian plural marker is probably regarded by the speaker as part of the stem, when embedded into the German ml.
Among the German-Hungarian examples, cases of Hungarian nouns that receive both a German preposition and a Hungarian case-marking suffix could be observed:
To the Hungarian noun egyetem ‘university’ in example (24), the suffix of the superessive case -en ‘on’ was added, and is at the same time preceded by the gp preposition of (StG auf) ‘on’ which is the “copy” of the Hungarian case marker. Since in the German ml the preposition an ‘on’ would be appropriate, the mixed prepositional phrase seems not to be well-formed from the perspective of the ml. Moreover, both the ml and the el would require a definite article. Thus, this example seems to be a case for “half-marking” (Kovács, 2001: 151, 165), a category which falls between the full agreement and non-agreement with the ml and the el.
3.5 Transfer of el Markers without el Nouns
Winford (2010: 176) maintains that only certain kinds of bound morphology can be borrowed, mostly in “situations of close typological fit between languages, where substitution of one inflection for another is facilitated”. He states further that borrowing of overt inflectional morphology appears to be very limited in contrast to bound derivational morphology.
The German-Hungarian example (25) illustrates the import of the Hungarian plural marker -ek accompanied by the ending of the causal-final case -ért (-ér in spoken Hungarian) ‘for’ in the German ml:
Because the resulting mixed construction resembles the German equivalent prepositional phrase für ‘for’ + definite article + proper noun (family name) + plural marker -s, this example is likely to be a case for what Muysken (2000) calls “congruent lexicalization”. The same pattern is applied in this example to the common noun Freind (StG Freund) ‘friend’ too, a construction closely related to the equivalent Hungarian construction. Because very few examples have been found in the data,5 it could be only tentatively suggested that a grammaticalization process is taking place here supporting Auer’s (1998) model of a fused lect. However, the German prepositional construction zu dene Freinde ‘for their friends’ could be a counterargument to this assumption.
4 The Insertion of Hungarian and Romanian Adjectives in the gp
el adjectives constitute a much smaller inventory in the data than nouns and even verbs. They tend to be integrated syntactically both as predicative adjectives and as attributive adjectives along with the head noun or without.
The German-Hungarian (26–27) and German-Romanian (28) examples show well-formed constructions, because the inserted unmarked forms satisfy the demand of the German grammar, i.e. no agreement markers on the predicative adjective with the subject. Moreover, the adjective forms would also satisfy the requirements of the el (see sections 2.2 and 2.3).
Regarding these examples, two aspects related to the inserted Hungarian adjectives deserve special attention. In example (26), the adjective gyáva ‘cowardly’ is additionally derived with the German derivational suffix -ig. Thus, the category adjective is in this case double marked. This seems to be a quite frequently used pattern in the speech community of Palota (see also examples 30–31). In example (27), there are two Hungarian adjectives that carry the Hungarian comparative markers -abb/ebb. This is a case of an el island, the integration of the adjectives being realized along with their original inflection. The examples provided by Földes (2005: 159–160) are similar to those shown here.
Inserted Hungarian attributive adjectives occur in cases in which the head nouns they modify are not switched (29–31) as well as in cases in which the whole noun phrase is switched (32):
The adjective következő ‘next’ in example (29) has been left unmarked, as it would be expected in Hungarian, but it is preceded by the German definite article that is marked for the accusative case. In examples (30–31), the switched adjectives are embedded in the German noun phrase according to the rules of the ml, being provided with the gp inflectional morpheme -e. It is worth noting that in both cases the speakers turn the Hungarian adjectives into German adjectives by adding the German derivational suffix -ig. Moreover, example (30) seems to be even a case for triple marking: the Hungarian root adjective bordó ‘wine-red’ is provided with the Hungarian derivational suffix -s meaning ‘more or less adj’ and then with the German derivational suffix -ig. In the case of szereplő-s ‘with casts’, it should be noticed that this adjective is not inflected according to the German ml. This fact might lead to the assumption that the function of the German derivational suffix -ig is to prepare the inserted Hungarian adjective for inflection in the ml noun phrase.
Finally, example (32) illustrates the case of a wholly switched Hungarian noun phrase with zero-marking that could be considered a bare form, since it lacks the required ml system morphemes in order to integrate it as a direct object in the German ml as well as the el personal possessive suffix -a marking the thing possessed, if the equivalent Hungarian sentence is considered.
