Boring as hell: a corpus study of intensifying post-modification of predicative adjectives in the ‘ADJ as NOUN’ frame

Boring as hell: acorpus study of intensifying post-modification of predicative adjectives in the ‘ADJ as NOUN’ frame Jaroslav Emmer (University of South Bohemia in České Budějovice) ABSTRACT This paper presents aphenomenon that has attracted little attention despite being relatively frequent and very productive in spoken language. The aim of this study is to prove that the ‘ADJ as NOUN’ frame, which is commonly used to form comparative phrases and similes, is also employed to intensify the meaning of predicative adjectives in both English and Czech but without the element of comparison. This study analyses data from two spoken corpora; the Spoken BNC2014 for English, and ČNK— ORAL_v1 for Czech. The corpus data serve as evidence confirming the existence of this pattern as adistinct entity with its own functions alongside being used in comparisons and similes. The results further show that both languages display semantic incongruence between the adjectives and nouns co-occurring in this structure. However, each language uses nominal elements from different lexical fields. Finally, the paper discusses potential issues of interpretation and the general motivation behind the usage of intensifying post-modification. KEYWORDS corpus data, intensification, post-modification, predicative adjectives, simile DOI https://doi.org/10.14712/18059635.2020.2.2 1 INTRODUCTION The initial idea of this study is based on corpus research of adjectival similes. The process of gathering simile items from various spoken corpora using the formalized correlative ‘(as) ADJ as NOUN PHRASE’ and ‘ADJ jak(o) NOUN PHRASE’ frames produced alot of irrelevant data. Sifting through the clutter, Ifound the vast majority of the items to be common comparative phrases, i.e., (as) much as NP, (as) quick as NP, podobný jak(o) NP; etc. What remained was alist of adjectival similes (conventional idiomatic comparisons) and other items that, upon closer inspection, did not look like any kind of comparison, neither in English, nor in Czech: (1) boring as hell (2) awkward as fuck (3) drahý jak(o) prase [expensive as apig] (4) hořký jak(o) kráva [bitter as acow] Searching for this phenomenon in traditional and reputable sources dealing with English, such as Huddleston and Pullum 2002; Quirk et al. 1985; Dušková et al. 2009, 128 LINGUISTICA PRAGENSIA 2/2020 yielded no results. Moon (2008: 5) describes the above-mentioned examples as ‘emphatic particles’, which is avague definition at best. In Czech, Čermák (2007: 401) views words such as blázen, prase, etc., as synsemantics, i.e. as examples of desemanticized items that occur in many intensifying similes and are very productive. However, it is precisely the desemanticized nature of the nominal element and indivisibility of the ‘as NOUN’ combination that speaks against the simile interpretation. The absence of this phenomenon in reputable sources may be attributed to mainly two reasons. Firstly, the phenomenon utilizes vulgar language, which is generally avoided in serious, authoritative texts. Secondly, it is not considered aphenomenon that would be distinct from similes, even though it is never explicitly listed among the examples of simile. As the term simile suggests, the feature that stands in the foreground is similarity. Čermák (2007: 385) notes that similarity is one of the most prominent semantic relations, however, it is largely based on our view and categorization of reality. Literal similarity is to be sought as the main relation at work in literal comparisons but not necessarily in established similes, where the function is primarily to highlight acertain feature that is shared by the topic (comparandum) and the vehicle (comparatum) within the given context. The purpose of an established adjectival simile is not to provide aliteral comparison but to highlight acertain feature of the topic with the use of aprototypical carrier of the said feature. In this sense, adjectival simile functions as aquality or degree modifier of the feature, often also involving speaker evaluation. The function of the ‘as NOUN/jak(o) NOUN’ in expressions such as expensive as fuck and horký jak svině is solely to emphasize and the idea of the frame involving comparison of the topic and the vehicle is inconceivable. In this paper, Iwill discuss both structural and semantic reasons why these constructions should not be treated as simile. 