<<

School of Language and Literature

English G3, Bachelor's Course Supervisor: Mikko Laitinen Course Code: 2EN10E Examiner: Christopher Allen Credits: 15 Date: 17 January 2014

Collective in English used in Sweden A corpus-based study on number concord with collective nouns

Petra Örlegård

Abstract The purpose of this thesis is to investigate how Swedes writing in English construct number concord with collective nouns. This was done by studying three primary corpora: the Swedish English Newspaper Corpus (SWENC) and the press sections of Frown () and F-LOB (British English). The findings were compared with the results in the Blogs in English by Swedes Corpus (BESC), Frown (American English) and F-LOB (British English). The SWENC contains texts from three online newspapers and one corporate newsletter in English, all of which are written by Swedes. The BESC contains texts from Swedes blogging in English. Frown and F-LOB contain texts from fifteen text genres in the 1990s. The results in the SWENC are discussed and compared with the press sections of F-LOB and Frown. The results are also compared with the BESC, Frown and F-LOB in order to see whether there is regional and stylistic variation. The results show that Swedes prefer singular verbal concord with collective nouns even though forms occur which seem to be closer to F-LOB Press (BrE) than Frown Press (AmE). In , the SWENC differs from the press sections of Frown and F-LOB in constructing pronominal number. This difference could be influenced by Swedish usage which allows both singular and plural pronominal number with collective nouns. The study also shows that plural forms of political parties that seem to constitute collective units show variation in Sweden whereas such nouns take only plural concord in both American and British English.

Keywords: , American English, British English, collective nouns, concord, corpus, news genre, plurality, singularity, Sweden

Petra Örlegård Table of contents

1 Introduction 1

2 Collective nouns and concord 3 2.1 Variation in number concord with collective nouns 3 2.2 Proper nouns that constitute collective nouns 5 2.3 Factors which agreement 6 2.4 Previous studies on agreement with collective nouns 9

3 Materials and methods 11

4 Results 14 4.1 Findings in the SWENC, F-LOB Press and Frown Press 14 4.2 of the news genre with the BESC, F-LOB and Frown 19 4.2.1 Comparison of the political parties with the plural forms of political parties 21 4.3 Discussion of the findings with previous findings 23

5 Conclusion 25

References 26

Appendix 28

Petra Örlegård 1 Introduction This thesis deals with collective nouns, more specifically with how Swedish journalists construct their choices of verbal and pronominal number with collective nouns. Collective nouns can be used with singular or plural agreement, as in illustrations (1) and (2). (1) In 2008 the couple were in trouble with the police. (Swedish Wire in the SWENC) (2) While Malmö resident Philipp Marra moved to Sweden about a decade ago simply because his partner got a job there, the couple has stayed on. (The Local in the SWENC) Arguably, in prescriptive grammar, a singular goes with a singular and a plural noun takes a plural verb. Collective nouns are, however, singular in form but plural in meaning as they denote “a group of people or animals or /…/ institutions” (Greenbaum, 1996:103). In fact, collective nouns can either constitute a group acting as a unit or a group acting as individuals. According to Collins Cobuild (1990:37), a singular verb is used when the is thought of as a unit whereas the plural verb is used when the collective noun is thought of as a number of entities. Consequently, number concord varies. Leech, Deuchar and Hoogenraad (2006:190) discuss this problem of number concord with collective nouns. They argue that one needs to distinguish grammatical concord from notional concord. Grammatical concord refers to prescriptive grammar and syntactic agreement, i.e. verbal and pronominal agreement is based on the number of the noun. Notional concord, on the other hand, follows semantic agreement, i.e. variation in agreement with collective nouns that is based on the semantic meaning of the collective, e.g. whether acts as a single unit or as a group of individuals (ibid). Levin (2001:159) discusses this in his doctoral dissertation and claims that there are a number of functional factors, e.g. syntactic, semantic and stylistic factors, behind the variable usage of number with collective nouns. However, plural concord with collective nouns is “common enough in speech, but still tend to be frowned on in serious writing” (Leech et al, 2006:190). Therefore, Leech et al (2006) claim that singular concord is more frequent in writing than in speech. Concord with collective nouns is assumed to be problematic for both native and non- native speakers of English (Levin, 1998; Bauer, 1994:61fff). These problems are brought about by the fact that there is regional variation, e.g. American English (AmE) and British English (BrE) (Levin, 2001:60; Biber et al, 1999:19), and by the fact that there is variation

1 within the group of collective nouns, e.g. government has a political reference whereas family refers to a number of people within a family. Not only that, there is variation as regards to which is used when a singular verb co-occurs with a collective noun (Aremo, 2003), as in illustrations (3) and (4) below. (3) The family is not happy in their house. (Invented example) (4) The government has to make its decision on Sunday. (Invented example) Even if a singular verb is used in both examples, the pronoun number varies. In addition, collective nouns in AmE prefer to take singular verbs and , though plural pronouns are also possible, e.g. with couple (Quirk et al, 1985:759), whereas the selection is clearly different in BrE: both singular and plural verbs and pronouns are possible with collective nouns (Levin, 2001:160). Levin (2001:159) however observes that plural forms of pronouns are equally common in both AmE and BrE in spoken English. Based on the fact that there clearly is variation in number concord with collective nouns in English, e.g. regional differences, differences between spoken and written English and between nouns, the Swedish perspective thus makes an interesting study. This thesis will therefore investigate whether the frequencies of verbal and pronominal number with collective nouns in the Swedish corpus show variation. Furthermore, this thesis will explore which factors affect the concord with collective nouns. The aim of this study is to explore the choice of concord with collective nouns. This is done through a corpus study which examines if there is variation in usage and if Swedish journalists are consistent when choosing number verbs and pronouns with collective nouns. The findings from the corpus will be analysed in order to find out whether concord with collective nouns is based on various linguistic factors as explored in Levin (2001) or on grammatical concord, i.e. singular pronouns and verbs. Levin (2001:53) argues that there are stylistic factors that affect number agreement and it is assumed that different genres show stylistic differences. This thesis will thus compare the news genre with blog texts to see if there are stylistic differences that could have affected concord. Moreover, this thesis will look at whether there is variation based on which collective noun is involved. The collective nouns that will be analysed are presented in Table 1 on page 3 and are taken from earlier research as well as English grammar books. In addition to the list of collective nouns, proper nouns that constitute collectives will be looked into and they include the names of corporations and organizations like IKEA, BBC, the UN, SAAB, VOLVO and SAS. The names of political parties are also included. Investigating the abovementioned problems will include the following aspects:

2 1. Do Swedish journalists follow the American English variety, the British English one or a mixture of both in verbal and pronominal concord with collective nouns? 2. Which number concord occurs more frequently with collective nouns in texts written in English by Swedish journalists? 3. How consistent are Swedish journalists writing in English in selecting the number verb and pronoun with collective nouns?

Table 1. List of collective nouns. Army Assembly Association Audience Board Class Club Commission Committee Community Company Congress Corporation Council Couple Crew Crowd Department Family Faculty Gang Government Group Jury Navy Majority Management Military Minority Office Organisation (z) Parliament Party Population Senate Society Staff Team The press The public The state

In order to answer the abovementioned questions, this thesis will have the following structure: Section 2 provides a background on concord and discusses variation in number concord, proper nouns that constitute collective nouns and factors that affect variation in agreement. The changes of number agreement during the twentieth century and the tendencies of AmE and BrE will also be discussed here. Section 3 discusses the materials and methods used for this thesis. The results of the study are then presented and discussed in section 4 and finally, section 5 provides conclusions and suggestions to future studies.

2 Collective nouns and concord

2.1 Variation in number concord with collective nouns Many researchers observe that there is variation in agreement with collective nouns in English (Biber et al, 1999:20; Aremo, 2003; Levin, 2001:30). For instance, collective nouns are used with both singular and plural concord and this variation is more noticeable in BrE (Biber et al, 1999:188; Levin, 2001:37). However, most collective nouns take singular concord (Biber et al, 1999:188; Levin, 2001:30). It has also been observed that the collective noun itself affects the number concord (Levin, 2001:37). According to Biber et al (1999:188), nouns like audience, board, committee, government, jury and public occur primarily with singular whereas staff occurs with plural concord most of the time. On the one hand, nouns like family and crew occur with both singular and plural concord in BrE; on the other hand, they are more likely to occur with singular concord in AmE (ibid). Swan (2005:519), however, notes that in

