Abstract
This study investigates the use of so-called pseudotitles, that is, determiner-less structures providing descriptive information in front of name noun phrases, as in linguist Allan Bell, in Bahamian newspaper language. Pseudotitles originated in American journalistic writing, but they have spread to numerous varieties of English worldwide and even to British English. A corpus of pre-independence and contemporary press news reports is analyzed quantitatively with a view to establishing not only the frequency of pseudotitles but also the constraints that govern their usage in Bahamian English. The study also considers the position and structure of equivalent appositives and their relationships with pseudotitles. It will be shown that, at least with regard to the feature investigated here, Bahamian journalists followed American norms even in British colonial times, which may be accounted for by the social history and current sociolinguistic situation of the country. At the same time, these norms have been modified and adapted to local linguistic realities, which presents another piece of evidence in favor of a nuanced view of linguistic Americanization.
Introduction: Pseudotitles and the Americanization of Varieties of English Worldwide
Pseudotitles are structures such as United Bahamian Party backbencher in example (1):
(1) United Bahamian Party backbencher Bernard Dupuch, who demanded in the House of Assembly recently that the question of independence for the Bahamas should be made an election issue, last night told a mass meeting at the party’s headquarters in Farrington Road, that 3,000 people in Abaco have already signed a petition stating firmly that they are against any move by the PLP Government towards independence. (Nassau Guardian and Bahamas Observer, February 8, 1968)
Such structures occur in positions in which we usually find titles such as Doctor, Bishop, or General, that is, right in front of a name noun phrase (NP). Unlike titles, however, pseudotitles do not serve as honorifics but merely provide descriptive information about the referent of the name NP they precede. Pseudotitles used to be considered “a hallmark of American journalistic style” (Bell 1988:326); Quirk et al. (1985:276), in fact, describe them as part of what they label “Timestyle,” that is, the American style of news reporting associated with Time magazine.
While they may have originated in American English, pseudotitles have spread to other varieties of English and even to British newspaper language. However, whereas in American English pseudotitles are not stylistically marked and occur in most papers, regardless of formality, 1 they constitute “one of the most clearly stratifying features in British newspaper language” (Jucker 1992:207), with more formal newspapers, such as the Times and the Guardian, shunning them, and popular papers, such as the Daily Mail and the Daily Express, making quite frequent use of them (Rydén 1975:37). New Zealand presents an interesting case of change in that, until the 1970s or 1980s, language use in the news media was thoroughly British in orientation. As Bell (1988:326–327) has documented, this “traditional colonial foundation […] has been weakening and refocusing towards the United States,” which includes a drastic increase in the use of pseudotitles. Meyer (2002a, 2002b, 2004), finally, has traced pseudotitles in the press sections of seven subcorpora of the International Corpus of English (ICE), finding that “the British-based stigma against pseudo-titles has virtually disappeared world-wide, even in those varieties […] that are British influenced” (Meyer 2002b:165). Interestingly, ICE Philippines and ICE New Zealand even surpass American English not just in terms of the frequency of pseudotitles but also in terms of their length and complexity.
The use of pseudotitles clearly represents an Americanism, that is, a “linguistic feature (of pronunciation, spelling, grammar, meaning, or pragmatic use) that is unique to, or occurs with greater frequency in, or is associated for whatever reason” with language use in the United States of America, “in contrast to the usage of the same language in another nation” (Algeo 1989:156). More specifically, pseudotitles may be described as a “diachronic Americanism,” that is, “an expression that originated in America, whatever its current use may be” (Algeo 1992:287). On account of its location, the Caribbean appears as a world region particularly susceptible to the spread of Americanisms, and in fact, what Christie (2003:20) says about Jamaica in the following could easily be extended to the Bahamas: Jamaica’s geographical proximity to the USA makes it particularly open to the influence of US English. This is fostered by both formal and informal contacts including business and cultural interchanges of various kinds, business, and vacation travel as well as travel for purposes of study, employment and professional contacts, for example. Family ties are also an important source of influence. Almost every Jamaican has at least one relative who resides permanently in the USA with whom more or less regular contact is maintained. The most direct cultural influences stem from popular entertainment, tourism, and goods and services that reflect US industry and lifestyles.
As for studies investigating the American impact on varieties of English worldwide, lexical influences have been at the forefront of attention (see, e.g., Bayard 1989; Meyerhoff 1993; Leek & Bayard 1995; and Vine 1999 on New Zealand English; Modiano 1996 on English in Europe; Taylor 1989 and Peters 2001 on Australian English; Awonusi 1994 and Igboanusi 2003 on Nigerian English; Trüb 2008 on South African English; and Sedlatschek 2009 on Indian English). Leech et al. (2009) explore grammatical changes in contemporary English. The authors systematically mine the Brown family of corpora for a variety of categories such as the subjunctive, modals and semi-modals, the progressive, and the passive; they identify Americanization as one of the major forces driving the changes described.
With regard to the Caribbean, previous research has shown that present-day English in Jamaica, the Bahamas, and Trinidad and Tobago shows a significant trend toward the use of American forms, once again especially in the area of lexis (Mair 2002, 2009; Bruckmaier & Hackert 2011; Hänsel & Deuber 2013). Deuber and Hackert (2013) present an overview of nine Caribbean territories; they investigate not only lexis but also grammar and orthography and find that the degree to which American features are employed varies considerably by level of language as well as by variety and depends on, among other factors, a community’s sociohistorical relations with the United States.
In this article, I investigate pseudotitles in Bahamian press news reports with a view to shedding light on the role that American linguistic norms play in Bahamian journalistic writing today. I will first outline the historical and sociolinguistic background of English in the archipelago before moving on to the presentation of the data employed and the linguistic variable investigated here. I will discuss the results of a number of quantitative analyses, first comparing the use of pseudotitles in various Bahamian newspapers to that in British and American ones, and then presenting the results of a Varbrul analysis that included both linguistic and extralinguistic factors that have been hypothesized to influence the occurrence of this feature. I will also take a look at the position and structure of equivalent appositives and their relationships with pseudotitles. Finally, some conclusions pertaining to the Americanization hypothesis will be drawn.
