The development of phrasal verbs in British English from 1650 to 1990A corpus-based study

Resum

Phrasal verbs or particle verbs are one of the most idiosyncratic features of the English language, as well as of other Germanic languages, such as German or Dutch. They pose many problems for non-native speakers, because their meanings have to be learned separately from the meanings of their verbal bases (give vs. give up), given that the union of the two elements of the compound (the verb and the particle) very often gives rise to new non-compositional forms very similar to idioms. This dissertation tackles some of the questions concerning the nature of phrasal verbs. First, I intend to delimit the concept of phrasal verb as conceived of in Present-day English. One of the topics often discussed in relation to this category is precisely that of the difficulty of establishing the boundaries between phrasal verbs and other related categories. Second, I aim at filling a gap in the literature of phrasal verbs by carrying out a corpus analysis of the development of these structures in the recent history of English, more precisely between 1650 and 1990. After comparing the recent history of phrasal verbs with their status in earlier stages of the language as described in the literature, a third aim of the present dissertation is to establish a relationship between these structures and the processes of grammaticalization, lexicalization and idiomatization. For these purposes, data have been extracted from several sources, in particular from ARCHER 3.1 (A Representative Corpus of Historical English Registers), although many illustrative examples, especially those involving Present-day English combinations, have also been obtained from the BNC (British National Corpus) or the Internet. Recurrent use has also been made of several dictionaries of English, most notably the Oxford English Dictionary, as well as dictionaries specialized in phrasal verbs and related structures. The dissertation is organized into seven chapters and contains five appendices. The outline of the chapters is as follows. Chapter 1 opens with a description of the aims and scope of the dissertation (1.1) and offers some theoretical preliminaries about the object of study (1.2). Here English phrasal verbs are introduced and compared with other related categories with which they bear certain similarities, especially prepositional verbs and phrasal-prepositional verbs. More precisely, phrasal verbs are described within the larger context of multi-word verbs and thus defined as discontinuous lexical items. This is one of the commonest views in the literature, though not the only one. In fact, one of the main debates over this category concerns the question of whether phrasal verbs belong with the lexicon, syntax or phraseology. Chapter 2 constitutes a more thorough analysis of the category of phrasal verbs. First, the two elements of the compound are examined individually. Section 2.2 deals with the verbal elements, which are commonly Germanic and monosyllabic or disyllabic verbs with the accent on the first syllable. In Section 2.3, in turn, I discuss the nature of the particles. Morphologically (2.3.1), phrasal-verb particles are generally described as adverbs in origin which have come to form a (more or less bound) unit with a verb and which, for this reason, behave differently from other adverbs. The semantics of the particles is the aim of Section 2.3.2, where I establish a distinction between five different semantic types of particles. Literal particles (2.3.2.1) express the meaning of `motion in general¿. Other particles add aspectual or aktionsart meanings to the verbs they collocate with (2.3.2.2). Distinguishing between aspect and aktionsart has proved difficult since the label `aspect¿ has been employed in the literature to refer to both concepts. In general, phrasal-verb particles can be said to portray either telic aktionsart meaning (especially up), iterative and/or continuative aspect (mainly along, away and on) and inchoative or inceptive aspect (e.g. away in some imperative sentences). In Section 2.3.2.3 I argue for the existence of a category of emphatic particles which function as `colloquializers¿, i.e. as elements which are added to simple verbs, not only to provide them with some kind of emphasis, but also with a familiar, colloquial tone. Some other phrasal-verb particles have developed metaphorical or figurative connotations from their original meanings of movement, most commonly the so-called `orientational metaphor¿ (Section 2.3.2.4). Finally, idiomatic particles (2.3.2.5) are those which form a semantic unit with the verb in such a way that their meaning can no longer be inferred. In general, although the semantic load of the individual elements of the compound is important, phrasal verbs and other multi-word verbs have often been described as showing