3 AmE, there are occurrences of plural pronouns with collective nouns and that family can take plural verbs. Levin (2001:65) also notes that plural concord is more frequent with family than with government in AmE. However, both nouns prefer singular concord in written BrE (ibid). The reason can be that government is more likely to be seen as a unit, e.g. that people in a government are collectively making decisions (Levin, 1998). In addition, it is noted in Vannestål (2007:112f) that plural personal pronouns co-occur more frequently than singular personal pronouns with collective nouns. Furthermore, Swan (2005:519) observes that singular and plural forms are sometimes mixed in a sentence in BrE, which is illustrated in (5). (5) The team plays its first game on Sunday and after that they are going to go to New Zealand. (invented example) Aremo (2003) states that variation in concord is also caused by the differences between various types of collective nouns, e.g. collective nouns that can constitute both common and proper nouns and collective nouns that only constitute common nouns. The differences are also illustrated by what the noun itself denotes. Nouns like council, government, senate and congress have a political reference and can be used as proper nouns spelt with an initial capital without a or as common nouns and are modified by the definitive determiner the (ibid). Other collective nouns like family, audience, club, crowd, gang and team only constitute common nouns and refer to specific groups of people. It should also be noted that collective nouns acting as proper nouns are restricted in flexibility with number concord (ibid). Thus, several researchers argue that syntactic and semantic factors influence the number concord with collective nouns (Levin, 2001:159; Corbett, 2006:156). According to Quirk et al (1972:190), the meaning of a sentence is affected if a noun takes either a singular or plural verbal concord, which is illustrated in (6) and (7) below. This variation is based on the fact that “the singular stresses the non-personal collectivity of the group and the plural the personal individuality within the group” (ibid). Thus, when the noun has a singular concord, it acts as a unit, but when the noun has a plural concord, it acts as a group of entities (Aremo, 2003; Biber et al, 1999:188). (6) The club has the resources, and it will not disappoint its supporters. (Aremo, 2003) (7) The club have the resources, and they will not disappoint their supporters. (Aremo, 2003; italics added) Aremo (2003) remarks that there are certain cases when singular concord is required and other cases when either plural or singular concord is possible. For instance, a collective noun

4 with a definitive determiner can take either a singular or plural verb, e.g. The government is/are but a collective noun without a determiner, i.e. a proper noun that behaves as a collective e.g. Congress, can only have singular verbal concord (Aremo, 2003). Furthermore, collective nouns which are preceded by a determiner, e.g. a, each or every as in “Each family”, usually require a -verb concord, i.e. they take a singular verb (Aremo, 2003). Thus, variation in number concord depends on the context, such as on the presence of a determiner, if the collective noun also acts as a proper noun and also on whether the collective noun denotes a unit or a group of individuals (ibid). Quirk et al (1972:190) also divide the collective nouns into three subclasses, i.e. specific, generic and unique. Specific collective nouns are e.g. army, committee and government. Generic collective nouns are e.g. the élite, the clergy and the public. Unique collective nouns often refer to proper nouns acting as collectives, e.g. the Arab League, Congress and the United Nations (ibid).

2.2 Proper nouns that constitute collective nouns Since most of the research on collective nouns is concentrated on common collective nouns, there is very little research on proper nouns that behave as collective nouns (Levin, 1999). Proper nouns such as the BBC, the music band Abba and England, referring to the national football team, are nouns that also act as collective nouns (Levin, 1999). It is observed that plural agreement is often found with such nouns and with the names of organizations and companies as well as sports teams, such as Manchester United (Levin, 1999; Biber et al, 1999:189). There is, however, a difference between BrE and AmE when it comes to agreement with sports teams (Biber et al, 1999:189). Sports teams usually occur with plural concord in BrE whereas AmE prefers singular verbs and singular personal pronouns to go with sports teams (ibid; Levin, 1999). However, if the names of sports teams have a plural form, e.g. New York Rangers, they often take plural agreement even in AmE (Biber et al, 1999:189; Levin, 1999). Singular verbal agreement is also preferred in AmE with national teams, but plural personal pronouns may occur (Levin, 1999; Quirk et al, 1985:318). Bauer (1994:89) claims that the use of plural verbs with sports teams in BrE clearly indicates that e.g. England denotes the national sports team and not the country itself. Bauer explains this by pointing out that, in British and New England Englishes, political collective nouns have a tendency to have singular concord, e.g. Britain is to appeal to the UN. The use of plural verbs with the names of countries, therefore, clearly distinguishes national sports teams from the countries that they represent (ibid). It is also observed in Levin (1999) that

5 plural agreement with sports teams is likely to occur if the full name of the sports team is used, e.g. the Orlando Magic, rather than only Orlando. However, not all proper nouns can be collective nouns and the process of proper nouns becoming collective nouns is called metonymy (ibid). Levin maintains that proper nouns become metonymic collectives if the noun itself represents a collective. Furthermore, Fries (1981) as cited in Levin (1999) argues that the use of plural verbs with proper nouns acting as collectives can be seen as a stylistic factor, e.g. if a company has a plural verbal agreement, it can be seen to signal that many people are working in the company. Levin (1999) however points out that singular verbal agreement is more frequent than plural verbal agreement with the names of companies and organisations. He also found that only 6 % of the British newspaper The Independent have plural verbal agreement with company names and that only one out of 210 tokens in The New York Times has a plural verb with company. Additionally, in his observation he found that on the one hand, plural forms of personal pronouns with such collectives were very rare. On the other hand, plural personal pronouns seem to be preferred with both names of organizations and companies as well as the noun company in spoken AmE (ibid).

2.3 Factors which affect agreement Research shows that singular agreement is preferred with collective nouns in formal written language in both BrE and AmE, whereas variation in number agreement was found to occur more frequently in spoken English (Levin, 1999). Levin suggests that in spoken language, people are more likely to have a relationship with the company that they are talking about, e.g. they might have worked in the company and are more likely to think of the company as a group of people and not as a unit. It can also be argued that grammatical agreement might be more influential in written language rather than in spoken language (ibid). In addition, semantic factors are found to have a very strong influencing factor on agreement, which is shown by the fact that certain verbs used with collective nouns, i.e. non- personal collectives, are very likely to take the singular form, e.g. verbs like consist of, contain and include (Levin, 1999) as in illustration (8): (8) While the party includes social conservative and religious fundamentalists, Hägglund wanted the party to be the conscience of Sweden's centre-right parties. (The Local in The SWENC) Reid (1991:210f) also suggests that semantic cohesion is more frequent than grammatical concord and claims that agreement is determined by its function. Although it has been

6 observed that plural verbal agreement with collective nouns occurs more frequently in BrE than in AmE, Levin (1998) argues that singular verbal concord is more common in most contexts even in BrE. Furthermore, Levin’s research shows that plural personal pronouns are more frequent than plural verbs in both AmE and BrE. A reason for this is the distance between the pronoun and its antecedent. The distance between a pronoun and a collective thus increases the probability of a notional concord (ibid). Verbs, on the other hand, often occur immediately after the noun, which is thus an influencing factor to why singular verb forms are more common (ibid). Since this thesis looks at how Swedish journalists construct number agreement with collective nouns, it has to be noted that both Teleman et al (1999) and Språkriktighetsboken (2005) point out that plural pronominal references are possible in Swedish when the collective noun is grammatically singular. Number concord with collective nouns in Swedish will be further discussed in section 4.3. Reid (1991:210f) illustrates variation in number agreement in The Great Gatsby by F. Scott Fitzgerald, which are shown in (9) and (10), and provides an example of contrasting sentences in (11): (9) My family have been prominent, well-to-do people /…/ (Reid, 1999:210f) (10) Her family is one aunt about a thousand years old. (ibid) (11) The Kent family /…/ have /…/ remind themselves /…/ Her family looks back… (ibid) Reid (1991:210f) claims that family in both cases of example (11) are acting independently; however, the first token in (11) is followed by themselves, which naturally requires a plural verb, whereas there is no pronoun in the second token that influences the verb form. Reid thus argues that writers are given the opportunity to change number concord according to the context, i.e. number concord may vary due to the different meanings that one might want to imply. The contrasting forms in agreement are likely to have been caused by category shift, e.g. player vs. team, person vs. people and not by the collective noun itself (ibid). Even if it is often claimed that plural agreement is used with collectives when they constitute a group of individuals and singular agreement when they constitute a unit, Levin (1998) points out that it is not always easy to determine why singular agreement is preferred in one case and not in others. He illustrates this problem clearly in (12): (12) Her family are quite well-off whereas Celie’s family is poor. (Levin, 1998) There is also variation in verbal agreement with collective nouns when a collective noun is the head of a noun phrase with a post-modifying noun, e.g. a group of soldiers were or a

7 group of former employees was. In these examples, the post-modifying noun is very likely to affect the number agreement and plural verb forms are in such contexts common even in AmE (Levin, 1998). It is claimed in Collins Cobuild English Grammar (1990:37) that a singular pronoun or determiner should be used if the verb that is used with the collective noun is in the singular and that the plural form should be used if the verb is in the plural. However, research has shown that agreement is not carried out in a straightforward way. Studies have shown that collective nouns are flexible in number concord with pronouns and can take both singular (it) and plural pronouns (they) and (its/their), even when the verb itself is either singular or plural (Biber et al, 1999:332). The variation is illustrated below: (13) The Hydro Board were playing tunes with their dams /…/ (Biber et al, 1999:332) (14) The Government has made it clear that it may not /…/ (ibid) Levin (1998) argues that it is easy to see why concord with collective nouns is problematic for non-native speakers of English. Svartvik & Sager (1996:353) suggest that non-native speakers of English should be consistent in their usage and obey grammatical concord, i.e. use singular verb and singular forms with singular collective nouns. Nevertheless, since plural personal pronouns often follow singular verbs in constructions with collective nouns, Levin (1998) suggests that non-native speakers of English should be allowed to use that type of construction as well, at least when there is a distance between the pronoun and its antecedent. Finally, non-native speakers should use singular verbs and pronouns with relative which, since it makes the noun an impersonal unit, whereas plural verbs and pronouns should follow relative that/who, since this refers to the individuality in the collective noun (Quirk et al, 1972; Levin, 2001:32). The difference is illustrated in (15) and (16). (15) The team are full of enthusiasm. (Swan, 2005:519) (16) A team which is full of enthusiasm has a better chance of winning. (ibid) Swan (2005:519) claims that which is in (16) is more natural than who are full of enthusiasm, which thus makes it more likely for constructions with which after a collective noun to take a singular verb. However, the collective noun in (15) co-occurred with a plural verb. Sentences in which collective nouns occur without a are thus more free to co-occur with either a singular or plural verb. Swan (2005:519) also explains the differences between the plural and singular forms and what they imply of the collective nouns:

8 Plural forms are common when the group is seen as a collection of people doing personal things like deciding, hoping or wanting. Singular forms are more common when the group is seen as an impersonal unit. Plural verbs are thus often used when collective nouns constitute a group of individuals, i.e. when it is clear that they are doing things individually, e.g. disagreeing (Leech & Svartvik, 2002:274) as in the family are arguing. In this example, it is clear that singular concord is not possible since the collective noun itself refers to the individuals who constitute a collective, not the collective as a unit. This is further confirmed by Quirk et al (1972:361) who point out that e.g. audience often co-occurs with a plural verb if the individual reactions of the audience are taken into consideration, e.g. The audience are enjoying the show. In contrast, the singular verb is often used when the size of the collective is taken into consideration, e.g. The audience was enormous. There are thus factors that affect the variation of number in cases when the same noun varies its number concord. Arguably, concord with collective nouns is not only notional and the choice of concord seems to some extent to be “grammatically determined” (Bauer, 1994:61). She did an investigation with a very limited data of very formal English in The Times. The conclusion that she could draw was that the number of singular concord occurrences with collective nouns seems to have increased in The Times during the twentieth century. However, it has to be noted that certain nouns were found exclusively with singular concord, e.g. the press and association (ibid). In addition, Bauer (1994:62ff) observes that government was used with either singular or plural concord at the beginning of the twentieth century. Later, which number concord was used depended on whether the noun referred to the British government or foreign governments. Plural concord was used for the British government and singular concord for foreign governments. However, at the end of the century, singular concord is always preferred with government, even if plural concord may occur in some instances. It has to be noted that government was the most frequent collective noun in the corpus (ibid). Aitchinson (2001:88) claims that frequent words seem to be more susceptible to change, which could perhaps explain the fact that the number agreement with government has changed during the twentieth century. Thus, number agreement with infrequent collective nouns is assumed to vary a lot.

2.4 Previous studies on agreement with collective nouns Corbett (2006:21fff) observes that agreement within the phrase is most canonical whereas agreement within the clause but not within the phrase is less canonical. In addition, there is

9 also agreement within the sentence but not within the clause, i.e. relative pronouns and antecedents. In other words, the shorter the structural distance between the collective noun and the targets (verbs and pronouns), the more likely it is for a canonical agreement to occur (ibid). The idea of canonical agreement is related to agreement hierarchy, i.e. how agreement is determined by syntactic and semantic factors (ibid, 2006:21). Phrases with a collective noun and a plural verbal agreement are called “less canonical instances of agreement” (ibid, 2006:24). Thus, he points out that “the committee has” is a more canonical instance of agreement than “the committee have.” It does not mean that plural concord with collective nouns is ungrammatical and illogical, but rather that the number agreement is not reached in a straightforward way as when a collective noun has singular concord (ibid). The degree of also needs to be taken into account as a that could affect agreement with collective nouns (Corbett, 2000:55; Levin, 2001:12f). For example, if the subject denotes an animate, the verb will be plural; if the subject on the contrary denotes an inanimate, the verb will be singular (Corbett, 2000:55). Biber et al (1999:303) note that inanimate nouns tend to occur with an of-construction rather than an s-genitive, as for impersonal collective nouns such as club, college and committee. Jespersen (1909-1949) claims that plural verbs are used only when the noun itself refers to people and not to institutions and animals. In Visser’s (1963:68) research, only a few instances of inanimate collective nouns were found with plural verbs which supports the fact that animacy might affect number agreement. However, Persson (1989) as cited in Levin (2001:12f) challenges the idea of restricting collective nouns to the degree of animacy and claims that the degree of animacy can be applied to other types of nouns that are animate and are thus not only limited to collective nouns. Thus, and mobility were added as semantic features of collective nouns (ibid). The problem with volition and mobility, however, is that it was found that there are nouns that lack volition and mobility which still take plural verbs in BrE (Dekeyser, 1975). Levin (2001:13) thus argues that animacy, mobility and volition should not be considered the defining features of collective nouns since they seem to apply to other types of nouns as well. Nevertheless, they seem to be factors which influence agreement with collective nouns. Corbett (2006:155) defines syntactic agreement as grammatical agreement and this occurs when the agreement is consistent with the number of the controller, e.g. The team wins. In the case of collective nouns, syntactic agreement would thus be reached by singular agreement. Semantic agreement is defined as notional agreement (ibid) and this occurs when a collective noun takes a plural agreement. Thus, Corbett argues that singular common nouns

10 reach both syntactic and semantic agreement, e.g. “The girl lives here.” With collective nouns, however, it is not possible to reach both syntactic and semantic agreement; thus, only one of the two agreements is reached when choosing either the singular or the plural form (ibid, 2006:156). Notion, i.e. semantic meaning, thus becomes more important when choosing number agreement with collective nouns. Variation has been found in several studies and number concord with collective nouns is not always carried out systematically. Summing up, singular concord is more frequent than plural concord in written English (Leech & Svartvik, 2002; Leech et al, 2006; Levin, 2001).

3 Materials and methods The primary material consists of examples of written English by Swedes. Since there was no available corpus, an effort of establishing a new corpus has been made. This corpus includes a selection of newspapers and a corporate newsletter written in English by Swedes and has therefore been named the Swedish English Newspaper Corpus (SWENC). The news genre was chosen due to the fact that it is a genre that shows “a drift towards a more ‘oral’ style” (Hundt & Mair, 1999) and “picks up new trends quickly” (Lindquist, 2009:63). Journalistic prose is also claimed to be innovative (ibid) and one of the more informal genres in F-LOB and Frown (Hundt & Mair, 1999). The SWENC thus provides data for English usage in Sweden. The establishment of a corpus was made in order to conduct an empirical study of English usage in Sweden. Using the corpus as a tool, this thesis will investigate whether the singular verb and pronoun pattern is more frequent than the plural one when it comes to collective nouns. The purpose is also to find out whether English in Sweden follows American English or the British English variety. In order to summarise and discuss the findings, the findings in the SWENC will also be compared with the findings in the press sections A, B and C of Frown (Freiburg-Brown corpus of American English) and F-LOB (Freiburg-LOB corpus of British English). The texts in F-LOB and Frown are from the 1990s. By doing the comparisons, it should be possible to see whether there are similar patterns between one of the two varieties and the Swedish one. It will therefore be interesting to see whether Swedish journalists go after one variety or are innovative in usage, i.e. differ from both varieties. To find concordance hits with collective nouns, the corpus software AntConc has been used. Using AntCorc, it turned out that it was easier to find out number concord by verbs such as has/have, is/are, was/were, consists/consist, includes/include, writes/write and the like.

11 Thus, there are several more hits for many of the collective nouns that are investigated here, but they have been omitted since they have past verbs that are not clear in number such as put, said and thought. The instances that have been included are tokens which has any of the following: (1) singular or plural verb, (2) singular or plural verb following relative pronouns and (3) singular or plural pronouns. Moreover, tokens with pronouns in the next sentence have been included if it is obvious that they refer to a collective noun in the preceding sentence. In order to implement this study, this thesis will also look into concord with collective nouns in Swedish and whether Swedes construct their agreement choices in English based on number concord in the . The main collective nouns that will be analysed in the corpus are provided in Table 1. In total, 41 collective nouns will be investigated in the corpus. The investigation will also look at the names of countries and organisations acting as collective nouns, e.g. proper nouns that act as metonymic collectives like organisations and corporations, e.g. IKEA, BBC, the UN, SAAB, VOLVO and SAS. The collective nouns have been selected because some are frequent and some less frequent; also, some collective nouns are more related to politics, e.g. senate and government while others refers to smaller groups of people, e.g. family, team and committee. It is argued that the noun itself plays a great role in the choice of number verb and number pronoun (Biber et al, 1999:188; Levin, 2001:40) and that the choice of verb as well can be an influencing factor for agreement (Levin, 1998). The results will be analysed in order to find out whether there are differences between collective nouns, within a collective noun and also which factors that affect the concord. As for limitations, this is an empirical research, but only three online newspapers and a corporate newsletter in English in Sweden have been used because of the fact that the number of newspapers available in English in Sweden is small. The SWENC includes texts from the following online newspapers/newsletters: Stockholm News, The Local, the Swedish Wire and the English section of Svenskt näringsliv. Since this thesis looks at how Swedish writers use the , it should be pointed out that articles by non-native Swedish writers are not included in the SWENC. A detailed description of the abovementioned newspapers and corporate newsletter are given below: Stockholm News is a news website written in English by Swedish journalists and the articles, which have been collected for the corpus, cover the cultural, political and economical sections as well as the Sweden section of the newspaper. The articles were written between