2 FORMAL CONSIDERATIONS As was already said, there is aformalized ‘(as) ADJ as NOUN PHRASE’ frame that is most frequently employed in comparative phrases (e.g. as good as, as long as) and adjectival similes (e.g. (as) white as snow, (as) thick as thieves). The only structural variation to be observed is absence of the initial as which is important when considering the third construction that uses the frame. This construction is best described as ‘intensifying post-modification’ (further only IP) because the function of the nominal element is only to intensify (emphasize) the sense of the adjective. Structurally, IP is different from comparative phrases and adjectival similes as it does not allow any variation— initial as is not possible. While this may seem trivial, it actually reflects an important feature of IP that formally separates it from adjectival similes. Adjectival similes and comparative phrases use the as … as string, which is sometimes referred to as acorrelative conjunction (Dušková 2009). While this description may be applicable to cases such as as clear as crystal in their full form, i.e. without the omission of the initial as, it is much less viable for cases such as stiff as aboard, where we miss the first correlate. The category of correlative conjunction might be JAROSLAV EMMER 129 understandable as aresult of comparison of English and Czech classification but feels somewhat arbitrary in the context of English. Huddleston and Pullum (2002: 1130) describe the initial as as a“degree adverb marking scalar equality”, which is often omitted when the adjective-head is followed by acomparative complement ‘as + NP’. The second as is best described as asubordinator, more specifically as prototypical comparative preposition (Huddleston and Pullum 2002: 1141). It follows from this that the ‘as + NP’ complementation is realized by apreposition phrase whose function depends on the type of phrase: (1) Simile: (as) white as snow (2) Comparative phrase: as much as (3) IP: boring as hell In (1) we can describe the function of the preposition phrase as introducing scalar equality with the use of an established vehicle for the particular simile. Scalar equality may also be applied to (2) when it is used literally, however, the idiomatic use disqualifies the scalar equality interpretation since the meaning becomes similar to that of although, e.g. As much as Iwould like to, Istill can’t help you. No comparison interpretation is applicable to (3) as there is no real comparison taking place. The preposition phrase functions as an adverb intensifier of the adjective-head and can be viewed as an amplifier within the intensifier classification framework of Quirk et al. (1985: 590–591), who further subdivide amplifiers into: “(a) MAXIMIZERS, which can denote the upper extreme of the scale, and (b) BOOSTERS, which denote ahigh degree, ahigh point on the scale. Both subsets, but especially boosters, form open classes, and new expressions are frequently created to replace older ones whose impact follows the trend of hyperbole in rapidly growing ineffectual.” Out of these two subclasses, IP would classify as abooster. While they do not at any point discuss the ‘as NP’ post-modifier, Quirk et al. (1985: 1414–1415) briefly mention simile and other figures of speech as means of emotive emphasis, however, they pay no further attention to these phenomena because they do not primarily belong to the field of grammar. Another structural difference is determination of the nominal element and its possible modification. In adjectival similes and comparative phrases, determiners and modifiers occur frequently and their presence depends on the grammatical category of number and countability of the nominal element and other contextual factors, e.g. (as) flat as apancake, (as) bright as abutton, (as) thick as two short planks, (as) big as our house, (as) funny as those people, (as) clever as my beautiful daughter, etc. IP examples do not exhibit any determination or modification of its nominal element, e.g. boring as hell, funny as fuck in English, or drahej jak(o) prase [expensive as apig], ledový jak(o) blázen [icy as amadman] in Czech. Such absence is conditioned by the semantic meaning of the nominal element because it has lost its literal reference (has been grammaticized) and therefore there is nothing to determine. 