Loengusari TÜling

Eesti kirjakeel on umbes 500 aasta vanune. Esimesed tekstid olid valdavalt kiriklikud, aga juba paarsada aastat tagasi kirjutati eesti keeles muu hulgas ka loodushoiust, ilmastiku muutumisest, vaktsineerimise vajalikkusest, sugulaskeeltest, kaugetest maadest, õhupallidest ja rakettidest. Kohati jäävad need tekstid tänapäeva lugejale arusaamatuks, sest nii sõnavara, vormistik kui ka kirjaviis on muutunud. Käesoleval aastal alustasime vana kirjakeele veebisõnastiku loomist. Sõnastiku eesmärk on avada nii tähenduste arengut kui ka morfosüntaktilisi, morfoloogilisi ja fonoloogilisi muutusi eesti kirjakeele vanemas ajaloos. Ettekandes näitan vanu tekstikatkendeid ning tutvustan loodavat sõnastikku.

Alzheimer's disease is the most common form of dementia, contributing to 60-70% of all dementia cases. There are over 50 million dementia sufferers worldwide, and due to ageing population, the number of dementia sufferers is expected to triple by 2050. Recent research has shown that the changes in speech and language may be the earliest signs of cognitive decline, and that these changes can appear as long as decades before dementia diagnosis. Promising results have been achieved in detecting changes in speech and language (for example, in acoustics, fluency, syntax complexity, information density and semantics) automatically, using natural language processing (NLP), signal processing (SP) and machine learning (ML). Developing the technology for automatic Alzheimer's disease detection from speech and language could make screening for Alzheimer's disease cheaper, faster and more accessible, as the potential signs could be spotted using, for example, a mobile application. Detecting dementia early would allow starting treatment early, and improve the life quality of the elderly. In my presentation, I will give an overview of the studies that focus on automatic Alzheimer's disease detection from speech and language.

Among languages of the world, quotative markers sometimes develop new functions and follow the grammaticalization path (verb ‘say’ >) quotative marker (> complementizer >) reason/purpose clause marker. The grammaticalization of quotatives into reason/purpose-clause linkers implies stages where both meanings can simultaneously arise within one construction, e.g.: my father always talked me out of it, saying/because I don’t have a gift for it anyway. The reported discourse describes the purpose/reason of an action or its lack and transfers the quoted speaker’s perspective.
In this talk, first I look at the cases of such ambiguity in the use of grammaticalized quotative particles in reported discourse constructions of six Finno-Ugric languages (Hungarian, Estonian, Finnish, Erzya, Udmurt, Komi). Second, I check for such use among quotative markers in English and Russian. I argue that to achieve a reason and purpose meaning quotative markers and RD need to follow particular syntactic conditions, e.g. possibility to use the quotative marker as a single quote-introducer, preference to indirect reported discourse, etc.

Since the advent of the digital age, digital tools and computational approaches to linguistics have been and continue to be developed for linguistics research and in support of majority languages. While not the most common target of such language technology, endangered minority languages also stand to benefit from such advances. Technological advances in linguistics research thus have the potential to have real-world, practical effects in addition to theoretical outcomes, even in the context of critically endangered languages. In my talk, I will present a case study of how this can work, specifically for Pite Saami, a critically endangered Uralic language spoken by around 35 individuals from Swedish Lapland and (historically) adjacent areas of Norway. I will provide an overview of digital, lexical and grammatical tools developed for automatic corpus annotation for Pite Saami, and then I will discuss how these are being used to support revitalization efforts and promote an increase in language use, thereby hopefully increasing the chances of the Pite Saami language surviving beyond current generations.

In this talk, we will present the outcome of our work on the morphological dictionary of Soikkola Ingrian. Ingrian is a highly endangered northern Finnic language spoken in the Leningrad oblast of the Russian Federation. We will talk briefly about the goals we pursued while compiling the dictionary, and then present the electronic version and discuss its functions. The electronic dictionary of Ingrian is freely accessible; the current version can be downloaded from http://www.ingrian.org/Ingrian-dictionary/.

