ABSTRACT. The term evolutionary linguistics has spread rapidly in linguistics since it was first innovated apparently by James W. Minett & William S-Y. Wang in 2005. It has even become part of the name of a conference series on language evolution, viz., the International Conference in Evolutionary Linguistics (CIEL), whose latest meeting was at Indiana University at Bloomington in August 2016. However, like that of its model, evolutionary biology, the meaning of the term has evolved beyond the concerns with phylogenetic emergence only, with evolution still interpreted as ‘(process of) change associated with unfaithful transmission or reproduction’.  Concerned almost exclusively with the phylogenetic emergence of language, biolinguists such as Berwick & Chomsky (2011, 2016) have focused on the genetic underpinnings of the emergence of Language in mankind and the nature of the faculty of language, assuming a saltatory switch from no Language to Language. On the other hand, gradualists such as Croft (2008) and Mufwene (2013) have extended the term both to phylogenetic emergence and to various aspects of language change (see also Croft 2000), while McMahon & McMahon (2012) would rather restrict its application to the former only.

Arguing that evolutionary biology is about more than phylogeny, Mufwene (2001, 2008), for example, has extended evolutionary linguistics to the study of speciation in modern languages (including the emergence of creoles) and of the vitality of modern languages (including language endangerment and loss). He argues that the same ecological factors that account for structural change, especially those associated with population movements, language contact, and changing population structures, also drive language shift and the speciation of the prevailing language.

Thus, in a broad sense, historical linguistics has been made part of evolutionary linguistics, in an approach that invokes the specific socioeconomic ecologies in which speakers evolve as actuators of change. This approach makes it critical to address the actuation question (Weinreich et al. 1968, McMahon 1994, Labov 2001, Mufwene 2014), while showing that the same communicative activities that drive structural change under specific ecological pressures also drive language speciation and affect differentially the vitality of languages. One particular question that arises is why historical or genetic linguists have shown such little interest in the current preoccupation with language endangerment and loss, when one may expect them to be better placed to provide historical perspectives on the subject matter? I will answer the question in my presentation. Meanwhile, I submit that traditional historical linguistics as part of evolutionary linguistics is enriched by the explanatory ambitions and the broader scope of concerns of the latter. Worth considering is also whether the traditional transformational description of change as A à B cannot be replaced by a variational, competition-and-change-selection-based account, viz., B prevailed over A under specific ecological conditions, as suggested by Darwinian, variational evolution and supported by facts in genetic creolistics.

References

Berwick, Robert C. & Noam Chomsky. 2011. The biolinguistic program: The current state of its development. In The biolinguistic enterprise: New perspectives on the evolution and nature of the human language faculty, ed. by Anna Maria di Sciullo & Cedric Boeckx, 19-41. Oxford: Oxford University Press.
Berwick, Robert C. & Noam Chomsky. 2016. Why only us: Language and evolution. Cambridge, MA: MIT Press.
Croft, William. 2000. Explaining language change: An evolutionary approach. London: Longman.
Croft, William. 2008. Evolutionary linguistics. Annual Review of Anthropology 37.219-234.
Labov, William. 2001. Principles of linguistic change: Social factors. Malden, MA: Blackwell.
McMahon, April. 1994. Understanding language change. Cambridge: Cambridge University Press.
McMahon, April & Robert McMahon. 2012. Evolutionary linguistics. Cambridge: Cambridge University Press.
Minett, James W. & William S-Y. Wang, eds. 2005. Language acquisition, change and emergence: Essays in evolutionary linguistics. Hong Kong: City University of Hong Kong Press.
Mufwene, Salikoko S. 2001. The ecology of language evolution. Cambridge: Cambridge University Press.
Mufwene, Salikoko S. 2008. Language evolution: contact competition, and change. London: Continuum Press.
Mufwene, Salikoko S. 2014. Language ecology, language evolution, and the actuation question. In Language contact and change: Grammatical structure encounters the fluidity of language, ed. by Tor Afarli & Brit Maelhum, 13-35. Amsterdam: John Benjamins.
Weinreich, Uriel, William Labov, & Marvin I. Herzog. 1968. Empirical foundations for a theory of language change. In Directions for historical linguistics: A symposium, ed. by Winfred P. Lehman & Yakov Malkiel, 97-195. Austin: University of Texas Press.

ABSTRACT. Mufwene (2015), argues that AAVE [African American Vernacular English] and WASE [White American Southern English] “appear to have been one and the same regional variety until Jim Crow was introduced in the late nineteenth century and triggered the Great Migration of African

Americans out of the South. AAVE was invented as a separate ethnolect in the North, where most White Americans were then getting their first exposure to American Southern English” (p. 59).

In the nineteenth century, 16,000 African Americans immigrated to Liberia. Of this number, 91% arrived in Liberia either prior to the American Civil War (1861-1865) or within seven years of its cessation (Singler, 1989). Singler (2004, 2015) argues that the Liberian Settler English (LSE) spoken today is a direct descendant of the language that the original Settlers brought with them from the United States. LSE shares a number of features with AAVE, including Urban AAVE, and WASE. However, it also shares many features with Urban AAVE to the exclusion of WASE (and for which no evidence of these features’ presence in prior WASE exists).

The present paper examines LSE and (Urban) AAVE, with particular reference to features shared by these two varieties to the exclusion of WASE. It explores the possibility of parallel development (subsequent to the African American migration to Liberia and to the Great Migration), the existence of “vernacular universals” (Chambers 2004), and post-migration influence (of AAVE on LSE, not vice versa), and it considers features that (1) the Settlers brought with them from the American South but which are not part of AAVE today or (2) are found in Urban AAVE but not in LSE. Additionally, it examines features that obtain in LSE and Urban AAVE alike but have undergone greater grammatical elaboration in Urban AAVE.

Key differences exist between LSE and Urban AAVE, an undeniable consequence of 150-200 years of separation. Ultimately, however, the weight of congruence argues against Mufwene’s assertion that (Urban) AAVE’s emergence as an ethnolect is relatively recent and Northern. Rather than being a post-Jim Crow, post-Great Migration phenomenon, Urban AAVE is—to use an LSE expression—the same old bumblebee in a brand-new suit.

References
Chambers, J.K. 2004. Dynamic typology and vernacular universals. In Bernd Kortmann (ed.), Dialectology meets typology: Dialect grammar from a cross-linguistic perspective, 127–145. Berlin: Mouton de Gruyter.
Lanehart, Sonja (ed.). 2015. The Oxford Handbook of African American Language. Oxford: Oxford University Press.
Mufwene, Salikoko S. 2015. The emergence of African American English: Monogenetic or polygenetic? Under how much substrate influence? In Lanehart (ed.), 57-84.
Singler, John Victor. 2015. African American English over yonder: The language of the Liberian Settler community. In Lanehart (ed.), 105-124.
_______.  2004. The morphology and syntax of Liberian Settler English. In Bernd Kortmann, Edgar W. Schneider, Clive Upton, Rajend Mesthrie & Kate Burridge (eds.), A Handbook of Varieties of English. Vol. 2: Morphology and Syntax, 879-897. (Topics in English Linguistics, ed. Bernd Kortmann & Elizabeth Closs Traugott.) Berlin, New York: Mouton de Gruyter.
_______.  1989. Plural marking in Liberian Settler English, 1820-1980.  American Speech 64.40-64.

ABSTRACT. In the Romance languages, subject clitics are solely found in contiguous central area: they are characteristic of the Gallo-Romance branch, i.e. French and Oïl dialects, Francoprovençal, a few northern and eastern Occitan varieties, Gallo-Italian dialects, and Raeto-Romance varieties (Swiss Romansh, Dolomitic Ladin of Val Badia). Furthermore, they exist in neighbouring Italian dialects (i.e. Veneto and Tuscan dialects) and non-Gallo-Romance Raeto-Romance varieties (Dolomitic Ladin, Friulian). In some varieties, they may be analysed as preverbal agreement markers, e.g. in colloquial French (Roberge 1986, Culbertson 2010), in some Francoprovençal and Occitan varieties as well as in many northern Italian dialects. However, Latin and the majority of the Romance languages do not feature subject clitics. The varieties mentioned above have developed them in a grammaticalization process. Different stages of this process are still exhibited by modern varieties. Especially Francoprovençal varieties differ to a great extent in the stage of grammaticalization reached. Subject clitics may occupy different structural positions reflecting their grammaticalization status: the canonical subject position in SpecTP/SpecAgrSP (i.e. same position as lexical subjects and free pronouns); the head position of the respective XP, adjoined to the finite V (e.g. Poletto 1995, Kato 1999, Culbertson 2010); a distinct but adjacent head position to the one with the finite V: D in Pers, V in Num (Roberts & Roussou 2003); the head position of the respective XP, incorporated to the finite V (i.e. complete loss of independent status). The situation in Gallo-Romance (in France, Switzerland, and northern Italy) is discussed on the basis of data from the national (ALF, AIS) and regional linguistic atlases (especially for Francoprovençal: ALLy, ALJA, TPPSR, ALAVAL, APV). The following hypotheses are explored: there is a correlation between the loss of rich agreement, the loss of the null-subject property (inherited from Latin in Romance) in some varieties, the development of clitic subject pronouns in all of these, the development of a complex interplay of pronominal systems (existence of a clitic or paradigmatic gap, its realization in a specific syntactic environment) and remaining verbal agreement, and the subsequent development of a new form of null-subject languages (NSL) with preverbal agreement in some of these varieties. The difference between the languages is where the EPP- and person/number-features are located: only on the (clitic) pronoun (non-NSL; isolating wrt to these features); only on the verb (NSL; agglutinating/inflecting); on both (null-subject status depending on the analysis of the clitic pronoun). I propose a Subject Agreement Cycle (cf. Givón 1971; van Gelderen 2011a, b) for Romance. My aim is to contribute to a better understanding of the linguistic change that has taken place by combining the grammatical analysis and the interpretation of the geolinguistic distribution of the phenomena involved in Romance languages and dialects.

Bibliography Culbertson, Jennifer (2010): “Convergent evidence for categorial change in French: From subject clitic to agreement marker” Language 86.1: 85-132. van Gelderen, Elly (2011a): The Linguistic Cycle. Language Change and the Language Faculty. Oxford: Oxford University Press. van Gelderen, Elly (2011b): “The grammaticalization of agreement” In: Narrog, Heiko & Heine, Bernd (eds), The Oxford Handbook of Grammaticalization [Oxford Handbooks in Linguistics]. Oxford: Oxford University Press. Givón, Talmy (1971): “Historical syntax and synchronic morphology: an archaeologist’s field trip” In: Papers from the Seventh Regional Meeting, Chicago Linguistic Society. Chicago IL: Chicago Linguistic Society, 394-415 Kato, Mary Aizawa (1999): “Strong and weak pronominals in the null subject parameter” Probus 11: 1-37. Poletto, Cecilia (1995): The Higher Functional Field. Evidence from Northern Italian Dialects [Oxford Studies in Comparative Syntax]. Oxford: Oxford University Press. Roberge, Yves (1986): “Subject doubling, free inversion, and null argument languages” The Canadian Journal of Linguistics 31: 55-79. Roberts, Ian & Roussou, Anna (2003): Syntactic Change. A Minimalist Approach to Grammaticalization [Cambridge Studies in Linguistics 100]. Cambridge: Cambridge University Press.

Linguistic atlases AIS = Jaberg, K[arl] & Jud, J[akob] (1928-1940): Sprach- und Sachatlas Italiens und der Südschweiz. 8 vol. Zofingen: Verlagsanstalt Ringier & Co. ALAVAL = Kristol, Andres (conception, direction) (1994–): Atlas linguistique audiovisuel du francoprovençal valaisan [anfangs: du Valais romand]. Réalisé par Gisèle Boeri, Federica Diémoz, Magda Jezioro, Raphaël Maître, Aurélie Reusser-Elzingre et al. Neuchâtel: Centre de dialectologie et d’étude du français régional, Université de Neuchâtel. ALF = Gilliéron, J[ules] & Edmont, E[dmond] (1902-1910): Atlas linguistique de la France. Paris: Champion. ALJA = Martin, Jean-Baptiste & Tuaillon, Gaston (1971-1981): Atlas linguistique et ethnogra¬phique du Jura et des Alpes du Nord (Francoprovençal Central) [Atlas linguistiques de la France par régions]. 4 vol. Paris: Éditions du Centre National de la Recherche Scientifique. ALLy = Gardette, P[ierre] (1950-1976): Atlas linguistique et ethnographique du Lyonnais. 5 vol. Lyon/Paris: Institut de Linguistique Romane des Facultés Catholiques de Lyon/Centre National de la Recherche Scientifique. APV = [Schüle, Ernest, Tuaillon, Gaston & Grassi, Corrado (premier comité scientifique), Telmon, Tullio et al.] (in prep.): Atlas des patois valdôtains. Aoste: Bureau Régional pour l’Ethnologie et la Linguistique (BREL). TPPSR = Gauchat, Louis, Jeanjaquet, Jules & Tappolet, Ernest (1925): Tableaux phonétiques des patois suisses romands. Relevés comparatifs d’environ 500 mots dans 62 patois-types [Glossaire des patois de la Suisse romande]. Publiés avec introduction, notes, carte et répertoires. Neuchâtel: Imprimerie Paul Attinger.

ABSTRACT. Grammaticalization is usually held to give rise to the phenomenon of layering, i.e. “the persistence of older forms and meanings alongside newer forms and meanings” (Hopper & Traugott 2003: 124), insofar as one and the same morpheme can develop different constructional patterns as in the classical examples of perfect in several European languages (cf. Harris 2003 and Drinka 2015 for a general picture):

(1) LAT nefāriōs ducēs captōs iam et comprehēnsōs tenētis (Cic., Cat. 3, 7, 16) ‘you have now captured the infamous leaders and hold them in custody’ OFR Et [chis empereres] avoit letres seur lui écrites (R. de Clari, p. 86, 1.11) ‘and this emperor had letters written on him’ FR Cet empereur avait des lettres écrites sur lui ‘This emperor had letters written on him’ FR Cet empereur avait écrit des lettres sur lui ‘This emperor had written letters on him’

In Old French the biclausal structure corresponds to the Latin pattern in which the object of the possessive verb is modified by a past participle featuring a resultant state. The modern French examples show the effect of layering of the verb avoir which either reflects the old biclausal pattern or profiles the new monoclausal perfect. Layering is often related to morphological differentiation because the erosive effect of grammaticalization can lead to the development of a morpho(phono)logical behavior which is different with regard to the source morphemes. This process is usually known under the label of divergence or split (Hopper & Traugott 2003: 118). For instance, while in its value as a full verb to go doesn’t admit the reduced form: I’m going to / *gonna London, the latter is possible with the auxiliary: I’m going to / gonna have cheesecake for dessert. Furthermore, one crucial change characterizing grammaticalization is the effect of constructionalization (cf. Traugott 2003) whereby the past participle loses its original agreement and the old biclausal structure gives rise to one word form of the embedded verb écrir: *Cet empereur avait écrites des lettres sur lui. Again, this can be related to the reductive effects of grammaticalization which also forces the identification of the subject of avoir with that of écrir. Against such general tendency towards reduction apparently connected with grammaticalization, in Titsch, a Walser German variety spoken in Gressoney (Aosta Valley), a morphological differentiation can be observed, in which the weak version of the past participle is used in the perfect construction while its strong correspondence is normally used in the passive construction resulting from the grammaticalization of the verb goa ‘to go’:

(2) a. mét der gallò heintsch … d’husgspònnto wollschtrangna gwäschet ‘with the gall they have washed the home-spun wool strand’ b. d’gròssò lougò ésch gwäschne kanget ‘the big laundry has been (lit. gone) washed’

Notice that in the passive construction the past participle gwäschn-e – but not the auxiliary kanget – agrees with the subject while this is not the case of the past participle gwäschet of the perfect. In this regard, the passive construction parallels the pattern of the typical copula construction displaying the agreement of the subject with the predicative expression:

(3) d’hannetò moss nid déck-é òn nid dénn-é si ‘The mashed potatoes should be neither thick-AGR nor thin-AGR’

Thus, the effects of grammaticalization have given rise to the reuse or the exaptation (cf. Author 2016) of a morphological differentiation – the traditional distinction between strong and weak conjugation of the Germanic verbs – which originally displayed a purely lexically-governed distribution. The latter has been remotivated by associating the morphological features with specific constructional patterns resulting from grammaticalization.

References Author. 2016. Co-opting exaptation in a theory of language change. In M. Norde & F. Van de Velde, eds. Exaptation in language change. Amsterdam, 57-92. Drinka, B. 2015. Sources of auxiliation in the perfects of Europe. In H. De Smet, L. Ghesquière & F. Van de Velde, eds. On Multiple Source Constructions in Language Change. Amsterdam, 129-174. Harris, A. C. 2003. Cross-linguistic perspectives on syntactic change. In B. D. Joseph & R. Janda, eds. The Handbook of Historical Linguistics. Oxford, 529-551. Hopper, P. J. & E. C. Traugott. 2003. Grammaticalization. 2nd ed. Cambridge. Traugott, E. C. 2003. Constructions in grammaticalization. In: B. D. Joseph & R. Janda, eds. The Handbook of Historical Linguistics. Oxford, 624-647.

ABSTRACT. Goal: Earlier diachronic research (Nuyts 2013) has revealed that the Dutch core modals moeten ‘must’, mogen ‘may’ and kunnen ‘can’ are currently involved in a process of syntactic re-autonomization (a considerable increase of uses without a main verb elsewhere in the clause), arguably a case of systematic degrammaticalization. This process started in the course of the New Dutch period. The new autonomous uses are grammatically very different from the original main verbal uses from which the auxiliary uses of these modals have developed (in the Middle Ages). And the process is not accompanied by a return to more ‘objective’ meanings (in Traugott & Dasher’s 2002 sense) – on the contrary, it appears to ‘focus’ on highly ‘subjectified’ meanings, the deontic one in particular. The present paper extends this investigation to the less central Dutch modals zullen ‘shall, will’, willen ‘will, want’, and (be)hoeven ‘need’, and it also looks at durven ‘dare’, not generally considered a modal in Dutch (unlike its cognate in German e.g.) but closely related. The question is: do these verbs show the same developments, with the same timing (historically), and the same or comparable grammatical and semantic properties?

Method: For the sake of comparability, this investigation uses the same method as the earlier study of the three core modals. We analyze the properties of the different verbs in corpus data from four different time periods: Old Dutch, Early Middle Dutch, Early New Dutch and Present Day Dutch. For each period we use a random sample of 200 instances of each verb (for PDD we use two separate sets of 200 instances, one with written and one with spoken language), selected from the available materials according to criteria such as representativity (e.g. in terms of text genres and regional spreading) and comparability across the periods.

Analysis: The analysis reveals some interesting similarities, but also some clear differences, with the developments in the core modals. To some extent this may be due to grammatical and/or semantic differences in the original forms from which the present verbs have emerged, but there are also clearly factors emerging in the course of the evolutions (e.g. the development of an important temporal meaning in zullen) which affect the re-autonomization process in some of these forms. As such the results throw new light on the earlier findings regarding the core modals, and lead to further reflection on the question to what extent the re-autonomization process in the Dutch modals is a matter of ‘group dynamics’, and on which factors play a role in it (e.g. analogy). We will situate the discussion in the context of the debate about the status of grammaticalization (Hopper and Traugott 2003) relative to other mechanisms and factors of language change.

References

Hopper, P. & E. Traugott. 2003. Grammaticalization. Cambridge: Cambridge University Press. Nuyts, J. 2013. De-auxiliarization without de-modalization in the Dutch core modals: A case of collective degrammaticalization? Language Sciences 36: 124-133. Traugott, E. & R. Dasher. 2002. Regularity in Semantic Change. Cambridge: Cambridge University Press.

ABSTRACT. The diachrony of the Germanic preterites has been given much attention in historical linguistics (Ball 1968; Tops 1974; Van Coetsem 1990; Hill 2010; Mailhammer 2007, among many other studies). As opposed to ‘idiographic’ studies which scrutinize the data with admirable philological detail, recent years have seen an upsurge in ‘nomothetic’ studies, that rely on statistical analysis (e.g. Cuskley et al. 2014). An exemplar study of the latter type is Lieberman et al. (2007). This study found that the regularization rate in English preterites is independent of time, and only depends on the frequency of the verb. According to this study, a verb’s regularization rate neatly scales with the square root of the frequency bin of that verb. A replication study for German by Carroll et al. (2012) has not been able to detect a similar time-independent ‘law’. This led to a sort of Mexican stand-off where English appears to be decidedly different than German in its obedience to the law of regularization, but with no clue as to who is the odd man out: is the time-independent nature of the English regularization rate the common case, and is German somehow off, or is the English exceptional in showing this law-like trend? Bringing more data to bear, we show that the results of Lieberman et al. (2007) are most likely a statistical artefact. Replicating the study for Dutch, which is both geographically and linguistically in-between its west-continental neighbours (Van Haeringen 1956, Hüning et al. 2006; Vismans et al. 2010; Ruigendijk et al. 2012; Smessaert et al. 2016), it becomes clear that the time-independent nature of the regularization rate fails to transpire across the board, and adding more detailed measurements in Lieberman et al.’s dataset breaks down the rate for English as well. We reassess the trends observed in the data, and give empirical corroboration for the suggestion by Carroll et al. that underlying the differences between English and German are demographic factors.

Ball, C. 1968. ‘The Germanic dental preterite’. Transactions of the Philological Society 67: 162-188. Carroll, R., R. Svare & J. Salmons. 2012. ‘Quantifying the evolutionary dynamics of German verbs’. Journal of Historical Linguistics 2(2): 153-172. Cuskley, C., M. Pugliese, C. Castellano, F. Colaiori, V. Loreto & F. Tria. 2014. ‘Internal and external dynamics in language: evidence from verb regularity in a historical corpus of English’. PLoS ONE: 9(8): e102882. Hill, Eugen. 2010. A case study in grammaticalized inflectional morphology: Origin and development of the Germanic weak preterite. Diachronica 27(3). 411-458. Hüning, M., A. Verhagen, U. Vogl & T. van der Wouden (eds.), Nederlands tussen Duits en Engels. Leiden: Stichting Neerlandistiek Leiden. Lieberman, E., J.-B. Michel, J. Jackson, T. Tang & M. Nowak. 2007. ‘Quantifying the evolutionary dynamics of language’. Nature 449(7163): 713-716. Mailhammer, R. 2007, The Germanic strong verbs: foundations and development of a new system. Berlin: Mouton de Gruyter. Ruigendijk, E., R. Vismans & F. Van de Velde (eds.). 2012. ‘Dutch between English and German’. Special issue of Leuvense Bijdragen – Leuven Contributions in Linguistics and Philology 98. Smessaert, H., J. van der Horst & F. Van de Velde. 2016. ‘Germanic Sandwich 2013’. Special issue of Leuvense Bijdragen – Leuven Contributions in Linguistics and Philology 101. Tops, G. 1974. The origin of the Germanic dental preterit. Leiden: Brill. Van Coetsem, F. 1990. Ablaut and the reduplication in the Germanic Verb. Heidelberg: Carl Winter. Van Haeringen, C.B. 1956. Nederlands tussen Duits en Engels. Den Haag: Servire. Vismans, R., M. Hüning & F. Weerman (eds.). 2010. ‘Dutch between German and English’. Special issue of Journal of Germanic Linguistics 22(4).

ABSTRACT. All the Semitic languages use a relative marker as at least one strategy of relativization, and all branches show reflexes or relics of reflexes of an interdental relative marker. The wide consensus among semitists that the relative pronoun was originally identical to the proximal demonstrative is based on the formal identity between the bases of the two in West Semitic, and on the wide attestation of the process Demonstrative > Relative in world languages (Heine and Kuteva 2002; Diessel 2009). In this paper we will review the evidence supporting these assumptions and argue that the current analysis of the Semitic relative is a procrustean bed, into which the Semitic evidence does not fit comfortably. We will show that there are a number of significant problems with the reconstruction of the relative pronoun, which, when taken together, make tracing its origin to the demonstrative highly unlikely. We will discuss the syntactic behavior and morhology of the relative marker to show that it is unlikely to be derived from the demonstrative. Instead, we will argue, that the opposite is true: the demonstrative in West Semitic is a secondary formation on the basis of the relative marker.

ABSTRACT. The Construct State in Bohairic Coptic

Linguists specializing in Afro-Asiatic languages have paid much attention to certain phenomena that are characteristic of the Afro-Asiatic language family such as the (i) the biconsonantal and triconsonantal root system, allowing for non-concatenative (using the traditional interoperation of concatenation) morphological inflection (McCarthy 1981), (ii) definiteness spreading throughout the noun phrase (Ritter 1991/1992), (iii) the construct state (CS), and (iv) the interaction between VSO and SVO canonical word orders (Ouhalla 1994). Following in their footsteps, I pursue investigation regarding the CS.

The CS is a syntactic noun phrase consisting of at least two nominal members that are in a genitive relationship (Benmamoun 2006). A CS nominal consists of a phonologically reduced head nominal, which is immediately followed by an embedded nominal phrase, the head and the specifier, respectively. The CS, although phrasal, acts as one prosodic unit (Siloni 2001).

Typically, the CS has been analyzed within various frameworks that have attempted to formulate a holistic account for the interaction between the syntax and the semantics in regards to the definiteness feature and definiteness spreading (Ritter 1991, Danon 2007, Heller 2002, Dobrovie-Sorin 2000/2003), while maintaining the observation that CS nominals, historically, show syntactic similarities to VSO patterns (Benmamoun 2003). These studies, while highly influential, were restricted to the Semitic languages, e.g Arabic and Hebrew. In this study I argue that what has been labeled as the short possessive construction (Haspelmath 2015) in another Afro-Asiatic language, Bohairic Coptic, should in fact be analyzed as a CS construction, with the idiosyncrasies being due to language variation, rather than as a separate phenomena.

Like the CS, the short possessive found in Bohairic Coptic displays (i) head-specifier linear order, (ii) phonological reduction in the head nominal, in which the definiteness marker on the head must appear in its phonologically reduced form when in the short possessive construction, (iii) strict adjacency between the head and the specifier, meaning that modifiers of the head must occur outside of the construction, and (iv) the functioning as one prosodic unit. It varies from the CS found in Semitic languages, however, in two ways: first, Bohairic Coptic only allows this CS construction when the head is definite, and second, Bohairic Coptic requires the presence of a particle, `n, between the head and the specifier. While the former is easily attributed to language variation, the latter, I propose, is to be treated as a historical remnant of the synthetic structure found in earlier stages of Egyptian, rather than a genitival preposition.

Misleadingly, the particle `n has been labeled as a genitival preposition because of its occurrence in the short possessive construction. However, when looking at the productive history of Egyptian we find evidence that `n does not have any strictly functional purpose in the nominal phrase of Bohairic Coptic. This is mainly because the Egyptian language is divided into two main stages, Earlier Egyptian, consisting of Old Egyptian and Middle Egyptian, and Later Egyptian, consisting of Late Egyptian, Demotic, and Coptic (Kramer 2012) characterized by a major change from a more synthetic to a more analytic, respectively, pattern in the nominal syntax and the verbal system (Junge 1985). While Earlier Egyptian maintains the VSO order in verbal formations, Later Egyptian follows more of an SVO pattern (Loprieno 1995). In this presentation I trace the short possessive construction through earlier stages of Egyptian and find that the particle `n, within Bohairic Coptic, occurs in phrases that have remained either strictly head initial, as in the case with the short possessive, or optionally head initial, as in the case of adjective phrases, even after the language had shifted from a more synthetic morphological system to a more analytic one.

Works cited Benmamoun, Elabbas (2003). Agreement parallelism between sentences and noun phrases. Lingua 113: 747-764.

