Variation and variability in L2 learning trajectories : Learning the Finnish existential construction

This is a sel -a chi ed e sion o an o iginal a icle. This e sion may di e om he o iginal in pagina ion and ypog aphic de ails. Au ho (s): Ti le: Yea : Ve sion: Copy igh : Righ s: Righ s u l: Please ci e he o iginal e sion: CC BY 4.0 h ps://c ea i ecommons.o g/licenses/by/4.0/ Va ia ion and a iabili y in L2 lea ning ajec o ies : Lea ning he Finnish exis en ial cons uc ion © 2022 The Au ho (s) Published e sion Lesonen, Si kku; S eink auss, Rasmus; Suni, Minna; Ve spoo , Ma jolijn Lesonen, S., S eink auss, R., Suni, M., & Ve spoo , M. (2022). Va ia ion and a iabili y in L2 lea ning ajec o ies : Lea ning he Finnish exis en ial cons uc ion. Jou nal o he Eu opean Second Language Associa ion, 6(1), 1-17. h ps://doi.o g/10.22599/jesla.77 2022 RESEARCH CORRESPONDING AUTHOR: Si kku Lesonen Uni e si y o Jy äskylä, Seminaa inka u 15, PO Box 35, FI-40014 Jy äskylä, FI [email p o ec ed] KEYWORDS: usage-based linguis ics; complex dynamic sys ems heo y; cons uc ion; L2 Finnish; L2 de elopmen TO CITE THIS ARTICLE: Lesonen, S., S eink auss, R., Suni, M., & Ve spoo , M. (2022). Va ia ion and a iabili y in L2 lea ning ajec o ies: Lea ning he Finnish exis en ial cons uc ion. Jou nal o he Eu opean Second Language Associa ion, 6(1), 1–17. DOI: h ps://doi.o g/10.22599/ jesla.77 Va ia ion and a iabili y in L2 lea ning ajec o ies: Lea ning he Finnish exis en ial cons uc ion SIRKKU LESONEN RASMUS STEINKRAUSS MINNA SUNI MARJOLIJN VERSPOOR ABSTRACT Taking an onomasiological app oach and a dynamic usage-based pe spec i e, his s udy explo es how ou beginning L2 lea ne s o Finnish de elop in exp essing exis en iali y (‘ he e is some hing somewhe e’) be o e and a e ins uc ion. Da a we e collec ed weekly o e a pe iod o nine mon hs and examined o con en ionalized and non-con en ionalized cons uc ions ha exp ess exis en iali y. As expec ed om a dynamic usage-based pe spec i e, bo h in e -indi idual a ia ion and in a-indi idual a iabili y we e iden i ied. The ini ial epe oi es o wo o he lea ne s we e qui e a iable, as hey used se e al di e en non-con en ionalized cons uc ions be o e se ling on mo e con en ionalized ones. In con as , he wo o he lea ne s did no independen ly y ou di e en ways o exp essing he a ge ed meaning bu s a ed o use he con en ionalized Finnish exis en ial cons uc ion only a e pedagogical in e en ion. As one would expec om a usage-based pe spec i e, some lea ne s’ ini ial epe oi es included some i em-based cons uc ions ha we e simila o each o he . As a as ins uc ion is conce ned, o all lea ne s he e was an inc ease in he use o he con en ionalized cons uc ion a e an explici in e en ion, bu he use was no mo phologically accu a e. The indings con i m wo commonly held hypo heses in dynamic sys ems app oaches: Lea ne s own hei own lea ning ajec o ies and ini ial ajec o ies a e some imes cha ac e ized by high deg ees o a iabili y because lea ne s need o y ou di e en s a egies be o e hey can adap o he equi emen s o he new si ua ion. *Au ho a ilia ions can be ound in he back ma e o his a icle 2Lesonen e al. Jou nal o he Eu opean Second Language Associa ion DOI: 10.22599/jesla.77 1. INTRODUCTION Adul second-language (L2) lea ne s who a e li ing in he a ge -language speaking communi y may ha e a need o ai ly complex linguis ic s uc u es a he e y beginning o hei L2 lea ning p ocess. In Finnish, o example, exis en iali y ( he e is some hing somewhe e) is such a case. I is con en ionally exp essed wi h he Finnish exis en ial cons uc ion (FEC; e.g., Suomessa on paljon jä iä ‘The e a e many lakes in Finland’), which is peculia in many ways (e.g., Huumo, 2003; I aska, 2010; Kajande , 2013). The e o e, i is o en conside ed o be a di icul s uc u e (I aska, 2010) and no augh in beginne -le el L2 Finnish cou ses. Howe e , be o e any pedagogical in e en ion, lea ne s may wan o exp ess his meaning, and so hey need o ely on c ea i e linguis ic solu ions ha may no be con en ionalized in he a ge -language speaking communi y. In his s udy, we aim o iden i y he c ea i e lea ne language cons uc ions ha de ia e om con en ional language o ms bu do no cause a b eakdown in communica ion (see Waa a, 2004, o de ini ion o lea ne language cons uc ion). The e o e, we adop an onomasiological app oach, meaning we s a ou in es iga ion om he meaning and explo e wha cons uc ions lea ne s use o exp ess i (Fe nández-Domínguez, 2019; G zega, 2012). We ace he de elopmen o he ways ou L2 Finnish lea ne s’ exp ess exis en iali y du ing a nine- mon h pe iod ha included wo pedagogical in e en ions ocusing on he FEC. We ocus on he indi idual de elopmen al p ocesses by aking a dynamic usage-based (DUB) app oach and in es iga e he e ec o eaching FECs. Mos s udies ha look a he e ec o ins uc ion do so wi h p e- and pos - es s o see how he ea men has a ec ed he L2 de elopmen o g oups o lea ne s (see No is & O ega, 2000; Spada & Tomi a, 2010 o me a-analyses). Al hough such g oup s udies can gi e an excellen indica ion o gene al ends, hey canno shed ligh on he ac ual de elopmen al p ocess (Lowie & Ve spoo , 2019). In a longi udinal s udy se ing, such as he cu en s udy, i is possible o s udy bo h a iabili y (di e si y in pe sonal epe oi es) and a ia ion (lea ne s’ indi idual ajec o ies) in L2 de elopmen (Lowie & Ve spoo , 2015). 