Substrate Languages Research Articles

A Mediterranean Thesaurus: Substratal Plant Names and Landscape Terms in Latin and Romance (An Experimental Approach)

Abstract In this paper we investigate lexical materials in Latin and Romance that appear to stem from ancient substrata in the Mediterranean. We focus on plant names and landscape related vocabulary, e.g., words for ‘berries’, ‘shrubbery, herbs and trees’, and words for ‘rocky’ and ‘swampy’ terrains, as both of these semantic fields are parts of the lexicon considered to be of very archaic origin. After investigating evidence for potential cognates in languages in the Mediterranean and exploring the geolinguistic distribution of the investigated words, we try to form conclusions about the involved strata and their chronological organization. We identify two important strata: an older Eurafrican layer (Hubschmid 1960) and a widely attested Euskaro-Caucasian layer associated with the arrival of Neolithic farming (see the discussion in Bengtson (2017a/2022, 2017b) and Bengtson/Leschber (2019, 2021, 2022)). The so-called ‘Mediterranean Thesaurus’ (in memory of Johannes Hubschmid) aims to establish a compilation of lexical data featuring words with unclear or highly disputable etymologies in the languages around the Mediterranean Sea. As a first step, we are collating words in an open-ended table and their geolinguistic/dialect-geographical distribution across the ranges of the Mediterranean languages, some of which are geolinguistically remote areas. We trace two main questions: (a) whether there are characteristic semantic fields that can be more often identified (such as plant names and topography); (b) whether we can highlight typical phonetic features and sound clusters in these words. Since the period of the Swiss and Italian etymological pioneers of the 20th century (Johannes Hubschmid, Giovanni Alessio, Carlo Battisti, Vittorio Bertoldi, Jakob Jud, Laura Lombardo, Alfredo Trombetti and many others), there has been great scientific progress in human genetics and archaeology that is increasingly revealing the prehistory of the Mediterranean. We recognize the work of these etymologists who – under much more demanding circumstances – offered suggestions about the origin of etymologically difficult words. Several hypotheses have been revisited in order to include or to reject the contribution of a specific substrate to the linguistic substrata of the macro-area. Our aim is to explore the prehistory of the Mediterranean area to understand the ethnogenesis of its peoples and their rootedness in millennia past, and to find evidence for a tentative stratification of substratum languages.

Loss of unreleased final stops among Mandarin-Min bilinguals: Structural convergence of languages in contact

The two languages of a bilingual speaker are interconnected and mutually influence linguistic forms and structures. This study presents a case in which two languages in contact exhibit phonotactic asymmetries but converge on abstract phonological units by bilingual speakers. The specific case examined here concerns the change-in-progress of unreleased final stops among young Mandarin-Min bilingual speakers in Taiwan. Phonotactically, obstruent finals are illegal in Taiwan Mandarin, whereas Taiwanese Southern Min (TSM), a local substratum language, allows obligatorily unreleased final stops. In the discrimination of stimuli modeled after TSM, bilingual listeners were consistently outperformed by Korean listeners, a non-native reference group without restrictions against obstruent finals. A follow-up production study revealed that final stops produced by the bilingual speakers were prone to deletion accompanied by vowel lengthening, similar to a long vowel in an open syllable, as well as frequent substitution. Furthermore, strong correlations were found between bilingual speakers’ perception and production accuracy, indicating a bidirectional co-evolution between perception and production during language development. Taken together, the results suggest that a loss of unreleased final stops is underway in TSM through the structural convergence of two interacting phonological systems within bilingual individuals.

Preterite Formations in Niya Prakrit and Khotanese: A Case of Grammatical Interference?

Some features of the North-Western Prakrit that was used as the administrative language of the kingdom of Kroraina (‘Niya Prakrit’) have in the past been more or less vaguely described as the result of language contact. One particularly striking feature that invites such an account is the innovative preterite of this Prakrit variety. We argue that, from a structural point of view, this formation and its morphosyntactic behaviour can be plausibly attributed to interference from Khotanese. In addition, a scenario involving Khotanese as a substratum language of Niya Prakrit may also be well accounted for from a historical and sociolinguistic perspective.

