Printer Friendly

Gender and inflection class in loan noun integration.

1 Assumptions and premises

In language-contact studies, special attention has been devoted to lexical borrowing, more specifically to loanword accommodation (see, among others, Brown 1999, Winter-Froemel 2008, Haspelmath & Tadmor 2009, Thomason & Kaufman 1988, Thomason 2014). Of all categories, it is stated that nouns are more easily borrowed (Whitney 1881, Moravcsik 1978, Hock & Joseph 1996) and Matras (2009: 168) attributes this fact to their referential properties.

Various factors, language internal and external, have been claimed to contribute to the transfer of nouns from one language to another. For instance, beside the vital role of socio-political and economic (language external) factors, which facilitate borrowing in contact settings, there are also language internal requirements which govern the process between the system that exerts a controlling influence (source language or donor) and the affected language (target or recipient), such as form similarities, structural and semantic equivalences (see, among others, Ibrahim 1973, Poplack, Pousada & Sankoff 1982, Winford 2005, 2010).

It is generally stated (Thomason 2001, inter allia) that loan nouns are firstly adopted without being analyzed, while an analysis, or a reanalysis, come at a second stage. There is usually more than one strategy according to which a word can be inserted in the recipient's morphology. For instance, Wichmann & Wohlgemuth (2008: 99) have proposed that verbs can be inserted in a language directly or indirectly. In the first case, verbs are transferred by taking on slight (or none) phonological modifications, while in the second case, loan verbs become compatible with the requirements of the recipient's morphology only with the support of some functional elements, for instance affixes.

In this paper, we deal with the integration of loan nouns in two Modern Greek dialects, Heptanesian and Pontic, in differently conditioned situations of linguistic contact. (141) Our study shows differences, but also similarities in the way the dialects in question handle inflection, more specifically, grammatical gender (unless necessary, hereafter simply gender) and inflection-class assignment in their loan noun integration, notwithstanding their contact with genetically and typologically unrelated systems: Heptanesian has been affected by the semi-fusional Romance, whereas Pontic has been influenced by the agglutinative Turkish. It is demonstrated that, independently of the properties of the donor, the integrated nouns bear an overt inflectional ending according to the recipient's standards, where the structure of a native noun is a combination of a stem and an inflectional ending. It is worth noticing that the inflectional part of the loan word may either be a Greek ending, added to the loan -when reanalyzed as a stem- or come from the reanalysis of the final segment of the loan into a Greek inflectional ending. Since the adaptation of loan nouns requires only the presence of inflection, in accordance with the morphology of Greek native nouns, but there is no use of extra material, as for instance, an integrating derivational suffix which would flag membership to the category of nouns, we assume that the items under examination enter the recipient by following a semi-direct insertion strategy. In contrast, the compulsory presence of an integrator would denote indirect insertion. As shown by Ralli (2012a,b, 2014), the latter applies to loan verbs, where the Greek verbalizer -iz- is, for instance, used in the Aivaliot (142) dialect for the accommodation of verbs of Turkish origin (e.g. Greek/Aivaliot kazad-iz-u 'to become rich' < Turkish past tense kazadi).

As exposed in the following sections, our investigation reveals: (a) the predominant role of the morphology of an inflectionally-rich language, that is, Greek, for the inflectional adjustment of nominal loanwords (see also Aikhenvald 2000, 2006 and Ralli 2012a,b, 2013 for similar claims); (b) a certain role played by a form matching of the endings between the native nouns of the donor and those of the recipient language; (c) tendencies of the recipient language, to classify its nouns by distinguishing between native and loans in terms of inflection class and apply neuter gender to -human loans. (143)

Our data are drawn from the available written sources (inter alia Angelopoulos 2008, Pomonis-Tzaglaras 2007, Kasimatis 1996, Kollas 1960 for Heptanesian and Dawkins 1916, Topharas 1932, Oekonomides 1958, Papadopoulos 1955, 1958-1961, Drettas 1997 for Pontic), the databases and the digitized oral material of the Laboratory of Modern Greek Dialects (www.lmgd.philology.upatras.gr) of the University of Patras.

The paper is organized as follows: after the introduction, section 2 investigates the notions of gender and inflection class in Greek and its dialects, and shows the correlation between the two features. In section 3, a sketchy description of the socio-historical background of the two dialects is offered, and certain properties are pointed out regarding their features of gender and inflection class. Heptanesian and Pontic data are analyzed in section 4, where claims and proposals are put forward with respect to the morphology of [+/- human] loan nouns. In particular, the interplay of semantic, morphological and phonological factors underlying gender and inflection-class assignment is thoroughly examined. In section 5, there is a recapitulation of the main arguments discussed in the paper. The paper ends with two appendices, Appendix I which provides a general picture of Greek nominal inflection classes and the segmentation of nouns into stems and endings (based on Ralli 2002), and Appendix II which contains maps of the Ionian/Heptanesian islands and the Pontus area.

2 On gender and inflection class

According to Corbett (1991: 1) gender is "the most puzzling of grammatical categories". It constitutes a distinctive feature of nouns and contributes to their classification.

The grammatical gender of a noun is distinct from natural gender (sex), the latter being based on the relevant attributes of its referent. However, it usually correlates with it for nouns expressing animacy (Dahl 2000), or more specifically 'humanness' for certain languages, as shown by Ralli (2002) for Standard Modern Greek (hereafter SMG). (144) Grammatical gender does not characterize every language, but when it exists, its assignment may depend on semantic and formal (phonological and morphological) criteria.

The notion of 'default gender' has been used in many different senses in the literature; it is connected to the less marked option, is usually called 'prototypical gender', and it is the category with most members (Corbett & Fraser, 2000). In this paper, we show that, in the dialects under investigation, there is a certain preference for assigning the neuter value to -human loan nouns, suggesting that it may be considered as the default gender value, in accordance with Ibrahim (1973), Poplack, Pousada & Sankoff (1982), Kilarski (2003) and Stolz (2009). In fact, the neuter value has been already proposed as the unmarked default gender option for SMG by Dressler (1997), Anastasiadi-Symeonidi (1994), Grandi (2002) and Christofidou (2003).

Gender is a fundamental morphological characteristic of SMG where nominal words are specified for one of a tripartite value system, that is, masculine, feminine or neuter. As Ralli (2000, 2002, 2005) has shown, SMG nouns are combinations of stems and inflectional endings and gender is a property of stems, actively involved in the process of inflection, in that inflected nouns inherit their gender value from their stems, as in (1). On the contrary, the feature values of case and number are inherited from the endings:

(1) a. [??]romos (145)        <       [??]rom          -os
       'road.MASC.NOM.SG'             'road.MASC'      NOM.SG

    b. proo[??]os             <       proo[??]         -os
      'progress. FEM.NOM.SG'          'progress.FEM'   NOM.SG

    c. vuno (146)             <       vun              -o
      'mountain.NEU.NOM/ACC/VOC.SG'   'mountain.NEU'
    NOM/ACC/VOC.SG


Like stems, derivational suffixes are also marked for gender. As a result, if a noun contains a derived stem, produced by the combination of a stem and a derivational suffix, the gender value of the morphologically-complex stem is that of the suffix:

(2) xorizmos               <          [[[xor   -iz ]   -m]   -os]
    separation.MASC.NOM.SG location   -ate     ion.MASC      NOM.SG
    'separation'                      'separate'


Ralli has further shown that for +human nouns, grammatical gender is closely related to the biological sex of the referent, in that male noun stems are masculine while female ones are feminine. In this respect, Ralli's assumption is slightly different from Dahl's (2000: 99-100) which relates gender to the more general feature of animacy. In fact, in Greek, the grammatical gender of nouns denoting animals is not predicted by sex distinctions (Ralli 2002: 531).

