- 1 I want to thank the two anonymous reviewers for their very thoughful and helpful comments. Any rema (...)
- 2 Used in a broad way, the term ‘intensifier’ includes “any device that scales a quality, whether up (...)
1This article1 explores the diachronic semantic development of the intensifier2 way in American English illustrated in (1)-(2):
(1) She didn’t care if he was old enough to be her grandfather – or great-grandfather by now; and Margalo did agree that he was incredibly handsome. But there was old, and there was way old, and Gregory Peck was definitely in the second category. [COHA, Bad Girls in Love, 2002]
(2) “These guys really bugged you, didn’t they?” “No”, I replied with a sigh. “They just have different interests. Way different.” [COHA, The Chicken Asylum, 2001]
- 3 Or ‘wicked cool’, as suggested by one reviewer.
2Way belongs to the category of amplifiers which scale upwards (Quirk et al. [1985: 590]). Quirk et al. [1985: 590] divide amplifiers into maximizers, which refer to the upper extreme of the scale, and boosters, which denote a high point on a scale. According to the Oxford English Dictionary (OED s.v. way adv.3), way in its intensifying use means ‘extremely, very’; it thus functions as a booster. In (3), way cool can indeed be paraphrased as ‘so cool’,3 but the collocation with wrong in (4) may rather suggest that way functions as a maximizer, being (more or less) synonymous with absolutely or completely:
(3) Lisabette (Clapping her hands.) “She’s going to be in the play. This is way cool!” (To Casey) “Isn’t that cool?” Casey (Smiling, but a bit reserved): “Yes, cool.” [COHA, Playscript of Anton in Show Business, 1999]
(4) “You got how much money of your own?” Matt stood up, hat in hand, and said mournfully, “Only three hundred and eighteen dollars, so I guess I’ve got a lot of gall to take up your time. You wouldn’t be interested in my little stake.” “Sit down, man, sit down,” boomed the contractor and waved his seegar in an expansive gesture. “You’re wrong, way wrong” [COHA, Love is Eternal, 1954]
3It typically modifies an adjective in the absolute (see (1)-(4)) or the comparative degree, as in (5), or an adjectival head already pre-modified by (amplifier) too (OED s.v. way adv.3), as in (6):
(5) In his eyes there’s this look. Like he was seeing something different, something important, something way bigger than Fairhope. [COHA, Doing Alien, 1994]
(6) “Don’t go out in this, Beck,” Lucy said. “Hit’s way too bad a night.” [COHA, The Voice at the Back Door, 1956]
4Way is thus quite specific within the class of amplifiers. It differs from very, the prototypical booster, which cannot modify an adjective in the comparative degree or precede the sequence <too ADJ[ECTIVE]> (*very cooler; *very too cool) (Quirk et al. [1985: 473]). Way is also distinct from amplifier much which does collocate with adjectives in the comparative degree (much longer; much more convenient) or adjectives pre-modified by too (much too expensive), but not adjectives in the absolute degree (e.g. *I must have read the directions much wrong). Way can also modifiy an adverb, a verb, or a noun, as in (7), (8) and (9), respectively:
(7) “You move reality generator too far,” he said amiably. “Mess up everything way bad.” [COHA, The Attack of the Ignoroids, 1999]
(8) “You don’t think I’ll do it,” she had told him as she raced through another set of sit-ups. “You way underestimate me.” [COHA, Assassins, 2000]
(9) When we recorded it originally I doubled up the drums and it sounded way Gary Glitter, way Clash. [OED s.v. way adv.3, 1990 21 July 14/6]
5Intensification in the English language is a very popular topic of research. Earlier studies have focused, among other issues, on synchronic structural, semantic and sociolinguistic properties of various individual intensifiers in several present-day varieties of English (e.g. Stoffel [1901]; Bolinger [1972]; Bäcklund [1973]; Altenberg [1991]; Paradis [1997], [2000], [2001]; Stenström [2000]; Ito & Tagliamonte [2003]; Tagliamonte & Roberts [2005]; Macauley [2006]; Xia & Tao [2007]; Tagliamonte [2008]), or explored the origins of their intensifying function, often accounted for in terms of grammaticalization (e.g. Peters [1992]; Adamson [2000]; Lorenz [2002]; Nevalainen & Rissanen [2002]; Méndez-Naya [2003], [2012]; Buchstaller & Traugott [2006]; Traugott [2008a], [2008b]; Breban & Davidse [2016]). It has been shown that degree adverbs, especially boosters, develop from items or constructions performing various types of modification, such as manner adverbs (e.g. fairly, pretty), spatial adverbs (e.g. right, downright), quantitative adverbs (e.g. much), qualitative adverbs (e.g. terribly), emphasizers (really), taboo/swear words (e.g. bloody) (see e.g. Fettig [1934]; Peters [1994]; Nevalainen & Rissanen [2002]; Napoli & Hoeksema [2009]; Mendéz-Naya [2006], [2008], [2014]).
6Interestingly, the amplifier way is said to be descendent of away (Kirchner [1955]; Bolinger [1972]; Méndez-Naya [2012: 369]; King [2016]; see also OED s.v. away adv. 9b) which belongs to the category of spatial adverbs. However, there has been – to the best of my knowledge – no diachronic corpus-based study conducted to investigate its origins. In an attempt to lay the groundwork of a general theory that can explain why a given source word grammaticalizes as one type of intensifier rather than another, King [2016] focuses on three intensifiers, including way, which were not chosen randomly but “specifically because they are transparently related to source words that have spatial meanings” (King [2016: 3]). King [2016: 3] provides the following examples as evidence of the origin of way as a reduced form of spatial away:
(10) Mitchell, that letter DeWitt wrote you, has gone away – way down yonder.
(11) Such a lovely blue sky away, way up so high!
- 4 In future research the study of the amplifier meaning of way could be extended to verbal and (non s (...)
7While I consider that the spatial reading illustrated in (10)-(11) may indeed come into play in the semantic development of amplifier way, I propose that the spatial meaning of away is not the only semantic source involved. This contrasts with most case studies on language change which “generally focus on just one construction, drawing straight lines between a construction and a single historical ancestor” (Van de Velde et al. [2015: 1]). In line with a more recent view that more than one source construction may be involved in language change, whether the latter be traced back to independent constructions with their own history or to different uses of the same construction (Van de Velde et al. [2015: 1-2]; see also Breban & De Smet [2019: 879]), I suggest that multiple sources are at the origin of the development of the degree meaning of way. My assumption is based on an extensive corpus-based analysis of the functional uses of several constructions involving (a)way in American English that may have shaped over time the development of the degree meaning illustrated in (1)-(6) above. In the present article, I will focus on patterns involving both way in various uses and an adjectival head, e.g. <way ADJ>, <(in) every way ADJ>, <a long/good/great way ADJ>.4 The data extracted from the Corpus of Historical American English suggest that (i) the role of the spatial meaning in the semantic development of amplifier way may not only be related to the spatial adverb away but also to other spatial constructions such as <a long/good/great way(s) PREP[OSITION]/ADV[ERB]spatial> and <a way(s) PREP/ADVspatial>, and that (ii) other source constructions need to be taken into account, i.e. the originally manner constructions <ADJ (in) every way> and <(in) every way ADJ>. It will thus be assumed that way followed various developmental paths which may have interlocked at some point.
8This will lead me to address the related issue of the mechanism(s) and causes of change involved in the semantic-pragmatic development of the amplifier way. I will propose that the rise of the degree meaning may result from the interplay between several mechanisms or causes of change, i.e. metaphorization, pragmatic inferencing, and also analogization, i.e. the mechanism related to analogical thinking (see e.g. Traugott [2011: 25]), based on speakers’ recognition of similarities with other extant degree modifier constructions, most prominently the degree modifier construction <far ADJ> (far better, far different).
