Postpositives in English: in search of adjectives available
Abstract
137
Full text
Postpositives in English: in search of adjectives available Pavlína Šaldová (Charles University, Prague) ABSTRACT Single postpositive adjectives as aminor type of noun postmodification in English are surveyed using acorpus sample to assess their retrievability, to provide an overview of the morphematic types, and their register distribution. Applying an ancillary test of positional mobility to characterize the postpositive occurrences, four broad groups are delimited. Postmodification by a-adjectives and adjectives in terminological compounds is infrequent, whereas the majority of the sample is constituted by adjectives in -able/-ible and in -ed, whose post-head position involves changes in semantic meaning, as well as occurrences which cannot be accounted for by the constraints stipulated in grammars. The frequency and patterning of available, responsible or necessary call for are-evaluation of the function of postposition, encompassing all frequent forms and taking into account reference and other textual factors. KEYWORDS adjectives, corpus, English, noun phrase, postposition DOI https://doi.org/10.14712/18059635.2021.2.2 1 INTRODUCTION Single postpositive adjectives in English are treated, in synchronic grammars, as an exception to the basic syntactic rule that adjectives precede nouns (ablack swan ~ *aswan black), and the postpositive use is described as subject to severe restrictions, such as the use of the superlative (the best result possible) or the temporal semantics of the head noun (years past) (Huddleston & Pullum 2002: 445, 552). Fischer characterizes adjectival postposition in Present-Day English as “amixed bag of remnants, with very little left in the way of general rules” (2006: 254), which leaves anumber of questions unanswered as to the interpretative status of the position (iconicity), the degree of positional and functional variation (where both preand post-head positions are attested), or the representation of adjective types. In literature, single postpositive adjectives have received varying degrees of attention focussing primarily on two (related) issues: the theoretical status of the postpositive placement and its relation to the attributive and predicative uses, i.e. whether pre-nominal adjectives are syntactically related to post-nominal adjectives, whether they are “both related to the postcopular adjectives ... [or whether they] are
138 LINGUISTICA PRAGENSIA 2/2021 simply variants of each other” (Alexiadou 2014: 90);1 and the interpretative effects of the position.2 Although an adjective should, by default, be able to function in any of the attributive (any intelligent person), postpositive (anyone intelligent), and predicative (anyone who is intelligent) functions, many items are restricted (absolutely, in agiven sense, or in various adjectival constructions) and classified accordingly into attributive-only and never-attributive adjectives (with asmall subset of postpositive-only adjectives) (Huddleston & Pullum 2002: 553).3 Given the intrinsic complexity combining the syntactic, semantic and interpretative aspects of the three positions (and functions) with the fact that not all adjectives have the same range of functions, it is felt that an approach to English postpositive adjectives using corpus data may stimulate afurther inquiry into the matter, as in most accounts (stemming from approaches which rely on constructed examples) relatively few, identical and decontextualized examples are used throughout. Acorpusbased survey may highlight some aspects which would otherwise remain unnoticed. This study thus takes an exploratory corpus-based approach to probe into aspects of frequency, productivity, heterogeneity and, to alimited extent, variability of the position of English adjectives appearing in postposition. It aims 1.to tackle the question of searchability of the construction in acorpus (Section 3); 2.to survey the range and representation of adjectives attested as single postpositives in the sample (Section 4.1), and the range of written registers in which they appear (Section4.2); 3.to use aheuristic criterion of positional fixedness/alternation to determine the proportion of various subtypes of postpositive constructions in the sample (Section4.3); and 4.to outline an approach to postpositive adjectives integrating their other modifying uses (Section 5). The main objective of this study is to determine the representation of the morphematic types of adjectives and their postpositive uses in acorpus sample. The results of this approach will serve to delineate areas for further inquiry, such as an account of why certain adjectives are frequent in this position and of instances where factors determining the post-head use are not easy to identify. 1 For the summary of the claims in favour of (‘reductionism’) and against (‘separationism’) derivationally relating the attributive and predicative use of adjectives, and other aspects under debate, see e.g. Alexiadou (2014). 2 There is ageneral agreement that “[w]hen two different positions are possible for adjective–noun combinations, the position the adjective occupies relative to the noun it modifies will have areflex on the way it is interpreted” (Alexiadou 2014: 93). Cf. Bolinger’s (1967) interpretational difference between ‘permanent’ vs. ‘temporary’ (navigable rivers vs. rivers navigable) or Ferris (1993). For an overview of interpretative categories, see Cinque (2010). 3 Matthews (2014: 7–14) points out that this division assumes that postmodifiers are not attributive (i.e. function) and proposes adifferent hierarchy involving ‘position’, by dividing adjectives into ‘modifying-only’ (attributive-only vs. postpositive-only) and ‘bipositional’ (default vs. never-attributive). Cf. also Quirk et al. (1985: 403) for central and peripheral adjectives.
