1. Introduction

In the present paper I intend to analyze how the English language influences the variety of Estonian spoken by immigrants and their descendants living in Canada, from now on referred to as Canadian Estonian. My analysis is based on data collected from the forum discussions of a Toronto-based on-line newspaper, Eesti Elu (“Estonian Life”).

The question the paper seeks to answer is what the characteristics of Canadian-Estonians’ written language use are and which contact phenomena are detectable in this variety. Besides loanwords, I am interested in features that are the result of English influence and, especially, the characteristics that are different from those present in the European variety of Estonian spoken in Estonia. Furthermore, I intend to provide an explanation for these phenomena mainly on the basis of speakers’ attitudes and the social characteristics of the contact situation, e.g. by identifying the driving forces behind borrowing.

In my analysis, I adopt the borrowing scale by Thomason and Kaufman (1988) and modified by Thomason (2001) to account for differences in the two varieties as regards contact with English. I combine the typologies of Haugen (1950) and Lanstyák (2006) for the classification of the corpus data.

2. Literature review

2.1. The models used for analysis and categorization of the data

In the first part of this section, I discuss theoretical frameworks concerning borrowing, concentrating mainly on Thomason’s (2001) borrowing scale which provides an explanation for the differences between the borrowed English material in the case of the European vs. the immigrant variety of Estonian.

Thomason (2001) discusses language contact from various aspects, providing an elaborated framework on the linguistic and social factors, mechanisms, and outcomes of contact induced language change. For the explanation of the different results of change, she develops a modified version of the earlier framework in Thomason and Kaufman (1988).

Thomason (2001: 70–71) distinguishes between four stages in her borrowing scale where phases are defined on the basis of social factors, especially intensity of contact between speakers of the source and the recipient language (each later stage also includes linguistic results of the earlier stages):

(i) Casual contact: only lexical borrowing (in non-basic vocabulary) of mostly nouns (verbs, adjectives, and adverbs also); no structural borrowing.

(ii) Slightly more intense contact: function words and slight structural borrowing.

(iii) More intense contact: non-basic vocabulary is also borrowed, the degree of the structural borrowing is moderate.

(iv) Intense contact: heavy borrowing in lexicon and in structure as well, even if it changes the typological characteristics of the recipient language.

 

The main criterion for the differentiation between the stages is the source language proficiency of the members of the borrowing community. In the first stage, borrowers are not inevitably fluent in the source language, and bilinguals need not be involved in the borrowing process. In the next phase, borrowers are bilinguals, however, the majority in the community is still monolingual. The third stage presents a state when borrowing is favored by different factors and the number of bilinguals increases. In the final stage, bilingualism is extensive among speakers of the recipient language and borrowing is favored.

In the case of Canadian Estonian, the borrowed features indicate that the immigrant variety reached the third or even the fourth stage (with word order changes and loss of congruence, etc.). However, there are differences also within the Canadian Estonian speech community as regards language use, which might indicate that not all of the Canadian Estonian speakers possess the same level in the knowledge of English and Estonian, and their community is not homogeneous. Canadian Estonians are split not only as regards their knowledge of Estonian but also as regards their attitudes towards the preservation of this language.

For the differentiation between the different types of lexical borrowing, I use Haugen’s system. According to Haugen (1950), the three main types of borrowing are loanwords, loan blends (or hybrids) and loan shifts (loan translations and semantic loans). The major criteria according to which he differentiates between the three groups are morphemic importation and morphemic substitution. His typology is represented in Table 1, formed on the basis of the above mentioned criteria.

	morphemic importation	morphemic substitution
loanwords	+	–
loan blends	+	+
loan shifts	–	+

Table 1. Haugen’s typology

In loan blends only “a part of the phonemic shape of the word has been imported, while a native portion has been substituted for the rest” (Haugen 1950: 214). In the CEst² database, the term Olympic Mängudel (Olympic + Mängudel > mäng ’game’, Mängudel ’on the Games’), represents this type of borrowing. In hybrids such as this, it is usually the second part which is not borrowed, because preserving the L1 part as the second (i.e. the inflected) part of the word makes morphological adaptation easier.

The category of loan shifts can be divided into two subgroups on the basis of the morphological make-up of the word. If the borrowed element consists of two stems, i.e. it is a compound, it is then a calque according to Haugen’s terminology. Whereas, if the word is morphologically simple, it is considered to be a semantic loan or semantic calque, and then “no formal structural element whatever has been imported, only a meaning” (Haugen 1950: 214). So the types of borrowing can be divided along morphological lines and on the basis of the degree of morphemic importation.