English in the Bahamas: Historical and Sociolinguistic Background
The Commonwealth of The Bahamas, as the country is officially called, “represents a crossroads for many varieties of English […], so that Bahamian English seems American to the British, British to Americans, Caribbean to those not from the Caribbean proper, etc.” (Holm & Shilling 1982:vii). In other words, English in the Bahamas has always been subject to varying influences. On the one hand, there is the inherited colonial norm, British English. Having been first settled by a group of British religious dissenters from Bermuda in 1648, the Bahamas remained a British colony for more than 300 years, and the British legacy is still clearly visible in the country’s institutional structure. Not surprisingly, then, educated British English still exerts considerable influence in all official administrative or political language use, in the educational system, and in the more conservative media.
On the other hand, the Bahamas received their main population influx in the wake of the American Revolutionary War, when thousands of loyalists and loyalist slaves fled the newly independent United States and resettled in parts of the New World that remained British. The loyalist immigration not only tripled the entire colony’s population and increased the proportion of blacks to three-quarters of the whole (Craton & Saunders 1992:179) but also introduced the predecessor of contemporary Bahamian Creole to the islands—most likely an early form of Gullah, that is, the creole that is still spoken on the Sea Islands off the coast of South Carolina and Georgia (Hackert & Huber 2007).
About 1,000 loyalists came to the Bahamas via New York. These immigrants, mostly “modest white families—former soldiers, townsfolk, and farmers” (Craton & Saunders 1992:183), settled in and around the capital of Nassau on New Providence and on other northern islands like Abaco. Other loyalists were from the American South; the planters among them tended to take their families and slaves directly to the largely unsettled islands in the southeast of the archipelago, where the British government had granted them lands and where they set up a plantation economy based on cotton, which, however, did not last (Craton & Saunders 1992:213–232). Even though there has always been a considerable amount of migration within the Bahamas, particularly from the more rural islands to the larger cities, different islands are still well known for their distinctive speech patterns. According to Holm (1980:51–52), for example, the speech of today’s white Abaconians is evocative of that of eighteenth-century New Englanders in their use of the archaic pronoun ye and of regional words such as lot ‘to plan’ or up along ‘away from the sea.’
American linguistic influence in the contemporary Bahamas emanates from a number of sources. Many Bahamians have friends or relatives in the United States and regularly visit the country, either for private reasons or to do business. Most Bahamian homes have access to U.S. television via cable or satellite. With regard to education, teaching materials are often American, and many Bahamians attend college or university in North America. Finally, those employed in the service industry regularly interact with tourists, most of whom come from the United States. In sum, the influence of American English in the Bahamas is considerable and may even have risen since independence in 1973.
Data and Methodology: The Newspaper Corpus
The corpus used for this study is an extended version of what will form the press section of ICE Bahamas (http://ice-corpora.net/ice/iceba.htm), which is being compiled at the University of Munich in cooperation with the College of the Bahamas. The ICE project, that is, the International Corpus of English, goes back to the late 1980s and Greenbaum’s (1996:3) proposal to “provide the resources for comparative studies of the English used in countries where it is either a majority first language (for example, Canada and Australia) or an official additional language (for example, India and Nigeria).” Caribbean Englishes fit neither of these descriptions, and, in fact, the standard varieties extant in the region moved into linguistic focus only very recently. In part, this comparative neglect may be owed to insecurity among researchers of World Englishes as to how to classify Caribbean varieties (Kachru 1992:3, for example, explicitly excluded them from his circle model); partly, it may have to do with the reluctance described by Youssef (2004:42–43) of Caribbean linguists themselves to view standard English as a significant component of their communities’ linguistic repertoire.
Each ICE corpus follows the same design and provides 500 samples of written and spoken English of 2,000 words each. To be contributors to ICE, language users must be adults and must have received formal education in English at least up to secondary school, since ICE is interested in standard varieties of English (Nelson 1996:28). People who do not meet these criteria can still be contributors if they enjoy public status. This is the case for politicians, writers, and—relevant in the present context—journalists (Greenbaum 1996:6). The ICE project is still under way, ICE Great Britain being the first corpus to have been completed. Corpora for India, Canada, Singapore, New Zealand, East Africa, Hong Kong, the Philippines, Ireland, Jamaica, and Nigeria have followed. Currently, corpora are being compiled for a number of other countries, among them Malta, Ghana, Sri Lanka, Fiji, Trinidad and Tobago, and the Bahamas.
The texts that will constitute the press section of ICE Bahamas were taken from the Nassau Guardian and the Freeport News online archives. These newspapers are the two most widespread and easily accessible Bahamian papers. Both are dailies. Having been published in the Bahamian capital since 1844, the Nassau Guardian is the Bahamas’ oldest newspaper (http://www.thenassauguardian.com/). The Freeport News was established in 1961 to serve the island of Grand Bahama in the north of the archipelago; it is now owned by the Nassau Guardian as well but is managed and edited locally (http://freeport.nassauguardian.net/). The years covered are 2002 to 2007. This material was supplemented by texts collected in 2010 from the Abaconian, a biweekly paper circulated freely in Abaco and its offshore cays (http://www.abaconian.com). In order to add a diachronic, pre-independence perspective, 1968 editions of the Nassau Guardian and Bahamas Observer (NGBO), the forerunner of the contemporary Nassau Guardian, were sampled. These were obtained via microfilm from the Library of Congress.
Only Bahamian news items were selected, that is, news that had appeared in the national or local sections of the papers, so as not to run the risk of encountering pieces distributed by international news agencies (see Mair 1992:79; Sand 1999:20). Also, sports news items were excluded because linguistic usage in the sports sections of any paper often differs quite dramatically from that observed in other sections of the same paper (see Rydén 1975:16–17; Bell 1988:327; Jucker 1992:239; Meyer 2004:345). For the current study, moreover, only “hard news reporting” (Bell 1988:327) was considered, which includes press news reports (ICE category W2C) but not editorials (W2E). This was done to ensure the comparability of the material investigated here with that used by Bell (1988) and Meyer (2002a, 2002b, 2004).
