sign in sign up
<<
Home , Inchoative verb, Labile verb

” It g rew a d ay of e xpectation ”

A d iachronic c orpus s tudy on the e volution of the v erb g row in British English

Rosaleena Luokkala

D e partme n t of Engl ish

Bache l or Degree Project English Linguist ics

Autumn 20 1 8

Superv isor: Caroline Gentens

” It g rew a d ay of e xpectation ” A d iachronic c orpus s tudy on the e volution of t he v erb g row in British English Rosaleena Luokka la Abstract English ha s an extraordinary number of labi le , that is, verbs that can be used both transitively with a sense and intransitively with an inchoative sense. This corpus - based stud y investigates the evolution of t he grow from exclusively intransitive to labile in British English in the Late Modern English period. A random sample of 500 instances of the verb grow was drawn from the period 1710 - 1780 a s well as from the p eriod 1850 - 1 920 of the Corpus of Late Mo de rn English Texts i n order to track diachronic changes . The instances in the samples were categorized according to their verb pattern and type of complement ( if any ) , and instances of the past grown were also categorized based on the auxiliary us ed ( be / have /none). The study suggest s that grow came to be used transitively when resultative intransitive constructions (e.g. be grown ( over )) were reanaly z ed as passives ; that the use of phrase complements with copular grow decreased and became archai c to make the di stinction betwe en copular and transitive uses less ambi gu ous ; and that the fact that the be - auxiliary was replaced by the have - auxiliary in constructions helped avoid ambiguity between intransitive and transitive use s o f gro w . Thus , t he study provi des some empirical evidence for Visser's ( 1963 ) hypothesis that th e change from be - to ha ve - perfect s played a central role in the acquisition of lability.

Keywords grow, labile ver b, , corpus study, diac hronic study.

Con tents

1. Introduction ...... 1 2. Background ...... 2 3. Method and material ...... 6 3.1 Corpus linguistics ...... 6 3.2 The Corpus of Late Modern English Texts ...... 7

3.2.1 Extraction of data from the corpus ...... 7 4. Results ...... 8 4.1 Intransitive instances ...... 8

4.1.1 Bare intransitive instances ...... 9

4.1.2 Copular i nstances ...... 9

4.1.3 Ambiguous intransitive instances ...... 12

4.1.4 Phrasal verbs ...... 13 4.2 Transitive instances ...... 14 4.3 Instances with auxiliary verbs ( have / be ) ...... 15 5. Conclusion ...... 16 References ...... 18

1 . Introductio n

Verbs are said to be “the most imp ortan t lexical and syntactic category of a language” (F ellbaum, 1990, p. 278). Whil e grammatical sentences in English can be compos ed without a referential noun, as in It is snowing , they must all contain at least one verb (Fellbaum, 1990, p. 278). Verbs a re al so considered “more complex” than (Pickering & Frisso n, 2001, p. 557). Th e importance and compl e xity of verbs make them a n interesti ng category of l anguage to study. Verbs are distinguished by the syntactic patterns th ey can occur in; the two m ain p atterns are intransitive and transitive. Modern English is extra ordinary in that it has so many verbs that can be used both int ransitively (wi thout an ) and transitively (with an object) without formal marking to disti nguish between the two synt actic patterns. In this study, these verbs will be referred to a s ‘la bile ver bs’. There are esti mated to b e over 800 labile verbs in Present - d a y En glish ( PDE) ( McMillion, 2006, p. 185). The number is considerably smaller in other Germa nic languages (McMill ion, 2006, p. 185; Haspelmath, 1993) and it was also much smaller in O ld En glish ( 55 according to V isser, 1963 , p . 99 ). For some reason th e number of labile verbs in English se ems to be increasing. Due to research lacking in this field, not much is known a bout the mechan isms of lability and why and how specific verbs become labile (Kulikov & Lavidas, 2014, p . 872). Grow is an example of a in PDE . Th e verb has existed sin ce the Middle Ages , an d i t was originally used exclusively intransitively, a s in The corn grew fast (grow, n.d. - a). The earliest example of intransiti ve grow in the Oxford English Di ctionary (O ED ) i s from circa 888 (gr ow, n.d. - c) . A transitive sense meaning “t o produce or cultivate” appeared in the 1700s (grow, n.d. - a). The earlie st ex ample of this transitive use in the OED is from 1774: “They likewise grow some Rice and Tobacco, whic h is sen t t hr oug h Virginia” (grow , n .d. - c). The aim of the study is to invest igate the evolution of the verb grow and the context s in which it came to be u sed trans itively in British English. The study attempts to answer the following question: h ow has the verb gr ow d ev eloped a transitive use in British English over the course of th e Late Modern English period? The question will be in vestigated using th e Cor pus of La te Modern English Texts . The verb grow has been studied prev iously by Alam (2003) and Qing - Hua (2011). T he y both examine the poly semy of gro w from a synchronic perspective and demonstrate in different ways that the polysemy is a result of an exten sion of the verb’s basic literal me aning, that of t he devel opment of a living thing towards maturit y. However, th e diachronic dev elopment of transi ti ve grow has not been studied in detail before. The rest of the the sis is organized in the following wa y : c h apter 2 will describe the theoretical background ne cessary for understanding the categorization of the corpus ins ta nces a s well a s the fin dings ; c hap ter 3 wil l present the method and material used in the study ; i n chapter 4, the categorization and the find ings of the study will be presented ; f i nally, c hapter 5 will con clude the thesis.

1

2 . Background

This pa rt will i nt roduce the con cept of transi tivity and discuss previous research on Engli sh labile verbs . T h is section will form the basis for understan ding the categorization of the corpus material . Transitivity refers to the abil ity of predicates to ta ke direct object s. A distinctio n is generally m ade between intr ansitive and transitive verbs: the former do not take objects, whereas the latter do. In En glish there are many verbs that can appear in both intransitive and tra nsitive verb patterns. The foll owing section de sc ri bing the ve rb patterns is b ased on McMillio n and the examples therein (n.d., p p . 30 - 35). The patterns are summarized in Table 1. Intr ansit ive verbs can be bare intransitive or copular. The bare intransitive pattern consists of a s ubject and a predicat or (i.e. a verb group ) . In this pattern t he subje ct either does or undergoes something, an d thus, the semantic role of the within this pattern can be that of an actor or that of an undergoer. The chair broke. The queen died. O scar coughed . The cop ul ar patte rn is seen as a sub - category of the int ransitive patter n. It consists of a subje ct, a predicator , and a subject predicative . The subj ect predicative is a complement that attaches , for example, a quality, characteristic , or location to t he subje ct. I n terms of s emanti cs, th is pattern often has the sen se of being or b ecoming. Harry is in the kitchen . Lucy became a doctor. Sh e grew anxious. Transitive verbs can be monotransitive, ditransitive, or complex transitive. The monotransit ive patte rn c onsists of a subjec t, a p redicator , and a di rect ob jec t . A direct obj ect is a complement in the form of a noun phrase or a subordinate claus e. The overall sense of this pattern is that the subject affects the direct object by d oing something to or with the object. Thus , the d irect object typica lly expres se s the patient role and can often take the place of the subject in a passive clause. Cha rlie despises him. They grew cabbages. I hope that Emily shows up . The ditransiti ve pattern includes a subj ect , a predicator, an i ndirec t object , and a dir ect obj ect. An indirect object often expresses the recipient role, takes the form of a noun ph rase , and is placed before the direct object. Semantically, the subject in a ditransi tive pattern ‘transfers’ t he di rect ob ject to the indi rect o bject .