division of labour or structural compensation, i.e. the semantics of the compound does not belong to one or another element of the combination, but is rather shared by both of them (Section 2.3.2.6). The view taken in the present dissertation is that phrasal verbs are discontinuous lexical items. For this reason, besides examining the characteristics of the individual elements of the combination, an analysis of phrasal verbs as a unit was mandatory. This is the aim of Section 2.4, where the idea that the verb and the particle of phrasal verbs form a single semantic and lexical unit is justified on the basis of four main criteria: replaceability by a single-word equivalent (e.g. put off ¿ postpone), derivation (e.g. breakinable, turn-off), coordination with another verb (e.g. chatting away and making her laugh) and elision (e.g. I cleaned up my room and mum [cleaned up] the kitchen). Section 2.4.1 deals with the description of the semantics of phrasal verbs, which are classified within six main semantic groups, namely literal (2.4.1.1), figurative (2.4.1.2), idiomatic (2.4.1.3), aspectual/aktionsart (2.4.1.4), reiterative (2.4.1.5) and emphatic (2.4.1.6). These do not constitute discrete categories, since sometimes the same verb-particle combination can be classified into more than one group. In Section 2.4.2 I review some of the main syntactic tests which have generally been applied for the distinction of phrasal verbs from other related categories, in particular prepositional verbs. The only tests which have proved satisfactory to identify phrasal verbs are the definite NP test (2.4.2.2), the test according to which phrasal-verb particles cannot occur before relative or interrogative forms (2.4.2.11), and also the test according to which an adverb phrase cannot be inserted between the verb and the particle (cf. Section 2.4.2.3). These are supported by the test of particle placement with object pronouns (2.4.2.1), the adverb insertion test (2.4.2.3) (either the insertion of all, clean, right, straight, the heck and the hell, on the one hand, and of other adverbs, on the other), preposing of the particle (2.4.2.4), passivization (2.4.2.5), particle coordination (2.4.2.7), verb gapping (2.4.2.10) and cleft-formation (2.4.2.12). Finally, the nominalization test (2.4.2.6), insertion of a directional PP (2.4.2.8), substitution by other particles (2.4.2.9) and replacement by an antonym (2.4.2.13) were rejected as reliable tests for various reasons. Chapter 3 depicts the diachronic evolution of phrasal verbs since their first appearance in Old English to the present-day. English phrasal verbs have not always had the structure with which we associate them in the contemporary language. In earlier stages they are attested in a wider variety of syntactic arrangements. Thus, in Old English (Section 3.2) particles could appear in preverbal or postverbal position and with or without intervening material between them and the verbal element. Although the preverbal pattern was predominant during most of the Old English period both in main and subordinate clauses, the use of the postverbal pattern became established in all types of clauses already during Early Middle English. Moreover, the syntax of Old and Middle English phrasal verbs was more flexible than that of their present-day equivalents in allowing various elements to appear between the verb and the particle (e.g. a negative marker, an infinitive marker, a modal verb, a stranded preposition, a direct object or an adverb). Similarly, Early Modern English phrasal verbs (Section 3.3) were still more permissive than their Present-day English counterparts as regards the order of clause constituents, as well as regards the type of intervening elements which can appear between the verb and the particle. Moreover, most scholars agree that idiomatic and aspectual/aktionsart meanings of phrasal verbs start growing in number from this time onwards. Section 3.4 shows that the literature on the development of phrasal verbs during the Late Modern English period and the 20th century is scarce. In general, those studies discussing this rather long period of the history of English point to a continuation of the tendencies found in earlier stages and to the fact that phrasal verbs seem to have grown in frequency both in British and in American English. The aim of Chapter 4 is twofold. On the one hand, I describe the main characteristics of ARCHER 3.1 (A Representative Corpus of Historical English Registers), the corpus used in the present dissertation (Section 4.2), and the methodology applied to the selection of examples (Section 4.3). On the other hand, I present the corpus results, as far as the linguistic characteristics of phrasal verbs are concerned (Section 4.4). In Section 4.2.1 I discuss the main advantages of ARCHER and the features which make it a useful resource for my analysis, namely the