Benmamoun, E (2006). The Construct State. Encyclopedia of Arabic Linguistics. Brill Academic Publishers.

Danon, G. (2007). Definiteness Spreading in the Hebrew Construct State. Lingua 118.7: 872-906.

Dobrovie-Sorin, C., (2000). (In)ndefiniteness spread: from Romanian genitives to Hebrew construct state nominals. In: Motapanyane, V. (Ed.), Comparative Studies in Romanian Syntax. Elsevier, Oxford, pp. 177– 226.

Dobrovie-Sorin, C., (2003). From DPs to NPs: a bare phrase structure account of genitives. In: Coene, M., D’hulst,Y. (Eds.), From NP to DP, vol. 2. The Expression of Possession in Noun Phrases. John Benjamins, Amsterdam, pp. 75–120.

Junge, F. 1985 "Sprachstufen und Sprachgeschichte," in Zeitschrift der Deutscher Morgenliindischen Gesellschaft. Supplement 6, 22, edited by W. Rollig, 17-35 (Stuttgart:Franz Steiner Verlag).

Haspelmath, M. (2015) The three adnominal possessive constructions in Egyptian-Coptic: Three degrees of grammaticalization. Egyptian-Coptic linguistics in typological perspec- tive, 261-288. Berlin: De Gruyter Mouto.

Heller, D. (2002). Possession as a lexical relation: evidence from the Hebrew construct state. In: Mikkelsen, L., Potts, C. (Eds.), Proceedings of WCCFL 21. Cascadilla Press, Somerville, MA, pp. 127-140.

McCarthy, J. (1981). A Prosodic Theory of Nonconcatentive Morphology, Linguistic Inquiry 12, 373-418.

Kramer, R. (2012). Egyptian. In Semitic and Afroasiatic: Challenges and Opportunities, Porta Linguarum Orientalium 24. 59-130. Wiesbaden: Harrassowitz.

Loprieno, A. (1995) Ancient Egyptian: A Linguistic Introduction. Cambridge: CUP.

Ouhalla, J. (1994). Verb Movement and Word Order in Arabic. In Verb Movement. Cambridge University Press. pp. 41-72.

Ritter, E. (1991). Two functional categories in noun phrases: Evidence from Modern Hebrew. In Syntax and Semantics 25: Perspectives on Phrase Structure: Heads and Licensing. San Diego: Academic Press.

Ritter, E. (1992). Cross-linguistic evidence for number phrase. Canadian Journal of Linguistics 37(2): 197-218.

Siloni, T. (2001). Construct States at the PF Interface, in Linguistic Variation Yearbook, Vol. 1, John Benjamins, pp. 229-266.

ABSTRACT. Bound yet free: the double life of POc *akin[i] and its Southeast Solomonic reflexes

Proto Oceanic (POc) is reconstructed as having a relatively rich system of valency changing devices. Unlike some other devices, the origin and development of the so called "long" or "remote" transitive suffix *akin[i] is not so well understood. This paper examines the richness of forms and functions of reflexes of POc *akin[i] in Southeast Solomonic (SES) languages in order to better understand the development of transitivity marking systems in these languages, and transitivity in Oceanic languages more broadly. Diachronically and synchronically there is a consensus that POc and many contemporary Oceanic languages had/have a system of marking transitivity that contrasts a short and a long transitive suffix; reflexes of *-i and *akin[i], respectively. The generally accepted view is that each marker introduced/introduces object argument with different semantic roles (Pawley, 1973). This distinction is often reflected in the Southeast Solomonic languages:

Lau 1) moa 'vomit' moa-si-a 'vomit at/on s.t., s.o.' moa-taini-a 'vomit s.t. (out)' (My field data)

However, *akin[i] and its reflexes do not completely fit into a neat paradigm of transitivity with *-i. Despite being both commonly labelled the transitive suffix, the histories of *-i and *akin[i] are essentially not parallel. This is exemplified by the ways in which these two markers are reflected in contemporary Southeast Solomonic languages. Whilst both are used as transitive suffixes, reflexes of *akin[i] also have prepositional uses. *Akin[i] reflexes commonly express the kinds of meaning reconstructable for POc, such as concomitant and instrumental. But in the SES languages we also find a range of other meanings expressed by reflexes of *akin[i]. For example they can (i) denote a purpose, as in 2), or (ii) signal a successful completion of an action, as in 3). Whilst in POc both transitive suffixes had an applicative as well as a causative function, in some SES languages reflexes of *akin[i] appear to have undergone a shift towards being the default marker of causation. Furthermore, reflexes of *akin[i] are often used to derive not only transitive, but also intransitive verbs, as shown in 4).

Owa 2) A aqau ka sitaria rai mata ni maofa. 'It was an aqau tree they split for dancing sticks.' (Mellow, 2014:107)

Lengo 3) a) A Mery te oni-oni-a na kei na. 'Mary is getting her basket ready.'

b) E oni-laghini-a ti na kei na a John? 'Did John finish readying his basket? (My field data) 'Are'are 4) Pita 'e po'o-ta'i. 'Peter turned (around).' (My field data)

Southeast Solomonic is an ideal subgroup within which to consider the history and development of *akin[i], and the transitivity marking systems in Oceanic, in more detail. Whilst *-i always seems to have functioned as a suffix and obligatorily occurred with verbs of the appropriate phonological shape (Evans, 2003), *akin[i] appears to have had a somewhat looser relationship with the verb. Although POc *akin[i] is reconstructed as likely occurring as a suffix with a number of verbs, it appears it likely occurred as a preposition with others (Evans, 2010). The SES data suggest that some originally suffixal uses may have become prepositional; some prepositional forms seem to include initial "thematic" consonants (as shown in 2) and 5)), usually interpreted as reflecting original word final consonants, protected by the suffixes from loss in absolute word-final position which took place in many Oceanic languages. Whilst both *-i and *akin[i] have been associated with introducing arguments with a particular semantic role and thus modifying the meaning of the verb, the semantic contribution of reflexes of *akin[i], in its bound as well as free form, appears to be more robust than that of *-i. One could say that perhaps *akin[i] has always carried a more specific range of meanings on its own than *-i. This could have enabled the speakers to use *akin[i] to express particular meanings that could be relatively independent of the verbal meaning, which could have facilitated the fluid nature of *akin[i] as more or less tightly "bundled" with the verb phonologically. And it is indeed the case that some languages allow variation, expressing the same meaning with either a preposition, as in 5a), or with a suffix, as in 5b), thus continuing the "double life" of the reflexes of *akin[i] into present.

Arosi 5) a) 'ari wou ta'ini-a b) 'a'a-ta'ini-a 'go away from' 'to glide away from' (Fox, 1978:1, 418)

Their existence as bound as well as free morphemes, in addition to the range of meanings and functions attributable to reflexes of *akin[i], suggest that these morphemes may have "double lives" in more than one sense in the SES languages, and that the transitivity paradigm they form with reflexes of *-i may be less neat than previously thought. An in depth examination of the reflexes of *akin[i] in these languages allows us not only to again a deeper understanding of the development of *akin[i] in the history of the SES languages, but also to reconstruct in more detail how the original system of transitivity worked in Proto Oceanic. At a more theoretical level it allows us to gain deeper understanding of processes of change in core aspects of grammar involving clause structure and grammatical relations.

References Evans, B. (2003). A study of valency-changing devices in Proto Oceanic. Canberra, A.C.T: Pacific Linguistics in association with the Centre for Research on Language Change. Evans, B. (2010). Proto oceanic *akin[i]: reconstructing change in progress. In J. Bowden, N. Himmelmann, & M. Ross (Eds.), A journey through Austronesian and Papuan linguistic and cultural space. Papers in honour of Andrew Pawley (pp. 179-196). Canberra: Pacific Linguistics. Fox, C. E. (1978). Arosi dictionary (Revised ed.). Canberra: Pacific Linguistics. Mellow, G. (2014). A dictionary of Owa; a language of the Solomon Islands: Book News, Inc. Pawley, A. (1973). Some Problems in Proto-Oceanic Grammar. Oceanic Linguistics, 12(1/2), 103-188.

ABSTRACT. The development of agreement in Germanic: evidence from a parallel text analysis

Parallel texts, i.e. the same text in various translations or adaptations, have recently been

discovered as a valuable data source in typological research (see e.g. Cysouw & Wälchli

2007, Mayer & Cysouw 2014). The data can be collected very economically, and, more

importantly, it is almost exactly comparable as the text is often nearly identical. In this paper

we try to extend the advantages of parallel text analysis to diachronic research. In a

comparison of a passage of the New Testament (Luke 2;1–2;20), we will explore diachronic

developments and typological differences of agreement in Germanic, both quantitatively and

qualitatively.

A quantitative analysis provides interesting results with respect to what we call the

“pervasiveness” of agreement: How many, and which, instances of agreement (attributive,

predicative, and anaphoric) can be found in the selected passage? The table below illustrates

this for nine Germanic versions (beginning with Old High German and extending into the 21st

century):

[Table included in PDF version of the abstract from the ICHL program index]

In the selected passage, we annotated all agreement relations (i.e., relations between a

controller and a target, in the terms of e.g. Corbett 2006). All relations for which we find

agreement in at least one of the morphological categories gender, number, or person were

annotated for all parallel texts.

In Continental West Germanic, the number of agreement relations is stable or even

increases in the history of German. The Modern Dutch version (Statenvertaling 1977)

displays a similar profile as German, while the number of agreement relations declines

markedly in Afrikaans. In Mainland Scandinavian, it is decreasing substantially in modern

Norwegian. This development is mostly due to the loss of agreement on the verb. In the

history of German, predicative agreement decreases as well, but this is not due to loss of

morphological differences on the verb, but rather to the loss of inflection in predicative

adjectives and participles, which, interestingly, is preserved to some degree in Mainland

Scandinavian. In Afrikaans, no predicative agreement at all takes place. Neither the verb nor

adjectives and participles show any agreement. On the other hand, West and North Germanic

concur on one development: We find an increasing use of attributive agreement morphology,

mostly due to the grammaticalization of the article, and in the case of German, due to more

frequent use of nominal constructions (“Nominalstil”).

On a qualitative level, a parallel text analysis allows to compare cases of “gender

resolution” and “lexical hybrids”. Examples (1) displays a gender resolution context (see

Corbett 2006: 243–253): The plural personal pronoun refers to Joseph and Mary:

(1) OHG: Thô sieM.PL thar uuarun, vvurđun taga gifulte, thaz siu bari

NON: En er .auN.PL varo .ar. .a fyldosc dagar Marie

‘And so it was, that, while they were there, the days were accomplished that she

should be delivered.’ (Luke 2;6)

In both Old High German and Old Norse, the personal pronoun may agree in all three

genders, thus, when referring to two persons of different sex, rules of gender resolution have

to be applied. In this example, German opts for the masculine, whereas Old Norse uses the

neuter plural (cf. Askedal 1973). The inclusion of further versions will allow to determine

whether this is a general difference between West and North Germanic.

Furthermore the analyzed section also contains a “lexical hybrid” (cf. Corbett 2006:

213). The German noun Kind and Scandinavian barn ‘child’ is grammatically neuter but

refers to a male child (namely, the newborn Jesus) in the following passage. All languages in

our study have the possibility to show either formal or semantic agreement, but only

Pennsylvania German and Norwegian Bokmål use the masculine personal pronoun:

(2) New High German: … dazu dasN KindN in der Krippe liegen. Da sie esN aber

gesehen hatten …

Pennsylvania German: … un datt voah’sN kindN am leiya im foodah-drohk. Vo si

eenM ksenna henn …

Norwegian (Bokmål): … og detN lilleN barnNetN som lå i krybben. Da de fikk se

hamM …

‘…and found the baby lying in a manger. When they saw him…’ (Luke 2;16-17)

In sum, a parallel text analysis allows for a very fine-grained investigation of the diachronic

developments in the Germanic agreement system. The inclusion of further versions will

eventually cover the entire Germanic language family, from its beginnings (with Wulfila’s

Gothic version) to all modern languages, including English and Insular Scandinavian. The

analysis of more than 20 versions, representing different languages, language stages and

dialects, will allow to establish a thorough typological profile of agreement and its differing

developments in Germanic.

References

Askedal, John Ole. 1973. Neutrum Plural mit persönlichem Bezug im Deutschen unter

Berücksichtigung des germanischen Ursprungs. Oslo: Universitetsforlaget.

Corbett, Greville G. 2006: Agreement. Cambridge: Cambridge University Press.

Cysouw, Michael & Bernhard Wälchli (eds.) 2007. Focus on parallel texts. Sprachtypologie

und Universalienforschung (Thematic Issue) 60.2.

Mayer, Thomas & Michael Cysouw (2014): Creating a massively parallel Bible corpus.

Proceedings of LREC 2014: 3158–3162.

ABSTRACT. The current study is an examination of the historical development of the argument structure of Spanish psychological verbs of liking, with a special emphasis on the verb gustar, 'to like, to please'. This verb and others of its type deviate from the canonical transitive model, since their formal subject (with the semantic role of theme or cause) does not coincide with the logical subject (with the semantic role of experiencer) (Batllorí 2012). These non-canonical 'subject' patterns tend to occur cross-linguistically with events that are spontaneous or less controlled than those with a canonically marked subject, and with verbs that most commonly indicate a stative meaning, as is the case with psychological states (Onishi 2001). The present study, by analyzing corpus data that range from the 13th to the 19th centuries, aims to understand the argument structure shift through time in psychological verbs of liking, due to the case of the experiencer semantic role evolving from nominative (1a) to dative (1b).

(1) a. Y con una pica, sin sueldo, sin algún entretenimiento ni mando, gustó de ser un particular soldado. 'And with a pike, and no salary or any entertainment or command, he liked being a soldier.' (16th c., Mateo Alemán, Guzmán de Alfarache)

b. Las alhajas le habían gustado mucho a doña Lupe por lo ricas y elegantes. 'Doña Lupe had liked the jewelry very much because it was rich and elegant.' (19th c., Benito Pérez Galdós, Fortunata y Jacinta)

The change in argument structure of gustar is said to have been influenced by great frequency of the dative experiencer verb placer (Batllorí 2012; Elvira 2006; Melis 1999). Quantitative analysis is needed to support this claim and shed further light on the topic. Additionally, the gap left by the lexical loss of the verb pagarse 'to be content, to like', as well as the influence of agradar 'to like', should be taken into account when investigating the surge of gustar as a psychological verb. I approach this study by following a usage-based onomasiological approach (Baldinger 1980), i.e., looking at designations of a particular concept, and investigating by which different expressions the concept can be designated or named. Quantitative aspects of onomasiological structure investigate whether some categories are cognitively more salient than others, namely, an onomasiological study can help understand the factors that determine the choice of one construction over another (Geeraerts 2009). Linguistic factors analyzed in the study are the following: type of construction, individuation features of the theme / cause, grammatical person, grammatical aspect, and time period. The frequency of syntactic constructions is likewise analyzed in order to better understand whether frequency effects at the usage level affect structural weight. The evidence presented in this study comes from one corpus compiled of 16 narrative Peninsular Spanish texts ranging from the early 13th century to the late 19th century, as well as by data compiled from corpora searches from Davies's Corpus del Español (corpusdelespanol.org), Biblioteca Virtual Miguel de Cervantes (cervantesvirtual.com), and Corpus Diacrónico del Español (www.rae.es). In conclusion, preliminary results show more variability between constructions in early Spanish than in later periods: during the 13th c., there seemed to be variation between passive constructions and constructions with a dative experiencer, with the latter construction gaining ground over time. Moreover, the relationships between constructions and time period, grammatical aspect, and individuation features of the theme / cause prove to be significant (p >.001).

References Baldinger, Kurt. 1980. Semantic theory: towards a modern semantics. Basil Blackwell, Oxford. Batllorí, Montserrat. 2012. Diacronía de los verbos psicológicos: una propuesta de entrada léxica. In G. Clavería, M. Freixas, M. Prat, & J. Torruella (Eds.), Historia del léxico: perspectivas de investigación, 341-374. Madrid/Frankfurt: Iberoamericana/Vervuert. Elvira, Javier. 2006. El desarrollo de la construcción biactancial estativa en español. Revista de Historia de la Lengua Española, 1, 45-66. Geeraerts, Dirk. 2009. Onomasiology and lexical variation. In Allan, K., & Brown, E. K., Concise encyclopedia of semantics. Boston: Elsevier. Melis, Chantal. 1999. Los verbos placer y pesar en la Edad Media: la expresión 'impersonal' de las emociones. In F. Colombo (Ed.), El Centro de Lingüística Hispánica y la lengua española. Volumen conmemorativo del 30 aniversario de su fundación, 87-105. Mexico: UNAM. Onishi, Masayuki. 2001. Non-canonically marked subjects and objects: Parameters and properties. In Aikhenvald, A. Y., Dixon, R. M., & Onishi, M. (Eds.), Non-canonical marking of subjects and objects. John Benjamins Publishing.

ABSTRACT. An enthralling yet little known fact about roots for the verb ‘succeed’ in Indo-European is that a large majority of these verbs across several branches of the language family occur with subject-like arguments marked in the dative or accusative case. These verbs form a subset of predicates that take non-canonically (i.e. non-nominative) case-marked subjects, a well-documented phenomenon across the Germanic languages (cf. Barðdal 2004, Barðdal et al. 2016) and other Indo-European languages more generally (cf. Danesi 2014, 2015 on Avestan and Greek, Fedriani 2009, 2014 on Latin, Hock 1990 on Sanskrit, inter alia). Moreover, these same verbs of success find their etymological source in metaphorical extensions of verbs that fall into a variety of semantic fields, including verbs of i) motion, ii) giving, iii) touching/contact, iv) aiming/reaching, v) growing, and vi) luck, among others. Examples of the motion success metaphor (1) and the growth success metaphor (2) are given below.

(1) hṓs hoi dólōi ou proekhṓree (Ancient Greek) since him.DAT craft.DAT not [< motion] ‘since he could not succeed by craft’ (Hdt. 1.205)

(2) him with ne speow (Old English) he.DAT thing not [< grow] ‘he did not succeed at all’ (Beo. 2852)

While semantic change is difficult to analyze in its lack of systematicity and regularity (relative to sound change), the fact that these semantic extensions recur in our dataset is not unexpected, as semantic extensions in general “reflect certain basic metaphorical extensions that all humans can construct, and so it is not surprising that they are found again and again in the histories of languages” (Fortson 2005: 658). Whether the extensions are universal or specific only to Germanic remains to be investigated; that is, the universality of the metaphors that lead to verbal polysemy is a separate issue. In this paper, we investigate instead the relation between a recurrent metaphor, namely, motion success, and the fact that such verbs are constructed with non-nominative subjects in several Indo-European languages. The relationship between argument structure and meaning change is an area that has received considerable attention (e.g. Kemmer & Barlow 2000, Hilpert & Koops 2008, Christiansen & Joseph 2016); in this paper, we explore this relationship via reconstruction of a subset of the verbs with non-canonical argument structure. With regard to success, the largest set of verbs with oblique subject-like arguments used to indicate success are derived from verbs of motion accompanied by a preposition/prefix or adverb. Verbs like Old Icelandic ganga ‘go (+ well)’ and Germanic cognates (from the Proto-Germanic verb *gangan-/gungan-), Latin succedō (< sub ‘under’ + cedō ‘step’), and Greek sym-baínō (< syn ‘with’ + baínō ‘step/go/walk’) form the foundation of our analysis. However, we note that other categories of ‘success’ metaphors produce verbs that take oblique subject-like arguments to a somewhat lesser degree, e.g. “touch success” (e.g. Old Icelandic taka < *takan-/tēkan- + refl. ‘touch’, Latin contingere < cum ‘together’ + tangere ‘touch’), “give success” (e.g. Old Icelandic gefast vel ‘give + refl. + well’, Old Russian ou-dati-sja < ‘at’ + ‘give’ + refl), “grow success” (OHG ge/spuon ~ OCS (ou-)spěti < PIE *speh₁- ‘to succeed, prosper’ [< ‘become fat, ripen’]) and “luck success” (Icelandic heppnast < heppni + refl. ‘chance’, auðnast < auðna + refl. ‘fortune, good luck’, lánast < lán + refl. ‘luck’ ~ Middle Dutch ge/lucken). The accumulation of these data suggests that the connection between metaphorical meaning and non-nominative subject marking is an aspect of Indo-European in general, a point that has received considerable attention more recently, cf. Fedriani 2011, Smitherman 2012, Barðdal & Smitherman 2013, Sigurðardóttir & Eythórsson 2016. We first reconstruct an argument structure construction meaning ‘succeed’ for Proto-Germanic with a dative subject and verb of motion (*gangan-/*gǣjan-; for reconstructed forms, see Kroonen 2013) based on the relatively secure position of having lexical cognates across Western and Northern Germanic of two morphologically related verb forms. From Proto-Germanic and the analogues of such motion success verbs in other Indo-European language branches, we reconstruct a more general schema for Proto-Indo-European, where the form of the verb is unspecified but where the verb’s basic/original meaning is motion and the accompanying argument structure selects for a non-nominative subject. These facts about Proto-Germanic and Proto-Indo-European are modeled in the form of a Constructicon, representing speakers’ knowledge about the interrelation between verbal polysemy and argument structure, namely in the occurrence of two separate constructions: one with the literal meaning of the motion verb and the other with a metaphorical meaning of the motion verb, alongside their respective canonical and non-canonical argument structure. As such, the goal is to contribute to a better understanding of Indo-European syntax and the syntax-semantics interface, since such a reconstruction connects the morphosyntactic patterns of case assignment with metaphorical extensions of verbs.

References Barðdal, Jóhanna. 2004. The Semantics of the Impersonal Construction in Icelandic, German and Faroese: Beyond Thematic Roles. In Focus on Germanic Typology [Studia Typologica 6], 105–137. Ed. Werner Abraham. Berlin: Akademie Verlag. Barðdal, Jóhanna & Thomas Smitherman. 2013. The Quest for Cognates: A Reconstruction of Oblique Subject Constructions in Proto-Indo-European. Language Dynamics and Change 3(1): 28–67. Barðdal, Jóhanna, Carlee Arnett, Stephen Mark Carey, Thórhallur Eythórsson, Gard B. Jenset, Guus Kroonen & Adam Oberlin. 2016. Dative Subjects in Germanic: A Computational Analysis of Lexical Semantic Verb Classes Across Time and Space. STUF: Language Typology and Universals 69(1). Christiansen, Bethany J. & Brian D. Joseph. 2016. On the Relationship Between Argument Structure Change and Semantic Change. Proceedings of the Linguistic Society of America 1(26): 1–11. Danesi, Serena. 2015. Oblique Subjects in Ancient Greek. A paper presented in Kviknes, Norway, in June 2015. Fedriani, Chiara. 2009. The “behavior-before-coding-principle”: Further evidence from Latin. Archivio Glottologico Italiano XCIV: 156–184. Fedriani, Chiara. 2011. Experiential Metaphors in Latin: Feelings were Containers, Movements and Things Possessed. The Transactions of the Philological Society 109(3): 307–326 Fedriani, Chiara. 2014. Experiential Constructions in Latin. Brill: Leiden. Fortson, Benjamin W. IV. 2003. An Approach to Semantic Change. In Brian D. Joseph & Richard D. Janda (eds.) The Handbook of Historical Linguistics, 648–666. Oxford: Blackwell. Hock, Hans H. 1990. Oblique subjects in Sanskrit? Experiencer Subjects in South Asian Languages ed. by M. K. Verma & K. P. Mohanan, 119–139. Stanford: CSLI Publication. Kroonen, Guus. 2013. Etymological Dictionary of Proto-Germanic. Leiden: Brill. Kemmer, Suzanne & Michael Barlow. 2000. Introduction: A Usage-Based Conception of Language. Usage-Based Models of Language, ed. by Michael Barlow & Suzanne Kemmer, 7–23. Stanford: CSLI. Sigurðardóttir, Sigríður Sæunn & Thórhallur Eythórsson. 2016. Two Types of Impersonalization in Icelandic. Paper presented at the 22nd Germanic Linguistics Annual Conference. Smitherman, Thomas. 2012. Metaphors Expressed by Argument Marking Patterns: An Historical and Typological View. Paper presented at Hitches in Historical Linguistics, Bergen, February 23–24, 2012.

ABSTRACT. Ledgeway (in press) explains that split intransitivity “represents the least constrained option and historically is also is the most widespread pattern in Romance which survives today in numerous Gallo-Romance and Italo-Romance varieties”, where unaccusative verbs align with BE and transitive/unergative verbs align with HAVE. Mackenzie (2006: 117) however notes that the membership of certain French verbs to the unaccusative category does not necessarily select their auxiliary, as opposed to Standard Italian for example, so that obvious unaccusative verbs such as manquer ‘to be missing’, exister ‘to exist’, or surgir ‘to arise’ surface in the standard language with the auxiliary avoir ‘have’. Moreover, many sociolinguistic studies suggest that in various parts of the Francophonie native French-speakers employ both auxiliaries, avoir and être, in the spoken language with the 20 or so verbs that, prescriptively, require the exclusive use of être:

(1) J’ai tombé (AVOIR) vs Je suis tombé (ÊTRE) : I fell/have fallen

This phenomenon plausibly points towards a generalisation of the avoir auxiliary in these dialects. This occurs as something notable in Montréal, Québec, Canada (Sankoff & Thibault 1980), Ontario, Canada (Canale, et al. 1978; Willis 2000), the Maritime Provinces, Canada (King & Nadasdi 2005; Balcom 2008), Vermont, USA (Russo & Roberts 1999), New England, USA (Stelling 2011), and Louisiana, USA (Papen & Rottet 1997), and in certain regions of France and Belgium (Remacle (1956), Aub-Büscher (1962), Bouzet (1963), Hendschel (2001), Pooley (1988), and Auger (2003)). Levitt (1979: 26) explains that Old French tended to use être with intransitive verbs, a construction inherited from the perfect tenses of the Latin deponent verbs, which were passive in form but active in meaning: on the model of secutus est ‘he followed’, it is plausible that constructions such as *est venutus becoming il est venu ‘he came/has come’ or *est intratus becoming il est entré ‘he went in/has gone in’ have developed. Until the second half of the 19th century, prescriptive French grammars allowed some motion verbs to be conjugated with both auxiliaries, depending upon whether the emphasized feature pointed to the change itself or the result of change, or whether the change was completed or not (Sankoff & Thibault 1980; Grevisse 2011). In 1675, 17th-century grammarian Ménage (in Levitt 1979) illustrates it in this manner:

(2) Monsieur a sorti (AVOIR): Monsieur has gone out (but he has returned) (3) Monsieur est sorti (ÊTRE): Monsieur has gone out (and has not come back yet)

However, it remains unclear whether the French speakers of the day made such aspectual distinctions. Besides there is also evidence that in the 17th and 18th centuries certain verbs were often freely conjugated with both auxiliaries interchangeably: among others, the verbs demeurer ‘to stay’, descendre ‘to go down’, entrer ‘to go in’, monter ‘to go up’, (re)partir ‘to leave’, retourner ‘to come back’, sortir ‘to go out’, and tomber (Levitt 1979: 26), which all show variation in Sankoff & Thibault’s 1980 paper on auxiliary alternation in Montréal French. But since there is no doubt that the Norman dialects carried over in Nouvelle-France by the settlers in the 16th-18th century acted as a substrate on Québécois French, it is legitimate to question whether this morphosyntactic alternation could have simply been inherited from these dialects, which at the time of the colonisation of Nouvelle-France might have exhibited this variation (semantic or not) between the usage of avoir or être. I recorded 12 native speakers of Montréal French in 2013 (and an additional 48 in spring/ summer 2016) and transcribed the compound tense tokens of the verbs that had shown alternation in Sankoff & Thibault (1980). A comparison of my preliminary results (2013) with those of 1980 reveals that the auxiliary alternation has overall significantly decreased. In this regard, Montréal French appears to be aligning itself with the European standard. Use of avoir correlates with the completion of lower educational degrees and lower socioeconomic classes. Since Ledgeway (in press) shows that among the Romance varieties with a structured 2-aux variation auxiliary distribution can be sensitive to mood (e.g. Romanian), tense (e.g. Sanleuciano and the Salentino dialect of Torre Santa Susanna), person (e.g. Ariellese and Catalan), and argument structure (e.g. Lengadocien Occitan), I test the impact of these factors on my data. While mood/tense (past infinitive) and person (1SG with pronominals) appear promising, argument structure seems to have no influence. Moreover, the Auxiliary Selection Hierarchy (Sorace 2000), which gives results for Standard Italian and Standard French, cannot yet not account for the distribution of the Montréal data. While factors such as lexical effect, frequency, and existence of transitive homonym appear to be at play in this variation.