9The framework of constructionalization (Traugott & Trousdale [2013]) will provide the main background for the discussion. Constructionalization is a basic concept in a Diachronic Construction Grammar approach to language change (Noël [2007]) which came about in the spirit of rethinking grammaticalization in constructional terms. While grammaticalization refers to “that part of the study of language change that is concerned with such questions as how lexical items and constructions come in certain linguistic contexts to serve grammatical functions or how grammatical items develop new grammatical functions” (Hopper & Traugott [2003: 1]), constructionalization is defined as the rise of form-meaning pairings, in other words ‘constructions’ (see e.g. Goldberg [1995], [2006]; Fillmore & Kay [1997]; Croft [2001]), out of previously non-aligned material (Traugott & Trousdale [2013: 22]). Constructionalization follows multiple small constructional changes of both form and meaning, the outcome of which is the creation of a formnew–meaningnew pairing (see e.g. Traugott [2015: 51]). Grammatical constructionalization encompasses much of what has been discussed in the grammaticalization literature, but it goes beyond it to consider more far-reaching morphosyntactic changes. Indeed, another key concept in Diachronic Construction Grammar is the concept of taxonomic hierarchy. Constructions operate in a network which involves inheritance from the more abstract to the less schematic constructions. Traugott [2008a: 30], [2008b: 236] has distinguished the following constructional levels, with different degrees of schematicity:
(i) macro-constructions: highly abstract, schematic constructions
(ii) meso-constructions: sets of similarly behaving constructions
(iii) micro-constructions: individual construction types
(iv) constructs: empirically attested tokens of micro-constructions
- 5 For example, the Ditransitive Schema has several subschemas, including the Cause-not-receive subsch (...)
10In later work, Traugott & Trousdale [2013] use the terms schema and subschema instead of macro- and meso-construction,5 and Traugott [2019] makes a distinction between a micro-construction, a micro-schema and a more abstract schema:
A “micro-construction” is a low-level substantive-type construction such as after all in its discourse-marker use, a micro-schema is a low-level substantive-type construction that has a substantive part and one or more open slots, e.g. all but X, what is X doing Y? Higher-level schemas consist of abstract slots (e.g. ditransitive SUBJ V OBJ1 OBJ2 such as I gave her a book); how many higher-levels schemas are posited depends on the level of systemic generalization that the researcher is interested in (e.g. ditransitives are a subset of transitives) (Traugott [2019: 127]).
- 6 The way-construction referred to in this paper differs from the way-construction illustrated in she (...)
11In this paper I will adopt a constructionalist perspective on language change which supports the idea that pattern matching is an important factor of change, because construction grammar highlights memberships of sets (for constructional analyses of other degree modifier constructions, see e.g. Traugott [2019: 132-138]; Trousdale [2012: 178-186]). I will discuss analogization and paradigmatization, i.e. alignment with other constructions in an extant schema (Traugott [2019]; Traugott & Trousdale [2010: 38]), as potential key factors of change in the development of the micro-schema <wayamp[lifier] ADJ>, possibly acting in concert with metaphorization and inferencing. Considering the topic of the present issue, the main focus in this paper will be on the diachronic semantic changes leading to the emergence of the amplifier way-construction.6 It will still occasionally address the question of morphosyntactic change(s) in order to support an analysis in terms of constructionalization.
12The organization of the paper is as follows. Section 1 provides information about the data and methodology. In Section 2 I examine corpus data concerning various micro-schemas involving way which may be at the origin of the degree meaning of the amplifier, and other micro-schemas in the network of degree modifiers in English that may have shaped this meaning. Section 3 provides a summary of the findings and some conclusions which enable us to couch the development of the amplifier way-construction in terms of grammatical constructionalization.
- 7 Labov [1994: 11] notes that “historical documents survive by chance, not by design” and that they “ (...)
- 8 To the best of my knowledge, there is, however, no corpus of historical American English that cover (...)
- 9 Jucker [2018] observes that fictional representations of spoken language are not substitutes for ev (...)
13The investigation of the historical development of amplifier way in Section 2 will rely on a dataset collected from the Corpus of Historical American English (COHA) which contains 385 million words of American English from 1810-2009, grouped by decades. In addition to the “bad-data problem” (Labov [1966: 100])7 which is a well-known issue in historical linguistics, there arises the issue of the type of material contained in the corpus. COHA covers only written genres, i.e. fiction, magazines, newspapers and non-fiction books. It is thus restricted in the type of material it contains that may be relevant for the present study of an amplifier which is more characteristic of spoken than written (American) English. Since conversations as such or other forms of spoken language are not included, it could be considered difficult to reconstruct the diachronic semantic path(s) of change leading to the present-day uses of the amplifier way on the basis of written data only.8 However, the written data in COHA include represented conversation in fiction. Although they differ from the speech language users produce in spontaneous conversation, they give us “considerable insight into interactional language use” (Traugott [2019: 132]; see also Jucker et al. [1999]; Culpeper & Kytö [2010]),9 thus giving credit to the present attempt to formulate hypotheses as to the origins and semantic development of amplifier way on the basis of data extracted from COHA. The data collected were not restricted to a particular genre so as to cover as fully as possible the history of the different micro-constructions or schemas that may have played a part in the semantic development of the micro-schema <wayamp ADJ>. Note that the genre and subgenre balance in COHA stays almost identical from decade to decade.
14As mentioned earlier, the investigation of the amplifier uses of way in PDE was restricted to constructions in which way collocates with adjectival heads, whether in the absolute or the comparative forms, or premodified by too (see Section 2.1.). For historical data, I first relied on the OED’s entries for way, and selected all the sequences which could have had a semantic bearing on the historical development of the micro-schema <wayamp ADJ>. What were considered to be relevant patterns were:
15(i) sequences in which an adverbial phrase including way modifies an adjective and expresses high degree, i.e. <(by) a long way ADJ>, <a good/great way ADJ> and <all the way ADJ> (OED s.v. way n.12d), as in (12):
(12) I’m sure it is a long way better to keep yourself to yourself. [COHA, The Squire of Sandal Side, 1886]
16(ii) sequences in which adverbial way modifies a spatial preposition or adverb (OED s.v. way adv.2a and 2b; s.v. away adv. 9b), i.e. <wayadv PREP/ADVspatial> ‘at or by a considerable distance, far’. An example is provided in (13):
(13) I’b been hunting you, eber since yesterday mornin’, way up to de Skippack creek, sixteen miles from here, as true as my name am Charles de Fust. [COHA, Herbert Tracy, or, The Legend of the Black Rangers, 1844]
17The pattern <wayadv PREP/ADVspatial> was thought to be worth investigating because in both this pattern and the amplifier micro-schema, way has a premodifying function and expresses high degree or some idea related to it (‘a considerable distance’). Furthermore, as mentioned in the Introduction, it is sometimes reported in the literature that the amplifier way is descendent of spatial adverbial (a)way (Kirchner [1955]; Bolinger [1972]; Méndez-Naya [2012: 369]; King [2016]; see also OED s.v. away adv. 9b). The lack so far of corpus-based analyses meant to trace down its historical development from the spatial construction <wayadv PREP/ADVspatial> led me to put this hypothesis to the test. The data in COHA showed 17 different (originally spatial) prepositions or adverbs collocating with the adverbial modifier way in this specific pattern, namely up, out, down, over, before, back, past, under, off, beyond, after, ahead, behind, above, beneath, below and away.
18(iii) sequences in which an adverbial phrase including way modifies a spatial preposition or adverb, i.e. <a long/good/great way(s) PREP/ADVspatial> ‘at or by a considerable distance, far’ (OED s.v. way n.12b), and <a way(s) PREP/ADVspatial> in which a way(s) is a reduced version of <a long/good/great/little way(s), according to the OED (s.v. way n.12b; s.v. way n. and int.P3). These are illustrated in (14) and (15):
(14) [T]he Israelites during the thirty-seven years of their journeying from Mount Sinai may have gone by those tracts of country in which the waters from Horeb could follow them, till in the thirty-ninth year of the Exodus they came to Ezion-gaber (Num. xxxiii. 36), which was a part of the Red Sea a great way down the Arabian side. [COHA, Autobiography of Frank G. Allen, Minister of the Gospel and Selections from his Writings, 1887]
(15) “Mr. Barnes -- he is with you?” It was Constance that spoke. “Yes; but --” “Where is he?” “We left him a ways down the road and --”. The sound of a horse’s hoof beats in front of the manor, breaking in on this explanation, was followed by hurried footsteps upon the porch. [COHA, The Strollers, 1902]
19There are reasons to believe that if the construction <wayadv PREP/ADVspatial> could possibly have favoured the semantic development of <wayamp ADJ>, as suggesteg by King [2016] (see also Section 2.2.1.), then the other spatial patterns <a long/good/great way(s) PREP/ADVspatial> and <a way(s) PREP/ADVspatial> could have had a similar influence. Note for that matter that the above-mentioned constructions of degree or extent <(by) a long way ADJ>, <a good/great way ADJ> and <all the way ADJ> are thought to be figurative uses of <a long/good/great way PREP/ADVspatial> (OED s.v. way n.12d). The same prepositions or adverbs which were found as collocates of adverbial way in the pattern <wayadv PREP/ADVspatial> were searched for as potential collocates of a long/good/great way(s) and a way(s) in the patterns <a long/good/great way(s) PREP/ADVspatial> and <a way(s) PREP/ADVspatial>.
