2.4 Inquadramento della ricerca proposta (in ambito nazionale ed internazionale) / National - international framing of the research program
The "Tree model" of linguistic change has been always predominant in the Indo-European linguistics from its initial formulation by A. Schleicher (1862) up to the present days. However, as J. Schmidt (1872) pointed out, the Tree model does not provide any kind of information about the isoglosses shared by languages that are not headed by an immediate protolanguage. These innovative isoglosses, which we suggest to term "branch-crossing" isoglosses, can have arisen from linguistic contacts between immediately unrelated IE languages and fall completely out of the Schleicherian approach. Schmidt's substantiated his alternative "Wave theory" with a first list of such isoglosses. Later on, other linguists have sporadically annotated such phenomena. However, a systematic classification and analysis of these isoglosses have never been made. While the Tree model constituted the framework for a number of larger systematic surveys of the Indo-European family (from Delbrück & Brugmann's Grundriß onwards), the Wave model lacks a comprehensive compendium of data.
The ultimate goal of the present research program should therefore consist in producing a survey that would contain all the data available, their detailed typological classification and historical analysis. Such survey must be provided with the geographical representation of the collected isoglosses, so that they could be captured more easily.
A METHODOLOGICAL PREMISE
One of the foremost tasks of this initial stage of the research program must be the comprehension of the different sources of isogloss and the creation of a heuristic procedure capable of distinguishing them. The question is not totally new, and is based on a decades-long debate concerning the model of the linguistic change (for instance, Tree model vs. Wave model, see Goebl 1983, Hüffler 1955) and the question of the dialectological approach to the IE languages suggested by Meillet (1922, 2nd ed.); see also Lazzeroni (1987) for a survey of the question.
The scholarly literature on this topic shows that innovative isoglosses shared by a group of languages con have different sources at their origin. In fact, not all types of isoglosses are equally informative and valuable for the sake of the present research. There are at least four types of possible sources:
Some general tendencies are, indeed, always present in the diachronic change of languages due to the principle of economy of the muscular energy and mental effort in the production of speech (see Martinet 1955). Instances of such development can arise spontaneously in unrelated languages. In terms of markedness degree it can be observed that unmarked outcomes are more widespread than the opposite (Nichols 1992, cf. already Meillet 1922, discussed in Lazzeroni 1987). E.g., the spirantization of /g/ in Ukrainian and Dutch, as well as the loss of word final /g/ in Middle Persian and Middle English is only a coincidence due to a natural weakening tendency.
Same substrat influence
Common (or typologically similar) substrate languages can lead to a shared innovation in a group of genetically unrelated languages. Thus, the increasing number of cases registered in Tocharian, but also in Middle and New Indo-Aryan languages (as opposed to the Iranian branch of the Indo-Iranian group), can be ascribed to typologically similar substrate varieties; an increase in the case-forms number registered in historical Baltic and Slavic languages is also correlated with a common (Ugro-Finnic?) substrate (see Kulikov 2011).
The drifts are processes that have a common genetic background but keep developing in a parallel way in two (or more) languages also after their genetic separation. The classic example of drift is the Umlaut phenomenon in German and English (with a very similar parallel evolution in French, see Belardi 1979), and especially its development into a morphological feature, namely the marker of the plural (which is the original example of the drift phenomenon as proposed by Sapir 1921).
These type of isoglosses is we are mostly focusing on here. A famous example, studied by Gusmani (1972), is the loss of /s/ (deletion or transformation into a /h/-like sound) in different positions in a series of languages — Greek, Indo-Iranian, Armenian and partly Slavic, Celtic and Anatolian languages — connected geographically but unrelated genetically (at least, not immediately related way).
The main criterion for the differentiation between natural outcomes, drifts, contact phenomena and genetic relationships must be the detection of the divergent vs. convergent nature of the isogloss (see Nichols 1992).
DATA KNOWN FROM THE EXISTING STUDIES
In the present research a contact-based approach to the Indo-European languages is not uncommon. Indeed, many branch-crossing possibly contact-induced isoglosses are widely known nowadays, although no general survey thereof can be found. Here some of them are listed in order to exemplify the data we are looking for (see also Gusmani 1972 for some other examples).