The categorization of the data into these three main types proves insufficient, as there are borderline cases in the corpus. As a result, I also employ a more elaborate typology of lexical borrowings Lanstyák’s (2006: 56). It enhances Haugen’s system with subtypes of borrowings and constructs a more complex system with seven main categories of primary loanwords, which are “words or fixed phrases that emerge under the influence of a word or fixed phrase of the source language and do not contain another source language loanword” (Lanstyák 2006: 37; my translation).

direct loans in the stricter sense	tax ‘tax’	direct transfer             indirect transfer
formal loans	joogurt ‘yoghurt’
hybrid stems	organisatioon ’organisation’
hybrid calques	Olympic Mängudel ‘at the Olympic Games’
homogenous calques	meditsiiniabi kaart ‘medicine card’
semantic loans	positsioon ‘position’
stylistic loans	luuser ‘loser’

Table 2. Lanstyák’s typology of primary loanwords (2006: 56)

The direct loans in the stricter sense (Lanstyák 2006: 38) are loanwords that exist e.g. both in CEst and CEng but are missing from EEst, and the form and the meaning of the two words are identical, or at least similar to each other. Formal loans (Lanstyák 2006: 39), on the other hand, are loanwords that exist e.g. in EEst and CEng as well with similar forms and similar or identical meanings, but the CEst form is closer to the CEng variant. As the data did not warrant such a distinction, I included these two categories under the label “pure” loanwords (cf. also Winford 2003: 45).

Hybrid stems (Lanstyák 2006: 40) are CEst loanwords that have equivalents both in CEng and EEst which are similar to the CEst word in form but neither of them is exactly the same as the CEst variant.

Hybrid calques correspond to the loan blend category in Haugen’s system where homogeneous calques and semantic loans are merged under the label loan shift.

A stylistic loan, in Lanstyák’s (2006) typology, is a loanword that has an equivalent in CEng and in EEst as well, but the CEst form is closer to the CEng form as regards its stylistic characteristics. As in most of the cases it would have been impossible to decide which register a loanword belongs to, I ignore this loanword type in my analysis. However, the word luuser ‘loser’ might be considered as a stylistic loan, since in EEst it is used as a slang word, while in CEst it might be neutral in style.

In my analysis, I use a system which is a combination of Haugen’s (1950), Lanstyák’s (2006) and Winford’s (2003: 45) typologies, and contains the following categories:

simple	“pure” loanwords	hybrid stems	semantic calques
derived		derived hybrids	– (derived calques)*
compound		compound hybrids	homogeneous calques

Table 3. Typology of loanwords in my analysis

There are two main criteria in the modified system I use: while I keep Haugen’s morphemic importation and morphemic substitution as decisive factors, I also analyze the morphemic makeup of the word, differentiating between simple, derived and compound stages. In case of the “pure” loanwords, the latter criterion is irrelevant as the internal structure of the word might be opaque for the borrowers, seeing that no substitution occurs.

In hybrids (hybrid stems, derived hybrids and compound hybrids), the substitution is partial. The case of hybrid stems is controversial, as in the corpus most of the examples which I classify as hybrid stems are in fact derivations in the source language. The reason I analyze them as hybrid stems is that in CEst it is not the derivative suffix which is recognized and substituted for its EEst counterpart (in which case the given word would be a derived hybrid), but only a part of the word’s phonemic structure is replaced, which does not correspond to the morphemic makeup of the word. For example, the CEst word organisatioon should have the form organisatsioon if the suffix – tion would have been substituted for its EEst counterpart – tsioon. As far as pronunciation is concerned, hybrid stems might be pronounced the same way as their EEst counterparts, i.e. both CEst organisatioon and EEst organisatsioon are realized as [organisatsio:n] and it is only the written form which reflects the English orthography. In this case, the borrowed element could be considered a semantic loan. However, the word might also be a hybrid in pronunciation, having the [š] as the English counterpart but also the [o:] as in EEst [organisašio:n]. On the basis of the written form, it is impossible to unambiguously determine whether phonological adaptation occurred and to what extent.

Compound hybrids are called hybrid calques in Lanstyák’s terminology, and the term corresponds to Haugen’s loan blend category.

Calques encompass Haugen’s loan translation and semantic loan category, and derived calques, as well. I include the derivational layer in the system as it is necessary to set apart hybrid stems and derived hybrids. Theoretically it is possible to have derived calques (Benkő 1998) in a contact variety, but the CEst database does not contain any. The derivational layer is used for finer distinction in my analysis, however, as derivation and compounding overlap with many borderline cases the derived categories are not essential to isolate.

2.2. Estonians in Canada (socio-historical background of the contact situation)

Immigration of Estonians to the New World occurred in five phases (Williams 1997):

(i) Colonial and Frontier Period, 1627–1896

(ii) Early Twentieth Century, 1897–1939

(iii) World War II, 1939–1945

(iv) The Refugee Period, 1945–1954

(v) Independence Period, 1991–Present.

Estonians came into contact with English speakers first in the 17^th century. Early 19^th century economic reforms and urbanization triggered emigration (Aruvald 1975), which culminated in the second half of that century. Alberta was popular among Estonian peasants as a destination as there were large uninhabited lands in that province and the landscape reminded them of the home country, the regions (Tartumaa, Võrumaa) they left due to lack of free lands. Still, they were not able to buy lands for financial reasons until 1908–1910 and had to earn their living in cities.