2

We wrote h im a postcard . Lucy gave Oscar a new sweater . A complex transitive consists of a subje ct, a predicator, a direct object , and an object predicative . An object predicative a scribes a property to the d ir ect obj ect and is typ ically re aliz ed by a noun ph rase or an phrase. Semantically, this pattern often denotes the result of an event or ca used motion , or has the sense of viewing things in a certain way. They appointed her mana ging director . Ch arlie pu she d Harry off the bus .

T able 1. V erb pa tterns: the f irst two a re intransitive and the last three are transitive. Based on McMillion ( n.d. , p. 35 ). S= subject ; P=predicato r; sP=subject pre dicative; dO=direct object; iO=indirect object; oP= object p redicativ e . Intransitive S P Copul ar S P sP Mo not ransitive S P dO D itr ansitive S P iO dO Complex transitive S P dO oP

Adverbials can be a d ded to any tr a nsitivity pattern to e xpress circumstances related to the event described in the clause , for example , t i me, place, manner, conditio n or re ason; the spe ake r’s view on what th ey are saying; or a semantic link between sentences. Adverbials are c ommo nly realized b y preposition phrases, adv erb phrases, noun phrases , or subordinate clauses. Adverbials can function as either modifiers or complements , but t here is some d isagreem e nt as to where t o draw the li ne between the two . However, i n McMilli on ( n .d ., p. 30), the term ‘ adverbials ’ is reserved for modifiers and obligatory elements are referred to as complements . Following McMillion ( n.d.), this study uses t h e term ‘ a dverbial ’ to refer to optio nal el ements (underlined in the examples) that can be ad d ed t o any clause pattern wi thout changing t he basic structure of the clause : M rs . Jone s died in her bed last Sun da y . (i ntr ansitive) Unfortunately , s he was ill last week end . (copular) They grew cabbages in their garden . (mo notran sitive) However , she b ough t him a new sweater . (d itransitive) He pu shed h im off the bus furiously . (co mplex transi tive) Pr epo sition phrases are sometimes modifiers , as in M rs . Jone s died in her bed , but can in some cases be seen as complements. Qui rk, Greenbaum, Leech and Sva r tvik (1985) mention quasi - adjec tival complements in the form of preposition phrases, such as Thi s machine is out of date and This dr ess seems out of fashio n . They c an be cons idered adjectiv al for three reasons: they are semanti call y similar to ( out - o f - date / obsolete ), the y can be coordinated with adjectives ( They’re happy and in good

3 health ) and they are used as complements w ith other copular verbs besides be ( They s eem in good hea lth ) (p. 658). Many verbs in English can b e used both intransitive l y a n d transiti vely with a shift in terms of sense from inchoative to causative. The p attern has an inchoative sen se, b ecau se the subject is pr ese nted a s undergoing a chan ge by itse lf without an external agent or causer. In the transitive patt ern, the s ubject of the intransitive pattern i s placed in object position and now there is a causative sens e: the new subject is present ed as do ing something t o th e object . The se verbs are sometimes referred to as ergative verbs, “double - functioned or amph ibious verbs” (Visser, 1963, p. 97), ambitransitive ve rbs (Dixon, 1994, p. 18), et c., but the term ‘labile v erbs’ is used in this study, because it is deemed t o be the most neu tral and transparent ( McMillion, 2006, p. 7 - 10 ). McMillion (2006) defines l a bile verbs in the fol lowing way:

1. They can occur in both transitive and intransitive patterns. 2. T he subjects of the intrans itive patterns denote t he same kind of ent i ties as t he objects i n correspondi ng transitive patterns. 3. The transitive pattern s exp ress a causative event in which the transitive subject causes some chang e in the direct object. Th e causative subject is ext ernal to the af fected o bject and has t he s e mantic ro le of actor or agent, and , depending on the individual verb, can include an i mate beings, events, natural forces, and conditions. 4. The intransitive pat terns express events in w hich the sub jects, usually inanimate part icipants , are interpret able as self - c aus ing or se lf - affecting, or in which an external cause is irrelevant. (p. 1 )

Mc Millio n more over uses the term in a restrictive sense to refer to verbs that can be used in both transitivity patterns without being marked either morphologically or lexically , for example break (intransitive ) and break (transitive ) (p. 9). According to Ha spelmath (1993), a caus ative - pair is defined semantically. For example, the stick brok e (i n choative) an d the gir l broke the s tick (causative) describe the same basic situation, but whereas the caus ative clause includes an agent who ca u ses t he stick to break, the inchoative clause presents the situation as occurring by itself, even if there may b e an agen t i n the obj ective situat ion (p. 90). He furthermore points out that the condition for a verb to appear in an inchoative - causative alt e rnati on pa ir is that it does not con tain any “agent - ori ented meaning components or o ther hig hly specific me anin g componen ts that make the spontane ous occurrence of the event extremely unlikely” (p. 94). In other words, the verb must be seman tically free o f any refer ence to an agent. This is exemplified by mean s of cut and tear , where the former i ncludes the age nt - o r iented me ani ng compon ent ‘by means of a sharp instrument’, whereas the latter does not, which is why it ca n be used both transiti vely and intra n sitiv ely ( p. 93). Lability can be seen in different ways depending on the theoretical perspective. Most generative theories on lability posit different lexical entries for the inchoative and causative verb meanings, which dictate the argument structure pat terns associated with them, and formulate lexical rules to derive one from the other (e.g. Levin & Rappaport Hovav , 1994 , p. 47). More recent accounts from cognitive and constructionalist perspectives have criticized the redundancy of these multiple lexica l verb entries, and

4 focused on the meanings th at can result from the combination of a verb's meaning with the meanings carried by the constructions it combines with (e.g. Goldberg , 1995). E nglish d iffers f rom other languages , including other Germanic la ngu ages, in that it has so m any verbs th at can be used transit ively and intransitively just by alternating the word order (Haspelmath, 1993, p. 101). Languag es use various ways to ma rk the causative - i nchoativ e contrast ; they can , for example, u s e affixatio n, suppletion or particles, or c han ge the stem vo wel. Engl ish can mark the contrast in these ways, for example raise – rise and lay – lie show a change in the stem vowel, whereas ki ll – die and teach – learn ar e examples of suppletion ( McMillion, 2006 , p. 17). How eve r, many causative - inc hoa tive alterna tions are expr ess ed using labile verbs, which means that there is no formal difference between the causative an d the inchoative. According to McMillion (2006), t he re are at least 800 labile verbs i n PDE : 500 - 700 single - w or d ones and 200 - 300 m ult i - word o nes, that is , phra sal verbs (p. 7), whereas Old English (OE) had relatively few labile ver bs in comparison. Visser (1 963) was one of the firs t to describe this incr ease in labile ver bs from OE to PDE: “Historically, the mos t r emarkable fact is that, wh ereas in Old English t he number of double - faced or amphibious verbs was far inferior to that o f intransitive verbs, in Pr es. D. English amphibiou s verbs far outnumber t he i ntransitive ve rbs” (p. 99 ). In other words, he says tha t intrans itive v erbs were far more prev ale nt than labile verbs i n O E , whereas the reverse is true for Mod ern E nglish (ModE) . He illustra tes the transitivization process by ju xtaposing, o n the one h and, 223 exclus ively intransiti ve verbs in O E with 58 in Mo d E , and, on the other han d, 55 labile v erb s in O E with 506 in Mo d E (p. 98 - 127). These lists are certainly not exhaustive an d have been r eviewed and contested , for example , by van Geld eren (2011), b ut the numbers do give an idea of the ex te nt of the change that has taken p la ce from OE to PD E. Vi sser (196 3) mentions the loss of t he prefix ge - as one of the reasons for the increase in lab ile verbs. Th e prefix use d to be one of th e disting uishing factor s between transitive and intransitive ve rb s in OE : growan w as the intransit iv e verb meaning “ to sp rout, flo uri sh ” and gegrowan the t ransitive equivalent meaning “ to produce ” (p. 127). Wh en the prefix disa ppeared , the transitive and i ntra nsitive forms becam e identical. Another reason mentioned by V isser is the ambi guity in constru ct ions such as it was c rumpled , whe re the verb phrase can be seen as an intransitive one which expresses a resulti ng state or as a passive c onstruction of a t ransitive verb suggesti ng the operation o f an agent : it was crumpled [ by someo ne ] (p. 131). This poin t will be relevant to the discussion below. Vi sser mentions sev eral oth er factors tha t have in hi s view contributed to the increase in l abile verbs . Howeve r, they are onl y hypot h es es which need to be i nvestigated and tested . I t is q uite possible tha t it is because there are so many way s in which verbs can become la bil e that t he labilizatio n process has been so extensive . Although there is some research show in g that there is an in crease in lability from O E to ModE , diachronic aspects of l abile verbs ha ve overall received very little att ention. Kulikov and L avidas (2 014 ) argue that “[a] syst em atic treatment of [labili ty] from a diachronic pers pective is lackin g: the rise, developmen t, and decline of labile verbs remain on the periphery of t ypological r es earch. In many cases, we cannot exp lain why and how the lability eme rges and disappears” ( p. 872). In other words, th ere is a need