References Aub-Büscher, G. 1962. Le parler rural de Ranrupt (Bas-Rhin): essai de dialectologie vosgienne. Paris: Klincksieck. Auger, J. 2003. The Development of a Literary Standard : The Case of Picard in Vimeu-Ponthieu, France. In B. D. Joseph, J. DeStefano, N. G. Jacobs & I. Lehiste (eds), When Languages Collide Columbus : The Ohio State University Press, 141–164. Balcom, P. 2008. On the learning of auxiliary use in the referential variety by speakers of New Brunswick Acadian French. Canadian Journal of Linguistics 53.1: 7–34. Bouzet, J. 1963. Syntaxe béarnaise et gasconne. Pau: Marrimpouey jeune, (§55). Canale, M., R. Mougeon & M. Bélanger. 1978. Analogical leveling of the auxiliary être in Ontarian French. In M. Suñer (ed.), Contemporary Studies in Romance Linguistics. Washington D.C.: Georgetown University Press, 41–61. Grevisse, M. 2011. Le Bon Usage: grammaire française refondue par André Goosse. 15th ed., Gembloux: Duculot. Hendschel, L. (2012). Il Croejhete Walone : contribution à une grammaire de la langue wallonne. King, R. & T.Nadasdi. 2005. Deux auxiliaires qui voulaient mourir en français acadien. In P. Brasseur & A.Falkert (ed), Français d’Amérique: approches morphosyntaxiques. L’Harmattan,103–11. Ledgeway, A. in press. From Latin to Romance Syntax: The Great Leap. In P. Crisma & G. Lon-gobardi (eds), The Oxford Handbook of Diachronic and Historical Linguistics. Oxford: OUP. Levitt, J. 1979. Variations in the use of the auxiliairies avoir and être in modern French, Geolinguistics 5, American Society of Geolinguistics, 25–32. Mackenzie, I. 2006. Unaccusative Verbs in Romance Languages. Basingstoke: Palgrave Macmillan. Papen, R. & K. J. Rottet. 1997. A Structural sketch of the Cajun French spoken in Lafourche and Terre-bonne Parishes. In A. Valdman (ed.), French and Creole in Louisiana. NY: Plenum Press, 71–108. Pooley, T. 1988. Grammatical and phonological variation in the working-class French of Roubaix. PhD thesis, University of London. Remacle, L. 1956. Syntaxe du parler wallon de La Gleize.  Tome 2: verbes, adverbes, prépositions. Paris: Les Belles Lettres. Russo, M. & J. Roberts. 1999. Linguistic change in endangered dialects: The case of alternation between avoir and être in Vermont French. LVC 11, Cambridge: CUP, 67–85. Sankoff, G. & P. Thibault. 1980. The alternation between the auxiliaries avoir and être in Montréal French. In G. Sankoff (ed.), The Social Life of Language, Philadelphia: UPP, 311–345. Sorace, A. 2000. Gradients in auxiliary selection with intransitive verbs. Language 76: 859-890. Stelling, L. E. 2011. The effects of grammatical proscription on morphosyntactic change: Auxiliary variation in Franco-American French. Arborescences: revue d’études françaises, n° 1. Willis, L. 2000. “Être ou ne plus être?” Auxiliary alternation in Ottawa-Hull French. Master’s thesis, Ottawa: Université d’Ottawa.

ABSTRACT. The habere + infinitive Latin construction has been identified the root of the modern synthetic, or inflectional, future (SF) paradigm in a majority of Romance languages. It has been overwhelmingly concluded through extensive analyses such as that of Pinkster (1985) and Fleischman (1982) that the emergence of another future paradigm, the periphrastic future (PF), did not enter Latin to replace the SF but rather fulfill different roles concerning register. In examining modern Romance – specifically the Western Romance languages Portuguese, Spanish, Catalan, and French – it is evident that the use of SF and PF rely on increasingly broader semantic contexts. With respect to the spoken language and written language not gaged towards formal literary register, the PF has become significantly more prominent than SF and arguably has entered into a cycle of semantic ascension that could allow for it to replace the SF altogether as discussed in Aaron’s (2014) study of Spanish future semantics. This stage of grammaticalization is not, however, emerging in the all Romance languages that employ both structures. If any comparisons were to be drawn with SF, one would argue that PF is primarily relied upon to express actions that have less finiteness in the future or may express an action that may or may not be realized. This purpose of PF is nearly identical to the conditions of the subjunctive mood. I propose that the evolution of PF and its role in Romance grammars is occurring in dialogue with the historic development of the expressing the future in the subjunctive in each of these languages. Portuguese maintains a regular spoken use of the future subjunctive in everyday speech whereas in Spanish, the future subjunctive has fallen out of use almost entirely save for formal legal documents and in preserved idioms. French and Catalan do not employ a future subjunctive and allow for the uncertainty to be expressed using the present and instead relies heavily on semantic interpretation to express the future aspect. Using the regularity – or absence – of this future aspect in subjunctive in each of the major Western Romance languages, I argue that there is a correspondence with the speakers’ evolved tendency to turn to the PF in order to express future indefiniteness that historically has been reserved for the future in the subjunctive mood. This investigation examines a corpus inventory of the future expressed in the subjunctive mood sorted into data samples representing the sixteenth century until modern day. In Portuguese and Spanish, where speakers have had more explicit emphasis on the future of indefinite actions in the subjunctive, the use of PF is less prominent in indicative future than in French and Catalan where the subjunctive future is expressed using the present. These relationships between PF and future subjunctive in turn affect the dynamic of PF and SF within the indicative mood and the speakers’ tendencies when describing future events with reference to their finiteness.

ABSTRACT. Some fifty years after the controversy on the roots of African American Vernacular English (AAVE) originated, the mood of the debate has calmed down substantially, although the issue itself, the relationship and balance between British English or creole / African elements which have shaped AAVE, still has not been fully resolved. As several contributions to Sonja Lanehart’s recent Handbook of African American Language (OUP 2015) indicate, by now most contenders are willing to accept compromise positions, but differences in attitudes, in detail and in the weight attributed to alternative perspectives remain. I start by briefly surveying the history of this discussion and the resources that have been identified and employed to investigate the issue empirically.

My own contribution (EWS, American Earlier Black English. University of Alabama Press 1989), well received and influential in its time, investigated a wide range of morphological and syntactic features in the transcripts of interviews with 108 ex-slaves from all across the South, compiled by the “Federal Writers’ Project” in the 1930s and 1940s and published as facsimile typescripts in many volumes. In this presentation I return to that source but look at not structural properties but, in line with the topic of the panel, ecological factors which can be extracted from these interviews. The texts constitute an important resource on “history from below”, since they record memories of very old African Americans of antebellum life patterns and their personal experiences before, during and after the Civil War. I screen these texts systematically for statements concerning plantation sizes, interethnic relations and attitudes, life and work experiences, war memories, and post-emancipation settings, and I provide a broadly quantifying documentation and some qualitative sample quotations of realizations of these ecological parameters. Obviously, some care and reluctance is called for in the interpretation of such statements, but I will use them to offer a broad assessment of what antebellum patterns of life and communication may tell us and imply for likely linguistic developments and the evolution of African American Vernacular English.

ABSTRACT. Specialized literature on Grammaticalization recognizes two major directionalities of language change: 1. Downgrading in the cline, also known as Unidirectionality Principle: lexical forms > grammatical forms; free forms > bound forms; etc. Givón (2009): “more than 90% of grammaticalization is down in the cline”; also Haspelmath 2004 and many other authors. 2. Upgrading in the cline: forms having syntactic grammatical status > forms having less grammatical status; bound forms > free forms; univerbation > deverbation, etc. This second cline is also labelled degrammaticalization, antigrammaticalization, pragmaticalization, lexicalization, subjectification, and grammaticalization also.

The proposal of this paper is to show that Grammaticalization is multi-dimensional rather than only down or up. The evidence comes from the Spanish historical syntax. In this language four directionalities are attested.

1. Down in the cline. Free syntax > Morphology. Existential construction ha + ý > Existential verb hay. Full verb haber ‘to have / there is’ + lexical adverb ý (Latin ibi) ‘in that place’ > univerbation, two bound morphemes. 2. Up in the cline. Grammar > Discourse. Full movement verb andar ‘to walk’ > (Inter)Subjective discourse marker of exhortation and confirmation ándale ‘go on / that’s right’ with an obligatory dative clitic. 3. Nor down nor up. Grammar > Grammar. Temporal mientras ‘while’ > Conditional mientras ‘while / if’. This third type of directionality might be considered “no directionality properly” (Fisher 2010), but it is Grammaticalization and it is a type of directionality. 4. Up and down. Grammar > Discourse > Grammar. Full movement verb vaya > Subjective discourse marker of surprise / disgust vaya > Intensive quantifier adjective, vaya + noun.

The paper will be focused on in Directionality 4, because, as far as I know, it has been completely ignored in the specialized literature on language change and on grammaticalization. I will analyze two changes, exemplified in (1) and (2). Both changes show the same diachronic path: an exit from Grammar towards Discourse, and a return to, or reinsertion in, Grammar, but in a very different function from the original grammatical value.

(1) a. Mandaron que Bartolomé de Madrid vaya a Valladolid e se le libren tres mill maravedis e que parta el lunes ‘They ordered that Bartholomew of Madrid goes to Valladolid and that…’ b. ¡Vaya!, no me sabía yo esas mañas ‘Well [lit. go]! I wasn’t aware of those bad customs’ c. Vaya carrazo que te has comprado ‘What a car you have bought yourself’

(2) a. Ya Plinio, en su Historia Natural, dice que las palmas datileras dan en las costas de España un fruto... ‘already Plinio, in his History Natural, says that the palms date give…’ b. ─¿Cómo vivirá esa gente? ─Dizque son artistas ‘How is that these people live? Supposedly, they are artists’ c. los transeúntes viles, amparados por la dizque ley, solían correr tras el ladrón ‘vile people, protected by the supposed-bad law, used to run…’

I will analyze the etymological context, the bridge context, and the shift context of each change. The paper emphasizes that this up and down cline is not uncommon in Spanish. So, this language throws theoretical light on a major and controversial topic in language change: directionality.

The data are culled from the Corpus Diacrónico del Español (CORDE) from Real Academia Española (www.rae.es), and from the Corpus Diacrónico y Diatópico del Español de América (CORDIAM) from Academia Mexicana de la Lengua (www.cordiam.org).

References.

Fischer. Olga. 2010. “On problems areas in grammaticalization. Lehmann’s parameters and the issue of scope”, in Formal evidence in grammaticalization research, A. van Linden, J. C. Verstraete and K. Davidse (eds.), Amsterdam-Philadelphia: John Benjamins, 17-42. Givón, Talmy. 2009. The genesis of syntactic complexity: Diachorny, ontogeny, neurocognition, evolution, Amsterdam-Philadelphia: John Benjamins. Haspelmath, Martin. 2004. “On directionality in language change with particular reference to grammaticalization”, in Up and down the cline. The nature of grammaticalization, O. Fischer, M. Norde and H. Perridon (eds.), Amsterdam-Philadelphia: John Benjamins, 17-44.

ABSTRACT. Formal erosion and functional extension are defining features of grammaticalization processes and as such are conventionally seen as undermining linguistic structures that are based on diagrammatic iconicity between form and function (cf. Haiman 1985). However, I argue that the tendency to drop dummy pronouns in English (i.e., "Looks like ...") consists of maximum functional extension, obligatorification (Lehmann 1985), and maximum erosion. Yet, the result is iconic: the absence of a referent is eventually expressed by the absence of morphosyntactic material indicating a referent. This suggests that iconicity may exist before a grammaticalization process starts and may subsequently be restored again toward the end of the process. If true, this would mean that iconicity is a more central determinant of linguistic structure than usually assumed. Finally, iconicity also provides more semantically grounded support for a model of language change such as the periphrasis-fusion-reduction cycle (Lüdtke 1980). Fusion and reduction do not only occur because of increasing routinization or predictability of the reducing element, but also in part because language users strive for more iconicity.

ABSTRACT. Czech complementizer zda(li) ‘whether, if’: The path of grammaticalization

In this talk, we will examine the change of zda from a modal particle in Old Czech to a present-day complementizer, with special focus on the role of the coalescence of zda with the polysemous enclitic particle li in the grammaticalization process. The importance of this coalescence has not been fully recognized yet due to a lack of quantitative studies. Our analysis draws on corpora of older Czech texts (Staročeská textová banka, DIAKORP and its unpublished part) which enable us to consider frequency effects on the stabilization of form-meaning pairing and also to assess the influence of genre on particular paths of development.

The origin of these particles is not fully clear. There is a consensus that li is older than zda and it developed into an interrogative particle (also known from present Russian; Boye & Kehayov Eds. 2016). According to Bauer (1960), it was further reanalysed as a conditional conjunction when the result of satisfied condition followed immediately (which conforms to the scenario proposed by Iatridou & Zeijlstra 2015).

In the case of zda, there are two ways of diachronic interpretation: one suggests that it relates to the emphatic particle že (<*ghe) and has functioned as a question word from the earliest stage, before it was recruited as a clause linkage marker and lost its place in questions. The other way assumes that zda was primarily an epistemic particle expressing the nuance of hope similarly to PD Czech snad 'perhaps' (sharing historical background with Slovakian azda; Machek 1997) and this particle constituted the basis for further development.

Depending on the perspective chosen, the accounts of how the coalescence happened may also differ: it was proposed (Bauer ibid.) that a special case of using li in multiple questions / complement clauses became the context of transition (cf. example 1 and 2): (1) kaký jest to byl pokrm, kde-li jest vzat 'what was the dish like, where.PART did it come from' (KristA, BOC) (2) Zda o volech péčě jest hospodinu anebo zda-li ne pro ny to praví? lit. 'PART is it about oxen that God is concerned or PART doesn't he say it for us?' in the context of a simile (BiblOl, BOC)

From this point of view, li is treated mere as an emphatic particle that strengthens the interrogative power of zda. As opposed to this, we argue that it was the newer particle zda that reinforced the older item. Firstly, repetitive contexts such as example (1) and (2) are quite rare (estimated on the basis of a random sample, they represent about 5 % of the instances of either main-clause or dependent-clause interrogative structures) and we assume that they do not have the desirable conservation effect (cf. Bybee & Thompson 1997). Secondly, li as a question word has been vanishing throughout the centuries and even in the earliest period (c. 1200 - c. 1400) direct questions account for a maximum of 25% of structures with li (about 65% are conditional clauses and about 10% complement clauses). Thirdly, the examples (1) and (2) should not be seen as analogical because the occurence of li in wh-questions and wh-clauses is not typical, it does not fit in the development towards conditional use.

Given that there are no interrogative particles in PD Czech except the expressive ones (recruited from question pronouns and adverbs), it seems probable that the reinforcement of the interrogative li served a specific purpose. As can be seen from the example (2), the question is expressive and has a persuasive function: either it forces the addressee to answer in its intention (with the opposite polarity) or no answer is expected at all. Such a pragmatic function was identified in more than 90% of cases of direct questions with zda-li or zda in our sample. In addition, the proportion of the compound to the simple form, roughly 80:20, indicates that zda-li as a product of reinforcement/renewal (Lehmann 1995, Traugott & Hopper 2003) was the preferred form in these constructions.

It may seem quite counterintuitive then what corpus data reveal about the further use of these forms: the coalesced form zda-li associated originally with a specific persuasive construction has generally prevailed in use over the simple and more polysemous form zda until the beginning of the 20th century (regardless of the type of construction). After that, though, there is a very strong preference for the use of zda in PD Czech (95-97% of all instances). We will argue that full grammaticalization of former modal particle zda was possible only through a stabilization of its use in a distinctive construction with the particle li: in any other context (purpose, condition, wish...) it had to compete with items that are more frequent. Moreover, the most salient collocations of the complementizer zda in PD Czech ('question', 'to find out', 'to decide' etc.; logDice, SYN2015) clearly show that the grammaticalization has taken place in an interrogative context. The rise of the shorter form at the expense of the longer in the recent period can be ascribed to the accelerating pace of speech.

The transition alone from a question word to a complementizer proceeded first, as our data suggest, in genres and registers that report or imply questions (instructions, administration, law documents). In our case, the loss of expressivity seemed to be a prerequisite of this transition. However, examples (3) and (4) from a book of legal opinions show that the original implication of the persuasive construction could have been maintained in varying degrees and exploited to reproduce someone's perspective.

(3) žádal jest k své straně nálezu, zdali by jáhen také neměl zase jeho žalobě odpoviedati (PrávJihlA; BOC) lit. 'he asked for an opinion for his side whether the deacon should not also respond to his accusation' (4) I jest otázka, zdali Petr z slibu pošlého [...] moci bude zproštěn býti (PrávJihlA; BOC) lit. 'So that's a question whether Petr can be released from his promise'

We will focus on such small steps more closely in an attempt to provide a full picture of the grammaticalization process.

References: Bauer, J. (1960). Vývoj českého souvětí. Praha: Nakladatelství Československé akademie věd. Boye, K., & Kehayov, P. (Eds.). (2016). Complementizer Semantics in European Languages (Vol. 57). Walter de Gruyter GmbH & Co KG. Bybee, J., & Thompson, S. (1997, September). Three frequency effects in syntax. In Annual Meeting of the Berkeley Linguistics Society (Vol. 23, No. 1, pp. 378-388). Hopper, P. J., & Traugott, E. C. (2003). Grammaticalization. Cambridge University Press. Iatridou, S. & Zeijlstra, H. (2015). If diachronically. Paper presented at the 22nd International Conference on Historical Linguistics. Naples, 27-31 July 2015. Křen, M. et al. (2015). SYN2015: A representative corpus of written Czech. Ústav Českého národního korpusu FF UK, Praha. URL: http://www.korpus.cz. Kučera, K. et al. (2015). DIAKORP: A diachronic corpus, version 6. Ústav Českého národního korpusu FF UK, Praha 2015. URL: http://www.korpus.cz. Lehmann, C. (2002, July). Thoughts on grammaticalization. Second, revisited edition. Seminar für Sprachwiss. der Univ. Erfurt. Lehečka, B. et al. (2016). Staročeská textová banka (BOC; Bank of Old Czech), version 2016. URL: http://vokabular.ujc.cas.cz/banka.aspx?idz=STB. Machek, V. (1997). Etymologický slovník jazyka českého. Praha: NLN.

ABSTRACT. Ditransitives have caused much debate as to whether DP DP and DP PP have a common underlying structure and, if so, which derives from which and if the theta-roles differ. The consensus at the moment (e.g. Harley 2002; Beavers 2011) seems to be that the DP DP and DP PP variants have separate structures and certain facts in Old English confirm this. Harley and others also see the light verbs involved as either `Cause-have’ or `Cause-go’ but mark the theta-role of the light verb as Agent. This class of verbs is varied in that some involve an Agent (They deliberately gave this couple obscure directions) or Causer (This gave me a headache) and some an Experiencer (They taught her Chinese) and some a Result (They gave it away to the orphanage). In this talk, I take some of Levin’s (1995) classes of alternating ditransitives and trace them back to Old and Middle English, using the Oxford English Dictionary, Bosworth & Toller’s Anglo-Saxon Dictionary, The Dictionary of Old English, The Middle English Dictionary, the Corpus of Contemporary American English, and the Corpus of Historical American English for data on verbs and the various stages. One of the interesting results that will be the focus of the talk is the causative origin of many of the ditransitive verbs in modern and older stages of English. Examples are bring, the loans divide and pass, and the verb feed, shown here. Fedan is not ditransitive but causative in Old English, as is clear from the inanimate Causer in (2).

(1) He hi fedde mid fætre lynde hwæte and hunige. he them fed with fatter fat wheat and honey `He fed them with fat, wheat, and honey.’ (DOE, Paris Psalter, 80, 15) (2) hu þis land mihte eall þone here afedan. how this land can all that army feed `How this land can feed this entire army.’ (DOE, Peterborough Chronicle 1085a5)

Feed continues to be causative into Modern English, as in (3), next to ditransitive, as in (4). The causative doesn’t allow the adverb deliberately or a second object, shown in (5). Likewise, in the Old English (1) and (6), the second object is introduced by a preposition mid `with’, so the object is an adjunct.

(3) The Earth Can Feed, Clothe, and House 12 Billion People (http://true-progress.com/the-earth-can-feed-clothe-and-house-12-billion-people-306.htm) (4) Had Robert deliberately fed the info to Asa (COCA fiction 2004) (5) *The earth deliberately feeds us great food. (6) And an þane dæg drehten angan erest fedan Israela folc and on that day lord began first feed Israel’s people in þam westenne mid manan, þam heofenlican mete. in the desert with mana, that heavenly food `And on that day the lord started first feed Israel’s people in the desert with mana, that heavenly food.’ (DOE, Napier 1883, no. 44: 'Sunnandæges spell' 70)

The causativity of feed fits its morphologically causative origin. The Gothic verb fôdjan derives from Germanic *fôđ-jan. The talk will also discuss light verbs and particles and show that ditransitives have a dual nature, causative and non-causative.

ABSTRACT. This paper argues within the generative framework (Chomsky (2001, 2008)) that there are two topic positions in clauses with clause-internal discourse particles in Old English (OE): one position is the Specifier of the CP headed by a finite verb, and the other the Specifier of the CP headed by a discourse particle such as þa ‘then.’ I will call these two topic positions the primary and the secondary topic positions, and assume with Kemenade and Los (2006) that the clause-internal discourse particle separates the secondary topic position and the focus position. The clause structure is schematically illustrated in (1).

(1) [CP XP Vfinite [CP YP þa [FP ZP F [TP

In this structure, XP and YP are the primary and the secondary topic, respectively, and ZP is in the focus domain. Let us consider a pair of examples below.

(1) Æðeldryð wolde ða ealle woruld-þincg forlætan ‘Æthelthryth desired to forsake all worldly things’ (ÆLS [Æthelthryth] 31) (2) Þær mihte wundor ða geseon, se ðe wære gehende, hu se wind and se lig wunnon him betwinan, ‘Then might he who was at hand see a miracle, how the wind and the flame strove between them;’ (ÆLS [Martin] 434)

In (1) the indefinite NP ealle woruld-þincg ‘all world-things’ stays at the complement to the infinitive; in (2) the indefinite NP wundor ‘miracle’ occupies the topic domain on the left of ða ‘then.’ The object can be assumed to move across the discourse particle in (2). This type of object movement, driven to satisfy a discourse requirement, is supported by a fact of floating quantifiers. Let us consider example (3).

(3) Ac he gebohte us þa ealle mid his deorwurðan blode of helle wite ‘but he redeemed us then all with his precious blood of punishment’ (WHom 13:45.1242)

This example involves the floating quantifier ealle ‘all,’ which is separated from its associated element us by the discourse marker þa. Ryu (2004) argues that in (4) below, the floating quantifier ealle ‘all,’ which is separated from its head noun by the verbal element, is a focus element.

(4) for þon þe he him wæs ær bæm lað ‘because he was hateful for both of them before’ (Or 139.19-20/Ryu (2004: 69))

If Ryu’s analysis is on the right track, it can be safely said that ealle ‘all’ in (3) is a focus element as well (see also Takami (2001) for floating quantifiers). Also, given that pronouns convey given information, it can be supposed that the movement of the pronoun us leaving the quantifier ealle behind is a kind of topicalization and it is discourse-driven. Suppose the general assumption that the Specifier position of the CP headed by a finite verb is a topic position, it can be safely said that example (3) has two topic elements, one is before the finite verb gebohte ‘redeemed’ and the other is before the discourse particle þa ‘then.’

References Chomsky, Noam (2001) “Derivation by Phase,” Ken Hale: A Life in Language, ed. by Michael Kenstowicz, 1–52, MIT Press, Cambridge, Mass. Chomsky, Noam (2008) “On Phases,” Foundational Issues in Linguistic Theory: Essays in Honor of Jean-Roger Vergnaud, ed. by Robert Freidin, Carlos P. Otero and Maria Luisa Zubizarreta, 133–166, MIT Press, Cambridge, MA. Kemenade, Ans van and Bettelou Los (2006) “Discourse Adverbs and Clausal Syntax in Old and Middle English,” The Handbook of the History of English, ed. by Ans van Kemenade and Bettelou Los, 224–248, Blackwell, Maldon. Ryu, Miyako (2004) “The Comparison between the Quantifier Floating in Japanese and English: Some Observations from a Historical Point of View,” Proceedings of the Fourth Annual Meeting of the Japanese Cognitive Linguistics Association, 263–273. Takami, Ken-ichi (2001) A Functional Analysis of English and Japanese Constructions, Ootori Shobou, Tokyo.

ABSTRACT. It has been repeatedly observed, on the basis of typological ‘big data,’ that there is a worldwide preference for suffixes as opposed to prefixes. This can be explained in several ways. The explanation of this preference is directly relevant to a question highlighted in Good (2008), namely, the relationship between language universals and language change: do synchronic structural universals constrain change, or do diachronic universals, ultimately motivated by synchronic usage factors, give rise to synchronic universals? In this paper, we argue that universally dispreferred structures can and do arise as the result of regular language change, given the right background structures as the particular ‘ecology’ in which change takes place. Specifically, we show that Ancient Egyptian-Coptic (Afroasiatic), shows a long-term diachronic macro-change from mixed suffixing-prefixing to an overwhelming preference for prefixing. The empirical basis for this study is a comparison of ten typologically-significant parameters in which prefixing or affixing is at stake, based on Dryer’s (2013) 969-language sample. We argue that each of the micro-changes implicated in this macro-change are better understood in terms of changes at the level of individual constructions, via grammaticalization, rather than in terms of a broad structural ‘drift.’

ABSTRACT. 1 Introduction In Old French (OF), several factors determine whether a noun occurs with a determiner or as a bare noun, and previous studies have shown that, in OF, bare Ns are favoured by predicate Ns, object Ns, non-count Ns, indefinite Ns, and plural Ns (Boucher, 2005; Buridant, 2000; Carlier, 2007, 2013; Carlier & Goyens, 1998; Dufresne, Tremblay, & Déchaine, to appear; Foulet, 1928/1974; Mathieu, 2009; Moignet, 1976; Stark, 2007, 2008). One factor not been considered in the previous literature is the impact of GENDER. We explore this question on the basis of data from two Anglo-Norman texts (Brendan (B) and Marie de France (MdF)). We show that bare Ns are also conditioned by GENDER, with masculine Ns leading the change.

2 Case, number and gender inflection in Old French OF nominal inflection is conditioned by CASE (subject/object), NUMBER (singular/plural), and GENDER (masculine/feminine). While both D and N inflect for CASE and NUMBER, only D inflects for GENDER.