130 LINGUISTICA PRAGENSIA 2/2020 3 SEMANTIC CONSIDERATIONS If we look at Čermák’s(2007: 102) ‘Cd-R-(Tc)-c-Ct’ description of the constituting elements of the frame that was designed for metaphors and similes, we can clearly see that the nomenclature does not fit the IP items. comparandum relator tertium*comparator*comparatum* The lecture is boring as hell Unlike in similes, IP does not compare— it merely intensifies the meaning of the adjective. Tertium comparationis is aterm used for ashared feature of the compared items and therefore does not apply here. As for the comparator and comparatum, since we are not comparing, semantically these terms too make little sense in IP items. In the example above, boring is apredicative adjective and as hell is aprepositional phrase functioning as an adverb of degree that complements the adjective. Adverbs of degree are aspecial semantic subcategory of adverbs and as Dušková et al. (2009) mention, they do not add any independent sense to the construction. Having established the definition of the post-adjectival nominal element, we should also note that the problem with the interpretation of IP comes from its transitional nature. Čermák (2007: 401) raises an interesting question of ambiguity when he mentions synsemantics. Czech uses similes where the nouns in ‘ADJ jako NP’ sequences potentially function as autosemantics, e.g. tlustý jak(o) prase, těžký jak(o) kráva, etc. However, these expressions may well be interpreted as instances of IP with the nominal element desemanticized, and it is hard to unequivocally decide the category based on ‘possible semantic congruence’. In the case of drahý jak(o) kráva, we can assume that buying acow would be an expensive purchase, but the expression cannot be considered aproper simile since expensive is not adefining (nor salient) feature of cow. The remaining category apart from IP is acomparative phrase, which is too context-dependent to be generally felicitous. In English, the boundary is much less fuzzy because expressions such as boring as fuck or dusty as shit can hardly be treated as similes or comparisons of any sort. Expressions with hell may become ambiguous if they contain an adjective congruent with hell and are isolated from the topic, e.g. hot as hell, but once the topic is provided, the expressions is disambiguated, e.g. she is hot as hell, where the expression uses the informal sense of hot (sexy) and triggers the purely intensifying IP reading. Another point to be considered is what makes IP the preferred intensifier in certain situations. Aneed for astronger emphasis is the likely explanation as the degree of intensity is arguably higher than that of ordinary boosters such as very or alot. This emphasis is mainly caused by the vulgar nominal element. One of the hypotheses is that IP primarily intensifies inherently negative adjectives both in English and Czech. It follows from this that IP is to be expected as the preferred intensifier in emotionally charged utterances where negative evaluation is the principal point of intensification, e.g. compare the intensity and emotive evaluation of very expensive and expensive as fuck. However, there are also instances of IP being used as abooster of JAROSLAV EMMER 131 inherently positive adjectives, e.g. gorgeous as fuck or dobrý jak(o) svině. These expressions are viewed by some as infelicitous because of the clash of positive and negative evaluation, as can be seen in other combinations, such as fucking good or hrozně dobrý. Ultimately, it comes down to subjective interpretation of the booster, and when the booster is viewed as afully grammaticized intensifier, the negative aspect may be lost completely. 