During both world wars, science institutions in Austria-Hungary, Finland, Estonia, and Germany made use of the situation in which representatives of different nations and languages were accessible in prisoner-of-war camps, and collected data for anthropological, linguistic and musicological research. Relations between the researchers and the researched were of various kinds, and the collected data were edited and published with different levels of rigour. The documentation of languages was probably the most successful discipline, which seemed less burdened by too high expectations. Due to technical progress making possible the recording of sound, these documentation activities produced touching sound documents, which, however, are of less extent and scientific value than the written documents. The Finno-Ugric languages of Russia as in the first half of the 20^th century are today documented in corpora which contain to a substantial degree contributions by prisoners of war. This refers equally to texts and lexical resources. 

Teen ettekandes ülevaate Läti riigi kõrghariduskeelepoliitika praegustest suundumustest tasakaalu otsimisel läti keele ja teiste kõrghariduses kasutatavate keelte vahel. Kõrgharidus on Lätis enamjaolt lätikeelne, kuid muidugi on üleilmastumise ja kõrghariduse rahvusvahelistumise mõjul hakatud inglise keelt valdkonnas kasutama tihemini kui varem. Nii nagu mitmel pool mujalgi on ka Lätis riigikeele ja inglise keele kasutamine kõrghariduses teema, mille üle arutletakse ja mis aeg-ajalt ka eri osapoolte vahel pingeid tekitab. Riigi keelepoliitika elluviijad on seni seisnud läti keele mitmekülgse kasutamise eest kõigis kõrgharidusastmeis (Rozenvalde 2018), kuid viimasel ajal on ka riigi tasemel tehtud ettepanekuid inglise keele eelistamiseks mõnes olukorras. Lisaks on riigi kõrghariduskeelepoliitikas pikka aega pingeid tekitanud vene õppekeele laialdane kasutamine erakõrgkoolides, millele riik püüab ideoloogiliselt neutraalseks jäädes piiri panna.

Praegu valmistab Läti riik kõrghariduspoliitikas ette mitut reformi, mille elluviimine mõjutaks keelekasutust Läti kõrgkoolides. Keelekasutuse vaatevinklist võibki kavandatavad ümberkorraldused jagada kaheks: esiteks muudatused, mille raames arutletakse läti ja inglise keele kasutamise üle kõrghariduses, ja teiseks reformid, mis mõjutaksid läti ja vene keele kasutamise tasakaalu erakõrgkoolides. Minu eesmärk ettekandes on võrrelda, kuidas seadusandjad põhjendavad neil kahel suunal kavandatavaid muudatusi. Selleks analüüsin ettekandes 2018.–2020. aastal riigi tasemel koostatud ettepanekuid muudatuste elluviimiseks kõrghariduses, mis on seotud keelekasutusega, ning muid nende ettepanekutega seotud dokumente: Läti haridus- ja teadusministeeriumi kontseptsioone, teatiseid ja pressiteateid, seaduste muudatusettepanekuid koos selgitustega, põhiseaduskohtu lahendeid. Analüüsiks kogutud materjalis põhjendavad reformide elluviijad oma ettepanekuid, et veenda nende vajalikkuses teisi ühiskonnaliikmeid, sh seadusandjaid. Analüüsin materjali argumentatsioonianalüüsi meetodi abil, mis võimaldab kirjeldada ja hinnata seisukohti ja argumente, mille on esitanud erinevad osapooled, kes riigi poliitikat ellu viivad (Boreus 2017; Eemeren & Henkemans 2016; Liakopoulos 2000).

Viited

Boreus, Kristina. 2017. Argumentation Analysis. In Kristina Boreus & Göran Bergström (eds.) Analysing Text and Discourse. Eight Approaches for the Social Sciences. Los Angeles, London, New Dehli, Singapore, Washington D.C., Melbourne, 53–85.