- 10 One reviewer suggested that I also consider the pattern <every which way ADJ, but no tokens were re (...)
20For purposes of historical reconstruction, the following constructions were also examined:
(i) the patterns <(in) every way ADJ> and <ADJ (in) every way> which were encountered when investigating some of the above-mentioned constructions. Manual pruning was required to eliminate false positives (e.g. privileged in every way imaginable in which the phrase in every way does not modify the adjective imaginable). From a semantic-pragmatic viewpoint, the data show that both <(in) every way ADJ> and <ADJ (in) every way> ‘ADJ in every respect’ can give rise to a meaning of high degree (see Section 2.2.3.), hence their relevance for the present study.10 One example is provided in (16):
(16) In George Manning the physical and intellectual man had been developed in rare harmony. He was taller and larger every way than his brother Henry […]. [COHA, Evenings at Donaldson Manor Or, The Christmas Guest, 1871]
21(ii) a construction which shows functional and semantic similarities with the pattern <wayamp ADJ>, i.e. <faramp ADJ> illustrated in (17):
(17) That’s an old trick. The scheme I’ve been thinking of is far easier, and I’ll confide it to you. [COHA, Accusation, 1817]
22This pattern is attested earlier than the amplifier way-construction, and it is thus considered to be a quite plausible analogical model.
23Though COHA is a large-scale resource for studying language change in English, it is not without its limitations. One of the shortcomings which complicated the task and increased the time required to complete it is part of speech tagging which sometimes turned out to be inaccurate. Davies [2012], the creator of COHA, acknowledges the problem and remarks that “tagging for nearly 100,000 ‘problematic’ types from the 1800s has been manually reviewed and corrected.” Still, some of the searches conducted for the present paper show that further efforts need to be made in that direction. For instance, part of speech tagging turned out to be an issue when searching for adverbial way in collocation with spatial prepositions or adverbs (see Section 2.2.). A wide search for way tagged as an adverb in collocations with a preposition returned fewer results for each of the possible prepositional collocates (e.g. down, up; see above) than a more specific search for adverbial way collocating with each of the individual prepositions isolated. But even the latter sometimes yielded fewer results than those obtained when conducting the same search without tagging way as an adverb. I therefore decided to keep tagging to a minimum in this particular case, which required extensive manual pruning of the whole dataset to eliminate numerous irrelevant data (e.g. nominal uses of way followed by a spatial particle as in on the way down).
24I relied upon the original division of the corpus into individual decades to examine step-by-step developments from 1810 to 2009 of each of the contructions that were likely to be relevant to the present study, namely <wayamp ADJ> – whether the adjective be in the absolute, in the inflectional or periphrastic (more/less) comparative form, or premodified by too (see Section 2.1.) – <a long/good/great way ADJ>, <all the way ADJ>, <wayadv PREP/ADVspatial>, <a long/good/great way(s) PREP/ADVspatial>, <faramp ADJc[o]mp[arative]> and <faramp too ADJ>. Table 1 provides the raw figures and, underneath, the relative frequencies of the intensifying construction(s) under study and the patterns that might have shaped its development.
Table 1. Raw figures and relative frequencies of the amplifier micro-schema <way ADJ> and the constructions that may be involved in its semantic development
25The patterns <a long/good/great way ADJ> or <all the way ADJ> were finally not included in Table 1 because the search in COHA of the phrases a long/good/great way and all the way as modifiers of any adjectival collocates yielded very few, if any, results – eight tokens of the pattern <a long way> and three of <all the way ADJ>, which are illustrated in (18) and (19), but none of <a good/great way ADJ>.
(18) The old-fashioned motive-power of the rod and ferule has made many a good scholar and true lover of learning, although the ideal educator is one who, without swerving from his disciplinary course, yet contrives to make it all the way alluring. [COHA, North American Review, 1871]
(19) If you plan for to-morrow, plan a bit of pleasure with it; that’s a long way better than expecting a headache. [COHA, St. Pierre’s Studies of Nature, 1836]
26The paucity of data concerning these constructions led me to discard them as potential sources of development of the construction <wayamp ADJ> (examined in Section 2), at least on the basis of the data retrieved from COHA. Still, it could be rather surprising considering the existence of another pattern of extent based on the nominal use of way, i.e. <half-way ADJ> (OED s.v. half-way A.adv). The latter was hit upon when searching for <wayamp ADJ>. Though it expresses moderate, not high degree, as shown in (20), it could support the idea that nominal uses of way in phrases of extent did have the semantic potential to give rise to the amplifer use of way.
(20) I don’t know where he is most of the time now and I lie awake telling myself what a fool I am to care -- but when I hear his key in the latch everything in the world is all right again. I go to sleep half way happy -- trying to make myself think this one will pass too. [COHA, When Ladies meet, 1932]
27Tokens of <half-way ADJ> in COHA are, however, limited to about 40. A possible conclusion is that there may have been signs of a budding development of a degree meaning out of such constructions of extent based on the nominal use of way, but the low frequency and productivity of these patterns may have prevented them from developing further.
28I will first present the data retrieved concerning the amplifier use of way in collocation with adjectival heads, and then address the issue of the semantic paths of development it may have embarked on.
29The amplifier use of way was investigated by running search queries for any element tagged as an adjective directly following adverbial way. This yielded tokens like way cool with the base form of the adjective in the absolute degree, and way better ‘better by a great amount’ with the inflected -er form of the adjective in the comparative degree, which drove the search for two variant patterns including periphrastic comparatives, i.e. <wayamp more ADJ> and also <wayamp less ADJ>. Finally, relying on the OED’s entry (s.v. way adv.3), I also searched for the pattern <wayamp too ADJ> in which amplifier way modifies an adjectival phrase composed of an adjectival head modified by another amplifier (too). The results show that amplifier way collocates with (a few) descriptive (e.g. high, short) and mainly evaluative adjectives such as cool, good (better/worse), interesting, handsome, expensive, silly, harsh, painful, cheap(er), smart(er), wrong, awesome, cynical, overpriced, overrated. Many of the collocations occur only once in the corpus (e.g. way interesting, way silly, way catastrophic) while others (e.g. way big(ger), way too big, way cool(er), way too easy) are more frequent – up to 18 tokens of e.g. way (too) big(ger).
- 11 The OED dates it back much later to 1965.
30Green’s Dictionary of Slang dates back the amplifier use of way in American English to the late nineteenth-early twentieth century. One of the examples illustrating its use as “general intensifier” ‘very, extremely’ is provided in (21):11
- 12 The earliest example provided in Green’s Dictionary of Slang actually dates back to 1888. It is not (...)
(21) Got a heavy date and I’m way late. [Green’s Dictionary of Slang s.v. way adv., 1932]12
31The earliest attestations of the construction <wayamp ADJ> in COHA also go back to that time, as shown in (22) and (23):
(22) You’ll excuse me for sayin’ you’re rather delayin’ your chance to insure your own [safe return]! For we’re way worn and weary, your style isn’t cheery, we’ve had quite enough of your game. [COHA, The Writings, 1896]
(23) Pass over them documents for Cherokee Hall, an’ don’t hold out nothin’ onto us. We-alls is’ way too peevish to stand any offishul gaieties to-day. [COHA, Wolfville, 1897]
32Note, however, that the search for the variant constructions <wayamp ADJcmp> and <wayamp too ADJ> returned only very few hits until the late twentieth century. One example is provided in (24):
(24) Miss Devlin in school had told them all that the sky was so many many miles away that everything in the sky looked very small, when it was really very very big, and the sky was way bigger than the earth. [COHA, Random House 1981, 1936]
33The frequency of the construction <wayamp ADJabs[olute]> is slightly higher over the same time period, but all in all the variant amplifier way-constructions show a low token frequency until the 1990s, as illustrated in Figure 1.
Figure 1. Relative frequency of the amplifier way in pre-adjectival position in COHA (per million words)
34Early tokens could nonetheless be regarded as innovative precursors of a construction which then took off in the late twentieth century, especially in the form of <wayamp too ADJ>, and, to a lesser degree, <wayamp ADJcmp>. Note for that matter that there are only two tokens of the pattern <wayamp less ADJ> in COHA; by contrast, the search for <wayamp more ADJ> yielded 28 results. In what follows (see Section 2.2.3., in particular) I will look into the kind of contexts which may have favoured the development of the specific constructions <way ADJ-er>, <way more ADJ>, and <wayamp too ADJ>.