12 2010-2012. Most of the sources that were quoted in this newspaper were in Swedish and the writers themselves translated those quotes. Quotes that were not translated came from international sources. The Swedish Wire consists of articles on economics, politics and business and some of the articles are about art and global news. The articles which have been collected cover the years 2012-2013. Johan Nylander is the only writer in this newspaper and he does all the writing himself. The Local is an online newspaper in Sweden and this newspaper also has editorial offices in six other countries in Europe. The articles from this newspaper cover the years 2012-2013 and the articles by Ann Törnkvist are in the news section whereas the texts by David Lindén are debate articles. Svenskt näringsliv is a Swedish organization for Swedish companies and is not a newspaper per se, but has a corporate newsletter in English. Swedish employees at Svenskt näringsliv write the articles in English. It should also be noted that newspapers often follow style guides, which is also confirmed through personal communication with Ann Törnkvist, the female reporter from The Local. She explains that the newspaper follows The Economist Style Guide, i.e. BrE, with the exception of the usage of z as in organization, which follows the American English variety. In addition, since newspapers have editors, it is not possible to know to which extent the articles were edited and whether the linguistic variables found in the corpus were added by the editor or made by the writer. In terms of validity, Hundt’s (1998) research as cited in Levin (2001:48) has shown that newspapers’ style guide were found not to influence agreement patterns, thus Levin (2001) points out that newspapers which have editors are still valid in a corpus study. In the case of Svenskt näringsliv, which is clearly different from the other authentic newspapers, the articles are always written in Swedish first and then translated into English. Instances of Swedish-influenced linguistic variables are then assumed to be more frequent in this specific newspaper. Since the number of Swedish writers writing in English is limited, it should also be pointed out that it has not been possible to collect more material in English within the news genre in Sweden. However, the material in the SWENC is extensive, consisting of approximately 165,325 words. The material in the press sections of F-LOB and Frown consist of 184,039 and 183,215 words respectively. On the one hand, it is more than enough to carry out an extensive empirical investigation on language variation. On the other hand, since collective nouns constitute a small subcategory of nouns, they are rare phenomena (Levin, 2001). In order to find variable instances on number concord with collective nouns, the size of this material might therefore not be sufficient. Thus, to expand the possibilities of finding a

13 linguistic variable, this investigation will look at a list of collective nouns as provided in Table 1 and also look into the Blogs in English by Swedes Corpus (BESC). This corpus consists of approximately 100,000 words and has been established for this study as well. Even though blogs are more informal than news, they could provide an interesting contrasting study that could indicate how Swedes construct agreement with collective nouns in everyday language. To compare the results in the BESC with the SWENC and the press subcorpora of F-LOB and Frown, this study will also look at both F-LOB and Frown which contain texts from 15 different sections and consist of 1 million words.

4 Results The results of the corpus study are examined below. The tables for all agreement findings in the SWENC, F-LOB Press, Frown Press, the BESC, F-LOB and Frown can be found in Appendix. The findings for the SWENC and the press subcorpora of F-LOB and Frown will be discussed in section 4.1. They will be compared with the results in the BESC, F-LOB and Frown in section 4.2. Section 4.2.1 compares and discusses the findings of the singular political parties and the plural political parties which provide differences in usage for the plural forms. Section 4.3 discusses the findings compared to previous studies and also discusses Swedish usage.

4.1 Findings in the SWENC, F-LOB Press and Frown Press The total results for agreement with collective nouns in the SWENC are illustrated in Table 2. 25 out of 41 collective nouns were found with agreement.

Table 2. Total results of the SWENC singular verbs plural verbs singular pronouns plural pronouns 92% 8% 61% 39%

Table 2 shows that 92% of the tokens took singular verbs and that 61% of the tokens took singular pronouns. Variation in agreement was thus found to occur more frequently in the pronominal category. The problem with analysing and comparing the results in the SWENC with the press subcorpora of F-LOB and Frown is that for some collective nouns there were only a few tokens and for others there were several tokens. The total results for verbal and pronominal concord with collective nouns in the SWENC and the press subcorpora of F-LOB and Frown are illustrated in Figure 1 below.

14 100% 80%

60% Singular verbs 40% Singular pronouns 20% 0% SWENC F-LOB Press Frown Press

Figure 1. Total results in percentage for singular verbs and pronouns.

Figure 1 indicates that singular verbal agreement is the preferred number of agreement within the news genre as observed in the SWENC and the press subcorpora of F-LOB and Frown. The SWENC has 92% singular verbal agreement whereas F-LOB Press has 90% and Frown Press 96%. For pronominal agreement, the results show more variation. The SWENC has more instances of plural pronominal agreement than the press sections in both F-LOB and Frown. The SWENC has 61%, F-LOB Press 78% and Frown Press 89%. Plural pronouns seem to be more common than plural verbs in all three corpora. Figure 1 however indicates the total results and does not say so much about variation in agreement with certain collective nouns. Variation in such nouns will thus be discussed next. The results for singular verbal agreement with the most frequent collective nouns in the SWENC, F-LOB Press and Frown Press are illustrated in Figure 2 below.

100% 90% 80% 70% 60% 50% 40% 30% SWENC 20% 10% 0% F-LOB Press Frown Press

Figure 2. Singular Verbal Agreement in Percentage for the most frequent collective nouns.

The results for singular pronominal agreement with the same nouns are illustrated in Figure 3 below. The most frequent collective nouns in the SWENC are government, committee, board,

15 company, party, office, organisation (organization), commission and names of organisations1. This was expected, as political collective nouns tend to be more frequent in news. Figure 2 indicates that singular verbal agreement is the preferred number of agreement for the most frequent collective nouns as well. Both the SWENC and F-LOB Press however show some variation in verbal number whereas Frown Press show 100% singular verbal concord for all the collective nouns in Figure 2. Office, organisation (organization) and commission were however not found with verbal agreement in the press sections of Frown. Organisation (organization), commission and names of organisations were found with 100% singular verbal concord in the SWENC whereas F-LOB Press shows 71% for names of organisations but 100% for commission. Agreement with organisation was not found in the press sections of F-LOB. Office was found with 87% singular verbal agreement in the SWENC but with 83% in F-LOB Press. Government, committee, board, company and party had some instances of plural verbal agreement in the SWENC but none in the press sections of F-LOB.

100% SWENC F-LOB press 50% Frown press

0%

Figure 3. Singular Pronominal Agreement for the most frequent collective nouns.

The results in Figure 3 indicate that there is more variation in pronominal number than verbal number for the SWENC and F-LOB Press. In contrast, Frown Press only shows singular pronominal agreement for all nouns except for office, organisation (organization) and commission which were not found with pronominal agreement. The SWENC shows more variation in pronominal agreement than the press sections of F-LOB. The results in percentage of singular pronouns for the SWENC are as follows: Government 44%, committee 67%, board 67%, company 88%, party 75%, office 50%, organisation 100%, commission 60% and names of organisations 70%. F-LOB Press shows 100% plural pronominal

1 See Table 1 in Appendix which provides a footnote of which organisations are included in this category.

16 agreement with board whereas organisation, company, party, office and commission were found with 100% singular pronominal agreement. Furthermore, government has 95% singular pronominal agreement and names of organisations 50%. It has to be noted that some collective nouns were found with more instances of verbal agreement than with pronominal agreement. The results nonetheless show that there are more instances of plural pronominal agreement for the most frequent collective nouns in the SWENC than for the press sections in F-LOB and Frown. Examples of variation in verbal and pronominal agreement with government in the SWENC and the press sections of F-LOB and Frown are illustrated below in examples (17)- (20). (17) But the Swedish government do not want any Eurobonds. (Stockholm News in The SWENC) (18) She claims that the government has done one thing during their reign: increasing the gaps in the Swedish society. (Stockholm News in The SWENC) (19) Despite the , the government has abandoned its reliance on private investment to pull the east out of economic difficulties. (F-LOB Press) (20) In fact, the government cites its struggle with the ethnic population as the reason for spending over half of its budget on defense. (Frown Press) As discussed earlier, the SWENC shows more plural concord than the press sections of F- LOB and Frown. (17) illustrates the plural verbal agreement with government whereas government in (18) take both singular and plural concord in the same sentence, i.e. a singular verb and a plural pronoun. Since the Alliance, consisting of four parties, is currently in the Swedish government, it can be assumed that (17) and (18) illustrate the four parties in government. The use of plural concord is thus influenced by stylistic and semantic factors. (19) and (20) however illustrate syntactic agreement in both F-LOB Press and Frown Press. The results for names of organisations in the SWENC and the press subcorpora of F- LOB and Frown have been briefly discussed. However, the SWENC also had numerous findings of SAS, SAAB and IKEA which have been divided into their own categories. The results are illustrated in Figure 4 below. Looking at Figure 4, it can be observed that singular verbal agreement is the preferred number concord for names of organisations as well as for SAS, SAAB and IKEA in the SWENC. For pronominal agreement there are more instances of plural forms but singular pronouns co-occur more frequently with such nouns. Examples of variation in pronominal

17 agreement for the names of organisations found in the SWENC are illustrated in (21) and (22).