Singular Plural Subject (li) mur-s (li) mur Object (le) mur (les) murs Table 1: OF masculine nominal paradigm Singular Plural Subject (la) rose (les) rose-s Object (la) rose (les) rose-s Table 2: OF feminine nominal paradigm

In OF, the -s morpheme is ambiguous between case and number. As a result, bare Ns are ambiguous in Old French. This is particularly true of masculine Ns. In the case of masculine bare Ns, in the absence of a determiner, –s-less forms are ambiguous between subject plural and object singular, and forms suffixed with –s are ambiguous between subject singular and object plural. Feminine bare Ns are not ambiguous in the same way, as in most declensions, the suffixe –s only marks plural. This raises the question of whether GENDER is a relevant factor when tracking the emergence of D. 3. Methodology. Our study relies on two Anglo-Norman texts from the 12th century: Le voyage de St-Brendan (c.1106-21) and Les lais de Marie de France (c.1154-1189), both drawn from the syntactically annotated corpus Les voies du français (Martineau, 2008). The data was extracted with Corpus Search. Only argument nominals in subject and object position were considered. All were coded manually for semantic class (mass, abstract, count), definiteness, number, gender and declension. Modern French translations of the target texts were used to determine whether a bare N should be construed as definite or indefinite. This yielded a corpus containing a total of 1650 occurrences (445 in B, and 1205 in MdeF.) For each text, the effect of the linguistic factors was analyzed using the variable rule programme GoldVarb Lion (Rand & Sankoff 1990; Sankoff et al. 2005).

4. Overall results Our study confirms the distribution of bare Ns found in previous studies. Goldvarb analyses show that DEFINITENESS, NUMBER, and GRAMMATICAL FUNCTION remain constant factor groups across the two texts. The study also reveal striking differences. First, the percentage of bare Ns is systematically significantly lower in MdeF. Second, there is a significant increase in range of the factor group SEMANTIC CLASS between the two texts, partly due to an increase of bare mass and abstract Ns, and partly due to the decrease of bare count Ns in MdeF. We take this to indicate that the overall decrease of bare N argument is due to the decrease of bare count Ns. Finally, for count Ns, while DECLENSION is never a significant factor, GENDER emerges as significant only in B, where feminine Ns favor bare Ns.

5. The source of the gender effect in Brendan In order to circumscribe the source of the gender effect found in B count Ns, we considered the distribution of bare Ns according to DEFINITENESS, NUMBER, and GENDER. As shown in Table 3, feminine Ns favor bare Ns in only two contexts--plural definites and singular indefinites--which identify these two domains as the locus of the gender effect. In the other two cases, there is no gender asymmetry either because the change is near completion (plural definites), or because it has not been introduced yet (plural indefinites are categorically bare).

Definite Indefinite Singular Plural Singular Plural Masculine 8% 7% 55% 100% Feminine 7% 33% 93% 100% Table 3 : Percentage of bare count Ns in Brendan

6. From Brendan to MdeF As shown in Table 4 and 5, the gender effect found in B’s singular indefinites and plural definites disappears in MdeF. In the case of singular indefinites,Table 4 shows that while masculine Ns are relatively stable between B and MdeF, feminine Ns show a sharp decrease from being almost categorically bare in B to less than 50% in MdeF. The change is abrupt and happens equally in both object and subject position. The consequence of this change is the loss of the gender effect in MdeF singular indefinites. Moreover, masculine Ns shows an important S/O asymmetry in both texts (a difference of 37% in B, and of 29% in MdeF). This S/O asymmetry is considerably reduced in feminine Ns (a 12% difference in B ; an 8% difference in MdeF).

Brendan MdeF Subject Object Subject Object Masculine 22% 59% 23% 52% Feminine 83% 95% 38% 46% Table 4 : Bare singular indefinites in B and MdeF

Brendan MdeF Subject Object Subject Object Masculine 3% 18% 10% 7% Feminine 33% 33% 33% 8% Table 5 : Bare plural definites in B and MdeF

The situation is quite different when we look at the distribution of plural definites, where the loss of the gender effect is due to the relatively sharp decrease of feminine bare Ns in object position (from 33% in B to 8% in MdeF). In contrast, in subject position, masculine and feminine Ns are relatively stable. We take this to indicate that the gender contrast is neutralized in object position.

7. Conclusion The above describes a general pattern of change for count Ns: masculine Ns lead the change in all contexts (definite plurals, indefinite singulars, and definite singulars), and show more S/O asymmetry than feminine Ns. This may appear puzzling, as there is no a priori reason why nominal classes should be a relevant factor, nor why masculine should lead the change. We argue that the gender asymmetry is a by-product of the fact that unlike feminine determiners, masculine determiners are generally inflected for case in OF: determiners are thus introduced (in subject position) to disambiguate syncretic forms in the masculine paradigm. Finally, the fact that DECLENSION does not emerge as a significant factor confirms the crucial role of GENDER as the locus of change in OF.

References. Boucher, P. (2005). Definite Reference in Old and Modern French: The Rise and Fall of DP, in Montserrat Batllori et al. (eds), Grammaticalization and Parametric Variation. New York: OUP, 95-108. Buridant, C. (2000) Grammaire nouvelle de l'ancien français. Paris: SEDES. Carlier, A. (2013) Grammaticalization in Progress in Old French: Indefinite Articles, in D. L. Arteaga (ed.), Research on Old French: The State of the Art. Dordrecht: Springer. (2007) From preposition to article: The grammaticalization of the French partitive, Studies in Language 31, 1-49. Carlier A. (2013). Grammaticalization in progress in Old French: indefinite articles. In Arteaga (Ed.), Research on Old French: The State of the Art. Dordrecht: Springer. Carlier, A., & Goyens, M. (1998). De l'ancien français au français moderne: régression du degré zéro de la détermination et restructuration du système des articles. Cahiers de l'Institut de Linguistique de Louvain-la-Neuve, 24(3-4), 77-112. Dufresne, M., Tremblay, M., & Déchaine, R.-M. (to appear). The variable use of determiners in Old French and the DP hypothesis. Linguistic Variation, Tortora (ed.)). Foulet (1928/1974) Petite syntaxe de l'ancien français. Paris: H. Champion. Martineau, F. (2008). Un Corpus pour l’analyse de la variation et du changement linguistique. Corpus, 7, 135-155. Mathieu, E. (2009). From local blocking to cyclic Agree: the role and meaning of determiners in the history of French, in Ghomeshi et al. (eds.), Determiners: variation and universals, John Benjamins, 123-157. Moignet, G. (1976) Grammaire de l’ancien français. Paris : Klincksieck. Rand, D., & Sankoff, D. (1990). GoldVarb 2.1 : a variable rule application for Macintosh (Version 2). Centre de Recherches Mathématiques, U. de Montréal. Stark, E. (2007). Gender, number, and indefinite articles: About the ‘typological inconsistency’ of Italian. In Stark et al. (Eds.), Nominal determination, Benjamins. (2008). Typological correlations in nominal determination in Romance. In Müller & Kinge (Eds.), Essays on Nominal Determination: From morphology to discourse management

ABSTRACT. This paper explores the genesis and development of two-way Number systems with a focus on plural-marking. Haspelmath (2005: 142) points out that cross-linguistically, plural-marking varies along at least two dimensions: animacy and optionality. With respect to the animacy hierarchy (Comrie 1989, Corbett 2000, Croft 2003, Haspelmath 2005, among others), speakers make number distinctions for certain NP types, the most important of these being the contrast between animate noun (especially nouns with a human referent) and inanimate nouns. With respect to optionality, Haspelmath (2005) distinguishes between non-occurrence, optional occurrence, and obligatory occurrence of plural markers. These two dimensions, when combined, will give us six values, summarized in Table 1.

Table 1: Cross-linguistic summary of plural markers: animacy and optionality (from Haspelmath 2005: 142) Feature Occurrence No nominal plural 28 languages Plural only in human nouns, optional 20 languages Plural only in human nouns, obligatory 39 languages Plural in all nouns, always optional 55 languages Plural in all nouns, optional in inanimates 15 languages Plural in all nouns, always obligatory 133 languages TOTAL: 290 languages

Table 1 shows a clear synchronic snapshot of variability in the marking of plurality on nouns. within languages that mark nominal plurality. The objective of this study is to examine whether it can be translated into a unidirectional diachronic pathway. Number-marking has developed in various ways. Given the variability in the development of number marking, the first question that we address is the following: what are some of the common paths that languages which did not express nominal plurality took to develop plural-marking? One of the common ways is through the semantic extension of morphological or lexical elements originally expressing some notion of Number. For example, distributive expressions, as in Southern Paiute (1), imply the existence of more than one entity, and over time, may come to express plurality.

(1) qa'nɪ / qaŋqa’nɪ house/house.DIST ‘house’/ ‘house scattered here and there’ > ‘houses’ Source: Sapir (1930-1:258)

Similarly, the Modern Korean plural-marking morpheme -tul is also argued to have evolved from a Middle Korean nominal form signifying ‘and other similar things’ (Park 2010; Lee & Ramsey 2011). The second question that we investigate is the following: when plural-markers come into a language, how are they integrated? Are they observed first on animate nouns, as implied by a diachronic translation of Table 1? If so, do they eventually extend to other semantic categories as well? Indeed, we find that in some languages which develop plural-marking, in particular when plural markers originated as distributives (2a), as in Cayuga, it is limited to human/animate nouns, as in (2b).

(2) a. kanyo:Ɂ kanyoɁshó̦:Ɂo̦h ‘wild animal’ 'game’ (distributive function) b. eksá:Ɂah eksá:Ɂahshó̦:Ɂo̦h ‘child’ ‘children’ (plural function) Source: Mithun (1988:228-9)

Similarly, in Korean, the new plural-marker started out being used exclusively on animate nouns as predicted, and, over time, extended its use to other nominal categories, including inanimate nouns, abstract nouns, and mass nouns (MacDonald 2014). Corbett (2000: 57) proposes an implicational ranking of semantic classes for languages that do not mark Number on all count nouns. He claims that if a language marks Number on nouns referring to inanimates, then it must also mark it on animate nouns, further implying number marking on nouns referring to humans. We could expect that Corbett’s Animacy Hierarchy is mirrored in diachrony, such that if Number is introduced as a category, it will appear first with human nouns, then be extended to other animates, then to inanimates. This appears to be supported in Korean (MacDonald 2014). We also return to the question of optionality raised above. We can also predict that when a noun system acquires plural-marking, it must go through a stage of optionality before becoming obligatory. In other words, we predict that one development pathway, if not the sole such pathway, corresponds to Table I above, beginning with no Number-marking at all and gradually expanding beyond animacy and optionality to generality and obligatoriness. So far, the trajectory of Number-marking in Korean and other languages appears to follow this pathway. Our presentation will focus on both animacy and optionality with respect to the diachronic development of two-way Number systems.

References Comrie, B. (1989) Language universals and linguistic typology. 2nd Edition. Oxford: Basil Blackwell. Corbett, G. (2000) Number. Cambridge: Cambridge University Press. Croft, W. (2003) Typology and universals. 2nd Edition. Cambridge: Cambridge University Press. Haspelmath, M. (2005) Occurrence of nominal plurality. In B. Comrie et al. (Eds.), The World Atlas of Language Structures. Oxford: Oxford University Press: 142-5. Lee, K.-M. & S. Ramsey (2011) A history of the Korean language. Cambridge: Cambridge University Press. MacDonald, D. (2014) The changing status of plural-marking in Korean. Proceedings of the 30th NorthwestLinguistics Conference. Simon Fraser University. Vancouver, Canada. Mithun, M. (1988) Lexical categories and the Evolution of Number Marking. In M. Hammond & M. Noonan (Eds.), Theoretical Morphology: Approaches in Modern Linguistics. San Diego: Academic Press, Inc.: 211-235. Park, S. (2010). Grammaticalization of plurality marker tul. Seoul International Conference on Linguistics. Seoul, Korea. June. Sapir, E. (1930-1931) The Southern Paiute Language. Boston: American Academy of Arts and Sciences.

ABSTRACT. This presentation concerns the reconstruction of adjective attribute marking system in Proto-Saami. I will suggest here a new explanation regarding the formation of adjective attribute forms in Proto-Saami. Traditionally it is assumed that the adjective attribute marker in Proto-Saami has been a suffix -s (e. g. Rießler 2006). It is still used in many adjectives. Based on my studies in oldest Saami adjectives I challenge this perception.

The Saami (also written in forms Sami or Sámi) languages are a northwestern branch of Uralic language family, other branches of which include for example Finnic (including Finnish and Estonian) and Ugric (including Hungarian). There are nine still living Saami languages spoken in northern parts of Scandinavia and Finland, and in Kola Peninsula in Russia. The languages are closely related and they date back to Proto-Saami. All the Saami languages are relatively small, the amount of Saami speakers is around 25 000 altogether. The biggest language is North Saami with about 20 000 speakers.

The Uralic languages are agglutinative languages, and the nouns have a rich case system. The adjectives have historically not agreed with their heads in attributive position. In the western part of Uralic languages (i. e. Finnic and Saami) adjective agreement systems have developed.

Saami developed a system for adjective attribute marking (examples 1 a–c), and it is a complex and dialectally varying system in modern Saami languages. The adjectives have two forms, a predicative (also called basic form), and an attribute form. The attribute form does not agree in case or number with its head.

1) a. Stuorra dállu big.ATTR house.NOM

b. Stuorra dálu-s big.ATTR house-LOC

c. Dállu lea stuoris. house.NOM is big.NOM

The adjective attribute marking system in Saami has been much studied (e.g. Nielsen 1933; Átányi 1943; Rießler 2006; 2011), but the focus of the study has been in the origin of the -s, which occurs in many adjective forms, not in the adjective attribution system in Proto-Saami. Thus, such an attribute system has not been reconstructed.

The material of this study consists mainly of Juhani Lehtiranta’s Common Saami Vocabulary (Yhteissaamelainen sanasto, 1989). I have collected all the adjectives that exist in the material (altogether around 140 words). In addition I have used the etymological database of the Saami languages (Álgu). The attribute forms have been collected from following languages; South Saami, Lule Saami, North Saami, Inari Saami, and Skolt Saami. Together they form a rather comprehensive basis for the reconstruction of the Proto-Saami adjective attribute marking.

The approach to the Proto-Saami adjective system in this presentation is etymological. I have divided the material according to the etymological background of the words. The oldest adjectives (i.e. words that date back beyond the Proto-Saami) in Proto-Saami suggest that the attribute marker has varied due to the phonological structure of the adjective. Derived adjectives have, and still have, their own attribute form. Non-derived adjectives have a system that has based on the phonological structure of the word (the examples are shown in North Saami orthography). It seems that -s has been more an adjective marking suffix than only an attribute marking suffix. The words ending with -a (examples 2 a–b) have the old form in attributive position, and the nominative form has suffix -s: ođas ‘new.NOM’.

2) a. Biila lea ođas. b. ođđa biila car.NOM is new.NOM new.ATTR car.NOM

3) a. Bárdni lea njoahci. b. njoazes bárdni boy.NOM is slow.NOM slow.ATTR boy.NOM

The old adjectives ending with -i show more variation, but a clear tendency is that the attributive form has gained the suffix -s (examples 3 a–b). This division suggests that already in Proto-Saami the adjective attribute marking system has been rather complex and has been based on the phonological structure of the words.

Abbreviations

ATTR attribute INESS inessive LOC locative NOM nominative

Literature

Átányi, István 1943: A lapp melléknevek alakpárjainak kérdéséhez. – Nyelvtudományi közlemények 51: 307–355. Lehtiranta, Juhani 1989: Yhteissaamelainen sanasto. Mémoires de la Société Finno-Ougrienne 200. Helsinki: Société Finno-Ougrienne. Nielsen, Konrad 1933: A note on the origin of attributive forms in Lapp. – Liber semisaecularis Societatis Fenno-Ugricae. Mémoires de la Société Finno-Ougrienne 67: 296–307. Helsinki: Suomalais-Ugrilainen Seura. Rießler, Michael 2006: Om samiskan -s attributiva adjektivform. – Andrea Amft and Mikael Svonni (ed.): Sápmi Y1K – Livet i samernas bosättningsområde för ett tusen år sedan: 135–150. Umeå: Umeå universitet. Rießler, Michael 2011: Typology and evolution of adjective attribute marking in the languages of northern Eurasia. Academic Dissertation. Leipzig: Institut für Linguistik.

Electronic sources:

Álgu = Sámegielaid etymologalaš diehtovuođđu – The etymological database of the Saami languages [online-database]. Helsinki: Kotimaisten kielten keskus. November 2006. Constantly updated. http://kaino.kotus.fi/algu/ Referred to 1.12.2016

ABSTRACT. Recent research has revealed that the formerly assumed very strong correlation between the presence of antipassive constructions and ergative alignment doesn't stand close scrutiny (e.g.Janic 2016). While only a comprehensive survey of the co-occurrence both of antipassive construction and ergative alignment in a language is suitable to either confirm or disprove such a hypothesis, much work still needs to be carried out in order to capture the full diversity of phenomena related to the kinds of processes that are involved in the derivation of antipassives and similar constructions. For example, it is unclear whether the relation between, e.g., accusative alignment and antipassives is such that the alignment type blocks off the formation of antipassives altogether. Another possible scenario is that antipassives and similar constructions rise and disappear in short cycles and are by nature not stable enough to become as strongly entrenched as, say, equivalence sets in alignment systems, which are known to be rooted in the area.

A case in point is the sudden appearance of morphologically derived antipassives within the history of Middle Welsh (cf. ex. 1), which seems to have disappeared towards the end of the Middle Welsh period.

(1) Ac ual byd yn ym-warandaw a llef yr erchwys, ... and as be.prs.hab.3sg ptcl antip-hear.vn with cry art pack ‘And as he was listening for the cry of the pack, ...’

In a similar way, the cognate of MW ym- in Old and Middle Irish, imm-(a- n ), develops reciprocal, collective simultaneous action and transitive functions within a narrow time frame of approx. 200 years, before disappearing at the end of the Middle Irish period. The most important kind of morphosyntactic changes caused by adding imm-(a- n ) to a verbal stem relates to argument realisation. Theses changes occur on morphosyntactic devices that are commonly involved in Old and Middle Irish argument selection, namely pronominal infixes, verbal endings, argument marking, and verbal agreement. Most importantly, the use of imm-(a- n ) with a finite verb does not reduce or increase the number of arguments of the predicate, i. e, the inventory of roles is preserved (dubbed “operational diatheses” in Kulikov 2011, p. 384). However, the derivation with imm-(a- n ) causes various patterns of argument demotion: – passive-like demotion of the a argument alone into a PP headed by do ‘to’ (2), – combined passive- and antipassive-like demotion of the a (PP headed by do ‘to’) and p (fri ‘against’) argument simultaneously (3), – and of the s argument (4)

(2) immá-ríacht do ⁊ Protésalús comrac dessi imm-reach.3sg.pst to.3sg and P. battle.acc two.gen ‘both he and Protésalús reached the battle of them both’

(3) ní-ma-nacige do frim-sa hi richiud nime ná fri harchangliu ná fri hapstlu neg-imm-see.3sg to.3sg to.1sg.emph in heaven heaven nor to arch-angels nor against apostles ‘he won’t see me nor the arch-angels nor the apostles in heaven’

(4) ima-sisedar doib imm-come_to_a_stand.prt.3sg to.3pl ‘... and they came to a stand’

Based on the the analysis of the repeated emergence and subsequent rapid loss of antipassive-like structures in Insular Celtic we conclude that the presence of such constructions is possibly not tied to a basic alignment type. We tentatively suggest that its infrequency is due to its instability over time, provided that the phenomenon has not merely been overlooked.

ABSTRACT. This paper explores how the oblique form came to replace the nominative in subject complements in Danish. In Modern Danish, pronominal subject complements invariably occur in the oblique form: "det er mig" ‘it is me.OBL.’, "det er dig" ‘it is you.OBL.’, "det er ham" ‘it is him.OBL.’ etc. This was not always so. In Old Danish, pronominal subject complements would occur in the nominative form, so at some point a change took place. The first attestations of subject complements in the oblique form are found in a manuscript from 1425, and the latest attestations in the nominative are seen in the early 18th century. Surprisingly, how the replacement of the nominative by the oblique form took place is vastly understudied. Insofar the change is treated at all in works on Danish, it is typically only touched upon and not fully explored, the primary concern being the timing issue and not the circumstances or the motivations for the replacement (e.g. Dyrlund 1895: 202-204; Mikkelsen 1911: 243; Brøndum-Nielsen 1965: 68; Brink & Lund 1975: 664). However, when any motivation for the change is mentioned (e.g. Falk & Torp 1900: 24), it takes its departure in the word order hypothesis originating from Jespersen (1894, and later), whereby in the course of Middle English and Early Modern English, new principles of distribution of case forms emerge; somewhat crudely put, according to this hypothesis, constituents to the left of the verb occur in the nominative and constituents to the right occur in the oblique form. Later studies on pronominal case distribution in English and French also point to the word order hypothesis, and they (as indeed does Jespersen himself) qualify the hypothesis by proposing that several other factors were also at play in the change, such as analogy with other syntactic constructions, phonetic resemblance with other case forms, and – what is of more interest to the matter at hand – text pragmatic features (Foulet 1935, 1936 on French; Quinn 2005 on English). As regards the later point, Foulet (for French) and Quinn (for English) have this in common: they both include text pragmatic features in their treatment of the subject complement. Quinn (2005: 242-248) consistently refers to the subject complement as the ‘focus pronoun’, and Foulet (1936: 45) points to the subject complement being ‘the most important word of the phrase’. This idea of text pragmatic features being influential on the choice of case form ties in with recent studies on the distribution of pronominal case forms in Modern Danish. According to Jørgensen (2000), in Modern Danish, pronouns in the nominative form have affinity to the theme, and pronouns in the oblique form have affinity to the rheme (cf. also Heltoft 1997). Taking inspiration from this, I propose that the replacement of the nominative with the oblique form in subject complements is embedded in a general change of the functions of case forms in Danish. Whereas Old Danish is a ‘case rich’ language and case forms are used to keep arguments/participants morphologically distinct (cf. Sigurðsson 2006), Modern Danish is a ‘case poor’ language with different principles for the distribution of case forms, these principles being partly syntactic, partly text pragmatic. In the course of history, this reorganization of what determines the distribution of case forms has manifested itself as a series of changes, the end result (so far) being the situation presented by Jørgensen. As for the change specifically relevant to the case form in subject complements, this could be outlined as follows: Stage 1: At this stage, Danish is still a case rich language. Morpho-syntactic features are more important than text pragmatic ones as regards the case system. Therefore the nominative may be used by participants with affinity to non-focus (e.g. subjects) as well as focus (e.g. subject complements). Similarly, the oblique form may also be used by participants with affinity to non-focus (e.g. quirky subjects, ethical dative, indirect object) as well as focus (e.g. direct object). Stage 2: Text pragmatic features gain influence on the distribution of case forms: Thus, the nominative becomes the form specialized for non-focus participants (subject). The oblique form becomes the generalized form to be used in all other functions regardless of text pragmatic status, non-focus (e.g. ethical dative, indirect object) as well as focus (e.g. direct object AND subject complement).

References: Brink, Lars & Jørn Lund (1975). Dansk rigsmål, lydudviklingen siden 1840 med særligt henblik på sociolekterne i København. København: Gyldendal. Brøndum-Nielsen, Johs. (1965). Gammeldansk Grammatik V: Pronomenerne. København. Dyrlund, F. (1895), anmeldelse af Kristian Mikkelsen Dansk Sproglære med sproghistoriske Tillæg 1894. Arkiv för nordisk filologi XI, 180-208. Falk, Hjalmar & Alf Torp (1900). Dansk-norskens syntax i historisk fremstilling. Kristiania: Aschehoug & Co. Foulet, Lucien (1935, 1936). L’extension de la forme oblique du pronom personnel en ancien francais. Romania 61, 257-315, 401-464, og Romania 62, 27-91. Heltoft, Lars (1997). Hvem opslugte hvo. Et bidrag til beskrivelsen af det danske kasussystems udvikling. I: F. Lundgreen-Nielsen, M.A. Nielsen & J.K. Sørensen (red.): Ord, Sprog oc artige Dict. Festskrift til Poul Lindegård Hjorth. København: Universitets-Jubilæets danske Samfund & C. A. Reitzel, 227-256. Jespersen, Otto (1894). Progress in language - with special reference to English. 2nd ed. by James D. McCawley, John Benjamins 1993. Jørgensen, Henrik (2000). Studien zur Morphologie und Syntax der festlandskandinavischen Personalpronomina - mit besonderer Berücksichtigung des Dänischen (= Acta Jutlandica LXXV:2, Humanities Series 73). Aarhus: Aarhus University Press. Mikkelsen, Kr. (1911). Dansk ordföjningslære med sproghistoriske tillæg. København: Hans Reitzel 1975. Quinn, Heidi (2005). The Distribution of Pronoun Case Forms in English. John Benjamins, Amsterdam. Sigurðsson, Halldór Ármann (2006). The Nom/Acc alternation in Germanic 2006. Hartmann, J. & L. Molnarfi (eds.): Comparative Studies in Germanic Syntax. John Benjamins, 15-50.

ABSTRACT. The Basque language is usually described as a language with no split ergativity (Trask 1979: 386): there is no split based on aspect/tense, nor on the animate/inanimate character of the noun. In this presentation, I will argue that Basque can indeed be described as a split ergativity language. The nature of this split lies on morphological number: plural phrases follow a neutral alignment, whereas singular phrases follow a typically ergative pattern. This would be a A-SO/ASO system in Dixon’s notation (1994: 109).

ABSTRACT. Grammaticalization studies of individual items have mainly focused on the evolution of grams and their properties. The discussion concerning loss of grammatical meaning and/or associated morphosyntactic properties have revolved around the proposed unidirection of the grammaticalization processes, and subsequently around how degrammaticalization should be understood and identified (Norde 2009, 2012, Börjars & Vincent 2011). Losses of grammatical properties in lexemes have until recently therefore mainly been studied from this perspective, setting out to answer the question whether a change qualifies as degrammaticalization or not. The question of whether the semantic and syntactic properties so often found to co-evolve in grammaticalization studies also disappear at the same time, has hardly been addressed at all. This study focuses on a case clearly not describable as degrammaticalization: the loss of future meaning of the Swedish auxiliary vilja. It investigates what happens with the raising properties associated with grammatical meaning in Swedish auxiliaries when that grammatical meaning is lost, but the form continues in use, albeit with meanings located on the more lexical stretch of the lexical-grammatical continuum. Swedish vilja has several cognates in the other Germanic languages such as Danish ville, English will, German wollen, Dutch willen, Icelandic vilja Norwegian vilja, tracable to the PIE root *wel-, attested in the sense of ‘want, desire’. Some of them have developed modal and temporal meanings while others hardly show any development from the original lexical meaning (Börjars & Vincent in prep). Swedish vilja did actually develop a temporal, future meaning and was used as an auxiliary in Old Swedish (Björkstam 1919, Lagervall 2014). This meaning has however been lost, and vilja is now used primarily in the original sense of ‘want, desire’, which has been in use continuously. Therefore, this cannot be degrammaticalization, since a layer of grammatical meaning was developed, used and lost, while the older layer ‘want, desire’ never went out of use. The corpus providing material for the study is an historical dictionary, The Swedish Academy Dictionary. Approximately 12 000 examples of vilja were found, dating from 1526 to 2000, and the 993 examples of vilja with inanimate subjects were extracted from them, following Zaenen et al. 2004 in the categorization of the animacy of subjects. The study focuses on inanimate subjects for semantic and syntactic reasons. Future irrealis is inherent in verbal complements to verbs or modals with the meaning ‘desire’ or ‘intention’, and it is therefore impossible to tell if a clause with an animate subject and vilja has a sense of prediction, or is purely intentional. To find more easily identified examples of prediction, the study was therefore limited to clauses with inanimate subjects. The aim of finding out if vilja had distribution in contexts taken to indicate subject raising also motivated the exclusion of animate subjects, as these contexts consist of passive clauses with inanimate subjects, or clauses with the expletive subject det in Swedish. Results indicate that the future meaning of vilja was frequently used during the 16th and 17th centuries, less so during the 18th century and hardly at all during the 19th century (Bylin sub.). Vilja with future meaning was possible to passivize over, with no change of meaning of the clause, a common indicator of subject to subject raising. This is no longer possible today, as the subject in a passivized clause with vilja now is animate and clearly shows a semantic relation to vilja (Laanemets 2013:176 ). It was also used with expletive subjects, both true expletives and wheather-its in affirmative and non-affirmative contexts. Examples are few, but the intriguing modern Swedish restriction on expletives with vilja as grammatical in non-affirmative contexts only, seems to have come about during the 19th century. The results thus seem to have two implications. First, semantic and syntactic properties seem to depend on each other not only when auxiliaries grammaticalize, but also when they vanish. Second, the loss of certain raising properties, and the subsequent restrictions on viljas distribution with expletive subjects might shed some diachronic light on the synchronic interpretation of the construction as an indicator of subject-to-subject raising (Postal 1974 13–14, 369–370, Eide 2005:175, Boye 210:75).