5 for research on the diachroni c aspects of lability , both in English and in other languages, in order to explain how verbs bec ome labile. The current study is thus a n attempt to tra ck the development of the verb grow f rom e xclusivel y i ntransit ive to labile.

3 . Method and m aterial

3.1 Corp us l inguistics A c orpus is a collection of texts from the real world. The wor d is nowada ys gen erally understood to refer to “a co llection of (1 ) mach ine - reada ble (2) aut hentic texts (including tr anscripts of spoken data) which is (3) sampled to be (4) repr esentative of a pa rticular language or language var iety” (McEnery, Xiao , & To no , 2006, p . 5; i talics in the o riginal) ; that is , a c orpus is not a ran dom collection of texts . There are different kinds of c orpora ; a corpus can be ge neral or specialized, written - only or spoken - only, and synchronic or diachronic . Re presenta tiveness is a crucial aspe ct in corpu s design, as the goal of a corpus is to be repr esentative of a language or language variety. The representativeness of a corpus is determi ned by bala nce and sampling. Balan ce has to do with t he ra nge of ge nre s or tex t categories i n cluded in the cor pus, and a balanced corpus usually includes a w ide rang e. Although corpus makers strive for balance in their corpus design, there is no way to rel iably measu re corpus balance, so t he notion of b alance re lie s on “in tuition and be st estimate s” ( McEnery et al., 2006, p. 16). Sampling, on the other hand, is associ ated with the notion that a corpus cannot cover everything within a given language and is always a sam ple of a larger populat io n . A samp le is considered represent ative when the finding s based on the sample can be generalized to the whole pop ulation (McEner y et al., 20 06, p. 19). There is some disagreemen t concerning whe ther corpus li ngu istics s hould be s ee n as an independent bra nch of linguistics or as a met hodology. I n this work , as in fo r example Lindquist (2009, p. 1) and McEnery et al. (2006, p. 7), it is vie wed as a methodology. Co rpus linguistics is n ot a bran ch of lingu istics in the same way as, for example, sociolinguisti cs is. While sociolinguistics studies the relation between lan guage and society, corpus linguistics does not reveal what is studied , but rather, how something is studied (Lindqu ist , 2 009, p. 1 ). It is a m ethodology th at can be use d t o study man y different areas of langu age, including aspects of sociolinguistics, s ynt ax, semantics , historical linguistics, etc. A lthough corpus lin guisti cs is often consider ed quantita tive, it is pos sible as well as desirabl e in many case s to combine q uan titative and qualitativ e elements in a corpus linguistic study. This study is one whi ch uses a mixed methods approach. The research questions of t his stud y are such that th ey must be examined qualitat ivel y; a purel y q uantita tive approach c annot answer q ues tions about how somethi ng, in this case transitive grow , has evolved. Quantitative findings will be presented, discussed, an d compared, but the emphas is i s on the qu alitative analysis of the co nte xt s in w hich transitive grow develope d.

6

3.2 The Corpus of L ate Modern English Text s T he Corpus of Late Modern English Texts, version 3.0 (henc efor th CLMET ; see De Smet , 2005 ; Diller , De Smet, & Tyrkkö , 2011) is a co llectio n o f public domai n te xts by Bri tis h autho rs published be tween 1710 - 192 0. V ersion 3.0 contains over 34 million words of ru nning text divided over three sub - corpora : there are 10 ,480 ,431 words in Period 1 (1710 - 1 780) , 11,285,587 in Period 2 (1 780 - 1850 ) , and 12,620,207 in Period 3 (1850 - 1920) . The corpus covers five genres: narrative fiction, narrati ve non - fiction, drama, l etters , and tre atise s , as well as a num ber of unclassifi ed t exts ; and it claims to be approxima tely genre - bal anced ( KU Leuven, n.d. ). The CLMET is used in this study, beca use it is t he most extensive corp us for British En glis h in th e L ate Modern English p eriod that wa s read ily accessible . This study examines and c ompares Period 1 and Period 3; Period 2 has been excluded due to a limited amount of space and time .

3. 2.1 Extra ction of data from th e c orpus The corpus material was extrac ted manually using Word smith 7.0 , a software used for finding words and p hrases in corpora . The search i ncluded the f ol l owing f orms : gro w , grows , grew , grown , and grow ing . Other, non - standard forms were not included in the search, because there were not ma ny in the corpus and th ey would not have added anything to the an alysis. For exampl e, the f o rm groweth ha d 4 and 3 hits in Pe riod 1 and 3, respe ctively. That is no sur pri se , because the suffi x - eth had mostly been replaced by - ( e ) s during the 17th century (OED, 2012). The form g rowed had no hits in P eriod 1 and only 11 in P e riod 3, which is a ver y low number in comparison to the frequency of the other forms (see Table 2) . The manual sear ch yielded in total 2 , 122 instances in P eriod 1 and 5 , 209 in P eriod 3 . Table 2 shows t he frequ enc ies of the form s that were included in the search . All t h e h its for each period w ere gathered into an Excel file, assigned a random number usi ng the RAND function , placed in ascend ing order based on the random number, a nd then a sample of 5 00 relevant l ines in each period was scruti nized .

Table 2. Absolute f req uencies of the f orms of grow i ncluded in the CLMET search . F orms of grow Period 1 Period 3 g row 606 980 g rows 274 406 g rew 576 1 , 620 g rown 401 1 , 16 2 g row in g 265 1 , 041 Total 2 , 12 2 5 , 20 9

All instances where grow could be understood as a ver b w ere con s idered relevant . Clea rly a djectival and nominal uses of grow , such as “he began to roar like a full - grown elephant” ( CLMET3_0_3_190.txt), “th e Eu ropeans have a growing asc endency over t he nations of Africa an d America” (CLME T3_0_1_52.txt), and “ He tells m e he [ … ] has taken to vine - growing” (CLMET3_0_3_283.txt), were excluded from further analysis , and a su ff icient number of lines was e xamined to come up to 500 relevan t insta nc e s in each period. Th is meant that 54 3 lines were examin ed for P er iod 1 and 566 for P eriod 3.

7

The categorization will be presented in the next chapter together with the results. T he a djectival and nominal instan ces will not be discusse d further .