In literature, gender is often claimed to be related with the feature of inflection class (see, among others, Corbett 1991, Aronoff, 1994, Ralli 2002). However, there is no agreement among scholars as to the dominance direction of this relation. For instance, Aronoff (1994) considers as 'normal' the dominance that goes from gender to inflection class, while Corbett (1991) for Russian and Ralli (2002) for SMG -human nouns provide strong evidence for the opposite case. We believe that since both grammatical gender and inflection class are not universal features, the direction of dominance between the two is language-dependent or even case-dependent. With respect to this, we will see in section 4 that the Greek dialectal data does not provide clear evidence.

According to Ralli (2000) Greek distributes its nouns into eight inflection classes (ICs) of varying productivity, two for masculine nouns (IC1 and IC2), two for feminine, (IC3 and IC4) and four for neuter (IC5, IC6, IC7, IC8). (147) She bases this division on the presence or absence of allomorphic variation of noun stems as well as on the form of the inflectional endings. For instance, an IC8 noun like soma 'body' (see Appendix I) is subject to the systematic allomorphic stem variation X~Xt (e.g. soma-0 in nominative singular but somat-os in genitive), and its paradigm differs from those of the nouns belonging to the other classes. Like gender, the inflection-class feature characterizes noun stems. However, as opposed to gender, inflection class is also a property of the endings. Thus, it functions like a matching device between stems and endings, ensuring the well-formedness of the inflected nominal structures. As stated in Ralli (1999), the peculiar character of inflection class, as compared to the other features of nominal inflection, is justified by the fact that its function is purely morphological: it does not participate in agreement and is not generally visible by any syntactic process. As such, Ralli postulates that, in inflectional structures, this feature does not percolate to the topmost nodes of inflected words but is only an indicator of the form of their paradigm and a matching device for the right combinations of stems and endings. Consider the following inflected words for an illustration of these assumptions:

(3) a. [??]romos            <   [??]rom -os
       'road.MASC.NOM.SG'       'road.MASC.IC1'   NOM.SG.IC1

    b. vuno                 <    vun -o


'mountain.NEU.NOM/ACC/VOC.SG'mountain.NEmC5'NOM/ACC/VOC.SG.IC5

The distribution of SMG nominal inflection into eight inflection classes is depicted in (4), where inflected nouns are given in the citation form, that is, in the nominative singular, while glosses for case and number are omitted as irrelevant for the argumentation. (148)

(4) SMG

a.   Masculine nouns
     skilos 'male dog'      < skil.MASC.IC1     -os.IC1
     pateras 'father'       < patera.MASC.IC2   -s.IC2
     maGitis 'student'      < maoiti.MASC.IC2   -s.IC2
     kafes 'coffee'         < kafe.MASC.IC2     -s.IC2
     papus 'grandfather'    < papu.MASC.IC2     -s.IC2

b.   Feminine nounsmitera
     'mother'               < mitera.FEM.IC3    -o.IC3
     tixi 'luck             < tixi.FEM.IC3      -o.IC3
     alepu 'fox'            < alepu.FEM.IC3     -o.IC3
     poli 'town'            < poli.FEM.IC4      -o.IC4

c.   Neuter nouns           < vun.NEU.IC5       -o.IC5
     vuno 'mountain'
     spiti 'house'          < spiti.NEU.IC6     -o.IC6
     kratos 'state'         < krat.NEU.IC7      -os.IC7
     soma 'body'            < soma.NEU.IC8      -o.IC8


Interestingly, most of the Greek dialectal varieties share similar gender and inflection-class properties. (149) There are some exceptions though. For instance, in Pontic, there are relics of the Ancient Greek third inflection class, which is preserved to mark definiteness in masculine nouns (see section 4.1). (150) Moreover, in Cappadocian (151), especially in its Southern variety, there is a significant levelling of inflection classes, and a tendency to lose the tripartite grammatical gender distinction in favor of the neuter gender form, principally observed in the use of the article (Janse 2004, forthcoming, Karatsareas 2009, 2011).

It is worth noticing that in both SMG and its dialects, the adoption of most nouns from another language is accompanied by the assignment of inflection, as also pointed out by Anastasiadi-Symeonidi 1994, Christofidou 2003, Melissaropoulou 2013a,b, 2014, Makri, Koutsoukos & Andreou 2013). We would like to claim that this is due to a minimum requirement imposed by the Greek system -whose inflectionally rich nominal words are combinations of stems and endings- in order to accommodate foreign nouns. However, there are few instances where borrowed nouns remain uninflected, mainly in the recent adoption of foreign terms ending in a consonant which denote technical objects. In these cases, inflection is indirectly expressed by the article, which does not contain an overt distinction between a stem and an inflectional ending and the morphosyntactic features are incorporated into the word itself. The following examples, drawn from SMG, depict integrated (5a) and non-integrated elements (5b), while the source languages are Turkish and French, respectively:

(5) a. SMG o parali-s    <     Turkish parali
       det.MASC.NOM.SG         wealthy.man.MASC-NOM.SG

    b. SMG to makijaz    <     French maquillage
       det.NEU.NOM.SG        'make-up'


Finally, as will be presented in the subsequent sections, the comparison of loan data from the dialects under examination, reveals resemblances, but also incongruities, in terms of inflection as far as gender and inflection class are concerned.

3 The dialectal data

3.1 Heptanesian

Varieties of the Heptanesian dialect are spoken on the islands of the Ionian sea, Corfu, Cephalonia, Ithaca, Zante, Paxi, Kithira and on the smaller islands of Othoni, Herikusa, Mathraki, Antipaxi. (152) These islands were the only part of Greece that was not conquered by the Ottoman Turks but went under Venetian rule for four or five centuries (ca. end of 14th -beginning of 19th c.), depending on the island. The imposition of Standard Italian as the official language used in administration and education (Fanciullo 2008) and of the Venetian dialect for broad communication on a daily basis endowed Heptanesian with a considerable number of foreign features (Papageorgiou 1994, Kontosopoulos 2001), which are mostly detectable in its phonology, morphology and vocabulary.

Comparing the two systems in contact, that is, Romance as donor and Heptanesian as recipient, it is worth pointing out that they share some features with respect to inflection (e.g. gender and number), but the Romance system is poorer in overt forms, since case has disappeared, and Romance does not display the wealth of inflectional paradigms that we observe in Greek. Moreover, while Romance has a two-gender value system, Greek displays a tripartite value one.

Interestingly, Matras (2009: 174) mentions the possibility of gender maintenance between contact languages having more or less similar gender systems, and claims that languages which assign gender to their nouns, equally assign gender to borrowed words as well. In this light, we expect Heptanesian to assign gender to loan nouns originating from Romance. And in fact, nouns inserted from Standard Italian and Venetian either preserve or modify their original gender value in order to adapt to the new morphological needs imposed by the target system. However, we will see in 3.2 that loans are allotted gender even if their source language (in our case Turkish) is not characterized by grammatical gender, suggesting that for the integration of borrowed words the morphology of the recipient language prevails over that of the donor.

As will be seen in the examples below, Heptanesian masculine loan nouns end in -os, -is, -as, -es in the citation form (6a), like those in SMG (4), but feminine loan nouns end only in -a (6b) and neuter nouns in -o, or -i (6c). In other words, there are no loans assigned to IC4, IC7 and IC8, which, nevertheless, include many examples of native nouns. This is not surprising as IC4 (feminine nouns in -i, e.g. poli-o 'town', see Appendix I) and IC7 (neuter nouns in -os, e.g. krat-os 'state') comprise +learned nouns, that is, those which are either remnants from Ancient Greek or are built according to Ancient Greek patterns. As for IC8, with some exceptions, it involves mainly deverbal nouns, which presuppose a combination of a verb stem and the derivational suffix -ma (e.g. jemizma 'filling' < jemiz 'to fill' + -ma). Note that in Heptanesian, there are loans in -ma containing a borrowed base, as for instance, premurarizma 'care, willingness', but they derive from verbs in -aro (e.g. premuraro 'to care') which are also derived structures on the basis of Italian nouns (e.g. premura 'attention, care, consideration, haste'). These examples will be excluded from our examination since they constitute cases of secondary/indirect transfer.