The treatise on the history of spanish by B. de Aldrete (1606) as the first textbook of romance philology

The paper analyses the role of B. de Aldrete’s treatise “Del Origen y principio de la lengua castellana o romance que oi se usa en España” (1606) in the development of Romance philology. The XVII-century author writes about the most important aspects of internal and external history of Spanish, such as: pre-Romance Spain and substratum languages; Roman conquest and romanization; Hispanic Latin; German conquests of Spain; Arabic conquest and the Reconquista; formation of kingdoms in the north and state-building processes; sociolinguistic situation in Spain; the role of Spanish in the New World; changes from Latin to Spanish in phonetics and morphology; sources of Spanish lexis; early written texts; territorial, social, functional variation of Spanish. Apart from the aspects of Spanish philology, B. de Aldrete pays attention to the formation and functioning of Pyrenean languages: Catalan, Galician, and Portuguese. However, B. de Aldrete does not limit himself to examining Ibero-Romance languages. Many aspects of the history of Spanish are shown against a wider, Romance background, bearing in mind the earlier tradition (the Antiquity, in the first place). He also confronts Spanish with other Romance languages and Latin. The analysis of the first treatise on the history of Spanish makes one reconsider B. de Aldrete’s contribution to the development of language description models and the bases of Romance philology. The treatise sets up a model of Romance philology as a full-fledged philological discipline.

Geographical Typology and Linguistic Areas, with Special Reference to Africa ed. by Osamu Hieda, Christa König, Hirosi Nakagawa (review)

Reviewed by: Geographical Typology and Linguistic Areas, with Special Reference to Africa ed. by Osamu Hieda, Christa König, Hirosi Nakagawa Jeffrey Heath Geographical Typology and Linguistic Areas, with Special Reference to Africa. Edited by Osamu Hieda , Christa König , and Hirosi Nakagawa . Tokyo University of Foreign Studies, Studies in Linguistics 2 . Amsterdam : John Benjamins , 2011 . Pp. vi + 320 . $143.00 (hardcover). This is a collaboration between a Japanese university and a collection of western linguists, brought together for a 2009 conference in Tokyo. It has the feel of an actual conference, complete with a blessing from the university president and a presentation of the university’s Center for Corpus-Based Linguistics. The chapters are broadly united by a language-contact focus, ranging from specific language pairs to broad areal linguistics. About two-thirds of the volume deals with grammatical effects of contact. The authors appreciate that native and foreign structures interact in a complex way, only occasionally resulting in perfect calques (isomorphism). Alexandra Aikhenvald’s pro-grammatic “Areal Features and Linguistic Areas: Contact-Induced Change and Geographical Typology” (pp. 13–39) proposes a leaky four-part typology of contact-induced processes: reanalysis; reinterpretation and extension (of inherited morphemes and categories); grammaticalization (see below on Heine’s chapter); and grammatical accomodation (contamination by a phonologically similar foreign form). Yaron Matras, in “Explaining Convergence and the Formation of Linguistic Areas” (pp. 143–60), advocates a more cognitive approach to borrowing and pattern transfer, with a top-down emphasis on communicative strategies. [End Page 397] Bernd Heine’s “Areas of Grammaticalization and Geographical Typology” (pp. 41–66) considers whether suspiciously similar grammaticalizations within an area are single-step calques (“polysemy copying”) or parallel multistage compressions with light prodding from contact languages. Much of the paper is concerned with African reflexive-object constructions (‘he cut him’, ‘he cut his body’, ‘he cut his head’, ‘he cut his life/soul’, or ‘he cut-reflexive’ with a verbal derivational affix). While ‘body’ reflexives dominate central and southern Africa, Heine finds that ‘head’ reflexives are dominant in two zones (western Sahel and Ethiopia), separated by an eastern Sahel zone with mostly ‘life/soul’ reflexives. Even Arabic dialects obey these regional patterns. Complicating factors are reflexive or logophoric syncretisms, and the discovery since the chapter was written of Russian-like transpersonal reflexives in some Dogon languages (Togo Kan, Tomo Kan). Tom Güldemann’s “Proto-Bantu and Proto-Niger-Congo: Macro-Areal Typology and Linguistic Reconstruction” (pp. 109–41) is a serious reflection on macrolevel reconstruction methodology. Niger-Congo is unusual in that several linguists have argued for a protolanguage with heavy-duty affixal morphology that has been whittled down in some of the descendants still in West Africa, such as Kwa, but preserved in mainstream Bantu, which spread far to the east and south. Outside of Africa, the usual pattern is for linguists to posit an ideal “golden age” protolanguage well endowed with free particles, each with a simple gloss, that later became compacted into morphologically messy daughter languages. Güldemann is cautiously sympathetic to the view that Proto-Niger-Congo was semisynthetic, and that mainstream Bantu developed its more baroque affixal morphology later under the influence of long-lost substratum languages. Matthew Dryer’s “Noun-Modifier Order in Africa” (pp. 287–311) reflects the transition in professional linguistic typology from Greenbergian worldwide ordering correlations to area-specific correlations. The latter imply that deep historical relations (genetic and contact) can trump synchronic functional pressures. Methodologically, this requires not only a large number of languages but also consideration of distributions across genetic stocks. Using detailed maps, Dryer shows that modifiers (adjectives, numerals, definites, possessors) have a stronger tendency to follow nouns in sub-Saharan Africa than in other continents, when intersecting features (e.g., verb-object versus object-verb order) are controlled for. Brief mention will suffice for the other grammatical chapters. Christa König’s “Case Marking and Linguistic Geography” (pp. 67–90) considers two pairs of adjacent Cushitic and non-Cushitic languages where structural case-marking systems have converged unidirectionally. Kazuhiro Kawachi wrestles inconclusively with definitions of linguistic area in “Can Ethiopian Languages Be Considered Languages in the African Linguistic Area? The Case of Highland...