100%

50% Singular verbs

Singular pronouns 0% Names SAS SAAB IKEA of org.

Figure 4. Results in the SWENC for names of organisations.

(21) The union sent a debate to American newspaper /…/ In this article they appeal to the American government /…/ They argue that it /…/ (Stockholm News) (22) Vattenfall estimates its costs to SEK 109 million only for Dagmar. (Stockholm News) The plural pronominal number is illustrated in (21) and (22) illustrates the singular verb and pronoun. (21) seems to indicate that a number of individuals in the union appeal to the government about something. (22) seems to indicate that Vattenfall denotes a unit, not a number of individuals. They can be compared with the findings of agreement with IKEA in the SWENC which are illustrated in (23)-(26) below. (23) “IKEA don’t do bribery” (Stockholm News) (24) “IKEA neither give or receive bribes,” says Helen Duphorn. (Stockholm News) (25) Despite being a company from the Netherlands, IKEA often accentuate the Swedish heritage, among else by selling Swedish food in their furniture stores worldwide. (Stockholm News) (26) Swedish furniture giant Ikea is getting both bigger and more profitable, according to its annual report. (Swedish wire) (23) and (24) illustrate the use of plural verbs with IKEA within quotes, which can indicate that plural forms are likely to occur in spoken language. However, it has not been possible to find out whether the person in question talked in Swedish and her quote was translated into English or if the quote is original. This nonetheless shows a preference for plural agreement in spoken language. In contrast, IKEA in (25) takes a plural verb even though it was not a quote. It seems as if the plural verb has been used to agree with the plural possessive pronoun

18 in “their furniture stores”. (26) shows singular verbal and pronominal concord. It seems as if IKEA in (26) acts as a collective unit since this example illustrates the economical situation of the furniture corporation, not the economical situation of the individuals working in the corporation. Examples (23)-(26) can indicate that variation in agreement usage by Swedes seems to be influenced by a number of factors: the preferences of spoken language, agreement with pronominal number and whether IKEA denotes a unit or a number of individuals, i.e. the notional meaning of the noun. As discussed above, the press sections of Frown only show singular concord for names of organisation, whereas F-LOB Press shows some instances of plural agreement. (27) and (28) illustrate variation in F-LOB Press. (27) BT - which makes profits of pounds105 a second - plan to axe 6,500 operator jobs by the end of 1992. (F-LOB Press) (28) Hellenic, which is owned by Leventis-Ioannou Group, is offering 25 per cent of its shares in the float. (F-LOB Press) Both (27) and (28) have a relative pronoun. (27) has a singular verb following the relative pronoun but a plural verb to agree with BT. Levin (2001) discusses the distance of the antecedent with its controller. It seems as if the plural verb has been used since there is a distance to the collective noun BT. (28) however illustrates the singular verb. There does not seem to be any difference in meaning in both examples and thus it can be assumed that variation has been influenced by stylistic factors.

4.2 Comparison of the news genre with the BESC, F-LOB and Frown Section 4.1 has discussed and illustrated the results for the SWENC and the press subcorpora of F-LOB and Frown. The findings will now be compared with the results for the BESC, F- LOB and Frown. Table 3 illustrates the total results of the SWENC and the BESC.

Table 3. Total results of the SWENC and the BESC. Verbs following Verbs relative pronouns Personal pronouns Singular Plural singular plural Singular Plural The SWENC 92% 8% 90% 10% 61% 39% The BESC 100% 0% 80% 20% 87% 13%

The number of tokens found for the BESC was much smaller than for the SWENC and some of the instances of agreement have so few tokens that they cannot be properly analysed. Agreement was found with 12 out of the 41 collective nouns that this thesis has looked into.

19 Even though blogs are assumed to be more informal than newspapers, the BESC shows 100% singular verbal agreement whereas the SWENC shows 92%. The BESC however shows some variation in pronominal number, but this variation is greater in the SWENC. This could however be due to the fact that the number of tokens was small and that blogs do not include as many collective nouns as newspapers. It can also be assumed that political collective nouns do not occur frequently in everyday language as it does in the news genre. There were however several findings of agreement with government but the noun seems to have occurred in political blogs. Team was found with more tokens in the BESC than the SWENC. This indicates that the frequency of tokens depends on the written genre. In sum, the SWENC shows more variation than the BESC. The results of F-LOB and Frown have also been analysed to see if they provide any contrasting differences in results as compared to the press sections of F-LOB and Frown as well as the SWENC and the BESC. The total results are illustrated in Figure 5.

100%

80%

60% Singular verbs 40% singular pronouns 20%

0% SWENC F-LOB Frown F-LOB Frown BESC Press Press

Figure 5. Total results of agreement in the six corpora.

Figure 5 indicates that F-LOB has more instances of plural verbs than the other corpora and that F-LOB Press (10%) has more instances of plural verbs than the SWENC (8%) and Frown Press (4%). In contrast, the SWENC has more instances of plural pronouns (39%) than all the other corpora but is very close to the results in F-LOB which has 33% plural pronouns. Nevertheless, the results show that the SWENC differs from both BrE and AmE within the news genre. Furthermore, Figure 5 seems to indicate that Swedish journalists are more flexible in constructing pronominal agreement than British and American journalists. In addition, the results show that F-LOB has more instances of plural pronouns than Frown as well as the press subcorpora. However, the SWENC has more instances of plural pronouns than the F-

20 LOB corpora which cover 15 different text genres in British English. It can thus be assumed that Swedish journalists are even more flexible in constructing pronominal agreement than British people do in general. The total percentages for singular verbs and pronouns thus show that the SWENC is closer to F-LOB Press in verbal agreement (2% difference) but differs from both F-LOB Press and Frown Press in pronominal agreement. Variation as illustrated in Figure 5 seems to have been caused by regional and stylistic differences. In sum, Swedish journalists writing in English seem not to follow BrE or AmE at all. The results also indicate that in AmE there is only 1% in difference of verbal number between the news genre and the other genres of writing. In contrast, Frown shows more instances of plural pronouns than the press subcorpora, which indicates that there is a difference in pronominal number based on written genre in AmE. There is also a difference in BrE between the news genre and the other written genres: F-LOB shows more instances of both plural verbs and pronouns than F-LOB Press. The comparison of F-LOB and Frown with the press subcorpora thus indicates that there are regional differences as well as stylistic differences in written genres. This consequently suggests a future corpus study of different genres of written English by Swedes.

4.2.1 Comparison of the political parties with the plural forms of political parties There is clearly a difference here as regards to the singular collective nouns of political parties and the plural forms of political parties when comparing the results of the SWENC, F-LOB and Frown. The results for the plural political parties are illustrated in Figure 6. Figure 6 shows that in the SWENC, the plural forms of political parties took 30% singular verbs and 77% singular pronouns. The results in the SWENC shows great differences to both Frown and F-LOB, which show a clear preference for plural concord with plural forms of political parties as well as the sports team Jays in Frown, which was the only token of sports teams found in all the corpora.

100% 80% 60% Plural verbs 40% Plural pronouns 20% 0% SWENC F-LOB Frown

Figure 6. Results for the plural forms of political parties.

21

The results can indicate that Swedish journalists are innovative in the usage of political forms of political parties, i.e. they seem to treat the plural forms of political parties as collectives and thus there is variation in number agreement. However, since this thesis is about agreement with collective nouns and not with plural forms of nouns, this is a new area that has not been studied and should thus make an interesting future study on its own. Figure 6 can be compared with the results for the singular political parties in Figure 7. In Frown, political parties were not used as singular collective nouns but as plural nouns, so the results for political parties in Frown are only provided as plural forms in Figure 6. Figure 7 indicates that singular forms of political parties in the SWENC had 89% singular verbal agreement and 59% plural pronominal agreement. Political parties in F-LOB Press also show 89% singular verbal concord but take 100% singular pronominal concord. Furthermore, political parties in F-LOB are also regarded as singular units with 85% singular verbal concord and 100% singular pronominal agreement.

100%

50% Singular verbs 0% Singular pronouns

Figure 7. Results for the singular forms of political parties as collective nouns.