35What follows is an attempt to reconstruct the paths of (semantic) development of the amplifier way in American English from source constructions involving different uses of way. For this purpose, I discuss some of the historical data presented in Table 1 (in Section 1), and examine the contexts of use of the constructions that may have shaped the degree meaning until the 1990s, when the amplifier function of way seemed well-established (see Figure 1 in Section 2.1.).
- 13 The variant constructions <wayamp ADJabs>, <wayamp ADJcmp> and <wayamp too ADJ are subsumed under t (...)
36The starting point of this historical investigation was the possibly spatial origin of amplifier way which is said to derive from adverbial (a)way (Kirchner [1955]; Bolinger [1972]; Méndez-Naya [2012: 369]; King [2016]). King [2016], in particular, seems to establish a historical connection between the earlier (originally and chiefly American English) spatial pattern <wayadv PREP/ADVspatial> (see OED s.v. way adv.2a and 2b; s.v. away adv.9b) illustrated in (25), and the more recent amplifier micro-schema <way ADJ>:13
(25) There’s Patrick, my brother, way over in Ireland -- the last time I saw him I wasn’t taller than that butter firkin. [COHA, Little Ferns for Fanny’s Little Friends, 1850]
37The modifying function of way in both constructions supports this hypothesis – in part at least. Evidence will be provided later of an intermediate stage in the evolution of the micro-schema <wayadv PREP/ADVspatial> which could account for the further expansion (Himmelmann [2004]) of adverbial way to other types of modifiers than particles, especially to adjectival collocates.
38Before addressing this issue, one needs to look into the origin of adverbial way in the construction illustrated in (25) to understand its possible historical semantic link with the micro-schema <wayamp ADJ> (see infra for a discussion of the kind of mechanism of semantic change which might be involved). The OED (s.v. way adv.1) observes that adverbial way is probably an aphetic form of away, i.e. the adverb has lost the unstressed vowel at the beginning. As shown in (26), away can occur in the same type of micro-schema as way in (25):
(26) [I]ts a present to me, says I from Starks &; Co., away up in York State, on the Black River. [COHA, The Select Letters of Major Jack Downing, 1834]
39In this construction, away mainly “modif[ies] adverbs of distance or time, as away back, away behind, away down, away up, emphasizing the extent, remoteness, etc., involved” (OED s.v. away adv.9b). In the micro-schema <wayadv PREP/ADVspatial>, way ‘at or by a considerable distance, far’ (OED s.v. way adv.2) is thus indeed very close to away in the construction <away PREP/ADVspatial>. Evidence is provided in (27) and (28) which supports the idea that <wayadv PREP/ADVspatial> derives from <away PREP/ADVspatial> as they show a possible intermediate stage in the phonological erosion of away which takes the form ’way:
(27) A wicked man was coming to take little Harry away from his mother, and carry him ’way off in the dark. [COHA, Uncle Tom’s Cabin,1852]
(28) “Now you see,” said the boatswain, “just so sure as you have gentle breezes from the south’ard, you’ll have a thundering Levanter at the back of ’em.” “Yes, yes,” said a tar, “I know that to my sorrow. I was up the Straits last v’y’ge, ’way up to Smyrna and Zante, after reasons, and we ketch’d one of these thundering Levanters, and was druv’ way to h -- ll, away up the Gulf of Venus.” [COHA, An Old Sailor’s Yarns, 1835]
40The fact that the full form away (away up the Gulf of Venus) in (28) occurs alongside the phonologically reduced one ’way (’way up to Smyrna and Zante) is particularly enlightening. In (29) and (30), the construction <way PREP/ADVspatial> is probably no longer interpreted as a phonological variant of <away PREP/ADVspatial>. A constructional change appears to have taken place at the morpho-phonological level which makes it possible to analyze <way PREP/ADVspatial> as a new construction, with way taking on a new function, namely modifier of spatial particles:
(29) “Oh mamma, see, see!” cried Grace, more than half frightened at the sight; “papa has gone away, way out, and Maxie with him.” [COHA, Elsie at Nantucket, 1884]
(30) “Where are the purple notches?” the doctor asked. “See yonder, away, way off?” Thaine pointed toward the misty southwest horizon where three darker curves were outlined against a background of pale purple blending through lilac up to silvery gray. [COHA, Winning the Wilderness, 1914]
- 14 Note, however, that some (few) figures in Table 1 are biased because the pattern <wayadv PREP/ADVsp (...)
41This pattern, which dates back to the 1830s according to the data in COHA (see Table 1 in Section 1; see also OED s.v. way adv. 2a and 2b), may result from repetition of the pattern <’wayadv PREP/ADVspatial>. Table 1 in Section 1 shows a significant increase in its frequency from the second half of the nineteenth century onwards.14 The fact that it was well-entrenched by the time way developed into an amplifier in the 1990s supports the assumption that this spatial construction could have been one of the sources of development of the micro-schema <wayamp ADJ>, at least in case of a semantic relationship between the two (see infra). As mentioned in Section 1, 17 spatial particles were found as collocates of way in this micro-schema.
42The same spatial particles were found in collocation of the phrases a long way(s), a good way(s) or a great way(s) in the other spatial pattern investigated, i.e. <a long/good/great way(s) PREP/ADVspatial>, which is illustrated in (31):
(31) The minister grew comparatively calm. His eyes, however, were soon greeted by a little, glimmering light, which, at first a long way off, was approaching up the street. [COHA, The Scarlet Letter, 1850]
43As mentioned above, the OED observes that the noun phrases a long way(s), a good way(s), a great way(s), which are used adverbially, are found in later use in the reduced form a way(s) (OED s.v. way n. 12b), just like a little way(s). Example (32) is one of the tokens retrieved of adverbial a way(s) in COHA, from the second half of the nineteenth century onwards:
(32) So this is Bob Morton’s law-office, is it? How Bob has got up in the world. Why! when I left home, his office was a way up in the loft of an old hulk of a house, down by the wharf. [COHA, The Tempter, 1871]
44There is no specific mention in the OED that the phrase a way expresses a considerable rather than a little distance – as a little way(s), for instance, does – but the OED seems to suggest that this is so by referring in the same entry (OED s.v. way n. 12b) to the use originally in American English – since the mid-nineteenth century – of the phrase a ways which indicates ‘a considerable or not insignificant distance’ (OED s.v. way n. P3b). Note for that matter that about 40% of the patterns <a way(s) PREP/ADVspatial> found in COHA clearly express a not insignificant distance through the use of quite as modifier of a way, as shown in (33):
(33) He watched them until he had floated quite a way below, when he began to hope that they had given up their designs upon him, and he might make his way back to his friends upon the island in safety. [COHA, Adrift in the Wilds, 1887]
45But it is difficult to consider that the construction <wayamp ADJ> or even the micro-schema <wayadv PREP/ADVspatial> originate from the pattern <a way(s) PREP/ADVspatial> since it shows an overall low token frequency, with only scattered uses until the 1880s at least (see Table 1 in Section 1), while the early uses of the micro-schema <wayadv PREP/ADVspatial> date back to the 1830s. However, uses of <a way(s) PREP/ADVspatial>, together with the micro-schema <a long/good/great way(s) PREP/ADVspatial> from which the former is derived and which has a much higher token frequency (see Table 1), may have reinforced the heightening effect of <wayadv PREP/ADVspatial>.
46What needs to be addressed now is the question of the mechanism of change from the spatial reading ‘at or by a considerable distance, far’ (OED s.v. way adv. 2a and 2b) of the construction <wayadv PREP/ADVspatial> to the degree meaning of <wayamp ADJ> in PDE. King’s [2016] study provides a preliminary answer. The author focuses on the change from certain types of adverbs to certain types of intensifiers, including (a)way. What is of interest for the present paper is not so much King’s [2016] discussion of what determines that way has become an amplifier and not, e.g., a downtoner (Quirk et al. [1985: 590]), as the fact that the author considers the amplifier way to be derived from adverbial spatial (a)way, and analyzes the semantic change in terms of metaphorical transfer. On the basis of the path image-schema used to depict the semantics of away, and the metaphorical interpretation of a scale in terms of a path (Johnson [1987: 122-123]), King argues that an increase in distance (away) is reinterpreted as a rise in degree. Example (34) leaves no doubt as to the possibility for the string <way PREP(/ADV)> to develop a (metaphorical) degree meaning:
(34) In Hollywood colossal is several degrees under supercolossal and supercolossal is way beneath terrific. [COHA, Murder on the Face of It, 1940; emphasis mine]
- 15 As observed in Section 2.1., the very first example provided in Green’s Dictionary of Slang of the (...)