2. BACKGROUND 2.1. INVESTIGATING L2 LEARNERS’ LINGUISTIC MEANS: THE ONOMASIOLOGICAL APPROACH Language is used in social in e ac ion o achie e ce ain communica i e goals ha a y om gene al in en ions such as eques ing help o sha ing emo ions o mo e speci ic goals (Schmid, 2020), such as ge ing ac oss he idea ha he e a e many lakes in Finland. P o icien speake s may choose among di e en kinds o con en ionalized linguis ic solu ions in he speech communi y o achie e a ce ain communica i e goal (e.g., Fe nández-Domínguez, 2019; Lesonen e al., 2021; Schmid, 2020). Con en ionalized links be ween communica i e goals and linguis ic o ms a e usually e e ed o as cons uc ions (e.g., Goldbe g, 2006). Fo example, i a p o icien speake wan s o con ey he meaning ha he e a e many lakes in Finland, he likely o m is he FEC (VISK [Comp ehensi e Finnish G amma Online] § 893, 894): Suomessa on paljon jä iä ‘The e a e many lakes in Finland’. Howe e , a beginning L2 Finnish lea ne may no ye ha e no iced o acqui ed he o m o his cons uc ion and may he e o e use c ea i e linguis ic solu ions o exp ess a simila meaning (e.g., *Se on paljon jä iä Suomessa ‘*I is many lakes in Finland’). To in es iga e such lea ne language cons uc ions, i is use ul o ake an onomasiological app oach. In an onomasiological app oach, he in es iga ion s a s wi h he meaning and he aim is o in es iga e wha kinds o linguis ic solu ions a e used o exp ess i (Fe nández-Domínguez, 2019; G zega, 2012). This app oach makes i possible o in es iga e L2 lea ne s’ make-do solu ions ha e lec hei his o y, goals, and abili ies a ha ime (La sen-F eeman, 2013). Using his app oach, we iew lea ne s’ L2 as i s own, aluable linguis ic a ie y, which in he end may o may no esemble he a ge a ie y. 2.2. TRACING L2 LEARNERS’ LANGUAGE CONSTRUCTIONS: A DUB APPROACH A DUB app oach, a combina ion o Complex Dynamic Sys ems Theo y (CDST) and a usage- based pe spec i e, sees language lea ning as an indi idual, dynamic p ocess in which L2 cons uc ions eme ge non-linea ly and become en enched in usage e en s in which an L2 lea ne uses he L2 o he pu poses o in e ac ion (e.g., Ve spoo & Beh ens, 2011; Ve spoo e al., 2012; Lesonen, 2020). Because all lea ne s and hei usage e en s a e unique, L2 lea ning 3Lesonen e al. Jou nal o he Eu opean Second Language Associa ion DOI: 10.22599/jesla.77 ajec o ies a e expec ed o be ai ly di e en om one indi idual o ano he (in e -indi idual a ia ion; Lowie & Ve spoo , 2015). To gain insigh s in o hese indi idual de elopmen al p ocesses, a longi udinal, case-s udy app oach is an app op ia e me hodology (La sen- F eeman & Came on, 2008). I has been shown in longi udinal s udy se ings ha indi idual lea ne s’ ajec o ies a y o a g ea ex en in he de elopmen o syn ac ic complexi y (e.g., Chan e al., 2015; Lowie & Ve spoo , 2019). Much less is known abou how indi iduals a y in hei abili ies o exp ess a ce ain meaning. This s udy aims o ill his gap by in es iga ing wha kind o indi idual ajec o ies ou L2 Finnish lea ne s show when de eloping hei linguis ic esou ces o exp ess he meaning ha is con en ionally exp essed wi h he FEC. In addi ion o in e -indi idual a ia ion, he e is in a-indi idual a iabili y because con en ionalized and non-con en ionalized cons uc ions may al e na e in he L2 lea ne s’ p oduc ion. Va iabili y¹ hus e e s o di e si y wi hin he lea ne ’s pe sonal epe oi e. In he con ex o CDST, i has been sugges ed ha eme ging L2 cons uc ions show conside able a iabili y, especially a he beginning o he lea ning p ocess because lea ne s a e ying ou di e en ways o making meaning (Ellis, 1994; Thelen & Smi h, 1994; an Dijk e al., 2011). As poin ed ou by an Gee and an Dijk (2002, p. 341), a iabili y “can be an essen ial ac o in p omo ing de elopmen ”. I is assumed ha ee explo a ion leads o a iabili y in pe o mance, which can be seen as a p e equisi e o de elopmen ( an Dijk e al., 2011). Fo L2 lea ne s, di e si y in cons uc ion use is impo an because i p o ides ma e ial o selec ion: The mos e ec i e exp essions can be selec ed o la e use. Such de elopmen was obse ed in Cancino e al.’s (1978) da a, which an Dijk e al. (2011) e-analyzed la e om a CDST pe spec i e. In his longi udinal s udy, a he beginning o he lea ning p ocess, he lea ne used di e en kinds o cons uc ions, some a ge like and o he s non- a ge like, o exp ess nega ion. In ecen yea s, many esea che s ha e adop ed a CDST/DUB app oach o s udy he ole o a iabili y in L2 de elopmen . I has been shown ha a lea ne ’s pe o mance luc ua es when measu ed wi h bo h speci ic and holis ic measu es (Spoelman & Ve spoo , 2010; Tilma, 2014; Ve spoo e al., 2008; Ve spoo e al., 2017). Resea ch has also demons a ed ha lea ne s whose language exhibi s subs an ial a iabili y may be mo e success ul han hei less- a iable pee s (Chan e al., 2015; Huang e al., 2021; Lowie & Ve spoo , 2019). Whe eas CDST emphasizes a iabili y in lea ning ajec o ies o e ime, usage-based linguis s a gue ha in i s -language (L1) acquisi ion, cons uc ions a e ini ially ela i ely ixed and ied o speci ic lexical i ems (i.e., i em-based cons uc ions) and he e o e show e y li le a iabili y ea ly on. I is assumed ha hese ini ial cons uc ions a e o e-lea ned exp essions. Only a e open slo s wi hin cons uc ions ha e been disco e ed does he amoun o a iabili y inc ease in