Sample size varies. Whereas the NGBO sample comprises 25,810 words, the samples collected from the Abaconian and the Freeport News add up to 18,827 and 17,842 words, respectively. At 27,068 words, the largest sample comes from the Nassau Guardian. Whenever token numbers are compared in what follows, word counts have been normalized to 20,000. As there is no way to search for pseudotitles or appositives automatically, all tokens were retrieved manually, that is, the texts were read in their entirety, and all pseudotitles and equivalent appositive structures were extracted, entered into a spreadsheet, and coded independently by two coders. 2 Examples will be quoted with newspaper title and publication date.
If Bahamian English were undergoing a process of Americanization, we would expect to find a shift in norm orientation from traditional, British colonial to American between pre-independence and contemporary Bahamian press texts. With regard to pseudotitles, this means that we would expect to find not only more pseudotitles in contemporary Bahamian newspapers than in the pre-independence sample but possibly also longer and more complex ones, as in ICE Philippines or ICE New Zealand (Meyer 2002b:165). These hypotheses will be tested in the following.
The Variable: Pseudotitles versus Titles and Appositives
The term “pseudotitle” was coined by Bell (1988:330), who described structures such as Chief of Internal Medicine and Infectious Diseases at the Princess Margaret Hospital, Dr Gomez, journalist Jessica Robertson, and lawyer Fayne Thompson in example (2) as appositional constructions consisting of a name NP plus a preceding descriptive NP upon which the rule of “determiner deletion” has operated (Bell 1988:328).
(2) Chief of Internal Medicine and Infectious Diseases at the Princess Margaret Hospital, Dr Gomez was a special guest on the Island 102.9 FM Parliament Street talk show, hosted by journalist Jessica Robertson and lawyer Fayne Thompson. (Nassau Guardian April 14, 2003)
As the term and the preceding description suggest, pseudotitles must be distinguished from titles on the one hand and appositives on the other, which is not always easy, as the boundaries between the categories are sometimes unclear (see Meyer 2002b:148).
Structurally, titles and pseudotitles are very similar. They occur in the same syntactic slot, and apart from exceptional cases such as the title the Honorable, both preclude the use of a determiner. Moreover, both are dependent on a head. In other words, neither titles nor pseudotitles can stand alone but obligatorily constitute part of an NP, as in Dr Gomez (title) and journalist Jessica Robertson (pseudotitle). Vocative use represents an exception to this rule. As Quirk et al. (1985:1320) note, most titles (but not pseudotitles) can be used as vocatives, and, in fact, this phenomenon can often be employed in order to distinguish the two categories.
Semantically and pragmatically, pseudotitles are clearly different from titles. The latter, in principle, fall into the following seven “more or less closed” classes (Bell 1988:329): Professional (Doctor, Professor, etc.), Political (Minister, Senator, Governor, etc.), Religious (Pastor, Bishop, Reverend, etc.), Honors (Sir, Earl, etc.), Military (Commander, Sergeant, etc.), Police (Inspector, Superintendent, etc.), and Foreign (Monsieur, Senorita, etc.). Whereas titles serve to honor the individual named in the following NP and are thus markers of “respect” or “status” (Quirk et al. 1985:773), the function of pseudotitles is rather to describe. Because they are honorifics, titles are capitalized. In principle, pseudotitles are not, but practice is much more mixed.
Despite these seemingly clear-cut orthographic, syntactic, and semantic-pragmatic differences between pseudotitles and titles, there is what Bell (1988:331) has termed a “titleness squish,” that is, a considerable area of overlap between the two categories. This overlap is owed not only to differences between varieties in the sense that governmental designations, for example, appear to be considered titles by American journalists, whereas alternative appositive structures seem to be preferred in British newspapers (e.g., Gov. George W. Bush of Texas vs. the Prime Minister, Mr John Major; Meyer 2002b:149–150). It is also found for individual nouns in individual varieties, which seem to “have the ability to slide out of” one category and “slip into” the other (Bell 1988:330). In order to resolve potentially ambiguous cases, I relied on syntax and classified tokens containing a modifier as pseudotitles, as their function is certainly more descriptive than honorific (see Bell 1988:330; Meyer 2002b:151). Table 1 presents a few examples from my Bahamian data that appear to possess such dual-class membership.
Nouns Occurring in Both Title and Pseudotitle Function in Bahamian Press Texts.
As noted by Bell (1988:333), in general, a pseudotitle is “brief like the title which it imitates: a single noun, perhaps with one preposed modifier.” He also points out, though, that the acceptability of longer pseudotitles “varies from place to place, time to time, and media to media” (Bell 1988:332). As will be shown in more detail in Table 3 below, in Bahamian newspapers, the majority of pseudotitles are only one or two words long. However, there are also some very long and quite complex ones, as examples (3) to (8) illustrate. 4
(3) Ministry of Health and Environmental Services Parliamentary Secretary Ron Pinder said on Sunday, that, acting on numerous complaints from the public, health officials will conduct inspections on local food and convenience stores. (Nassau Guardian July 1, 2003)
(4) Rotary Assistant Governor for Grand Bahama Donald Ward headed a group of Rotarians forming the welcoming party for three exchange students participating in the Rotary International Exchange Programme. (Freeport News August 16, 2004)
(5) Youth Pastor and host of the radio program, “Plugged In,” Duerre Thomas explained the M.A.D. acronym means ‘making a difference’ […]. (Freeport News February 13, 2007)
(6) Vice-President of Grand Bahama Island Tourism Board and Chairman of the Organizing Committee for Caribbean Tourism Conference 25 Terrance Roberts […] explained that it was about recognition and to discuss the action plan put in place by the committee. (Freeport News December 5, 2002)
(7) Senior Education Officer in charge of Language Arts at the Ministry of Education and part of the adult reading programme, Daphne Barr, concurred at the training seminar that there are other adults in the programme learning to read […]. (Freeport News October 27, 2005)
(8) A number of prominent people in the community gathered at the scene, including […] close friends former prime minister Perry Christie and his wife Bernadette […]. (Nassau Guardian November 19, 2007)
Pseudotitles stand in a relationship of “systematic correspondence” (Quirk et al. 1985:57) with equivalent appositive structures, that is, with NPs that are either identical in reference with the name NP they co-occur with, as in the permanent secretary to the Ministry of Health, Mr. R. H. Robertson (NGBO May 7, 1968), or include the reference of the concomitant name NP (Quirk et al. 1985:1301), as in Rodger Johnson, an executive director at Grand Bahama Power Company (Freeport News December 5, 2002). Appositives have been found to be very common in press texts, presumably because they “convey information in as economical a form as possible” (Meyer 1992:118). As the previous examples show, they may be preposed or postposed. A determiner is necessary if an appositional construction is preposed (otherwise we are dealing with a pseudotitle) but optional if postposed. Appositives are often separated from the name NP they co-occur with by means of a comma; pseudotitles, like titles, in principle are not. Again, however, practice may be mixed. Table 2 summarizes the criteria for distinguishing between titles, pseudotitles, and appositives as just outlined.