47Now the question is what contexts of use may have triggered host-class expansion (Himmelmann [2004: 32]), i.e. construction-internal expansion of the class of elements adverbial way can collocate with. More precisely, what may have triggered the kind of constructional change (see e.g. Traugott & Trousdale [2013]; Traugott [2015]) consisting in the expansion of adverbial way to adjectival collocates in the pattern <wayamp ADJ>? Quite importantly, the data in COHA show that the semantic and pragmatic contexts of use of the micro-schema <wayadv X> – whereby X is any type of modified item – are expanded. In particular this semantic-pragmatic context expansion (Himmelmann [2004: 32-33]) takes the form of a gradual metaphorical change in the semantics of the prepositional or adverbial collocates of way in the pattern <wayadv PREP/ADV>. While these collocates take on an original spatial – whether allative, i.e. “movement towards a location”, or locative – meaning in early uses of the construction, as in “papa has gone away, way out” in (29) above, they start to show not only temporal but also even more abstract meanings which sometimes make the string <wayadv PREP/ADV> semantically akin to the pattern <wayamp ADJ>,15 as in (35) to (37), and may thus pave the way for the expansion of the construction <wayadv X> to adjectival collocates:
(35) “In such a matter as this, the opinion of an expert is everything. I am going to have one of the principal musicians of the town go and try them all, and tell me which we ought to have.” “And while he’s about it,” said Alice, “you might ask him to make a little list of some of the new music. I’ve got way behind the times, being without a piano so long.” [COHA, The Damnation of Theron Ware, 1896]
(36) It will surprise everybody and cause big drop […]. A drop only of about six points. This is one of the highest averages for the (September) month ever known. Is way above the ten-year average. [COHA, New York Times, 1905]
(37) If we got outside we would freeze to death, as it was way below zero. [COHA, New York Times, 1927]
48The (originally spatial) prepositions behind in (35), above in (36) and below in (37) take on a metaphorical meaning, so that the phrases way behind the times, way above the ten-year average and way below zero can be paraphrased as ‘way outdated’, ‘way superior to the ten-year average’ and ‘way inferior to zero’, respectively. Interestingly, in the informal phrase way off base ‘completely mistaken’ in (38), it is the string off base, which is originally an instantiation of <preposition + noun (phrase)>, which receives a metaphorical interpretation, so that the meaning of the whole is akin to adjectival meaning (‘mistaken’):
(38) And every time he fails he takes it out on me because I didn’t get to Elmer sooner. But you can see perfectly well he’s way off base, trying to make out I could have done a thing to prevent what happened. [COHA, The Aggravation of Elmer, 1955]
49Uses of the construction <wayadv PREP/ADV> with various abstract metaphorical interpretations of the prepositional or adverbial collocates, or the string <preposition + noun (phrase)>, such as those illustrated in (35) to (38) above, occur as from the late nineteenth-early twentieth century, and could thus be considered as intermediate stages in the development of the more recent adjectival pattern <wayamp ADJ> out of the original spatial construction <wayadv PREP/ADVspatial>.
50The data in COHA show the same kind of constructional change through the semantic-pragmatic expansion of the construction <a long/good/great way(s) PREP/ADV>, which is illustrated in examples (39) to (42):
(39) When he was yet a great way off, his father saw him, and had compassion, and ran, and fell on his neck, and kissed him. [COHA, Female Scripture Biographies, 1817]
(40) You say the negro blood is mighty poor stuff, and the race is a long way behind ours. [COHA, What Answer? 1868]
(41) The day will come, I believe, when Mr. Eakins will be rated, as he deserves, far above the painters of mere pretty effects, and a good way above even men of similar artistic skill who devote themselves to less characteristic and less vital themes. [COHA, The New York Art Season,1881]
(42) “You know it all better than I do. I’m only trying to say that I get some kind of joy out of being near people like that, out of having some reason for doing anything that they do. That must be why I go to market like this, for of course there is no other reason.” “Poor kid,” said Mark, looking down at her with tenderness. “What a long way off base you are.” “And not you?” asked Jessie. “Certainly, sometimes. But I told you last night --” “Yes,” she said. “I remember.” “And I am much nearer to these people than you,” he said. [COHA, East Side, West Side,1947]
51While behind takes on its original spatial meaning in (39), behind in (40) and above in (41) do not; they are reinterpreted metaphorically, so that a long way behind and a good way above mean ‘way inferior (to)’ and ‘way superior (to)’, respectively. And just as “way off base” in (38) above was interpreted as ‘completely mistaken’, “a long way off base” in (42) expresses the same idea. Note, however, that the prepositions or adverbs which a way(s) collocates with take on mainly a spatial meaning, sometimes a temporal one, but not the kind of metaphorical reading permitted by e.g. a long way behind in (40) above. The assumption put forward is that although the constructions <a long/good/great way(s) PREP/ADV> and <a way(s) PREP/ADV> are less likely to have been direct sources of development of the micro-schema <wayamp ADJ>, they could still have had an indirect impact on it – both with literal and metaphorical readings of the particles – thus enabling to reinforce the heightening effect of the construction <wayadv PREP/ADV> via metaphorization again. But while it is the notion of increase in distance captured by away in the spatial construction <away PREP/ADVspatial> which is, by hypothesis, metaphorically reinterpreted in terms of high degree in the construction <wayamp ADJ> (see King [2016]), it is the notion of long distance denoted by the phrases a long/good/great way which is thus reinterpreted.
52When searching for the amplifier micro-schema examined in Section 2.1., I came across a pattern showing way as part of a phrase functioning as modifier of an adjective, i.e. <(in) every way ADJ>. The OED cites the use of nominal way in adverbial phrases such as (in) any way which indicate “a respect or particular in which things can be judged or compared” (OED s.v. n. 21), but not the specific phrase (in) every way. I set out to explore the latter as a potential source of development of the amplifier micro-schema because of formal and semantic similarities. I thus searched for the patterns <(in) every way ADJ> but also <(in) every way more/less/too ADJ> in order to establish potential parallels with the different forms of the amplifier patterns, i.e. <wayamp ADJabs>, <wayamp ADJcmp>, and <wayamp too ADJ> (see Section 2.1.). I also examined the pattern <ADJ in every way>. The searches returned a total of 604 tokens, fairly distributed (see Table 1 in Section 1) between the pattern <(in) every way (more/less/too) ADJ>, with the phrase (in) every way in premodifier position of the adjectival head, and the pattern <(more/less/too) ADJ (in) every way>. Note that just as I retrieved only two tokens of the construction <way less ADJ>, the search for the patterns <(in) every way less ADJ> and <less ADJ (in) every way> yielded only two results too. I will therefore no longer mention less when referring to the constructions <wayamp ADJcmp>.
53The examination of the whole dataset revealed that the original reading ‘in every respect (more/too) ADJ’/‘(more/too) ADJ in every respect’, which is illustrated in (43), could in specific contexts give rise to an inference (Traugott & Dasher [2002]) of high degree, as in (44):
(43) The whole cycle is given in the last twenty years of Republican rule here at home, when the representatives of the party resolved to hold on to office after they had done their work, for all sorts of patriotic reasons, but at bottom because they found it profitable and pleasant to hold on. They were forced to pay roundly for the privilege; the possession was assured to them only by concessions of the largest kind to importunate interests. This was bad, bad in every way; in morals, because the honor and profit of administering the people’s trust should be conferred by an honest majority of the people, and not bought of an interested class; as practical politics, because the ally is sure to sell out to the highest bidder whenever it becomes safe to do so. [COHA, Reform: The Democratic Machine, 1886]
(44) He had said, as plainly as he could look it, that he didn’t want me here; that I was only a trouble to him; that I made him unhappy by remaining; that he would be much better in every way if I were gone. [COHA, The Last Woman, 1909]
54In (43) the fact that the representatives of the party held on to office is said to be bad in two respects, i.e. “in morals” and “as practical politics”. Nevertheless, it could also be argued that “this was bad, bad in every way” also invites the inference that it was very bad, although this kind of implicature probably remains backgrounded in the utterance. By contrast, the high degree meaning seems to come to the fore in the context of (44), which could be described as a “bridging context” (Heine [2002: 86]), i.e. “a specific context giving rise to an inference in favour of a new meaning” so that the “target meaning [is] foregrounded”.