young child en’s language p oduc ion. (Dąb owska & Lie en, 2005; Tomasello, 2003.) In L2 lea ning, howe e , indings ha e been mixed. In some cases, lea ne s use i em-based cons uc ions ini ially (Eskildsen, 2009; Eskildsen & Cadie no, 2007; Mellow, 2006); in o he cases, he e is mo e a iabili y in he ea ly cons uc ions (Lesonen e al., 2017; Lesonen e al., 2020; Roeh -B ackin, 2014). The di e ences be ween L1 and L2 may be explained by he ac ha L2 lea ne s a e al eady con e san wi h slo - inding and need less ime o disco e hem. L2 lea ne s can use hei L1 when p oducing L2 exp essions (e.g., Cadie no, 2004; Ja is & Pa lenko, 2008). Angelis and Selinke (2001) poin ou ha mul ilingual lea ne s may ans e om languages o he han he L1. In e language ans e ( ans e o he L2 om o he languages han he L1) has been explained by a cogni i e mode o “ o eign language alk”: The use o a language o he han he L1 may be p e e ed because i sounds mo e o eign han he L1 (de Angelis & Selinke , 2001). 2.3. INSTRUCTION AS AN EXTERNAL FORCE IN DYNAMIC L2 DEVELOPMENT An L2 class oom o e s oppo uni ies o meaning ul in e ac ions, as well as he chance o de elop s a egies o make use o a o dances ou side he class and o answe lea ne s’ ques ions on how he L2 wo ks. In CDST and DUB, ins uc ion may be seen as an ex e nal o ce ha may play a ole in L2 lea ne s’ de elopmen (e.g., La sen-F eeman, 1997; La sen- F eeman & Came on, 2008; Nicolescu & Pe escu, 2013). Because in a complex dynamic sys em “all pa s a e connec ed o all o he pa s” (de Bo & La sen-F eeman, 2011, p. 10), p edic ing linea cause and e ec ela ionships may be e y di icul . Howe e , i should be no ed ha in a dynamic sys em, no all connec ions a e equally s ong (de Bo & La sen-F eeman, 2011). Exis ing esea ch sugges s ha he connec ion be ween ins uc ion and L2 de elopmen is 4Lesonen e al. Jou nal o he Eu opean Second Language Associa ion DOI: 10.22599/jesla.77 ela i ely s ong: I has been shown ha lea ne s de elop hei L2 esou ces when hey lea n languages in ins uc ional se ings (e.g., No is & O ega, 2000; Spada & Tomi a, 2010). S udies on he e ec o ins uc ion ha e mainly ocused on g oups and he e ec o di e en eaching app oaches. Some s udies ha e ound ha a he g oup le el, explici ins uc ion is mo e bene icial han implici ins uc ion (No is & O ega, 2000; Spada & Tomi a, 2010), while o he s show ha a p edominan ly implici p og am is mo e e ec i e han an explici p og am (Rousse- Malpa , 2019). In one longi udinal case s udy om a DUB pe spec i e, Tilma (2014) ound ha he lea ne who ecei ed ins uc ion ha emphasized o m did be e ini ially on complexi y and accu acy measu es han he lea ne who ecei ed ins uc ion ha emphasized meaning, bu hese di e ences diminished o e ime, wi h no signi ican di e ences a e a yea . In he cu en s udy, we ocus on he ins uc ion o he FEC. T adi ionally, he FEC is nei he in oduced no augh ea ly on because i is conside ed a di icul s uc u e o L2 lea ne s, as i is peculia in many ways (I aska, 2010). The e b (mos o en he e b olla ‘ o be’) does no ag ee wi h he subjec , and he subjec comes a e he e b. The mos di icul aspec is he o m o he subjec because i can be in he pa i i e case (I aska 2010; Kajande 2013; Fo FEC, see VISK § 893.) Al hough L2 lea ne s may ind he FEC icky, hey s ill migh ha e a communica i e need o his cons uc ion ea ly on. As poin ed ou by Aal o e al. (2009), i is a he he communica i e need o he lea ne han he complexi y o he s uc u e ha makes lea ning easy o di icul : Lea ning a complex cons uc ion may be easy i he lea ne needs he cons uc ion equen ly. By selec ing he cons uc ions o he analysis wi h an onomasiological app oach, we can ace he lea ne s’ need o FECs because use o non-con en ionalized cons uc ions can be seen as an indica o o he lea ne s’ communica i e need. In he cu en longi udinal s udy, we examine he ole o wo pedagogical in e en ions on ou indi idual lea ne s’ ajec o ies. The speci ic ocus is on how ins uc ion a ec s a iabili y in indi idual lea ning ajec o ies. Va iabili y is in es iga ed in wo ways. Fi s , an indi idual lea ne ’s language use is in es iga ed wi h espec o he equency o con en ionalized FECs and non- con en ionalized cons uc ions ha a e used o exp ess exis en iali y. A la ge epe oi e o di e en kinds o exis en ial cons uc ions (bo h FECs and non-con en ionalized cons uc ions) was conside ed a high deg ee o a iabili y. Second, he FECs a e in es iga ed wi h espec o accu acy. Al e na ion o a ge like and non- a ge like o ms wi hin he FECs is conside ed a high deg ee o a iabili y. We assume ha his kind o a iabili y in lea ne language is educed a e he pedagogical e en because ins uc ion may help he lea ne s o ind he con en ionalized linguis ic means o exp ess exis en iali y. Knowing he con en ionalized cons uc ion educes ee explo a ion ha leads o a iabili y in pe o mance ( an Dijk e al. 2011). 