PAVLíNA ŠALDOVá 139 2 PRELIMINARIES 2.1 CONSTRAINTS AND CONDITIONS In an English noun phrase, unlike in Romance languages,4 the default position of single adjectives and phrases without their own post-head dependents is before the head noun, “while others occur postpositively, i.e. after the head of the NP”5 (Huddleston & Pullum 2002: 445), i.e. atypical mistake and amistake typical of beginners, alight AP and aheavy AP, respectively. For the sake of simplicity, ‘pre-/post-head position’, or ‘pre-/postmodification’ are descriptors used in this study without any assumptions concerning the relation between position and function (i.e. whether the postpositive uses are more like attributive or predicative functions). The post-head uses of adjectives are regarded as a“minor type of postmodification” (Quirk et al. 1985: 1293–96), subsuming heavy APs (where the adjective is complemented or modified, including comparative constructions or coordination: amistake typical of beginners, amistake both typical and common), fixed phrases (the president elect), and the so-called ‘postposed mode qualifier’ based on the French model (Lobster Newburg). Other postpositive occurrences are treated as an exception to the basic syntactic rule and as subject to severe restrictions, and can be summarized broadly as follows: (a) certain adjectives (proper, absent, concerned, present, involved) occur both in the preand post-head positions, but with adifference in sense: aconcerned face ~ the area concerned (Quirk et al. 1985: 418–19); (b) certain adjectives (appointed, desired, required; following, past, preceding; and positive) can occur in either position: at the appointed time ~ at the time appointed, in years past ~ in past years (ibid: 419); (c) adjectives in -able/-ible can have postposition when the noun is modified by another adjective in the superlative, by only, or by the general ordinals last, next: the best possible use ~ the best use possible, the only suitable actor ~ the only actor suitable (ibid: 418); (d) temporary vs. permanent attributes: the stars visible ~ the visible stars (ibid: 419); (e) aadjectives: the house ablaze, the boats afloat; (f) institutionalized expressions, neoclassical style (cannot be modified): heir apparent, attorney general, Asia Minor (ibid: 418). Other, more flexible, context-dependent and register-specific conditions apply in cases where the two positions alternate. Matthews points out instances of poetic li4 In Spanish, the pre-head position (explicative (nonrestrictive) function, subjective, inherent property) and the post-head position (specifying (determinative) function, objective, non-inherent property) are, to some extent, in asymmetric relation; e.g. el fierro lobo vs. un lobo grande (Zavadil and Čermák 2010: 444–45). For acomparison of Romance and Germanic languages, cf. Cinque (2010). 5 Postposition is obligatory with indefinite pronouns as NP heads (someone young, something new) and will be considered here only where relevant.
140 LINGUISTICA PRAGENSIA 2/2021 cence, such as aline from Tennyson Of that great order of the Table Round, where there is no implication “that there could be any reason, other than metrical convenience, why the adjective should be so used” (2014: 171). Observing the effects of positional mobility of adverbs (Why did you abruptly back away? vs. Why did you back away abruptly?), Bolinger notices that “the stereotyped position of most English adjectives makes selective contrast impossible (except by means of contrastive stress)” (1952: 1122).6 However, “[w]efind an occasional adjective that does … compare closely with the descriptive adverb in admitting two positions … In He belonged now to the wórld everlásting we find aselective contrast, everlasting being opposed to temporal” (ibid: 1130). Selective contrast by postposition is achievable also in parallel constructions (“In the body national as in the body natural …”, ibid.: 1131). Bolinger’s (1967) distinction between relatively enduring ‘characteristic’ and ‘occasion’ use has been shown neither necessary nor sufficient for the account of the postpositive uses, e.g. by Ferris (1993: 48), who postulates different intensional patterns (eligible students = relation of ‘qualification’, the students eligible = relation of ‘assignment’ (ibid: 51)).7 James stipulates that “given an appropriate context, the man happy or the man sick can be perfectly acceptable” (1979: 687), on the condition that “the existence of areferent or referents describable by the NP [is] presupposed” (ibid.). Blöhdorn presents the so-called ‘Attributive Postmodification Filter’ (2009: 161–162) combining three criteria which render apost-head light adjective grammatical: “the adjective receives afocus value within the phrase structure” (following Firbas (1964)); “the adjective is a‘contextually-non-bound item’” (following James (1979) and Sgall et al. (1986));8 and “the determiner has to be ‘quantificational’ (Milsark 1977)”. The adjective thus receives a‘focus semantic value’ (Blöhdorn 2009: ibid), which can be observed, with varying degrees, in the four types of postpositive uses discussed in Section 4.3. 2.2 MINOR TYPE OF POSTMODIFICATION Postmodification by adjectives (both light and heavy APs) is regarded as “aminor type of postmodification” (Quirk et al. 1985: 1293–96), and ‘minor’ is likely to mean ‘infrequent’. In contrast to other forms of postmodification (finite and nonfinite clauses, PPs, or appositives), no quantification of APs as postmodifiers is attempted in Biber et al., the only mention being that “[c]ertain adjectives, such as involved, available, and concerned, tend to be postposed after anoun head” (1999: 519). ‘Minor’ also means that less attention is paid to this structure. The only two studies which use corpora and can provide some quantitative assessment seem to be Blöhdorn (2009) studying single postpositive adjectives in the Brown and Frown corpora, 6 Ferris (1993: 45): “No amount of determination to treat the adjectives as emphatic permits them to become postnominal … *achild intelligent could do that in five minutes or *the juggler skilful wanted apay raise.” 7 For the discussion of the the interpretative effects, see Cinque (2010). 8 Cf. also Svoboda (1970) and Svoboda (1987).