An interesting fact from that period is that Lutheran Estonians did not settle in rural Quebec due to the strong position of the Roman Catholic Church in that province. However, some of the Estonian immigrants chose Montreal as their destination, but in the 1970s, because of the political changes in Quebec, they left for Ontario, creating a 10,000 population in Toronto, which is considered to be the city with the biggest Estonian population outside Estonia (Williams 1997).

The second period of immigration includes the years of the independent Estonian Republic, when immigrants left as they sought better economic opportunities and escaped from the civil war in 1918–1920.

During the Refugee Period, Estonians left for Sweden and Germany to escape the Soviet occupation. But since Sweden maintained a neutral and, in a way, friendly relationship with the Soviet Union and since Germany was in ruins, a large number of Estonians chose to immigrate to North America. Between the years 1947 and 1952 around 13,000 Estonians settled in Canada, and their number reached 20,000 in the early 1970s (Aruvald 1975). This coincided with the baby-boom period of the Canadian Estonian community, as these refugees were mostly in their 20s and 30s at the time of immigration, and had children in subsequent years (Kulu 1992: 96).

In the late 1940s the main cultural center for the Estonian immigrants was Montreal, changing for Toronto in the 1950s due to economic reasons. Job opportunities and the strong Estonian community in Toronto (newspapers Our Life and Free Estonian, a cultural center called “Estonian House”, a theatre and various youth organizations, etc. attest to this) attracted new immigrants to Toronto.

During the years of the Soviet occupation, the immigration of Estonians to Canada was insignificant. The new wave started after 1991, and in six years around two or three thousand Estonians arrived in the New World (Williams 1997).

According to census data (Canada Census 2006), most recently 8,485 people claimed Estonian as one of their mother tongues (of whom 8,240 people declared Estonian as their single mother tongue, while 240 as one of their multiple mother tongues).³

In conclusion, language contact between Estonian immigrants and speakers of Canadian English has had a range of outcomes, depending on various factors such as age, occupation and residence of the immigrant; and the strength of the social network within the community.

3. The corpus and methodology

I collected the data from the forum pages of the Eesti Elu (Estonian Life), an online newspaper of Canadian Estonians, in March, 2007. On March 19, 2007 the forum contained over 1,766 entries under 188 topics.

The comments were partly in English and partly in Estonian, with some of the participants including information in both languages to reach more people; especially announcements and invitations were bilingual.

Data collection involved the pinpointing of the phonological, morphological, syntactic and semantic features that seemed to have evolved or changed under the influence of English. I had the whole corpus re-read and the selected examples checked by a native speaker of Estonian from Estonia. I excluded all ambiguous data, for example, when the immigrant or homeland dweller status of the person was not clear, or if the data seemed to be mistyped.

In the next phase of my research, I grouped the collected data into main categories on the basis of their grammatical status in the case of loans, and on the basis of morphemic importation and morphological structure for the classification of loan blends and loan shifts. I analyzed instances of syntactic transfer, defining congruence differences, word order discrepancies and argument structure deviations.

Drawbacks caused by the choice of forum discussions as a source of data were the following: it was impossible to decide which generation of Canadian Estonians members of the discussion belonged to. So when I refer to CEst in this paper, I mean the language use of Canadian Estonians in the forum of the Eesti Elu newspaper with no reference to generation, age, gender or language proficiency, as finer distinction is impossible to make on the basis of this type of data.

The language use characteristic of forums and special features of web-language in general presented the other main problem in my research. Especially orthographic and stylistic mistakes, as compared to the EEst standard, occurred, but these can be attributed to informal language use, and they are not necessarily indicators of English influence or language attrition. However, some of the orthographic mistakes might imply insecurity in the informants’ knowledge of Estonian, especially the errors in marking the length of the consonants, e.g. the differentiation between voiceless and voiced stops (t and d in particular).

Important factors influencing the language use in the forum discussions were the way Canadian Estonians characterized their variety and related to the EEst variety. Canadian Estonian speakers considered their variety as a distinct one which had both “old” Estonian characteristics and new contact phenomena. One of the participants claimed that it was Canadian Estonians who preserved a pure version of the Estonian language, not having undergone Russian influence and not willing to be Americanized like the young nowadays in Estonia (cf. also Vahtre 2004: 279).

One of the comments (from an EEst speaker) sums up the attitudes and relations between the two speech communities, how they reflect on the two main types of contact influence, lexical and structural transfer.

(1) “Teie ei seedi meie slängi või isegi kõnekeelt, meile teeb jälle nalja, kuidas raskete õigekirja- ja lauseehitusvigade saatel või siis lihtsalt niisama kohmakas keeles slängi ja russisme-anglitsisme-fennisme kritiseeritakse.”
’You cannot stand our slang or even our spoken variety, on the other hand, it’s funny for us how you criticize slang and the Russicisms, Anglicisms and Fennicisms while you write making serious mistakes in orthography and sentence structure or in a language which is clumsy anyway.’

