CC BY
45
Review
Grammaticalization in Slavic Languages: from Areal and Typological Perspectives
Edited by Motoki Nomachi, Sapporo: Slavic Research Center, Hokaido University,
2010, 132 p.
[the 2nd edition is available online at the web page of Slavic Research Center,
Hokkaido University:
http://src-h.slav.hokudai.ac.jp/coe21 / publish/no23_ses/contents.html]
Milivoj Alanović
Department of Serbian Language and Linguistics Faculty of Philosophy University of Novi Sad
0. This volume of articles represents the result of collaborative research that was centered upon two studies: “The Functional AppZroach to the Typology of Slavic Languages: Research in the Semantic Category of Possessiv-ity” (2008-2009), carried out under the auspices of a Japan Society for the Promotion of Science (JSPS) Grant-in-Aid for Research Activity Start-up, and “A Comprehensive Study of the Kashubian Syntax” (2009-2012), supported by a JSPS Grant-in-Aid for Young Scientists (B). The volume contains at least two series of papers: four chapters and three additional texts, and one essay and two review articles.
1. The opening chapter (“Linguistic Area and Grammaticalization Theory”, p. 1-26) contains the paper Is Europe a Linguistic Area? written by Bernd Heine and Motoki Nomachi. The authors discuss a single central
Slovene 2012 №1
Milivoj Alanović
I 265
question: Is it really possible to argue that there is a European linguistic area, and, if so, does it have clearly defined boundaries? In order to address these questions, the autors, took two grammatical categories or phenomena as their points of focus: the rise of auxiliary constructions and possessive perfects.
The first example concerns the structure of the verb phrase in European languages; more specifically, the example presents a case in which a specific word behaves like a lexical verb, on the one hand, and like a functional category expressing distinctions of tense, aspect, modality, etc., on the other.
1.1. The initial part of this chapter is concerned with such a case of “doublets”, dealing with a set of four constructions associated with verbs for “threaten” in European languages. The example of the Portuguese verb amea-gar “to threaten” illustrates these constructions. The fact that these constructions are found in different language families in Europe, including families that are genetically unrelated (e.g., Indo-European and Finno-Ugric), leads to the conclusion that the presence of these constructions across Europe must be the result of language contacts.
According to the authors, the chronology suggests that the grammati-calization of “threaten”-constructions must have originated in French, since it is attested there first, and was subsequently replicated in other languages of Western Europe, where it is attested several hundred centuries later. The diffusion of this grammaticalization process in Central and Eastern Europe appears to be a more recent development; it is weakest in Eastern Europe.
The example of “threaten”-constructions confirms that contact-induced grammatical replication is clearly structured, proceeding unidirectionally from less grammatical to more grammatical structures, in the present case from lexical verb to auxiliary, and that contact-induced grammatical change has both a language-internal and an external component.
1.2. The second part of the chapter is confined to some morphosyntactic properties of the possessive perfects that are found essentially only in Europe. Possessive perfects (“have”-perfects) can be considered to be a paradigm areal property of European languages. On typological grounds, therefore, the rise of these constructions must have been due to historical factors. The possessive perfect of modern European languages, as shown in the article, has its roots in early Latin. That possessive perfects spread via replication from Romance languages to Germanic is a uncontroversial hypothesis.
In the Romance and Germanic languages, predicative possession is built on the action schema [X has Y], relying on a more-or-less transitive “have”-verb. In some other languages, different conceptual schemas have been employed: in the Celtic languages it was the goal schema [Y is to X] and in North Russian and Estonian it was the location schema [Y is at X].
2012 №1 Slovene
266 I
Grammaticalization in Slavic Languages: from Areal and Typological Perspectives
Finally, the authors conclude that all those “peripheral” languages that have had a history of intense contact with Germanic or Romance languages have created a more advanced possessive perfect, such as Breton with French, Estonian with German, and North Russian presumably with Scandinavian languages. But, with the exception of the southwestern dialects of Macedonian, none of these languages has developed a stage-3 perfect, because the replicated categories tend to be less grammaticalized than the categories that provided the model.
At the very end of this first chapter there is an important conclusion: the replication of possessive perfects followed the same sequence of stages as observed in the case of “threaten”-auxiliation, allowing for implicational predications of the form. Thus, if a language has reached stage X then it has also reached all preceding stages. That fact suggests that diachronically the sequence of grammaticalization was stage 0 > 1 > 2 > 3.
2. The second chapter, “Language Contact and Grammaticalization” (p. 27-48), contains one paper, Grammaticalization and Language Contact between German and Slovene, written by Alja Lipavic-Oštir, which investigates the influence of German on the Slovene language with respect to the changes in grammar caused by language contact.