4 PHRASEOLOGICAL CONSIDERATIONS Since simile is aphraseological unit, it only makes sense to consider what makes IP different to simile with regard to fixedness. Instances of ‘as hell’ or ‘jak(o) blázen’ function as degree modifiers and can be headed by both verbs, e.g. hurt as hell, bolí jak(o) blázen, and adjectives, e.g. hot as hell, horký jak(o) blázen. Therefore, these phraseological units can be recorded as ‘V/ADJ + as hell’ and ‘V/ADJ + jak(o) blázen’, with the preposition phrase being semantically equal to alot or hodně and other near-synonyms of similar degree of intensity. In similes, on the other hand, the vehicle is tied to the tertium and isolating the vehicle breaks the idiom, e.g. isolating the vehicle as abat from its standard tertium blind and attaching it to adifferent adjective would not result in an idiomatic construction (grey as abat) and would most likely be interpreted as an odd literal comparison. The ‘as abat’ string does not carry any idiomatic meaning of its own, unlike IP, which is semantically an independent phraseological unit. This is not to argue that individual adjectival similes always have strictly fixed lexical components, however, the variability is fairly limited to only ahandful of items at most, e.g. white as asheet/aghost/snow, and even those variants may differ pragmatically. It follows from the above that this significant difference has important implications for language acquisition and learning. Similes are learnt as asingle unit of vocabulary, i.e. the tertium and the comparatum are both integral parts of the simile, whereas IP is just aprepositional phrase, paradigmatically independent of the adjective, with the function of intensifying verbs and adjectives. 5 DATA AND METHODOLOGY The primary methodology of this study is based on corpus linguistics and the research can be described as originally corpus-driven because it is focused on aphenomenon that was discovered while analysing corpus data for adifferent purpose. Consequently, once the studied structure was identified, acorpus-based approach was employed for additional data extraction. Given the colloquial nature of the researched phenomenon, the following spoken corpora were chosen: 1. The English Spoken BNC2014, which contains 11,422,617 tokens in 1,251 texts; 2. The Czech ČNK— ORAL_v1, which contains 5,368,392 tokens in 1,546 texts. 132 LINGUISTICA PRAGENSIA 2/2020 The original simple query "(as)? _{A} as (_{ART})? (_{A})? _{N}" used for simile retrieval in the Spoken BNC2014 returned quite afew IP items. The part of speech inspection showed that as is tagged _II (general preposition) and the intensifier is tagged _NN1 (singular common noun). The _NN1 tag is formally justified for anumber of reasons: (i) the lemma can function as anoun outside the frame, (ii) the lemma follows apreposition (see Huddleston and Pullum 2002, for their re-evaluation of the traditional idea that prepositions are conventionally followed by nominal elements), (iii) the lemma is part of aformalized frame and occupies the nominal slot. The ‘PREP + NOUN’ string is not an unusual form to function as an adverbial intensifier, e.g. changed beyond recognition, reduced to ashes, etc. However, recognition and ashes maintain their semantic motivation alongside the intensifying function, whereas fuck or hell become desemanticized and their function is only intensification. Therefore, the noun tag is formally acceptable but feels somewhat less appropriate when considered semantically. There are also cases where the nominal element is shifted to the pre-modifier slot and possibly changes its form, e.g. crazy as fuck×fucking crazy, stupid as shit×shit stupid. This behaviour can also be observed in some adjectival similes, e.g. clear as acrystal×crystal(-)clear, solid as arock×rock(-)solid, or even literal comparisons where it is sometimes the default variant, e.g. knee(-)deep×deep as to reach the knees. This tendency can be explained by the function of the nominal part, which is degree modification. Huddleston and Pullum (2002: 583) state that “the degree function is by far the most common, and in AdvPs and predicative AdjPs it is virtually the only possibility apart from that of the focusing adverbs only, even, etc.”