PAVLíNA ŠALDOVá 141 and Schönthal (2013) focussing on postmodification by adjectives (both light and heavy) in the BNC.9 Schönthal’s (2013: 36) estimate based on extrapolation is that postmodifying adjectives (both light and heavy APs) represent 0.78% of all adjectives in the BNC. In his sample of 2,602 post-head adjectives,10 light APs constitute 8%. These numbers, however, include items such as International or United as highly frequent constituents of names of companies or sports clubs, and also indefinite compound pronouns as heads of NPs. Discarding pronominal heads, the proportion of single postpositive adjectives amounts to some 10% of the post-head adjectives. Figures reported in Blöhdorn (2009: 9–21), the only study focusing specifically on single adjectives searching among all adjectives in the Brown and Frown corpora (1 million words each, the total of 153,895 adjectives), suggest that post-head adjectives (both light and heavy APs, including modifiers of indefinite pronouns) represent almost 2% of all adjective uses. Discarding pronominal heads, the light AP uses amount to 0.2% of all adjectives in the corpora and around some 17% of postmodifying uses.11 These slightly disparate proportions reflect differences in methodologies and the scope of the two studies. Moreover, any generalizations will correlate with the frequency of adjectives in specific functions, which are register sensitive (cf. Biber et al. 1999: 506). The same applies to the occurrences of adjectives as types: Schönthal (2013: 41) indicates that five adjectives concerned, available, present, possible, responsible represent 82.7% of all single postpositives in his sample.12 Blöhdorn (2009) identified 384 tokens and 145 types, the most frequent being general, incorporated and available (80, 27 and 26 instances). As is evident, amore systematic description should be sought, which would take into account the heterogenous nature of the adjectives occurring postpositively. 9 Markus (1997) used the LOB corpus only to illustrate selected combinations. 10 Schönthal (2013) classifies the postpositive uses into eight categories. His examples include e.g. attorney general, members present, but also something surprising, …handed acoconut fresh from the tree, an egalitarian sense of solidarity quite unlike the stratified society outside, her tendency to play parts older than herself, confronted with truths too horrible to face, the date so fixed etc. 11 The percentages are not given in the study. The total of 438 single postpositives is reported. 54 tokens, however, can be discarded as they are not English NPs, and, in addition, it can be assumed that anumber of items listed in Tables at pp.18–20 do not represent instances of adjectival postposition (e.g. wide, high, still or west), so the number of instances can be expected to be lower. 12 Schönthal’s (2013: 37) thirty most frequent post-head adjectives (both light and heavy APs): available, concerned, United, International, involved, necessary, responsible, similar, different, full, new, wrong, likely, special, capable, interested, due, present, appropriate, possible, relevant, suitable, greater, inherent, payable, important, comparable, specific, common and equal.
142 LINGUISTICA PRAGENSIA 2/2021 3 EXTRACTION OF THE SAMPLE The written section of The British National Corpus (BNC)13 was used to retrieve asample representing the range of adjectives used in the post-head position. Single postpositive adjectives are not only infrequent (cf. 2.2) but also inherently difficult to retrieve, as asequence of positions tagged ‘N(oun)’— ‘A(djective)’ without any further limitation is too general (210,055 hits in the subcorpus). To ensure manageability of the data in this first approximation, athird position was added to narrow the range of syntactic patterns retrieved. In this study achoice was made to restrict the N-Asequences by afollowing V(erb) in any form, yielding 6,301 hits.14 Potential false positives were further filtered out by constraining the adjective position to instances in lower case (to filter out parts of proper nouns, e.g. South African, Ltd., Inc., titles, and block language), and by using anegative list of adjectives which are not plausible light postmodifiers (usually resultatives and expressions of measure).15 This query yielded 1.800 lines, out of which 30% had to be removed after manual checking, resulting in the final sample size of 1.247 concordances (cf. Table 1). The procedure included reading the lines and determining, with the help of The Oxford English Dictionary Online (OED), information on word-class status, whether the items could be regarded as adjectives and whether they could be interpreted as light postmodifiers.16 13 The search was limited to the written component as single postpositives (both adjectives and participles) generally occur in formal written styles (James 1979: 688). Furuta (2012: 515) reports that single past participles are six times more frequent in the written section (65.17/pmw) of the BNC than in the spoken component (9.66/pmw). In addition, in stances of spoken language are notoriously difficult to analyse. Text HPY (Ferris 1993) had to be removed from the search as its Chapter 5 focuses on single postpositives and contains anumber of speculative and ungrammatical examples, which, unless discarded, appear among the concordances. 14 An alternative or simultaneous option (third position in the query) to prevent excess quantity of false positives can be to retrieve the N-Asequence followed by apunctuation mark (over 30,000 hits), should alarger sample be desired. The representation of adjective types thus obtained does not differ from the results of the query used herein. 15 The regular expressions in the query: [tag="N.*"& word!="term|word"][tag="AJ0" & word="[a-z].*"& word!="old|young|senior|junior|long|high|wide|tall|thick|thin|deep|wo rld-wide|billion-worth|metric|western|international|daily|weekly|great|full|right|alone |fresh|even|happy|safe|fine|stent|arid|special|freezing|entertaining|jnr|saline|sure|only |open|pregnant|clear|intensive|prone|slow|clean|good|dry"] The list can be extended further, especially with short adjectives (resultatives) and items appearing in constructions such as five months pregnant, two meters wide, etc. Inversely, these adjectives can be used in additional searches to justify/problematize their inclusion in the list. 16 Some excluded instances represent potentially interesting recurring patterns, i.e. This suggests (…) that the spectrum function Eis independent of the energy production processes for all wave numbers large compared with those at which the production occurs.