An acoustic study of front rounded vowels in Shetland dialect.

This paper presents an acoustic analysis of front rounded vowels (FRVs) in the dialect spoken in the Shetland Islands, the northernmost locality of the British Isles. FRVs are typologically marked and estimated to occur in only 6.6% of the world’s languages [I. Maddieson, in Haspelmath et al. The World Atlas of Language Structures (2005)]. Their occurrence in the Shetland dialect is, at least partly, attributable to a Scandinavian substratum language. There is significant variation across the archipelago regarding several aspects such as (1) the number of lexically contrastive FRVs, (2) phonetic quality (close to half-close), (3) contrastive length, and (4) lexical distribution and support. This paper presents an investigation of three speakers from one locality in which FRVs have retained firm lexical support. The issues addressed concern the dialect’s overall acoustic vowel space (based on F1, F2, and F3), the position of FRVs within the acoustic space, and what the contrasts among FRVs and other adjacent vowels appear to rest on acoustically. Special focus is directed to phonetic contexts that support the greatest number of vowel contrasts and display the most crowded acoustic vowels spaces.

Must in Singapore English

Abstract In Singapore English, the modal verb must is predominantly deontic, in contrast with the usage profile of must in most native varieties of English. I account for this variation in terms of three linguistic factors prevalent in the contact ecology of Singapore English. First, the deontic and epistemic modalities are expressed by different modal expressions in Chinese, the main substratum language. Second, the perfect form have V-en has experienced a sharp decline in Singapore English, which contributes to the decline of the epistemic must. Third, the grammaticalization path that must has undergone in English is not followed by the equivalent expressions in Chinese. These three factors exert convergent pressure on must towards the deontic modality. The paper demonstrates the importance of computerized corpora in the study of contact-induced linguistic change.

Multifunctionality in Pitkern-Norf’k and Tok Pisin

This paper investigates claims as to the role of categorial multifunctionality in the Pitkern-Norf’k (PN) language. It concludes that the grammar of this construction is far less productive than that of either English or Tahitian (the principal substratum language). Essentially, multifunctionality in PN is a subset of English. In this, it is very different from Tok Pisin, which developed an independent system of categorial multifunctionality over the first 80 years of its existence. The paper makes some suggestions as to why the productivity of the derivational lexicon is so different in the two languages.