2.4. TARGETED L2 CONSTRUCTION The con en ional FEC exp esses he ac ha someone o some hing exis s o is in a pa icula place (VISK § 893,894). Examples 1 and 2 show wo p o o ypical examples o he exis en ial cons uc ion (see Appendix A o he glossing). (1) Ke okse-ssa on ain yksi kei iö Floo -INE be(3SG) only one ki chen The e is only one ki chen on each loo . (2) Tampe ee-lla on paljon jä -i-ä Tampe e-ADE be(3SG) many lake-PL.PAR The e a e many lakes in Tampe e. As shown in Examples 1 and 2, he noun ph ase (NP) e e ing o he place is opicalized: The exis en ial cons uc ion ells some hing abou his NP (VISK § 894). This NP is mos o en in he inessi e case (ending -ssA, ypically exp essing ha some hing is in o inside some hing, VISK § 1238) o in he adessi e case (ending -llA, ypically exp essing ha some hing is on some hing, VISK § 1238). O he cases can be used as well (wi h ci ies, he choice o -ssA and -llA -ending is no limi ed by his in/on dis inc ion). The mos equen e b in exis en ial cons uc ions is olla ‘ o be’, bu o he e bs can also be used. In a p o o ypical exis en ial cons uc ion, he e is no cong uence be ween he p edica e and he subjec : The p edica e is always in he hi d-pe son- singula o m. The subjec (o en) e e s o some hing ha has no been men ioned be o e in 5Lesonen e al. Jou nal o he Eu opean Second Language Associa ion DOI: 10.22599/jesla.77 he discou se. I can be in he pa i i e case i i is in he plu al (example 2; excep wi h nouns ha only ha e a plu al o m), i i e e s o an uncoun able/mass noun (example 3), o i he e b is nega ed (example 4; VISK § 893, 894). (3) Kupi-ssa on kah i-a Cup-INE be(3SG) co ee-PAR The e is co ee in he cup (4) E u i-ssa ei ole islanni-n ku ssi-a E u -INE NEG be Icelandic-GEN cou se-PAR The e is no Icelandic cou se in E u . 3. THE CURRENT STUDY This s udy aces he de elopmen al ajec o ies o cons uc ions exp essing exis en iali y in ou L2 lea ne s o Finnish. The da a collec ion las ed nine mon hs, s a ing be o e and inishing a e wo ins uc ional in e en ions. The aim was o answe he wo esea ch ques ions p esen ed below. 1. Wha lea ning ajec o ies do he ou lea ne s show when exp essing he meaning ha is con en ionally exp essed wi h he exis en ial cons uc ion in Finnish? a. To wha ex en is he e a ia ion be ween he lea ne s’ ajec o ies (i.e., Do he lea ne s use simila o di e en kinds o linguis ic solu ions o exp ess exis en iali y?)? b. To wha ex en does each lea ne show a iabili y in he cons uc ions hey use o exp ess exis en iali y? 2. Wha impac do pedagogical in e en ions ha e on he ou lea ne s’ ajec o ies? a. How do he in e en ions a ec he equencies o FECs? b. How do he in e en ions a ec he accu acy o o ms wi hin he FECs? 4. METHOD 4.1. PARTICIPANTS The pa icipan s, ou adul L2 lea ne s o Finnish, we e en olled in di e en English-medium p og ams and we e aking he same Finnish cou ses a he same Finnish uni e si y a he ime o he s udy (see sec ion 4.4.). Backg ound in o ma ion on he pa icipan s is p esen ed in Table 1 . To ensu e ha he pa icipan s’ da a we e compa able in e ms o Finnish language p o iciency a he beginning o he s udy, h ee expe ienced a e s e alua ed he lea ne s’ w i ing p o iciency a he beginning o he s udy (see Appendix B). The a ings show ha a he e y beginning o he s udy, in weeks 1 and 3 (w i en ex s 1 and 2), Jungo’s and Khadiza’s w i ing p o iciency was e alua ed as sligh ly highe han Lena’s and Al a o’s, p esumably because hey had al eady been in Finland o some ime. Howe e , by week 5 (w i en ex 3), Lena and Al a o had caugh up wi h hei pee s. The pa icipan s’ o e all p o iciency de elopmen o e PARTICIPANT AGE L1 OTHER LANGUAGES TIME OF RESIDENCE BEFORE THE STUDY EXPLICIT INSTRUCTION BEFORE THE STUDY Lena 23 Ge man Englisha,b, F encha, Icelandica,b 0 0 Jungo 22 Chinese (Hunanese) Manda in Chinesea, Englisha 2 yea s 1 Finnish cou se o 5 ECTSc, 20 hou s o sel s udy Al a o 30 Spanish Englisha, F encha,b, Russiana 0 0 Khadiza 31 Bangla Englisha Hindi, U du 4 yea s 0 Table 1 Backg ound in o ma ion on he pa icipan s. aLea ned in an ins uc ional se ing. b Lea ned in he a ge -language communi y. c Eu opean C edi T ans e Sys em. 6Lesonen e al. Jou nal o he Eu opean Second Language Associa ion DOI: 10.22599/jesla.77 he nine-mon h pe iod is e iden o hei success ul comple ion o a B1 le el Finnish cou se a he end o he pe iod o obse a ion. 4.2. DATA COLLECTION The da a we e collec ed om he ou uni e si y s uden s weekly, w i en and spoken da a al e na ing, o e a pe iod o 9 mon hs. Bo h ypes o da a we e collec ed because i was desi able ha he da a collec ion e lec ed ac ual usage-e en s ha lea ne s ake pa in in eal li e. Al hough w i en and spoken a ian s can be qui e di e en in he Finnish language, he exis en ial cons uc ion is no subs an ially di e en in hese di e en modes. In o he wo ds, i was no expec ed ha he lea ne s would use di e en cons uc ions o exp ess his meaning in one mode e sus he o he . On hese g ounds, he decision was made o c ea e one co pus consis ing o w i en and spoken samples. The numbe o da a poin s is shown in Table 2 . The da a a e ee esponse da a. The pa icipan s we e asked o alk o w i e abou opics like summe plans o hei home own. The opics we e usually amilia o he pa icipan s because hey we e selec ed in acco dance wi h cou se con en . The w i en da a we e handw i en and he w i ing samples we e on a e age 99 wo ds long ( ange: 40–176 wo ds; SD = 33). The spoken da a we e eco ded in a language s udio, usually wi h a eco de (Roland R-05) bu once wi h a sma phone. The spoken da a consis ed o monologues and dialogues. The o he pe son in a dialogue was ei he ano he L2 speake om he same class o an L1 speake o Finnish (mos ly he i s au ho o he pape ). The speaking samples we e on a e age 260 wo ds long ( ange: 63–629 wo ds; SD = 128). In he i s hal o he da a collec ion, he da a we e collec ed du ing he lessons o he pa icipan s’ language cou se. The e was a ime limi o app oxima ely 20 minu es. In he second hal , he da a we e collec ed ou side he class and he e was no ime limi . 