PAVLíNA ŠALDOVá 143 The retrieval of the target combination is complicated not only by interfering systemic factors inherent in an analytical type of language, both at the syntactic and word-formation levels, and the interface of the two (e.g. the word class status of some adverbs/adjectives, participles/adjectives, etc.), but also by the fact that the query is fully dependent on word-class labels determined by the tagger (the CLAWS system17 in our case). The query yields instances of inconsistent tagging or ambiguous (portmanteau) word class status, mainly of certain adverbs/adjectives and –ing and -ed participles/adjectives. Based on the exploratory character of this study, most -ed and -ing items yielded by the query were included in the sample so that areas of potential overlap remain visible and can motivate searches of analogous items (cf. also the discussion of example (3.b) below and in 4.1). The interfering false positives are syntactic patterns with complex transitive verbs such as keep, leave, make, find, hold, etc. illustrated in (1a); absolute adverbial constructions (with their mouths open); instances of apposition (1b); or elliptical/fragmentary structures (1c): (1) a. the number of observations necessary to render an inductive inference legitimate b. Some of the vowels … are central and some are not; the feature central would identify… c. Atlantic event magnificent says PM Another complicating factor is the tagging of adjectives due to the highly productive pattern of conversion from adjectives to nouns, where lexicalized nouns are tagged incorrectly as adjectives (2a-b), including adjectives used as afused modifier-head (partial conversion). Salient are -ic(ical) nouns as in (2c), but many other suffixed or simple results of conversion into nouns (2d) are also tagged incorrectly, including interesting cases not recognized as nouns by the OED (2e,f). Instances interesting from the typological point of view are given in (2g) where the first element is aplural modifier/gender marker to azero-marked plural noun. (2) a. The surface integral may be converted to avolume integral. b. party faithful; Dublin deaf; the Newry select c. the study of an inner city problematic; abastard classical took over d. the Hacienda massive; raspberry cordial; Pisco sour; development on aTeesside common; forest fair; McCartney sound-a-like e. Unlike the novelistic of cinema …, the television novelistic is … f. outputs from the lower frequency bistable are …;18 asimple CMOS astable g. the dilettanti rich followed limply behind; women religious Another complex issue is the discrimination between homonymous adjectives and adverbs (viewed either as an instance of conversion or aterritory of functional over17 For the description and discussion, see Reference Guide for the British National Corpus (XML Edition) at http://www.natcorp.ox.ac.uk/docs/URG/posguide.html. 18 The 4013 i.c. contains two independent bistables so a2-bit latch may be made using only one chip.
144 LINGUISTICA PRAGENSIA 2/2021 lap). Thus abreast (3a) is tagged as an adjective on analogy with asleep; country-wide in (3b) is tagged as an adjective in contrast to nation-wide tagged as an adverb. (3c) shows one more example where ahomonymous adverb is tagged as an adjective. (3) a. where rocks … came so near together that two horses abreast could hardly pass b. the sample of schools country-wide showed… c. the supply of alcoholic liquor … to any private friends of the holder of the licence bona fide entertained by him at his own expense Compounds represent another systemic difficulty, especially the highly productive adjective-centred compound adjectives ‘noun +adjective’ (4a), or some ‘noun+noun’ compounds (4b): (4) a. tax free, lead free, price sensitive, world famous, crystal clear, water soluble, cost effective, the most computer illiterate, H pylorus positive b. video nasty, DNA duplex Other instances of false positives involve phrases from languages other than English (5a), Latin-based binomial nomenclatures (5b), and excerpts representing older stages of English (5c). (5) a. le français fondamental b. Nucellum lapillus, crocus biflorus c. cunnes wise; Shakespeare’s Sonnet 54: O, how much more doth beauty beauteous seem All other examples used in the discussion below that are not part of the main sample are also taken from the written section of the BNC, unless indicated otherwise. The same holds for any additional searches concerning the variability of the position (Sections 4 and 5). 4 RESULTS AND ANALYSIS The sample of 1,247 postpositives contains 150 types of adjectives (Table 1), out of which 50 adjectives appear with afrequency higher than one (cf. the superscript figures in Column 3, Table 1). Ten most frequent adjectives represent 80% of the sample conforming to the Zipfian distribution (Table 6 in Section 5). 4.1 MORPHOLOGY Table 1 outlines abroad division of the postpositives in the sample based on their morphematic structure.19 19 The division is inevitably asimplification, not discriminating between un/productive suffixes (e.g. payable vs. possible, extant, etc.).