Criteria for Distinguishing between Titles, Pseudotitles, and Appositives.
It is important to note at this point that, even though in terms of syntactic structure pseudotitles behave like titles, the following analyses are restricted to the variation between pseudotitles and appositives, which is indicated in Table 2 by means of shading. Only the latter two categories can actually be considered “way[s] of saying the same thing” (Labov 1972:323), that is, variants of a linguistic variable. This is owed to the fact that, as already indicated, while titles merely serve as honorifics, pseudotitles and appositives actually have descriptive function, so that in purely semantic terms there is a difference between Minister Loretta Butler-Turner (title) and Minister of State for Social Development Loretta Butler-Turner (pseudotitle; Nassau Guardian November 21, 2007) but not between the latter and the Minister of State for Social Development, Loretta Butler-Turner (appositive) or Loretta Butler Turner, (the) Minister of State for Social Development (appositive).
There are 281 pseudotitles and 258 individually occurring appositive structures in the sample. A further twenty-eight appositives occur in conjunction with a pseudotitle, as in policewoman Blenda Lightbourn, a late witness and a participant in the raid and arrest (NGBO May 22, 1968), or another appositional construction, as in the Governor, Sir Ralph Grey, Patron of the Bahamas Red Cross (NGBO February 21, 1968). The total number of titles (including Mr., Mrs., etc.) is 667. This includes 131 titles co-occurring with other structures, as in Civil Rights leader, Dr. Martin Luther King (NGBO June 21, 1968), but as just noted, titles do not form part of the envelope of variation as defined here. All of the following statistics will be restricted to the alternation between pseudotitles and individually occurring appositives as semantic equivalents; the co-occurrence of different types of structure will be taken up in the discussion of the Varbrul analysis performed on the data.
Results 1: The Distribution of Pseudotitles in British, American, and Bahamian Newspapers
Figure 1 presents the normalized frequencies of pseudotitles in the four Bahamian papers sampled here compared to the numbers obtained by Meyer (2002b:157) for ICE Great Britain (ICE GB) and ICE U.S.A.
Pseudotitles in British, American (Meyer 2002b:162), and Bahamian Newspapers, per 20,000 Words.
Figure 1 shows considerable differences between the samples displayed. As expected, there are more than twice as many pseudotitles in ICE U.S.A. than in ICE GB. While both Nassau papers show even more pseudotitles than ICE U.S.A., the Abaconian and the Freeport News pattern in between ICE GB and ICE U.S.A. The differences are statistically significant overall (chi-square = 39.7339, p = .000169) as well as for individual pairs such as NGBO versus ICE GB (chi-square = 25.2737, p = .0005). The pair ICE GB versus ICE U.S.A (chi-square = 15.8049, p = .07022) just barely misses the .05 significance mark. What is definitely not significant is the difference between the contemporary Nassau Guardian and its colonial forerunner, NGBO (chi-square = 0.5294, p = .4669). In sum, if Bahamian journalists now follow American norms, they apparently also did so before independence, at least in Nassau.
Obviously, these tests are not particularly informative yet, as Figure 1 displays only pseudotitles but disregards alternative structures. This is important to note because the differences in the frequencies with which pseudotitles occur in the various samples displayed in Figure 1 might simply be an artifact of an underlying difference in the number of contexts in which they could have occurred. Figure 2 displays the proportions of pseudotitles and appositives in all samples investigated for this study. Again, the counts for ICE GB and ICE U.S.A. have been taken from Meyer (2002b); numbers are all normalized to 20,000 words.
Pseudotitles versus Appositives in British, American (Meyer 2002b:157), and Bahamian Newspapers, per 20,000 Words.
As Figure 2 shows, with the exception of the historical Bahamian press sample, the numbers of contexts in which pseudotitles could have occurred, that is, the sums total of pseudotitles and appositives per 20,000 words, are almost identical across the samples, and, in fact, the differences displayed are not statistically significant (chi-square = 2.3678, p = .6685). This changes if the NGBO is included (chi-square = 31.4609, p = .007595). There is, however, no correlation between the sums total and the frequencies of pseudotitles across the various samples (cor-coef = 0.5591148, p = .2487). The NGBO stands out in that it has by far the most contexts in which pseudotitles could have occurred, but in terms of the proportion of pseudotitles (69/161 = 43 percent), it quite closely resembles two of the other Bahamian papers, that is, the Abaconian (46/93 = 49 percent) and the Freeport News (46/105 = 44 percent) and also, albeit to a lesser extent, ICE U.S.A. (59/110 = 54 percent). By far the highest proportion of pseudotitles occurs in the contemporary Nassau Guardian (80/112 = 71 percent), the lowest in ICE GB (23/101 = 23 percent).