(6)    Heptanesian                             Italian/Venetian
    a. Masculine                               Masculine
       avokatos     'lawyer'                   avvocato
       dzeneralis    'general of the army'     generale
       arkivistas    'archivist, file clerk'   archivista
       lavorantes    'worker'                  lavorante

    b. Feminine                                Feminine
       infermiera    'female nurse'            infermiera

    c. Neuter                                  Neuter
       apartamento   'apartment'               appartamento
                                               Feminine
       tsenturi       'belt'                   cintura


3.2 Pontic

Pontic is an Asia Minor dialect originally spoken in a geographical area which is spread over 400 kilometers (from Inepolis to Colchis) in the northeast coast of Asia Minor, as well as in parts of the inland located about 100 kilometers from the coast (Tombaides 1996). The emigration of the 19th century led to the establishment of Pontic communities in Caucasus, whereas the population exchange in accordance with the Lausanne Treaty in 1923, following the Greek-Turkish War (1919-1922), resulted in the subsequent massive movement of Pontic people to mainland Greece.

Since then, the dialect has been used by second and/or third generation refugees, mainly in northern Greece, but can also be found in few cities of Ukraine and Georgia. It is noteworthy that the dialect is still spoken in certain dialectal enclaves in the western part of Trebizond (Tonya and Ophis), by Muslim Pontics, who were exempted from the population exchange for religious reasons. It is usually called Muslim Pontic or Romeyka (Mackridge 1990).

Pontic preserves a number of archaic features, typical of earlier stages in the history of Greek, retains a number of shared features with the rest of Greek varieties of Asia Minor and exhibits contact-induced features from Turkish, which has affected Pontic both on the vocabulary and the structural level.

Unlike the Greek-Romance case, contact between Greek and Turkish instantiates the interaction between two typologically different systems, i.e. the fusional Greek and the agglutinative Turkish. Turkish does not have inflection classes and in terms of gender, we deal with a 'battle' between a gendered language (Greek) and a morphologically gender-neutral one (Turkish). (153) Thus, whereas in the Greek-Romance pair Greek as a recipient language could possibly accept some gender features of the donor, a similar influence should not in principle be possible when it comes to Turkish loanwords. Nevertheless, as already said in 3.1, Turkish loan nouns also receive gender and inflection class according to the inflectional needs of Greek, demonstrating the importance of morphology in this language. (154)

In Pontic, loan nouns are almost exclusively accommodated as masculine in -is or -as in the citation form (rarely in -es) that is, as nouns inflected according to IC2 (7a), and, despite some rare exceptions (e.g., tsopanos 'shepherd' in (7a)), adaptation of Turkish nouns according to IC1 (ending in -os in the citation form) is not common. Feminine borrowings are adapted as nouns ending in -i, -a, -e (IC3), while neuter loan nouns end in -in (IC6).

(7)    Pontic                                 Turkish

    a. Masculine
       tsopanos    'shepherd'                 coban
       pekiars     'unmarried man'            bekar
       hovardas    'spender, womanizer'       hovarda
       kioses      'bald man'                 kose

    b. Feminine
       orospi      'prostitute                orospu
       balduza     'bride                     baldiz
       kaxpe       'prostitute                kahpe

    c. Neuter
       kartali-n    'hawk'                    kartal
       kindi-n      'the time of nightfall'   ikindi
       poi-n        'height'                  boy


For the same reasons exposed for Heptanesian, IC4, IC7 and IC8 are not found among loans, although many native nouns are inflected according to these inflection classes. However, as opposed to Heptanesian, IC5 is also missing from loans due to a historical evolution, according to which many ancient nouns in -on (today's nouns of IC5) got restructured into nouns in -ion in early medieval period (Browning 1969), and with the ultimate loss of /o/, they emerged as IC6 nouns in -in. While native Pontic nouns in -on can still be found (e.g. aeropon 'breeze'), together with the more recent ones in -in (e.g. a[??]elfin 'brother'), loan nouns are uniquely adapted as those in -in. (155)

4 The interplay of semantic, morphological and phonological factors

In what follows, we demonstrate that gender assignment to nominal loanwords and their integration into Greek inflection classes are subject to various criteria, that is, phonological, morphological and semantic, separately or conjointly. As already mentioned in section 1, Ralli (2002) has shown that for gender assignment in SMG, the role of semantics is important, since +human noun stems are assigned masculine or feminine value, depending on whether they are male or female (based on sex). However, there is no one-to-one correspondence between gender and inflection class, since, as shown in Appendix I, both +human masculine and feminine nouns are distributed into more than one inflection class. In contrast, for SMG -human nouns the dominance relation suggests a direction from inflection class to gender.

The same priority to semantically driven gender seems to apply to borrowed +human nouns as well (reanalyzed as stems), although not with the same faithfulness, as will be depicted below. Intriguingly though, we will see that in -human loans gender can both dominate or be dominated by inflection class.

4.1 +Human loan nouns

(8)    Heptanesian                                   Italian

    a. imbresari.MASC-os   'agent'                   impresario.MASC
    b. dzenerali.MASC-s    'general of the army'     generale.MASC
    c. arkivista.MASC-S    'archivist, file clerk'   archivista.MASC
    d. abitante.MASC-s     'dweller'                 abitante.MASC
    e. infermiera.FEM-o    'nurse.woman'             infermiera.FEM
    f. insenianta.FEM-o    'female teacher'          insegnante.FEM

(9)    Pontic                                        Turkish

    a. pekiar.MASC-s       'unmarried man'           bekar
    b. hovarda.MASC-s      'spender, womanizer'      hovarda
    c. kiose.MASC-s        'bald man'                kose
    d. tsopan.MASC-os      'shepherd'                coban
    e. orospi.FEM-o        'prostitute'              orospu
    f. balduza.FEM-o       'bride's sister'          baldiz
    g. kaxpe.FEM-o         'prostitute'              kahpe


As (8-9) illustrate, in both Heptanesian and Pontic, borrowed nouns are accommodated following the semi-direct strategy (see section 1), since beside inflection they do not need any other particular integrator for their accommodation. Interestingly, when a foreign +human noun is transferred to these systems, it becomes subject to the needs of Greek morphology. That is, it undergoes a reanalysis into a stem and an ending (8a), or only into a stem (e.g. 8c, 9b), in order to adapt to Greek inflection, and it is assigned masculine or feminine gender, depending on whether it denotes a male or a female entity. In addition, it is incorporated into a Greek inflection class, most often on the basis of a form matching (156) between its own ending and the final segment of a corresponding Greek stem (e.g. 8c) or inflected word (e.g. 8a). (157)

In Heptanesian (8), +human male loanwords are accommodated as masculine nouns in -os, -is, -as, -es in the citation form of nominative singular (8a-d), while +human female ones are feminine, ending in -a (8e,f). More particularly, Romance masculine nouns ending in -o (e.g. avvocato, impresario) are adapted to Heptanesian analogically to native Greek IC1 inflected nouns, which also end in -o in the accusative case (-os in the nominative, see the inflection of the native noun anoropos 'man' in Appendix I). An assignment to IC1 triggers a reanalysis of the Romance word and a split into a stem and an ending, following the pattern of equivalent Greek nouns, which identifies the -o as the mark of the accusative singular (e.g. avokat-o, imbresari-o). As a result, more Greek endings can be added to the stem, that is, the endings of the entire IC1 paradigm, among which, the -os of the nominative case (e.g. avokat-os, imbresarios) in the singular number.