4. Discussion

One of the most puzzling questions I faced analyzing the corpus was the low number of lexical borrowings, both direct and indirect, even though the presence of structural transfer should have implied heavy lexical borrowing, involving even basic vocabulary, as suggested by Thomason and Kaufman’s framework (1988). I explain this phenomena by the attitudes of the community despising direct transfer from English on the one hand, and, in connection with this, by the participation of EEst speakers in the discussions, on the other.

Winford (2003) also emphasizes the role of the community norm in determining the extent of lexical transfer claiming that “rate of borrowing (especially of nonce loans) is dependent on the norms of the community behavior, rather than on lexical need” and that borrowing “seems to be motivated by accommodation to the conventions of use in the social networks of the speakers” (2003: 40).

4.1. Lexical borrowings

4.1.1. Direct loans

4.1.1.1. Phonological adaptation

As regards the phonological and morphological adaptation of loanwords, little can be said since the written form of borrowings reflects the pronunciation only to some extent and only in a number of cases. Words are written either using the English original orthography (trafficus ‘in the traffic’) or changed to comply with the rules of Estonian (gei ‘gay’).

However, the case of hybrid stems, as in the example organisatioon ‘organisation’, indicates that there is an in-between category. Hybrid stems show partial phonological adaptation, represent a typical contact phenomenon, and serve as a unique form of morphological adaptation.

There are some general tendencies that occur throughout the corpus and are easier to identify. The phonological adaptation of English loanwords in CEst involves the substitution of sounds missing from the Estonian sound inventory and modification of the phonotactic characteristics of words to comply with the rules of the recipient language. Devoicing of oral stops (helptesk,⁴ luuser < loser and paar < bar), lengthening (internett < internet, rokk < rock) and monophthongization (trennima < train, beeb(i) < baby) were the most prominent changes.

4.1.1.2. Morphological adaptation

The analysis of morphological adaptation is hindered by the fact that loanwords often occur in an unmarked form in texts, either because they require zero nominative case even according to the rules of standard EEst, or they are inserted into the discourse as nonce-borrowings similar to code-switches, which are also phonologically unassimilated.

In analyzing the morphological adaptation of English loanwords into CEst, I use the typology Pedaja (2006) developed for EEst on the basis of morphological naturalness. The main characteristics of morphological adaptation of loanwords in CEst comply with general borrowing tendencies, as loanwords in CEst enter the open and productive morphological classes in the system, mainly the 0:i type both in the cases of nominals and verbs (luuser : luuseri,  rokkima ‘to rock, to include rock music’). This latter phenomenon can be explained by the hypothesis that claims that “verbs often get borrowed as non-verbs — not necessarily as nouns, but simply underspecified for part-of-speech membership” (Wichmann and Wohlgemuth 2005: 14). The reason behind the prevalence of the 0:i type both in nouns and verbs might be the fact that in borrowing verbs the stem is an i-stem noun (Pedaja 2006: 75), e.g. CEst trennima ‘to train’ < trenn : trenni and tippima ‘to tip’ < tipp : tippi. It is interesting that these verbs do not occur in EEst, only in analytic structures such as trenni tegema ‘lit. to do training’ and tippi andma ‘lit. to give a tip’.

There are cases where the paradigm the loanword is assigned to differs in EEst vs. in CEst (although there is variation within EEst as well). For example, the word marketing occurs in the form marketingut in the partitive case in the CEst corpus, while in Pedaja’s list and also according to my EEst informant, it has the partitive form marketingi in EEst. The reason behind this discrepancy can be that CEst speakers assigned it to the 0:u type by analogy to the EEst -ing type, e.g. looming ‘creation’ which shows the 0:u alternation and has the partitive form loomingut.

Simplification is also present in the morphological adaptation processes, grade alternation is usually avoided, if it is present, it involves the less marked weakening type (Pedaja 2006: 79), e.g. in CEst link – lingi ‘link’.

Nouns are borrowed as stem morphemes, while adjectives and verbs vary in this respect. There are examples with and without thematizing suffixes as well. Pedaja (2006: 75) suggests that borrowing adjectives without a thematizing suffix is getting increasingly productive in EEst. In my corpus, out of the 6 adjectives, two were borrowed without a suffix (digi ‘digital’, privaat ‘private’), the others were added the -ne ending, e.g. sentraalne ‘central’.

Verbs were usually borrowed without a suffix, conforming to the direct insertion strategy (Wichmann and Wohlgemuth 2005). There was one example in the corpus in which a thematizing suffix was added to the stem (importeerida as opposed to EEst importida).

While compounds and derived words are mostly borrowed with importation (Pedaja 2006: 88), there are cases in which one of the parts of the compound or the ending is recognized and changed for its recipient language counterpart. These are hybrid compounds and derived hybrids, respectively.

Although derived hybrids also represent borderline cases, they can be considered rather as direct transfers. In these cases, the –ize and –ic endings of the words are recognized and replaced with –eeri- and –ne, respectively, e.g. Ameerikaniseerida ‘Americanize’ and fantastine ‘fantastic’ (cf. EEst fantastiline).