This article deals with some examples of grammaticalization in Slovene that have appeared under the influence of German. The results of this research are presented in twelve sections: Gender Leveling, Analytic Perfect, Possessive Perfect, Analytic Genitive, Modal Passive, Subject pronouns, Article, Onikanje, Combinations of Numerals, Verbal Prefixes, Za as an Infinitive Marker, and Recipient Passive. As we can see, the phenomena of the processes of grammati-calization in Slovene under the influence of German differ among themselves according to whether completely new categories are grammaticalized (for example, articles and the modal passive), or whether grammaticalization has influenced new forms for expressing existing categories.
2.1 Gender leveling is a pattern by which human female participants in speech acts are referenced with masculine morphology in adjectives and participles. This means that the case morphology that marks masculine gender adopts the function of marking feminine gender, which is a kind of generalization. Endings for the masculine gender adopt the role of expressing the natural feminine gender, for example, Dvanest let sem bil (MASC) star (MASC)... (“I was twelve years old.” [female speaker born in 1919]).
2.2. For expressing past actions, Slovene dialects as well as standard Slovene use the analytic form of the verb biti [to be] + active resultative l-partici-ple. The structure is stable, so the auxiliary biti is not dropped, and neither are the morphological suffixes of the participle, which differ with regard to gender and number.
Slovene 2012 №1
Milivoj Alanović
I 267
2.3. Considering the example Imam zgrajeno stanovanje iz Ytonga (“I have a flat built of Ytong [concrete brick]”), the author shows that possession is no longer the primary meaning of these constructions — both cases are about expressing possession (“to have a flat”), but the meaning in the foreground is also that I built the flat and I have it.
2.4. The prepositional phrase with the German von (das Haus von meinem Vater [the house of my father]) and the Slovene od (hiša od mojega očeta [the house of my father]) can replace the synthetic genitive in both languages. The introduction into common usage can be explained through a process of grammaticalization where each grammatical sign obtains a new grammatical function, which can be described as expressing possessivity. Both processes of grammaticalization are compared to the synthetic genitive in both languages and the possessive adjective in Slovene.
2.5. The modal passive, in German consisting of the modal wollen [want] and the auxiliary haben [have] in combination with the main verb in the past perfect participle and expressing the grammatical meaning of “want to get something done”, is also partly grammaticalized in Slovene, and because of that the following dialectal usage is possible: Lase je hotela imeti pobarvane (“She wanted to have her hair dyed”).
2.6. In Slovene dialects, particularly those along the border, the use of a personal pronoun appears in places where there should be a zero pronoun, since this is a pro-drop language: ...jaz (PERS.PRON., 1. PERS.SG) sem se z Liksu spoznala... (“... I met Liksa...”)
2.7. The noun does not express the categories of definiteness or indefiniteness; however, these categories are morphologically marked in adjectives (velik avto [a big car] versus veliki avto [the big car]). The noun has no article, but such cases can be found in Slovene.
2.7.1. The definite article appears in the dialectal forms to (N. SG.), ti (M. SG.), ta (F. SG.), and others (ta mlajša hčera [the younger daughter]); these appear in most dialects. The definite article is proclitic and unstressed, and as in standard German, it may appear with superlative forms of adjectives (najtavečji [the biggest]).
2.7.2. Slovenian dialects (from Carinthia, Haloze, and Slovenske Gorice in Styria) also use an indefinite article formed on the basis of the number “one,” and it is stressed: Ona je bila en tak lepi otrok “She was such a beautiful child”.
2.8. Onikanje is the use of the third-person plural pronoun and/or corresponding verbal forms in polite address. It is a phenomenon that clearly made its way into Slovene through German influence but gradually eroded during
2012 №1 Slovene
268 I
Grammaticalization in Slavic Languages: from Areal and Typological Perspectives
the twentieth century. It is less common to encounter individuals using oni-kanje today, but it has not completely died out.
2.9. In Slovene, numerals from twenty-one onwards are formed in the ones-tens order (petindvajset [five-and-twenty]). German also uses this method of formation (fünfundzwanzig [five-and-twenty]). The structure or formation of numerals was, obviously, received from German, but it has not fully gram-maticalized, since grammar books from the mid-twentieth century still allow both types of formation.