PAVLíNA ŠALDOVá 145 The most numerous (and questionable in terms of the word class identification) are (un-)-ed forms (480). “Participles have always been considered as exhibiting features of two word classes, namely those of verbs and adjectives… it does not come as asurprise that constructions containing participles may have similar ambivalent nature.” (Schönefeld 2015: 424) The traditional division of -ed forms into adjectival and verbal passives has been modified by Embick (2004) adding aresultative type, and by Sleeman (2011) adding afourth type of ‘eventive property’ reading, and proposing several criteria along the scale between fully adjectival (stative) and fully eventive types of participles to distinguish among ‘adjectival’, ‘resultative’, ‘eventive property’ and ‘fully eventive’ types. morphology types tokens -able -ible 24 available252 payable45 receivable4 imaginable3 allowable2 applicable2 detectable2 recoverable2 acceptable achievable assessable attainable chargeable liable obtainable suitable unenforceable unimaginable unquestionable responsible44 possible23 visible5 invisible permissible 397 (un-)-ed 36 concerned428 interested8 unemployed5 unknown3 excreted2 raised2 unalloted2 unseen2 appalled bereaved collected combined deferred designated detailed devoted driven excited listed machined missed selected stored tested textured unaddressed unbeaten unclothed undetected unionised unmounted unscathed unsecured unserved unspecified used 480 -ing 11 outstanding23 everlasting2 approaching deteriorating disappearing existing occurring paying pending prevailing remaining 34 -al 10 general8 martial6 fiscal3 personal3 royal3 floral2 temporal2 digital financial spiritual 30 -ous 4fuliginous gracious nervous previous 4 -ent 6 present105 apparent4 adjacent3 absent magnificent violent 115 -ic(al) 7 artistic economic ecclesiastical electronic exotic grammatical scientific 7 -ful 2beautiful wonderful 2 other 20 corporate18 necessary14 rogatory7 extant4 absolute2 intact2 angry correct emeritus genteel golden guilty implicit designate morose opposite presumptive roman supreme utilitarian 61 a4 alive7 asleep4 afloat3 awake216 compounds 12 countrywide6 post-xx4 dark-ringed fitted-out left-right magneto-optical near-impossible par-excellence pre-1900 socalled upstream worthwhile 20 simple 14 proper49 due6 galore5 certain4 simple4 overt3 new2 real2 barren dead dire free ill usual 81 types 150 1247 Table 1.Postpositive adjectives in the sample (superscript values = number of occurrences where n >1)
152 LINGUISTICA PRAGENSIA 2/2021 pronouns (with azero article expressing indefinite reference in all our instances), the resemblance may be said to translate to the position of the modifier (restrictor). In addition, adjectives attested in this position are morphologically complex.33 All need not be present in all instances, as shown in (11c), probably due to the presence of other quantifiers (some). (11) a. (all) things artistic, digital, electronic, existing, exotic, scientific, unseen, wonderful; her taste in all things artistic was polished b. matters ecclesiastical, economic, grammatical; whatever part of the brain it is that deals with matters grammatical can be regarded as… c. some idea of the basics of processors and things digital will help Moving the adjective before the noun seems more plausible with matters due to its abstract sense, as things following an adjective imply ‘things in the physical sense’ (~electronic things). The pre-head position thus seems constrained by the meaning of the whole construction. (ii) Postpositive past34 occurs with temporal nouns (cf. *approaches past), and positive only with proof. As the adjective in proof positive can occupy both positions, the choice is motivated stylistically (to match other legal terms). To this category anumber of adjectives illustrated in (12a,b,c), of French and Latin provenance, can be subsumed, with stylistic preferences for each position as they appear with arestricted terminological set of head nouns.35 Example (12d) illustrates two postpositive adjectives, with general director designate also attested (Google). (12) a. body corporate (cf. body politic) ~ corporate body b. Dr E.Ramey, professor emeritus of physiology at … ~ Mr A.Chandler, emeritus professor of business history at … c. entente floral ~ floral entente d. the new director general designate’s previous work ~ the general director designate Relying on analogy with the examples above, we may tentatively add to this group those instances displaying varying effects of alternation, which can be regarded historically as relics. They testify to the stages of development where single postmodifiers were more common, pertaining to the poetic style (hyperbaton) as in (13a,b), or resulting from analogies with Latin sources (13f). 33 Simplex adjectives appear in coordination: Seduced by all things fine and feudal… or are rare (all things purple). Adjectives denoting nationalities are frequent, but they have been filtered out by the query using lower case adjectives only. 34 Not attested in the sample as past in days past or years past is tagged as an adverb by CLAWS tagging. 35 In Google Books N-gram Viewer, body corporate, professor emeritus, director general and their alternants display converging tendencies.
PAVLíNA ŠALDOVá 153 Example (13c) shows anow lexicalized combination pastures new (‘afield or area of thought or activity’), alternating with new pastures in (13d). Aslightly different degree of lexicalization is displayed by body beautiful (13e),36 and possibly also in life everlasting in (13f). It is evident that changing the position in instances (13c,e,f) results in the loss of allusion to their sources (as in, to acertain degree, body corporate), representing adjustment to the regular pattern of modification. (13) a. thy rocks, and streams, Slumbers! — while tribes fuliginous invade The soft, romantic, consecrated scenes b. shady the cedars around that place; the ripe fruit golden grows there c. Agood run in pastures new would do you aworld of good. d. It is now time to move on to new pastures. My ambition now is to move into film. e. This is aculture which glorifies the body beautiful, placing great emphasis on physical fitness, sexuality and youth. (OED) f. How can an infinite intelligence and life everlasting37 exist within finite bounds? Another discernible pattern, illustrated already in (6b), is the contrastive use of adjectives based on employing antonymic relations syntagmatically for comparison. Such comparison can be marked morphologically (more in (14a)), lexically (preferable in (14b)), syntactically via contrastive coordination (14c) or via acombination of synonymic and antonymic relations in the text, as in (14d). Adjectives following proper names express some temporary aspect of the otherwise definite entity, and when not explicitly in an antonymic contrastive relation, they can be preceded by the indefinite article (14e).38 (14) a. Ifear the world invisible is more visible to him than to the rest of us. b. Francis decided that Lacuna angry was almost preferable to Lacuna gloating. c. The last few days he had been getting on her nerves. George fit and well could get you down, but George ill was anightmare. d. She couldn’t remember ever seeing Lydia so cast down. [intervening 16 clauses] If the presence of Beuno was all that would cheer up Lydia then Betty felt that he should be here. Lydia morose seemed apainfully unnatural phenomenon. 36 “originally used for emphasis; now used to imply the obsessive pursuit of an idealized form or vision” (OED, “beautiful”, adj1c) 37 Possibly acalque of vitam aeternam in the Apostolic Creed. All quotations in OED provide instances of everlasting in the pre-head position only: (“everlasting”, adj1) e.g. 1500 forgyuenes of synnes ayen rysynge of flesshe, & euerlastynge lyf. 1554 Tyll Ioy and euirlestand lyfe. This combination is lexicalized into aplant name for Anaphalis margaritacea: Southern folk medicine employed the use of life-everlasting or everlasting root to induce sweating… (OED), cf. also other plant names: mountain everlasting, mouse-ear everlasting, etc. Cf. also world everlasting. 38 This use is different from the use of proper names as acommon noun, as aHercules, aMary, etc. Overall, proper names represent aspecific group with respect to modification.