There are fuzzy boundaries between direct and indirect transfers, especially as hybrid calques are concerned, as they show characteristic features of both types. In the majority of the hybrids in the corpus, in 11 out of 13, the second part of the compound was the Estonian element, e.g. day care susteemis ‘in the day-care system’, customer service treeningutel ‘on customer service trainings’ and Refugee staatust ‘refugee status (partitive)’, but abstrakset cristianityt ‘abstract Christianity (partitive)’ and parkimis loti taga ‘behind the parking lot’. The congruence is usually missing from these examples, which might promote the tendency of borrowing adjectives without a thematizing suffix.

4.1.2. Indirect transfer

Among the indirect transfer phenomena, the homogeneous calque type prevails. Some calques were used throughout the corpus, e.g. täisaegne (also in the form täis-ajalise) from the word full-time; or the expression Hästi tehtud! from Well done! and Proovige ära! from Try it out! Some of the calques are created in order to make the meaning of the expression clear to EEst readers and are not exact equivalents of the CEng original Jello sees maadlemist ‘jelly-wrestling’ (lit. ‘wrestling in Jello’), meditsiiniabi kaart ‘medical card, lit. medical help card’. Abbreviations a.m. and p.m. are also calqued, as e.l. ‘before midday’ and p.l. ‘after midday’ as in pühapäev 12:00 e.l. – 4:00 p.l. ‘12:00 e.l. – 4:00 p.l. Sunday’.

Semantic loans also form a significant group if we include among them all the verbs which were calqued not only with their meaning but also with their argument structure. These verbs are responsible for the changes in the morphosyntactic features of CEst, especially argument patterns and case marking. It is interesting to see that it is mainly verbs that are transferred as semantic loans, while nouns are borrowed as “pure” loanwords. Moreover, the nouns that can be analyzed as semantic loans in CEst are loanwords in EEst as well, so the phonetic similarity might have encouraged Canadian Estonians to use them as Estonian equivalents of a given English word (although they proved to be “false friends” in some of the cases): eksemplaar ’copy’ in the meaning ’example’, paragrahv ‘article, passage’ in the meaning ‘paragraph’ and positsioon ’position (place)’ used in the meaning ’workplace’. This strategy, however, is present in case of the verbs as well, e.g. registreerima ’register’ in the meaning ’register at a school’ or genereerima ’generate’ in the meaning ’to generate profit’.

It is interesting to see that it is not loanwords that occur to the greatest extent in the data as English interference phenomena. The fact that semantic and syntactic transfer features are more common in the language use of Canadian Estonians can be explained by the intensity of the English influence and the unique communicative situation (the participation of L1 EEst speakers in the discussions).

4.2. Structural transfer

The most important structural changes affect the passive construction, the use of verbs, the case system, and the structure of sentences.

4.2.1. The passive construction

The passive construction has two distinct uses in Estonian, the subjective action passive and the resultative (stative) passive (Erelt 2003: 102). The first type is also called subjectless passive (Erelt et al. 2000: 364), its main function is to “impersonalize an agent expressed as the grammatical subject of the clause” (Erelt 2003: 54): Joostakse. ‘Someone is running’ (Erelt 2003: 102). The second type of passive, the resultative (stative) expresses a condition that is the result of an action. Uksed olid avatud (Erelt et al. 2000: 365 and Erelt 2003: 103) ‘The doors were open’.

In the CEst corpus, out of the six instances where the passive form used differs from the EEst standard variant, four apply the second type of passive. The reason I claim that these examples can be explained by English influence is that the structure of the resultative passive matches the English model.

In the following example, the resultative passive is used instead of the action type as the result of the direct translation of the English structure (be plus past participle) which provides a form expressing the present perfect action passive in Estonian.⁵

(2)                           on see kutsutud                          “ühiskondliktöö”

                  be  it    call                                  community work

                  PRES    PASS.PAST.PARTP     NOM⁶

                  ‘it is called “community work”’

The first type of passive occurs in example (2) with the indication of the agent by means of the NP + poolt ‘by someone’ phrase. However, in this case the partial object required by the verb rules out the use of first passive, as the appearance of the agent in the surface structure is only allowed with the total object in the first type of passive.

(3)                          Seda trükiti Torontos, Ontarios, Kanadas, Estoprint, Ltd. poolt.

                  ‘It was published in Toronto, Ontario, Canada by Estoprint Ltd.’

In case of the passive, of the two forms present in EEst, the resultative passive is taking over the role of the action passive in CEst, as English structures are rather similar to the resultative type. The action passive also loses its importance in expressing general subjects as it is replaced by the English-like third person plural or the word meaning ’people’.

There are other cases where not only the form of the passive is contact induced, but the use of the passive itself diverges from the Estonia Estonian language use. For instance, the following sentence See on tehtud vabatahtlike poolt meie kõigi heaks ‘It is done by volunteers for the benefit of all of us’ would more apply to the EEst norm as ‘Volunteers do this for the benefit of all of us’. Although in English both sentences are possible, in EEst passive is not used in cases where the agent of the action is present in the sentence.