On the properties of Saramaccan FU

The formfu(variantu) fulfills several functions. It is a preposition selecting NPs and clauses, tensed or infinitival. It is a mood marker occurring either between the subject and the verb, or before the subject. It is a complementiser selected by predicates of thewant-class; as such it is in a paradigmatic relationship with other complementisers in the language. Finally, in some contexts,fuappears to function as a case marker, rather than as a preposition, as it may be associated with several thematic roles. The first objective of the paper is to provide a detailed inventory of all the functions offuand a detailed description of its properties for each of its functions. This will be done mainly on the basis of published sources. The proposal thatfucan head various syntactic projections (P, Force, Fin, Mood, Kase) will be shown to account for its multifunctional character. The second objective of the paper is to discuss the origin of the properties offu. First, we consider the grammaticalisation scenario proposed in the literature. In this scenario, the prepositionfuwould have been reanalysed as a complementiser. We argue that this scenario is not an optimal one. Second, we consider a relexification scenario along the lines of Lefebvre (1998b). Although the form of the lexical item in question is derived from Englishfor, as has been noted by several authors, most of its other properties cannot be derived from this lexical item. A comparison of the properties offuwith those of corresponding lexical items in one of the substratum languages of Saramaccan, Fongbe (e.g. Smith 1987), yields a different conclusion: while the form of the Saramaccan lexical entry is derived from English, the bulk of its semantic and syntactic properties are derived from those of corresponding substratum language lexical entries. The properties of the creole lexical entry thus appear to follow from the re lexification account of creole genesis. In this case, however, two substratum lexical entries (nú, preposition and complementiser, andní, mood marker and complementiser) appear to have been relexified on the basis of a single superstratum formfor, yielding the creole lexical entryfucumulating all the functions of the two substratum entries. Some details distinguish the creole lexical entry from the two substratum ones. It will be shown that the make up offuhas also involved some reorganisation of the original lexicon, and some innovation.

Creolization and Contact (review)

Reviewed by: Creolization and contact ed. by Norval Smith, Tonjes Veenstra Don E. Walicek Creolization and contact. Ed. by Norval Smith and Tonjes Veenstra. (Creole language library 23.) Amsterdam: John Benjamins, 2001. Pp. 323. ISBN 1588111113. $110 (Hb). The twelve papers in this volume were inspired by a 1995 workshop on language contact and creolization at the University of Amsterdam. They cover five general topics: relexification, second-language learning, koine-formation, contact linguistics, and sociolinguistics. The collection focuses on contact between social groups and individuals rather than merely on interactions of communities. Claire Lefebvre begins the first section by describing how relexification, dialect leveling, and re-analysis apply to the genesis and development of functional category lexical items in Haitian. She shows that both superstratum and substratum languages (specifically Fongbe) are reflected in creole lexical entries. In the next chapter, ‘Voodoo chile’, Norval Smith compares Haitian to Saramaccan. Asking why these two languages have different end-results despite shared Eastern Gbe influence, Smith examines work by Lefebvre. He suggests that both universalist and substrate approaches partially explain creole genesis. In Ch. 3, Anthony Grant provides an informative overview of research on language intertwining. He argues that this process is distinct from creolization and offers a detailed look at three languages: Berbice Dutch, Zamboangueño, and Cotabateño. In the next paper, Maarten Mous explains paralexification. He suggests that this process of adding word forms to lexical entries leads to language intertwining. In the first contribution on second-language learning, Ulla-Brit Kotsinas describes learner varieties of Swedish. She identifies several ‘deviant’ features of pidginized Swedish, characteristics she sees as [End Page 189] evidence of compensatory strategies rather than L1 interference. Kotsinas also presents evidence of creolization among first-generation immigrants. In the next chapter, responding to Kotsinas, Pieter Muysken looks at the grammatical properties of Negerhollands in terms of L2 acquisition of Dutch. Muysken recognizes features that can be attributed to the acquisition of Dutch as a second language as well as some that seem to reflect West African patterns. The section on koine-formation and convergence begins with Jeff Siegel’s article on creole genesis and Fiji Hindi. He describes a sociolinguistic model in which mixing, leveling, and simplification each play key roles. Frans Hinskens responds to Siegel in the next chapter by questioning how empirical the latter’s sketch is. He includes questions about Siegel’s methodology and suggestions for further development of the model. In Ch. 9, Silvia Kouwenberg points out that there is no coherent theory of convergence in historical linguistics. Criticizing analyses that identify convergence only retrospectively, she shows that it serves primarily as a descriptive concept. Sarah Thomason contributes the volume’s first chapter on contact linguistics. She compares pidgin/creole genesis with more typical contact-induced change. Relatedly, she holds that borrowing plays a significant role in mixing but not in genesis. In Ch. 11, Ellen Prince approaches Yiddish not as a creole but as a language in contact. She advances the theory that contact effects result from pragmatic and semantic intentions in which speakers target concepts as expressed in the contact scenario. The final chapter, ‘Social stratification and network relations in the formation of Sranan’, focuses on sociohistorical factors operating at the microlevel. Highlighting insightful sociolinguistic studies written in Dutch on this topic, Jacques Arends dispels myths about plantation life and lists examples of overlooked network relations. As in previous chapters, this exciting work reminds the reader that similar phenomena could be investigated for other creoles. This excellent and impressive volume approaches the problems it addresses at a variety of levels and will surely motivate provocative discussion and research. Don E. Walicek University of Puerto Rico, Río Piedras Copyright © 2004 Linguistic Society of America