References Björkstam, Harald. 1919. De modala hjälpverben i svenskan I. Tör, lär, mon, må, måtte och vill. Lund: Håkan Olssons boktryckeri. Boye, Kasper 2010. Raising Verbs and Auxiliaries in a Functional Theory of Grammatical Status. In: K. Boye & E. Engberg-Pedersen (ed.) Language usage and language structure. (Trends in Linguistics. Studies and monographs. 213.) Berlin: De Gruyter Mouton. 73–104. Bylin, Maria. Sub. Futural betydelse och hjälpverbssyntax hos vilja 1526–1950: Uppkomster och förluster – går sambandet mellan semantik och syntax åt båda hållen? Börjars, Kersti & Vincent, Nigel. 2011. Grammaticalization and directionality. In: B. Heine & H. Narrog (ed.) The Oxford handbook of grammaticalization. Oxford: Oxford University Press. 163–176. Börjars, Kersti & Vincent, Nigel. In prep. Modelling step change: the history of WILL-verbs in Germanic. Eide, Kristin Melum. 2005. Norwegian Modals. (Studies in Generative Grammar. 74.) Berlin: Mouton de Gruyter. Laanemets, Anu. 2013. Passiv i moderne dansk, norsk og svensk. Et korpusbaseret studie af tale- og skriftsprog. Tartu: University of Tartu press. Lagervall, Marika. 2014. Modala hjälpverb i språkhistorisk belysning. Göteborg: Institutionen för svenska språket, Göteborgs universitet. Norde, Muriel 2009. Degrammaticalization. Oxford: Oxford University Press. Norde, Muriel 2012. (De)grammaticalization and (de)subjectification. In: J. Van der Auwera & J. Nuyts (eds.) Grammaticalization and (Inter-)Subjectification. Brussel: Royal Flemish Academy of Belgium for Sciences and the Arts Contactforum. 37-63. Postal, Paul M. 1974. On Raising. Cambridge, Mass.: The MIT Press. The Swedish Academy Dictionary, http://g3.spraakdata.gu.se/saob/ Zaenen, Annie, Carletta, Jean, Garretson, Gregory, Bresnan, Joan, Koontz-Garboden, Nikitina, Tatiana, O´Connor, M. Catherine & Wasow, Tom. 2004. Animacy Encoding in English: Why and How. I: Proceedings of the 42nd Annual Meeting of the Association for Computational Linguistics (ACL ’04), Workshop on Discourse Annotation. S. 118–125. Barcelona.

ABSTRACT.  Abstract: Present perfect is considered as a construction which refers to as a marker of prior events that are INCLUDED within the overall period of the present in contrast with the preterit that marks situations assigned to a past occurrence which is CONCLUDED from the present. The aim of this paper is to establish the development of present perfect in Persian. My intention is to analyze this construction in Persian in terms of development, functions, meanings, and usage. For this purpose, I deal with this construction from the perspective of form, meaning, and its specific uses in Persian both diachronically and synchronically. From a synchronic point of view, the resultative, experiential and current relevance meanings of this construction can be covered by the compound verb form. From a diachronic point of view, this form has undergone formal and functional changes during three main eras of Persian: Old Persian, Middle Persian, and Modern Persian, however; we can find similarities among these forms as well. Fundamental to the present study are four assumptions: First, contemporary present perfect construction has been shortened in form in the majority of contexts especially in spoken texts, usually with a change in syllable stress, although their functions remain constant. Second, unlike English, past adverbials may accompany present perfect in current Persian. Third, the findings lend support to the claim that evidentiality in many contexts if not all is a part of their functions. Finally, in its present use, Persian speakers delete the copula from this construction both in spoken and in informally written texts on the one hand and use an indicative form in place of perfect in many cases on the other.

Introduction: It is generally believed that undertaking investigations on present perfect constructions, even if studied by several researchers including (Comrie, 1976; McCoard, 1978; Dahl, 1985; Givon, 1982; Bybee (1994), Mahmoodi Bakhtiari, 2002; and Taleghani, 2008), Dahl (2011) and Jugel (2016), could hardly be described as plain sailing. This circumstance, in the main, arises from the complex structures of perfect, which, according to many linguists, are bound up with tense since they do show the features of tense at least as relative tenses and are indubitably tied to aspect. Such a dual relationship becomes even more appreciable when it comes to focusing on Persian since, few if any, investigations have been conducted in this regard so far. The data reported in this study provides convincing evidence that present perfect in modern spoken Persian has been reduced from compound to simple in terms of form. It has a significant interaction with the past tense and perfect aspect on the one hand and exhibits various time reference ambiguities on the other. It is therefore of interest to further investigate the possible developments of the present perfect from Old Persian to its current use and describe some salient properties of these constructions.

Present Perfect versus preterit: Concerning the distinction between the English simple past and the present perfect in terms of location in time, (Comrie 1985,p. 78) argues that the perfect is not distinct from the past" since both state an occurrence in the past.”

Present perfect with an inclusion of the present moment can be the main reason why some linguists including McCoard (1978) considered its semantic meaning as indefinite past, unlike preterit which is taken as definite past, as:

1-She has eaten lunch. (Which implies that perhaps she does not need to eat food at the moment,) versus:

2-She ate lunch. (Which does not necessarily mean she is not hungry.)

With a typological view, as is well known, and unlike Persian, present perfect in English does not tolerate the presence of any temporal expressions explicitly referring to the past. Such an expression in any Persian past imperfective sentences triggers the use of either simple past tense or present perfect depending on the speaker's intention and verb forms. Moreover, the choice between present perfect and simple past in English depends too much on how relevant the situation is considered to be for the present moment by a speaker and this is a subjective judgment. But when this intuition of relevance is very clear, especially in the presence of adverbial expressions like "already", the correlation between past imperfective and present perfect is very clear: 

  3- He has already eaten lunch.

Literature review: Concerning present perfect in Persian, some research has been performed that can be viewed in 3 perspectives:

a. General research regarding the form of this construction

b. Those research about the form of this verb in Iranian languages

c. Specific research on its form in standard Persian

From the first group, there can be found some information in traditional grammar books in general and the history of Persian language written by Khanlari (1970). Other grammar books have somehow repeated data regarding the form and meaning of this verb.

From the second type, I can refer to Gharib (2003) and Kalbasi (2003). The former investigates the present perfect in Soghdi, a north-eastern Iranian language. His finding shows there has been a kind of present perfect like what we presently have in the north of Iran, made of the main verb participle and helping verb of /dār/ or  /δā r-/, as in:

/wit-u δārām/     I have seen     /wit-u δāre/          you have seen          /wit-u δārt/            s/he has seen

The latter investigates present perfect in different Iranian dialects. She studies this form in 60 dialects and claims that there are 31 different forms for present perfect.

Next reference I refer to for this section is Shokri (1999) who studies present perfect in different dialects of Mazandaran province, north of Iran. She finds out that there is a kind of present perfect with the verb “have” in that area that is used with transitive verbs only, as in Rashti dialect:

 /be-xând-ə dâr-am/ I have read.

The last reference I examined is Rezayati and Ghayoori (2012) who claim there has been found 3 types of present perfect in Shahnameh:

a. Past participle + a present form of “be”:  /basteam/ I have tied

b. Past participle without helping verb:  /nebeshte/ (S/he has) written

c. Past participle + a form of verb “istadan”:  /shodasti/ you have become

Present Perfect form in different stages of Iranian languages: From remained sources of the past, including petroglyphs, we can distinguish three main eras for the development of present perfect form. This classification focuses mainly on the development of the form of this verb rather than the historical period, however; since the language change happens inevitably during the time, we have to refer to the history of each period for our purpose.

  a. Old Iranian period started about 1000 BC till the end of Achaemenid era, 400 BC, including Old Persian and Avestan,

b. Middle Persian, about 400 BC till Sassanid Iran, – 900 AD, almost beginning of Islam

c. New Persian, since 900 AD up to now.

Present perfect in Old Persian: According to Khanlari (2003) and Sadeghi (1978) there was no compound form of Present perfect (like the one in contemporary Persian, (participles +helping verbs) in Old Persian and Avestan. In that time, tenses, moods, and aspects were identified by a combination of prefixes, roots, and endings.  Therefore, present perfect in those languages were formed by duplicating of roots with a change in the first syllable (Abolghasemi (2004: 150), for instance:

/dā-darəs-a/ = “I have seen”   /da-dā-þa/ = “You have created”     /vā-vərəz-ōi/ = “He has exercised”.

In addition, a kind of present perfect with a long construction, was used in Old Persian for state events and also for completed events in past, from objective adjective plus ending –ta, plus linking verb of “budan”(be)  and “agent”.

Bagheri (2015: 96) claims that for describing past events there were different tools such as types of past tenses. Moreover, using the past participle followed by a nominal in possessive case is another tool for describing the past events. For instance, instead of 3, sentence 4 is used:

/ima, taya manā krtam/                            

in ra mᴂn kᴂrdᴂm

this.D.O. I do-P-1SG

3- I did this.                                                                   

in kᴂrdeye mᴂn ᴂst

this done I is

4- This is what I have done.

Based on her claim, the past participle in Old Persian was mostly made from the weak root followed by the inflectional ending of “-ta”.

She claims that in the remained scripts from Old Persian, it is obvious that using past participle for events occurred in the past, is preferred than using simple past, since plenty of phrases such as the following can be seen: (p. 95)

ima: tya: manā: katam     this that I  doPP           5- This is what I have done.

She believes that past stem of Persian verb system is the natural continuation and developed form of Old Persian past participle. Given this form is used very commonly for events happening in the past, in the Middle Persian, it is used as the stem of the past verb. She claims that the only difference between Old Persian PP and the Middle Persian past stem is the phonolo gical development.

Accordingly, the vowel /a/ was deleted from the end of suffixes “-ita” and “-ta”.  (p. 196)

Present perfect in Middle Persian: Basically, Middle Persian languages are divided into West Branch and East Branch. There are two important languages representing Middle Persian: Pahlavi and Parti, which are from the west branch of Iranian languages taken from Old Persian. They are more important since they have a direct relationship with New Persian and different Iranian dialects. Their importance arises from this fact, as Sadeghi (1979) claims, that Dari Persian and New Persian are their continuations.

From remained Parti texts, one can understand that present perfect had been used very rarely, however; available evidence shows its form as an objective adjective (past participle) plus present forms of auxiliary išt- (literally, stand and be). Examples from (Rezayee Baghbidi, 2002):

/izγad ištām/:  I have escaped   /izγad ištēh/:   You SG have escaped    /izγad ištēd/:    S/he has escaped

In Pahlavi (Middle Persian), the transitive and intransitive present perfect were formed differently. For intransitive form past root of the verb plus ēst- (stand, be) were used, like the following:

/raft ēstēm/     I have gone.  /raft ēstē/        You SG have gone.   /raft   ēstēd/       S/he has gone

A difference in form between Parti and Pahlavi, two Middle Persian languages can be seen as for the latter, the separate agentive subject appear at the beginning of the present perfect form while such markers cannot be seen in the former.

Present perfect in Dari Persian: Khanlari (2003) believes that Dari Persian is a member of New Persian languages which historically begins from the 8th and 9th century with Arabic handwriting.

Present perfect in this version of Persian had been made of past root of the verb plus unpronounced /ah-/ and enclitics of “be” verb. This ending (ah-) is a replacement for ēst of Middle Persian. The conjugation of the verb xordæn for example is as follows:

xorde æm       I have eaten xorde ee        You SG have eaten.      xorde æst       S/he has eaten…

These forms are used in written texts. In contemporary spoken Persian, the last vowel of past participle (e) is usually deleted and the form resembles past simple form of the Persian verbs, however; the stress pattern changes while the speaker means the present perfect function of the construction.

   Present Perfect form in Contemporary Persian: The perfect in contemporary Persian appears to be constructed by so-called “shortened infinitive „ (masdar-e-morakham), which is the same as the past stem of any Persian verb and the unpronounced bound morpheme of ha, an adjectival suffix,which is henceforth called past participle , as the shared element of all perfect constructions, plus the existential verb form of astan (to be),from Early New Persian, reduced to the enclitic present or past copula. It is a verbal suffix that marks person and number so that it makes the omission of the subject possible since the verb forms are finite. For third person singular, however, it can be retrieved especially in the written language because the subject is absent and the verb undergoes zero morpheme. As an example, the conjugation of the verb kardan (to do) for present perfect is as follows: karde-am, (I’ve done)                                            karde-im, (We‘ve done)…

Persian perfect and stress pattern: There is a phonological rule for stress pattern of verb forms in Persian that in positive forms, the stress usually falls on the last syllable of the first constituent. Accordingly, for simple past forms, stress must fall on the last syllable of the past stem. If this rule is violated and stress falls on the last syllable of the whole form, then the construction plays the function of perfect. As an example, /ʹdidᴂm/ (I saw) is past simple but /didʹᴂm/ (I have seen) is taken as present perfect. Samei (1995) adopts this rule, restricted it for the two verb forms of /daʃtᴂn/, (to have), and /xordᴂn/ (to eat), only but it cannot be confined to these cases.

  Current relevance and Resultative functions: Shariat (1988) defines present perfect as an event that started from the past but the same action or its result can be witnessed at present, for instance:

6- The window has opened. (and it is still open)

He claims that present perfect has two meanings of (soboot) and (hodoos).

His terminology is different from what is used in linguistics these days, however, he identifies two important functions for present perfect:

 By “soboot” he means an action that happened in the past but has not finished yet:

Bᴂhāreh zire derᴂxt istādeh ᴂst    Bahareh under tree stand.-3SG.PP-copula

7- Bahareh has stood under the tree.

The two functions he identifies remind us of the current relevance function of perfect (soboot) together with resultative function (hodoos).

Conclusion: We proposed four arguments to present an explanation for the syntactic and semantic function of present perfect in Persian. The function and semantic properties of the Persian Perfect are associated with the anterior or perfect aspect. This category typically expresses features both associated with tense and aspect. It is, however, difficult to assign Persian perfect as a tense or as an aspect since it shows both the features of tense on the one hand and the characteristics of an aspect on the other.

The data yielded by this study provides convincing evidence that present perfect in Persian, as a category for stating the prior event, and in case of stative, depicting a present state usually of the subject, resulting from a change of state, has the tendency towards the amalgamation of preterit and the present perfect with a differentiation of stress shift in some cases but with the same function that perfect provides. It is not limited to perfect forms, but it is manifested by indicative forms specifically when the degree of remoteness of the event is not too much.

The second argument is that the Persian present perfect, with two crucial time points of the situation located in the past and the present moment with equal importance, enjoys a hierarchical structure rather than a horizontal pattern adopted by other scholars. We have modified theory of time event and the perfect in order to adapt it to Persian perfect with two main features of actual state of relevance to the present time on the one hand, and having reference to a prior event on the other hand, and consider its current relevance, as the main function on the top of a hierarchical model and resultative, experiential, evidential functions and perfect of recent past as the entailment from the main function which are arranged horizontally. Concerning perfect in Persian, with a common element of the participle, we have a principal function which stands on the top of a hierarchy and there are other sub-functions that fall under this principal. The principal function is “current relevance” since without it, we cannot logically have resultative, experiential, hot news functions and so on. We can assume the concept of entailment for this situation. Current relevance entails resultative and other functions.

The third argument is that Persian admits using past adverbials such as “yesterday” or “last week” with present perfect forms whereas they do not include the present moment. To some extent, the feature of admission of past adverbials refers to the behavior of the verb and partly to the orientation of the event structure and its relevance to the present.

The fourth argument is that in Persian perfect bear the feature of evidentiality. Indirect information, usually described under the label of evidentiality in many contexts, is depicted in perfect progressive more obviously than other forms of perfects.

 References:

Abolghasemi, M. (2004). History of Persian language. SAMT publication: Tehran.

Bybee, J., Perkins, R. & Pagliuca, W. (1994). The Evolution of Grammar: Tense, Aspect, and Modality in the Languages of the World. Chicago IL: University of Chicago Press.

Comrie, B. (1976). Aspect, An introduction to the study of verbal aspect and related problems. Cambridge: Cambridge University Press.

 Cmorie, B. (1981). “On Reichenbach’s Approach to Tense”. Papers from the Regional Meetings, Chicago Linguistics Society, 17, 24-30.

Dahl, E. (2011). Tense and aspect in Indo-iranian part 2: In language and Ilnguistics: Compass 5:5, 282-296.

Gharib, B. (2003). Present perfect and transitive pluperfect in Soghdi and their similarities with some new Iranian dialects. Translated by: Mitra Faridi. Journal of Dialectology: No. 2, PP: 56-65.

Givon, T. (2001). Syntax: An Introduction, Vol I. John Benjamins.

 Jackson, A. V. W. (1892). Avesta Grammar, W. Kohlhammer, Stuttgart.

Jugel, T. (2016). The perfect in Iranian languages.  International conference on: The function and semantics of the perfect in Indo-European Languages. Uppsala, Uppsala University.

Kalbasi, I. (2003). Present perfect in Iranian dialects. Journal of Dialectology: No. 2, PP: 66-89.

Khanlari, P. N. (2003). History of Persian language. 2nd Edition, Farhang-e-nashr-e Now: Tehran.

Mahmoodi Bakhtiary, B. (2002). Tense in Persian, Its Nature and Use. Lincom     Europa.

McCoard, R. W. (1978). “The English Perfect: Tense-Choice and Pragmatic Inferences”, North-

Radden, Gunter, and Rene Dirven.  (2007). Cognitive English Grammar. Amsterdam/Philadelphia: Benjamins.

Rezayati, M. & Ghayoori, M. (2012). Specific form of present perfect in Shahname. Journal of Fonoon-e Adabi, No.6, PP 1-12.

Rezayee Baghbidi, H. (2002). Parti Grammar, Tehran: Persian language and literature Farhangestan.

Sadeghi, A. (1978). The development of Persian language. Iran Azad University: Tehran.

Shokri, G. (1999). Present perfect in Mazandarani dialects. Tehran: Farhangestan Letter, No. 4.

Taleghani A. H. (2008). Modality, Aspect and Negation in Persian, University of Toronto, John

ABSTRACT. In this paper I combine methodologies from corpus linguistics with an experimental-like setup more affiliated to psycholinguistic research. The resulting methodology allows us to gain more insight into cognitive motivations of language use in speakers from the past, and how such individual motivations may trigger change at the communal level. Consequently the method also makes it possible to better assess their similarity to present-day speakers (the uniformitarian principle). One such cognitive motivation thought to be relevant in the early stages of grammatical constructionalization (grammaticalization) (Traugott & Trousdale 2013) is covered by the evasive concept of ‘extravagance’ (the desire to talk in such a way that one is noticed, Keller 1994, Haspelmath 1999). The methodology and its usefulness for operationalizing this notion of extravagance is tested on a case study on the Early Modern English extension of the [BE Ving]-construction. While its function of encoding ongoingness or ‘progressiveness’ had already been established for past tense adverbial clauses by late Middle English (Petré 2016), its use in the present tense at this point remained essentially stative. Only in the Early Modern English period did [BE Ving] extend its progressive function to present tense main clauses. Throughout Early Modern English, [BE Ving] remained in competition with the simple present in expressing ongoing situations, as seen from the equivalence of (1) and (2).

(1) For it is not my Superiours now that I am speaking of. (Richard Baxter, 1681) (2) But I now speak of the Reason of it as a Covenant in genere. (Richard Baxter, 1673)

On the basis of extensive data from the EMMA corpus (Early Modern Multiloquent Authors), I will provide evidence that [BE Ving] in this novel use indeed correlates with various features that are assumed to reflect extravagant language use. Interestingly, a comparison of two generations that are among the earliest to use this innovation with some frequency, suggests that there might be a shift from extravagance being signalled by coercion (of the still stative semantics of [BE Ving] into a progressive reading) to its being an entrenched characteristic of the construction itself (prior to its presumable wearing out in later stages).

References Haspelmath, Martin. 1999. Why is grammaticalization irreversible? Linguistics 37: 1043-1068. Keller, Rudi. 1994. On language change: The invisible hand in language. London & New York: Routledge. Petré, Peter. 2016. Grammaticalization by changing co-text frequencies, or why [BE Ving] became the ‘progressive’. English Language and Linguistics. 20(1). 31-54 Traugott, Elizabeth C. & Graeme Trousdale 2013. Constructionalization and constructional changes (Oxford Studies in Diachronic and Historical Linguistics). Oxford: Oxford University Press.

ABSTRACT. Recomplementation or complementiser doubling is a phenomenon whereby two, seemingly identical, finite complementisers are licensed in the same clause separated by an XP. As the scientific interest in the left periphery has increased (Rizzi, 1997), so has the interest in recomplementation. The phenomenon, rather than being deemed substandard as in previous times, is now seen as a clear indication of the complexity of the CP-area in the languages under scrutiny. So far, not much research has been done on the type of elements that may be sandwiched in the recomplementation structures, apart from stating that they are topic-like elements.

Recomplementation is attested in a range of Romance languages, such as Old Spanish (Fontana, 1993; Uriagereka, 1995; Wanner, 1998) and Modern Spanish (Villa-García, 2015), Old French (Salvesen 2014) and Child French (Roehrs & Labelle, 2003), Old Tuscan (Paoli, 2003) and certain dialects of Southern Italy (Ledgeway, 2005), and Catalan colloquial language (González i Planas, 2010). In Modern Spanish, the sandwiched XP has been argued to reside in a specTopic-position (Villa-García, 2015) in an articulated left periphery. The same is probably true for Old Spanish:

mas dixo que pues que ell era emperador.que tomaua but said comp since comp he was emperor comp took

otros .v. annos pora acabar aquello que començara. other five years to finish that which had.begun

“But he said that since he was emperor, he took another five years to finish what he had begun” (Estoria de Espanna)

In (1), the phrase in italics is an adjunct clause that belongs to the complement clause, and both preceding and following the adjunct clause, marked by bold face, there is an instance of the complementiser que, the first one is thought to reside in Force and the second in Topic.

In my talk, I analyse the sandwiched elements in a small corpus of Old Spanish recomplementated sentences taken from the ISWOC/PROIEL database (Jøhndal, Haug, & Nøklestad, 2007-2016), and I show that the sandwiched element must be an A-topic (Frascarelli & Hinterhölzl, 2007) to trigger recomplementation. On the basis of this observation, and the fact that also scene setting elements trigger recomplementation, I argue that the sandwiched element contains a restrictive feature, and that it is this feature which triggers the apparition of the second complementiser.

The findings imply both that the topic marker que in Old Spanish was productive in the language of the 13th and 14th under these specific circumstances, and that the seemingly identical phenomenon that is found in spoken Spanish today (which is optional) is somewhat different in syntactic nature, since it appears not to be restricted to A-topics.

Cited works: Fontana, J. (1993). Phrase Structure and the Syntax of Clitics in the History of Spanish (PhD), University of Pennsylvania. Frascarelli, M., & Hinterhölzl, R. (2007). Types of topics in German and Italian. In K. Schwabe & S. Winkler (Eds.), On Information Structure, Meaning and Form (pp. 87-116). Amsterdam: John Benjamins. González i Planas, F. (2010). Cartografia de la recomplementacióen les llengües romàniques. (Master), Universitat de Girona. Jøhndal, M. L., Haug, D., & Nøklestad, A. (2007-2016). Parallell corpus from the project Pragmatic Resources in Old Indo-European Languages. from University of Oslo Ledgeway, A. (2005). Moving through the left periphery: the dual complementiser system in the dialects of Southern Italy. Transactions of the Philological Society, 10(3), 339-396. Paoli, S. (2003). COMP and the left-periphery: comparative evidence from Romance. (PhD), The University of Manchester. Rizzi, L. (1997). The Fine Structure of the Left Periphery. In L. Haegeman (Ed.), Elements of Grammar (pp. 281-337): Springer Netherlands. Roehrs, D., & Labelle, M. (2003). The Left Periphery in Child French. In J. Quer, J. Schroten, M. Scoretti, P. Sleeman, & E. Verheugt (Eds.), Romance Languages and Linguistic Theory 2001 (pp. 279-294). Amsterdam: John Benjamins. Salvesen , C. M. (2014). Le complémenteur que et la périphérie gauche: analyse diachronique Syntaxe & Sémantique, 15, 47-80. Uriagereka, J. (1995). An F position in Western Romance. In K. É. Kiss (Ed.), Discourse Configurational Languages (pp. 153-175). Oxford: Oxford University Press. Villa-García, J. (2015). The Syntax of Multiple-que Sentences in Spanish. Philadelphia: John Benjamins. Wanner, D. (1998). Les subordonnées à double complémentateur en roman médiéval. Atti del XXI Congresso Internazionale di Linguistica e Filologia Romanza, I, 421-433.

ABSTRACT. Background: Korean, like other Eastern Asian languages, is considered to be a classifier language. A predominant property of classifier languages is that they lack plural-marking (Allen 1980, Chierchia 1998); however, Korean poses an interesting problem for this claim since it appears to have an optional plural-marker -tul (Kang 1994, Baek 2002, Kim 2005). Some researchers propose that it is not a plural-marker at all, but rather a marker of information structure marking distributivity (Park 2008) or focus (Song 1975). Korean -tul has been studied extensively; however, there is little consensus as to its distribution or function. My research takes a new approach to the analysis of -tul by examining the historical development of this morpheme.

Research Question: The goal of this paper is twofold. First, I hope to shed light on the modern-day uses of -tul through investigating its past. Second, I propose a comparative analysis of North and South Korean dialects which leads me to conclude that -tul functions differently in these two dialects of Korean. The question this paper addresses is whether the development of -tul in South Korean is a language-internal change, a language-external change (through extensive contact with English, a mass-count language), or some combination of language-internal and language-external changes.

Methodology: The South Korean study is based on a corpus analysis which investigates the historical use and development of -tul. The study comprised 125 newspaper articles which covered, approximately, a 100-year period (1924 – 2011). I analyzed the articles for data relevant to the distribution of -tul, the number of instances of -tul in the article, the type of nouns which -tul attached to, as well as cases where a plural interpretation was clear, but -tul was not used. The North Korean study was also a corpus analysis, which investigated the distribution and use of -tul in newspaper articles written by 5 North Korean writers. This study preliminarily investigated 10 newspaper articles, all of which were written in 2014. The same criteria which were applied to the South Korean data were also applied to the data from North Korea.