Swedish journalists seem to prefer singular verbs and plural pronouns with singular political parties even though plural verbs occur frequently. In contrast, they prefer singular pronouns and plural verbs with the plural forms of political parties. (29) and (30) illustrate variation in number agreement with the singular and plural forms of political parties in the SWENC. (29) illustrates the singular verb with the plural form of the party the Greens and (30) illustrates the plural verbal and pronominal concord with the singular party the Liberal Peoples Party. (29) The Greens has two spokespersons instead of a party leader. (Stockholm News) (30) The Liberal Peoples Party today present a new defence policy document, where they go against the stated policy /…/ (Stockholm News)

22 4.3 Discussion of the findings with previous findings Levin (2001) mentions that stylistic, semantic and syntactic factors influence agreement. Syntactic agreement seems to be the preferred form of verbal agreement in the SWENC as well as in the press subcorpora of F-LOB and Frown. Instances of semantic and stylistic agreement have however been found. Examples (27) and (28) illustrate stylistic differences even though they do not show any difference in meaning. Variation as observed in (23)-(26) can also have been influenced by stylistic and semantic factors. The findings of agreement in the SWENC thus suggest that Swedish journalists construct their own stylistic sentences with semantic meaning for both verbal and pronominal agreement. Semantic agreement, however, seems to occur more frequently with pronominal references than for verbs. Another working hypothesis is that Swedish journalists prefer to use singular verbs with collective nouns, since there is no distinction of the verb form in the Swedish language. Teleman et al (1999) however mention that plural pronominal forms are possible with collective nouns in Swedish, e.g. Centerns partistyrelse själva ville inte ta beslut om arbetet2. It can be argued that the reflexive själva denotes the plural of “the board themselves”, not “the board itself”. However, Teleman et al (1999) also states that in this example, semantic agreement has been used and that grammatical singular agreement is the preferred number concord in Swedish. Nevertheless, the results of pronominal number in the SWENC can indicate that Swedish journalists writing in English seem to be influenced by Swedish usage. Språkriktighetsboken (2005) illustrates the problem of agreement with collective nouns in Swedish, as in the example below: (34) Besättningen var missnöjda med lönen. De hade jobbat hårt den senaste tiden.3 (italics added) Even if plural verbs are not distinguishable from singular verbs in Swedish, it seems as if Swedes think that plural pronouns should co-occur with collective nouns and this is thus mirrored in the results of the SWENC. Språkriktighetsboken (2005) also points out that the plural form of the predicative missnöjda as in example (34) has been preferred so that it can agree with the plural pronoun. It is however noted in both Språkriktighetsboken (2005) and Teleman et al (1999) that semantic agreement has been criticized. The results in F-LOB and Frown as well as F-LOB Press and Frown Press seem to follow earlier academic research. Earlier academic research has shown that plural forms of

2 This can be translated as: The Board of the Center Party did not want to make the decision about the work on (their) own. 3 This can be translated into The crew was/were not happy with (their) wage. They had worked hard lately.

23 agreement are more common in written BrE than in AmE (Quirk et al, 1985; Levin, 2001; Biber et al, 1999). Nevertheless, the results indicate that singular verbal agreement is the preferred number agreement in written English in both AmE and BrE, which was observed in Levin (2001) as well. Instances of plural pronouns have been found in Frown and Frown Press but occurred more frequently in F-LOB and F-LOB Press. This follows what Aremo (2003) and Levin (2001) have observed. The results in the SWENC as well as Frown and F-LOB also show that the names of organisations most frequently take singular concord, which follows what Levin (1999) has observed. Moreover, Levin (1999) notes that singular verbs are more common than plural verbs since they often occur immediately after the collective noun. This is observed in all the corpora as well. Leech et al (2006) argue that notional concord is more frequent in speech than in writing and the results of all the corpora support this claim, showing a preference for grammatical concord in the written materials examined here. Biber et al (1999) argue that board and committee occur primarily with singular, which is supported by the results in all the corpora. According to Greenbaum (1996), plural verb forms are used with collective nouns when they constitute a number of individuals whereas singular verb forms are used when they constitute a single unit, i.e. a collective. The findings in the SWENC as well as the press subcorpora of F-LOB and Frown however do not indicate much difference in meaning between plural and singular verbs since singular verbal concord is the preferred number agreement. There are still a few examples of plural verbal agreement which show that the collective noun represents a number of individuals, as in (23) and (24). Summing up, the results indicate that the SWENC is closer to F-LOB Press in verbal agreement patterns but differ from both F-LOB Press and Frown Press in pronominal agreement patterns. Swedish journalists seem to prefer singular verbal agreement with collective nouns in English, but they are more flexible when it comes to pronominal agreement. Since collective nouns can co-occur with plural forms of predicative and pronominal references in Swedish (Teleman et al, 2009), it can be assumed that Swedish journalists construct agreement with collective nouns based on Swedish usage. Vannestål (2007:112f) also claims that plural personal pronouns are more common than singular personal pronouns. The results however show that singular pronouns are more common than plural pronouns. However, like Levin (2001) claims, plural forms seem to be more frequent in

24 spoken English. A research about how Swedes construct number agreement in spoken English would thus provide an interesting comparative study.

5 Conclusion The aim of this study was to investigate how Swedes construct number concord with collective nouns in English within the news genre. This was done by searching for concordances of collective nouns in three primary corpora and three secondary corpora: the SWENC and the press subcorpora of F-LOB and Frown as well as the BESC, Frown and F- LOB. The results in the corpora have been analysed and compared to see whether Swedish journalists writing in English follow BrE or AmE, whether they prefer singular or plural concord and if they are consistent in choosing number agreement. Conclusions that can be drawn from the investigation are as follows: Singular concord seems to be more frequent with collective nouns in Sweden. Nevertheless, there are several instances of plural pronominal agreement with collective nouns in the SWENC. It seems as if Swedish journalists prefer to use plural pronouns rather than plural verbs with collective nouns, i.e. they seem to be consistent in choosing verbal number but vary pronominal number based on stylistic and semantic factors. This study thus indicates that Swedish journalists writing in English seem to follow verbal agreement patterns in BrE but construct pronominal number differently. Since plural pronominal agreement occurs with collective nouns in Swedish, Swedish journalists writing in English might be influenced by Swedish usage. Nevertheless, significantly more research is needed to understand how Swedes construct number concord with collective nouns. Since this investigation was done by analysing the SWENC, which is a relatively small corpus of online news in English by Swedes, it is not possible to know whether the results can be generalized to other genres of English in Sweden as well. When it comes to future studies, further research with a significantly larger corpus is needed in order to find specific agreement patterns between different genres. This thesis consequently suggests future studies in spoken and other written genres in English in Sweden. In addition, Swedish journalists seem to be innovative when it comes to plural forms of political parties that seem to constitute collective nouns and which differ from both BrE and AmE usage. This is an area that also could make an interesting future study, i.e. a study of plural forms of nouns that act as collective nouns.

25 References

Primary sources Blogs in English by Swedes Corpus (BESC). Compiled by Oscar Svensson, Linnaeus University. The Freiburg-Brown Corpus (Frown). Compiled by Christian Mair, Albert-Ludwigs- Universität Freiburg. The Freiburg-LOB Corpus (F-LOB). Compiled by Christian Mair, Albert-Ludwigs- Universität Freiburg. Swedish in English Newspapers Corpus (SWENC). Compiled by Sanna Eriksson, Felix Fernebring, Elias Gustafsson, Jenny Gustafsson, Petra Gartsjö and Petra Örlegård, Linnaeus University.

Secondary sources Aitchinson, Jean. 2001. Language change: process or decay? 3rd edition. Cambridge: Cambridge University Press. Aremo, W.B. 2003. On some uses of the singular collective nouns congress, council, government, management, parliament and senate in contemporary English. In ‘Lekan Oyeleye & Moji Olateju (eds.). Readings in language and literature. 1st edition. Nigeria: Obafemi Awolowo University Press Ltd. 67-74 Bauer, Laurie. 1994. Watching English Change. Harlow: Longman. Biber, Douglas, Stig Johansson, Geoffrey Leech, Susan Conrad & Edward Finegan. 1999. Longman grammar of spoken and written English. Harlow: Longman. Collins Cobuild English Grammar. 1990. London: Collins. Corbett, Greville. 2000. Number. Cambridge: Cambridge University Press. Corbett, Greville. 2006. Agreement. Cambridge: Cambridge University Press. Dekeyser, Xavier. 1975. Number and case relations in 19th century British English: a comparative study of grammar and usage. Antwerp: Bibliotheca linguistica. Greenbaum, Sidney. 1996. The Oxford English Grammar. Oxford: Oxford University Press. Hundt, Marianne. 1998. New Zealand English grammar – fact or fiction? Amsterdam: John Benjamins. Hundt, Marianne & Christian Mair. 1999. “Agile” and “Uptight” Genres: The Corpus-based Approach to Language Change in Progress. International Journal of Corpus Linguistics 4(2): 221-242