154 LINGUISTICA PRAGENSIA 2/2021 e. Tenacious tackling and agreat display from Peter Canavan at full forward … Then just on half time the Ranch landed the blow from which UCG never recovered. ACanavan free was floated into the square, where corner forward Stephen Ramsey rose to fist the ball to the net. Examples pertaining to iii. the only day suitable in Group 4 are characterized as adjectives in -able or -ible which require an attributive superlative or only (the best result possible and *the result possible) (Huddleston & Pullum 2002: 445), or the general ordinals last, next (Quirk et al. 1985: 418–19). Our sample is, however, more varied, and only some of the instances can be explained by the constraints suggested above. The remaining 519 instances can be divided broadly into six sets based on their morphematic structure: -able/-ible (397); -ed (52); -ing (34); Latin and French participial extant and adjacent (cf. present) (2); necessary (14); and compounds (20). The 397 -able/-ible instances display anoticeable dominance of available (63%), followed by payable and responsible (11% each). As these three adjectives are clearly very different in terms of their usage, amore detailed discussion is deferred to Section 5. Another striking observation is that not all instances of the post-head uses of these adjectives are limited by the constraints stipulated above. Thus instances in (15a,b) contain only and the superlative, whereas instances in (15c,d,e) do not. The same applies to the other morphematic groups described below. (15) a. The only kind of central heating possible would be solid fuel. b. thousands of gallons of water at the highest pressure possible were pumped through the channels c. enough staff available would achieve training aims d. in the short-term the finance available would not necessarily allow all desirable changes… e. The use of UHF rather than VHF will further restrict the coverage possible. f. the rent achievable does not cover the outgoings Among the 52 -ed adjectives, the most represented is interested (15%). As examples in (16) indicate, the constraints are not identifiable in these instances either. In addition, interested could be seen as patterning with concerned, as the pre-head uses can vary in meaning (‘in aposition to gain from asituation or be affected by it’), though the senses may overlap (16a). (16b) and (16c) illustrate the variable position without asemantic difference.39 (16) a. Any committee interested is invited to send arepresentative. (~As an interested party, Iwas not allowed to vote.) b. Another issue unaddressed is the meaning of certainty. (~all the unresolved and unaddressed issues of the time) c. the spies were about and aDominican in some way unknown secured acopy (~apriest of some unknown religion) 39 Other factors involved, which may be related to the NP complexity, are disregarded.
PAVLíNA ŠALDOVá 155 The pre-head position of some -ed forms is notoriously restricted and needs to be surveyed in more detail, as well as the constraints on the single posthead -ed forms. Similar patterns are observed with other ‘participial’ modifiers, -ing (17a,b), extant and adjacent (17c-e). (17) a. the only giants remaining being the crocodiles, alligators and caymans b. The wet and windy conditions prevailing were not conducive to good hockey. c. among the best Roman versions extant are the Temple of Fortuna Virilis d. most of the work extant is in Palermo e. the people of the country adjacent will not pay Compound postmodifiers, the last morphematically distinct type, also call for further inquiry, representing an area overlapping with adverbs where aword-class distinction may be hard to draw, as examples of postand pre- (18a), countrywide, upstream, or par excellence attest. In (18b) the post-head use may be motivated by contrast, in (18c) by immediate analogy. (18) a. English and European works of art pre-1900; Darwinism post-Fisher is called neo-Darwinism. b. all the angles left-right have been turns from left to right looking from the base line c. Let Madrid and amission near-impossible40 bring what it may. Abrief survey in Group 4 indicates that postpositive uses of adjectives may be constrained or motivated by arange of factors. Type (i) things artistic represents aconstruction where things has abstract meaning (similar to some-/anything). With Type (ii) entente floral ~ floral entente the postpositive use reflects classificatory systems and older patterns/sources with the adjective receiving ‘afocus value’. This value is also present in patterns of contrastive antonymic relations, which allow simplex adjectives (ill, free) to appear in the post-head position. Type (iii) the only day possible subsumes -able/-ible adjectives and ‘participial’ forms (-ing, -ed, extant) with the superlative and other ordinals, but it has been shown in (15c-f), (16), (17b,d,e) that not all instances are determined by this constraint. Overall, they represent adjectives with adistinct morphematic structure where the factors stipulated in literature cannot be resorted to in all instances and as such should be submitted to further scrutiny. 5 TEN MOST REPRESENTED ADJECTIVES Comparison of the ten most featured adjectives (80% of the sample) given in Table 5 offers another heuristic,41 as these adjectives display several different tendencies of use (Table 6). 40 After the name of aU.S. television series Mission: Impossible. 41 This approach can only be hinted at here due to adifferent focus of this paper.