time-span covered (1650-1990) and the selection of genres it contains, which allow comparisons between written text types and registers closer to the spoken language. The drawbacks of the corpus are described in Section 4.2.2. These relate mainly to the fact that the American component of the corpus is incomplete and also that different tagging systems were employed by the compilers in the various genres of the corpus, which gives place to some inconsistencies. The most important disadvantage of ARCHER in relation to the contents of this dissertation is the fact that one of the texts is repeated and, for this reason, one of the tokens of phrasal verbs is duplicated. The aim of Section 4.3 is the description of the methodology employed for the corpus analysis. The searches were performed by looking for the individual particles by means of WordSmith Tools. These were stored in a Microsoft Excel database and coded according to linguistic parameters, which facilitated calculations and the creation of tables and graphs representing the tokens. In the second part of this chapter, I present the main linguistic characteristics of phrasal verbs in ARCHER. I discuss the morphological features of the relevant constructions, considering first the verbal element (cf. 4.4.1.1) and the particles (4.4.1.2) separately, and then the combinations as a single unit (cf. 4.4.1.3). A total of 660 verbal bases were found in ARCHER for the creation of phrasal verbs. The analysis revealed that in earlier stages of the language polysyllabic verbs of non-Germanic origin (e.g. massacre, overture, surrender, etc.) are quite frequently attested in phrasal verbs, probably due to the influence of Latin and French as languages of prestige in medieval times. As far as the particles are concerned, the following 33 items were found in the corpus to serve this purpose: aback, aboard, about, above, across, ahead, along, apart, around, aside, astray, asunder, away, back, behind, by, counter, down, forth, forward(s), home, in, off, on, out, over, past, round, through, to, together, under and up. Most of these particles seem to remain stable or increase in frequency over time (back, down, off, on, out, round, through and up), while the only particle whose use decreases drastically is forth. Interestingly, a number of adverbs attested in the language since earlier stages (across, ahead, along, around and past) started being used as phrasal-verb particles only from the 18th century onwards. Judging from the evidence of the type/token frequency and hapax legomena, the productivity of phrasal verbs in ARCHER turned out higher in the first half of the 19th century. As regards the frequency of the combinations at issue, the corpus results show a slight decrease during the first half of the 19th century, though in general phrasal verbs seem to grow up in frequency. Section 4.4.2 provides a semantic analysis of the combinations in the first and last subperiods of ARCHER (i.e. 1650-99 and 1950-90) which shows that, while reiterative combinations decrease and literal uses tend to remain rather stable, aspectual/aktionsart/emphatic combinations, as well as metaphorical and figurative combinations, seem to increase over time. The corpus analysis also rendered interesting results about the syntactic characteristics of phrasal verbs in the period under scrutiny (Section 4.4.3). The syntactic arrangement of the phrasal combinations in ARCHER is very similar to that of contemporary phrasal combinations, although a few isolated cases deserved separate mention (e.g. These I had never before observed down, in any other case in which I had formerly operated). To close this chapter, in Section 4.4.4 I discuss phrasal-verb nominalizations in ¿ing (e.g. a dipping down of a process; Section 4.4.4.1) and other derivatives (e.g. a setter forth; Section 4.4.4.2) in the corpus. Results demonstrate that, while the use of nominalizations has decreased, that of other derivatives has increased. Considering that the productivity of phrasal verbs in processes of word-formation constitutes a sign of the unitary character of phrasal verbs, the increase in the frequency of derivations suggests that phrasal constructions have increased in unity over time. Chapter 5 presents the corpus findings regarding the cross-genre and gender distribution of phrasal verbs in ARCHER. The main aim of this chapter is to check whether it is possible to relate phrasal verbs to the spoken, colloquial language or whether their occurrence in particular text types is conditioned by other factors and, if so, which ones. In Section 5.2 I provide a general description of the various genres in ARCHER and their main characteristics in order to ascertain where they can be situated along the formal-informal and oral-written continua. My aim in Section 5.2.7 is to establish a more fine-grained categorization of these text types, based on