4.3. DATA SELECTION AND ANALYTIC PROCEDURES We adop ed an onomasiological app oach: We sea ched o he o mal e baliza ions o a gi en concep (G zega, 2012), in ou case, exis en iali y. The i s au ho selec ed all cons uc ions ha con ained he meaning con en ionally exp essed wi h he FEC. In his s udy, Goldbe g’s (2006) de ini ion o a cons uc ion (a con en ionalized pai ing o o m and unc ion) has been b oadened o include he L2 lea ne s’ eme gen o m-meaning mappings, which migh no ye seem con en ional om he poin o iew o p o icien language use s (see Waa a, 2004). Two ca ego ies we e o med. One is con en ionalized FECs (e.g., Saksassa ei ole saunoja ‘The e a e no saunas in Ge many’). The second ca ego y is non-con en ionalized cons uc ions (e.g., *Se on kolme opiskelijaa samassa huoneessa ‘*I is h ee s uden s in he same oom’; *Jy äskylä on kaupunki paljon sil an kanssa ‘*Jy äskylä is a ci y wi h many b idges’). Al hough he non- con en ionalized cons uc ions migh some imes look qui e di e en om he con en ionalized cons uc ion, selec ing hem o he analysis was no p oblema ic; he link be ween he a ge ed meaning and he o m was clea in i ually all cases. As o FECs, we included hose ha con ained all he necessa y elemen s in he igh places, e en i hey con ained inaccu acies (e.g., using he nomina i e case ins ead o he pa i i e case o a subjec ). A e he manual selec ion o he cons uc ions, he equencies o he cons uc ions we e no malized because o he di e ences in ex leng h. In he p ocess o no maliza ion, he equencies we e calcula ed pe 100 wo ds. To answe ou i s esea ch ques ion (wha lea ning ajec o ies he ou lea ne s showed), line g aphs we e c ea ed o isualize he equency de elopmen o con en ionalized FECs and non-con en ionalized cons uc ions. FECs we e also analyzed in e ms o hei accu acy, wi h he esul s p esen ed in ba g aphs. To answe ou second esea ch ques ion (wha impac pedagogical in e en ions had on he ou lea ne s’ PARTICIPANT WRITTEN DATA SPOKEN DATA TOTAL Lena 17 18 35 Jungo 18 17 35 Al a o 16 17 33 Khadiza 16 12 28 Table 2 Numbe o da a poin s. 7Lesonen e al. Jou nal o he Eu opean Second Language Associa ion DOI: 10.22599/jesla.77 ajec o ies), he ins uc ional se ing was analyzed i s . Because he i s au ho was p esen a e e y lesson (ei he as a eache o he i s i e mon hs o an obse e du ing he las ou mon hs), we ha e p ecise in o ma ion abou wha was augh du ing he classes and he kinds o lea ning ma e ial and ac i i ies ha we e used. A e analyzing he ins uc ion o he exis en ial cons uc ion in he pa icipan s’ cou ses, changes in he lea ne s’ de elopmen al pa hs we e compa ed wi h he iming o he pedagogical in e en ions. 4.4. INSTRUCTIONAL SETTING The ou s uden s we e aking he same h ee L2 Finnish cou ses a a Finnish uni e si y. The h ee cou ses we e a Common Eu opean F amewo k o Re e ence o Languages (CEFR) le els A1, A2 and B1. Each cou se consis ed o 70 con ac hou s plus addi ional independen wo k. All h ee cou ses we e augh by a quali ied language eache who was an L1 speake o Finnish, he i s wo cou ses by he i s au ho and he hi d one by a colleague. The h ee cou ses aimed o de elop lea ne s’ skills in ou di e en unc ions: Social in e ac ion, elling and desc ibing, unde s anding and sea ching o in o ma ion, and de eloping as a language lea ne . The eaching app oach was p ima ily meaning- ocused, wi h some explici a en ion o linguis ic o ms. Du ing he pe iod o obse a ion, in o al 55 di e en FECs we e in oduced in he lea ning ma e ial and on he boa d. FECs we e speci ically deal wi h wice du ing he pe iod o obse a ion. In weeks 15 and 16, s uden s’ awa eness o he FEC was aised o he i s ime. In hese lessons, he ocus was no so much on he exis en ial cons uc ion, bu on he use o he pa i i e case. The exis en ial cons uc ion was p esen ed as one con ex in which he pa i i e case is used. In week 15, s uden s’ a en ion was d awn o he FEC and i was poin ed ou ha his cons uc ion co esponds o he English he e is/ he e a e. In week 16, he s uden s did a w i en exe cise ha included 13 di e en exis en ial cons uc ions. In his exe cise, he s uden s needed o o m he co ec pa i i e o m o he subjec wi hin each FEC. The pa i i e o m was he a ge ed o m, and no o he case would ha e been a ge like. In weeks 28 and 29, he main pedagogical in e en ion ook place. The exis en ial cons uc ion was explici ly augh , i s elemen s we e poin ed ou , and he o m o he subjec was discussed du ing he class. The e we e also speaking and w i ing exe cises ocused on he FEC. Figu e 1 shows he iming o he in e en ions. The i s au ho was he eache in he i s pedagogical in e en ion in weeks 15 and 16. The main pedagogical in e en ion in weeks 28 and 29 was augh by a colleague. 