4 . R es u lts

This part will e xp lain how the c orpus material w as categorized a nd pr esent the results of the categorization as well as the findings regarding the context s in which transitive grow d eveloped. The tables p resent the observed absolute frequenc y of each item in the sa mple ( N ) , the ir relative share in percent (%) , and the norma liz ed frequenc y p er million words of the estimat ed absolute frequenc y in the subcorp us (pm w) 1 to make it possible to compar e the two subcorpora and to see the diachronic chang es . When absolute fre quenc ies are discus s e d in t his study, they a lways refer to t he frequency in the s ample, not in the whole subcorpus.

4.1 Intransitive i nstances Period 1 and 3 have 459 and 407 one - word (i.e. not phrasal) intransitive instance s of grow , res pectively (Tab le 3). If ph rasal verbs like grow up ar e added, Perio d 1 has 4 95 intransitiv e insta nces, whe reas Period 3 has 478. T he freq uencies in t able 2 show that the overall use of grow has increased between the two periods, which is probably why intr ansitive use in g eneral has also incre ased. The on ly kind of int ra n sitiv e us e which has n ot increa sed i s copular use with noun phrase complement s.

Table 3. Frequencie s for c ategories of intransitive grow .

Period 1 Period 3 I ntransitive grow N % pmw N % pmw

Bare intransitive 93 20.3 34.7 12 3 30.2 8 9.7 Copular 359 78.2 133.9 280 6 8. 8 204.2 - A P c omplement 32 5 70.8 121.2 25 3 62.2 184.5 - NP complement 14 3.1 5.2 4 1 2.9 - PP complement 15 3.3 5.6 1 7 4.2 12.4 - AdvP complement 0 0 0 2 0.5 1.5 - Non - finit e complement 5 1.1 1.9 4 1 2.9 Ambiguous 7 1.5 2.6 4 1 2.9 Total 459 100 171.2 407 100 2 96.8

1 T he estimated absolute frequency i n the subcorpus was calculated by multiplying the absolute fr equency of the category in the sample by the total number of hits in the corpus and then dividing that amount by the total number of hits in the sample, e.g. for bare intransitives in Period 1: 93 x 2,122 / 543 ≈ 363.44. The normalized frequency was then calculated by dividing th e estimated absolute frequency by the size of the subcorpus (see section 3.2), and then multi plying that amount by a million , e.g. 363.44 / 10,480,431 x 1,000,000 ≈ 34.7.

8

4.1.1 Ba re i ntransitive i ns tances Ba re intra nsitive u ses are genera ll y relatively e asy to di stinguish from copular, phrasal , and transitive ones. There are 93 and 123 bare intransitive instances of grow in Period 1 and 3, respectively (Table 3). This category includes uses of gr ow t h at do not take a compleme n t but may be m od i fied by an adverbial . “All things ar e growing or decaying” (CLMET3_0_3_189.txt) i s an example of bare intransitive grow that is not modified by an adverbial, but cases where an adver bial is present are much m ore frequent . In the f ollowing case, th ere is an adve rbial of manner: “good Grass, which grows in littl e Tufts about as big as on e can hold in one's hand” (CLMET3_0_1_61.txt ), and here ther e is an adverbial of degr ee: “ Pi ne apples grow i n such a bundance th at they may be p u rchased, at the firs t hand, for the value of an English farthing” (CLMET3_0_1_29.txt). Th ese are classified as m odifiers and not as complements, because the y coul d be left out without changing the bas ic m e aning of grow .

4.1.2 Copu lar i nst ances Copular us e s of grow are very fre qu en t in both pe riods: Period 1 includes 359 instances of copular use, whereas Per iod 3 co ntains 280 (Table 3). These numbers ma ke copular u se by far the most frequent use of g row in th e sa mples. Be cause the CLM ET is quite he av y on ‘ literary ’ texts s uch as narrati ve fiction , it is likely that copular grow is overrepresen ted in the corpus vis - à - vis other corpora, because it seems th at the copul ar use of grow is more common in fiction than in ot he r gen res ( Biber, Johansson, Leech , Conrad, and Finegan , 19 99 , pp. 43 8 - 4 3 9). Semantically, c op ular grow usua lly denotes a gradual evolution into a state o r condition, although a sudden change may sometime s be impli ed (Visser, 19 63, p. 203). Copular grow characte ristically has a comple ment t hat either describes the subject o f t he clause or renames i t. I n the samp les, copular grow has five types of subject pr edicative complements: adjective phrases (AP), nou n phrases (NP), preposit ion phrases (PP) , phrases ( AdvP), and non - finite phrases. Many dictionarie s do n ot distinguish betwe en b are intran si t ive and copular mean in gs. The OED, f or example, does not d ifferentiate between the two, but category 12 seem s to be de dicated to copular grow :

To become or come to b e by degrees, sometimes with inclusion of the l iter al sense of increase o f ma gnitude or q u antity. a. with adje ct ive or (arch.) n. complement. b. wit h adverb or adjectival phrase formed with a pr epos ition. Now rare. (grow, n.d. - c)

Category 12.a. i s rather s elf - explanator y, whereas 12.b. is not clearly defi ne d , but it does see m s im i lar to quasi - adjecti va l compleme nts mentioned by Quirk et al . (1985). C opular grow most often takes a complem ent in the f orm of an AP: there are 325 and 253 in Period 1 and 3, res pectively (Ta ble 3). The AP complement of c op u lar gro w is o ften a sim ple ad j ective, such as dark , old , or po or , which is at t imes preceded by a modi fier, such as so , too , or quite . Compa ratives expr essed infl ectionally, such as “the noise grew l ouder and l ouder” (CLMET 3_0_1_48.txt), are rather freq ue n t in the samp les, where as t ho s e expressed periphra st ically, su ch as “The older I have grown, the more distinctly have I perceived that […]” (CLMET3_0_3_ 241.txt), also occur. Comparative adjectives ma ke an impli cit compariso n to an earlier state (Biber e t a l.,