The question which arises now is if, instead of the proposed reanalysis of a Romance loan in -o into a stem and an ending, and the subsequent adaptation of -o as an exponent of accusative case and singular number, this Romance word final -o is transferred as such in Heptanesian, that is, as an inflectional marker. In other words, one may wonder whether Heptanesian, beside lexical items as a whole, borrows structure as well. Following a rather 'retentionist' position (see, among others, Meillet 1921, Field 2002), we believe that, since the two linguistic systems, Greek and Romance, are not fully compatible on the nominal inflectional level, the reanalysis of borrowed nouns occurs at a second stage, that is, after nouns enter the recipient system as items with no structure. (158) In fact, contrary to Romance whose inflectional suffixes fuse the features of number and gender, Greek ones fuse together number and case -gender being a feature of stems. Thus, the lack of exact correspondence between the two languages could not facilitate the transfer of structure. (159)

Romance masculine nouns ending in -e and -a are integrated in Heptanesian as IC2 items, where they are reanalyzed as stems, analogically to other masculine nouns ending in -e and -a in the accusative case (-es and -as in the nominative case, see the inflection of tamias 'cashier'and kafes 'coffee' in Appendix I). Contrary to IC1 nouns, the endings -e and -a are not reanalyzed as pieces of inflection though, since in Greek, e and -a are seen as part of the stem. Once integrated as stems, these loan nouns receive the appropriate Greek endings of IC2, as for instance, the -s inflectional ending of the nominative case in the singular number (e.g. tamia-s, kafe-s). It should be noted that, very often, Romance loans in -e undergo a slight phonological modification of their final -e, adopted as nouns in -is (e.g. Romance generale > Heptanesian dzeneralis). We believe that, that this change occurs because, beside the fact that /e/ and /i/ are not distant vowels, nouns in -es are few and most of them non-native. (160)

Phonological shape regarding the final segments between the donor and the recipient can also trigger membership to IC3 for Romance feminine loans (8e), where matching of Romance and Greek final segments prompts assignment to the particular inflection class. However, even if the Romance feminine noun does not end in -a (8f), integration into the Greek system is still made as such, since this category constitutes the most frequent stem type of Greek feminine nouns today, as already shown by Anastasiadi-Symeonidi (1994) and Ralli (2005).

It is worth noticing that the effect of form-matching factor in loan noun integration is so strong that, sometimes, it may override the semantically-driven gender assignment, and produce a small number of double-gender formations in the singular number, where masculine forms exist side-by-side with neuter ones. As depicted in (10a-c), a word final vowel matching in the singular number between Romance masculine nouns in -o and Greek nouns of either IC1 in the accusative case or IC5 in the syncretic forms of nominative, accusative and vocative cases triggers assignment of two different gender values in the Heptanesian loans, that is, masculine and neuter:

(10) a. Hept. no[??]ar-o                    as native anorop-o
              'notary.MASC-ACC.SG'          'man'
              and

     b.       no[??]ar -o                   as native vun-o
              'notary.NEU-NOM/ACC/VOC.SG    'mountain'
              vs.

     c. Ven.  nodar-o
              'notary-MASC.SG'


In contrast, in the plural number (11a-c), there are no double gender formations because the inflectional suffixes of the two contact languages match only if the Greek nouns belong to IC1 in the nominative case:

(11) a. Hept. no[??]ar -i                   as native an[??]rop-i
              'notary.MASC-NOM.PL'          men'
              vs.

     b. Ven.  nodar -i
              'notary-MASC. PL'
              but

     c. Hept. *no[??]ar -a                  as native vun-a
              'notary.NEU-NOM/ACC/VOC.PL'   'mountains'


Again, the fact that neuter gender assignment occurs only in the singular number but is impossible in the plural (see *no[??]ara in 11c), strongly supports the role played by phonology for the integration of nouns, since the -i plural ending in Romance (see 11b) does not coincide with the -a plural ending of neuter nouns in Greek/Heptanesian, as shown in (11c):

As argued in section 3.1, the gender properties of the target language override those of the source. As far as Pontic is concerned, Turkish loans are assigned a grammatical gender, as opposed to Turkish which is genderless, at least grammatically. As depicted in (9), Pontic human male loans become masculine, whereas human female loanwords are allotted the feminine value. Moreover, similarly to Heptanesian, the phonological matching of the endings between the donor and the recipient also plays a vital role for ascribing these nouns to a particular inflection class: on the one hand, Turkish male nouns in -a and -e are accommodated in Pontic as IC2 masculine nouns, that is, as nouns ending in -as and -es in the citation form of nominative singular (9a-c). On the other hand, female nouns ending in -i and -e acquire inflection according to IC3 (9e-g), even when the original item ends in a consonant (9f). In the latter case, the loan noun ends in -a; thus, as assumed above, it is accommodated in accordance with the most frequent type of feminine nouns in Greek.

Phonology is also important for male loans ending in Turkish in a consonant (9a), but only in conjunction with a native Pontic phonological rule which deletes final unstressed /i/s. In fact, a Turkish word like bekar 'unmarried man' is accommodated in Pontic under the hypercorrected form pekiar(i)s, where an /i/ is believed to precede the ending -s, ultimately thought to have been phonologically deleted. As such, the form is ascribed to IC2 (see also Melissaropoulou forthcoming for the same assumption).

Comparing now the inflection of masculine loans in Pontic to the native one in effect, we observe an indubitable preference for IC2, that is, for the inflection class of masculine nouns in -as/-is in the citation form, as opposed to the inflection of native nouns where IC1 (i.e. nouns ending in -os, e.g. anorop-os 'man') is equally or even more productive. (161) Linguists dealing with this phenomenon (Hatzidakis

1907, Kyranoudis 2009, Malikouti Drachman & Drachman 1989) have attributed the low productivity of the -os masculine loans (IC1) to the position of stress. They have argued that the vast majority of loans of Turkish origin are stressed on the ultimate or the penultimate syllable, while Greek native nouns ending in -os in the citation form often bear stress on the antepenultimate (provided that their length is more than two syllables).

However, the position of stress does not seem to be crucial for loan nouns in SMG and other Asia Minor dialects affected by Turkish (e.g. Aivaliot (162)), where the same loans display endings of IC1 (-os in the nominative singular):

(12) a. Pontic     kolayuzi-s         'driver'    <    Tr kilavuz
                                                       'guide'
                   vs.

     b. SMG        kolauz-os          'follower'

     c. Aivaliot   kulayuz-us (163)   'follower'


We would like to propose that the different usage of the two inflection classes in Pontic, as compared to SMG and other dialects, serves classificatory purposes: IC2 prevails in loans, while IC1 involves mostly native nouns and is only exceptionally used for loans (7a and 9d). This is also supported by the fact that Pontic has maintained a large number of archaic features, among which, traces of the Ancient Greek third inflection class (e.g. the nominative singular forms lik-on 'wolf and pap-on 'grandfather' in 13b) in order to express definiteness. As shown below, masculine nouns in the definite form are inflected according to this inflection class, as opposed to forms of the common IC1 (-os in the nominative singular, as in 13 a) which are used to denote the notion of indefiniteness:

(13) a. lik-os 'wolf.MASC-NOM.SG.INDEF          b. lik-on wolf-NOM.SG.
        pap-os 'grandfather.MASC-NOM.SG.INDEF'     DEF'pap-on
                                                   'grandfather-
                                                   NOM.SG.DEF'


Crucially, this inflectionally realized distinction between definite and indefinite forms applies only to native IC1 animate masculine nouns and excludes loans.

Io the question now why a classification into native and borrowed nouns is not depicted on the inflection of feminine nouns as well, a possible answer could be found in the fact that, as opposed to masculine nouns which inflect according to IC1 and IC2, that is, whose inflection shows variability, Pontic feminine ones belong to one single inflection class, that is, IC3.

4.2 -Human loan nouns

So far, we have seen that, following the properties of the target language, humanness triggers specific gender assignment to loan nouns of both dialects, and that the particular inflection class to which they belong may be determined by a phonological matching of the final segments between the donor and the recipient, as well as by a language-specific strategy (in Pontic) for classifying borrowed nouns distinctively from native ones.