Analysis: What the South Korean research found was that in both the 1924 and 1946 data, there were very few cases of –tul found in the data sets. The cases that were observed were limited to use with human nouns and the use of –tul did not extend to animate non-human nouns or inanimate nouns. In the early data, –tul did not seem to be functioning as a plural marker in these early newspaper articles. Instead, -tul seemed to be functioning as a way to place emphasis or focus on the noun to which it attached. In the later data (from 1970, 1995, and 2011), -tul is used more frequently and its use is extended to additionally include non-human nouns, and later concept-denoting abstract nouns. This is illustrated in Figure 1 below The distributional patterns that were found in the latest data from South Korea differed significantly from the North Korean data. While I only analyzed North Korean data from 2014, there was far fewer instances of -tul being used and the use of -tul appeared to be limited to certain categories of nouns; namely human, animate, and a few inanimate nouns. This pattern matched most closely to what was observed in the South Korean data in the 1970s.

Discussion: Lee (1989), and Baik (1992) have both made claims that the use of modern -tul as a plural-marker was borrowed from English. Baik (1992:25) claims that during the period 1965 – 1985, which fits in well with the period where we see an increased use of -tul, there was a period of intensive contact between English and Korean. While a borrowing hypothesis is possible, it is important to remember that Korean -tul existed (and was used on plural nouns) long before contact with and influence by the English language. However, it is also important to note that in the newspaper articles from 1970 and later (the claimed period of extensive language contact between English and Korean) we see a substantial increase in the use of -tul. It is also worth noting that while South Korean dialects did undergo heavy influence from English, the same cannot be said for North Korean dialects. This is mirrored in the data as North Korean newspaper articles do not exhibit the extensive use of -tul seen in South Korean. My presentation will focus on both the North Korean and South Korean newspaper data and will outline to use and distribution of -tul in both data sets. The discussion will then focus on the question of the changing status of -tul from an internal/external change debate and from a historical linguistic perspective.

[NOTE: The figure does not appear in this document, please see pdf version] Figure 1: Instances of -tul in South Korean newspapers (by year)

References Allan, K. (1980). Nouns and countability. Language 56: 541-67. Baek, M.-H. (2002). Handkwuke pokswu uymi yenkwa. [A study on Korean plural senses.] Discourse and Recognition. 9(2): 59-78. Baik, M. J. (1992). Language shift and identity in Korea. Journal of Asian Communication 3(1): 15-31. Chierchia, G. (1998). Reference to kinds across languages. Natural Language Semantics 6:339-405. Corbett, G. (2000). Number. Cambridge: Cambridge University Press. Kang, B.-M. (1994). Plurality and other semantic aspects of common nouns in Korean. Journal of East Asian Linguistics 3: 1-24. Kim, C.-H. (2005). The Korean plural marker tul and its implications. PhD Dissertation (University of Delaware). Lee, S. (1989). The subversion of Korean. English Today 20: 34-37. Park, S. (2010). Grammaticalization of plurality marker tul. Seoul International Conference on Linguistics. Park, S.-Y. (2008). Plural marking in classifier languages: a case study of the so-called plural marking –tul in Korean. Toronto Working Papers in Linguistics 28, 281-295. Song, J.-J. (1997). The so-called plural copy in Korean as a marker of distribution and focus. Journal of Pragmatics 27, 203-224.

ABSTRACT. Expanding the canon: The unrecognized role of alignment in the typology of suppletion

Corbett’s canonical approach to morphological typology offers a rigorous way to examine “the challenging topic of suppletion” (2007: 9). This paper presents data that favor expanding his approach to include a new parameter: alignment. I also present a semantic basis for non-alignment. The Hungarian data expand our understanding of the typology and development of suppletion, specifically regarding what types of suppletion are possible and the ways in which suppletion can arise. The Hungarian copula van exhibits overlapping suppletion — where separate lexemes share suppletive forms (e.g., Spanish fui preterit of ser “be” and ir “go”; contrast non-overlapping suppletion, where forms are not shared, as in English went, belonging only to go). Hungarian shows that cases of overlapping suppletion must be further divided into two types: aligned and non-aligned. Certain forms of van are identical to forms of lesz “become”. Some of the overlapping forms are aligned, i.e., they belong to the same categories in both verbs (e.g., the infinitive lenni). Other forms, however, are not aligned. For example, the present forms of lesz are also the future forms of van. Thus leszek means “I become” or “I will be”. This case of overlapping suppletion differs from more typical cases, where the shared forms are exponents of the same category. (For the existence of multiple forms in some cells, see Thornton 2011 on overabundance.) For Corbett, canonicity in suppletion centers on how regular the semantics are and how irregular the morphology is. Accordingly, his tenth criterion states that non-overlapping suppletion is more canonical than overlapping suppletion and subdivides overlapping suppletion according to directionality, which concerns whether shared forms are regular in one of the paradigms (2007: 26). Thus Latin nullius exhibits directional overlapping suppletion with nullus “none” and nemo “nobody”, whereas Galician fun “I was” ~ “I went” illustrates nondirectional overlapping suppletion between ser “be” and ir “go”. Corbett views directional overlapping suppletion as less canonical than nondirectional overlapping suppletion. The Hungarian case is a mixture of aligned directional overlapping suppletion and non-aligned directional overlapping suppletion. Given the significance of semantic regularity in Corbett’s approach (and in others’ treatments), aligned overlapping suppletion should be considered more canonical than non-aligned overlapping suppletion, since non-alignment obscures semantic patterns. Overlapping suppletion is too rare to establish how alignment and directionality relate to each other, but the centrality of semantic regularity in suppletion favors considering non-alignment to be less canonical than alignment. Thus I propose this two-part reformulation of Corbett’s statement:

Criterion 10a: non-overlapping > non-directional overlapping > directional overlapping Criterion 10b: aligned overlapping > non-aligned overlapping

The semantic relationship between becoming and being maps onto present and future, respectively: if something wet is presently becoming dry, it will be dry in the future (if the process is completed). Other languages reflect this relationship. Elsewhere within Uralic, Mari exhibits a similar connection: “The … verbs ulam ‘is’ and lijam ‘becomes’ … are interchangeable in many contexts, but the present of lijam often refers to the future” (Kangasmaa-Minn 1998: 231). The Proto-Indo-European root *bhuh2- “become” has reflexes in copulas (e.g., Spanish fui ‘I was’, English be). The identification and explanation of the various types of semantic relationships found in suppletive paradigms allows us to go beyond the traditional assumption that suppletive verbs reflect a straightforward replacement of forms from one root with equivalent forms from another. Non-aligned overlapping suppletion in Hungarian not only illustrates the value of the canonical approach for morphological typology but also reinforces the capacity of semantic analysis to make sense of the seemingly intractable phenomenon of suppletion.

Table 1. Hungarian copula van

present past future subjunctive conditional

sg pl sg pl sg pl sg pl sg pl 1 vagyok vagyunk voltam voltunk leszek leszünk legyek legyünk volnék~lennék volnánk~lennénk 2 vagy vagytok voltál voltatok leszel lesztek légy~legyél legyetek volnál~lennél volnátok~lennétek 3 van vannak volt voltak lesz lesznek legyen legyenek volna~lenne volnának~lennének

infinitive present participle past participle future participle potential lenni való~levő/lévő volt leendő lehet

Table 2. Hungarian lesz “become”

present past (no synthetic future) subjunctive conditional

sg pl sg pl

sg pl sg pl 1 lesz-ek lesz-ünk lett-em lett-ünk

legy-ek legy-ünk len-n-ék len-n-énk 2 lesz-el lesz-tek lett-él lett-etek

légy/legy-él legy-etek len-n-él len-n-étek 3 lesz lesz-nek lett lett-ek

legy-en legy-enek len-ne len-n-ének

infinitive present participle past participle future participle potential lenni való~levő/lévő volt, lett leendő lehet

References Corbett, Greville G. 2007. Canonical typology, suppletion, and possible words. Language 83.8-42. Kangasmaa-Minn, Eeva. 1998. Mari. In Daniel Abondolo (ed.), The Uralic languages, 219-248. London: Routledge. Thornton, Anna M. 2011. Overabundance (multiple forms realizing the same cell): A non-canonical phenomenon in Italian verb morphology. In Martin Maiden, John Charles Smith, Maria Goldbach & Marc-Olivier Hinzelin (eds.), Morphological autonomy: Perspectives from Romance inflectional morphology, 358-381. Oxford: Oxford University Press.

ABSTRACT. The principle of Uniformitarianism was borrowed into Linguistics from Geology where it was made broadly known in the writings of Charles Lyell in his multi-volume Principles of Geology (1830 to 1833). As the principle is applied in Linguistics, it can be summed up, very roughly, in the notion that the present must be used to describe the past and, as an explanatory corollary, that forces operating around us today are the same ones that have been operating across the ages. Fox (1995:95) phrases it as follows: “the same principles operate consistently, regardless of time and place, and … regularities that can be observed in current and attested languages are equally valid for reconstructed languages” While some analysis of this principle exists, a great deal of indeterminacy still remains as to what exactly is meant by the notion that forces – or principles -- operating today are the same ones as in the past. There is a certain lack of precision in the literature as to what forces these may be and, perhaps more crucially, what is meant by “same” in regard to the past. Christy (1983, quoted in Joseph and Janda 2003:30) states, for example, that “what is observed in the present can be proposed for the past but what is not observed in the present cannot simply be banished from the realm of possible.” On the other hand, Lass (1997, quoted in Trask 2000:354) writes that “no linguistic state of affairs can have been the case only in the past.” The present paper will provide a review of some of the ways in which Lyell’s geological principle (actually four interlocking principles) of Uniformitarianism has been interpreted in the far more shallow time depth of Historical Linguistics, pointing out ambiguities and contradictions in the literature. It will also make some suggestions, influenced by the authors’ work in Cognitive Linguistics, as to how the term might be constrained in order to be made meaningful in language history.

References Christy, T. Craig. 1983. Uniformitarianism in Linguistics. Benjamins. Fox, Anthony. 1995. Linguistic Reconstruction: An Introduction to Theory and Method. Oxford University Press. Joseph, Brian D. and Richard D. Janda, eds. 2003. The Handbook of Historical Linguistics. Blackwell. Lass, Roger. 1997. Historical Linguistics and Language Change. Cambridge Lyell, Charles. 1830-3) Principles of Geology. London: John Murray Trask, R. L. 2000. A Dictionary of Historical and Comparative Linguistics. Edinburgh University Press

ABSTRACT. This paper attempts to advance our understanding of the intrinsic nature of the English progressive. Although Anderwald (2016: 2) notes that the prescriptive rule restricting occurrence of mental-state verbs in the construction has been acknowledged since the eighteenth century in grammar writing, practice of employing sentences such as I am loving and He was loving as illustrative models for the progressive was, in fact, substantially common in late modern grammars, including Lowth (1762: 56), Webster (1784: 24) and Blanch (1799: 63). These remarkably eminent grammars unanimously introduce the present progressive simply as the form that specializes in designating an actual situation existing at the moment of speech. This insight conforms to observations presented in Myers (1952: 177-178), Visser (1973: 1970), Ohye (1982), Kakietek (1997), Hübler (1998), Arnaud (2003), Granath and Wherrity (2013) and Killie (2014). If states can be defined straightforwardly in terms of cognitive homogeneity, the progressive seems to have been essentially aspect-neutral. It is certain that the progressive with a stative verb has been infrequent in written English, but this fact seems explainable if we follow those grand grammarians’ discernment. Examining diachronic as well as synchronic actual usage data of the construction, it is claimed that the intrinsic function of the periphrastic form is to narrow down the speaker’s viewing scope to the actual ‘here and now’ phenomenon that is foregrounded, superimposed on the back-grounded situation which the stem verb of the participle designates. This characterization leads not only to identifying one basic core meaning of the construction, which Kranich (2010: 72) has expressed skepticism about achieving, but also to explaining diverse facets of the progressive. For instance, the progressive is found largely within the context of the speech-based discourse, conceivably because the construction requires the speech-participants to share the same real-time and mental-space. I’m liking it may at times suggest temporariness because it is basically about an actual ‘here and now’ feeling. The speaker, however, may not necessarily be conscious of the boundaries of the fondness, which would entail possibility for a stative construal for the situation expressed by the verb like. I’m hoping if you could …. can be used for polite requests or declination because conveying delicate matters as if they are entertained merely at the speech moment could mitigate the mental burden. The oddity of It is being 5 o’clock may be attributable to superfluity of narrowing the scope for an already momentary situation. The construction is occasionally used for emphasis because confining induces focusing, which might enhance the magnitude of attention. Colossal success of Murray’s grammar compilation, which first adopted the prescription in the fifth edition, published in 1799, with its repercussion might have suppressed use of the progressive with a stative verb in norm-conscious writing, but there does not seem to be particular necessity to defend the rule in describing the spontaneous progressive.

References Anderwald, Lieselotte (2016) “I’m Loving It – Marketing Ploy or Language Change in Progress?,” Studia Neophilologica. Arnaud, René (2003) Letter-writers of the romantic age and the modernization of English: a quantitative historical survey of the progressive. Accessed at http://www.univ-pau.fr/ANGLAIS/ressources/rarnaud/index.html Blanch, Mercy (1799) A Short Introduction to English Grammar, etc. Volume I. London: printed by C. Law and F. Jollier. Comrie, Bernard (1976) Aspect. Cambridge: Cambridge University Press. Granath, Solveig and Michael Wherrity (2013) “I’m loving you – and knowing it too: Aspect and so-called Stative verbs,” Rhesis. International Journal of Linguistics, Philosophy, and Literature, Linguistics and Philosophy, 4.1: 6-22. Hübler, Alex (1998) The Expressivity of Grammar. Grammatical Devices Expressing Emotion across Time. Berlin: De Gruyter. Kakietek, Piotr (1997) The Syntax and Semantics of English Stative Verbs. Warszawa: Wydawnictwo Energeia. Killie, Kristin (2014) “The Development of the English BE + V-ende/V-ing Periphrasis: From Emphatic to Progressive Marker?,” English Language and Linguistics, 18 (3): 361-386. Kranich, Svenja (2010) The Progressive in Modern English: A Corpus-Based Study of Grammaticalization and Related Changes. Amsterdam and New York: Rodopi. Lowth, Robert (1762) A Short Introduction to English Grammar. London: A: Millar, R. and J. Dodsley. Murray, Lindley (1799) English Grammar (the 5th edition). London: Longman and Rees; York: Wilson, Spence, and Mawman. Myers, Louis McCorry (1952) American English: a Twentieth-century Grammar. New York: Prentice-Hall. Ohye, Saburo (1982) Doushi (1) “The Verb (1).” Tokyo: Kenkyusha. Webster, Noah (1784) A Grammatical Institute of the English Language. Hartford: Hudson and Goodwin. Visser, Fredericus Theodorus (1973) A Historical Syntax of the English Language Part 3 (Second Half). Leiden: Boston and Köln Brill.

ABSTRACT. Grammaticalization framework suggests itself as a predicative power for language change. This paper considers the grammatical functions of the locative construction fi(ih) in the Gulf Arabic Pidgin (a variety spoken by workers coming from the Indian subcontinent and south Asian countries to the Arabian/Persian Gulf States). In Gulf Arabic, there are 1) the preposition fi ‘in; into; inside’ and 2) the locative construction fi(ih) ‘there is/are’ which assumes only an existential function. In Gulf Arabic Pidgin, the locative construction fi(ih), however, has assumed several grammatical functions, including: 1) a possessive marker (i.e., have-constructions); 2) equative/predicative copula; and 3) preverbal predicative marker. The aims in this paper are two-fold: first, to show how a grammaticalization framework can possibly account for the grammatical innovations of fi(ih) in the Gulf Arabic Pidgin; and second, to suggest that these grammatical innovations might be results of an ongoing grammaticalization process of LOCATIVE>TMA/PROGRESSIVE. Earlier studies conducted on this pidgin serve as data sources for this project. The paper implicates the effectiveness of grammaticalization framework and its principles as an analytical apparatus that can account for and interpret language change in Arabic-based contact languages. From another aspect, locating the developments of fi(ih) from a locative construction to a copula and TMA marker in the currently documented historical background and sociolinguistic circumstances, specifically (non)nativization, of GAP generates more questions than answers. Copulas and TMA markers are typical morpho-syntactic features of creoles (Seigel, 2006; Winford, 2003) and scarcely reported as morpho-syntactic features of pidgins, even rapidly developed (Tagliamonte, 2000) and expanded/elaborated pidgins (Seigel, 2006; Winford, 2003). GAP was only classified as a pidgin without further classifications based on subtle developments, like the ones discussed in this paper that might reflect its current status in the Gulf States. On the one hand, such subtle developments might indicate a crucial shift related to the status of a language. Adding the findings and the scrutiny of this paper to the existing attested structural findings of GAP would result in a reconsideration of the pidgin’s status in the Arabian Gulf States since this kind of language development might run counter to some of pidgin universals, specifically in terms of copula and TMA grammatical developments. That is, where would it be positioned in terms of the pidginization-creolization continuum? Would it be an extended and elaborated pidgin? Or, it would be more than that when considering the other factors (e.g., length of stay). On the other hand, scholars interested in contact linguistics might want to incorporate the new findings and views of language in contact situations from different places in the world. Doing so might lead into reconsidering the defining grammatical features of pidgins where it is possible to note the developments of copula and TMA markers. Such findings and views might also steer those scholars to reevaluate the (non-) nativization distinguishing aspect between pidgin and creole languages.

ABSTRACT. Time expressions are a common source of causal, concessive and conditional conjunctions (Kortmann 1997:137-211). In this paper we focus on the development of the Basque intransitive adverb gero ‘later’ as a transitive adverb (‘after’) and as a subordinating conjunction. The Basque time adverb gero ‘later’ when postposed to a subordinate verb gives rise to clauses with a number of meanings, including temporal (1), causal (2) and conditional (3) (Artiagoitia 2003). (1) Profesionaletan hasiz gero 18 neurketa irabazi ditu 25 partidatan professional.loc.pl begin.inst later measure win aux game.loc.pl ‘Since he started in the professional leagues, he has won 18 points in 25 games’ (Berria 2003-10) (2) Krediturik eman nahi ez didazunez gero beste banku batera credit.part give want no aux.instr after other bank one joango naiz hemendik aurrera go.fut aux here.abl front.all ‘Since you do not want to give me credit, I will go to another bank from now on’ (Euskara Institutua (a) ) (3) Bost gradu gehiago igoz gero, izotz guztia urtuko litzateke five degree more climb.instr later ice all melt.fut aux ‘Rising by 5 more degrees/if it rose by 5 more degrees, all the ice would melt’(Berria 2004-10-19) The use of the intransitive adverb gero (‘later’) as a transitive adverb that can select a subordinate clause (‘after’) may have arisen by syntactic reanalysis, as in (4) (=3) and (5). The subordinate clause contains an inflected participle in (4) and a conjugated verb in (5). (4) [Bost gradu gehiago igoz] gero izotz guztia urtuko litzateke [Bost gradu gehiago igoz gero] izotz guztia urtuko litzateke ‘Five degrees going up, later all ice would melt’ ‘Once it rises by five more degrees, all ice would melt’ (5) [joango naizenez], gero ekarriko dut [joango naizenez gero] ekarriko dut go.fut aux-instr later bring-fut aux ‘As I will go, later I will bring it’ ‘Since I will go, I will bring it’ The reanalysis would be along the same lines as may be assumed for the origin of participial resultatives in -ta from coordinated sentences with eta ‘and’ (e.g. ikusi eta ‘see and’ ikusita ‘(having) seen’, Krajewska 2012:16). Starting from the transitive usage of the adverb gero, its development as a causal or conditional conjunction can be explained as a grammaticalization process associated with the conventionalization of an invited inference (see Geis & Zwicky 1971, Levinson 1995, Traugott & Dasher 2002: 16-17, 80, Pérez Saldanya 2014, among others). By this inference, what happens first and is presented as topic or theme will tend to be interpreted as the cause of what happens next and is presented as rheme. The interpretation will be causal if subordinate and main clause have a factual character, as in (2), and will be conditional (conditioned cause) if both clauses are presented as possible, as in (3). As we will show in this paper, the temporal origin of the conjunction gero determines the types of causal constructions in which it is used, at least at the earlier stages in the grammaticalization process. First, it explains the fact that it is used in causals that, like conditionals (Haiman 1978), have a thematic character. In addition, it explains why typically these are speech-act causals (Sweetser 1990: 76-86); that is, causals that justify the illocutionary force of the main clause, as in (6), or some kind of modality (intention, necessity, etc.), as in (7). (6) Ene penea daukusun gero, arren, berba bat esazu my pain you.see.comp later please word one tell.me ‘Since you see my pain, please, tell me a word’ (J.P. Lazarraga, 16th c.) (7) Horrein gaitz ziraden gero, eginen dut berzerik so hard you.are later do.fut aux other.part ‘Since you are so difficult, I will do something else (B. Etxepare, 16th c.) In this paper we consider the use of gero as a conjunction in a Basque corpus that includes texts from the first books written in this language (in the 16th century) to the present (Euskara Institutua (a) , Sarasola et al. 2015), analyzing the readings that arise in clauses placed to the left or to the right of the main clause. A question that may be asked is whether the evolution of gero may have been influenced by contact with Romance. This development has some similarities with that of Spanish pues (que) ‘since’(< Lat post ‘after’). An obvious difference is that the Spanish conjunction lost its temporal interpretation, becoming purely causal and eventually mostly linked to rhematic causals. In this respect Bq gero is more like English since, although it has an even wider range of values. References Artiagoitia, Xabier. 2003. “Adjunct subordination”. In: Hualde, J.I. & J. Ortiz de Urbina (eds.) A Grammar of Basque, 710-762. Berlin: Mouton de Gruyter. Euskara Institutua, UPV/EHU (a) [Basque Institute]. “Kausa perpausak”. Sareko Euskal Gramatika [Basque Grammar Online]. www.ehu.eus/seg ISBN: 978-84-693-9891-3 Euskara Institutua, UPV/EHU (b). Euskal klasikoen korpusa [Corpus of Basque Classics]. http://www.ehu.eus/ehg/kc/. Geis, Michael & Arnold Zwicky (1971). “On invited inferences”, Linguistic Inquiry 2: 561-566. Haiman, John (1978). “Conditionals are Topics”, Language 54: 564-589. Krajewska, Dorota (2012). “The diacrony of resultative constructions in Basque”. M.A. Thesis, Univ. del País Vasco/Euskal Herriko Unib. Kortmann, Bernd (1997). A typology and history of adverbial subordinators based on European languages. Berlin: Mouton de Gruyter. Levison, Stephen (1995). “Three levels of meaning”. In: Frank R. Palmer (ed.), Grammar and meaning: Essays in honor of Sir John Lyons. Cambridge: Cambridge University Press, pp. 90-115. Pérez Saldanya, Manuel (2014). “Oraciones causales”. In: Concepción Company Company (dir.), Sintaxis histórica de la lengua española Tercera parte: Preposiciones, adverbios y conjunciones. Relaciones Internacionales, vol. 3. México: Fondo de Cultura Económica / Universidad Nacional Autónoma de México, pp. 3447-3610. Sarasola, Ibon, Pello Salaburu, Josu Landa & Josu Zabaleta (2015). Ereduzko prosa gaur [Standard Prose Today (Corpus)]. http://www.ehu.eus/euskara-orria/euskara/ereduzkoa/ Sweetser, Eve (1990). From Etymology to Pragmatics: Metaphorical and Cultural Aspects of Semantic Structure. Cambridge: Cambridge University Press. Traugott, Elisabeth C. & Richard B. Dasher (2002). Regularity in Semantic Change. Cambridge: Cambridge University Press.

ABSTRACT. El objetivo de esta investigación ha sido identificar el origen de las cláusulas de gerundio de posterioridad (GP) en el español. Para esto, se ha propuesto un modelo evolutivo, basado en la ruta de gramaticalización de combinación de cláusulas subordinadas de Hopper y Traugott (1993), en la escala de informatividad de Kortmann (1991) y en la tipología de cláusulas verbales no finitas de Haspelmath (1995). Se analizaron cualitativa y cuantitativamente los rasgos sintácticos y semántico-pragmáticos de las oraciones de gerundio extraídas de un corpus de aproximadamente 350,000 palabras de textos de autores peninsulares e americanos, periodísticos y literarios, del español del siglo XIX. Se partió de que en un corte sincrónico se hallarían fósiles de estructuras en distintos estadios de evolución (layering).

Los resultados apoyan la hipótesis de que las cláusulas de GP son estructuras cosubordinadas, es decir, constituyen estructuras de un grado de trabazón sintáctica intermedio que muestran rasgos cruzados de coordinación y subordinación. Las GP se derivan de las cláusulas de adición-circunstancia acompañante (Ad/ Ac), que son de naturaleza predicativa. También, se han reunido datos a favor de que, a partir de las cláusulas GP, surgieron las cláusulas de consecuencia y finalidad, que son plenamente subordinadas. Es decir, a medida que se gramaticalizan, van adquiriendo nuevos rasgos sintácticos subordinados y valores semántico-pragmáticos: Ad/ Ac > posterioridad > consecuencia > finalidad.

Esto comprueba que tipológicamente el español pertenece a aquellas lenguas que tienen un sistema de grados de subordinación que incluye la parataxis, la cosubordinación y la subordinación. Esto explica por qué el GP no ha tenido cabida en la normativa del español, que se ha considerado una lengua de solo dos grados de trabazón de cláusulas: parataxis y subordinación. La normativa debe tomar en cuenta la diacronía de la forma, pues un acercamiento sincrónico no alcanza a explicar muchos usos presentes en el sistema lingüístico.

Se propone que una recategorización coherente de las cláusulas de gerundio debe contemplar seis clasificaciones “temporales”, tres para cada nivel: Cosubordinadas: Precedencia Concomitancia Subsecuencia Subordinadas: Anterioridad Simultaneidad Posterioridad-Ø

Las cosubordinadas tienen la capacidad de expresar concomitancia y secuencialidad, pero no enmarcan temporalmente al predicado principal como lo hacen las subordinadas. Lo que se ha denominado posterioridad es una casilla vacía en el nivel de la subordinación plena. Lo que sí existe es la subsecuencia, que es un valor cosubordinado, que ha sido mal llamado posterioridad.

Por último, los resultados sugieren una ruta de evolución cíclica. Según se van subordinando las estructuras oracionales de gerundio, pierden su capacidad deíctica y se vuelven más abstractas y continuas, o sea, tienden a funcionar como deverbales y atributivas: [+ nominal/ adjetival] > [+ verbal] > [+ nominal/ adjetival]. Si la ruta evolutiva propuesta es correcta debe haber una vuelta o aumento de frecuencia del uso adjetival restrictivo del gerundio y del gerundio sustantivo. En realidad, estos usos no normativos están presentes en el español actual o, mejor dicho, siguen presentes.

References: Haspelmath, M. (1995). The converb as a cross-linguistically valid category. En M. Haspelmath y E. Konig (Eds.), Converbs in Cross-Linguistic Perspective: Structure and Meaning of Adverbial Verb Forms —Adverbial participles, gerunds—. New York: Mouton de Gruyter.

Hopper, P., Traugott, E. (1993). Grammaticalization. Cambridge, USA: Cambridge University Press.

Kortmann, B. (1997). Adverbial Subordination. A Typology and History of Adverbial Subordinators Based on European Languages. (Empirical Approaches to Language Typology). Berlin/New York: Mouton de Gruyter.