The differences in the distribution of pseudotitles versus appositives are again statistically significant not just overall (chi-square = 55.1007, p = .0000001245) but also if individual pairs, such as ICE GB versus ICE U.S.A. (chi-square = 19.8315, p = .008458), NGBO versus ICE GB (chi-square = 10.2772, p = .001347), Nassau Guardian versus ICE U.S.A. (chi-square = 7.0145, p = .008085), or Nassau Guardian versus NGBO (chi-square = 21.484, p = .003568), are considered. Note that the latter two pairs did not show significant differences between their members when only frequencies of pseudotitles per 20,000 words were analyzed but show them now when the distribution of pseudotitles versus appositives is taken into account. This is particularly interesting with regard to the pair of press samples from Nassau, which showed very similar (and very high) frequencies of pseudotitles per 20,000 words in Figure 1. In Figure 2, by contrast, the historical NGBO appears much more conservative than its modern successor in that, despite a high frequency of pseudotitles, appositive structures are still preferred to an even greater extent.
As noted in the introduction, pseudotitles began as “a characteristic feature of current English journalistic style” (Rydén 1975:15) in the United States and from there spread to other English-speaking countries. In that process, they also seem to have developed in length and complexity, at least in postcolonial varieties of the language. Thus, Meyer (2002b:163) finds, first, that ICE New Zealand and ICE Philippines contain “significantly more pseudo-titles that were lengthier than five words” than both ICE GB and ICE U.S.A. and, second, that “in New Zealand and Philippine press reportage, innovative forms are evolving, making these varieties different from the others.” In his view, this “not only shows that a grammatical construction can be borrowed from one variety into another but that once the construction is borrowed, the constraints on its usage can change, leading to new forms” (Meyer 2002b:148).
Table 3 presents raw frequencies of pseudotitles by length 5 in the various Bahamian newspapers sampled for this study. It shows that there is a drastic drop in frequency between one- and two-word pseudotitles and three- and four-word ones, and then an additional drop for pseudotitles of five and more words. Figure 3 displays the normalized frequencies of short versus long pseudotitles in the various Bahamian press samples and in ICE GB and ICE U.S.A. The binary distinction into short (one to four words) and long (five or more words) pseudotitles was adopted in order to make my Bahamian data comparable to Meyer’s (2002b:162), who distinguishes only between these two categories. 6 Together with the examples presented in the section circumscribing the linguistic variable under investigation, it suggests that the development of pseudotitles is comparatively advanced in the Bahamas. Clearly, in Bahamian newspapers pseudotitles tend to be longer than in either British or American papers. In fact, ICE GB contains not a single pseudotitle that has five or more words. The differences in the distribution of short versus long pseudotitles are, once more, statistically significant (chi-square = 32.6843, p = .004348).
Pseudotitles by Length in Four Bahamian Newspapers, Raw Frequencies.
Short (One to Four Words) versus Long Pseudotitles (Five or More Words) in British, American (Meyer 2002b:162), and Bahamian Newspapers, per 20,000 Words.
A very interesting finding is that the Abaconian and the Freeport News all of a sudden appear quite different in that the proportion of long pseudotitles is significantly greater in the latter (15/46 = 33 percent) than in the former (2/46 = 4 % chi-square = 10.3906, p = .001267). The Abaconian, in fact, now patterns with the historical Bahamian press sample, that is, the NGBO (7/69 = 10 percent). Actually, if conservatism is indicated by the fact that long pseudotitles are avoided to an even greater extent than pseudotitles in general, as in ICE GB, then the Freeport News appears as the most progressive paper of the sample, featuring by far the highest proportion of long pseudotitles. Considering the fact that the Abaconian and the Freeport News showed very similar patterns with regard to both the frequency of pseudotitles overall and the distribution of pseudotitles versus appositives, this result appears surprising at first sight. If the social history and current sociolinguistic situation of the islands hosting the two newspapers are taken into account, however, the surprise disappears.
The Abaconian is published in Abaco, one of the northernmost Bahamian islands, whose distinctive character, as noted in the section describing the sociohistorical background of English in the Bahamas, is owed to its loyalist heritage. Whites in Abaco have always had a larger share of the population than in the Bahamas as a whole (Craton & Saunders 1998:60, 196); they call themselves “Conchy Joes,” and many of them vehemently opposed Bahamian independence and even tried to secede from the Bahamas and form their own British colony when it happened (Craton & Saunders 1998:359–360). Grand Bahama, by contrast, which is where the Freeport News appears, remained very sparsely settled until the mid-twentieth century, when the only inhabitants were a couple hundred descendants of slaves who had been left behind after the abolition of slavery in 1834 (Craton & Saunders 1998:59). Apart from two brief interludes of smuggling activity during the American Civil War and the prohibition era, the only economic activities practiced at the time were farming and fishing. In 1955, the Bahamian government, together with an American financial investor, decided to develop Grand Bahama into a tourist destination for wealthy Americans. Freeport was founded in the same year, and hotels, airports, a cruise ship harbor, and an oil-bunkering facility were built (Craton & Saunders 1998:323–327, 332–336). Today, Grand Bahama is probably the most industrialized and urbanized island of the entire archipelago. It appears that the differences in the histories and current sociopolitical, demographic, and economic situations characterizing the two islands are subtly reflected in language use, with the use of pseudotitles, as an Americanism symbolizing an orientation toward “modern” lifestyles, participating in this differentiation.
Results 2: Factors Influencing the Occurrence of Pseudotitles in Bahamian Newspapers
In order to test which factors favor or inhibit the use of pseudotitles in relation to equivalent appositives in Bahamian press texts, a Varbrul analysis was conducted, the results of which are displayed in Table 4. Varbrul is a statistical computer package that builds on raw frequencies in any corpus of performance data and uses maximum likelihood estimates in the context of a logit-additive model in order to determine the strength and direction of the various factors affecting the application of a variable linguistic rule, such as the occurrence of a pseudotitle rather than an equivalent appositive (see, e.g., Young & Bayley 1996). The program version used here is GoldVarb 3.0b3 (Sankoff, Tagliamonte & Smith 2005).