2.10. In Slovene, the category of verbal aspect exists, and for the formation of aspectual contrasts they are often word changes (počim [I crack (once)] — pokam [I am cracking]), or various prefixes used. The prefixes for forming aspectual contrasts often resemble German verbal prefixes: Ger. aus-/Slov. iz- [out], Ger. ausarbeiten = Slov. izdelati [to elaborate, draw up], Ger. hinein-/ Slov. v- [in], Ger. sich hineinmischen = Slov. vmešavatise [to get involved], etc. These prefixes are not necessarily translations of German prefixes but appeared by chance, since they are based on semantically equivalent metaphors.
2.11. The German zu is a particle that is sometimes used to mark the infinitive, for example, Bücher zu schreiben [books to write]. Standard varieties of Slavic languages do not use such a device, but it is encountered frequently and not only in colloquial Slovene: Kje okoli Postojne je kaj dobrega za pojesti? (“Where around Postojna is there something good to (PREP) eat (INF)?”).
2.12. The recipient passive structure is a replica structure from German Die Bücher habe ich geschenkt bekommen “The books were given to me as a gift”). The recipient passive in Slovene has the structure: verb dobiti [to get] + participle -n with endings for gender and number. Examples can be found in different varieties of Slovene: Prakso sem dobilplačano (colloquial Slovene) “I got paid for my practical work”).
Finally, the author concludes that in these described processes, language signs or combinations become grammatical signs, or existing grammatical signs, which acquire new grammatical roles. In these processes, Slovene proves to be a language that was not only strongly under the influence of German, but also that of Italian (abandoning the aorist and the imperfect), while the influence of Hungarian as well as the possible influence of Croatian are extremely locally restricted.
3. The third chapter (Nominal Morphology, p. 49-79), which includes the article Grammaticalization of the Masculine and Non-masculine Personal Category in the Polish Language, written by Alina K^pinska, discusses the development of the category of gender in Polish, centred on the masculine and non-masculine categories of meaning. This chapter discusses these categories empirically from the perspective of the grammaticalization of gender.
Slovene 2012 №1
Milivoj Alanović
I 269
3.1. In the first part of this article, the author claims that in the nominative plural the category of masculine personal gender is determined clearly against other ending changes of this case, which are caused by equalization within its isofunctional endings. They are based, according to the author, on the fact that the initial neuter ending -a became obsolete, that the older masculine ending was limited only to masculine personal forms, and that endings that were becoming obsolete were substituted with endings that were originally feminine. The domination of the initially feminine ending -e in all forms with the exception of masculine personal forms distinguishes the whole group as a non-masculine personal one. Therefore, this is how the category of masculine personal and the category of non-masculine personal gender (which is distinct in the plural) came into being.
3.2. In the second part of this article, we learn that the grammaticaliza-tion appears in noun inflection and occurs not only in the masculine personal category but also in the masculine animate category (that is, in a small portion of the categoriy). Due to grammaticalization, one masculine gender is divided into three: the masculine personal, the masculine animate, and the masculine inanimate.
Changes of inflectional endings are primarily conditioned by syntax. Those that took place in the nominative plural are the same alterations in isofunctional endings as in the remaining cases of the plural, and they are not comprehensive but limited and partial and do not encompass only one semantic group. The process of equalization of endings of the same grammatical case is analogous to changes concerning the remaining endings of the plural, such as the dative, the instrumental, and the locative. This process resulted in the demorphologization of gender in the above-mentioned grammatical cases. In present times, there is only one ending, that is, dat. -om, loc. -ach, and instr. -ami and recessive -mi. Equalization of singular accusative and genitive endings is limited to only a group of masculine gender nouns; there is no such equalization in the neuter or feminine gender.
The plural nominative forms of feminine and masculine non-personal forms became similar as a result of two processes. The first is the acquisition by masculine non-personal nouns ending in hard vowels of the ending -y // -i (after the vowels k, g) in the place of the older ending -i // -y (after hardened consonants) or of -owie, which is confirmed by feminine forms such as zony, corki, gwiazdy [wives, daughters, stars] and masculine forms such as koty, domy, obioki [cats, houses, clouds], formerly koci, domowie, obiocy. The second process is the extension of the scope of usage of the ending -e, typical of masculine and feminine nouns ending in a soft vowel having a hardened consonant at the back, to all masculine non-personal nouns whose endings are as described above, that is, those that temporarily might have had the ending -owie (oitarzowie [altars],
2012 №1 Slovene
270 I
Grammaticalization in Slavic Languages: from Areal and Typological Perspectives
wroblowie [sparrows], etc). The ending -owieis still rare in the inflection of masculine personal nouns ending in the above-mentioned way, compared to the more common ending -e, for example, krolowie, stryjowie, mqiowie, but cesarze, tiumacze, kowale, ziodzieje [kings, uncles, husbands, but emperors, translators, smiths, thieves]; in the inflection of some nouns, there are two correct forms: the older one with the ending -owie and the newer one with the ending -e, for example, samurajowie and samuraje [samurai].