ABSTRACT. Extensive variation in verb tense and mood has been reported in both the protasis and the apodosis of conditional clauses, i.e., if p (protasis), then q (apodosis), synchronically (e.g., Poplack 2001), diachronically (e.g., Mugler 1980), and across languages (Haiman & Kuteva 2002). This paper focuses on the variation in the apodosis of conditionals in Mexican Spanish (MS) and analyzes the factors conditioning the possible verb forms in synchrony and diachrony, with a special interest in the imperfect subjunctive. I show that the variation observed in synchrony in MS is the retention of a pattern found in the history of the language and can be explained by a combination of semantic and syntactic factors.

In modern day Spanish, when the protasis is in the imperfect subjunctive, prescriptively the apodosis should have a verb in the conditional (C), but other forms may also occur in this clause: the imperfect subjunctive (IS), the imperfect indicative (IN), the present tense (PR), or a tacit () form, as in (1): 1) Si tuviera dinero iría C)/ fuera (IS)/ iba (IN)/voy (PR)/_____ () ‘If I had money I would go’ Rojo (1986) finds that from the writing of The lay of the Cid (Poema del Mio Cid) until 1230 the most common structure was the –se imperfect subjunctive in protasis and the conditional in the apodosis, but after 1250 this pattern decreased considerably, not reemerging until the 1800’s (Bartol Hernández 2013).

Cross-linguistically, one of the most cited conditioning factors for the form used in the apodosis of conditional clauses is priming, i.e. the use of one form due to its previous presence in the discourse, which has been well established in psycholinguistic studies. For instance, Haiman and Kuteva (2001) look at conditionals in different language families and state that clause symmetry is the driving force behind this type of change. For example, in Spanish the IS in the protasis triggers its use in the apodosis. On the other hand, semantic explanations trying to account for this variation are based mainly on the reality of the event: if it is [+real] the conditional verb form is more likely to appear in the apodosis (e.g., Poplack 2001; LeBlanc 2009), and if the event is [–real] there is a higher probability for the use of the imperfect subjunctive. Similarly, temporal reference of the event is also used to account for this variation where a present/future time of reference triggers the conditional and a past time of reference triggers the subjunctive (Mugler 1980). On the other hand, authors such as Rojo (1986) argue that the binary [+ real] semantic distinction is not enough and claims that factors such as type of speech (i.e., direct or indirect) and genre (e.g., poem, play) also play a role in verb form choice.

This study draws from synchronic and diachronic data in order to analyze the variation in verb form and conditioning environment that appear in in the apodosis of conditional clauses when the protasis is in the imperfect subjunctive. Due to the low frequency of the phenomenon, the synchronic data comes from 216 sociolinguistic interviews (Corpus sociolingüístico de la Ciudad de México-PRESEEA and Corpus Monterrey-PRESEEA), from Corpus de Referencia del Español Actual (Mexico) and from twitter (Mexico). The diachronic data was taken from the Corpus Diacrónico del español - México (CORDE) and starts from the XVI century after the Spanish colonization. More than 500 tokens were coded according to the following independent variables: probability of the event from the speaker’s point of view (+real/potential/-real), polarity of the sentence (positive/negative), temporal time of reference (past/present/future), text type (e.g., play/novel/etc.), position of the apodosis (before/after the protasis), intervening material between the protasis and the apodosis, mode (written/oral). The data was then submitted to a multinomial regression analysis using the statistical package RStudio.

The data distribution for the MS synchronic data indicates that the conditional is the most frequent form, followed by the imperfect subjunctive, and other forms (e.g., present indicative) constituted the rest of the data. The diachronic data shows a similar distribution of patterns for the XVI-XVIII centuries as reported in previous studies (e.g., more IS in the apodosis in the XVII century). However, in MS, unlike what happens in other dialects of Spanish (e.g., Iberian Spanish), we don’t see a dramatic increase of the conditional in the apodosis in the XIX century. Instead, the increase was subtle and the imperfect subjunctive, even though its used decreased, has been kept at a constant rate. With regards to the statistical analysis, the multivariate logistic regression indicates that the imperfect subjunctive is favored when the event is [-real], when the sentence has negative polarity, when the time of reference is past, and when the apodosis comes after the protasis. These results show that the variation in the apodosis does not represent a neutralization in meaning between the forms; rather, the contrast in meaning remains.

This paper provides evidence for the relevance of semantic and syntactic factors in the variation of verb form in the apodosis of MS conditionals. The current study also shows the importance of connecting diachronic data with synchronic studies since synchronic linguistic variation could be mistaken for language change. In the case of MS, the use of the IS is by no means an innovation, but rather a retention of an old pattern that has persisted in this dialect. Moreover, the results are in line with what has been found in languages like French (Poplack 2001; LeBlanc 2009) where [+real] events favor the use of the conditional. With further research in other Romance languages (and/or other language families), a broader generalization could be made about conditional sentences. Finally, explanations that are based solely in one factor such as Haiman and Kuteva’s (2001) priming do not give us enough information about the factors conditioning variation.

References Bartol Hernández, José Antonio. 2013. "Habría dado vs. hubiera dado en la apódosis de las oraciones condicionales irreales a comienzos del siglo XIX." Moenia 19:443-467. Haiman, John, and Tania Kuteva. 2002. "The symmetry of counterfactuals." Complex sentences in grammar and discourse: Essays in honor of Sandra A. Thompson:101-124. LeBlanc, Carmen L. 2009. "Conditional morphology in si-clauses: A Canadian-French reanalysis." The Canadian Journal of Linguistics/La revue canadienne de linguistique 54 (2):317-337. Mugler, France. 1980. "Concerning the Usage and Evolution of the Conditional Sentence in Latin." Glotta 58 (1):119-132. Poplack, Shana. 2001. "Variability, frequency, and productivity in the irrealis domain of French." Typological studies in language 45:405-430. Rojo, Guillermo. 1986. "On the Evolution of Conditional Sentences in Old Spanish in Studies in Romance Linguistics." Publications in Language Sciences (24):167-188. RStudio Team (2016). RStudio: Integrated Development for R. RStudio, Inc., Boston, MA URL http://www.rstudio.com/.

ABSTRACT. Characterization of the subject and direct object of transitive sentences in the history of Spanish Javier Puerma Bonilla Universidad de Granada

Rodrigo Flores Dávila Universidad Nacional Autónoma de México

This presentation aims at diachronically characterizing and analyzing both the subject and the direct object when expressed through explicit NPs in transitive constructions. Although Spanish has been characterized as a language whose word order is SVO (Hawkins 1983; Dryer 1997; Gutiérrez-Bravo 2007, 2008), a diachronic approach seems determining to accurately characterize the constituents that actively participate in the transitive construction. The ongoing research aims at providing groundbreaking information regarding relative word orders throughout the history of Spanish. In doing so, we intend to analyze both the formal and the semantic aspects that have not been dealt with yet in relation with the explicit subject and the lexical direct object. This research is base don the analysis of a corpus of up to 1007 entries dealing with transitive constructions that stem from a word universe of up to 15000 per chronological interval. Intervals correspond to 400 years each: 13th century > 17th century > 20th century. Corpus data stem from up to 12 literary works, organized in up to 4 textual genres: novelized prose, essayistic and historiographical prose, juridical prose, and scientific prose. Our research is based on a study of general Spanish, however, American varieties have been included from the 17th onwards. Throughout the history of Spanish several word orders have been documented. In them, a explicit subject, a transitive verb, and a lexical direct object have always been present. That said, up to 6 different word orders can be tracked since the 13th century: (1a) SVO, (1b) SOV, (1c) VSO, (1d) VOS, (1e) OSV, (1f) OVS.

(1) a. Dizen que un cuervo avía su nido en un árbol en el monte [Calila, 143, XIII] ‘It is a raven had its nest in a tree the mount’ b. E Judas cuando esto oyó pesól mucho [GEI, 406, XIII] ‘And once Judas heard that, he did regret it’ c. cerró el criado la puerta con llave y se la dio a su señor [Desengaños, 110, XVII] ‘the servant locked the door up and gave it to his master’ d. que me non aya enbidia ninguno [Calila, 308, XIII] ‘may nobody envy me’ e. Estas materias solo yo las alcanzo, aunque el Tiempo no las ignora [Rey Gallo, 215, XVII] ‘I alone know these subjects, although time does not ignore them’ f. Esto dixo la Hormiga [Rey Gallo, 167, XIII] ‘That said the ant’

Formal characteristics (structural weight) and semantic characteristics (lexical features) of both explicit subject and lexical direct object appear to determine the positions such constituents occupy in transitive constructions of Spanish. Given that the quantitative distribution of all 6 word orders differs throughout the history of Spanish language, on the one hand, we analyze whether such sematic and formal features are determinant when it comes to foster or inhibit possible word orders; on the other hand, which ones are the main word orders in the history of Spanish.

References

Gutiérrez Bravo, Rodrigo. 2007. “Prominence Scales and Unmarked Word Order in Spanish”, Natural Language and Linguistic Theory 25, 235–270. ____. 2008. “La identificación de los tópicos y los focos”, Nueva Revista de Filología Española 56, 363-401. Hawkins, John A. 1983. Word order universals, Nueva York: Academic Press. Dryer, Matthew S. 1997. “On the six-way word order typology”, Studies in Language 21, 69–103.

ABSTRACT. In this contribution, the influence will be discussed certain registers of world language English may have on the language of specific groups of speakers of other Western European languages. An example may be the spread of the clipped -o (as in nympho, dipso subsequently commo, journo and finally lefto, creepo) in informal registers of languages such as Dutch, Swedish and even German. In all these languages, the new ‘suffix’ became productive in its own way. Especially the case of German is interesting since in German there already existed a clipping pattern ending in -i, which can be described in terms of a hypocoristic suffix. Traditionally English and Dutch prefer monosyllabic clipped forms: English Dutch (1) ad (2) Jap(anner) Japanese pub mees(ter) teacher

Also in German one finds ample examples of monosyllabic ‘Kurzwörter’ (3) Prof(essor) Alk(ohol)

In contrast to Dutch in German a disyllabic pattern ending in an unstressed vowel occurs rather frequently: (4) Abo(nnement) Alu(minium) Next to these patterns one finds truncated forms followed by a suffix -i, which should be interpreted as a hypocoristic suffix: (5) Schoki (Schokolade) Pulli (Pullover)

This last type of ‘clipped forms’ is the most common and most frequent pattern in German

Similar forms occur in English (6) Chevvy (Chevrolet) ciggy (cigarette)

In Dutch, this type of clipped forms does not exist.

However, in modern Dutch, Swedish and other languages one finds disyllabic clipped forms ending in unstressed -o frequently. In addition, the same new suffix even can occur after full words in an informal register. Notice that the syllable before -o must be stressed, resulting in a final trochee. (7) lesbo (lesbian) (8) suffo (suf ‘dull’) Limbo (Limburger) lullo (lul ‘penis’)

The pattern behind these recently coined forms is borrowed from modern American English, were one finds examples such as: (9) commo (commissary) (10) weirdo journo (journalist) sicko

Most remarkable is that the same development even occurred in German, however the productivity of this clipping + suffixation process is hampered by the similar -i hypocoristic suffixation:

(11) Realo (Realist) (12) Fundamentalo (fundamental) Anarcho (Anarchist) Kloppo (Jürgen Klopp) The process usually starts with the direct borrowing of a few words from a pop music register or (social) media language into the vernacular of Dutch youngsters, followed by a process of reanalysis and a subsequent new Dutch, Swedish or German word formation process. The speakers who introduce this kind of language change cannot be described as bilingual. They have a certain knowledge of English, which makes it possible for them to grasp the meaning of this informal ‘suffix’ when used in mass media, popular music, social media and other forms of popular culture. However, what seems more important for the process of borrowing is the stylistic value or connotation of the ‘suffix’. A similar development can be found when one takes a closer look at a process such as blending, which became very popular in the jargon of PR, commercial name giving and science in Western European languages recently. References ELSEN, Hilke (2011): Grundzüge der Morphologie des Deutschen. Berlin-Boston: de Gruyter. HAMANS, Camiel (1997a): Clippings in modern French, English, German and Dutch. Raymond Hickey & StanisławPuppel (eds.). Language History and Linguistic Modelling. A Festschrift for Jacek Fisiak on his 60th Birthday. Berlin/New York, Mouton De Gruyter, 1733–1742. HAMANS, Camiel (1997b): Im Westen nichts Neues: over de opkomst van een beschaafd morfeem. Ariane van Santen en Marijke van der Wal (eds.). Taal in tijd en ruimte. Leiden, SNL 237–245. HAMANS, Camiel (2004): The relation between formal and informal style with respect to language change. C.B. Dabelsteen & J.N. Jorgensen (eds.). Languaging and Language Practicing. Copenhagen, University of Copenhagen, Faculty of Humanisties. Copenhagen Studies in Bilingualism vol. 36, 168–195. HAMANS, Camiel (2012): From prof to provo: Some observations on Dutch clippings. Botma, Bert & Roland Noske (eds.), Phonological Explorations: Empirical, Theoretical and Diachronic Issues. Berlin: Mouton de Gruyter, 25–40. HINSKENS, Frans (2001): Hypocoristische vormen van reductievormen in het hedendaagse Nederlands. Neerlandica extra muros 49: 37–49. JESPERSEN, Otto (1942): A Modern English Grammar on Historical Principles. Copenhagen: Ejnar Munksgaard. MARCHAND, Hans (1969): The categories and types of Present-Day English Word-Formation. A Synchronic-Diachronic Approach. Second, completely revised and enlarged edition. München: C.H. Beck. VIJVER, Ruben van de (1997): The duress of stress: On Dutch clippings. Jane Coerts & Helen de Hoop (eds.), Linguistics in the Netherlands 1997. Amsterdam: John Benjamins: 219–230.

ABSTRACT. The relexification of diminutives is, we claim, one of the most productive ways for the creation of new words in Spanish. This phenomenon consists of the reanalysis of a lexical item composed of root + diminutive suffix, whereby the original semantic value of the suffix is bleached over time and its form is subsequently reanalyzed as part of the new root carrying new meaning (e.g., tortilla). A classic example of this semantic shift, not unique to Spanish, was employed long before in Latin (e.g., AURIS + -CULA = oreja ‘ear’). In contrast to the relexified diminutive, an ad hoc diminutive is a form that retains the separate meanings of root and diminutive suffix (e.g., perro + ito = perrito ‘little dog’). This study provides a quantitative analysis of the entirety of diminutives, relexified in the history of Spanish, by utilizing three separate dictionaries: 1) Stahl and Scavnicky’s Reverse Dictionary of the Spanish Language (1973) for isolation of potential diminutive forms; 2) the online Diccionario de la lengua española de la Real Academia Española (DRAE) at to assess the semantic variation of each entry and to eliminate words from the count that do not originate in diminutives (e.g., mecánica); and 3) the Diccionario etimológico de los sufijos españoles y otros elementos finales (2002) to construct the approximate timing of entry of the variety of diminutive forms, both ad hoc and relexified, into the Spanish lexicon. This study has resulted in the first complete, quantitative and diachronic classification of relexified diminutives in modern Spanish which uses several dictionaries to provide a more comprehensive description of a given lexical phenomenon. Through combined use of these dictionaries as well as frequency data, the study will show that words ending in -illo, -illa are significantly predominant among relexified diminutives, followed by those ending in -ete, -uelo, -ito, -V-jo, and -ico, in decreasing order. Another interesting finding is that not until the fifteenth century did -ito began ousting –illo as an ad hoc diminutive, much in the same way –ELLU replaced –V-(C)ULU in post Classical Latin. The DRAE data of this study also confirms the fact that –V-(C)ULU was the precursor to –ELLU as the ad hoc diminutive of choice, in that the number of Spanish –V-jo, -V-ja forms relexifying in Latin were shown to number comparably much higher than –illo, -illa forms relexifying during that time More generally, this presentation will demonstrate that present-day relexified diminutives in Spanish have witnessed multiple origins: Classical Latin, Post-Classical Latin, and Romance languages other than Spanish. It will also be shown that not all present-day diminutive forms, particularly the extremely productive ad hoc suffix –ito, have originated in other historically diminutive forms, as in the case of Latin hypocoristic –ITTU. The findings of this study suggest that the future of relexified diminutivization will continue the recycling path, perhaps at first with words ending in –ito, becoming the very next common relexified diminutive in the near future (already evident in such forms as burrito), much like –illo was previously, and as V-(C)ULU was before that. The next interesting question to follow is, if –ito is to become a new form characteristic of relexification, then what other, new form would take its current place as the next ad hoc diminutive of choice?

References

Alvar, Manuel / Pottier, Bernard. 1983. Morfología histórica del español. Madrid: Editorial Gredos. Butler, Jonathan, Lowell. 1971. Latin -INUS, -INA, -INUS and -INEUS: From Proto-Indo-European to the Romance Languages. Berkeley: University of California Press. Crowhurst, Meagan J. 1992. Diminutives and Augmentatives in Mexican Spanish: A Prosodic Analysis. Phonology 9, 221-253. DIRAE Diccionario inverso basado en el Diccionario de la lengua española de la Real Academia Española. González Ollé, Fernando. 1962. Los sufijos diminutivos en castellano medieval. Madrid: Editorial Gómez. González Ollé, Fernando / Casado Velarde, Manuel. 1992. Spanisch: Wortbildungslehre. Formación de palabras. In Holtus, G. et al (Eds.) Lexicon der Romanistichen Linguistik, VI-1. Tübingen, Max Niemeyer, 91-109. Gooch, Anthony. 1970. Diminutive, Augmentative and Pejorative Suffixes in Modern Spanish. A Guide to Their Use and Meaning (2nd ed.). Oxford: Pergamon Press. Grandgent, C. H. 1907. An Introduction to Vulgar Latin. Boston: D.C. Heath and Co. Publishers. Hasselrot, Par Bengt. 1957. Études sur la formation diminutive dans les langues romanes. Acta Universitatis Upsaliensis 1957/11, 344. Lapesa, Rafael. 1981. Historia de la lengua española. (9th ed.) Madrid: Editorial Gredos. Lázaro Mora, Fernando A. 1999. La derivación apreciativa. In Bosque, Ignacio / Demonte, Violeta (Eds.) Gramática descriptiva de la lengua española Madrid: Editorial Espasa Calpe, S. A. Vol. 3, 4645-4682. Lefebvre, Claire. 1999. Creole genesis and the acquisition of grammar. Cambridge Studies in Linguistics 88. Cambridge University Press. viii and 461. Mendoza, Martha. 2005. Polite Diminutives in Spanish. In Lakoff, Robin T. / Ide, Sachiko (Eds.) Broadening the Horizon of Linguistic Politeness Amsterdam: John Benjamins. 163-173. Meyer-Lübke, Wilhelm. 1895. Grammaire des langues romanes: Morphologie. Paris: H. Welter. Pharies, David. 2002. Diccionario etimológico de sufijos españoles. Madrid: Editorial Gredos. Placencia, María Elena 1996. Politeness in Ecuadorian Spanish. Multilingua - Journal of Cross- Cultural and Interlanguage Communication 15/ 1, 13-34. Ryan, John M. / Parra-Guinaldo, Víctor. Classification and History of Relexified Diminutives in Modern Spanish: A Lexicographic Approach. In Maria Victoria Rodríguez Domínguez et al., Words across History. Advances in Historical Lexicography and Lexicology. Las Palmas de Gran Canarias: Servicio de Publicaciones y Difusión Científica de la ULPGC. Pp. 364- 380. 2016. Stahl, Fred A. / Scavnicky, Gary E.A. 1973. A Reverse Dictionary of the Spanish Language. Chicago: University of Illinois Press. Teschner, Richard V. 1999. Camino Oral. McGraw-Hill. Travis, Catherine E. 2004. The Ethnopragmatics of the Diminutive in Conversational Colombian Spanish. Intercultural Pragmatics 1/2, 249-274. Weidhase, Rodsel. 1967. Beitrage zur Kenntnis der Spanischen Suffixe. Tubingen.

ABSTRACT. As shown by Meira and Gildea (2009), the broadest typological survey of adjective classes (Dixon 2004) does not account for the large category of adverbs to which property concepts belong in Cariban languages. Furthermore, even the two major typologies of parts of speech (Croft 2001 and Hengeveld 1992) leave out the possibility of a system such as the one found in the Cariban family consisting of four major lexical categories (verbs, nouns, adverbs and postpositions) without a basic lexical category dedicated to the function of adnominal modification. This paper builds on the hypothesis that such a system developed from postpositional phrases (Meira and Gildea ibid.) by providing an additional source for the phenomenon in secondary predication from the Yawarana language (spoken in Venezuela and Brazil by about 9000 people). The typological uniqueness of the parts of speech system found in Cariban languages lies in the combination of the semantic and functional potential of one of its lexical classes. Throughout the family, this lexical class hosts close to a 100% of all the property concepts (in the 13 semantic types typically associated with adjectives, cf. Dixon 2004:3-5) in addition to concepts of manner, time and space and had been described as functioning without further derivation in verbal modification (1) and stative (copular) predication (2). Regardless of semantics, in order to function as adnominal modifiers, members of the class are always derived through nominalizing suffixes (3a) (the same suffixes are used in argument function (3b)). (1) Ekammajö-kö-öne tü tamjöne-ene. (2) ködöije-nñe sü'na na=to yawö. 3O.tell-IMP-INTENS INTENS fast-INTENS sick-PL dog 3.COP=COL thus ‘Tell it fast!!!’ (ConvDmgzFem.035) ‘(the) dogs are sick’ {DescCaz.065} (3) a. kün-a'seuwü-i=cho yaawö [jooj-ato sotto] yaawö 3S.DIS-laugh-RPP=COL thus lot-NZR people thus ‘A lot of people would laugh’ {Yek:CtoCania.019} b. iyö judum-ato=ke kün-ajojo-i che'a tüwü yaawö DEM.IN black-NZR=INSTR 3/3.dis-touch-PRP also 3SG thus ‘he also touched that with the black [thing]’ { CtoKms.511} The functional potential of dependent clauses had not been thoroughly explored, but it was expected to be similar to the potential of basic lexemes in the parts of speech system (cf. Van Lier 2009 on predicting this correlation). Thus, nominalized clauses and adverbial clauses were expected but the existence of a verbal adjective or participle required further inquiry. In a documentary corpus of Ye’kwana, a functional analysis of the constructions containing a reflex of the *t-V-tjô form –previously identified as a participle that could function as a stative or resultative predicate (4)– revealed that it could also function as a verbal modifier (a converb, (5)) and in the previously overlooked function of secondary predication (6). Further analyses revealed this was also the case in textual examples of other Cariban languages (Tiriyó and Carib of Surinam). (4) ¿Tü-madicha-jö'-e=jünka nai? (5) iyö nuunö t-öne'ma w-öönetü-i AZR-ripen-PL.O-PTCP=no? 3.COP-INTER DEM moon AZR-see.PTCP 1S-dream-PRP ‘Isn't [it] ripened?’{DescTab.080} ‘I dreamt I was watching the moon’ (6) t-akoicha-ane minñaatö kün-moomü-i yaawö AZR-finish.PTCP-INTENS liana 3/3.DIS-roll_up-PRP thus ‘He rolled the liana until the end (lit. finished)’ {CtoTapMor.089} Syntactic sources for the development of a typologically unique property concepts class 2/2 This discovery led back to a reassessment of the functions of the basic and derived adverbs which confirmed secondary predication was also possible with lexemes of this class. (7) Nudö-ödö ewü w-akö'-jö'-a yaawö. alive-AUG 1SG 1SG-cut-PL.O-NPST then ‘I cut them (the anteater hands) alive.’ {DescCaz.148} In the languages analyzed by Meira and Gildea as well as in Ye’kwana, a high number of the adverbial lexemes share endings that seem to be former derivational morphemes, like the essive morpheme (=me, =pe or =je depending on the language) or a circumfix of the type t(ɨ)-VERB-Ce but also other non-longer synchronically productive morphemes. So presumably Cariban adverbs originated as postpositional phrases which ended replacing the original noun root hypothesized to have referred to an object or person possessing a property (e.g. ‘the beautiful one’). However, in at least two cases, Ye’kwana texts have revealed that the original noun root still exists as an independent but inalienable noun referring directly to the property: PERS-nñata-dü ‘PERS’s beauty-POS’ coexists with inñataje ‘beautiful(ly)’ and PERS-ködöi-chü ‘PERS’s sickness-POS’ coexists with ködöije ‘sick(ly)’ (these forms differ from the synchronic essive in that the possessive suffix is missing and not preserved as it is in the synchronic use of the essive, e.g. y-oi-chü=je ‘his-mixture-POS=ESS’). This means it is not necessarily the case that the lexicalized postpositional phrases displaced their source object noun root and that, if both expressions can be preserved, probably each had or developed its own array of semantic and functional potential. Meira & Gildea (2009:127) hypothesized the postpositional strategy for expressing property concepts was born through an innovative copular construction that did not allow a nominal complement due to its intransitive origin. It is the case that, since verbs take bare nouns as arguments, nouns in non(-core)-argument functions are always found in a postpositional phrase. Hence, to express a secondary predicate, property concepts being nominal and being used as something other than a core argument of verbs would also have been syntactically required to occur in a postpositional phrase. The latter provides a second syntactic argument that historically supports the functional analysis that is made of Cariban languages as having a lexical class of predicate modifiers semantically associated with properties. It is possible that similarly functionally motivated systems of parts of speech exist in other languages but have not yet been described. Although probably most reference grammars of endangered languages include a section on the different types of dependent clauses –including adverbial ones– it is fairly unlikely that more than a few even mention secondary predicates. It is however this less widely explored function that was necessary to understand some of the examples found in the spontaneous annotated corpus of Ye’kwana. Hopefully the case of Ye’kwana will inspire other documenters to look beyond the most well established typological categories and typologists and historical linguists to encourage detailed work in more endangered languages. References Meira, S., and S. Gildea. 2009. “Property Concepts in the Cariban Family: Adjectives, Adverbs, And/or Nouns.” In The Linguistics of Endangered Languages- Contributions to Morphology and Morphosyntax, ed. by L. Wetzels, 95–133. LOT Occasional Series.Utrecht,The Netherlands: LOT. Lier, E. H. van. 2009. “Parts of Speech and Dependent Clauses: A Typological Study.” http://dare.uva.nl/record/314300.