55What defines this type of bridging context is first of all the use of a gradable adjective, better in (44), in collocation with the phrase (in) every way. Note for that matter that the vast majority of the 175 adjectives collocating with (in) every way in the constructions <(more/too) ADJ (in) every way> and <(in) every way (more/too) ADJ> in COHA are gradable (evaluative) adjectives, e.g. good, bad, strong, beautiful, attractive, fine, great, wrong, uncomfortable, interesting, big, pleasant, poor, vicious, important, handsome, magnificent, stupid. One condition was thus met for the phrase (in) every way to turn into a degree modifier. Secondly, and quite relatedly, about 15% of the examples of the two constructions show the adjectival head being premodified by degree modifiers like so, very, completely, utterly, etc. This is the case in (44) where much intensifies the meaning of better. Another characteristic of a bridging context possibly giving rise to the meaning of high degree in the constructions <(more/too) ADJ (in) every way> and <(in) every way (more/too) ADJ> is the use of similar degree modifiers in the larger context of the constructions, as in (45) or (46):
(45) Is it so very wonderful that Abby should grow? To be sure -- certainly not -- she was very fair when I saw her last -- when I left this part of the world, I mean. Very -- So upright, and so graceful and free in her carriage... Free in her carriage? For a child, I mean -- so modest, and so remarkable in every way -- so attentive, so quiet. [COHA, Rachel Dyer: A North American Story, 1858; emphasis mine]
(46) The Place Vendome is small, surrounded by high houses, and the stately Column seems dwarfed by them. But for its historic interest, and especially that of the material employed in its construction, I should not regard it very highly. Far better placed, as well as more majestic and every way interesting, is the Obelisk of Luxor. [COHA, Glances at Europe in a Series of Letters from Great Britain, France, Italy, Switzerland &c. D, 1851; emphasis mine]
56Although the OED does not report the use of (in) every way as degree modifier, some examples do suggest that its high degree meaning may even have been semanticized at some point. In (47) and (48), the literal meaning of the phrase in every way is apparently not available:
(47) When you combine quality with style, you will be a better shopper. You will then have a garment which will be finer in every way as far as material and workmanship are concerned. [COHA, Your Clothes and Personality, 1937]
(48) The cabin stood in a valley, or canon, in the shadow of gigantic pine-trees, rising straight as a flagpole to the altitude of nearly two hundred feet. They were forest giants, impressive in their lofty stature, and Ben regarded them with wonder and awe. They were much smaller in every way than the so-called big trees to be found in the Calaveras and Mariposa groves. [COHA, The Young Explorer, 1880]
57In (47), the garment is considered to be finer only with respect to “material and workmanship”. The literal reading of “in every way” is clearly ruled out, and the idea is that the garment will be much finer if you combine style with quality. In (48), the meaning of the adjective small(er) makes it difficult, if not impossible, to think of different ways in which the pine-trees could be smaller “than the so-called big trees to be found in the Calaveras and Mariposa groves”. The only possible interpretation seems to be that “the pine-trees were much, much smaller”. These thus appear to be examples of Heine’s [2002: 97] “switch contexts” which are “incompatible with the source meaning”.
58Two syntactic features concerning the patterns <(more/too) ADJ (in) every way> and <(in) every way (more/too) ADJ> need to be discussed at this point. First, in the pattern <(in) every way (more/too) ADJ> illustrated in (49) below, the phrase (in) every way occurs in pre-adjectival position, and is thus syntactically closer to amplifier way than it is when occurring in postmodifier slot.
(49) All these enormities were new things to me. I had seen nothing, I had imagined nothing, so every way terrible as came within my notice under the squalid roof of this poor needle - woman. [COHA, Atlantic Monthly, 1865]
59This is not to say, however, that in the pattern <(in) every way (more/too) ADJ> the phrase (in) every way systematically expresses high degree and cannot be interpreted literally. In (50), for instance, the speaker/writer enumerates all the features that make a prince worthy of a princess:
(50) It is taken for granted that the prince should be in every way worthy of the princess: handsome, brave, highly born, virtuous and wise. [COHA, A Study of Two Worlds: Persephone, 1959]
60Still, the pre-adjectival slot may have facilitated the rise of a pragmatic inference of high degree, especially in very expressive contexts such as illustrated in (46) above in which the speaker/writer makes abundant use of degree modifiers in the immediate context. “Analogy with other similarly behaving degree modifier micro-constructions […] which were well established by that time” (Trousdale [2012: 182]) may thus have reinforced this type of inference (see Section 2.2.4. for a discussion of the role analogy is likely to have played in the development of the micro-schema <wayamp ADJ>).
61Secondly, in about 33% of the tokens retrieved the preposition in is omitted (see examples (46) and (49) above). Examples (51) and (52) illustrate the use of every way (without the preposition in) combined with its use in premodifier slot:
(51) Nothing would seem to be plainer, than that Aurelian spared Zenobia because she was a woman; because she was a beautiful and every way remarkable woman. [COHA, Letters of Lucius M. Piso, from Palmyra, to his Friend Marcus Curtius at Rome, Volume 2, 1837]
(52) The berries were not only much more numerous than the year before, but they were every way larger and finer. [COHA, Needle and Garden, 1865]
62One might assume that every way is a morphologically reduced version of in every way and that the phrase moved to premodifier slot over time, and then consider these as possible signs of constructional change. But the data show that every way was in fact more frequent than in every way in the early nineteenth century, and this continues until the late 1860s. The full form in every way then becomes in its turn much more frequent than every way both in <in every way (more/too) ADJ> and <more/too ADJ in every way>, and this situation continues until the 2000s. The time gap between the first half of the nineteenth century and the late twentieth century, when the token frequency of amplifier way took off, suggests that the phrase every way with the preposition in omitted probably had little impact on the syntactic development of the amplifier. However, the persistence of the high degree meaning associated with (in) every way both in premodifier and postmodifier slot could be considered as a sign of semantic inheritance. It is in fact noteworthy that the use of both every way and in every way in premodifier position falls dramatically in the early twentieth century, as shown in Figure 2, that is to say at the time when way started to be used as amplifier (see Figure 1 in Section 2.1.), and that there are in COHA no tokens at all of the pattern <every way (more/too) ADJ> and only 5 of <in every way (more/too) ADJ> over the 1980-2009 period. Recall that, as shown in Figure 1 (in Section 2.1.), the frequency rate of <wayamp ADJ> really takes off around the 1990s. This might suggest that it has replaced the phrase every way as premodifier of adjectival collocates.
Figure 2. Relative frequency of the pattern <(in) every way ADJ> in COHA (per million words)
63It is noteworthy that while the search for the specific pattern <wayamp ADJcmp> returned only four hits – with the inflected or periphrastic form of the comparative – up until the 1990s (see Table 1 in Section 1), the search for the pattern <(in) every way ADJcmp> did return many more results (40 tokens) before it fell into disuse in the 1980s. Since the pattern <(in) every way ADJcmp> may also have invited the inference that ‘if something or someone is more X in every way, it or s/he is X to a high degree’, one could assume that this specific pattern may have favoured the development of the construction <wayamp ADJcmp> as from the 1990s. By contrast, only two tokens were found of the construction <(in) every way too ADJ> in COHA, which means that this specific pattern cannot account for the rise and, later, the spectacular growth of the construction <way too ADJ> from the 1980s onwards. Therefore, other contexts of use of way in collocation with too will have to be looked into.
64Two other way-based patterns recorded in the OED might bear out the assumption that the amplifier way is, historically, semantically related to the phrase (in) every way. These are <in no way ADJ> and <no way(s) ADJ> originally ‘in no manner or not at all ADJ’ (OED s.v. way n. and int. P6; no way adv.A), illustrated in (53) and (54):
(53) The scenery was not without beauty; wooded hills of gentle slope every where bordered the river; and its islands and alluvions, still of considerable extent, are no way inferior to the lands of the Ohio. [COHA, Nuttall’s Journal, 1827]
(54) They were burned in the furnace downstairs without being counted or checked up, and without witnesses save a clerk named Frank D. Leland of Olean, who was in no way legally responsible for their care. [COHA, New York Times, 1907]
- 16 I did not conduct a quantitative analysis of the patterns <(in) no way ADJ> and <ADJ (in) no way be (...)