the terminology employed by Culpeper & Kytö (2010). These scholars distinguish between speech-related text types, on the one hand, and writing-based and writing-purposed text types, on the other, a classification which proved very useful for the text types in ARCHER. In Section 5.3 I survey the use of phrasal verbs across the different genres. The output of my analysis provided very high and low frequencies of phrasal combinations both in texts which are close to the oral language and in those which approach the written medium, especially during the earlier periods. A more in-depth examination of the relevant texts proved that there are factors beyond the degree of formality of a text type which may condition the frequency of occurrence of phrasal verbs, such as the topics discussed in the narration, the format of the text, the particular idiosyncrasies of the writer and the changing conventions of texts over time. Finally, gender variation is the aim of Section 5.4, which shows that phrasal verbs tend to be favoured by females from 1750 onwards (except in the period 1850-99). Nevertheless, the difference in usage between female and male writers turned out to be too scarce to permit categorical conclusions. In the first part of Chapter 6 I analyze the extent to which the processes of grammaticalization, lexicalization and idiomatization may have affected the historical development of phrasal verbs. In Section 6.2 I define grammaticalization as a diachronic process of language change whereby, in certain linguistic contexts, a full lexical item becomes a grammatical morpheme, or a grammatical item increases its grammatical features. Section 6.3 introduced the concept of lexicalization, defined as a diachronic process of linguistic change by means of which a syntactic construction or word formation comes to be used by speakers as a new contentful form. This new form possesses both formal and semantic characteristics, unpredictable from those of the constituents of the original construction or word formation pattern. In Section 6.4 I argued that idiomatization can be defined as a process of semantic change whereby an item or group of items acquire new abstract or idiomatic meanings, thus implying a movement on a scale from concrete to abstract. This process is directly related to both grammaticalization and lexicalization and is very often derived from them, but it also shows a certain degree of independence. On the one hand, phrasal verbs can be seen as a case of lexicalization, considering that the verb and the particle of a phrasal verb form a single lexical item (6.5.1). First, a number of particles have grammaticalized as markers of aspect (6.5.2). Second, in particular combinations (e.g. run away) the verb seems to have undergone a certain degree of grammaticalization by losing much of its original lexical content, the particles thus carrying the primary meaning of the compound (6.5.3). Finally, Section 6.5.4 argued that some combinations have undergone idiomatization, since they have lost (part of) their original connotations of movement in favour of more abstract, idiomatic meanings. This section also reviewed some of the ways in which idiomatic meanings can be acquired by phrasal verbs, namely the semantic development from literal combinations (6.5.4.1), the reduction of elements (6.5.4.2), syntactic reanalysis (6.5.4.3), direct formations (6.5.4.4) and analogy (6.5.4.5). The purpose of the second part of this chapter is to propose a model for the understanding of the relationship between phrasal verbs and the processes of grammaticalization, lexicalization and idiomatization. Taking as a point of departure a model of graded categorization which includes both the syntactic and the semantic aspect of phrasal verbs, in Section 6.6 I suggest that phrasal verbs should be regarded as lexicalized and idiomatized units to various degrees: the extent to which lexicalization has affected the combinations influences their syntactic fixity, whereas idiomatization brings about the emergence of opaque meanings. In view of this, a semantic and a syntactic cline are established to represent the development of phrasal verbs. The various available syntactic tests summarized in Section 2.4.2 are applied to phrasal combinations to check their degree of lexicalization, whereas their degree of idiomatization is measured in terms of the transparency of the individual members of the compound. References ARCHER 3.1 = A Representative Corpus of Historical English Registers 3.1. 1990-1993/2002/2007/2010. Official web: http://www.alc.manchester.ac.uk/subjects/lel/ research/projects/archer/ BNC = British National Corpus. http://davies-linguistics.byu.edu/personal/ Culpeper, Jonathan & Merja Kytö. 2010. Early Modern English dialogues. Spoken interaction as writing. Cambridge: Cambridge University Press. OED = Oxford English Dictionary. http://www.oed.com/