4.2.2. Argument structure

Three main changes concern the use of verbs; besides the modification of the argument structure, there is a tendency that involves the evolution of the main verb into an auxiliary verb (’do’):

(4)                          No estieks peaksite te kõik eesti keelt oskama ja ma küll ei tea, et äkki te teetegi seda.

              ’Well, first of all, you should all speak the Estonian language and I don’t know maybe you do’.

In the following instances, the number of the arguments has not been changed, but the form, i.e. the case the argument receives, matches the English model rather than the EEst one. In example (4), the illative case is replaced by comitative under the influence of the English counterpart of the verb (fall in love with).

(5)                           Olin Eestis suvekuudel, ja armusin Gin Long Drink’iga

                  ’I was in Estonia in the summer months and fell in love with Gin Long Drink.’

The argument structure is also modified in earlier loan verbs which coincide in form with their English equivalent even if the source language is not English (example 5). The verb diskrimineerima ‘discriminate’ requires partitive case but in the sentences in the corpus (three tokens from different speakers), the equivalent of against is used, the postpositional phrase with the head vastu ‘against’.

(6)                          diskrimineerite korralikude, töökate eurooplaste vastu

                  ‘you discriminate against decent and hard-working Europeans’

4.2.3. Agreement in the case system

Besides the changes in the choice of the case there is another tendency which concerns the case marking system. In the Estonian language, “[d]emonstratives, adjectives (including ordinals), present participles and ja-agent nouns agree with their heads in case and number” (Erelt 2003: 113). In the CEst corpus, the EEst rules for agreement are modified, giving way possibly to a new system. In EEst out of the fourteen cases, four (the abessive, comitative, terminative and the essive) have partial agreement, which means that the attribute is in the genitive and agrees only in number. In the other cases, there is full agreement, i.e. the attribute agrees in case with the head. In CEst, however, the system has started to change. Partial congruence dominates in all of the cases, except for the genitive in which there is no agreement at all. Rules for the agreement between the attribute and the noun are thus modified in a way that the genitive becomes the default category.

The CEst agreement patterns diverge from the EEst variety to different extents. If the noun has an attribute and a determinant as well, CEst forms vary from partial agreement to no agreement. In the former case the attribute and the noun agree in case, while the determinant is in genitive (I included the EEst versions in brackets):

(7)                          partitive

                  niisuguse     enne           sõjaaegset           keelt

                  such    pre    war                                       language

                  GEN            PART                                   PART

                  (niisugust enne sõjaaegset keelt)

                  PART          PART        PART

                ’such pre-war language’

In the latter case, there is no agreement; neither the attribute, nor the determinant agree with the noun in case:

(8)                          allative

                  selle          uue      impeeriumile

                  such          new     empire

                  GEN         GEN    ALL

                  (sellele       uuele    impeeriumile)

                  ALL         ALL    ALL

                  ’to such a new empire’

4.2.4. Object marking

Case marking is modified not only as a result of argument structure changes, but also because the rules of object marking undergo change. The use of the three cases, nominative, partitive, and genitive, are rearranged following general simplification patterns and under the influence of English, which reinforces non-standard features in CEst.

Figure 1 shows the directions of change, in which N, P and G stand for the nominative, partitive and genitive, respectively.

Figure 1. Directions of changes in object marking

Marking of the total and the partial object are interchanged in CEst, instead of P, G is used (in 7 instances and N in 5 instances, out of which 2 are plural), while P is used in all of the cases (7) where G should have been used to mark the total object.

P instead of G (total object):

(9)                          Lugesin labi neid kommentaare   (neid kommentaare instead of need

kommentaarid).

               ‘I read the comments.’

G instead of P (partial object):

(10)                      Kui otsid töö koha…                   (töökoha instead of töökohta)

               ‘If you are looking for a job…’

In the imperative, G is used instead of N, and N replaced P in the imperative and in negation.

G instead of N (imperative):

(11)                      too kaasa kohvri                          (kohvri instead of kohver)

               ’Bring a suitcase with you!’

N instead of P (imperative):

(12)                      Või otsi ÜEKN sait!                     (sait instead of saiti)

               ‘Or search for the site of ÜEKN!’

N instead of P (negation):

(13)                      …ei ole olnud võimalus käia LCBOs       (võimalus instead of võimalust)

               ’…there wasn’t any chance to go to the LCBO’

However, in the majority of the sentences the object marking abides by the EEst rules. Nevertheless, the above-mentioned examples were written by different participants of the forum discussions and might indicate the onset of change. Similar tendencies occur in Sweden Estonian (Maandi 1989), which can be explained by various factors. In the 1950s several thousands of the Sweden Estonians immigrated to Canada (Maandi 1989: 228). This might have contributed to the same tendencies in the two varieties or might have reinforced the internal change that is ongoing even in EEst, as in colloquial EEst there are examples for the change of the partial object for the total one (P for G) and the other way round (N or G for P), the increase in the use of P being the prevailing tendency (Erelt et al. 2000: 379).

4.2.5. Order changes in the phrase and in the clause

The order of the constituents is also modified in CEst, not only in phrases, leading to the dominance of post-head modifiers, but also in simple and complex sentences, creating new syntactic structures, e.g. piping and pseudo-clefting.