65Although the specific meaning of (in) every way obviously differs from that of (in) no way, the two patterns share an original manner reading as well as a degree function. This common feature supports the hypothesis that the development of <wayamp ADJ> was influenced, in part at least, by the existence of other way-constructions expressing degree which preceded it in time. A quick look at the first ten adjectives collocating with (in) no way in COHA16 suggests that the pattern <(in) no way ADJ> is used as from the early nineteenth century, just like the pattern <(in) every way ADJ>. Among the tokens retrieved there are also a few examples of the pattern <no ways ADJ>, which also date back to the nineteenth century, according to the data in COHA. One example is provided in (55):
(55) “Make yourself easy, madam,” said he, “his lordship is no ways dangerous, I know him too well, and he knows me full as well.” [COHA, The Tennessean, 1827]
- 17 As one reviewer observed, it may be interesting to draw a parallel between, on the one hand, the us (...)
66In examples (53) to (55) no way(s) ‘not to any extent, not at all’ functions as a minimizer. It is the negative no which turns no way(s) into a negative maximizer (Quirk et al. [1985: 597]) scaling downwards, in other words a minimizer. One could therefore assume a historical connection between the older negative maximizer no way(s) and the more recent amplifier way which scales upwards, at the opposite end of the scale, with the constructions <(in) every way ADJ> and <ADJ (in) every way> half-way along the path, as semantic precursors of <wayamp ADJ>.17
67In this section I propose that analogization, i.e. the mechanism of analogy leading to the development of a new micro-schema on the basis of an exemplar model, is most probably involved in the development of the amplifier way in American English. Recently, analogy-driven changes have received a great deal of attention (see e.g. Fischer [2007], [2008], [2009]; De Smet [2013]; De Smet et al. [2018]). In analogy, the behaviour of one expression is modelled after the behaviour of another which it resembles (Antilla [2003]; Fischer [2007]; De Smet [2013]). Fischer [2009: 16] argues that one needs to consider changes in terms of the network that the construction/item operates in, and that the influence of paradigms in the grammar system is likely to be stronger than the influence of context (Fischer [2013: 521]). This idea has gained ground in the wake of recent research on constructional grammaticalization (e.g. Traugott [2007], [2008a], [2008b], [2015]; Trousdale [2008], [2010], [2012]), and more generally constructionalization (Traugott & Trousdale [2013]). As Traugott [2019: 130] observes, “if, as DeSmet (2013) proposes, ‘language users are tireless at inferring regularities from usage’ (DeSmet 2013: 8), analogy will necessarily be a crucial factor in the development of new expressions”. In this paper I argue that analogy and analogization most certainly played a crucial role in the emergence of the construction <wayamp ADJ>.
68In PDE, the constructions <wayamp ADJcmp> and <wayamp too ADJ> are very close to the constructions <much/faramp ADJcmp> and <much/faramp too ADJ>. In (56), “way bigger” could thus be paraphrased as far bigger, and in (57) “way older” is rephrased as “much older” in the following context:
(56) Miss Devlin in school had told them all that the sky was so many many miles away that everything in the sky looked very small, when it was really very very big, and the sky was way bigger than the earth. [COHA, Random House, 1981]
(57) “So, how old are you?” Amy asked suspiciously. “In Earth years?” Dorothy smirked. “Way older than you.” Small surprise, Dorothy acted much older than anyone Amy knew, except maybe Dad. [COHA, Kansas, she says, is the name of the star, 2006]
69The amplifiers much and far differ from way in terms of register (among other things). Much occurs in formal contexts, far in both formal and informal ones, and way is considered to be informal in the constructions <wayamp ADJ> and <wayamp too ADJ> – and slang when modifying an adjective in the absolute degree, e.g. way cute (OED s.v. way adv.3). In (56), way seems to be best suited because of the conversational tone of the passage and the repetitions (“so many many miles away”, “really very very big”) which, like “way bigger”, suggest a high degree of expressivity.
70The focus here will be on the constructions <faramp ADJcmp> and <faramp too ADJ> because far can also modify a spatial preposition or adverb, just like (a)way (see Section 2.2.1.), as shown in (58):
(58) Soon upon a mountain plain I gaz’d with wonder new. There high a castle rear’d its head: And far below a region spread […]. [COHA, The Sylphs of the Seasons with other poems, 1813]
71Examples of the pattern <far PREP/ADVspatial> ‘Xspatial by a long distance’ are found from the early nineteenth century onwards in COHA. The parallel between (a)way and far based on their function as modifiers of both spatial prepositions or adverbs, and adjectival heads, and the spatial meaning of adverbial far and (a)way explain why I took a closer look at the amplifier far-construction, which is illustrated in (59):
(59) The practice of dedicating children to God, is, we perceive, sanctioned by the usage of high antiquity; but, what is far better, it is conformable to reason and Scripture. [COHA, Female Scripture Biographies, 1817]
- 18 One reviewer observed that far also occurs in the superlative pattern. This is not the case, howeve (...)
72I propose that semantic analogization based on this construction may have been one of the factors of change steering the semantic development of the amplifier way. Semantic analogization is a mechanism of analogical extension (De Smet [2013: 65]) based on semantic similarity between syntagmatically related source and target constructions (De Smet [2013: 69]). De Smet [2016] argues that in order for analogy to be able to operate, the analogical model needs to be accessible, which (among other things) depends on its degree of entrenchment. A comparison between the frequency of <faramp ADJcmp> and <faramp too ADJ> on the one hand, and <wayamp ADJcmp>18 and <wayamp too ADJ> on the other, shows a sharp contrast between the two types of construction.
Figure 3. Relative frequency of the constructions <faramp ADJcmp> and <faramp too ADJ> compared with <wayamp ADJcmp> and <wayamp too ADJ> (per million words)
- 19 The fact that the schemas <wayamp ADJcmp> and <wayamp too ADJ may have been attracted to and come t (...)
73Figure 3 provides evidence of the much higher frequency rate of the amplifier far-constructions than the way-constructions from the early nineteenth century to – at least – the late twentieth century, by which time the frequency rate of the construction <wayamp ADJ> rose significantly (see Figure 1 in Section 2.1.) and may be interpreted as a sign of entrenchment. Taking discourse frequency as a proxy to entrenchment, I argue that since the constructions <faramp ADJcmp> and <faramp too ADJ> were well-entrenched when way started to show signs of development into an amplifier in the late nineteenth-early twentieth century (see Section 2.1.), they served as analogical templates for the development of the amplifier way.19 Similarly, the extant micro-schema <far PREP/ADVspatial> may have spearheaded the further development of the micro-schema <way PREP/ADVspatial>, which is itself one of the possible triggers of the development of the degree meaning of way (see Section 2.2.1.). The co-existence in the English language of the micro-schemas <far PREP/ADVspatial>, <faramp ADJcmp> and <faramp too ADJ> most certainly drove some language users to use the innovative constructions <wayamp ADJcmp> and <wayamp too ADJ> as (metaphorical) extensions of <way PREP/ADVspatial> on analogy with the (hypothetically metaphorical) extension of the micro-schema <far PREP/ADVspatial> to the micro-schemas <faramp ADJcmp> and <faramp too ADJ>. If so, semantic and paradigmatic analogization are further causes of change in the development of the degree modifier constructions <wayamp (more/too) ADJ(-er)>, alongside metaphorization (see Section 2.2.1.) and inferencing (see Section 2.2.2.). At least, one can hardly discard the assumption that the far-constructions favoured the spread of the (degree modifier) way-constructions in the community of American English language users.
74There still remains to explain why the amplifier way, by contrast with far, can collocate with various evaluative adjectives in the absolute degree, such as cool in (60) (see Section 2.1.):
(60) On Christmas morning under the tree, Amanda is thrilled with her trendy new pair of Guess jeans and her Songbird Barbie doll. Joey thinks his new Nike cross-trainers are way cool, and he’s in love with his Disney 101 Dalmatians jogging gear. [COHA, Santa’s Sweatshop, 1996]
75As mentioned above, the amplifier far pre-modifies adjectives in the comparative degree and adjectives in the absolute degree which are already pre-modified by amplifier too. Only a restricted set of adjectives in the absolute degree can occur with it without too being used as premodifier, e.g. different, preferable, inferior, superior, wrong, short, distant, remote, wide, adrift, astray, astern. Two examples are provided in (61) and (62):
(61) I attended my studies in school, with far different feelings and different motives, from what I had ever done before. [COHA, Memoir of Mrs Ann H. Judson, Late Missionary to Burmah, 1829]
(62) The implication was that the Kennedy set-up was far preferable to that at other airports like Rome’s. [COHA, New york Times, 1985]
- 20 The constructs <far different>, <far preferable>, <far inferior>, <far superior are not unrelated t (...)