– loss of the qualitative distinction between /a/ and /o/, everywhere except for the Hellenic and Italic groups
– RUKI-law in Indo-Iranian, Balto-Slavic (Lubotsky 1999, Beguš 2012) and perhaps in Germanic (see Prescott 2012).
– loss of the aspirated stops, everywhere except for Sanskrit, Classical Greek and Armenian
– creation of a back spirant /h/ (or the like), from different sources
– palatalization of velar stops, as well as of other consonants
– creation of a new adjectival declension (Balto-Slavic, Germanic)
– adding of pronominal elements to nominal stems (Prakrits, Balto-Slavic)
– increasing productivity of the derivational velar suffix (coming from IE *-ko) in denominal adjectives and derived nouns, (Middle Iranian, Slavic, Latin, Greek, Germanic; importantly lacking in the more ancient Hittite, see Ciancaglini 2012)
– creation of "augmented" past forms of the verb, unknown to the IE, but attested in many singular IE languages
– creation of the infinitive form of the verb, which cannot be reconstructed at the proto-IE stage
- reduction of the nominalization morphology variability
– evolution towards a rigid word order (from the almost completely free word order supposed for Proto-IE)
– evolution towards the generalization of the transitive construction, with a strong subjecthood (see Bauer 2000, Kulikov 2012, Comrie 2006)
– evolution from the existential possessive construction (such as Latin mihi est liber 'I have a book') towards the lexicalized transitive possessive predicate (such as English verb to have), correlated with the rigid word order (see Keidan 2008; cf. Baldi & Cuzzolin 2010)
– development of a separate, morphologically marked, lexical class for primary adjectives (see Alfieri 2009, 2011)
A great number of long-distance lexical isoglosses within the IE languages have been known for many years (having started from such important data collections as Schmidt 1872). A correspondence between Sanskrit namas 'homage, offering' and Novgorodian Old Russian namъ 'loan' could be mentioned among the latest discoveries in this respect (see Patri 2001, Keidan 2009).
The ground ideas of the present research have been initially exposed by the principle investigator at an international symposium of Indo-European studies in Saint Petersburg (see Keidan 2013).
Some methodological questions and preliminary results of the research are at the focus of an international workshop organized by the principle investigator in 2014, see http://asiatica.wikispaces.com/CBC5
2.5 Sintesi del programma di ricerca e descrizione dei compiti dei singoli partecipanti / Synthesis of the research program and description of the duties of each participant
SYNTHESIS OF THE RESEARCH PROGRAM
Direction of analysis
Before entering the present research program each component of the research team has been working on specific sub-fields of the Indo-European studies having often reached some interesting results; such results would acquire a new significance if considered in the light of the branch-crossing hypothesis that we are focusing on here. Usually, when an unexpected isogloss is observed within the IE languages in isolation, the scholar does not have at his/her disposal any theoretic framework that would allow to treat it in a general way. Such cases remain — so to say — isolate curiosities; see, e.g., the Avestan/Old Russian lexico-syntactic isogloss, which, besides being very interesting, did not receive any general explanation by its discoverer (Patri 2003).
In our approach, on the other hand, what appears as an isolated fact if viewed per se, could be re-interpreted in a more informative way, after it has been contextualized within the general framework of the "second-generation" IE linguistic area, which constitutes the main hypothesis of our research program. Thus, for example, the development from /s/ to /h/ (or the like), occurred in many "second generation" IE languages, must be seen as a contact-based phenomenon, which makes it a meaningful fact, rather than a fortuitous coincidence (cf. Gusmani 1972).
This, in a certain way, goes in the opposite direction with respect to the traditional approach of the IE studies. The traditional comparative method is directed from the attested linguistic data "backwards", i.e. to the reconstructed, essentially unattested, common stage. Our approach, on the other hand, is headed from the attested linguistic data "onwards", i.e. from the proto-language towards later linguistic stages, when some instances of convergent development could have occurred.