There are several examples in the corpus where the order of the components reflects the English model. For instance, in the following examples the structure of the phrase containing a determiner is changed:  üks veel tõrjuv põhjus ‘one more parrying reason’ and Üks veel asi ‘one more thing’ where the EEst versions would have the order veel üks tõrjuv põhjus and veel üks asi (Erelt et al. 2000: 422) as in EEst the adverbial modifier of the head precedes the head in the phrase containing a determiner.

The structure of complex sentences has also undergone changes in CEst as compared to the EEst variety, producing a new mixed type of sentence transformation. In EEst, the transformation known as wh-movement is not possible, i.e. the constituent of the embedded sentence cannot be moved to the main sentence. In CEst, however, piping does occur, so instead of the structure mida EKN-i rahvas arvas, et kus võiks ESTO 2008 toimuda ‘what did the EKN people think where the ESTO 2008 would take place?’, example (13) follows the English model:

(14)                      …kus EKN rahavas arvas, et ESTO 2008 võiks toimuda?

                  ’…where did the EKN people think the ESTO 2008 would take place?’

Transformation occurred in the following examples as well, in which the English construction was copied by CEst. These sentences can be identified as subtypes of pseudo-clefting in which the structure “what is (adjective) is that”, or more generally “what VP is that” is adopted in CEst. The EEst variant of the example (14) would omit the clefting-like part of the structure and would sound as kurb on (see) et with the structure “(adjective) is that” where the element see is a correlative and optional (Erelt 2003: 119).

(15)                      Tiiu, mis on kurb on et sa ei ole seda Eesti ühiskonda Kanadas kunagi näinud.

                  ’Tiiu, what is sad is that you have never seen this Estonian community in

                  Canada.’

In conclusion, it seems that the changes discussed in this section occur in the areas of grammar where the EEst system differs from the CEng one to the greatest extent. Structures undergo either mere simplification or are modified along English patterns. As common in contact situations, the two linguistic systems, CEng and CEst, become more similar in a way that the more complex subsystems of CEst are liable to change and conform to the English pattern.

According to Fenyvesi (2005), changes in a contact variety might arise as results of language attrition, interference from the dominant language, or as a combination of these two factors (under multiple causation). As regards CEst, it is not always possible to decide which tendency can be attributed to which features, however, there are cases in which one explanation seems more plausible than another. Changes in agreement between the attribute and the noun simplified the system in CEst, in this case English influence could not have been directly involved, as attributive agreement is missing from English, this tendency is rather a language attrition feature in CEst.

It is probable, however, that changes in the argument structure of verbs are attributable uniquely to English influence, as the new structure is closer to the English form than to the EEst one.

Changes in the object marking system can be explained by multiple causation, as some of these tendencies (e.g. the use of P instead of G and vice versa) are present also in non-standard EEst varieties, so the CEst form could have evolved as result of internal development. However, the use of N instead of P seems to have emerged as the result of contact with English. Fenyvesi (2006) observed about the language use of American Hungarians that in a number of cases the non-standard forms occur when “non-standard usage happens to correspond to contact-induced usage” (2006: 172).

The nature of my data source provided limited possibilities as regards the analysis of the status of CEst. General statements that would apply to the entire CEst speech community cannot be made, as the choice of informants was not representative, and the forum discussions rarely revealed any information concerning the generation, age, social status or gender of the participants, which would be vital in order to conduct a sociolinguistic study on the CEst variety. Although the treatment of the CEst speakers as a homogenous community is questionable, I think the forum serves as a frame for language use with its norm negotiated in the discussions. The analysis of the lexical and structural features different from the EEst standard provides insights into the changes CEst has undergone. Moreover, on the basis of these features, it is possible to claim at least that there are members of the CEst speech community whose language use shows features characteristic of language shift.

5. Conclusion

In this paper, I have investigated contact induced phenomena in Canadian Estonian, analyzing the language use in forum discussions. On the basis of this analysis, it seems plausible to claim that Canadian Estonian shows features characteristic of language shift that are typical of the fourth stage of Thomason’s modified borrowing scale (Thomason 2001). What is interesting, however, is that increased structural transfer is not accompanied by heavy lexical borrowing.

Comparing the socio-historical background of the contact situations and identifying driving forces behind interference and borrowing, I have sought to explain the differences between the contact phenomena in CEst and EEst.

In CEst, structural transfer prevails. The limited nature of lexical borrowing can be accounted for by the purist attitudes of the CEst community and the involvement of EEst speakers in the forum discussion. That is why I think that the social and sociolinguistic characteristic of the contact situation should be studied in more detail. Winford draws attention to the fact that there have been few studies carried out on how “degrees of social solidarity and accommodation…[…]…influence patterns of lexical or other borrowing” (Winford 2003: 39).

Partly resulting from the limitations of the corpus, only a number of features and the main tendencies of CEst could be studied in this paper. For a more thorough analysis, oral data would be required from speakers belonging to different generations.