5. RESULTS Table 3 shows he numbe o cons uc ions used o exp ess exis en iali y du ing he pe iod o obse a ion o each pa icipan . Nex , he esul s o each indi idual lea ne a e p esen ed sepa a ely. Two igu es o each lea ne a e p o ided. The i s igu e illus a es he lea ning ajec o ies ( esea ch ques ion 1) and shows he no malized equencies o bo h con en ionalized FECs (Thick, g ey line) and PARTICIPANT NUMBER OF CONSTRUCTIONS NUMBER OF CONSTRUCTIONS PER 100 WORDS NUMBER OF FECS PER 100 WORDS NUMBER OF NON- CONVENTIONALIZED CONSTRUCTIONS PER 100 WORDS Lena 39 15.7 8.7 7.0 Jungo 54 33.9 16.7 17.2 Al a o 30 9.3 7.9 1.4 Khadiza 22 11.2 10.5 0.7 Table 3 Numbe o cons uc ions used o exp ess exis en iali y. Figu e 1 Timing o he pedagogical in e en ions. 8Lesonen e al. Jou nal o he Eu opean Second Language Associa ion DOI: 10.22599/jesla.77 non-con en ionalized cons uc ions ha we e used o exp ess exis en iali y (o he lines). The second igu e illus a es he impac o pedagogical in e en ions ( esea ch ques ion 2). I shows he equency o FECs wi h a non- a ge like subjec ( ed ba s), FECs wi h he a ge like subjec (g een ba s), and non-con en ionalized cons uc ions ha we e used o exp ess exis en iali y (g ey ba s). The iming o ins uc ion is indica ed wi h a ows. The blue a ows indica e he i s pedagogical in e en ion and he ed a ows indica e he iming o he main pedagogical in e en ion. Fo each lea ne , accu acy pe ains o he subjec o m only. Accu acy was ela i ely high o he opicalized NP and o he p edica e (84–100%). I should be no ed ha he e is no always a one- o-one co espondence be ween he equencies o exis en ial cons uc ions and hei subjec s in he igu es because some imes he e was mo e han one subjec in di e en o ms wi hin one cons uc ion o non-Finnish wo ds we e used whose accu acy could no be e alua ed. 5.1. LENA As he igu es show, Lena ied ou many di e en kinds o cons uc ions o exp ess he meaning ha is con en ionally exp essed wi h he FEC. The non-con en ionalized cons uc ions used mo e han once included he ollowing ypes: kanssa ‘wi h’ (example 5), se on ‘i is’ (example 6), and on olemassa ‘exis ’ (example 7). (5) The i s example om week 10 Jy äskylä on kaupunki *paljon *sil a-n kanssa Jy äskylä be(3SG) ci y *many *b idge-GEN wi h Jy äskylä is a ci y wi h many b idges (6) The i s example om week 11 Tal ella se ei ole au inko Suome-ssa päi ä-llä. Win e -ADE i NEG be sun Finland-INE day-ADE In he win e , i is no sun in Finland du ing he day (In he win e , he e is no sun in Finland du ing he day) (7) The i s example om week 20 *ole ole-ma-ssa *mon a *sukke-ja *be be-3.INF-INE *many *sock-PL.PAR exis s many socks In some weeks Lena used bo h he con en ionalized FEC and non-con en ionalized cons uc ions o exp ess exis en iali y. This kind o non-linea i y in he de elopmen was e iden , o example, in week 16, when Lena used he con en ionalized FEC (example 8) and he possessi e cons uc ion o exp ess exis en iali y (example 9). (8) The i s example om week 16 Be liini-ssä on eloku a ea e i Be lin-INE be(3SG) mo ie. hea e The e is a mo ie hea e in Be lin (9) The second example om week 16 * ee-llä ei *on *ko eiini * ea-ADE NEG *be(3SG) *ca eine Tea has no ca eine (The e is no ca eine in ea) The e seemed o be a gene al endency in Lena’s ajec o y o he con en ionalized FECs o ake o e om non-con en ionalized cons uc ions o e ime ( Figu e 2 ). A e he main pedagogical in e en ion, he use and p opo ion o he FEC inc eased. Howe e , inaccu a e o ms o he subjec also linge ed a e he main pedagogical in e en ion in which he o m o he subjec was explici ly emphasized (see Figu e 3 ). 15Lesonen e al. Jou nal o he Eu opean Second Language Associa ion DOI: 10.22599/jesla.77 AUTHOR AFFILIATIONS Si kku Lesonen o cid.o g/0000-0002-5274-8518 Uni e si y o Jy äskylä, Seminaa inka u 15, PO Box 35, FI-40014 Uni e si y o Jy äskylä, Jy äskylä, FI Rasmus S eink auss o cid.o g/0000-0002-3643-8704 Uni e si y o G oningen, NL Minna Suni o cid.o g/0000-0002-4869-1846 Uni e si y o Jy äskylä, FI Ma jolijn Ve spoo o cid.o g/0000-0001-8828-8405 Uni e si y o Pannonia, Hunga y REFERENCES Aal o, E., Mus onen, S., & Tukia, K. (2009). Funk ionaalisuus oisen kielen oppimisen läh ökoh ana [Func ionalism as he basis o eaching a second language]. Vi i äjä, 3, 402–423. de Angelis, G., & Selinke , L. (2001). In e language ans e and compe ing linguis ic sys ems in he mul ilingual mind. In J. Genoz, B. Hu eisen, & U. Jessne (Eds.), C oss-linguis ic in luence in hi d language acquisi ion: Psycholinguis ic pe spec i es (pp. 42–58). Mul ilingual Ma e s. DOI: h ps://doi. o g/10.21832/9781853595509-004 de Bo , K., & La sen-F eeman, D. (2011). Resea ching second language de elopmen om a dynamic sys ems heo y pe spec i e. In M. Ve spoo , K. de Bo , & W. Lowie (Eds.), A dynamic app oach o second language le elopmen . Me hods and echniques (pp. 5–23). John Benjamins. DOI: h ps://doi. o g/10.1075/lll .29.01deb Cadie no, T. (2004). Exp essing mo ion e en s in a second language: a cogni i e ypological pe spec i e. In M. Acha d & S. Niemeie (Eds.), Cogni i e linguis ics, second language acquisi ion, and o eign language eaching (pp. 13–49). De G uy e Mou on. DOI: h ps://doi.o g/10.1515/9783110199857.13 Cancino, H., Rosansky, E., & Schumann, J. (1978). The acquisi ion o English nega i es and in e oga i es by na i e Spanish speake s. In E. M. Ha ch (Ed.), Second language acquis ion: A book o eadings (pp. 207–230). Newbu y House. Council o Eu ope. (2001). Common Eu opean F amewo k o Languages: Lea ning, eaching, assessmen . Camb idge Uni e si y P ess. Chan, H., Ve spoo , M., & Vah ick, L. (2015). Dynamic de elopmen in speaking e sus w i ing in iden ical wins. Language Lea ning, 65(2), 298–325. DOI: h ps://doi.o g/10.1111/lang.12107 