Our approach continues the tradition of the study of branch-crossing phenomena started already by Schmidt's "Wave theory" (1872), Schuhardt's (1884) study on the common features of Slavic, Germanic and Romance, by Trubeckoj and especially by Jakobson (1931), whose notion of "Eurasian Sprachbund" partly inspired our notion of "second-generation IE standard".
The "second-generation IE standard" will also serve as a kind of statistical reference point: the accumulation of shared feature will be considered a significative fact that must drive the analysis, rather than being its casual sub-product. With this in mind not only we can re-consider those isoglosses between particular languages that have been observed but never explained so far, but we are also provided with a heuristic tool that allows us to find out new isoglosses never captured by the preceding scholarship.
A very good recently suggested example of this kind of approach is the unification of the RUKI-law with some phonetic changes in Germanic by Prescott (2012): only overturning the scholars' viewpoint some apparently unrelated phenomena (indeed, never even compared to each other) have been generalized as parts of one and the same common drift or convergence.
In the classical Comparative method, once the mail phonological correspondences have been established between semantically identical cognates and the phonology of the proto-language hypothesized, they can serve as a heuristic for suggesting some semantically less obvious etymologies. Similarly, we hope that our approach will also reach the same quantity-to-quality turn.
Our notion of "second-generation IE standard" must be thought of as an instance of a linguistic area or Sprachbund (see Thomason 2001). In order to define better the distinctive features of this areal standard we must classify and analyze the isoglosses under consideration. We must rule out those that are purely genetic and those too casual and natural to be really relevant.
If two languages can be proven to have developed a common feature that cannot be explained genetically, then it must be considered a true contact-based branch-crossing innovation; otherwise, if diachronically we observe a divergent development of the shared feature, the genetic origin is the default explanation (see Nichols 1992).
Thus, this specific kind of Sprachbund should be constituted of common drifts, common substrate influences and especially of the isoglosses derived from the contact phenomena. Among the latter, the systematic ones, especially the morphological and phonological ones, will be privileged over the casual phenomena or phenomena that do not affect the core structure (such as lexical borrowings).
The "second-generation IE standard" is, therefore, bound much more on typology than on chronology: it is supposed to include the "middle" phases of the IE languages (such as Middle Indian or Middle Iranian), but also the oldest attested forms of such languages as Slavic or Baltic.
The common innovations are to be observed regardless the geographical vicinity of the languages. What we are trying to develop is a topological representation of isoglosses, while their geographical interpretation must be seen as a consequence.
TASK OF THE TEAM MEMBERS
A common task for the research team will consist in collecting and classifying of the primary linguistic data: branch-cross isoglosses as they have been published in the preceding scholarly literature, and also newly discovered (or re-considered) facts. Each member, however, will focus the analysis on his/her specific fields of competence, but will apply the common new approach to them.
Artemij Keidan will coordinate the team work and will be especially focusing on the methodological and theoretical problems, such as:
– the classification of the isoglosses
– the definition of the essential features of the "second generation IE standard"
– the preparation of the database that will contain the collected data
Furthermore, he will analyze the material coming from such IE languages as: Prakrits and Late Sanskrit (including the indirect information provided by the indigenous grammatical tradition), Early Germanic languages (especially Gothic), Slavic languages (Especially Slavonic and Old Russian). He will be concerned mostly with phonological and morphological isoglosses. For this reason, he will examine under a statistical point of view the available data on the phonological systems of the IE languages in order to highlight some common features never observed before. For morphology, he will study the birth of new adjectival declination endings that show a strong parallelism in various genetically unlinked "second generation" IE languages.
Claudia A. Ciancaglini has devoted her research activity overs last few years to the areas of Iranian (Ancient and Middle), Greek and Tocharian. Within the present research program she will concerned with the history of some innovation in the derivational morphology that are attested in many branches of the "second generation" IE languages but are extremely rare in the oldest IE languages, so that many scholars do not believe they did really belong to the Proto-IE.