Factors Influencing the Use of Pseudotitles versus Appositives in Bahamian Newspapers.
Note: Log likelihood = −270.342, chi-square/cell = 1.0449.
The individual factors examined are displayed in column 1; column 2 shows token frequencies per factor; and column 3 gives the proportion of tokens that feature a pseudotitle rather than an appositive. The factor weights in column 4, finally, assess probabilities of choosing a pseudotitle over an equivalent appositive. These probabilities were established in a stepwise logistic regression analysis (“step-up/step-down”), in which each factor group was systematically added to and removed from the analysis in order to determine if its presence would make a significant contribution to the model of variation. Factor weights greater than .5 indicate favoring effects, those less than .5 disfavoring ones. Square brackets around individual factor weights show that Varbrul discarded this factor group in the stepwise regression analysis because it did not contribute significantly to the observed variation. Factors were culled from the literature; where necessary, they were adapted to the realities of the corpus employed for the investigation reported here.
The entire data set is described by an “input probability,” best understood as a measure of the rule to be applied overall, that is, apart from influence of the various factors. Two other pieces of information are calculated by the program: the log likelihood and the chi-square/cell. Both represent statements as to the goodness-of-fit of the model generated. A chi-square/cell value of one is generally deemed a good value (Patrick 1999:239).
The factor group “Paper” confirms the results shown in Figure 2 above, that is, that the Nassau Guardian is the Bahamian paper that most favors pseudotitles over equivalent appositives. Pseudotitles are also favored by constructions involving political nouns. This category includes not only ministerial terms and other governmental designations, as in former Minister of Labour Peter Graham (NGBO February 8, 1968) but also terms denoting political occupations of other kinds, as in United Bahamian Party backbencher Bernard Dupuch (NGBO February 8, 1968). 7 Just as in Rydén’s (1975:17) data, the most frequent political noun in the Bahamian newspapers is minister (N = 107). In accordance with Bell (1988:332) and Meyer (2002b:149) (as well as with the criteria specified in the section circumscribing the variable context), I categorized Minister as a title and excluded it from the analysis if it appeared either on its own or in the phrase Prime Minister (N = 41) but considered it part of a pseudotitle (N = 48) if it was accompanied by modifiers, as in the example above. 8
The results for noun type displayed in Table 4 differ from those obtained by Rydén (1975:16–17), who, in a study of “Noun-Name Collocations in British English Newspaper Language,” found that political nouns occurred in pseudotitles (instead of in an appositive construction) only 13 percent of the time, whereas other nouns did so 26 percent of the time. Not only are pseudotitles thus much more frequent in Bahamian journalistic writing than in British newspapers (as shown in Figures 1 through 3 above), but with regard to noun type the order of constraints is also reversed. This may have to do with what Bell described in the late 1980s, when he noted that “expressions such as Prime Minister Lange have had full acceptance as titles in the United States media for many years” but “remain deviant” in the British quality press (Bell 1988:332). The situation must have been the same in the 1970s, when Rydén’s data were collected. That Bahamian usage appears American compared to those data is not surprising.
That the occurrence of a pseudotitle or an equivalent appositive is related to the length of the construction is obvious, and Varbrul here once more confirms what Figure 3 already indicates. Length, however, is also related to the structural complexity of an NP. I distinguished four types in this respect: (1) constructions containing no modification, such as Labourer Edward Morley (NGBO February 8, 1968; in this case, the descriptive NP Labourer consists of a head only); (2) constructions featuring premodification only, such as Freeport businessman Lambert Bowe (Freeport News November 19, 2007; here, Freeport is the premodifier, businessman the head of the descriptive NP); (3) constructions containing postmodification only, such as leader of The Swingers, Anthony “Huck” Williams (Freeport News October 27, 2005; head = leader, postmodifier = of The Swingers); and (4) constructions showing both premodification and postmodification, such as Rotary Assistant Governor for Grand Bahama Donald Ward (Freeport News August 16, 2004; premodifier(s) = Rotary Assistant, head = Governor, postmodifier = for Grand Bahama). Whereas the presence of a premodifier does not disfavor the occurrence of a pseudotitle (and, in fact, slightly favors it compared to the absence of any kind of modification at all), postmodification, either on its own or in conjunction with a premodifier, strongly inhibits pseudotitles.
This is entirely in accordance with the findings of both Bell (1988:336) and Meyer (2002b:160), and there is a strictly linguistic reason for the phenomenon. As Bell (1988:328) notes, in contrast to appositives, pseudotitles are not independent NPs but are embedded under the name NP they precede. As such, they are subject to the principle of “end-weight” (Quirk et al. 1985:1361–1362), which stipulates that heavier constituents be placed at the end of a structure rather than at its beginning. Obviously, the longer and more complex a pseudotitle, the more unbalanced the NP containing it will be. An appositive, by contrast, is independent of the name NP to which it is related, which, as noted in the section circumscribing the linguistic variable, is often indicated in writing by means of a comma. The principle of end weight thus does not apply in the same way to appositives, which means that they can be not only longer but also more complex than pseudotitles. Postmodifiers, which most typically consist of prepositional phrases or relative clauses, are obviously longer and more complex than premodifiers, which most frequently consist only of adjective phrases, participle constructions, or other nouns. This explains the strong inhibiting effect that the occurrence of postmodification has on the use of pseudotitles not just in the Bahamian data but also in the data investigated by Rydén (1975:22–27), Bell (1988:336), and Meyer (2002b:160).