The interplay of relexification and levelling in creole genesis and development

This paper bears on the relationship between relexification and levelling, two processes that play a role in Creole genesis and development. It is a well-documented fact that situations where creoles develop involve several substratum languages and one (or at least one major) superstratum language. Relexification of several substratum languages on the basis of a single/major superstratum language provides the creole community with a common lexicon, hence a common language. As a mental/cognitive process, relexification is an individual process. Consequently through relexification, the specificities of the various substratum languages are reproduced in the incipient creole, thus creating what might be referred to as different dialects of the new language. Relexification of several substratum languages, on the basis of a single superstratum language, can be viewed as the major source of variation in an incipient creole. This provides a sound explanation for the fact that different substratum languages may contribute different features to a particular creole. Dialect levelling, as discussed in the literature on dialects in contact, refers to the reduction of variation between dialects of the same language, in situations where these dialects are brought together. On the basis of three sets of data from Haitian creole, it is shown that, in creole development, dialect levelling operates on the output of relexification. The role of levelling in creole genesis and development accounts for the fact that the properties of some specific lexical entries of the creole may depart from those of the corresponding lexical entries in the individual substratum languages.

Substratum Semantics in the Verbal Lexicon of Haitian Creole

The aim of this paper is to document the presence of substratum semantics in the verbal inventory of Haitian creole on the basis of a comparison of a sample of verbs in Haitian, French (its lexifier language) and Fongbe (one of its substratum languages). The paper begins with a comparison of the meanings of a sample of Haitian, French and Fongbe verbs. Although the phonological representations of the Haitian verbs are derived from the phonetic representations of French verbs, the details of their semantics do not correspond exactly to those of French, but rather to those of Fongbe. Idiomatic expressions in Haitian are often expressed with similar verbal phrases in Fongbe, whereas they are rendered by a simple verb in French. Aspectual properties of verbs (stative/non-stative) constitute another facet of verbal semantics, and I compare the aspectual properties of Haitian, French and Fongbe verbs. Haitian verbs, like Fongbe verbs, are often not specified for aspectual properties, in contrast to French verbs. Thematic properties constitute yet another topic in the semantics of verbs, addressed here from the point of view of the verbs' agentiveness. On the basis of these various types of data, it is argued that the bulk of Haitian verbs' semantic properties have been carried over into the creole from the substratum lexicons. This situation argues in favor of the claim that the process of relexification plays a central role in the formation of a Creole's lexicon. Some cases of acquisition of French verbs are also reported. Such cases have triggered a reorganization in the properties of the verbal lexical entries originally relexified from the Haitian substratum languages.

Portuguese in Sri Lanka: influence of substratum languages

Sri Lanka Portuguese Creole is considered to fall within the group of Indo-Portuguese Creoles, which were Portuguese-based Creoles that once flourished around coastal India and Sri Lanka. Sri Lanka Portuguese Creole arose when the Portuguese established trading posts and fortresses in Sri Lanka from 1517 onwards. The purpose of this paper is to examine critically the influence of substratum (indigenous) languages on Sri Lanka Portuguese Creole.

The Tense, Mood, and Aspect System of Haitian Creole and the Problem of Transmission of Grammar in Creole Genesis

It is often assumed that creolization involves a break in the transmission of grammar. On the basis of data drawn from the TMA system of Haitian creole, as compared with those of its source languages — French, the superstratum language, and Fongbe, one of the substratum languages — this paper argues that creolization does not involve a break in transmission of grammar. The properties of the Haitian creole TMA system are shown to reflect in a systematic way those of its contributing languages. While the syntactic and the semantic properties of the TMA markers of the creole parallel those of Fongbe, the markers' phonological form appears to be derived from phonetic strings found in the superstratum language. This systematic division of properties is predicted by the hypothesis that relexification has played a major role in the formation of the creole. The fact that the lexical entries of the creole have phonological representations which are derived from phonetic strings found in the superstratum language is the visible signal that creolization involves the creation of a new language. The fact that the lexical entries of the creole show semantic and syntactic properties that parallel those of the languages of the substratum argues that there has been no break in the transmission of grammar in the formation of the creole.