One such feature is, e.g., the derivational suffix in velar (IE *-ko-). This suffix, exhibiting many different and hardly clusterable functions (the most important of which seems to be the adjectival derivation), is very well attested in Classical Sanskrit, in the Middle Iranian varieties, in Slavic and Germanic languages, and also in Latin and Greek, etc. However, it is almost completely lacking in the oldest phases of the IE family (two examples in Gathic Avestan and only one in Hittite; very few in Old Persian). Interestingly enough, when present, the forms in *-ko- (with its various outcomes) are typically bound to the "low" sociolinguistic variety of the language. Thus, in Old Persian this derivational suffix is only used for the so-called "non-Aryan" ethnonyms (i.e., names of nomadic populations of different origin that inhabited the Central Asian steppes), while in Young Avestan they are attested in words that denote diseases, demons, sins and "Ahrimanic" (i.e. evil) creatures.
Moreover, the way it becomes more and more grammaticalized and morphonologically reanalyzed is strikingly similar in the later IE languages (see Ciancaglini 2012). In many IE languages this suffix agglutinates to the preceding thematic vowel, forming new outcomes according to the vowel quality (such as -ik-, -uk-). All this makes it a very good example of a common branch-crossing feature of the "second generation" IE standard.
Luca Alfieri, being a specialist in Indo-Iranian languages as well as Armenian, Latin and Greek, will cover these field of the IE family within the task of the data collection.
Furthermore, he will be more specifically concerned with a typical though unrecognized case of “second generation isogloss”: the IE parts-of-speech system. As already observed in Alfieri (2009, 2011), the Proto-IE parts-of-speech system is only preserved in Vedic and Avestan and differs from that found in Latin, Germanic and in all of the other Western IE languages. While Vedic displays two classes of simple items in the lexicon (roots and primary nouns) and has only about a score of primary adjective stems, usually encodes the quality predicate through a verb-like strategy and the quality modifier through a nominalized form of a verbal root, Latin shows three classes of primary lexemes (noun, verb and adjective), and typically encodes the quality predicate and the quality modifier through a primary adjective stem. Therefore, one of the most typical hallmarks of the IE family, namely the noun/verb/adjective division, must be re-considered as a “second generation” isogloss only.
This fact will be demonstrated through the study, on the one hand, of the independent birth of the adjectival declensions in the IE languages and, on the other, the lexicalization pattern occurred in all of the Western and modern IE language whereby a derived stem class (the nominalized forms of the verbal roots employed as modifiers) has become a simple stem class (the adjective class), producing the loss of the root function and the change in the parts-of-speech system.
In case a scholarship will be granted, the scholarship holder will devote his whole activity to the reviewing and the collection of the data from the preceding scholarly literature. This cannot be considered a purely mechanical procedure, since it presuppose an appropriate knowledge of the theoretical framework that we are based on, and the capacity of applying it to the analysis of particular linguistic facts. The resulting data will populate the projected database of isoglosses.
A laborer will be hired in order to provide a computer-based support for the data collecting task. One of the goals is to provide a geographically mapped representation of the isoglosses. The topological data obtained from the comparison of isoglosses, together with the geographical interpretation thereof, will be build into a database that has to be projected. This database will be published on an internet site as an open resource, similarly to such projects as the following (some of them are also oriented to the study of contact phenomena):
– The World Atlas of Language Structures (Max Plank Gesellschaft)
– The Indo-European Phonological Inventory Database (Pavia)
– TriMCo Dialectal Corpus (Mainz)
– The Atlas of the Araxes-Iran Linguistic Area (Max Plank Gesellschaft)
– The Universals Archive (Konstanz)
– The Tower of Babel: An Etymological Database (Moscow)
L. Alfieri, 'La categoria dell’aggettivo in vedico', Archivio Glottologico Italiano 94/1 (2009): 3–49
— 'A Radical Construction Grammar approach to Vedic adjective', Rivista degli Studi Orientali 84 (2011): 247–261
Ph. Baldi & P. Cuzzolin, New Perspectives on Historical Latin Syntax: Constituent syntax. 3. Quantification, Numerals, possession, anaphora. Berlin, New York: Mouton De Gruyter, 2010