However, according to Haugen (1950: 217) in language-contact situations, there may also be a clear tendency to assign loanwords to one particular gender, unless specific analogies intervene to draw them into another class. Haugen's premise applies to -human loan nouns in both Pontic and Heptanesian, where the neuter value seems to occupy a predominant position (see relevant examples in 14-15 and 19 below), as opposed to many native -human nouns which can bear a masculine or a feminine value as well. This runs against to the usual gender assignment to SMG -human nouns, where a gender value is triggered by their specific inflection-class feature (Ralli 2002).

With respect to Heptanesian, Romance -human loans, which fall into the neuter category, are distributed into two inflection classes, IC5 (nouns ending in -o in the citation form, as in 14) and IC6 (those ending in -i, as in 15).

(14) Neuter nouns in -o (IC5)

a. so[??]isfatsi-o    'satisfaction'    <   Ven. sodisfazion.FEM
b. kapar-o            'down payment,    <   It. caparra.FEM
                         deposit'
c. apartament-o       'apartment'       <   It. apartamento.MASC
d. ajut-o             'aid, support'    <   It. aiuto.MASC

(15) Neuter nouns in -i (IC6)

a. stratoni-o       'alley'         <   It. stradone.MASC
b. edukatsioni-o    'education'     <   It. educazione.FEM
c. tsekini-o        'old Venetian   <   Ven. zechin.MASC
                    gold coin'


In accordance with what we have seen in section 4.1, where a phonological matching of the Romance and Heptanesian final segments could dictate loan accommodation into a particular inflection class (see 10-11), -human loans of Romance origin ending in -o perfectly match native IC5 neuter nouns, which also end in -o. As such, they are incorporated into IC5 (14c,d), which triggers neuter gender in accordance with the general rule in Greek, where in -human nouns gender is elicited by inflection class. As shown above, the same loans could also be ascribed to IC1, since they match IC1 masculine native nouns in the accusative case (see +human native anorop-o 'manACC.SG' and 10a,b). However, this possibility is excluded from loan integration of human nouns because of a language-internal tendency for 'neuterization' , which overrides phonology and favors neuter gender assignment. Significantly, the same tendency renders neuter even those Romance loans which do not exhibit a form similarity with the native Heptanesian ones, as depicted by examples like (14a-b) and (15). Thus, contrary to SMG, where inflection class is responsible for assigning gender to -human nouns, in dialectal loans, inflection class is determined by a tendency for neuterization. Nevertheless, the choice of the particular type of inflection class, that is whether it would be IC5 or IC6, is rather ad hoc, since both inflection classes seem to be of almost equal productivity, at least in Heptanesian.

Phonology still plays a certain role though to the inflectional accommodation of -human loans, since certain feminine nouns ending in -a keep their Romance feminine gender, parallel to the neuter one; they are further reanalyzed as stems in order to adapt to the Greek inflectional paradigms of IC5 (neuter) or IC3 (feminine):

(16) a. burl.NEU-o / burla.FEM-o  < It. burla.FEM      'trick, joke'

     b. spitseri.NEU-o /          < It. spezieria.FEM  drugstore
          spitseria.FEM-o
     c. beladzi.NEU-o /           < It . bilancia.FEM  'weighing scale'
          beladza.FEM-o


Again, this peculiarity is justified by the fact that Romance feminine nouns ending in -a match the IC3 native Greek feminine nouns as far as the stem final segment is concerned: (164)

(17) a. Italian citta.FEM 'town'

b. Greek xora.FEM 'country'

It is important to notice now that the neuterization tendency can be diachronically confirmed, as observed by Hatzidakis (1907), who has shown that, in the medieval period, there is a shift of a number of Ancient Greek -human masculine and feminine nouns towards the neuter value (see also Browning 1969). (165) These nouns had first acquired the ending -ion, typical of neuter nouns, which, in subsequent periods, had been reduced into -in, and -i, as pointed out by Georgacas (1948: 243) and Horrocks (2010: 175-176). (166)

(18) Ancient Greek      Medieval Greek     Modern Greek
     Masculine pus      Neuter po[??]ion   Neuter po[??]i   'leg'
               Kormos          kormion            kormi     'body'
     Feminine  kefali          kefalion           kefali    'head'
               Trapeza         trapezion          trapezi   'table'


Neuter gender is also very frequent in Pontic loans, where most Turkish -human nouns ending in a consonant (21a), or in -i [i], -i [m],-u [y] (20b-e) are integrated as IC6 neuter nouns in -in:

(19) Pontic neuter nouns                                  Turkish
     a. kartali-n               'hawk'                    kartal
     b. kindi-n                 'the time of nightfall'   ikindi
     c. kamtfi-n                'lash, whip'              kamci
     d. poi-n                   'height'                  boy
     e. sjutsi-n/sigin          'bayonet'                 sungu


Note though that neuterization may not be the real reason for gender assignment to examples like (19b-e) because incorporation into IC6, prompted by phonology, and consequently neuter gender, could be triggered by the inflection-class feature, in accordance with Ralli's (2002) claim for SMG, where gender assignment to -human nouns is elicited by inflection class.

In fact, most of the endings of the original Turkish items match, or are reminiscent of, the Greek final vowel -i of native neuter nouns (the Turkish vowel harmony being absent in Pontic, all -i, -i, and -u Turkish vowels are pronounced as /i/). A phonological motivation may also hold for those Turkish items ending in a consonant (19a), since the absence of /i/ in Turkish is most probably perceived by native speakers as the result of application of the above mentioned rule of final unstressed vowel deletion (section 4.1) which characterizes many dialects. (167) Therefore, it would be reasonable to suppose that by hypercorrection, /i/ is thought to be the original ending before deletion in Pontic, and ultimately, the entire loan is further reanalyzed into a neuter stem in order to receive the appropriate Greek inflection.

Form matching may further explain why Turkish nouns ending in -a are incorporated into IC3 paradigms (compare the final vowels of the Greek feminine native noun latria 'adoration' to the Turkish genderless sevda 'love'). Like in the previous case, inflection class assists loans to be allotted the feminine value.

(20) Pontic feminine nouns                          Turkish
     a. sevda-o              'love'                 sevda
     b. zurna-o              'musical instrument'   zurna
     c. yuturma-o (168)      'lie'                  uydurma


Nevertheless, there are also Turkish loans which do not display any form similarity between the donor and the recipient, as far as the final segment is concerned, as for instance, the following noun:

(21) Pontic kuzi.NEU-n 'lamp' < Turkish kuzu

Loans of this category receive neuter gender due to the neuterization tendency and are incorporated into IC6. Moreover, contrary to examples like (19-20), where inflection class seems to dominate gender, examples which are not subject to the form-matching factor (e.g. 21) suggest an opposite dominance relation of the two features, since, for those cases, inflection class is rather determined by gender.

Interestingly, all neuter loans in Pontic are exclusively assigned to IC6, proving that this is a very productive inflection class. It is important to stress though that the other inflection classes, that is, IC5, IC7 and IC8 are also common in Pontic but they are reserved for native nouns (22), together with IC6:

(22)  a. IC5 aerop-on       'breeze'
      b. IC6 a[??]elfi-n    'brother'
      c. IC7 jel-os         'laugh'
      d. IC8 votaniazma-n   'gardening'


The fact that the dialect excludes loan noun integration from those particular paradigms further supports another language-internal property, already mentioned in 4.1, a tendency to differentiate native nouns from loans by classifying them into different inflection classes.

To sum up, Heptanesian and Pontic data confirm the close relation between the features of gender and inflection class. However, the examination of loan nouns does not lead to clear conclusions about the dominance of one feature over the other. On the one hand, when form matching is crucial into assigning inflection, gender seems to be triggered by inflection class. On the other hand, in cases where the neuterization tendency applies, gender has the dominant role. Moreover, the high frequency of--human borrowings allotted the neuter gender can be used as an indicative criterion for identifying neuter as the unmarked gender value of -human nouns, and thus, confirming the hypothesis about neuter having properties of an unmarked default value, as has been claimed by Anastasiadi-Symeonidi (1994), Dressler (1997) and Grandi (2002) for SMG.