9

19 99 , p. 445), and the fr e quency of such compl em ents is we ll aligned with t he inherent semantics o f copular grow that often expresses an increase in magnitude or quantity (grow, n.d. - c). The samp les also in clude compari sons com posed of as + adjectiv e + as : “I onl y b egin to be afraid th at it should grow as necessar y to shoot one's self here, as it is to go into the army in F rance” (CLMET3_0_1_44.txt). AP complem ents often consist of an adjective and a comp lem ent in turn, as in the afor em e ntioned examp le. NP com plem en t s are rather infrequ en t in the s amples: Period 1 has 14 NP complements, whereas Period 3 has only 4 (Table 3). This is the only category o f intransi tive grow which shows a decrease between the two periods. It is no surprise , t hough , since thi s use is n ow considered archaic ( gr ow, n.d. - c ). The NP compleme nt can be a simple one, as in “She is g rown a woman you see” (CLMET3_0_1_39.txt); or a mo re complex one, as in “Lowriver is a pleasant, genteel lit tle village, which […] has bee n growing, f or t he las t se ve n years, with each su cc eeding sum mer, more and more a place of favourite resort with the i nhabitants of Babel” (CLMET3_0_3_332.txt). This ha s been cat egorized as an NP com plement, because the ele men ts between the copular ver b an d the comple ment noun can be considered adverbial . Sometimes the distinction be tween a subject predicative and a direc t object is not clear - cut. For example, in the cas e of “ life [ … ] grew the object of my abhorrence ” ( CLMET3_0 _1_23.txt ) , it may not be immediately obvious that the NP fol lowing grow funct ions as a subject predicative of copular grow and not as a d irect obje ct of transitive grow . The data show that the use of copular grow with NP complements decreased around the s ame time as tra nsitiv e gr ow appeared in the acti ve . Co nsidering that dir ect objects are almost always NPs, it would have been easier to confou nd copular and transitive grow as long as NPs coul d equally well function as subject predicatives and d irec t objects. T he t hird k ind of complement that copu la r grow is found with is a PP . There are 15 PP complements in Period 1 and 1 7 in Per iod 3 (Table 3) . Not all grammarians agree that a copular verb ca n take a c omplement in t he form of a PP and some would ca t egorize th e se in stances a s b are intransitive. Fo r examp le, Biber et al. (1999) consider that some verbs, such as be , get , happen , and remain , can ta ke ‘ obligatory adverbials ’ , whereas grow , among ot hers, cann ot (pp. 142 - 14 3). Overall, grammars and dictiona r ies do not o ffer much help i n categorizing grow f ol lowed by a PP, because they generally contain examples of prototypical use and not so many examp les of exceptional or atypical use. However, in th is study a ny essential e lement, regardless of its grammati c al form, without which th e sen se of the clause is inc omplete has been classified as a complement following Huddleston & Pullum ( 2002 , p. 221 ) . This means that a PP has been classified as a compleme nt of copu lar grow in in stances where the PP functions as a subject p redicativ e complemen t desc ribing th e s ubject and/or where om ittin g the PP would change t he meaning of grow and the entire clause . Optiona l elem ents are considered mo difiers that can be re moved with out a change i n meaning. This distinction is imp ortant for the categ orization, b ecaus e it mean s t hat instances such a s “The hair of the people […] grows in great abundance” (CLMET3_0_1_29.txt) hav e been categorized as bare i ntransitive, whereas i nstances s uc h as “You mu st have passed a lamentable scene of anxiety; [… ] but I think we g ro w in s pirits ag ai n” (CLMET3_0_1_44.tx t ) have been classified as c opular, although grow is in both instances follow ed by a P P headed by the pre position in .

10

Grow has a basic me aning which de notes an increase in size, quantit y or degree or devel o ping toward s maturit y (gr o w, n .d. - d; Qing - Hua, 201 1 ; Alam, 2003). In instances such as “she had grown in all the secrets of all urement” (CLMET3 _0_3_244.txt ), omitting the PP wou ld entail a crucial chan ge in the meaning of grow and thus in the meaning of t h e whole cla use: she had g r ow n would mean that the p erson had matu red or increa sed in size, but in this instance the person has developed in a v ery particul ar way, namely in all the secret s of alluremen t (cf. alluring / knowledgeable in a ll the secrets of al l urement ). I n fact, m any o f t he PPs in the se instan c es could be re placed by an AP and retain their meaning. Thus, many of the in stances can be s een as quasi - adjectival, following Quirk et al. (1985). Fo r example , in “he is grown of less consequence” (CLMET3_0_1_44 . txt), the P P could b e rep l ac ed by the AP less impo r tant , and in “ my father and mother, growing in years” (CLMET3_0_1_7. txt), the PP could be re place d by older without a semantic c hange (although there is a stylistic change). Ev en in cases where grow has t h e basic mea ning of i ncrea s in g or developing toward s maturity, the re may be a P P which emphasizes the result of growing , as in “it wi ll g row to a gigantic siz e” (CLMET3_0_3_244.txt ), where the meaning of the PP could be expressed by the AP gigan t ic in size . Some of the P P s ar e not easily replace d by APs, as in “the Speculat ions about Infinites have run so high, and grown to such strange Not ions” (CLMET3_0_1_1.t xt), but the PP is, ne ver theless, an essential part of the clause and cannot be left out without cha nging the mean i ng o f grow . The fourth kind of complement of copular grow in the samples is a n AdvP. T here are no AdvP complements in Period 1 and only 2 in Period 3 (Table 3) : “ H ow thoroughly I've grown away from t hem — ceased to be one of them ” (CLMET3_0_3_265.txt) and “ Ernest and I had been growing apart for some time ” (CLMET3_0_3_228.txt) . They are similar to PP compl eme nts and have been categorized as copul a s for the same reason , namely that omitting the AdvP woul d result in a change in meaning. Another cate g ory of copular g row found i n the samples is made up of the verb with a non - finite complement . T here are relatively few instances in this category: 5 in Period 1 and 4 in Peri od 3 (Table 3 ). If phrasal ver b s wit h a no n - finite compl e me nt are added to these c ounts, there a re 6 in each period. The non - finite complements of grow are all to - infinitival , s uch as “I feel very often that I grow to correct twenty t hings in myself” (CLME T3_0_1_44.txt ) and “he grew to be an afflu ent trade sman” (C LM ET3_0_3_332.txt). Th e se instances a re difficult to label, because grow used in this way comes close to functionin g a s an . Some dictionaries, such as the OED, Oxford Learner’s Dicti onaries, and Collins English D i ction ary, m ention co nstru c ti on s of grow with the t o - , w hich the OED considers rare (grow, n.d. - c). This use has the sense “to gradual ly begin to do somethi ng” (grow, n.d. - d). Collins English Di ctionary is the only o ne out of the three dictionari e s tha t clas sifies th is us e a s copular alongside AP complements:

Y ou use grow t o say that someone or something gradually changes until they have a new quality, feelin g, or attitude. I grew a little afrai d of the guy next door . [VERB adjec tive] He's growin g old. [VERB adjectiv e] He gr ew to love his work. [ V ERB to - infinit ive] (grow, n .d. - b)

11

The other two dictionaries do not take a stance and classi fy this use as merely intransitive without specifying the ty pe. However, if we con sider that th ere are two intra n sitiv e verb pattern s, the ba re intransitive which i s characterize d by not havi ng a complement and the copular which is characterized by having a s ubject predicative complement, grow followed by a non - fin ite phrase comes close r to the latt er .

4.1.3 Ambigu ou s in transitive i nstance s Sometimes it is h ard to distinguish b e tween bare int ransitive and copular uses. In unclear cases it is necessar y to delv e into the semant ic s of the context in which the verb ap p ears more in detail. C ases of grow followed by a wor d that can be either an a djective or an ad ve rb are particularly d ifficult to ca tegorize, because such words can function either as subject predicat ive complements o r as modifiers. Like , wild , low and hig h are a few examples of words appear ing in the sample s that cause difficulties due to their dou bl e - function. “What ha v e you been doi ng to grow so lik e Murillo's Madonna?” (CLMET3_0_3_272.txt) has been categorized as a c opula r instance , because so like Muril lo’s Madonna se ems to function as a compl emen t and denote the result of becoming , where as “they saw w ond er growing like a flow e r inside” (CLM ET3_0_3_306.txt) has been categorized as bare intransitive, because like a fl ower seems to function as a modifier a nd denote the m anner i n which wonder is g rowi ng. However, “One sort, which is by far th e most Numerou s s or t of any in the Wood s , grow Somethi ng like birch” (C LMET3_0_1_61.txt) has been placed in the ambiguous category, because it was deemed impossible to t ell from the co ntext i f the intended mean ing is that this type of tree grows to be like birch or in a ma nn er similar to birch. Using the same strategy as abov e, “The sky had grown wilder since he stood there l ast hour” (CLMET3_0_3_319.txt) has been classif ied as a , since wilder functions a s a complement th at i ndicates the resul t of be coming, wherea s i n “There are, indeed, g rowing wild in the wood a few s orts of Fruit” (CLMET3_0_1_61.txt), the adverbial m eaning of wild , t hat is, in a wild manner seems to be the more rele va nt one, so it has been c lassified as bare int ransitive. Low and high a re more diffic ult t o categorize, becaus e it is often u nclear whether the word denotes the result or the place of grow ing, as i n “As this button ripens and withers, others grow lower and lower every year ” (CLME T3_0_1_88.txt ) and “Eig hty Ears of Wheat have be en found to sp rin g from one Root, but g r ows not very h igh” (CLMET3_0_1_83.txt). In these cases, the context does not reveal t he intended meaning conclusively, so t hey have been classifi ed as ambiguous. Ambigu ity and confu sion can a lso be caused by the omis sion of - ly in ad ve rbs that would in st a ndard PDE end in this suffix. The practice of omitting - ly from was quite com mon in Earl y Modern English (OED, 201 2 ) and it is even done in colloquial an d non - s tandard PDE ( Hughes & T rudgill, 1987, p. 20), so it is likely tha t it w as also done in t he Late Modern English period in question. There are some instances in the s amples wh ere an adje ctive may have been used a s an adverb. “Ginger gr ows plentiful he re” (CL MET3_0_1_83.t xt) and “C ommerce, in the pride of her prosperity , g ro ws n ice about her ro a ds” (CLMET3_0_ 3_332.txt) are two examples where it is unclear if the word fo llowing g row should be seen as a complement or as a modifier. The lat ter