3.3. In the third part of the article, Gender Classification of Polish Nouns, the author discusses gender classification, which is a result of a certain categorization within the lexical system formed by the whole set of Polish nouns. According to this categorization, each noun has a selective grammatical gender that is ascribed to a particular semantic and morphological group, and its position in this and not in any other group is shown by a proper paradigm connected with the generalised meaning. This meaning is usually determined only on the basis of the most general, minimalistic semantic references and on the basis of negation of the marked element, which in the case of Polish declension is the masculine personal and, broadly, the masculine gender. Each grammatical gender has its own inflectional paradigms, which means that grammaticalization took place in the Polish language.
The author concludes that the gender of the noun is expressed by certain grammatical means. These means — both morphological and syntactic — are the exponents of the category of gender. They include: 1) morphological characteristics, that is, the gender pattern of inflection (the whole paradigm), which even in the sing. nom. co-identifies (along with the general meaning) the appropriate noun gender; 2) congruence as a type of syntactic relation determining mutual agreement of noun forms and parts of speech whose gender is an inflectional category in the narrow sense, that is, those that are inflected by gender; and 3) syntactic rules of collocation with some types of cardinal or group numerals.
According to the author, the term grammaticalization may also refer to the gender classification of Polish nouns, and grammaticalization itself may be treated as the formation of an inflectional paradigm specific for each gender. Currently, the only gender that, due to inclusion of names of male persons, has positive references to extra-linguistic reality is the masculine personal gender. Due to grammaticalization, it has its own, specific inflectional exponents at the same time. The masculine animate gender emerges to a lesser extent (because the borders are fuzzy today) on the basis of positive semantic and inflectional criteria. The third and last masculine gender is distinguished as an element unmarked in comparison with the first two. The other two genders — the feminine and the neuter that have their own paradigms — are distinguished as unmarked, non-masculine elements in opposition to the three masculine genders treated jointly.
Slovene 2012 №1
Milivoj Alanović
I 271
4. The fourth chapter (Verbal Morphology, p. 81-104), written by Olga Mišeska-Tomić, contains one paper, The Macedonian “Have” and “Be” Perfects, which describes and analyses the use of the perfect form inherited from Common Slavic (sum “to be” + /-participle), and the two new perfect forms of non-Slavic origin (sum + n/t-participle and ima “to have” + n/t-par-ticiple) in the Macedonian language.
All of these forms are sanctioned by the norm, but unevenly spread. The perfects with “be” auxiliaries plus /-participles are the only perfects used in the eastern and northwestern Macedonian dialects; the “have” perfects are intensively used in the southern and western parts of the territory where Macedonian is spoken and have spread to the northeast with increasingly diminishing strength; while the “be” perfects with passive participles are intensively spoken in the southwest.
As the present article shows, Macedonian actually has two systems of perfects, to which the author refer as the “A system” and the “B system.”
4.1. In the A system, forms of the “be” auxiliary combine with /-participles inflecting for gender and number. Constructions of this system have kept the original function of the perfect as an expression of the result of actions, but are also used as exponents of evidentiality — a modal category that expresses the subjective relationship of the participants in the speech event to the narrated event.
There are two types of “be” perfects with /-participles in Macedonian — the present “be” perfect and the past “be” perfect, both of which have /-partici-ples of transitive and intransitive verbs, which inflect for gender and number, but unlike the latter do not occur in an attributive position and are used exclusively with “be” auxiliaries.
4.1.1. The present “be” perfect, presented in the second part of the article, Forms and Functions of “Be” Perfects with /-particip/es, either (a) represents an event which had taken place at an indefinite point of time in the past and has relevance in the present, or (b) expresses evidentiality. When it refers to the third person, singular or plural, it features only /-participles, whereas when it refers to the first and second person, singular and plural, the /-participles are accompanied by the “be” auxiliary clitics sum [be.1Sg], si [be.2Sg], sme [be.1Pl] or ste [be.2P] ’ (1st sum čita//čita/a sme čita/e; 2nd si čita//čita/a ste čita/e; 3rd čita//čita/a/čita/o čita/e).
4.1.2. The past “be” perfect, in which the /-participles co-occur with imperfect forms of the “be” auxiliary - bev [1Sg], beše [2/3Sg], bevme [1Pl], bevte [2Pl], bea [3Pl] — basically refers to a past event that had taken place before another event in the past (1st bev pročita//pročita/a bevme pročita/e; 2nd beše pročita//pročita/a bevte pročita/e; 3rd beše pročita//pročita/a/pročita/o bea pročita/e).