Latin and Punic in Contact? The Case of the Bu Njem Ostraca

The ostraca of Bu Njem come from a military outpost on the North African fringes of the Empire. Vernacular languages were spoken in the area. The ostraca record, among other things, contact between soldiers and the local population, and contain various African (Punic or ‘Libyan’) words and names, some of them previously unrecorded. The soldiers themselves have in many cases African names, or names with a special African connection, and it is likely that many were recruited locally. If so they may not have been fluent Latin speakers, and consequently the Latin which they wrote raises unusual questions. Is it Latin at all, or perhaps a pidgin or Creole? Or, on the contrary, is the language merely bureaucratic and formulaic Latin of no great interest? Do we, at last, have some hard evidence for a regional variety of Latin, in this case perhaps influenced by a substratum language or languages?.

The Language Bioprogram Hypothesis and Singaporean English

In his rather controversial book "Roots of Language" 1981, Bickerton claims that 'many of the prerequisites for human language were laid down in the course of mammalian evolution, and that the most critical of those prerequisites... was the capacity to construct quite elaborate mental representations of the external world....' (Bickerton 1981:294-295). The Language Bioprogram Hypothesis (LBH) concerns 'a biological' or 'a species-specific program for language, genetically coded and expressed in ways still largely myste rious, in the structures and modes of operation of the human brain' (Bickerton 1984:173). Among the important distinctions of the LBH are those between state and process, punctual and non-punctual, specific and non-specific and causative and non-causative. Such dis tinctions are clearly evident when a pidgin becomes a creole. The in nate punctual/non-punctual distinction is manifested in Bickerton's investigation of past tense marking in 2 decreolizing situations- Guyana and Hawaii. Verbs which were used non-punctually received a lower marking for past tense than verbs that were used punctually. What are the factors which influence variation in past tense marking of Chinese Singaporeans? The substratum language may have a strong influence since in Chinese there is no morphological change to the past form of the verb. Or it could be that the innate Language Bioprogram has an influence as well. Naturally, other fac tors such as teaching and general language learning strategies cannot be ignored. This paper is based on an empirical investigation of recordings of spontaneous speech of a hundred ethnically Chinese Singaporeans who have had English-medium education and discusses variation in past tense marking in Singaporean English as well as pedagogical im plications of the LBH for second language acquisition.

Question words in pidgin and creole languages

Question words in pidgin and creole languages

クサントスのレートーオン出土の三言語併用碑文とリュキア語研究の現状 -特にその統語構造を中心に-

The discovery of the trilingual inscription from Letoon by H. Metzger, 1973, has made a new epoch in the long history of the study of the Lycian language. This paper aims to clarify the main characteristics of Lycian and its position among the Anatolian languages chiefly through a syntactic analysis of the Lycian text. Lycian shares with the other Anatolian languages such features as paratactic constructions and the use of sentence introductory particles but shows, on the other hand, a striking contrast with them in word order and the use of preposition: it is the only VOS and prepositional language among the Indo-European Anatolian languages. This syntactic type of Lycian seems to correlate with its case system, which is characterized by the loss of morphological distinctions of many case endings through phonetic changes and thus is functionally the least effective among the Anatolian case systems. This fact well agrees with the general constatation that in the Indo-European history of syntactic development the languages with the VO word order tend to weaken or lose their case systems whereas those with the OV word order usually keep or strengthen them. With regard to the case endings themselves, especially in plural, however, Lycian shows a quite archaic and conservative character compared with the other Luwian languages, which radically restructured their systems by creating new endings on the basis of acc. pl. -nz. Hence we can infer that Lycian separated very early from the other south Anatolian languages and went its own way in the peculiar development toward VOS order. The writer also suggests that this development was probably caused by the influence from an unknown substratum language which seems also to have exerted a similar influence upon the Greek language.