Eemeren, Frans H. & A. Francisca Snoeck Henkemans. 2016. Argumentation. Analysis and Evaluation, 2^nd edition. New York & London: Routledge.

Liakopoulos, Miltos. 2000. Argumentation analysis. In Martin W. Bauer & George Gaskell (eds.) Qualitative researching with text, image and sound. A practical handbook. Los Angeles, London, New Dehli, Singapore: Sage Publications, 152–171.

Rozenvalde, Kerttu. 2018. Multilayered language policy in higher education in Estonia and Latvia: Case of National Universities. Summary of doctoral thesis. Riga: University of Latvia.

In this talk, we introduce the project bwrite – Academic Writing in the Baltic States: Rhetorical Structures through Cultures and Languages – which aims to map the academic writing traditions in the national languages of the Baltic States – Estonian, Latvian, and Lithuanian.
As writing is one of the primary means for the evaluation of students, dissemination of research, and the evaluation of researchers, it has become central in the academe. Specifically at the time when English as the lingua franca of science is gaining more ground, investigating and supporting academic writing in local languages is vital. Academic writing in different languages, however, is tied to specific conventions. For example, it has been found that more than language structure, L2 writers are affected by their cultural background (Connor et al., 2008). In other words, the way we communicate our thoughts dependents on (local) rhetorical conventions (e.g. what is considered to be polite etc.). Thus, in this project we keep to the assumption that the way academic texts are written is culture-dependent.
While quite a large body of research is available on writing traditions in languages with large populations of speakers (e.g. Martín & León Pérez, 2014; Siepmann, 2006), less is known about writing traditions associated with languages spoken by fewer people. As such, we are largely missing an understanding of rhetorical conventions when it comes to academic writing in the national languages of the Baltic States.
The current project aims to map rhetorical conventions of academic writing in Estonian, Latvian and Lithuanian. To achieve our ambitious aim, our first step is to build a model with features that captures any given writing tradition. We will present our initial feature model containing 5 features: Argumentation, Stance, Rhetorical Structure, Self-Promotion, and Coherence/Cohesion, and proceed to elaborate how these features are manifested on the macro-level, meso-level, and micro-level (Fetzer, 2013) across a range of academic genres.
 

Ettekandes võrreldakse eesti keele mõju ühelt poolt vanausuliste ja muudest põlisvenelastest elanike vene keele variandile ning teiselt poolt nõukogude aja sisserändajate ja nende järglaste vene keele variandile. Küsimus, kas eri sotsiolingvistilises olukorras kahe keele kontaktide tulemused on sarnased või mitte, ei ole sugugi triviaalne. Mõne keeleteadlase arvates on määravad formaalsed tegurid (keelte struktuur) ja sellest lähtudes tulemused peaksid olema enam-vähem samasugused. Samas on teada, et sotsiolingvistilised tegurid (keelte staatus, sotsiolingvistiline olukord, keeleideoloogiad jms) ja individuaalsed tegurid (hoiakud, kõnelejate taust, keelte oskus) võivad osutuda formaalsetest teguritest kaalukamaks. Ettekandes pakutakse kognitiivse kontaktlingvistika vaatenurka: kuigi nii formaalsed kui sotsiolingvistilised tegurid on olulised, inimkognitsiooni omaduste tõttu ilmuvad kontaktidest johtuvad muutused kindlas järjekorras; järjekord sõltub tähendusetüübist (konkreetne vs. abstraktne). Eesti keele mõju põlisvenelaste keelevariantidele on ulatuslikum, on toimunud mõningaid struktuurimuutusi (nt rektsioonis); nüüdisvene keeles on stuktuurimuutusi vähem, kuid võib oletada, et aja möödudes ja puristliku hoiaku puudumisel võib nüüdisvene keelekuju nii-öelda järele jõuda.