This paper has illustrated various types of noun and adjective morphologies in a language contact situation characterized by structural convergence of sometimes genetically and typologically different languages. Based on a select number of examples from the German-Hungarian-Romanian speech community of Palota (North-West-Romania), a closer look at the morphological integration of Hungarian and Romanian nouns as well as adjectives has revealed a variety of patterns ranging from full agreement with the ml (mixed ml + el constituent), agreement with both the ml and el (double marking) to agreement with the el (el islands) and non-agreement with both the ml and the el (bare forms). It was also possible to observe that, despite of typological differences, code-mixing is taking place at the level of both derivational and inflectional morphology, Hungarian (and also Romanian) derivational and inflectional suffixes being transferred in the German ml along with or without the el stem. The linguistic outcomes of the contact can be seen, on the one hand, as maintenance of certain elements of the ml. But at the same time, the increase in the use of double morphology and of grammatical elements from the el, which in some cases develop functional specializations, as well as of whole el islands seem to support the ongoing shifting process of present-day gp towards a composite language, a “fused lect” (Auer, 1998).
As to the linguistic complexity of the gp matrix language, it was shown that the gp, being the local variety of a German Sprachinsel, is characterized in its nominal system by simplification in nearly all domains: reduction of the plural allomorphy, loss of case endings, simplification of the adjective inflection paradigm. The frequent use of unmarked noun and adjective forms from the contact languages that in some cases turn out to be bare forms, as well as the use of more analytic expressions by the integration of el elements where the el would use synthetic expressions can also be taken as a decrease in the system complexity of a mixed language. Opposed to the decrease, there is, however, also an increase of complexity located in the additional use of German derivational adjective suffixes on Hungarian adjectives in order to prepare them for attribute inflection in the German ml, as well as in adopting agglutinative noun morphology in a fusional ml. These and other aspects such as gender assignment to nouns from languages with more or less similar gender systems or variation in the use of marked and unmarked forms are to be considered responsible for irregularity, redundancy and lack of transparency (see Trudgill, 2009, 2010, 2011).
Returning to the relation between typology, language change by contact and complexity suggested by Trudgill (2009, 2010, 2011), the language shift process in the speech community of Palota seems to favour both simplification and complexification of morphology, thus supporting the observations made by Fenyvesi (2005) and de Groot (2005, 2008) for Hungarian in contact spoken outside Hungary. Since this shifting process at the micro level goes hand in hand with a language shift at the macro level determined by extralinguistic factors, the case of Palota seems also to support Adamou’s (2012: 165) conclusion that the outcomes of language contact are strongly related to the type of contact rather than to typological factors.
The topic and most oft he data of this paper were presented at the Workshop Languages abroad, in Gothenburg, in October 2012. I wish to thank Christiane Andersen for the invitation and the participants for their helpful comments.
suffix deriving an adjective
suffixial definite determiner
(unspecified) inflectional suffix
AdamouEvangelia. 2012. Verb morphologies in contact: evidence from the Balkan area. In VanhoveMartineStolzThomasUrdzeAina and OtsukaHitomi (eds.) Morphologies in Contact143–162. Berlin: Akademie Verlag.
AuerPeter. 1998. From codeswitching via language mixing to fused lects: Toward a dynamic typology of bilingual speech. Interaction and Linguistic Structures 6: 1–28.
ChamoreauClaudine. 2012. Spanish diminutive markers -ito/-ita in Mesoamerican languages. A challenge for acceptance of gender distinction. In VanhoveMartineStolzThomasUrdzeAina and OtsukaHitomi (eds.) Morphologies in Contact71–90. Berlin: Akademie Verlag.
De GrootCasper. 2005. The grammars of Hungarian outside Hungary from a linguistic-typological perspective. In FenyvesiAnna (ed.) Hungarian Language Contact Outside Hungary. Studies on Hungarian as a minority language351–370. Amsterdam: John Benjamins.
GrootDeCasper. 2008. Morphological complexity as a parameter of linguistic typology: Hungarian as a contact language. In MiestamoMattiSinnemäkiKaius and KarlssonFred (eds.) Language Complexity. Typology contact change191–215. Amsterdam: John Benjamins.
HaspelmathMartin. 2008. Loanword typology: Steps toward a systematic cross-linguistic study of lexical borrowability. In ThomasStolzBakkerDik and PalomoRosa Salas (eds.) Aspects of Language Contact. New Theoretical Methodological and Empirical Findings with Special Focus on Romancisation Processes43–62. Berlin: Mouton de Gruyter.
SzabolcsiAnna. 1994. The Noun Phrase. In FerencKiefer and KissKatalin É. (eds.) The Syntactic Structure of Hungarian179–274. San Diego: Academic Press.
TrudgillPeter. 2009. Sociolinguistic typology and complexification. In SampsonGeoffreyGillDavid and TrudgillPeter (eds.) Language Complexity as an Evolving Variable98–109. Oxford: Oxford University Press.
TrudgillPeter. 2010. Contact and Sociolinguistic Typology. In HickeyRaymond (ed.) The Handbook of Language Contact299–319. Malden: Wiley-Blackwell.
WilkinsDavid P. 1996. Morphology. In HansGoeblNeldePeter H. and StarýZdeněk (eds.) Contact Linguistics. An International Handbook of Contemporary Research 109–117. Berlin: Walter de Gruyter.