156 LINGUISTICA PRAGENSIA 2/2021 concerned 428 34% available 252 20% present 105 8% proper 49 4% payable 45 4% responsible 44 4% outstanding 23 2% possible 23 2% corporate 18 1% necessary 14 1% of the total 1.247 1001 80% Table 5.Ten most frequent adjectives in the sample Drawing on Matthew’s suggestion that “[w]emust look for wider similarities, with modifiers that either can or have to be postpositive” (2014: 169), adjectives from Table 5 (mainly Groups 3 and 4 above) can be assessed from the point of view of their prehead and post-head uses, including acomparison of both light APs and heavy APs (19a), (19b) and (19c) respectively. Corporate can be left out of this account as it takes no dependents. (19) a. necessary skill b. do the work necessary c. the food necessary for his diabetic diet BNC written AN N Aheavy N Alight light AP of posthead uses concerned 310 767 1.808 70% proper 4.116 28 319 92% present 8.295 247 464 65% outstanding 1852 57 149 72% possible 7.903 73 392 84% available 1.945 3.161 2.640 46% responsible 655 1.033 122 11% necessary 3.348 993 166 14% payable 1 490 135 22% Table 6.Comparison of the nine most represented adjectives42 The following brief observations can be made: After the head noun, 1. concerned, proper, present, possible, and outstanding are used prevailingly as light APs; 2.on the contrary, responsible, necessary, and payable are used prevailingly in heavy APs; 3.interestingly, available, the most frequently used 42 These figures represent sequences A-N and N-A, disregarding any additional modification.
PAVLíNA ŠALDOVá 157 single adjective and without alexicalized difference, shows no clear pattern (n=5.801, 54% heavy vs 46% light APs). As for the first group (prevalence of light APs), light concerned and proper contrast with the heavy APs in meaning (adetective concerned with arecent case; clinical societies concerned about brain damage; W. continued in atone proper for instructing young ordinands) showing aclear difference between the two post-head uses. Outstanding is also used dominantly as alight AP, and its strong collocates indicate that the post-head occurrences display anarrower, specialized meaning than the pre-head uses, as it takes financial terms as head nouns only.43 In the second group (prevalence of heavy APs: responsible, necessary, and payable44), responsible instantiates the sense difference in preand post-head positions, but the light uses in the post-head position are related to the heavy uses of this adjective (as is the case of outstanding or present above). These two uses can, tentatively, be perceived as related, due to the strong prevalence of the heavy pattern with for (89%). What these adjectives appearing dominantly in heavy APs seem to share is that in light uses, the ‘missing’ dependent is identifiable in the context, as shown in (20). As necessary (among other adjectives) has been shown to have developed asecondary post-determiner deictic meaning indicating relations in the textual situation (Davidse et al. 2008), asimilar mechanism can be hypothesized to apply in the post-head light uses of some adjectives. (20) The most negative aspects of American policing … came together in the beating of Rodney King in …, which was followed by rioting when the white policemen responsible were acquitted in April 1992. [i.e. responsible for the beating] Available as the most frequent postpositive adjective without adifference in meaning is unique in the relatively equal proportion between the post-head positions, contrasting sharply with other adjectives, where one of the post-head patterns dominates. The fact that available is so pervasive in comparison with other adjectives which appear in the post-head position mainly due to the lexicalized difference in meaning calls for further scrutiny. 6 CONCLUDING REMARKS The paper has surveyed possibilities of extracting postpositive adjectives in acorpus. The main difficulties in the retrieval stem from interfering syntactic patterns (e.g. complex transitive verbs) and afunctional overlap between certain adjectives 43 The pre-head collocates of outstanding, not found with the adjective in postpositive use, are example, success, performance, achievement. The head nouns shared in both uses include bill, balance, loan, debt, amount, bond, share, mortgage. 44 Payable seems to be constrained to the post-head position, patterning with outstanding in that its strong collocates are amount, interest, tax, sum, fee, rent. Only one pre-head occurrence is attested in the BNC (the payable insurance money).
158 LINGUISTICA PRAGENSIA 2/2021 and adverbs (e.g. nation-wide, cf. Section 3). Overlapping of adjectival/verbal functions of -ed/-ing forms is reflected in inconsistent tagging of their word-class status. Given the fact that the definition and tagging of an adjective is inherently problematic, retention of participial forms in the sample serves the purpose of highlighting the similarities with -able/-ible forms. Our results will allow to refine corpus searches using specific adjectives or specific morphematic structure (e.g. asurvey of un-forms shared by the two frequent forms) to inform further research on postposition. The dominant types of adjectives in the post-head position are -able/-ible forms and forms relatable to participial uses,45 indicating that adjectives with this morphematic make-up represent the productive type of postpositives. Other adjectives (apart from Latinate compounds and a-adjectives) are attested mainly as instances of older poetic use and in antonymic contrastive patterns. As such they are found in environments providing the ‘focus value’, in which even asimplex adjective can appear postpositively. In terminological compounds, the ‘focus value’ correlates with the adjective function as differentiae, and it would be interesting to survey the developmental tendencies of this alternation. It has become clear that in order to explain the function and the range of related morphematic types of adjectives in post-head position, textual examples of -able/ -ible adjectives, -ed and -ing forms, in relation to adverbs, should be studied further to account for occurrences that cannot be explained by the presence of the superlative or ordinals. The fact that available is used so widely without alexicalized difference and without displaying apreference for either light or heavy uses, or the fact that necessary is also quite prominent may point in the direction of some secondary textual meaning, in asense tentatively parallel to the deictification of certain postdeterminer adjectival uses (Davidse et al. 2008). More detailed analysis is needed to determine whether and how the post-head uses of adjectives (subsumed under Group 4 (iii) above) correlate with the use of determiners and collocate with expressions of quantity (James’s (1979) presupposition constraint). The deliberately broad perspective adopted in this study, rummaging through Fischer’s “mixed bag of remnants” (2006: 254) and Bolinger’s “scraps and left-overs” (1952: 1132), hopes to have sustained implicit relations of apattern in all things artistic as an intermediary between the placement of adjectives after indefinite compound pronouns and productive patterns which make the placement of certain morphematic types of adjectives after nouns available. Acknowledgements This research has been supported by the research grant GAČR (Grant Agency of the Czech Republic) No. GA19-05631S “Adjectival postposition in English”. 45 Cf. Haspelmath (1994: 167) on analogy in formation of passive participles and -able/-ible adjectives, or Sleeman (2011).