. This shift from postto pre-modifying slot may then be understood as achange to amore ‘suitable’ and prototypical form for an item with the degree function, such as IP. Even informal Czech provides several (mostly) vulgar expressions that show this tendency: hovno/ prd platný, kurva drahý, etc. For the extraction of IP items in the Spoken BNC2014, the query was simplified to “_JJ as _NN1” based on the analysis of the previously obtained results, which showed that there is no pre-adjectival as and no pre-modification of the nominal element. This query returned 138 matches with 12.08 ipm (instances per million words), half of which were IP items. The query “_JJ as _NN2” returned 50 matches and only asingle item, rough as arseholes, could be considered IP. The list of IP items shows three nominal elements (Table 1): Query No. of items Frequency (ipm) JJ as fuck 47 4.11 JJ as hell 24 2.10 JJ as shit 3 0.26 Table 1.The number and frequency of nominal elements occurring in English IP In the case of ČNK— ORAL_v1 extraction, the query “[tag="A.*"] [word="jak.*"] [tag="N.*"]” returned 700 matches with 110 ipm. The ”jak.*” part of the query allows JAROSLAV EMMER 133 for both jak and jako variants to be included in the resulting list of items. The jak/ jako variation is of little semantic significance and is better interpreted as registerconditioned, but there also may be other factors, e.g. speaker preference or prosodic features of the surrounding elements. The frequency breakdown according to the rightmost lemma yielded items of which only the four most frequent nominal elements are listed in Table 2 (sviňa and svině are dialectal variants of the same lemma): Nominal element No. of items Frequency (ipm) prase 29 4.56 sviňa 20 3.14 kráva 19 2.99 blázen 14 2.20 svině 11 1.73 Table 2.The number and frequency of nominal elements occurring in Czech IP 6 RESULTS The following lists provide aset of English adjectives post-modified by fuck, hell or shit in the Spoken BNC2014 and aset of Czech adjectives post-modified by prase, sviňa, kráva, blázen or svině in ČNK— ORAL_v1. The type of evaluation is based on the adjective itself— outside the IP construction. Since the data provide only alimited number of examples that do not show any significant collocational preferences, there is no conclusive evidence of certain nominal elements in IP selecting specific adjectives. It follows from this claim that the selection of nominal items in IP remains aspeaker preference rather than acollocational preference of the adjective. 1. Spoken BNC2014 list of adjectives intensified by fuck, hell and/or shit: (1) Inherently negative evaluation: awkward, bored, boring, clever (sarcasm), cocky, corrupt, creepy, disrespectful, dodgy, dyslexic, expensive, grim, guilty, hard (‘difficult’), lazy, offensive, posh, rough, slow, stupid, stush (derogatory), trippy, ugly, vain, weird. (2) Neutral evaluation: civil, quiet. (3) Inherently positive evaluation: brainy, clever, cool, funny, smart. 2. ČNK— ORAL_v1 list of adjectives intensified by prase, sviňa/svině, kráva and/or blázen: (1) Inherently negative evaluation: blbý, dlouhý (trvání), drahý, hořký, ledový, líný, mafiánský, mas(t)ný (‘drahý’), nahnaný, nachlámaný, náladový, narvaný, nasraný, nezáživný, ostrý, poďobaný, podrážděný, pomalý, přísný, slaný, složitý, 134 LINGUISTICA PRAGENSIA 2/2020 strmý, šílený, špatný, těžký, tlustý, utahaný, uvalený, vožralý, vyděšený, vzteklý, zadlužený, zaprášený, shnilý, zklamaný, zpocený.1 (2) Neutral evaluation: opálený, pálivý, rychlý, silný, studijní, velký, známý.2 (3) Inherently positive evaluation: chytrý, dobrý, levný, přehledný, vděčný, výborný.3 It should be noted that many of these adjectives are polysemous which may result in varying inherent evaluation even before entering context, e.g. clever, hard, funny in English, or dlouhý, mastný, ostrý in Czech. Additionally, arigorous pragmatic account would show that the implications of quite afew of these adjectives can cause ashift in evaluation, e.g. hard work is often used positively when praising someone for their efforts, or levné auto usually suggests that the car is of low quality (negative). The data reveal that both English and Czech exploit the same formalized frame for the purpose of intensifying the adjective-head and that the nouns drawn to fulfil the role of an intensifier come from the same area of lexis— ‘bad language’. These expressions are considered taboo, either inherently or situationally. This observation is hardly surprising because taboo expressions are intense by definition and as McEnery (2005: 1) observes: “(…) such words have powerful