The co-occurrence of pseudotitles with other constructions is also restricted (Bell 1988:336; Rydén 1975:28). In coding the data, co-occurring constructions were not counted twice but assigned to the appropriate category and coded for the kind of co-occurrence relationship that obtained. As the present study focuses on pseudotitles, these always took precedence over other constructions; appositives, in turn, took precedence over titles. Thus, both Former National Democratic Party chairman Mr. Holland G. Smith (NGBO January 6, 1968) and policewoman Blenda Lightbourn, a late witness and a participant in the raid and arrest (NGBO May 22, 1968) would have been counted as pseudotitles but coded for co-occurrence with a title (Mr.) and an appositive (a late witness …), respectively; Mr. Preston Albury, PLP representative for Eleuthera (NGBO February 21, 1968), by contrast, would have been counted as an appositive co-occurring with a title. As titles such as President, General, or Bishop and what Bell (1988:329) has labeled “M-terms,” that is, Mr., Mrs., and so forth, are usually mutually exclusive because they occur in the same structural slot, they were collapsed into a single category.
As shown in Table 4, the presence of a title or M-term inhibits the occurrence of a pseudotitle, even though the co-occurrence of the two is by no means impossible, as these examples show: attorney Mrs. Smith (Abaconian January 15, 2012), Administrator Dr. Christie (Freeport News April 28, 2003), COB Dean of Social and Educational Studies Dr. Thaddeus McDonald (Nassau Guardian November 19, 2007). Again, structural reasons explain why a pseudotitle and a title or an M-term do not usually go together: as the name implies, pseudotitles occur where we otherwise find titles. As also noted by Bell (1988:329), however, “two titles may in fact be aggregated in front of a name,” as in Field-Marshal Earl Haig. The same appears to hold for the combination of title or M-term and pseudotitle. The presence of other constructions, as in policewoman Blenda Lightbourn, a late witness and a participant in the raid and arrest (NGBO May 22, 1968) affects the occurrence of pseudotitles only marginally.
The form of the name NP to which a pseudotitle or appositive is related apparently has no influence on the choice of construction. Other than in Bell’s (1988:336) New Zealand data, last name only does not inhibit the occurrence of a pseudotitle, and first name + last name does not favor it. Two possible reasons for this finding are (1) the small number of tokens containing last names only and (2) differences in the use and function of particular forms of address between the Caribbean and other English-speaking countries. Unfortunately, very little research has been done on this topic to date (though see Mühleisen 2005), so this interpretation remains entirely speculative at this point.
That different kinds of coordination affect the occurrence of pseudotitles has been observed by Rydén (1975:27–28). Coordination may occur in the descriptive NP (Bell 1988:333), where it may involve not only a single head with multiple modifiers, as in Labour and Welfare Minister Milo Butler (NGBO June 21, 1968; head = Minister, modifiers = Labor, Welfare), but also multiple heads sharing a single determiner and/or modifier, as in the appositional construction Jim Pickett, Steel HQ’s chief designer and builder (Freeport News February 13, 2007; modifier = chief, heads = designer, builder). There may also be two entirely independent NPs, as in Youth Pastor and host of the radio program, “Plugged In,” Duerre Thomas (Freeport News February 13, 2007; NP1 = Youth Pastor, NP2 = host of the radio program, “Plugged In”). Finally, coordination may also occur in the name NP, as in Rotarians Derek Hendfield, David Long, Dreswell Coakley (Freeport News August 16, 2004). On the whole, coordination is rare in the NPs analyzed here; apart from descriptive NPs in which multiple heads share a single modifier, which, however, are exceedingly rare, it does not inhibit the use of a pseudotitle. In this respect, the Bahamian data differ from those analyzed by Rydén, who finds that “normally, co-nouns are postpositive” (1975:27).
Finally, there is the relationship between pseudotitles and captions. If, as observed by Bell (1988:329), a pseudotitle “compresses an expression while simultaneously increasing its emphasis,” then one would expect to find such structures in captions even more frequently than in news language in general. A similar claim is, in fact, made by Rydén (1975:37), who maintains that “most amenable to the construction are captions, newsflashes and other kinds of condensed language calling for a high degree of spatial and syntactic economy” but does not provide any numbers. In the Bahamian data, pseudotitles are no more frequent than appositives in photo captions; rather, the reverse is the case. As indicated by the square brackets, however, this factor group did not contribute significantly to the observed variation.
Results 3: A Closer Look at Appositives
So far, this study has only contrasted pseudotitles with a general class of appositive structures, be they preposed, as in the winner of the laptop, John Haestad (Abaconian February 1, 2010), or postposed, as in Ted Doyle, chief of Government information services (NGBO May 7, 1968). Even though preposed and postposed appositives of the same syntactic type, as in the winner of the laptop, John Haestad versus John Haestad, the winner of the laptop, are semantically identical, they still appear to pattern very differently, at least in British newspapers. As Jucker (1992:219) has put it, “the crucial decision facing newspaper journalists, or their editors, is not whether to use the Labour Leader, Neil Kinnock or Labour Leader Neil Kinnock but whether to use Labour Leader Neil Kinnock or Neil Kinnock, (the) Labour Leader.” He finds that “up-market papers strongly prefer postposing, whereas the mid-market papers strongly prefer preposing” (Jucker 1992:220). There is, thus, a stylistic divide separating preposed and postposed descriptive NPs in British newspaper language, with postposed constructions representing the more formal, conservative pattern. If we assume that during Americanization it is particularly such conservative patterns that are replaced by others, this would mean that we should find a higher proportion of postposed appositives in the pre-independence NGBO than in its modern successor, the Nassau Guardian. That this is, in fact, the case is shown very clearly in Figure 4, which displays not only the proportions of preposed and postposed appositives but also the different relationships linking them with pseudotitles in Bahamian newspaper language.
Preposed versus Postposed Appositives and Their Relationships with Pseudotitles in Bahamian Newspapers, per 20,000 Words.