Dąb owska, E. (2001). F om o mula o schema: The acquisi ion o English ques ions. Cogni i e Linguis ics, 11(1–2), 83–102. DOI: h ps://doi.o g/10.1515/cogl.2001.013 Dąb owska, E., & Lie en, E. (2005). Towa ds a lexically speci ic g amma o child en’s ques ion cons uc ions. Cogni i e Linguis ics, 16(3), 437–474. DOI: h ps://doi.o g/10.1515/cogl.2005.16.3.437 Ellis, R. (1994). The s udy o second language acquisi ion. Ox o d Uni e si y P ess. Eskildsen, S., & Cadie no, T. (2007). A e ecu ing mul i-wo d exp essions eally syn ac ic eezes? Second language acquisi ion om he pe spec i e o usagebased linguis ics. In M. Nenonen & S. Niemi (Eds.), Colloca ions and idioms 1: Pape s om he i s No dic con e ence on syn ac ic eezes (pp. 86–99). Joensuu Uni e si y P ess. TEXT 1 TEXT 2 TEXT 3 TEXTS 1–3 TOGETHER RANGE MEDIAN RANGE MEDIAN RANGE MEDIAN RANGE MEDIAN Lena 1 1 1 1 2–3 3 1–3 1 Jungo 1–2 2 1–2 2 2–3 2 1–3 2 Al a o 1 1 1–2 1 2–3 3 1–3 1 Khadiza 1–2 2 1–2 2 2–3 2 1–3 2 APPENDIX B. THE PARTICIPANTS’ WRITING PROFICIENCY AT THE BEGINNING OF THE STUDY No e. The ange and median come om he a ings o h ee e alua o s. The c i e ia o he Finnish Na ional Ce i ica es o Language P o iciency es ing we e used. These c i e ia a e p o ided by he Cen e o Applied Language S udies and he Finnish Na ional Agency o Educa ion, (Uni e si y o Jy äskylä, 2019). The scale is om 1 o 6, 1 being he lowes and 6 he highes le el, co esponding o le els A1-C2 in he Common Eu opean F amewo k o Re e ence o Languages (CEFR) (Council o Eu ope, 2001). The leng h o he ex s anged be ween 40 and 159 wo ds (M = 91; SD = 41). Pa icipan s’ L2 w i ing p o iciency a he beginning o he s udy. 16Lesonen e al. Jou nal o he Eu opean Second Language Associa ion DOI: 10.22599/jesla.77 Eskildsen, S. W. (2009). Cons uc ing ano he language - usage-based linguis ics in second language acquisi ion. Applied Linguis ics, 30(3), 335–357. DOI: h ps://doi.o g/10.1093/applin/amn037 Fe nández-Domínguez, J. (2019). The onomasiological app oach. In M. A ono (Ed.), Ox o d esea ch encyclopedia o linguis ics (pp. 1–26). Ox o d Uni e si y P ess. DOI: h ps://doi.o g/10.1093/ ac e o e/9780199384655.013.579 Goldbe g, A. E. (2006). Cons uc ions a wo k – he na u e o gene aliza ion in language. Ox o d Uni e si y P ess. G zega, J. (2012). Lexical – seman ic a iables. In J. M. He nández-Campoy & J. C. Conde-Sil es e (Eds.), The handbook o his o ical sociolinguis ics (pp. 271–292). Blackwell. DOI: h ps://doi. o g/10.1002/9781118257227.ch15 Huang, T., S eink auss, R., & Ve spoo , M. (2021). Va iabili y as p edic o in L2 w i ing p o iciency. Jou nal o Second Language W i ing, 52(4), 100787. DOI: h ps://doi.o g/10.1016/j.jslw.2020.100787 Huumo, T. (2003). Inc emen al exis ence: he wo ld acco ding o he Finnish exis en ial sen ence. Linguis ics, 41(3), 461–493. DOI: h ps://doi.o g/10.1515/ling.2003.016 I aska, I. (2010). Eksis en iaalilauseen subjek ijäsen edis yneiden suomenoppijoiden ki joi uksessa [Subjec cons i uen o he exis en ial sen ence in he w i ing o ad anced lea ne s o Finnish]. Sananjalka, 52(1), 67–88. DOI: h ps://doi.o g/10.30673/sja.86697 Ja is, S., & Pa lenko, A. (2008). C osslinguis ic in luence in language and cogni ion. Rou ledge. DOI: h ps://doi.o g/10.4324/9780203935927 Kajande , M. (2013). Suomen eksis en iaalilause oisen kielen oppimisen polulla [Pa hs o lea ning Finnish exis en ial sen ences] (220) [Doc o al disse a ion, Uni e si y o Jy äskylä]. Jy äskylä S udies in Humani ies. h ps://jyx.jyu. i/handle/123456789/42649 La sen-F eeman, D. (1997). Chaos/complexi y science and second language acquisi ion. Applied Linguis ics, 18(2), 141–165. DOI: h ps://doi.o g/10.1093/applin/18.2.141 La sen-F eeman, D. (2013). Complexi y heo y/dynamic sys ems heo y. In P. Robinson (Ed.), The Rou ledge encyclopedia o second language acquisi ion (pp. 103–105). Rou ledge. La sen-F eeman, D., & Came on, L. (2008). Complex sys ems and applied linguis ics. Ox o d Uni e si y P ess. Lesonen, S. (2020). Valuing a iabili y: Dynamic usage-based p inciples in he L2 de elopmen o ou Finnish language lea ne s (184). [Doc o al disse a ion, Uni e si y o G oningen]. G oningen Disse a ions in Linguis ics. h ps:// esea ch. ug.nl/en/publica ions/ aluing- a iabili y-dynamic-usage- based-p inciples-in- he-l2-de e Lesonen, S., S eink auss, R., Suni, M., & Ve spoo , M. (2020). Lexically speci ic s. p oduc i e cons uc ions in L2 Finnish lea ne s. Language and Cogni ion, 12(3), 526–563. DOI: h ps://doi. o g/10.1017/langcog.2020.12 Lesonen, S., S eink auss, R., Suni, M., & Ve spoo , M. (2021). Dynamic usage-based p inciples in he de elopmen o L2 Finnish e alua i e cons uc ions. Applied Linguis ics, 42(3), 442–472. DOI: h ps:// doi.o g/10.1093/applin/amaa030 Lesonen, S., Suni, M., S eink auss, R., & Ve spoo , M. (2017). F om concep ualiza ion o cons uc ions in Finnish as an L2: A case s udy. P agma ics & Cogni ion, 24(2), 212–262. DOI: h ps://doi.o g/10.1075/ pc.17016.les Lowie, W., & Ve spoo , M. (2015). Va iabili y and a ia ion in second language acquisi ion o de s: A dynamic ee alua ion. Language Lea ning, 65(1), 63–88. DOI: h ps://doi.o g/10.1111/lang.12093 Lowie, W., & Ve spoo , M. (2019). Indi idual di e ences and he e godici y p oblem. Language Lea ning, 69(S1), 184–206. DOI: h ps://doi.o g/10.1111/lang.12324 Mellow, J. D. (2006). The eme gence o