effects on hearers and readers (…)“ One of the original questions was whether the taboo expressions used in IP intensify only inherently negative adjectives or potentially also neutral or even positive adjectives. Based on the list above, it is safe to assume that IP in English is predominantly applied to inherently negative adjectives with only afew exceptions. There is no evidence that the data of fuck, shit or hell exhibit different collocability related to negative/positive evaluation. As for Czech, IP again seems to favour inherently negative adjectives over neutral or inherently positive ones. Despite the identical taboo nature of the items used in IP, there are noticeable differences between English and Czech in the viable lexical items and their respective lexical fields. English intensifiers fuck and shit are readily recognized taboo— even in their original literal meanings, which relate to sex and bodily functions. Hell is also generally considered taboo, but there are different extra-linguistic conditions associated with the literal meaning, namely cultural and religious aspects, and one’sphilosophy might easily remove the literal meaning of hell from the taboo domain. Czech intensifiers are predominantly from the ‘animal’ lexical field. By itself this would not make them taboo items, but since they are more frequently used figuratively to describe people, the negative (or even offensive) connotations are prevalent. Therefore, these words are readily recognized as pejorative by most speakers of Czech. The literal meaning of kráva is ‘cow’. Similarly, the literal meaning of both svině and prase is ‘sow’ and ‘pig’ respectively with gender being the only referential 1 Stupid, long (duration), expensive, bitter, icy, lazy, gangster, greasy, scared, drunk, moody, stuffed, pissed (off), dull, sharp, pockmarked, irritated, slow, strict, salty, complex, steep, crazy, bad, heavy, fat, worn out, absurd, drunk, frightened, mad, indebted, dusty, rotten, disappointed, sweaty. 2 Tanned, spicy, fast, strong, study, large, (well-)known. 3 Smart, good, cheap, clear, grateful, excellent. JAROSLAV EMMER 135 difference between the two. However, the dialectal variant sviňa shows an interesting tendency in comparison to svině. See Table 3 and Table 4 for the illustration of the collocational profile of sviňa and svině with the -5/+5 window span: # Lemma frequency MI T-score logDice 1. jak 73 7.011 8.478 6.106 2. být 63 3.356 7.162 2.455 3. ten 60 2.920 6.723 2.019 4. .56 3.325 6.737 2.424 5. .. 34 4.329 5.541 3.427 6. a25 3.075 4.407 2.174 7. já 19 3.220 3.891 2.318 8. on 18 3.142 3.762 2.241 9. jako 16 3.491 3.644 2.589 Table 3.The collocation profile of the IP nominal element sviňa (104 hits) # Lemma frequency MI T-score logDice 1. ten 97 3.134 8.727 2.712 2. být 83 3.274 8.169 2.853 3. .49 2.653 5.887 2.231 4. ,48 3.498 6.315 3.076 5. a38 3.200 5.494 2.778 6. jak 35 5.471 5.783 5.044 7. jako 33 4.056 5.399 3.632 Table 4.The collocation profile of the IP nominal element svině (145 hits) We can see that svině is more frequent than sviňa and is not bound to occur alongside jak(o) (68 out of 145 hits). Sviňa, on the other hand, occurs in combination with jak(o) in 89 out of 104 hits. This can be treated as evidence of sviňa being used predominantly as an integral part of the ‘ADJ + jak(o) + sviňa’ phraseological unit. There is also anoticeable preference for jak in combination with sviňa unlike in svině, where jak and jako are evenly distributed. It should be noted that this preference for jak may well be attributed to speakers’ preference, but the frequency difference seems significant enough to rule out complete arbitrariness and the MI score strongly suggests fixedness of the combination. The last frequent Czech intensifier blázen comes from adifferent lexical field and its literal meaning is ‘madman’. Under the lexeme blázen, Slovník spisovného jazyka českého (1960–1971) lists ‘jako blázen’ as afixed combination which “označuje velkou míru děje” [denotes ahigh degree of action]. Lastly, there are afew items in the Czech corpus that could be considered nominal elements of IP, e.g. prdel (5 hits), sviň (4 hits), hovado (3 hits). They attach to various inherently negative adjectives and are themselves similar to the prototypical cases,