As far as these relationships are concerned, I followed Meyer (2002b:153–154) and distinguished three types. Appositives that can be directly converted into a pseudotitle stand in a relationship of “exact equivalence” with the latter, for example, Tyler Smith, first place winner (Freeport News November 19, 2007). These structures, by definition, can only occur in postposition, because, as outlined in the section circumscribing the variable context, if they occurred before the name NP we would be dealing with a pseudotitle. If only a determiner, including possessive determiners and numerals, would have to be deleted in order for a structure to function as a pseudotitle, it was counted under the heading of “determiner deletion,” for example, Jonnajah Bootle, the second participant (Abaconian February 1, 2010). If only a part of the structure would be usable as a pseudotitle, this was counted as a case of “partial equivalence.” Often, such structures contain relative clauses as postmodifiers, as in Mrs. Rolle, a teacher who has taught adults to read for the last 10 to 15 years (Freeport News October 27, 2005), or genitive NPs in determiner position, as in Geraldo Capo, the project’s developer (Nassau Guardian May 2, 2003); they are thus considerably more complex than structures related to pseudotitles either via exact equivalence or via determiner deletion (see Meyer 2002b:159).
Postposed appositives are about three times as frequent as preposed ones in the 1968 NGBO (70:22). This ratio declines to roughly two to one in the Abaconian (32:15); the contemporary Nassau Guardian features approximately equal numbers of preposed and postposed appositive structures (15:17). Interestingly, we find about six times as many postposed appositives than preposed ones in the Freeport News (51:8). 9
As displayed in Figure 4, even though both exact equivalence and determiner deletion occur in all papers except the Abaconian, where there are no exact-equivalence structures, these two types play a minor role among appositive structures in Bahamian press texts. Whereas determiner-deletion structures, however, tend to occur before the name NP whose referent they describe, partially equivalent appositives in the vast majority of cases occur after it. NGBO contains by far the most of these structures, followed by the Freeport News. Partially equivalent appositives are least frequent in the Nassau Guardian, and they occur almost equally frequently in preposed as in postposed position in that paper. As just noted, partially equivalent appositives tend to be more complex than both those that are exactly equivalent to pseudotitles and those where only a determiner would have to be deleted.
What Figure 4 confirms, thus, is a movement in Bahamian press language away from complex, postposed descriptive NPs in noun-name appositional constructions on to more compact, preposed constructions, be they appositives or pseudotitles, in between pre-independence and contemporary samples. Bahamian journalists are thus taking part in a more general development in contemporary English, which is the densification of content in individual syntactic units such as NPs (Leech et al. 2009:249–252). This development appears to be particularly pronounced in informational text types such as journalistic writing, and it appears to be describable as part of a process of Americanization affecting varieties of English worldwide. As Leech et al. (2009:234) put it, “before the mid-twentieth century, AmE and particularly the Press variety (both in the US and in the UK) were innovative in spearheading this tendency towards more compact noun phrase structure with greater density of information […] this tendency was then passed on to other varieties of written English, where a ‘catching-up’ phase has been under way.”
Conclusion
This paper has investigated the use of pseudotitles, that is, determiner-less structures providing descriptive information in front of name NPs, in Bahamian newspaper writing. This feature is frequent in journalistic language; it clearly originated in the United States but has been spreading to varieties of English worldwide. The data were taken from a corpus of contemporary and historical Bahamian newspapers; they were analyzed by means of both descriptive and inferential statistics. Finally, equivalent appositive structures were considered.
Even a cursory examination of the frequency of pseudotitles in Bahamian press news reports revealed that the Americanization of journalistic writing in the Bahamas is definitely not a recent phenomenon. Bahamian journalists, at least in the capital, Nassau, used pseudotitles liberally even when the country was still under British colonial rule, which does not come as a surprise if one takes into account the social history and current socioeconomic situation of the Bahamas. If we look at the distribution of pseudotitles versus appositives, we see the same trend—magnified, in fact, in the Nassau Guardian, which has a higher proportion of pseudotitles than even contemporary American newspapers.
While pseudotitles may have originated in the United States, they have not only spread to the Bahamas but also developed there. Particularly journalists from the two contemporary urban Bahamian papers investigated here, the Freeport News and the Nassau Guardian, literally seem to be stretching the construction to its limits not just in terms of length but also in terms of complexity. Whether this apparent partiality to determiner-less structures is in any way related to Bahamian journalists’ linguistic background as creole speakers is purely speculative at this point. It may be worth noting, however, that the determiner systems of creoles are very different from the system we know in standard English, with zero forms playing a much more important role (see, e.g., Holm 2000:214).
An examination of some factors governing the occurrence of pseudotitles showed that while there are general constraints common to all varieties of English, Bahamian newspaper language exhibits some notable differences, too. Other than in British English, for example, political nouns are quite conducive to the occurrence of pseudotitles; other than in New Zealand English, last name only does not inhibit the occurrence of a pseudotitle, and first name + last name does not favor it. The examination of equivalent appositive structures revealed a movement in Bahamian press language away from complex, postposed descriptive NPs in noun-name appositional constructions toward more compact, preposed constructions, be they appositives or pseudotitles.
With regard to the question of American influence on other varieties of English, Leech et al. (2009:254) note that caution is often warranted “regarding the too easy assumption of ‘Americanization.’” It is certainly true that pseudotitles originated in American journalistic usage and that, owing to the particular closeness of the Bahamas to North America, Bahamian journalists took over this feature even in colonial times. This does not indicate a simple convergence on American norms, however. On the one hand, Bahamian English seems to be taking part in the general trend toward the linguistic condensation of information in the NP that has already been described for both British and American English (e.g., Leech et al. 2009:249; Biber, Grieve & Iberri-Shea 2009:184). On the other hand, Bahamian journalists seem to be following their own route in at least some aspects of this process in the sense that they are adapting the feature to local linguistic realities, which presents another piece of evidence in favor of a nuanced view of linguistic Americanization.
Footnotes
Acknowledgements
I am grateful to Annette Hillers-Chen for help in obtaining thehistorical newspaper data, to Natascha Wanninger and Diana Wengler for help in preparing the data for analysis, and to Janina Kraus for statistical assistance. I also sincerely thank the editors and anonymous reviewers for their helpful comments. All remaining faults and errors are my own, of course.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.