second language syn ax: A case s udy o he acquisi ion o ela i e clauses. Applied Linguis ics, 27(4), 645–670. DOI: h ps://doi.o g/10.1093/applin/aml031 Nicolescu, B., & Pe escu, T. (2013). Dynamical sys ems heo y – a powe ul ool in he educa ional sciences. P ocedia – Social and Beha io al Sciences, 76, 581–587. DOI: h ps://doi.o g/10.1016/j. sbsp o.2013.04.168 No is, J., & O ega, L. (2000). E ec i eness o L2 ins uc ion: A esea ch syn hesis and quan i a i e me a-analysis. Language Lea ning, 50(3), 417–528. DOI: h ps://doi.o g/10.1111/0023-8333.00136 Roeh -B ackin, K. (2014). Explici knowledge and p ocesses om a usage-based pe spec i e: he de elopmen al ajec o y o an ins uc ed L2 lea ne . Language Lea ning, 64(4), 771–808. DOI: h ps://doi.o g/10.1111/lang.12081 Rousse-Malpa , A. (2019). E ec i eness o explici s. implici L2 ins uc ion: a longi udinal class oom s udy on o al and w i en skills (176). [Doc o al disse a ion, Uni e si y o G oningen]. G oningen Disse a ions in Linguis ics. h ps:// esea ch. ug.nl/en/publica ions/e ec i eness-o -explici - s- implici -l2-ins uc ion-a-longi udin Schmid, H-J. (2020). The dynamics o he linguis ic sys em. Usage, con en ionaliza ion, and en enchmen . Ox o d Uni e si y P ess. DOI: h ps://doi.o g/10.1093/oso/9780198814771.001.0001 17Lesonen e al. Jou nal o he Eu opean Second Language Associa ion DOI: 10.22599/jesla.77 TO CITE THIS ARTICLE: Lesonen, S., S eink auss, R., Suni, M., & Ve spoo , M. (2022). Va ia ion and a iabili y in L2 lea ning ajec o ies: Lea ning he Finnish exis en ial cons uc ion. Jou nal o he Eu opean Second Language Associa ion, 6(1), 1–17. DOI: h ps://doi.o g/10.22599/ jesla.77 Submi ed: 21 June 2021 Accep ed: 10 Janua y 2022 Published: 01 Ma ch 2022 COPYRIGHT: © 2022 The Au ho (s). This is an open-access a icle dis ibu ed unde he e ms o he C ea i e Commons A ibu ion 4.0 In e na ional License (CC-BY 4.0), which pe mi s un es ic ed use, dis ibu ion, and ep oduc ion in any medium, p o ided he o iginal au ho and sou ce a e c edi ed. See h p://c ea i ecommons.o g/ licenses/by/4.0/. Jou nal o he Eu opean Second Language Associa ion, is a pee - e iewed open access jou nal published by Whi e Rose Uni e si y P ess. Spada, N., & Tomi a, Y. (2010) In e ac ions be ween ype o ins uc ion and ype o language ea u e: A me a-analysis. Language Lea ning, 60(2), 263–308. DOI: h ps://doi.o g/10.1111/j.1467- 9922.2010.00562.x Spoelman, M., & Ve spoo , M. (2010). Dynamic pa e ns in de elopmen o accu acy and complexi y: a longi udinal case s udy in he acquisi ion o Finnish. Applied Linguis ics, 31(4), 532–553. DOI: h ps:// doi.o g/10.1093/applin/amq001 Thelen, E., & Smi h, L. (1994). A dynamic sys ems app oach o he de elopmen o cogni ion and ac ion. MIT P ess. Tilma, C. (2014). The dynamics o o eign e sus second language de elopmen in Finnish w i ing (233). [Doc o al disse a ion, Uni e si y o Jy äskylä]. Jy äskylä S udies in Humani ies. h ps://jyx.jyu. i/ handle/123456789/44632 Tomasello, M. (2003). Cons uc ing a language. A usage-based heo y o language acquisi ion. Ha a d Uni e si y P ess. Uni e si y o Jy äskylä. (2019, June 1). Yleise kieli u kinno . Ki joi amisen a ioin ik i ee i . [Na ional ce i ica es o language p o iciency. E alua ion c i e ia o w i ing.] h ps://www.jyu. i/hy k/ i/lai okse / solki/yki/yleis a/ ie oakieli u kinnois a/a ioin i an Dijk, M., Ve spoo , M., & Lowie, W. (2011). Va iabili y in DST. In M. Ve spoo , K. de Bo , & W. Lowie (Eds.), A dynamic app oach o second language de elopmen . Me hods and echniques (pp. 55–84). John Benjamins. DOI: h ps://doi.o g/10.1075/lll .29.04 an an Gee , P., & an Dijk, M. (2002). Focus on a iabili y: New ools o s udy in a-indi idual a iabili y in de elopmen al da a. In an Beha io and De elopmen , 25(4), 340–374. DOI: h ps://doi.o g/10.1016/ S0163-6383(02)00140-6 Ve spoo , M., & Beh ens, H. (2011). Dynamic sys ems heo y and a usage-based app oach o second language de elopmen . In M. Ve spoo , K. de Bo , & W. Lowie (Eds.), A dynamic app oach o second language de elopmen . Me hods and echniques (pp. 25–38). John Benjamins. DOI: h ps://doi. o g/10.1075/lll .29.02 e Ve spoo , M., Lowie, W., & an Dijk, M. (2008). Va iabili y in second language de elopmen om a dynamic sys ems pe spec i e. The Mode n Language Jou nal, 2(2), 214–231. DOI: h ps://doi. o g/10.1111/j.1540-4781.2008.00715.x Ve spoo , M., Schmid, M. S., & Xu, X. (2012). A dynamic usage based pe spec i e on L2 w i ing. Jou nal o Second Language W i ing, 21(3), 239–263. DOI: h ps://doi.o g/10.1016/j.jslw.2012.03.007 Ve spoo , M., Lowie, W., Chan, H. P., & Vah ick, L. (2017). Linguis ic complexi y in second language de elopmen : a iabili y and a ia ion a ad anced s ages. Reche ches En Didac ique Des Langues e Des Cul u es, 14(1), 1–27. DOI: h ps://doi.o g/10.4000/ dlc.1450 Waa a, R. (2004). Cons ual, con en ion, and cons uc ions in L2 speech. In S. Niemeie & M. Acha d (Eds.), Cogni i e linguis ics, second language acquisi ion, and o eign language eaching (pp. 51–75). De G uy e Mou on. DOI: h ps://doi.o g/10.1515/9783110199857.51 VISK= Hakulinen, A., Vilkuna, M., Ko honen, R., Koi is o, V., Heinonen, T. R., & Alho, I. (2004). Iso suomen kielioppi. [Comp ehensi e Finnish G amma Online.] Suomalaisen Ki jallisuuden Seu a. VISK, Elec onic Da abase, a ailable a h p://sc ip a.ko us. i/ isk. Accessed 25 Janua y 2022.