26 Jespersen, Otto. 1909-1949. A modern English grammar on historical principles I-VII. Copenhagen: Ejnar Munksgaard. Leech, Geoffrey & Jan Svartvik. 2002. A communicative grammar of English. 3rd edition. Harlow: Pearson. Leech, Geoffrey, Margaret Deuchar & Robert Hoogenraad. 2006. English Grammar for today: A New Introduction. 2nd edition. New York: Palgrave MacMillan. Levin, Magnus. 1998. Manchester United are my team: On concord with collective nouns. Moderna Språk 92:14-18 Levin, Magnus. 1999. Abba were a sausage on the cream cake of pop: Agreement in British, American and Australian English. Paper presented at MAVEN 2 in Lincoln, September 1999. Levin, Magnus. 2001. Agreement with collective nouns. Lund: Lund University. Persson, Gunnar. 1989. On the of collective nouns in English. In Odenstedt, Bengt & Gunnar Persson (eds.) Instead of flowers. Papers in honour of Mats Rydén on the occasion of his sixtieth birthday, August 27, 1989. Stockholm: Almqvist & Wiksell International. 179-195. Quirk, Randolph, Sidney Greenbaum, Geoffrey Leech & Jan Svartvik. 1972. A grammar of contemporary English. London: Longman. Quirk, Randolph, Sidney Greenbaum, Geoffrey Leech & Jan Svartvik. 1985. A comprehensive grammar of the English language. London: Longman. Reid, Wallis. 1991. Verb & noun number in English: A functional explanation. New York: Longman. Språkriktighetsboken. 2005. Svenska språknämnden och Norstedts. Jan & Sager, Olof. 1996. Engelsk universitetsgrammatik. 2nd edition. Stockholm: Almqvist & Wiksell. Swan, Michael. 2005. Practical English Usage. 3rd edition. Oxford: Oxford University Press. Teleman, Ulf, Staffan Hellberg & Erik Andersson. 1999. Svenska Akademiens Grammatik 3: Fraser. Stockholm: Svenska Akademien. Vannestål, Maria Estling. 2007. A University Grammar of English: with a Swedish perspective. Lund: Studentlitteratur. Visser, Frederik Theodor. 1963. An historical of the English language. Part I. Leiden.

27 Appendix

Table 1. The SWENC. Total Verb Relative pronoun Personal pronouns tokens Singular Plural Singular Plural Singular Plural N % N % N % N % N % N % government 60 91 6 9 1 100 0 0 8 44 10 56 75 family 4 100 0 0 0 0 0 0 0 0 1 100 4 team 0 0 1 100 1 100 0 0 0 0 0 0 2 committee 9 90 1 10 0 0 0 0 2 67 1 33 11 board 7 88 1 12 1 50 1 50 2 67 1 33 12 gang 0 0 1 100 0 0 0 0 0 0 1 100 1 company 25 96 1 4 3 100 0 0 15 88 2 12 39 couple 1 50 1 50 0 0 0 0 0 0 1 100 3 community 1 100 0 0 0 0 0 0 0 0 0 0 1 party 32 97 1 3 5 100 0 0 9 75 3 25 40 congress 2 100 0 0 0 0 0 0 1 50 1 50 4 group 1 100 0 0 0 0 0 0 0 0 1 100 1 association 1 50 1 50 0 0 0 0 0 0 0 0 2 jury 0 0 0 0 1 100 0 0 0 0 0 0 1 staff 1 100 0 0 0 0 0 0 0 0 0 0 1 council 1 100 0 0 0 0 0 0 1 50 1 50 2 majority 2 67 1 33 0 0 0 0 0 0 1 100 3 military 2 100 0 0 0 0 0 0 0 0 0 0 2 the public 1 100 0 0 0 0 0 0 0 0 0 0 1 office 13 87 2 13 1 100 0 0 1 50 1 50 16 parliament 1 100 0 0 0 0 0 0 0 0 0 0 1 organization/ organisation 5 100 0 0 1 50 1 50 1 100 0 0 7 the state 16 100 0 0 1 100 0 0 2 100 0 0 16 management 3 100 0 0 0 0 0 0 0 0 0 0 3 commission 15 100 0 0 0 0 0 0 3 60 2 40 17 names of org.4 21 100 0 0 0 0 0 0 7 70 3 30 24 SAS 5 100 0 0 0 0 0 0 3 100 0 0 7 SAAB 29 97 1 3 0 0 0 0 4 67 2 33 31 IKEA 8 73 3 27 0 0 0 0 2 67 1 33 10 political parties5 34 89 4 11 3 100 0 0 7 41 10 59 50

4 Includes the names of organisations and corporations: H&M, Tiger of Sweden, Björn Borg, Nordea, Vattenfall, The Swedish Federation of Trade, The International Committee of the Red Cross, The Union, European Investment Bank (EIB), Volvo Cars. 5 I have chosen to separate the political parties from “party” as I wanted to collect all instances of political parties into one group so that SDU (Sweden Democrats Youth) could be compared with the other political parties. Includes SDU (Sweden Democrats Youth), the Green Party, the Liberal Party, the Social democratic party, the Centre party. They are thus also comparable to the political parties Labour and Tory in the FLOB corpus.

28 the opposition 6 100 0 0 0 0 0 0 0 0 0 0 6 the Alliance 4 80 1 20 0 0 0 0 2 67 1 33 7 the Riksdag 1 100 0 0 0 0 0 0 0 0 0 0 1 The UN 1 100 0 0 0 0 0 0 1 100 0 0 2 TOTAL 312 92 26 8 18 90 2 10 70 61 44 39 402

Table 2. The BESC. Total Verb Relative pronoun Personal pronouns tokens Singular Plural Singular Plural Singular Plural N % N % N % N % N % N % government 7 100 0 0 1 100 0 0 5 100 0 0 9 family 1 100 0 0 0 0 0 0 0 0 0 0 1 team 6 100 0 0 1 50 1 50 0 0 1 100 8 company 2 100 0 0 1 100 0 0 0 0 0 0 3 community 1 100 0 0 0 0 0 0 1 100 0 0 1 party 3 100 0 0 1 100 0 0 2 100 0 0 5 group 1 100 0 0 0 0 0 0 0 0 0 0 1 army 1 100 0 0 0 0 0 0 0 0 0 0 1 majority 1 100 0 0 0 0 0 0 0 0 1 100 2 minority 0 0 0 0 0 0 0 0 1 100 0 0 1 organisation 1 100 0 0 0 0 0 0 1 100 0 0 1 Apple 5 100 0 0 0 0 0 0 3 100 0 0 5 TOTAL 29 100 0 0 4 80 1 20 13 87 2 13 38

Table 3. F-LOB Press. Total Verb Relative pronoun Personal pronouns tokens Singular Plural Singular Plural Singular Plural N % N % N % N % N % N % government 32 100 0 0 3 100 0 0 21 95 1 5 45 family 2 67 1 33 0 0 0 0 0 0 1 100 4 team 5 71 2 29 0 0 0 0 1 50 1 50 8 committee 2 100 0 0 0 0 0 0 2 100 0 0 3 board 2 100 0 0 0 0 0 0 0 0 1 100 3 company 8 100 0 0 3 100 0 0 4 100 0 0 14 club 2 100 0 0 0 0 0 0 0 0 1 100 3 couple 0 0 4 100 0 0 0 0 0 0 3 100 5 society 3 100 0 0 1 100 0 0 1 100 0 0 5 community 2 100 0 0 0 0 0 0 0 0 1 100 3 audience 4 100 0 0 1 100 0 0 1 50 1 50 5 crew 0 0 1 100 0 0 0 0 0 0 0 0 1 party 11 100 0 0 1 100 0 0 7 100 0 0 15 congress 2 100 0 0 1 100 0 0 1 100 0 0 3

29 group 3 100 0 0 0 0 0 0 1 50 1 50 4 association 4 80 1 20 1 100 0 0 1 100 0 0 6 the press 1 100 0 0 0 0 0 0 0 0 0 0 1 army 2 100 0 0 0 0 0 0 1 100 0 0 3 crowd 0 0 0 0 0 0 0 0 0 0 1 100 1 jury 3 100 0 0 0 0 0 0 0 0 0 0 3 staff 0 0 2 100 0 0 2 100 0 0 2 100 6 council 25 100 0 0 1 100 0 0 11 100 0 0 29 department 6 100 0 0 0 0 0 0 4 100 0 0 8 the public 2 67 1 33 0 0 0 0 0 0 3 100 5 office 5 83 1 17 0 0 0 0 1 100 0 0 7 parliament 3 100 0 0 0 0 0 0 0 0 0 0 3 corporation 0 0 0 0 1 100 0 0 0 0 0 0 1 commission 5 100 0 0 0 0 0 0 1 100 0 0 5 names of org. 5 71 2 29 2 100 0 0 2 50 2 50 10 political parties 16 89 2 11 0 0 0 0 8 100 0 0 18 TOTAL 155 90 17 10 15 88 2 12 68 78 19 22 227

