Projektist

A typological shift can be observed at the level of a single language or group of languages by studying a single structural feature, a group or type of features, or the overall picture of the grammatical structure of a language (see, e.g., Hewson 1996, Hintz 2016, Heine & Kuteva 2006). This project has a broader approach: we investigate the structural differentiation of the western subgroup of the Uralic language family – the Southern Finnic (SF) languages. This study combines perspectives of areal and intragenetic typology and relates them to the context of the languages of Europe and the Central Baltic (CeB) area as well as the broader Uralic family.

The Finnic and Saami languages form the westernmost groups of the Uralic language family, together with Mordvin, Mari, and Permic they contrast with the eastern Mansi, Khanty, and Samoyed languages (see Saarikivi 2022). Considering both western and eastern Uralic languages in general, the phonology and grammar of many Uralic languages appear to be typologically similar: they have relatively simple syllable structures, word initial stress, front-back vowel harmony, agglutinative morphology, right headed NPs, SOV word order, etc. (see Comrie 1988, Abondolo & Välijärvi 2023). However, Finnic and Saami stand out as having shifted towards becoming fusional languages, whereas other Uralic languages show more stability as agglutinative (Laakso 2021).

The Southern Finnic languages (Estonian, South Estonian, Livonian, Votic) are atypical Uralic languages in several respects. For example, they reveal complex phonological alternations, basic SVO word order, innovative means for marking reported evidentiality and presenting commands (Metslang 2009, Laakso 2021). Their phonological systems, for example, show three-way quantity oppositions as well as tonal contrasts and specific ways of interaction between word prosody and intonation (Lehiste et al. 2008, Balodis et al. 2016, Pajusalu 2012, 2022). These unique phenomena have not yet been systematically studied. Additionally, there is a lack of studies that would include a significant amount of features belonging to different structural domains (phonology, morphology, syntax) and would take a micro-areal perspective aiming at maximally good coverage of the language varieties in the area.

Areal typology contributes to wordscale typology by identifying microvariation in a particular area (Daniel 2012, Hickey 2017). The SF area with its southernmost varieties (Courland and Salaca Livonian; Kraasna, Lutsi, Leivu, Mulgi, Võro, and Seto South Estonian) and non-cognate contact varieties (Baltic, Slavic, and (historical) Germanic varieties) together form the CeB language area, which is characterised by a number of shared features in phonology, morphology, and syntax (Norvik et al. 2021, Wiemer et al. 2014 for shared features with Finnic, Baltic, and Slavic). It can be seen as a part of the wider Circum-Baltic area, the similar language features and innovations of which have attracted greater attention since the works by Stolz (1991) and Dahl & Koptjevskaja-Tamm (2001a, b). No bundles of features have been found uniting the entire Circum-Baltic area, which has led to discussion of a contact superposition zone and convergence on the microlevel (Koptjevskaja- Tamm & Wälchli 2001). The same also holds for the CeB area (Norvik et al. 2021). This could reflect the fact that there are differences to what extent

Finnic and Saami have been influenced by western non-cognate contact languages (see, e.g., Larsson 2001, Tauli 1964, Korhonen 1996 [1980]).

As a result of language contact, structures and inventories of languages tend to become more similar. This may be caused by an increase in formal similarity (matter borrowing), structural similarity (pattern borrowing), or their combination (Matras & Sakel 2007, Gardani 2020). The borrowability of the various elements of language is different; however, borrowability scales can be established (Thomason & Kaufman 1988, Matras 2009, Gardani 2012). The internal preconditions of a language determine the likelihood of adapting innovations from a contact language and predict the position of an innovation on a borrowability scale (see, e.g., Thomason & Kaufman 1988). The study of a multitude of SF special features in comparison with the neighbouring Indo- European languages will provide valuable data for testing borrowability scales and identifying various types of language-internal preconditions of change.

A language is also influenced by factors related to its role in society: the presence of a standard language, its development and areas of use; the use and status of different languages in society; the use of the language as an L1 or L2. In this respect, SF languages represent different situations. (see, e.g., Keevallik & Pajusalu 1995, Metslang 2022, Heinsoo & Saar 2015)

Typological shift may also occur as a result of language-internal factors (Aikhenvald 2006). For example, the ternary alternation of phonological quantity (Q) increased the possibility of expressing inflectional forms fusionally without case suffixes, cf. Estonian Q2 kooli [ko:li] ‘of school’ and Q3 kooli [ko::li] ‘to school’; in this alternation of long and overlong quantities, a tonal contrast is also significant (Lippus et al. 2009, 2011, Asu et al. 2016: 141). Influence of Baltic or Scandinavian contacts is plausible in the development of the tonal contrast; however, for the phonologisation of this complex prosodic alternation, there should be language-internal compatibility (Pajusalu 2012, Prillop et al. 2020).