12 interpretati on seem s slightly mo re relevan t, but these have, nevert heless, been p lac ed in the ambiguous ca t egory. Verse is particularly difficult to interpret, because it often uses non - stand ard word or der, metaphors and vague l a nguage. For example, i n “Shou'd there a Flow' r be plac'd f or me, To nourish, as it blooming g rew ” (CLMET3_0 _1_ 85.txt), it is not cle ar if blooming denotes the result or the manner of growing. Another example o f v erse t hat is difficult to interpret is this: “To Covent - Garden 'ti s sent (as yet swe et), There fades, and shrink s, and grows past all end uring” (CLMET3 _0_ 1_3.txt). Here, past a ll enduring c ou ld be a complement in th e form of an AP or a PP or a modifier, which is why it has been categorized as ambigu ous. Overall, Period 1 has 7 ins tances t hat h ave been cate gor ized as ambiguous, whereas Perio d 3 has 4. Mos t o f them consist of grow followed by a word that could be eithe r adjectival or adverbial, but some of them are like the latter and cannot be interpr eted without engaging in poetry analysis .

4. 1.4 Phrasal v erb s There are 36 and 71 phrasal ve rbs in Period 1 a nd 3, respectively. Th e normalized fr equencies indicate that phrasal verbs consisting of grow and a particle have increased almost fourfold , mostl y thanks to grow up . T his is an interesting d evelopment , a nd althoug h it may be connected to the labilizati on o f grow , it would requ ire much more t ime and space t han is currently available to investigate .

T able 4 . Fre quencies for c ategories of phrasal gro w .

Peri od 1 Per iod 3 P hra sal v erbs N % pmw N % pm w g row up 22 61.1 8.2 45 63.4 32.8 grow upon 7 19.4 2.6 4 5.6 2.9 grow on 1 2.8 0.4 3 4.2 2.2 grow into 6 16.7 2.2 1 1 15.5 8.0 grow to 0 0 0 2 2.8 1.5 grow out of 0 0 0 6 8.5 4.4 Total 36 100 13.4 71 100 5 1.8

The samples include six different phrasal verbs: grow up , grow upon , grow on , grow int o , grow to , and grow out of . The number s for each phrasal verb ca n be se en in T able 4. Although t hey are all int r ansitive , t hey have been kept separate from other intrans itiv e uses of grow in the categorization . Most instances of grow up are bare intransit ive, such as “Miss grew up under the i nfluences of such a direct ress” ( C LME T3_0_1_9.txt), but som e of t hem seem more like copulas, for example “their childre n will grow u p se lfish in the v ery core of their souls” (CLM ET3_0_3_216.txt). All the instances of grow into are copular, such as “The post at t he confluence was gr aduall y growi n g i nto the great c amp of a few months la ter” (CLMET3_0_3_312.txt). One instanc e of grow to seem s to also have the sam e copular meaning as grow into . It is no surprise that there are no in stances of transitive grow int o of the t ype grow + NP +

13 into + NP, be cau s e this use is more co mplex and p resu mably developed after the simple monot ransitive gro w . A ll the other phrasa l v erbs in the samples a re bare intransitive. There are at leas t three di fferent meanings of grow out o f in the s amples, and onl y one type has be e n placed in th e categ ory of phra sal verbs. Instances such as “it grows out of fashion l ike a fardingale” (CLME T3_ 0_1_44.txt), where it refers to the addressee’s humour, have been clas sified as copular due to their quasi - adj ectival nature, whereas “h e h ad a candle growin g out o f the top o f hi s head” CLMET3_0_3_281.txt) has been c ategorized as bar e intransitive, bec aus e the adverb phrase s eems to function as a spatial adverbial . Instance s such as “there have grown ou t of the a malgamate langu ages, the ‘ ana pto t ic’ languages” (CLMET 3_0_3_189.t xt) have been classified as phrasal though , because the y ha ve acquired an idio mat ic meaning “to develo p or come into existence from” (grow, n .d. - a).

4 .2 Transitive i nstances There are not many t ransitiv e i nstances in the sample s, and th ey are all monotransiti ve. Period 1 includes 5 instances of grow that can be understood as transitive an d they are all passive, whereas Period 3 in cludes 22 po ssibly transitive instan ces, of whi ch 10 are pass ive an d 12 are ac tive (Table 5). The nu mber s in c lude e ven the ambiguous cases .

Table 5. Frequencies fo r c a tegories of t ransitive grow .

Period 1 Pe riod 3 Transitive grow N % pmw N % pmw

Passiv e 5 100 1.9 10 45.5 7 .3 Active 0 0 0 12 54.5 8.8 Total 5 100 1 .9 22 100 16.0

Two o f t he transitive in stances in Period 1 ar e unambiguous a nd c annot be understood as anything but passive forms of transitive grow : “Cloves are now grown no where but on the island o f Amboina ” (CLMET3_0_ 1_29.txt) and “ unless mor e corn is either usually g rown , or usually imp ort ed into the coun try” ( CLMET3_0_1_51.t xt). These can both easily be turne d into active phrases where the subject is semantically understood as the causer of the e ve nt (e.g . they grow cloves nowhere but on the island ). Instances such a s “o ne of the towers of an old ruined a bby, g rown over with ivy” (CLMET3_0_1_23.txt) an d “thy p ath from thy door to t hy bowling - green shall never be grown up” (CLMET3_0_1_36.txt) are mor e p robl ematic though , because they cannot be tu rned into active p hras es. T hey also r esist the addition of a causative a gent by someone but yiel d to the incho ative test by itself (McMillion, 2006, p. 5). T hey have been ca tegorized as transitive , because the OED li sts this use as trans itive with the fo llowing descri pti on : “ passive . Of land, etc.: To be covere d with a growth of somethi ng. A lso with over . So †to b e grown about (i.e. surr ounded by a growth), to be grown up (i. e. crowded with a growth),” and adds that “ [t]hese uses seem to have arise n partl y from the ind ire ct passive o f phrases like to gr ow over , and partly from the intran sitiv e perfect conjugated wi th be ” (grow, n.d. - c). In other words, to be grown over