 

Works Cited

• Aruvald, Priit, ed. 1975. The Estonians in Canada. Volume I. http://www.ajalugu.ca/en/2.html (Date of access: 22/02/2007)


• Aruvald, Priit, ed. 1975. The Estonians in Canada. Volume II. http://www.ajalugu.ca/en/2.html (Date of access: 22/02/2007)


• Benkő Loránd. 1998. A történeti nyelvtudomány alapjai [Basic principles of historical linguistics]. Budapest: Nemzeti Tankönyv­kiadó.


• Canada Census 2006. Topic based tabulations. Statistics Canada. http://www12.statcan.ca/english/census06/data/topics/RetrieveProductTable.cfm?ALEVEL=3&APATH=3&CATNO=&DETAIL=0&DIM=&DS=99&FL=0&FREE=1&GAL=0&GC=99&GK=NA&GRP=1&IPS=&METH=0&ORDER=1&PID=89186&PTYPE=88971&RL=0&S=1&ShowAll=No&StartRow=1&SUB=701&Temporal=2006&Theme=70&VID=0&VNAMEE=&VNAMEF=#FN1 (Date of access: 02/02/2008)


• Erelt, Mati, Tiiu Erelt and Kristiina Ross, eds. 2000. Eesti keele käsiraamat [Handbook of the Estonian language]. Tallinn: Eesti Keele Sihtasutus.


• Erelt, Mati (ed.). 2003. Estonian language. Tallinn: Estonian Academy Publishers.


• Fenyvesi, Anna. 2005. Hungarian in the United States. In: Anna Fenyvesi, ed. Hungarian language contact outside Hungary: Studies on Hungarian as a minority language. Amsterdam: Benjamins, 265–318.


• Fenyvesi, Anna. 2006. Contact effects in Toledo, Ohio, Hungarian: Quantitative findings. In: Greg Watson and Pekka Hirvonen, eds. Finno-Ugric language contacts. Frankfurt: Peter Lang GMBH, 157–177.


• Haugen, Einar. 1950. The analysis of linguistic borrowing. Language, 26: 210–232.


• Kulu, Hill. 1992. Eestlased maailmas [Estonians in the world]. Tartu Ülikool. Majandusgeograafia kateeder: Tartu.


• Lanstyák, István. 2006. Nyelvből nyelvbe. Tanulmányok a szókölcsönzésről, kódváltásról és fordításról [From language to language: Essays on lexical borrowing, code switching and translation]. Pozsony: Kalligram.


• Maandi, Katrin.  1989.  Estonian among immigrants in Sweden. In Nancy C. Dorian, ed. Investigating obsolescence: Studies in language contraction and death, Cambridge: Cambridge University Press, 227-241.


• Pedaja, Kati. 2006. Uute laensõnade kohanemine eesti keele morfoloogilise süsteemiga [The adaptation of new loanwords to the Estonian morphological system]. Tartu, MA-Thesis.


• Thomason, Sarah G. and Terrence Kaufman. 1988. Language contact, creolization and genetic linguistics. Berkeley: University Press.


• Thomason, Sarah G. 2001. Language contact: An introduction. Edinburgh: Edinburgh University Press.


• Vahtre, Lauri. 2004. Eesti ajalugu  [History of Estonia]. Tallinn: Ilo Kirjastus.


• Williams, Michael. 1997. A history of the Estonian community in North America. http://www.geocities.com/Vienna/8921/colonial.html (Date of access: March 8, 2007)


• Winford, Donald. 2003. An introduction to contact linguistics. Oxford: Blackwell.


• Wichmann, Søren and Jan Wohlgemuth. 2005. Loan verbs in a typological perspective. http://loanverb.linguist.de/Wichmann_Wohlgemuth_verb-borrowing-typology.pdf (Date of access: February 23, 2007)

 

Notes

^* Derived calques are derived CEng words which have both the stem and the derivative suffix “translated” into CEst. The Eesti Elu corpus under investigation does not contain any examples for this rare type of borrowing. ↩

¹ I cite all of the examples verbatim from the corpus without making any orthographic changes or adding any missing diacritic marks. ↩

² Throughout this paper, CEst, CEng, and EEst stand for Canadian Estonian, Canadian English, and Estonia Estonian, respectively. ↩

³ The figures were indicated in this way in the source, hence the error in the sum of the numbers. ↩

⁴ The letters <d>, <b> and <g> that are part of the Estonian alphabet do not refer to voiced stops, they are shorter (voiceless) allophones of the phonemes p, t and k respectively, occurring in voiced environments. ↩

⁵ The EEst form uses the action passive and the cases assigned to the arguments of the verb are also different from the CEst example:
seda     kutsutatakse ühiskondliku-ks töö-ks
it          call               community work
PART PASS.PRES TRANSL ↩

⁶ Throughout this paper, the following abbreviations are used in the glosses: ALL = allative, GEN = genitive, NOM = nominative, PART = partitive, PARTP = participle, PASS = passive, PAST = past, PRES = present, TRANSL = translative. ↩