B. Bauer, Archaic Syntax in Indo-European. Berlin, New York: Mouton, 2000.
G. Beguš, 'The RUKI-Rule and Some Related Topics in Indo-Iranian Phonology'. 2012 [online]
W. Belardi, Dal latino alle lingue romanze. Il vocalismo. Roma: Bulzoni, 1979
K. Brugmann & B. Delbrück, Grundriß der vergleichenden Grammatik der indogermanischen Sprachen. Strassburg: Trübner, 1886–1900
C. Ciancaglini, 'Outcomes of the Indo-Iranian Suffix *-ka- in Old Persian and Avestan', G. P. Basello & A. V. Rossi (eds.), Dariosh Studies II. Persepolis and its Settlements: Territorial System and Ideology in the Achaemenid State: 77–86. Napoli, 2012
H. Goebl, '«Stammbaum» und «Welle». Vergleichende Betrachtungen aus numerisch-taxonomischer Sicht', Zeitschrift für Sprachwissenschaft 2 (1983): 3–44
R. Gusmani, 'A proposito del passaggio s > h in iranico, greco ecc.', Archivio Glottologico Italiano 57 (1972): 10–23
O. Höfler, 'Stammbaumtheorie, Wellentheorie, Entfaltungstheorie', Beiträge zur Geschichte der deutschen Sprache und Literatur 77 (1955): 424–476
R. Jakobson, 'Über die phonologischen Sprachbünde', Travaux du Cercle Linguistique de Prague IV (1931): 164–183
A. Keidan, ‘Branch-crossing Indo-European isoglosses: a call for interest’, Materialy čtenij pamjati I. M. Tronskogo 17 (2013): 406–417.
— ‘Le iscrizioni novgorodiane su corteccia di betulla in ottica comparatistica’, Incontri Linguistici 32 (2009): 175–196
— 'Predicative possessive constructions in Japanese and Korean', Rivista degli Studi Orientali 81 (2008): 339–367
L. Kulikov, 'Voice and valency derivations in old Indo-Aryan in a diachronic typological perspective: the degrammaticalization of the middle and other trends in the vedic verbal system', L. Milazzo (ed.), Usare il presente per spiegare il passato. Teorie linguistiche contemporanee e lingue storiche: 161–191. Roma: Il calamo, 2012
— 'The Proto-Indo-European case system and its reflexes in a diachronic typological perspective: Evidence for the linguistic prehistory of Eurasia', Rivista degli Studi Orientali 84 (2011): 289–309
R. Lazzeroni, 'Per una definizione dell'unità indo-iranica', Studi e saggi linguistici 8 (1968): 132–159
— 'Meillet indoeuropeista', A. Quattordio Moreschini (ed), L'opera scientifica di Antoine Meillet: 83-95. Pisa: Giardini, 1987
A. Lubotsky, 'Avestan compounds and the RUKI-rule', H. Eichner et al. (eds.), Compositiones indogermanicae in memoriam Jochem Schindler: 299–322. Praha: Enigma Corporation, 1999
A. Martinet, Économie des changements phonétiques. Bern: Francke, 1955
A. Meillet, Les dialectes indo-européenns. Paris: É. Champion, 1922
J. Nichols, Linguistic Diversity in Space and Time. University of Chicago Press, 1999
S. Patri, 'Une correspondance irano-slave en novgorodien', Historische Sprachforschung 114/2 (2001): 284–297
— 'La syntaxe de *-kwe “si, et” en novgorodien et en vieux slave, topicalisation et connectivité, Indogermanische Forschungen 108 (2003): 279–304
Ch. Prescott, 'Germanic and the ruki Dialects', Copenhagen Studies in Indo-European 4: 381–389. Copenhagen: Museum Tusculanum Press, 2012
E. Sapir, Language. An Introduction to the Study of Speech. New York: Harcourt, Brace, 1921
A. Schleicher, Compendium der vergleichenden Grammatik der indogermanischen Sprachen. Weimar: H. Böhlau, 1862
J. Schmidt, Die Verwandtschaftsverhältnisse der indogermanischen Sprachen. Weimar: H. Böhlau, 1872
H. Schuhardt, Slawo-deutsches und Slawo-italienisches. Graz: Leuschner & Lubensky, 1884
S. Thomason, Language Contact. Edinburgh University Press, 2001.