5 Conclusions

In this paper, we have argued that the factors regulating gender and inflection-class assignment to loanwords are mainly subject to the internal properties of the Greek language, where the rich inflectional morphology is of utmost significance. Our claims invoke Ralli's (2013, 2014) assertion, according to which the language's intra-linguistic actuality primarily constrains loan word integration on the grounds that borrowed words are exposed to all necessary modifications, so that they fit the target's word pattern. Moreover, we have shown that the phonological coincidence of certain endings between the two languages in contact can also govern loan noun accommodation and indicate that the role of certain properties of the donor language cannot be neglected. Crucially, we have maintained that the analysis of dialectal data shows the existence of an old tendency of Greek to assign neuter to -human nouns and divulges a penchant for classifying loan nouns in specific inflection classes, differentiating them explicitly from native nouns. Finally, we have confirmed the often invoked close relation between gender and inflection class, and provided hints for stressing the importance of the study of language contact which may function as a test bed for theoretical proposals about system compatibility and endo-systemic tendencies.

Abbreviations

ACC      accusative
DEF      definite
FEM      feminine
Hept     Heptanesian
INDEF    indefinite
It       Italian
MASC     masculine
NEU      neuter
NOM      nominative
PL       plural
SG       singular
SMG      Standard Modern Greek
Tr       Turkish
Ven      Venitian
VOC      vocative


Angela Ralli & Marianna Gkiouleka & Vasiliki Makri, University of Patras

Angela Ralli

University of Patras

University Campus

Rio Patras 26504.

E-mail: ralli@upatras.gr

References

Aikhenvald, Alexandra. 2000. Classifiers: a typology of noun categorization devices. Oxford: Oxford University Press.

Aikhenvald, Alexandra. 2006. Grammars in contact: a cross-linguistic perspective. In Aikhenvald, Alexandra & Dixon, R. M. W. (eds.). Grammars in contact. a cross-linguistic typology. Oxford: Oxford University Press. 1-66.

Anastasiadi-Symeonidi, Anna. 1994. [TEXT NOT REPRODUCIBLE IN ASCII] [Neological borrowing in Modern Greek: direct loans from French and English. Morphophonological analysis]. Thessaloniki: AUTH.

Anastasiadi-Symeonidi, Anna & Cheila-Markopoulou, Despoina. 2003. [TEXT NOT REPRODUCIBLE IN ASCII] [Synchronic and diachronic tendencies in gender in Greek. A theoretical proposal]. In AnastasiadiSymeonidi, Anna & Ralli, Angela & Cheila Markopoulou, Despoina (eds.). To yevoc [Gender]. Athens: Patakis. 13-56.

Angelopoulos, M. 2008. [TEXT NOT REPRODUCIBLE IN ASCII] [The old Corfiot language]. Corfu: Eminerosi Publ.

Aronoff, Mark. 1994. Morphology by itself. Cambridge, MA: MIT Press.

Aronoff, Mark. 1998. Gender agreement as morphology. In Booij, Geert & Ralli, Angela & Scalise, Sergio (eds.). Proceedings of the first Mediterranean conference of morphology. Patras: University of Patras. 7-18.

Brown, Cecil. 1999. Lexical acculturation in Native American languages. New York: Oxford University Press.

Browning, Robert. 1969. Medieval and Modern Greek. London: Hutchinson University Library.

Christofidou, Anastasia. 2003. [TEXT NOT REPRODUCIBLE IN ASCII]. [Gender and inflection in Greek: a natural approach]. In Anastasiadi-Symeonidi, Anna & Ralli, Angela & Cheila Markopoulou, Despoina (eds.). To yevoc [Gender]. Athens: Patakis. 100-131.

Carstairs, Andrew. 1997. Allomorphy in inflection. London: Croom Helm.

Corbett, Greville. 1991. Gender. Cambridge: Cambridge University Press.

Corbett, Greville & Fraser, Norman. 2000. Default genders. In Unterbeck, Barbara & Rissanen, Matti & Nevalainen, Terttu & Saari, Mirja (eds.). Gender in grammar and cognition (Trends in Linguistics: Studies and Monographs, 124). Berlin: Mouton de Gruyter. 55-97.

Dahl, Osten. 2000. Animacy and the notion of semantic gender. In Unterbeck, Barbara & Rissanen, Matti & Nevalainen, Terttu & Saari, Mirja (eds.). Gender in grammar and cognition (Trends in Linguistics: Studies and Monographs, 124). Berlin: Mouton de Gruyter. 99-115.

Dawkins, Richard. 1916. Modern Greek in Asia Minor: a study of the dialects of Silli, Cappadocia and Pharasa: with grammar, texts, translations and glossary. Cambridge: Cambridge University Press.

Dressler, Wolfgang. 1997. On productivity and potentiality in inflectional morphology. In Clasnet Working Papers, 7. 3-22.

Drettas, Georges. 1997. Aspects pontiques. Paris: Association de Recherches Pluridisciplinaires. Fanciullo, Franco. 2008. Gli italianismi del neo-greco. In L 'Italia Dialettale, 69. 1-41.

Field, Fredric. 2002. Linguistic borrowing in bilingual contexts. Studies in language companion series, 62. Amsterdam: John Benjamins.

Georgacas, Demetrius. 1948. On the nominal endings -ic, -iv in later Greek. In Classical Philology, 43, 4. 243-260.

Grandi, Nicola. 2002. Morfologie in contatto. Le costruzioni valutative nelle lingue del Mediterraneo. Milano: Franco Angeli.

Hatzidakis, Georgios. 1907. [TEXT NOT REPRODUCIBLE IN ASCII] [Medieval and Modern Greek B']. Athens: Sakellarios.

Haspelmath, Martin & Tadmor, Uri. 2009. Loanwords in the world's languages: a comparative handbook. Berlin: Mouton de Gruyter.

Haugen, Einar. 1950. The analysis of linguistic borrowing. In Language, 26. 210-331.

Hock, Hans & Joseph, Brian. 1996. Language history, language change, and language relationship: an introduction to historical and comparative linguistics. Berlin & New York: Mouton de Gruyter.

Horrocks, Geoffrey. 2010. Greek: a history of the language and its speakers. Oxford: Wiley-Blackwell.

Ibrahim, Muhammad Hassan. 1973. Grammatical gender: its origin and development. The Hague: Mouton.

Janse, Mark. 2004. Animacy, definiteness, and case in Cappadocian and other Asia Minor dialects. In Journal of Greek linguistics 5. 3-26.

Janse, Mark. To appear. Cappadocian. In Tzitzilis, Christos (ed.). Modern Greek dialects.

Karatsareas, P. (2009). The loss of grammatical gender in Cappadocian Greek. Transactions of the Philological Society, 107, 2: 196-230.

Karatsareas, Petros. 2011. A study of Cappadocian Greek nominal morphology from a diachronic and dialectological perspective. Cambridge: University of Cambridge. (Doctoral dissertation.)

Kasimatis, Giannis. 1996. [TEXT NOT REPRODUCIBLE IN ASCII] [The influence of Venetian on the Heptanese islands]. Athens: Meleti Fedros.

Kilarski, Marcin. 2003. Gender assignment in Danish, Swedish and Norwegian: A comparison of the status of assignment criteria. In Nordlyd, 31, 2. 261-274.

Kollas, Ioannis. 1960. [TEXT NOT REPRODUCIBLE IN ASCII] [Corfiot: Special edition with 5,000 linguistic idioms of Corfu, with some pictures]. Corfu: Anagnostiki Eteria Kerkiras.

Kontosopoulos, Nikolaos G. 2001. [TEXT NOT REPRODUCIBLE IN ASCII] (3rd edition). [Dialects and idioms of Modern Greek]. Athens: Grigoris.

Kyranoudis, Panagiotis. 2009. [TEXT NOT REPRODUCIBLE IN ASCII] [Morphology of Turkish loanwords in the Greek language]. Thessaloniki: Instituto Neoellinikon Spudon (Idryma Manoli Triantafyllidi).