PAVLíNA ŠALDOVá 159 REFERENCES Alexiadou, A. (2014) The syntax of adjectives. In: Carnie A., Y.Sato and D.Siddiqi (eds) The Routledge Handbook of Syntax, 89–107. London, New York: Routledge. Biber, D., S.Johansson, G.Leech, S.Conrad, E.Finegan (1999) Longman Grammar of Spoken and Written English. Harlow, Essex: Longman. Blöhdorn, L.M. (2009) Postmodifying attributive adjectives in English: an integrated corpus-based approach. Frankfurt am Main: Peter Lang. Bolinger, D. (1952) Linear modification. PMLA 67, 1117–1144. Bolinger, D. (1967) Adjectives in English: attribution and predication. Lingua 18, 1–34. Cinque, G. (2010) The Syntax of Adjectives. AComparative Study. Cambridge, MA: MIT Press. Davidse, K., T.Breban and A.Van Linden (2008) Deictification: The development of secondary deictic meanings by adjectives in the English NP. English Language and Linguistics 12/3, 475–503. Embick, D. (2004) On the structure of resultative participles in English. Linguistic Inquiry 35/3, 355–392. Ferris, C. (1993) The Meaning of Syntax: aStudy of the Adjectives in English. Harlow: Longman. Firbas, J. (1964) On defining the theme in functional sentence analysis. Travaux Linguistiques de Prague 1, 267–280. Fischer, O. (2006) On the position of adjectives in Middle English. English Language and Linguistics 10/2, 253–288. Furuta, Y. (2012) Postpositive Past Participles Used on Their Own. International Journal of Social Science and Humanity 2/6, 514–517. Hallman, P. (2021) On Passive and Perfect Participles. In: Grohmann, K.K., A.Matsuya and E.-M. Remberger (eds) Passives CrossLinguistically, 64–97. Leiden, Boston: Brill. Haspelmath, M. (1994) Passive participles across languages. In: Fox, B. and P.J.Hopper (eds) Voice: Form and function. Typological Studies in Language 27, 151–177. Amsterdam: Benjamins. Huddleston, R. and G.K.Pullum (2002) The Cambridge Grammar of the English Language. Cambridge: Cambridge University Press. James, D. (1979) Two semantic constraints on the occurrence of adjectives and participles after the noun in English. Linguistics 17, 687–705. Jones, S. (2002) Antonymy: ACorpus-Based Perspective. Hoboken: Taylor and Francis. Markus (1997) “The men present” vs “the present case”: word order rules concerning the position of the English adjective. Anglia— Zeitschrift für englische Philologie 115/4, 487–506. Matthews, P.H. (2014) The positions of adjectives in English. Oxford: Oxford University Press. Meltzer-Asscher, A. (2010) Present participles: Categorial classification and derivation. Lingua 120/9, 2211–2239. Milsark, G, (1977) Toward an Explanation of Certain Peculiarities of the Existential Construction in English. Linguistic Analysis 3, 1–30. Quirk, R., S.Greenbaum, G.Leech and J.Svartvik (1985). AComprehensive Grammar of the English Language. London: Longman. Reference Guide for the British National Corpus (XML Edition) edited by Lou Burnard, February 2007. Available at http://www. natcorp.ox.ac.uk/docs/URG/ [last accessed 15July 2021] Schlüter, J. (2008) Constraints on the attributive use of ‘predicative-onlyʼ adjectives: Areassessment. In: Trousdale, G. and N.Gisborne (eds) Constructional approaches to English grammar (Topics in English Linguistics 57), 164–179. Berlin: Walter de Gruyter. Schönefeld, D. (2015) Aconstructional analysis of English un-participle constructions. Cognitive Linguistics 26/3, 423–466. Schönthal, D. (2013) Adjectives postnominal. An analysis of the nature of adjectival postmodification within the English nominal group. MA thesis, Cardiff University. Sgall, P., E.Hajičová and J.Panevová (1986) The meaning of the sentence in its semantic and pragmatic aspects. Dordrecht: Reidel— Prague: Academia.
160 LINGUISTICA PRAGENSIA 2/2021 Sleeman, P. (2011) Verbal and adjectival participles: Position and internal structure. Lingua 121(10), 1569–1587. Svoboda, A. (1970) Anote on N + Adj. Sborník prací filozofické fakulty brněnské univerzity A18, 87–92. Svoboda, A. (1987) Functional perspective of the noun phrase, Brno Studies in English 17, 61–86. Zavadil, B. and P.Čermák (2010) Mluvnice současné španělštiny: Lingvisticky interpretační přístup. Praha: Karolinum. SOURCES The British National Corpus, version 2 (BNC World). Distributed by Oxford University Computing Services on behalf of the BNC Consortium. Ústav Českého národního korpusu FF UK, Praha 2001. Accessed at WWW: http://www.korpus.cz [last accessed 12April 2021]. The Oxford English Dictionary Online. March 2021. Oxford University Press. https://www.oed. com [last accessed 12April 2021]. Pavlína Šaldová Department of English Language and ELT Methodology Faculty of Arts, Charles University Nám. Jana Palacha 2, 11638 Praha ORCID ID: 0000-0003-0043-7520 [email protected]