ABSTRACT. South America is the most recently populated of all continents (O’Connor & Kolipakam, 2014). However, its rampant linguistic diversity, which consists of ca. 108 separate language families, remains still an unsolved puzzle for linguists interested in language change. Previous macro-comparative research has been based specially on lexicon, cf. Greenberg (1960, 1987), Suárez (1974), Kaufman (1990, 1994), and Loukotka (1968). Their approaches have not led to a wide consensus since many of their claims cannot be fully corroborated in the framework of the comparative method. The processes that might have been involved in this massive diversification are as yet unknown. For this talk, I focus on Shawi, an Amazonian language spoken in the Eastern Slopes of the Northern Andes of Peru. Before the arrival of the Spaniards, this region was characterised by very intense language contact and multilingualism, reflected in substantive Carib and Arawak vocabulary borrowed into modern Shawi. Although this linguistic richness was almost entirely wiped out of the Andean flank of the Shawi speaking area, languages like Awajún (Chicham) and Quechua IIB still play a significant role. Unfortunately, Shawi is a very underdocumented language, both grammatically and sociohistorically, an issue which was completely neglected until very recently (Rojas-Berscia, 2013, 2015; Rojas-Berscia & Ghavami Dicker, 2015; Valenzuela, 2011, 2015). Shawi is a language within a small language family (Kawapanan), together with its almost extinct sister language Shiwilu. This almost-isolate status, nevertheless, ignores the actual social history of the language. One potential problem is that, based on the traditional comparative method, previous studies seek a single history for all parts of the lexicon, while different words, henceforth just tokens, might have different histories. An alternative approach is to look at smaller groups of tokens that are more tightly related, such as pronouns or numeral systems. Shawi and Shiwilu show a pronominal system similar to so-called unrelated languages: Table 1: Pronominal systems in Kawapanan, Puelche and Aymara Shawi Shiwilu Puelche Aymara 1 ka kwa kwa naya 2 kema kenma kenma huma 3 ina/saya/pasu nana şaşa Hupa 4 kiya kuda kişa jiwasa Is this as simple coincidence or a deep historical signal? To answer this question, we gathered pronouns for 121 languages (Rojas-Berscia & Roberts in prep.). We sorted the forms into cognate sets, and then aligned the segments within each set (by hand, but starting from an initial automatic coding by LingPy). We used Bayesian phylogenetic inference to generate a set of trees that modelled the data well. However, instead of modelling innovations and death of whole cognate sets, we modelled the innovations and death of individual segments at particular points in the words. The final data had 1108 sites for 40 cognate sets. Summary trees were generated using maximum clade credibility and densitree. We compare the results to other accounts of the linguistic history of South America. As it will be shown, relatedness crosscuts the boundaries of traditional language families, i.e. small subsystems are related in different ways. Numeral systems also show a particular history. Kawapanan numeral systems share a formal mould, namely [numeral-stone classifier], with numeral systems from the Northern Pacific Coast (Mochica-isolate) and the northern-central eastern Slopes (Cholón-Hivito and Muniche) (Eloranta & Rojas-Berscia sub.), see Table 2. This can be explained by the fact that before the arrival of the Spaniards a very important trading route united all the speakers of these languages: the salt-trading route. Surprisingly enough, the classifier for salt stones lexicalised in all the numeral systems of the aforementioned languages. 1 Radboud Universiteit Nijmegen - Centre for Language Studies; Max Planck Institute for Psycholinguistics - Department for Language and Cognition; Pontificia Universidad Católica del Perú - Departamento Académico de Humanidades; Australian National University - The Wellsprings of Linguistic Diversity Table 2. Stone-based numeral systems in Kawapanan, Cholón-Hivito, Muniche and Mochica Shawi Shiwilu Cholón Híbito Muniche Mochica 1 a'-na a'-la a-ta/če e-tsí wü-tsa'a pac-pong 'twenty'; çoc-pong 'thirty', etc. 2 ka-tu ka-tu ip-ta op-tšē útspa< úp-tsa'a? 3 ka-ra ka-la ič-ta ú-tsi úts-ma<út-s-ma/ú-tsa'a-m? stone napi PK *lapi lapi PK *lapi ta tše tsá'a 'grain, corn' pong Classifier for stones, eggs or grains, small things -ra -la -ta -tše One of the main claims behind this first analysis is that South American linguistic history, as exemplified by Shawi as a language shaped by various waves of influence of different languages, does not necessarily fit the baselines of the traditional historical comparative approach. Pronominal and numeral systems in South America are related to each other in their own way. A methodology that looks at individual forms/tokens, i.e. a token-based approach, and traces back their history by looking at their geographical distribution, as well as correlations with socio-historical/archaeological/genetic phenomena, may shed more light on what we today consider a consented chaos. References: Eloranta, R. & L.M. Rojas-Berscia (sub.) Stones and grains as counting devices: Grammaticalization and lexicalization in sal-trading societies of Northern Peru. LUCL & Max Planck Institute for Psycholinguistics. Greenberg, J. H. (1960). General classification of Central and South American languages. En A. Wallace (Ed.), Men and cultures: Fifth international congress of anthropological and ethnological sciences (1956) (pp. 791-794). Philadelphia: University of Pennsylvania Press. Greenberg, J. H. (1987). Language in the Americas. Stanford, California: Stanford University Press. Kaufman, T. (1990). Language history in South America: What we know and how to know more. En D. L. Payne (Ed.), Amazonian linguistics: Studies in lowland South American languages (pp. 13-67). Austin: University of Texas Press. Kaufman, T. (1994). The native languages of South America. En C. Mosley & R. E. Asher (Eds.), Atlas of the world’s languages (pp. 46-76). London: Routledge. Loukotka, Č. (1968). Classification of South American Indian Languages. (J. Wilbert, Ed.). Los Angeles: Latin American Center, University of California. O’Connor, L., & Kolipakam, V. (2014). Human migrations, dispersals, and contacts in South America. En P. Muysken & L. O’Connor (Eds.), The Native Languages of South America (pp. 29-55). Cambridge: Cambridge University Press. Rojas-Berscia, L. M. (2013). La sintaxis y semántica de las construcciones causativas en el chayahuita de Balsapuerto (Licenciate Thesis). Pontificia Universidad Católica del Perú, Lima. Rojas-Berscia, L. M. (2015). Mayna, the lost Kawapanan language. LIAMES, 15, 393–407. Rojas-Berscia, L. M., & Ghavami Dicker, S. (2015). Teonimia en el Alto Amazonas, el caso de Kanpunama. Escritura y Pensamiento, 18(36), 117–146. Rojas-Berscia, L.M. & S. Roberts (in prep.) The history of pronouns in South America. Max Planck Institute for Psycholinguistics. Suárez, J. (1974). South American Indian languages. En The new Encyclopaedia Britannica (15th Edition). Chicago: Encyclopædia Britannica. Valenzuela, P. M. (2011). Contribuciones para la reconstrucción del Proto-Cahuapana: comparación léxica y gramatical de las lenguas jebero y chayahuita. En W. F. H. Adelaar, P. M. Valenzuela, & R. Zariquiey (Eds.), Estudios en lenguas Andinas y Amazónicas. Homenaje a Rodolfo Cerrón-Palomino (pp. 274-304). Lima: Pontificia Universidad Católica del Perú. Valenzuela, P. M. (2015). ¿Qué tan «amazónicas» son las lenguas kawapana? Contacto con las lenguas centro-andinas y elementos para un área lingüística intermedia. Lexis, 39(1), 5-56.

ABSTRACT. While all languages of western and central Arunchal Pradesh in northeastern India belong to the large Trans-Himalayan (Tibeto-Burman) phylum, the area is nevertheless one of extreme linguistic diversity. However, Blench and Post (2013) have made claims that there are multiple language isolates and micro-phyla in the West and East Kameng region of Arunachal Pradesh, but the data used in such assertions are almost exclusively lexical, and comparisons largely impressionistic. Unfortunately, general information resources (e.g., Glottalog) have adopted some of these, despite the fact that such claims remain undemonstrated nor indeed are even supported by the actual data. One such problematic case is the classification of Koro Aka. The Koro Aka are a very small group of people who have self-invisibilized within the larger Aka tribal group of the Kameng region of western Arunachal Pradesh. They consider themselves to speak a ‘dialect’ of Aka, a term which usually refers to the larger Hruso Aka people. A commonly heard belief in the area is that Koro and Hruso Aka are exactly the same people, “just a little bit different in dialect”. Syntactically similar, the languages have zero mutual intelligibility, and not many obvious cognates in common vocabulary (1). Hruso and Sajolang-Miji have also been claimed to be isolate languages by Blench and Post, but the fact that they both are clearly related to each other and to Bangru (Levai) as well, and that this Hrusish group they constitute belongs clearly within Trans-Himalayan, was demonstrated recently by Bodt and Liebherr (2015). Koro Aka on the other hand is clearly not closely related to Hruso or Miji (Anderson and Murmu 2010), despite local constructs in which Koro and Hruso are “dialects of the Aka tribe”. Koro does however have parallels with both Milang (Holon) spoken far to the east, and to Proto-Tani and indeed Proto-Trans-Himalayan, in some cases grammatical ones. There is however no evidence for a putative ‘Siangic’ family that is not part of Trans-Himalayan, which Blench and Post (2013) proposed for the shared correspondences such as (2), which they feel justifies positing an entirely new linguistic phylum. The two languages actually show very little common vocabulary not also found in other groups, and indeed all of the ones in (2) have parallels in other Trans-Himalayan languages, some not spoken in Arunachal Pradesh. If the two authors had considered grammatical data in their analysis (3), an obvious answer would emerge: Proto-Tani, Koro and Milang appear to be three sister languages in a ‘Macro-Tani’ group. Obvious cognates in case marking and TAM morphology, some with known Proto-Trans-Himalayan etyma, make the micro-phylum claim indefensible. The lesson learned is that one must consider all types of data, including key grammatical systems and cognates therein, when doing comparative-historical linguistics. This seems obvious but increasingly ones find proposals in linguistic taxonomy that clearly do not follow this. Nevertheless the basic tenets of the comparative method still remain, and we must systematically use all data sets that play roles in helping our understanding the historical development of individual languages and language groups, even as purely lexical based statistical methods and analyses increasingly expand their role in linguistic phylogenetics or taxonomy as practiced today. As such, expanded data sets must be used that include grammatical/syntactic forms of words to establish similar computationally and statistically valid observations for these type of data as well which in turn feed our models of linguistic phylogeny and comparative historical linguistic methods as a whole. This is one of several similar cases we discuss in our presentation from this region.

References Anderson, G. D. S. and G. Murmu. 2010. Preliminary notes on Koro: a 'hidden' language of Arunachal Pradesh. Indian Linguistics 71: 1-32. Blench, R. and M. Post. 2013. Rethinking Sino-Tibetan phylogeny from the perspective of North East Indian languages. In Thomas Owen-Smith and Nathan W. Hill (eds.) Trans-Himalayan Linguistics. Berlin: de Gruyter, pp. 71-104. Bodt, T.A. and Lieberherr, I., 2015. First notes on the phonology and classification of the Bangru language of India. Linguistics of the Tibeto-Burman Area, 38 (1): 66-123. DeLancey, S. 1984. Etymological notes on Tibeto-Burman case particles. Linguistics of the Tibeto-Burman Area 8 (1): 59-77. Matisoff, J. 2003. Handbook of Proto-Tibeto-Burman. University of California Publications in Linguistics 135. Berkeley: University of California Press. Post, M. and Y. Modi. 2011. Language contact and the genetic position of Milang (Eastern Himalaya). Anthropological Linguistics 53 (3): 215-258. Sun, Jackson T.S. 1993. A Historical-Comparative Study of the Tani (Mirish) Branch of Tibeto-Burman. Ph. D. Dissertation, University of California, Berkeley. Tayeng, A. 1976. Milang Phrase Book. Shillong: Directorate of Information, Arunachal Pradesh.

Examples (see attached)

ABSTRACT. When reconstructing the histories of words in a language family, it is customary to reconstruct each word to the lowest-level subgroup that covers all of the languages that show cognates of that word. Yet identifying the subgroups in a language family itself involves examining the distributions of cognates and making assumptions about how broadly they were once shared. One way around this potential circularity is to reason on the basis of geography: if the languages on the periphery of a certain geographic area reflect cognate set A for a given meaning, while the languages in the center reflect cognate set B, then it is likely that cognate set A represents the ancestral state, which once held throughout the region, and that B is a later innovation. This type of inference allows linguists to determine how “high up” in a language family a given word should be reconstructed, but without needing to construct a genealogical tree beforehand, and even without assuming that this genealogy must be tree-like. In this presentation, we propose a computational (hence easily scaled-up) version of this type of reasoning, and test it with data on 950 lexical isoglosses from 20 Germanic languages. In our approach, rather than using geography to determine which languages are “between” which others, we use linguistic distances, in the following manner: Each language is represented as a vector of 0s and 1s indicating its membership in various isoglosses, and the distance between two languages is the Euclidean distance between these two vectors; given two languages A and B, a language C is defined to be “between” A and B if C is at least as close to A and to B as A and B are to each other (a definition inspired by Hammarström & Güldemann 2014). For each isogloss, we take every language that is not already in the isogloss, and check whether it lies between two languages that are already in the isogloss; if so, we assign this language a value of “NA” (rather than 0), indicating that it may have once been part of the isogloss (but we don’t know for sure). We then use Bayes’ theorem to estimate a probability (between 0 and 1) for each NA value, aggregating the evidence from each language that is known to belong to the isogloss (i.e. each value of 1) and each language that is known not to belong (i.e. each value of 0). The result is a (probabilistic) reconstruction of the “original scope” of each isogloss. This enables us to infer when a word can be plausibly reconstructed to Proto-Germanic, even if it is not synchronically attested in all Germanic languages in our sample: for example, in our data, something resembling PGmc *fellan ‘skin’ is only attested in four languages (Gothic, Dutch, Afrikaans, and Limburgs), but is reconstructed with more than 92% confidence throughout all the Germanic languages. Likewise, a word attested within a geographically-restricted subset of Germanic languages will generally only be reconstructed to that subset: for example, forms similar to [flɛɾk] for ‘wing’ are only found (in our dataset) in Afrikaans, Low Saxon, and varieties of West Frisian, yet this is reconstructed (correctly) with 95% confidence to Dutch, and with 67% confidence to Limburgs, both of which are in the same geographical area. Crucially, it is reconstructed with low confidence (if at all) to the other languages.

ABSTRACT. In this talk I will discuss some aspects of the reorganization of voice distinctions in Late Latin, focusing on the actualization of a number of ‘new’ structures which became available for the encoding of voice in the transition from Latin to (Italo-)Romance, their various stages and the parameters determining their implementation, in relation to (i) the role played by aspect, the continuum of control (Lehmann 1988) and the person hierarchy, (ii) the diachronic relationship between auxiliation and serialization, (iii) the direction(ality) of the changes.

The ultimate goal is to detect patterns of invariance (i.e., persistence), of Latin inheritance and principled differences (i.e., divergence) in the type and extent of variation and further developments in this area of Romance morphosyntax.

More specifically, I will investigate the grammaticalization of lexical verbs of motion (come), activity (do/make) and change of state (become) as markers of the passive voice in the transition from Latin to (Italo-)Romance, in relation to the status of serial/light verbs, — whether intermediate stages in the auxiliarization process (Rosen 1977; Giacalone Ramat 2000; Giacalone Ramat & Sansò 2014, among others) or a different syntactic category (Butt 2003, Butt & Lahiri 2013) — and to the linearity of the relationship between serial verb and auxiliary (following the path verbal lexeme > serial > auxiliary) (Heine 2003, Hopper & Traugott 2003 and recent discussion in Bisang 2011; Börjars & Vincent 2011, among others).

I will argue that, although characterized by maximal desemanticization on a par with auxiliaries, the serial uses of the verbs under investigation, both in Late Latin (e.g., coctus factus cook.pp.m.sg become.pp.m.sg ‘cooked’) and in some early Italo-Romance vernaculars (e.g., O. Lombard strangosada facta anguish.pp.f.sg make.pp.f.sg ‘anguished’) seem to exhibit a different type rather than a reduced degree of decategorialization. I will also show that the relationship serial verb-auxiliary is non-linear: the same lexeme, in fact, can have simultaneously auxiliary and serial uses, the latter developing, for some verbs, after their auxiliary uses (Cennamo 2006; 2007).

I will also investigate the reanalysis of the reflexive morpheme as a voice modulator. This appears, rather, to follow a linear path, proceeding from anticausative to passive and optionally to an impersonal/indefinite reinterpretation. This last stage is attested to different extents in Romance, and not in the same range of constructions in every language, depending on the referential scope of the reflexive morpheme and its degree of grammaticalization, as witnessed, for example, by Italo-Romance (Cennamo 1993; 2014; 2016). These changes indeed give evidence for the passive and impersonal reinterpretation of reflexive patterns as reflecting two different diachronic paths, which at some point merge. While the passive reflexive is a Late Latin development, related to changes in the encoding of voice in the transition from Latin to Romance, the impersonal function of the reflexive with one-argument verbs is a Romance phenomenon,  and  one not equally attested across the Romance languages or even within the Italian dialects. It probably reflects a stage where the reflexive pronoun acquires a non-reflexive pronominal value, such that Latin se becomes equivalent to is ‘he’, while also coming to refer to first and second person participants (se = nos ‘we’, vos ‘you’) (Cennamo 1993: 81; 2014).

References

Bisang, W.  2008. Grammaticalization and the areal factor- the perspective of East and mainland Southeast Asian languages. In M.J. López & E. Seoane (eds), Rethinking Grammaticalization. Amsterdam/Philadelphia: Benjamins, 55–88.

___ 2011. Grammaticalization and linguistic typology. In H. Narrog & B. Heine (eds), 105–117.

Börjars, K. & Vincent, N. 2011. Grammaticalization and directionality. In H. Narrog & B. Heine(eds), 163–176.

Butt, M. 2003. The light verb jungle. Harvard Working Papers in Linguistics 9: 1–49.

Cennamo, M. 1993. The Reanalysis of Reflexives: a Diachronic Perspective. Naples: Liguori.

___ 2006. The rise and grammaticalization paths of Latin fieri and facere as passive auxiliaries. In W. Abraham & L. Leisiö (eds), Passivization and Typology. Amsterdam/Philadelphia:Benjamins, 311–36.

___ 2007. Auxiliaries and serials between late Latin and early Romance. In D. Bentley & A. Ledgeway (eds), Sui Dialetti Italo-Romanzi. Saggi in onore di Nigel B. Vincent. The Italianist, Special Supplement 1,  63–87.

___ 2014. Passive and impersonal reflexives in the Italian dialects: synchronic and diachronic aspects. In P. Benincà, A. Ledgeway & N. Vincent (eds), Diachrony and Dialects. Grammatical Change in the Dialects of Italy. Oxford: Oxford University Press, 71–95.

___ 2016. Voice. In M. Maiden & A. Ledgeway (eds), The Oxford Guide to the Romance  Languages. Oxford: Oxford University Press, 967-980.

Giacalone Ramat, A. 2000. Some grammaticalization patterns for auxiliaries. In J.C. Smith & D. Bentley (eds), Historical Linguistics 1995, vol. 1. Amsterdam/Philadelphia: Benjamins, 125–38.

Giacalone Ramat, A. and Sansò, A. (2014) “Venire (‘come’) as a passive auxiliary in Italian” in M. Devos & J. van der Wal (eds), Come and Go off the Beaten Grammaticalization Path, Berlin: Mouton de Gruyter, 21-44.

Heine, B. 1993. Auxiliaries. Cognitive Forces and Grammaticalization. Oxford: Oxford University Press.

___ 2003. Grammaticalization. In B. Joseph & R. Janda (eds), The Handbook of Historical Linguistics. Oxford: Blackwell, 576–601.

Heine, B.  & Kuteva, T. 2011. The areal dimension of grammaticalization. In H. Narrog  & B. Heine  (eds), 291–301.

Hopper, P. & Traugott, Elisabeth C. 2003. Grammaticalization. Cambridge: Cambridge University Press.

Ledgeway, A. 2011. Grammaticalization from Latin to Romance. In H. Narrog & B. Heine (eds), The Oxford Handbook of Grammaticalization. Oxford: Oxford University Press, 717–726.

Narrog, H. & Heine, B. (eds) 2011. The Oxford Handbook of Grammaticalization. Oxford: Oxford University Press.

Rosen, C.  1997. Auxiliation and serialization: on discerning the difference. In A. Altsina, J. Bresnan & P. Sells (eds), Complex Predicates. Stanford: CSLI.

ABSTRACT. This paper aims at contextualizing the study of Texas German within the larger fields of historical linguistics and language contact.  It will show how each of the three main large-scale studies of Texas German (those by Fred Eikel, Glenn Gilbert, and the Texas German Dialect Project) was very much a product of its scholarly era.  To take up just the last-named of these projects, for instance, recent TGDP studies have engaged with scholarly questions that were not part of earlier discourse, e.g. the possible creole status of Texas German, the role of social networks in Texas German language change, and phonological markedness and sound change in Texas German.  It will also discuss some possible future directions for the study of Texas German.

ABSTRACT. How and why do people and communities shift from speaking a minority language to become bilingual in the majority language, and then shift entirely to the majority language? An emerging view connects shift to changes in community structure captured in Warren’s “Great Change” (1978). The heart of this approach attributes shift to a change from local control of tightly interconnected institutions to more external or ‘vertical’ control of those increasingly less-interdependent institutions. This model has been tested using data from Cherokee in North Carolina (Frey 2013), American varieties of German (Lucht 2007, Salmons 2005a, b), Frisian (Bousquette 2016), Norwegian (Natvig 2017), Finnish (Johnson forthcoming), and Somali (Brown & Carpenter forthcoming). While previous approaches to language shift prove unfalsifiable (‘ethnolinguistic vitality’) or invoke community-specific factors that do not correspond to the actual progress of shift (World War I for American German), verticalization provides a general and falsifiable theory.

After introducing evidence on the timeline of shift, I show that verticalization in Texas German communities correlates with shift to English, including:

• Increasing centralization of school policy with regard to both public and private schools,
• Economic changes that make German-language publications unviable,
• Changes in religious institutions that make it difficult and eventually impossible to provide German-speaking clergy to congregations,
• Technological and related changes that reduce local autonomy and increase interactions across communities,

These and related social-demographic changes reduce local autonomy and disrupt formerly German-speaking social networks.

In short, Texas German provides important evidence for verticalization as the driving force in language shift.

ABSTRACT. In 1851 an article appeared in the Geographisches Jahrbuch (Geographic Yearbook, Gotha) claiming to establish definitive connections, using language observations, among the Comanches, Shoshones, and Apaches. Titled "Über die Verwandtschaft der Schoschonen, Komantschen und Apatschen" (“On the Relationship among the Shoshones, Comanches, and Apaches”), the article, authored by Heinrich Berghaus, was based on lexical data gathered by a young German settler in Texas and included an original list of 366 Comanche words and their German translations. Probably because it was published only in German, Berghaus’s article has been almost completely ignored by U.S. scholarship, even though it offers valuable insights into Native American languages and stands as a monument to German language and its outreach in Texas .

The first substantial, published, Comanche dictionary was therefore a Comanche-German dictionary. This presentation provides a sketch of the background to Berghaus’s article and lexicon. It further assesses the nature and value of the article’s contents and offers an appraisal of its significance both for linguistic research with respect to Comanche and as a document that foregrounds language contact between German and Comanche in mid-19^th century Texas.

ABSTRACT. In 1851 an article appeared in the Geographisches Jahrbuch (Geographic Yearbook, Gotha) claiming to establish definitive connections, using language observations, among the Comanches, Shoshones, and Apaches. The author of the article, Heinrich Berghaus, was a well-established cartographer but had no history of original anthropological research. Titled “Über die Verwandtschaft der Schoschonen, Komantschen und Apatschen” (“On the relationship of the Shoshones, Comanches and Apaches”), the article was based on lexical data gathered by Emil Kriewitz, a young German settler in Texas,and included a list of 366 Comanche words and their German translations, along with much valuable ethnographic detail. It appeared at a time when geography, ethnology, and linguistics were forming and differentiating as modern disciplines. Here we consider, for the first time, Berghaus’s and Kriewitz’s influence on the course of ethnology and linguistics in North America.

ABSTRACT. Background to my study of Texas German 1961-1970:

LSA Linguistic Institute, Austin, summer 1961

Preparation of my dissertation “The German Dialect Spoken in Kendall and Gillespie Counties, Texas”, Harvard University, 1963

Influences of Schlerath, Lenneberg, Chomsky, Halle, Labov, Seifert, Kurath, Wenker, Gilliéron Questionnaire, tape recordings, graduate-assistant fieldworkers, preparation of the Linguistic Atlas of Texas German at the Deutscher Sprachatlas, Marburg

We were dealing with language attrition and death. Nevertheless, at the time there were an estimated 70,000 fluent speakers in the 30-county area, divided between West (“Hill Country”) and East.

Greatest surprises: In the West we discovered the still lively Alsatian dialect area, descendants of Castro’s Colony, in Medina County, very different from the rest of the “Texas German” koiné. The East had unexpectedly many Slavic speakers – Czech and Polish were widespread, Wendish (Sorbian) had disappeared in the previous generation after a period of multilingualism including SG/TG.

Although outside the 30-county area, almost no German-speaking communities had survived as of the 1960s, as shown by the results of a mail survey we distributed throughout the state in 1966 (now deposited with Boas’ TG project in Austin), a geographical study of German surnames, especially in connection with land ownership (Census and other sources) might well indicate where the German communities were once located and where language shift has been completed.

In the study of TG syntax: We focused on the dative/accusative merger (first described by Eikel), compiling the data on a geographic basis and considered possible explanations for it. “Lieber Gott, ich bidde Dich, mach ein gutes Kind aus mich” (1964, David Frederick Walter, one of my graduate students from Fredericksburg, remembering a childhood prayer). Was this a reflex of the speakers’ original dialects, a general tendency in the West Germanic languages or of languages in general, or the influence of English? Or a combination of two or more of these factors? In any case, TG was reenacting what happened in Old English in the centuries before the Norman conquest.

Within the central 30 counties, the area of Comfort and Boerne merits special attention because of its “Freethinker” (Freidenker) heritage. Standard German held on there longer than most other places. An 80-year old woman told me in 1962 there are two kinds of German, the German in the books and the German we use here. The first says ‘wuerde’ and the second ‘taete’ (e.g., sie sagte, dass sie kommen wuerde vs. sie sagte, dass sie kommen taete), and she gave me many more examples using other alternates. She said these are two ways to say the same thing and that it is the same language. By the way, in no form of TG that we encountered did we observe syntax like, *sie sagte, dass sie wuerde kommen. Why would English have influenced dat/acc but not the German word order in subordinate clauses?

In phonology: We looked especially at the loss of front rounded vowels (also noted by Eikel) and the acquisition of retroflex R, which was unfortunately only mapped in part of the area. The same explanations could be invoked for the unrounding. Retroflex R is almost certainly from English.

Explanations for the historical longevity of bilingualism in Central Texas, up to seven generations: sparse population spread out over long distances; poor schooling, which weakened the use as of education as a tool to enforce monolingualism in English; the presence of sizeable numbers of Spanish speakers, a “colonial language” like English (Kloss), which acted to retard the quick abandonment of German; the influence of religion (or no religion -- die Freidenker) and politics (“Treue der Union” in Comfort, which by the way is said to be named after Bad Comfort in the Rheinland).

The future: We are fortunate to have an unusually large amount of data on Texas German stretching from Eikel in the 1950s to Boas currently, almost 70 years. After Pennsylvania German, more is known about TG than any other form of immigrant German in the U.S. Almost certainly, the few “Semi-Speakers” who employ TG on occasion will be replaced in the next generation by English-monolingual people who nostalgically look on it as the language of their ancestors. Therefore in the future, and probably in most cases right now, we should study TG via the analytical tools described by the growing literature on language attrition and death. It is not a process to regret but a textbook example of processes we can learn from. This is exactly what Boas proposes to do in adapting Trudgill’s analysis of the development of English in New Zealand to explain the evolution of German in Texas.

ABSTRACT. The presentation will focus on terms like Sprachinsel, linguistic island, language island and related expressions. These have been in use for a long time in traditional German dialectology and in recent sociolinguistic research. However, in many cases these expressions focus too much on an insularity of the speech community in question. This concept of insularity is in many cases (e.g. when it comes to Texas German) not adequate as it may block a more nuanced view of the actual linguistic situation.

With regard to German speech minorities, it is obvious that terms connected with a concept of insularity only fit in rare situations, for example when the focus is on German settlements in Northern Italy. In many other situations – for instance regarding Texas German - terms like colony, foreign language unit or linguistic minority provide a more appropriate description of the sociolinguistic reality.

If the focus is on different German language groups in North America, the classic concept of a Sprachinsel / linguistic island becomes blurred. The talk will argue for a more specific use of this term and related ones. This will be done using linguistic data from several German speech minorities. Also, the historic dimension of the term linguistic island and new formations of speech communities will be analyzed.