14 or up b oth seem to derive f rom the intr ansitive pe rfect us ing the be - auxiliary. T here is also one inst anc e which is a mbiguous b ecause of t he auxiliary used : “They breed wit h me excessively, and are gr own to the size o f small perch” (CLMET3_0_1 _44.txt), where the word they refers to gold fish. This could b e unders tood either as a passive trans itive with an ex t ernal causer : they are grown by som e one ; or as an intransitive perf ect w ith an inchoative sense : they have grown by t he m selv es . M ost of t he transi tive instan ces in Period 3 are un ambiguous and this co ncerns b oth the passive and the active ones. Period 3 in clude s tran sitive inst ances suc h as “Now a good deal of dhourra is grow n” (CLMET3 _0_3_332.txt) and “He has grown the whi te moss rose” (CL MET3_0_3_202.txt). T he instance “they sat down to […] the sa lad grown a nd gathered by their host's own hands” (CLMET 3_0 _3_2 38.txt) is interes ting, bec ause although it is pa ssive, the ‘c auser ’ is mad e explicit with the use of an agent ive by - phrase which w ould become the subjec t in a corresponding active phrase: the sal ad which their host’s own hands had gro wn and gathere d . The re is al so one inst ance of p assive transitive grown over and on e of grown up used in the sa me way as in Period 1 . Period 3 also incl ude s one ambiguous ins tance that ca n be seen either as transitive or as i ntransitive: “A little ma n with a pink face a nd large red e ars was sit ting in a fat pink chair, as if he had been grown there” (CLMET3_0_3_3 01 .txt). If it is seen as transitive, it seems that the little ma n i s likened to a plant and ha s been grown by someo ne , and if it is seen as intran sitive, grown fun ctions as a n adjectiv al compleme nt to the copula be . There is clea rly an overlap between intr ansitive perfects and passive transitives , which are both constr uct ed u sing t he auxil iar y be and the pa st part icipl e grown . Overall, i t does not seem to be a coincidence tha t th e instances in Pe riod 1 that coul d be understood as tran sitive are all passive , because the passive is simila r to the intransitive in that the patient , or the ‘caus ee’ , of an active tra nsit ive p hrase takes the place of the subject in a pass ive phrase a nd the agent , or th e ‘causer’ , loses its subje ct status (Has pelmath and Müller - Bard ey , 2 001, p. 5). It also does not seem to be a coincidence that the fir st instances t hat can be seen as tr ansi tive are ambiguous or b elong to the grown over/up category that is semantically clos er to intransi tive verb s .

4.3 Instance s with a ux i liary v e rbs ( have / be ) In the Late Modern Englis h period, there exist ed two competing auxiliarie s to m ark the perfect : be and have . While transitive perf ects in the active voice always s elect have as the auxilia ry and transit ive ver bs in the pass ive voice always sele ct be , intransitive perfect s could and can to some extent still be formed using e ither auxi liary. The data i n this st udy show a shift fro m the auxiliary be to have in pe rfect constructions fro m Pe riod 1 to 3 (T abl e 6). While be is used in 49 instances i n Per iod 1, it is used in only 1 3 in Period 3; and where as have is used in only 15 instance s in Period 1, it is used in 63 in Period 3. T he normalized frequencies indica te that t he use of be has h alv ed while the u se of have has increa sed eightfold between the t wo per iods , 1710 - 1780 and 1850 - 1920 . The findings ar e aligned with previous studies sh owing that the have - perfec t prevailed over the be - perfect between 1750 - 1800 (Kytö, 1997 , pp. 3 2 - 33 ). Los (20 15) ar gues that the hav e - per fe cts devel oped from passiv e trans itive exp ressions, such as the enemy has been bound [ by someone ] , whereas be - perfects d eveloped from

15 resul tatives of in transitive verbs ; change - of - state verbs in particular (p. 74) . She suggests that the be - per fect i s “an ex ten sion o f a construction with t he copula be ” (p. 7 4). T here are instances of grow in the samples which are perhaps bes t understood as com plements of t he copula be : “A little ma n […] was sitting in a fat p ink chair, as if he had been grown th ere” (CLMET3 _0 _3_ 301.tx t) discussed abo ve and “she h ad managed the housekeeping without a servant since Christiane had been grown up” (CLMET3_0_3_ 332.tx t) , although d ue to their form, t hey could also be understood as pa st perfect passives which would make them transit iv e. I n the latter case , a transit ive readi ng does not seem possible , though, since the phrasal verb grow up cannot be used transitive ly , bu t i n the f ormer case, either an intra nsitive or a transitive interp ret ation is possible . Either in terpretation is al so possible in instances suc h as “ They [ …] ar e grown to the size o f small perch” (CLMET3_0_1 _44.txt) discussed a bove. The data show th at the use of the be - auxiliary decrease d by 50 % f rom Period 1 t o Period 3 . In addition , those intransitive instances in P eriod 3 tha t use the be - auxiliary are alm os t all of the type th at could even in PDE be c o ns tructed using the be - a uxiliary , because the y are unambiguous , such as “ Why, Cousin Rut h, you are grown ” (CLMET3_0_3_ 2 0 6 .tx t) . It therefore seem s that have - perfects r eplaced be - perfects in cas es where be could lead to confusion between intransitive and tr ansi tive uses of grow . This finding is in line with Visser ’ s (1963) hypothesis that the change from be - to have - perfec ts played a part in the process of labilization .

Tabl e 6. Fr equencies for a uxiliary verb s with in transitive perfe ct grown .

Period 1 Period 3 Auxiliary verb N % pm w N % pmw

Be 49 64.5 18.3 13 15.7 9.5 Have 15 19.7 5.6 63 75.9 45.9 No au x iliary 11 14.5 4.1 7 8.4 5.1 Uncl ear 1 1.3 0.4 0 0 0 Total 76 100 28.3 83 100 60.5

5 . Conc lusion

The study set out to investigate the contexts in which the verb grow developed a transitive use in British English in the Late Modern English period. Durin g the inve stigation it became clear that t he categoriza tion of labile verbs l ike grow is complicated, bec ause in many cases there could be two possible categories. The distincti on is particul arl y difficult to make be twee n three pa irs that have be en identif ied as ambiguous: bare intran sitives and copulas; copul as an d transitives; an d intransitive perfects and pas sive tr ansitives. In the first case, the ambiguity arises f rom words and ph ra ses that could be interpreted ei ther as mod ifiers or as complements, de pending on whether they express dispe nsable information rega rdi ng, for exampl e, the manner or place of growi ng (bar e intransitive), or indispensable information regard ing the result o f growing (copul a). This ambigu ity is not crucial though, because they ar e both