Table 4. Frown Press. Total Verb Relative pronoun Personal pronoun tokens Singular Plural Singular Plural Singular Plural N % N % N % N % N % N % government 13 100 0 0 2 100 0 0 5 100 0 0 17 family 5 83 1 17 0 0 0 0 1 50 1 50 6 team 6 100 0 0 1 100 0 0 2 67 1 33 7 committee 2 100 0 0 0 0 0 0 1 100 0 0 3 board 14 100 0 0 1 100 0 0 5 100 0 0 16 company 3 100 0 0 3 100 0 0 6 100 0 0 10 club 3 100 0 0 0 0 0 0 0 0 0 0 3 society 2 100 0 0 0 0 0 0 0 0 0 0 2 community 3 100 0 0 0 0 0 0 1 100 0 0 3 audience 1 100 0 0 1 100 0 0 0 0 0 0 2 crew 1 100 0 0 0 0 0 0 0 0 1 100 1 party 2 100 0 0 2 100 0 0 3 100 0 0 7 congress 9 100 0 0 0 0 0 0 4 80 1 20 11 group 3 100 0 0 0 0 0 0 3 75 1 25 5 association 1 100 0 0 0 0 0 0 0 0 0 0 1 class 0 0 0 0 1 100 0 0 0 0 0 0 1 senate 4 100 0 0 0 0 0 0 0 0 0 0 4 the press 1 100 0 0 0 0 0 0 0 0 0 0 1 army 1 100 0 0 0 0 0 0 0 0 0 0 1 crowd 5 100 0 0 0 0 0 0 0 0 0 0 5 jury 1 100 0 0 0 0 0 0 0 0 0 0 1

30 staff 1 100 0 0 0 0 0 0 1 100 0 0 2 council 3 100 0 0 0 0 0 0 0 0 0 0 3 department 6 100 0 0 0 0 0 0 2 100 0 0 7 assembly 2 100 0 0 0 0 0 0 2 100 0 0 3 majority 0 0 3 100 0 0 0 0 0 0 0 0 3 military 1 100 0 0 0 0 0 0 0 0 0 0 1 the public 5 100 0 0 0 0 0 0 0 0 0 0 5 parliament 1 100 0 0 0 0 0 0 0 0 0 0 1 organization 0 0 0 0 1 100 0 0 0 0 0 0 1 the state 3 100 0 0 0 0 0 0 1 25 0 75 3 names of org.* 7 100 0 0 0 0 0 0 5 100 0 0 11 TOTAL 109 96 4 4 12 100 0 0 42 89 5 11 147

Table 5. F-LOB. Total Verb Relative pronoun Personal pronouns tokens Singular Plural Singular Plural Singular Plural N % N % N % N % N % N % governme nt 67 83 14 17 5 83 1 17 45 80 11 20 109 family 15 65 8 35 1 100 0 0 5 36 9 64 32 team 9 69 4 31 1 100 0 0 1 25 3 75 16 committee 15 100 0 0 2 100 0 0 9 82 2 18 22 board 5 83 1 17 0 0 0 0 0 0 2 100 7 company 29 85 5 15 4 80 1 20 22 88 3 12 54 club 9 100 0 0 1 100 0 0 0 0 1 100 10 couple 10 1 17 5 83 0 0 2 0 0 0 9 100 14 society 15 100 0 0 2 100 0 0 11 92 1 8 27 community 14 100 0 0 3 100 0 0 4 44 5 56 19 audience 11 92 1 8 1 100 0 0 0 0 4 100 14 crew 1 33 2 67 0 0 0 0 0 0 0 0 3 party 27 96 1 4 3 100 0 0 15 88 2 12 39 congress 3 100 0 0 1 100 0 0 1 100 0 0 4 group 27 93 2 7 3 100 0 0 9 56 7 44 35 association 8 89 1 11 1 50 1 50 2 67 1 33 12 class 2 100 0 0 1 100 0 0 3 100 0 0 6 senate 1 100 0 0 0 0 0 0 0 0 0 0 1 the press 3 100 0 0 0 0 0 0 0 0 0 0 3 army 6 100 0 0 0 0 0 0 2 67 1 33 9 10 crowd 1 100 0 0 0 0 1 0 2 67 1 33 5 jury 3 100 0 0 0 0 0 0 0 0 0 0 3 10 Staff 1 6 15 94 0 0 2 0 0 0 8 100 21 council 34 97 1 3 1 100 0 0 12 100 0 0 39

31 department 11 55 9 45 0 0 0 0 5 22 18 78 32 assembly 0 0 0 0 0 0 0 0 1 100 0 0 1 majority 0 0 2 100 0 0 0 0 0 0 0 0 2 military 1 100 0 0 0 0 0 0 0 0 0 0 1 navy 2 100 0 0 0 0 0 0 0 0 0 0 2 the public 5 71 2 29 0 0 0 0 1 20 4 80 9 office 15 88 2 12 0 0 0 0 1 25 3 75 19 parliament 6 100 0 0 0 0 0 0 1 100 0 0 7 population 9 90 1 10 0 0 0 0 0 0 0 0 9 organisation/z 6 100 0 0 1 100 0 0 2 100 0 0 9 corporation 4 100 0 0 1 100 0 0 4 100 0 0 6 the state 3 100 0 0 0 0 0 0 2 100 0 0 4 management 2 50 2 50 0 0 0 0 0 0 0 0 4 commission 6 100 0 0 0 0 0 0 7 100 0 0 12 political parties6 17 85 3 15 0 0 0 0 8 100 0 0 21 names of org. 37 84 7 16 3 100 0 0 30 79 8 21 58 TOTAL 431 83 88 17 35 81 8 19 205 67 103 33 700

Table 6. Frown. Total Verb Relative pronoun Personal pronouns tokens Singular Plural Singular Plural Singular Plural N % N % N % N % N % N % government 29 100 0 0 2 100 0 0 9 100 0 0 34 family 19 95 1 5 0 0 0 0 8 73 3 27 27 team 8 100 0 0 2 100 0 0 2 40 3 60 12 committee 4 100 0 100 1 100 0 0 5 83 1 17 9 board 14 100 0 0 1 100 0 0 6 100 0 0 17 company 26 100 0 0 5 100 0 0 21 84 4 16 48 club 5 100 0 0 0 0 0 0 2 100 0 0 6 couple 1 33 2 67 0 0 0 0 0 0 5 100 6 society 11 100 0 0 5 100 0 0 4 100 0 0 17 community 12 100 0 0 1 100 0 0 4 100 0 0 14 audience 7 100 0 0 2 100 0 0 2 50 2 50 9 crew 3 75 1 25 2 100 0 0 0 0 3 100 6 party 9 100 0 0 2 100 0 0 5 100 0 0 14 congress 18 100 0 0 0 0 0 0 7 88 1 12 23 group 17 100 0 0 3 60 2 40 5 56 4 44 24 association 5 100 0 0 0 0 0 0 2 100 0 0 6 class 3 100 0 0 2 100 0 0 0 0 0 0 5 senate 4 100 0 0 0 0 0 0 0 0 0 0 4 the press 6 100 0 0 0 0 0 0 2 100 0 0 8

6 Labour and Tory are included in this category

32 army 2 100 0 0 1 100 0 0 5 100 0 0 7 crowd 7 100 0 0 0 0 0 0 0 0 0 0 7 jury 1 100 0 0 1 100 0 0 0 0 0 0 2 staff 7 100 0 0 0 0 0 0 2 67 1 33 9 faculty 0 0 1 100 1 100 0 0 0 0 0 0 2 council 4 100 0 0 0 0 0 0 1 100 0 0 4 department 8 100 0 0 1 100 0 0 4 100 0 0 11 assembly 6 100 0 0 0 0 0 0 4 100 0 0 8 majority 2 33 4 67 0 0 1 100 1 100 0 0 7 minority 4 100 0 0 0 0 0 0 1 100 0 0 5 military 4 100 0 0 0 0 0 0 1 100 0 0 4 navy 1 100 0 0 0 0 0 0 0 0 0 0 1 the public 10 100 0 0 0 0 0 0 1 100 0 0 10 office 5 100 0 0 1 100 0 0 1 100 0 0 7 parliament 1 100 0 0 0 0 0 0 0 0 0 0 1 population 5 100 0 0 1 100 0 0 2 100 0 0 6 organization 5 100 0 0 1 100 0 0 3 100 0 0 8 corporation 4 100 0 0 0 0 0 0 0 0 0 0 4 the state 8 100 0 0 0 0 0 0 3 100 0 0 9 management 2 100 0 0 0 0 0 0 0 0 0 0 2 commission 3 100 0 0 0 0 0 0 0 0 0 0 3 Names of org. 29 100 0 0 2 100 0 0 18 82 4 18 42 TOTAL 319 97 9 3 37 93 3 7 131 81 31 19 448

Table 7. Plural forms of political parties in F-LOB.

Liberal democrats, Tories, the socialists Verb Relative pronoun Personal pronouns Total tokens Singular Plural Singular Plural Singular Plural N % N % N % N % N % N % 0 0 17 100 0 0 0 0 0 0 6 100 19 Table 8. Plural forms of political parties and sports teams in Frown.

Jays, Republicans, democrats Verb Relative pronoun Personal pronouns Total tokens Singular Plural Singular Plural Singular Plural N % N % N % N % N % N % 0 0 21 100 0 0 1 100 0 0 6 100 25

33 Table 9. Plural forms of political parties in he SWENC.

Social democrats, the liberals, the greens, the Sweden democrats, the moderates, the Christian democrats Total Verb Relative pronoun Personal pronouns tokens Singular Plural Singular Plural Singular Plural N % N % N % N % N % N % 13 30 31 70 1 20 4 80 1 77 12 23 51

34