Currently the past “be” perfect is most frequently used to denote a past event that had taken place before another event in the past.
2012 №1 Slovene
272 I
Grammaticalization in Slavic Languages: from Areal and Typological Perspectives
4.2. The B system, presented in third and fourth part of this article, Forms and Functions of the “Have” Perfects and Forms and Functions of the “Be” Perfect with Passive Participles, uses two types of constructions — one with forms of the “have” auxiliary plus invariable past participles of transitive or intransitive verbs, and another with forms of the “be” auxiliary, plus inflecting passive participles of verbs that, as a rule, are intransitive. Constructions with “have” auxiliaries plus participles of transitive verbs are used as exponents of the experiential perfect, or as exponents of the perfect of result, while the constructions with “be” auxiliaries plus inflecting passive participles are, as a rule, used as exponents of the perfect of the result of intransitive verbs.
4.2.1. The two sets of “be” perfects have corresponding sets of “have” perfects — present “have” perfects and past “have” perfects. Both sets have invariant participles ending on the suffixes -no or -to, formally corresponding to the Old Church Slavonic neuter form of the past passive participle.
4.2.2. To express the result of the action of an intransitive verb, Macedonian employs the present and past tense forms of the “be” auxiliary, plus inflecting passive participles marked by -n/-t for M. Sg, -na/-ta for F. Sg, -no/-to for N. Sg, or -ni/-ti for all persons plural. These constructions are typically used in passive clauses.
4.3. In the fifth part of the article, Origin and Spread of the Macedonian Perfects, the author discusses the origin and spread of these categories.
While the Macedonian “be” perfects with l-participles were inherited from Proto-Slavic, the “be” perfects with passive participles and the “have” perfects developed at the time when the Balkans were part of the Ottoman Empire (second half of the fourteenth century to the beginning of the twentieth century).
4.3.1. The “be” perfects with passive participles and the “have” perfects developed as a result of language contact with non-Slavic languages spoken in the Balkan territories adjacent to the territories where Macedonian was spoken: Greek, Albanian, Aromanian, and Megleno-Romanian in the case of the “have” perfects and Aromanian and Megleno-Romanian in the case of the “be” perfect with passive participles.
The author underlines that Macedonian is the only Slavic standard language that has advanced to the final stage of grammaticalization of the “have” perfect. The patterns of the latter two perfects were equivalent to corresponding patterns of the non-Slavic model languages, but used auxiliaries that had already been grammaticalized and participles that had existed in Old Church Slavonic.
5. In addition to the above articles, this volume also includes three other texts.
Slovene 2012 №1
Milivoj Alanović
I 273
5.1. The first of them, О доприносу српског префикса од- семан-тичком и граматичком лику глаголских лексема (p. 105-106), written in Serbian, is an essay by Professor Milka Ivić. This essay relates the verbal prefix “od” to the grammaticalization of verbal aspect in Serbian. In this short but very inspiring article, the author underlines that this prefix means: (a) the end of an action (одседети); (b) the cancelation of an earlier state of facts (одвезати); and (c) a new psychological state of facts (одбацити предлог).
5.2. The two remaining texts are a review article and a book review.
5.2.1. The first is a review (p. 107-124) of Bernd Heine and Tania Kuteva’s book The Changing Languages of Europe (Oxford, 2006) by Paul Wexler (Tel-Aviv). Wexler’s review is a beneficial addition and gives a critical overview of the book in which he stresses Relexification theory in the context of gram-maticalization.
5.2.2. The second (p. 125-132) is a review of Zuzanna Topolinka’s comparative Polish and Macedonian grammars Полски~Македонски: Граматичка конфронтација 8. Развиток на граматичките категорији (Skopje, 2008) by Angelina Pančevska (Skopje). As we can see, this work, which deals with the development of grammatical categories, is of great interest because it covers many of the themes dealt with in this collection.
6. The present collection of articles offers a credible series of scientific investigations and valuable interpretations on the dynamics of changes in Slavic languages in their full complexity. These highly interesting articles have shown us some new research perspectives and, at the same time, they will be an important and indispensable manual for all further scientific investigations of Slavic languages based on the field of the theory of grammaticalization.
Dr. Milivoj Alanović
Odsek za srpski jezik i lingvistiku
Filozofski fakultet
Univerzitet u Novom Sadu
Dr Zorana Đinđića 2
21000 Novi Sad
Serbia