One of the central tasks of this project is to take a closer look at the proposed shift towards a (more) fusional morphological type, which at least partly could be attributed to contacts. For Estonian, extensive typological changes in prosody and inflection are traditionally explained by the influence of contact languages, especially Middle Low German (Prillop et al. 2020: 30). The formation of fusional traits in Livonian is related to Latvian or older Germanic contacts (Grünthal 2015). Languages differ to the extent that the features representing different types are present; this also concerns the SF languages (Viitso 1990, Wälchli 2000, Laakso 2021). As a result of an increase in synonymy and redundancy, complexity is seen to increase (Dahl 2004).

To study typological shift in SF and Ingrian, new comprehensive datasets are used, such as Grambank and UraTyp (https://uralic.clld.org/; see Norvik et al. 2021, Skirgård et al. (submitted), and the description of the datasets below). As an example, in a study based on the UraTyp data, a division into phonology, morphology, and syntax revealed that close cognate languages may show different groupings depending on the domain (see Norvik et al. 2022a). An in-depth micro-areal study is needed to give further insight into the behaviour of close contact languages. The combined use of both quantitative and qualitative research methods makes it possible to explain the development and nature of typologically rare features in SF languages, and contribute to the general understanding of complex typological changes.

This project focuses on three SF languages – Estonian, South Estonian, Livonian – and also Ingrian as a Northern Finnic (NF) language with several innovative typological features similar to SF. All of the languages, excl. Estonian, are severely endangered, which makes their study an urgent task. Presently, there are only around twenty L2 speakers of Courland Livonian. Courland Livonian forms the basis of Standard Livonian (see Ernštreits 2012, 2013). The second main variety of Livonian, Salaca Livonian, became extinct in the 19th century but its main data and grammatical overview have recently been published (Winkler & Pajusalu 2016, 2018). The sociolinguistic situation and structure of Livonian have been extensively studied during the last decade (e.g., Pajusalu 2014, Norvik 2015, Kallio 2016, Kiparsky 2017, O’Rourke 2018); however, there are few comparative analyses of Livonian (cf. Wälchli 2000, Grünthal 2003, 2015).

In this project, there is also a special research emphasis on South Estonian language enclaves: Leivu and Lutsi in Latvia and Kraasna in Russia (see Balodis & Pajusalu 2021, Weber 2021). The Kraasna variety was spoken until the first decades of the 20th century; the last fluent L1 speakers of Leivu and Lutsi died in the relatively recent past, but local communities are active in preserving their heritage languages (Balodis 2021). South Estonian was the first language that diverged from Proto-Finnic (Sammallahti 1977, Kallio 2014, Prillop et al. 2020), Livonian was the next to diverge after South Estonian (Viitso 1985). Leivu was the first dialect that separated from the South Estonian language unity (Kallio 2021); Leivu also has old contacts with Livonian (Viitso 2009). Therefore, Leivu and Lutsi are important for understanding the formation of the SF language area both from historical and typological perspectives (see Norvik et al. 2021).

Ingrian is one of the youngest NF languages (Kallio 2007: 243; Pajusalu 2014; Prillop et al. 2020) as well as geographically the closest NF language to South Estonian. Interestingly, Ingrian and the SF languages, incl. South Estonian, share certain (morpho)phonological and morphosyntactic innovations and peculiar typological features (see, e.g., Markus et al. 2013, Saar 2017, 2019). As Ingrian is one of the least studied Finnic languages, its relationship to other languages in the family requires closer analysis.

The joint study of different language domains and their possible interrelationship has received relatively little attention in typological studies (see, e.g., Plank 1998, Luraghi 2017, Klumpp et al. 2018, Miestamo 2018). Emergence of new large-scale databases such as Grambank and UraTyp opens up new possibilities, including area-specific studies. This project allows us to shed more light on the SF area as there are no previous studies systematically investigating the languages of this area as we intend to do here.

Header image: Leivu language field notes recorded by Heli Tõugjas from Kārlis Būmanis in Ilzene, Latvia, 1956. (Source: University of Tartu Archives of Estonian Dialects and Kindred Languages, record number T0258)