Mackridge, Peter. 1990. Some pamphlets on dead Greek dialects: R. M. Dawkins and Modern Greek dialectology. In The annual of the British School at Athens 85. 201-212.

Makri, Vasiliki & Koutsoukos, N. & Andreou, M. 2013. [TEXT NOT REPRODUCIBLE IN ASCII]. [Borrowing of nouns and gender assignment in Heptanesian varieties.] In Ralli, Angela (ed.). Patras working papers in linguistics: morphology and language contact, 3. Patras: Laboratory of Modern Greek Dialects. 58-72.

Malikouti-Drachman, Angeliki & Drachman, Gaberell. 1989. [TEXT NOT REPRODUCIBLE IN ASCII] [Stress in Greek]. In Studies in Greek Linguistics, 9. 127-143.

Matras, Yaron. 2009. Language contact: Cambridge textbooks in linguistics. Cambridge: Cambridge University Press.

Meillet, Antoine. 1921. Linguistique historique et linguistique generale. Paris: Champion.

Melissaropoulou, Dimitra. 2013. Lexical borrowing bearing witness to the notions of gender and inflection class: a case study on two contact induced systems of Greek. In Open journal of modern linguistics, 3, 4. 367-377.

Melissaropoulou, Dimitra. 2014. Reorganization of grammar in the light of the language contact factor: a case study on Grico and Cappadocian. In Janse, Mark & Joseph, Brian & Ralli, Angela & Bagriacik, Metin (eds.). Studies in modern Greek dialects and linguistic theory: proceedings of the 5th International Conference of Modern Greek Dialects and Linguistic Theory. 311-334.

Melissaropoulou, Dimitra. To appear. Loanwords integration as evidence for the realization of gender and inflection class: Greek in Asia Minor. In Ralli, Angela (ed.). Contact morphology in modern Greek dialects. Cambridge: Cambridge Scholars Publishers.

Moravcsik, Edith. 1978. Language contact. In Greenberg, Joseph & Ferguson, Charles & Moravcsik, Edith (eds.). In Universals of human language, 1. Stanford: Stanford University Press. 93-122.

Oekonomides, Demosthenes. 1958. [TEXT NOT REPRODUCIBLE IN ASCII] [Grammar of Pontic Greek]. Athens: Academy of Athens.

Papageorgiou, Leandros. 1994. [TEXT NOT REPRODUCIBLE IN ASCII] [Linguistic comparison of Epirus and Corfu]. In Modern Greek dialectology. Proceedings of the 1st panhellenic conference on Greek dialectology, 1. 326-370.

Papadopoulos, Anthimos. 1955. [TEXT NOT REPRODUCIBLE IN ASCII] [Historical grammar of Pontic Greek]. Archeion Pontou, Parartima 1. Athens: Epitropi Pontiakon Meleton.

Papadopoulos, Anthimos. 1958-1961.[TEXT NOT REPRODUCIBLE IN ASCII] [Historical dictionary of Pontic Greek]. Archeion Pontou, Parartima 3. Athens: Epitropi Pontiakon Meleton.

Pomonis-Tzaglaras, G. 2007. [TEXT NOT REPRODUCIBLE IN ASCII] [Vocabulary of idioms of the old local dialect]. Zante: Platyforos.

Poplack, Shana & Pousada, Alicia & Sankoff. David. 1982. Competing influences on gender assignment: variable process, stable outcome. In Lingua, 57. 1-28.

Ralli, Angela. 1999. Inflectional features and the morphological module hypothesis. In Working papers in theoretical and applied linguistics, 6. English Department of Aristotle University of Thessaloniki. 111-142.

Ralli, Angela. 2000. A feature-based analysis of Greek nominal inflection. In [TEXT NOT REPRODUCIBLE IN ASCII] 11-12. 201-227.

Ralli, Angela. 2002. The role of morphology in gender determination: evidence from Modern Greek. In Linguistics, 40:3. 519-551.

Ralli, Angela. 2005. [TEXT NOT REPRODUCIBLE IN ASCII] [Morphology]. Athens: Patakis

Ralli, Angela. 2012a. Verbal loanblends in Italiot and Heptanesian: a case study of contact morphology. In L'Italia Dialettale: Rivista di dialettologiaItaliana, 73. 111-132.

Ralli, Angela. 2012b. Morphology in language contact: verbal loanblend formation in Asia Minor Greek (Aivaliot). In Stolz, Thomas & Vanhove, Martine & Otsuka, Hitomi & Urdze, Aina (eds.). Morphologies in contact (Studia Typologica). Berlin: Academie Verlag. 185-201.

Ralli, Angela. 2013. Romance verbal loans in Modern Greek dialects. In Janse, Mark et al. (eds.). Proceedings of the 5th international conference on Modern Greek dialects and linguistic theory (MGDLT5). Patras: University of Patras. 430-444.

Ralli, Angela. 2014. On the Romance influence on Greek dialects. In Patrizia del Puente (ed.). Atti del terzo convegno "Parole per parlare e per parlarne" . Potenza: University of Potenza. 259-281.

Stolz, Christel. 2009. Loan word gender in Maltese, with a special focus on gender copy. In Comrie, Bernard et al. (eds.). Introducing Maltese linguistics: proceedings of the first international conference on Maltese linguistics. Amsterdam & Philadelphia: John Benjamins. 321-353.

Thomason, S. (2001). Language contact: An introduction. Edinburgh: Edinburgh University Press.

Thomason, Sarah. 2014. Contact-induced language change and typological congruence. In Besters-Dilger, Juliane & Dermarkar, Cynthia & Pfander, Stefan & Rabus, Achim (eds.). Congruence in contact-induced language change: language families, typological resemblance, and perceived similarity (linguae & litterae 27). Berlin: Mouton de Gruyter. 201-218.

Thomason Sarah & Kaufman, Terrence. 1988. Language contact, creolization and genetic linguistics. Berkeley: University of California Press.

Thornton, Anna. 2001. Some reflections on gender and inflectional class assignment in Italian. In Schaner-Wolles, Chris & Rennison, John & Neubarth, Friedrich (eds.). Naturally! Linguistic studies in honour of Wolfgang Ulrich Dressler presented on the occasion of his 60th birthday. Torino: Rosenberg & Sellier. 479-487.

Tombaides, Dimitris. 1996. H [TEXT NOT REPRODUCIBLE IN ASCII] [The Pontic dialect]. In Tombaides, Dimitris (ed.). [TEXT NOT REPRODUCIBLE IN ASCII] [Studies in Pontic dialect]. Thessaloniki: Kodikas. 222-233.

Topharas, Konstantinos. 1998 [1932]. H [TEXT NOT REPRODUCIBLE IN ASCII] [Grammar of Pontic]. Thessaloniki: Kyriakidis.

Wichmann, Soren & Wohlgemuth, Jan. 2008. Loan verbs in a typological perspective. In Stolz, Thomas & Bakker, Dik & Salas Palomo, Rosa (eds.). Aspects of language contact: new theoretical, methodological and empirical findings with special focus on Romancisation processes. Berlin: Mouton de Gruyter. 89-121.

Whitney, William D. 1881. On mixture in language. In Transactions of the American Philosophical Association, 12. 1-26.

Winford, Donald. 2005. Contact-induced changes: classification and processes. In Diachronica, 22. 373-427.

Winford, Donald. 2010. Contact and borrowing. In R. Hickey (ed.), Handbook of language contact (pp. 170-187). Chichester: Wiley-Blackwell.

Winter-Froemel, Esme. 2008. Studying loanwords and loanword integration: two criteria of conformity. Newcastle Working Papers in Linguistics, 14. 156-176.