76The amplifier way collocates with similar adjectives encapsulating the idea of a comparison between two entities20 – which, quite interestingly, also underlies the pattern <way/faramp ADJcmp> – as shown, for instance, through the constructs way different (see Example (2) in the Introduction), way overcrowded, way overpriced or way overrated found in COHA, but it has expanded its collocational profile to a wider range of adjectives in the absolute degree, as evidenced by the results of the search query <wayadv ADJ> (see Section 2.1.) which yielded tokens like way cool, way rich, way cynical, way posh, way young, etc. Though the construction <faramp ADJabs> was most probably an analogical model for the development of <wayamp ADJabs>, the fact that it shows a much lower type frequency than the construction <wayamp ADJabs>, with many fewer evaluative adjectives as possible collocates of far, suggests that other factors of change are involved. Indeed, the construction <wayamp ADJabs> also (partially) matches other preexisting degree modifier micro-schemas such as <very ADJ>, <so ADJ>, <really ADJ>, <well ADJ> (for a study of these patterns, see e.g. Ito & Tagliamonte [2003]; Tagliamonte & Roberts [2005]; Breban & Davidse [2016]). In the late twentieth and early twenty-first centuries the construction <wayamp ADJabs> may thus have aligned its behaviour through attraction (De Smet et al. [2018]) not only to the construction <faramp ADJabs>, but also to the bigger constructional family of amplifiers modifying a wide range of adjectives in the absolute degree. Paradigmatic analogization, i.e. extension to an existing set of alternatives based on a “semantic, formal and/or distributional relation of similarity” (De Smet [2013: 69]) could therefore be involved in the emergence of the degree modifier construction <wayamp ADJ> (see Aaron [2016] for similar conclusions concerning some degree modifiers in Spanish).
77In the previous section, I identified several possible semantic sources and mechanisms or causes of change involved in the semantic development of the amplifier way. In the present section I provide a summary of the findings, and conclude not only that multiple source constructions may be at the origin of the constructionalization of the amplifier way and, more particularly, of its high degree meaning, but also that different mechanisms or causes of change may have interacted at different stages of its developmental path. It thus questions a simplistic linear development. The development of the amplifier construction is couched in terms of grammatical constructionalization, with special emphasis on “the architecture of construction grammar [which] demands thinking in terms of […] not only individual substantive constructions but also abstract schematic ones” (Traugott [2015: 52]).
78I first presented the construction <wayadv PREP/ADV> as one of the possible sources of development of <wayamp ADJ>. I concurred with King’s [2016] view that the metaphor high degree is increase in distance (see King [2016]) hypothetically underlies the semantic development of the construction <wayamp ADJ>, and I provided evidence of a constructional change in the pattern <wayadv PREP/ADV> through context expansion (Himmelmann [2004]) which may have further favoured the development of the amplifier construction by foregrounding the degree meaning. I also showed that the data in COHA suggest that the micro-schema <wayadv PREP/ADV> may not only be historically related to away, but also to the constructions <a long/good/great way PREP/ADVspatial> and, to a lesser degree, <a way(s)adverbial PREP/ADVspatial>.
79Another possible source of semantic development of the amplifier way was identified, which is yet not rooted in the spatial domain, i.e. the adverbial phrase (in) every way. Evidence was provided of contexts in which a degree meaning arises inferentially out of the original manner reading, and other contexts in which the degree meaning even appears to be semanticized. I have also shown that an interesting parallel could be drawn between the sharp decrease in the frequency of the pattern <(in) every way ADJ> in the early twentieth century and the early uses of the construction <wayamp ADJ> at the same period.
80A summary of the chronological development of the way-constructions investigated in this article is provided in Figure 4.
Figure 4. A chronological overview of the amplifier way-construction and its possible sources in COHA
- 21 Note that because the particles collocating with a way(s) mainly have a spatial meaning and no meta (...)
81The solid lines in Figure 4 indicate steady use of a pattern while the dotted ones point to (more) scattered uses.21
82Note that the shift to amplifier function involves subjectification, i.e. the tendency of meanings to become increasingly based in the speaker’s subjective attitude towards the proposition (Traugott [1995]; Traugott & Dasher [2002]). The shift in meaning is not only from concrete (lexical) to more abstract (grammatical) meaning but also towards a more subjective meaning in the sense that it expresses high degree from the point of view of the speaker.
83Language-internal factors provide additional clues as to what drove the semantic change of way. The prior existence within the larger degree modifier schema of the booster set composed of <faramp ADJcmp>, <faramp too ADJ>, but also, for instance, <very/so/really/well ADJabs>, and even <faramp ADJabs>, to a certain extent, was a most likely trigger for the change, which was based in semantic and paradigmatic analogization. I argued that the far-constructions played a central part in this process of analogization considering the semantic and syntactic similarities between far and (a)way, two spatial adverbs which can modify a spatial preposition or adverb. The higher frequency rate of <wayamp ADJcmp> and <wayamp too ADJ> in PDE as compared to <wayamp ADJabs> could in fact be related to the high frequency rate of <far ADJcmp> and <far too ADJ> (see Table 1 in Section 1). Still, further research is needed to account for the fact that it is the construction <wayamp too ADJ> which is much more frequent than <wayamp ADJcmp> (NF= 0,6 vs. 0,2) whereas the pattern <faramp too ADJ> is far less frequent than <faramp ADJcmp> (NF= 5,1 vs. 36,9).
84The study of the amplifier far-constructions brings us back in turn to the hypothetical spatial origin of way, showing how metaphorization and analogization – not only on the model of <far ADJ> but also <far PREP/ADV> – may have been intertwining mechanisms or causes of change. But the fact that amplifier way has a large collocational profile, modifying adjectives both in the comparative and the absolute degree, including a wider range of adjectives in the absolute degree than far (way rich, way old; *far rich, *far old), suggests that analogization on the model of <faramp ADJcmp> and <faramp too ADJ> only gives part of the picture of the development of the construction <wayamp ADJ> at the macro-level. This now brings us back to the pattern <(in) every way ADJ> which showed a wide range of evaluative adjectives involved, whether in the absolute or the comparative degree. The amplifier way may thus have inherited (part of) its collocational profile from it.
85Figure 5 represents a constructional network of the different constructions that might at some point have favoured the semantic development of the amplifier way-construction.
Figure 5. A constructional network of the hypothesized sources of development of the micro-schema <wayamp ADJ>
- 22 An exemplar is “an entrenched item stored in memory, typically a construction, to which another wit (...)
86Figure 5 shows quite clearly how, by hypothesis, multiple source constructions interacted in the course of development of the amplifier way, both on a micro- and a macro-level. On the micro-level the construction <wayadv PREP/ADV> appears to be a major source of development as it allows for exemplar-based22 analogization both from a syntactic and a semantic viewpoint. Indeed, way already takes on the function of modifier in this pattern and it has a general heightening effect. I propose that the development of the amplifier way-construction which the micro-schema <wayadv PREP/ADV> was giving way to was further favoured by the amplifier-like function of the phrase (in) every way. Note that the micro-schema <wayadv PREP/ADV> itself may have taken its source both in (a)way-based constructions and the micro-schema <far PREP/ADV>, all of which instantiate a higher-level shema, i.e. <X PREP/ADV> whereby X is any type of modifier. On the macro-level the change to amplifier way certainly involved some interaction with another higher-level schema, i.e. the degree modifier construction <X ADJ>, and micro-schemas that instantiate it such as <far ADJ>.
87This study thus departs from most traditional diachronic case studies which “[draw] straight lines between a construction and a historical ancestor”, and suggests instead that the rise of the amplifier way can be envisaged as the outcome of “the blending of clearly distinct lineages” (Van de Velde et al. [2015: 1]). A corollary assumption is that several mechanisms or causes of change interacted, i.e. metaphorization, inferencing, and analogization, possibly mutually reinforcing each other. This is in line with the recent view that one can “come closer to a true understanding of the developments in question by considering multiple causes acting in concert or even independently” (Joseph [2015: 677]).