16 intransitive and thus inchoati ve . In the second case, the a mbiguity s t ems from NPs wh ich could be interpreted either as subj ect predicatives of copula r grow or as direct object s of transitive gro w . The data indicate tha t the use of copular grow with an NP complement decreased from Period 1 to Period 3, as tr ansitive grow was em erg ing . It is possible that NP complemen ts of copular grow disappe ared because they co uld in some cases be mis i nterpreted as dir ect objects of transit ive grow . In the third cas e, the ambiguity is caused by the au xiliary be : intrans it ive perfects cou ld be formed using the auxiliary be and a past participle , while passive transitives are form ed the same way . This ove rlap renders c ert ain instance s highly a mbiguous and means that th ey can be interpreted as transitive and thu s causative or as intransitive an d thus inchoative . This ambiguity appears to have been cruc ial enough to have led to a c hange in the auxili a ry . The data show t hat over the Late Moder n E nglish pe riod, there w as a shift from be to h ave as the perfect auxiliary of intra nsi ti ve grow ; while be was the most common choice in Period 1, have had ta ken ov er by Period 3, at the sam e time as transitive g row was increasin g. Th e have - perfect may have r eplac ed the be - perfect to avoid ambiguity betw een int ransitive and transitive gro w , wh ich seems to provide so me preliminary evidence, a l beit limited to this one verb only , for Visser ’ s (1963) hy pothesis about the role of the auxiliary in the acquisition of labili zati on . The study ma kes no claims a bout verbs other than grow . The sample for P er iod 1 only includes passive f orms of transitive gr ow . Passive t ransitives are inherently closer to intransiti ves than activ e t ra nsi tives are, becaus e in t he passive voice the agent becomes irrelevant and th e object takes the place of the subject. Th ree o ut of the five transitive insta nces in Period 1 are al so semantic ally inchoative and cannot be turned into acti ve phrases . They ar e, nevertheless , classified as trans itive in line with the O ED, bec ause only transitives can occur in the passive voice. Activ e forms of transitiv e grow appear only af ter Period 1, as there are already 12 o f them in Period 3. The fact that the re are only p assive fo rms of transitive grow in P eriod 1 suggests that th e trans itivization of grow took place via the passive which c an be s een as a category in between prototy pic al intransiti ve and transitive use s . The se are the ways in which the data in dicate that g r ow has d eveloped a transitive use i n British English in the Late Modern English period . The findings can be seen as ten t ative a nd would ideally be confirmed in fut ure studies . It would be interesting to exten d the study to in clude all occu rre nces of grow in all th ree periods of the CLMET , t hen replicate the study using other corpora with texts from ar ound the same period , a nd ev en tu ally extend the study to other lab ile verbs.

17

Reference s

Alam, Y. S. (2003). The organization of the lexicon: T he polysemy of grow and d isambiguation. Proce edings of th e ACL 2003 Workshop on the Lexico n and Figurative La nguage , 49 - 54 . doi : 10.3115 /1118 9 7 5.1118982 Biber, D., Johansson, S., Leech , G., Conrad, S., & Finegan, E. (1999) . Longma n g ramma r of s poken an d w ri tt e n English . Harlow: Pea rson . De Smet, H . (2005). A C orpus of Lat e Modern English Texts . ICAME J ournal 29 , 69 - 82. Retrieved November 2 0, 201 8, from http://clu.uni .no/icame/ij29/ij29 - page69 - 82.pdf Dill er, H . , De Smet, H ., & Tyrkkö, J . (201 1). A Europ ean d atabase o f desc riptors of English electronic te xts. The E urop ea n English Messenger , 19 (2 ) , 21 - 35. Retrieved November 20, 2018, from http s://co re .ac.uk/download/pdf/34490684.pdf Dixon, R. M. W. (1994). E rgativi ty . Ca mbridge: Cambridge University Press. Fe llbaum, C. ( 1990) . as a semantic net. International Journal of Lexicog raphy , 3 ( 4 ) , 278 – 3 01 . https://doi.org/10.1093/ijl/3.4.278 Go ldberg, A . ( 1995 ) . Constructions: A construction grammar approach to argument structure . Chicago: Chicago University Press. Grow. (n.d. - a). In American Heritage Dictionary of the Englis h L a nguage . Retrieved Nov em ber 20, 2018 , from https ://ah dictionar y.com/ w ord/search.html?q=grow Grow. (n. d. - b) . In Collins English Dictionary . Retrieved Nov ember 20, 2018, from w ww .co llinsdic tiona ry.com/dictionary/english/grow Grow. (n.d. - c). In Oxford Engl is h Dictionary . Retrieved N ovember 20, 2 018, from w ww.oe d.com Gro w. (n. d . - d). In Oxford Learner’s Dictio nar y . Retrieved November 20, 2018, fr om https://www .oxfordlearnersdiction ar ies .com/def initi on/english/grow Haspelmath, M. (199 3) More on the typology o f i n choative/causative ve rb alternation s . I n B. Com rie & M . Polins ky (E ds .), and t ransitivity (pp. 87 – 120). Philadelphia : Benjamin s . Haspelmath, M. , & Müller - Bardey, T. (200 4). Vale nce c hange. In G. E. Booij, C. Lehmann , & J. Mugda n ( E ds.), Morp ho l o gi e: ein i nternationa l es Handbuch zur Flexion u n d Wo rtb ildung (pp. 1130 - 1145 ). Berl in: De Gr uyter Mouton . Huddleston, R. , & Pullum, G. K. (200 2). The Cambridge g rammar of t h e Engli sh l anguage . Cambridge: Cam bridge University Press. Hughes , A. , & Trudgi ll , P. (1987). English a c cents and d ialects: An i nt roduc tion to s ocial and r egional v arieties of Br itish English ( 2 nd ed .) . London: Edward Ar nol d. KU Leuven . (n.d.). Corpus p age . Ret rieved Novembe r 20, 2018, from https://perswww.kuleuven.be/~u0044428/clmet3_0. htm Kuli ko v , L. , & Lavidas, N. (2 014). I ntroduction. Lingui stic s, 5 2 (4), 871 - 877. https://doi.org/10.1515 /lin g - 2014 - 0010 Kytö, M . (1997). Be/have + p ast participle: the choi ce of t he auxi l iary with intransitives from Late Middle to Modern English. In M . Ri ssanen, M . Kytö , & K . Heik konen ( E ds. ) , English in t ransiti on: C orp us - based studies in linguistic varia tion and genre styles (pp . 19 – 85 ) . Berlin: De Gruyter Mouton . Levin, B., & Ra ppaport Hovav, M . (1994). A preliminary analysis of c au sative verbs in English. Lingua 92 (C) , 35 - 77. doi : 10.1016/0024 - 3841(94)90337 - 9

18

Lindquis t , H. (2 009). En glish l anguage: C orpus l inguist ics and the d escription of Eng lish . Edinburg h: Edinbur gh Unive rsity P ress. Los, B. (2015 ). Hist orical s yn tax of English . Edinburgh: Edi nburgh Univers it y Press . McEnery , T., X iao, R. , & Tono, Y. (2006). Corp us - b ased l a nguage s tudies: A n a dv anced r esource b ook . Lon don : Routl edge. McMillion, A. (n.d.). A v ery s hort i ntroduc tion to English s yntax . Hande d out to English I students at the Engli sh Department, S t ockholm University, in September 2014. McMill ion, A. (20 0 6 ). Labile v er bs in E nglish : Their m eaning, b ehavior and s tructu re . Stockholm: Enge lska insti tutionen. Ox ford Eng l ish Diction ary. (2012 ). Gram ma r in Early Modern English . Retrieved N ovember 20, 20 18 , from h ttps://publi c.o ed.com/blog /grammar - in - early - moder n - english/ Pick ering, M. J. , & Frisson, S. (2001). Proce ssing ambiguous ver bs: Evide nce from e ye movements. Journal of Experimental Psyc holo gy: Learning, Memory, and Cognition, 27 (2), 556 - 573. doi: 10 . 1037/0278 - 7393.27.2.55 6 Qing - Hua, L. (2 011). The co gnitiv e appro ach to the me anings of “grow”. US - China Foreign Lang uage, 9 (9), 575 - 580. Qui rk, R., Gr eenbaum, S., Leech, G . , & Svartvik, J. (19 85) . A c omp reh ensive g rammar of the English l an guage . Lond on: L o ngman . Van Gelderen, E. (2011). Valency changes in the h i story o f Engl is h. Journal of Historical Linguistics, 1 (1), 106 - 143. doi: 10.1075/j hl.1.1.05v an Visser, F. (1 963). An h is torical s yntax of the Engl ish l anguage. Volume 1: Syntactical u ni ts with o n e v erb . Lei den : Brill.

19

Stockholms unive r sitet 10 6 91 S tockholm Telef on: 08 – 1 6 2 0 00 www. su .s e