Appendix I: Nominal inflection classes in Modern Greek

              Singular                          Plural

Class I       skilos.MASC          'dog'
Nominative    skil-os                           skil-i
Genitive      skil-u                            skil-on,
Accusative    skil-o                            skil-us,
Vocative      skil-e                            skil-i

Class II      pateras.MASC         'father',
              maoitis.MASC         'student'
Nominative    patera-s, maoiti-s                pater-es, maoit-es
Genitive      patera-o, maoiti-o                pater-on, maoit-on
Accusative    patera-o, maoiti-o                pater-es, maoit-es
Vocative      patera-o, maoiti-o                pater-es, maoit-es

Class III     mite'ra.FEM          'mother',
              tixi.FEM             'luck'
Nominative    mitera-o, tixi-o                  miter-es, tix-es
Genitive      mitera-s, tixi-s                  miter-on, tix-on
Accusative    mitera-o, tixi-o                  miter-es, tix-es
Vocative      mitera-o, tixi-o                  miter-es, tix-es

Class IV      poli.FEM             'town'
Nominative    poli-o                            pol-is
Genitive      poli-s/pole-os                    pole-on
Accusative    poli-o                            pol-is
Vocative      poli-o                            pol-is

Class V       vuno.NEU             'mountain'
Nominative    vun-o                             vun-a
Genitive      vun-u                             vun-on
Accusative    vun-o                             vun-a
Vocative      vun-o                             vun-a

Class VI      spiti.NEU            'house'
Nominative    spiti-0                           spiti-a
Genitive      spiti-u                           spiti-on
Accusative    spiti-0                           spiti-a
Vocative      spiti-0                           spiti-a

Class VII     kratos.NEU           'state'
Nominative    krat-os                           krat-i
Genitive      krat-us                           krat-on
Accusative    krat-os                           krat-i
Vocative      krat-os                           krat-i

Class VIII    soma.NEU             'body'
Nominative    soma-0                            somat-a
Genitive      somat-os                          somat-on
Accusative    soma-0                            somat-a
Vocative      soma-0                            somat-a


Appendix II: Maps of the Heptanesian (Ionian) islands and Pontus

(140) For Angela Ralli and Vasiliki Makri this paper is the product of research conducted within the project "Morphology in language-contact situations: Greek dialects in contact with Turkish and Italian", implemented under the "ARISTEIA" Action of the "OPERATIONAL PROGRAMME EDUCATION AND LIFELONG LEARNING" and co-funded by the European Social Fund (ESF) and National Resources. M. Gkiouleka's research has been conducted within the framework of the Research Funding Program THALIS, co-financed by the European Union (European Social Fund - ESF) and Greek national funds through the Operational Program "Education and Lifelong Learning" of the National Strategic Reference Framework (NSRF). This study is the result of close collaboration and discussion by the three authors. However, for academic purposes, A.R. is mainly responsible for sections 1 and 2, M.G. for section 3.2, V.M. for section 3.1, while the remaining sections are the outcome of joint work. We sincerely thank Metin Bagriacik and an anonymous reviewer whose comments made this paper benefit greatly.

(141) Pontic is an Asia Minor dialect, spoken in Pontus (Black Sea area), and Heptanesian is the dialect of the islands of the Ionian sea. See Appendix II for maps and sections 3.1 and 3.2 for more information about these dialects.

(142) Aivaliot was once spoken in western Asia Minor. In 1922, after the end of the war between Greece and Turkey (1919-1922), Aivaliots, if not killed, were forced to leave their homeland (Lausanne Treaty 1923). Today, few hundreds of speakers can be found in refugee enclaves on the Aegean island of Lesbos.

(143) Such constraint is not generally applicable to Greek native -human nouns, although one can find traces of a tendency for neuterizing -human nouns in the history of Greek (see section 4.2).

(144) Throughout the rest of the paper, we will use both SMG to refer to SMG only and Greek as an umbrella term for SMG and all Greek dialects.

(145) In this paper, examples appear in a broad phonological transcription. Glosses are given only when they are relevant to the argumentation. For details about the segmentation of SMG inflected forms into stems and inflectional endings, as well as about the features characterizing these elements, see Ralli (2000, 2005).

(146) Neuter nouns display syncretic forms in the cases nominative, accusative and vocative in both singular and plural.

(147) IC6 and IC8 differ only in the genitive case of the singular number. Similarly, IC5 and IC6 display similar endings except for the syncretic nominative, accusative and vocative cases of the singular number. In the spirit of Carstairs (1997), the paradigms of the three inflection classes may be considered to constitute parts of a macro-paradigm.

(148) For the form of the entire inflectional paradigms, the reader is referred to Appendix I.

(149) With the exception of IC4 which comprises +learned inflected forms and is absent in most dialects.

(150) Note that, in Pontic, the division into eight inflection classes is often blurred due to many cases of heteroclisis affecting the plural number and the genitive case. Moreover, for IC6 and IC8 there is a number of slightly different endings from those of SMG (see Appendix I) due to historical evolution.

(151) Cappadocian was spoken in about 32 Greek-speaking settlements in central Asia Minor before 1923, when the exchange of populations between Greece and Turkey took place. Today, there are few remaining native speakers, in certain parts of Northern Greece (in the areas of Karditsa, Volos, Kilkis, Larisa, Thessaloniki, Chalkidiki, Kavala, and Alexandroupoli), all of them descendants from Cappadocian refugees. For details about Cappadocian, see Dawkins (1916) and Janse (forthcoming).

(152) The dialect of the Ionian island of Lefkada does not belong to Heptanesian, due to geographical and historical reasons (proximity to the Greek mainland and late Venetian occupation, compared to the other islands). This dialect shows similarities with the northern dialectal varieties of Continental Greece.

(153) Nevertheless, in accordance with Aronoff (1998) and his proposal of 'covert gender', one could claim that there are inherent gender properties in Turkish which are not morphologically realized.

(154) Loan nouns are also inflected for case and number, as stated in the introduction. In this paper, these features are not accounted for as irrelevant for the discussion.

(155) Note that according to Papadopoulos (1955) and Oekonomides (1958) the final -n of neuter nouns does not appear in some areas of Pontus. For instance, it is absent in the variety of Romeyka (areas of Ophis and Tonya).

(156) For the impact of phonology in gender assignment in Italian, see Thornton (2001).

(157) For the important role of phonology into assigning gender to loans in the Asia Minor dialects, see also Melissaropoulou (forthcoming).

(158) According to the so-called 'retentionist' position, the structure of loan nouns may be fully integrated in the recipient's morphology if the latter is structurally compatible with that of the donor.

(159) Transfer of structure is not impossible though in verbs, as maintained by Ralli (2012a,b)

(160) Nouns in -es are usually adopted from Turkish (e.g. ceftes 'meat ball') or French (e.g. kafes 'coffee')

(161) The same tendency also exists in SMG and other dialects. However, there are no accurate statistics for the exact productivity of inflection classes in Modern Greek dialects.

(162) See footnote 3.

(163) In Aivaliot, unstressed /o/ is raised to /u/.

(164) As shown in (15) and Appendix I, IC3 feminine nouns have a 0 inflectional ending in the nominative singular, but an overt ending in plural and the genitive case.

(165) Note that some of the original masculine or feminine nouns still exist today with a rather lexicalized meaning (e.g. trapeza 'bank').

(166) As exhibited in (19) below, Pontic still keeps the medieval form of these neuter nouns, since they end in -in.

(167) It exists in the Asia Minor Aivaliot, Cappadocian and Pontic, as well as in the Northern Greek dialects. However, in the Northern dialects and Aivaliot the rule generalizes deletion, applying to all unstressed /i/s and /u/s, independently of the position they hold in the word.

(168) In Pontic, there was a metathesis of the initial [u] and [j].
COPYRIGHT 2015 The Slovak Association for the Study of English
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2015 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Ralli, Angela; Gkiouleka, Marianna; Makri, Vasiliki
Publication:SKASE Journal of Theoretical Linguistics
Article Type:Report
Date:Sep 1, 2015
Words:9829
Previous Article:Jakobson's zero and the pleasure and pitfalls of structural beauty.
Next Article:The meaning of diminutives: a first-order perspective.
Topics:

Terms of use | Privacy policy | Copyright © 2021 Farlex, Inc. | Feedback | For webmasters |