“A Good Deal of Intensity”: On the Development of Degree and Quantity Modifier Good

Abstract

This paper explores how the originally descriptive adjective good (e.g., “a good man”) developed degree modifier (e.g., “a good scolding”) and quantity modifier (e.g., “a good many people”) uses. The work is innovative in exploring the intensification potential of unbounded rather than bounded adjectives and in distinguishing between degree and quantity modification, the latter only recently gaining attention in the cognitive-functional literature. The developmental path of good will be linked to its construal in terms of scalarity, the process of subjectification, and the categorial shift from modification to submodification.

Keywords

degree modification quantity modification scalarity intensification subjectification submodification

1. Introduction

As rightfully observed by the guest editors of this special issue (see Claridge & Kytö 2021), the discussion about and study of degree phenomena has largely focused on the intensification of adjectives by means of adverbs. Although the multitude of studies, in general as well as in this special issue, vouches for the value of this focus, this paper aims to explore less well-trodden ground. The focus here lies with intensification conveyed by adjectives. This is of course not the first study on adjectival intensifiers, but so far studies have mainly looked into intensification through specific adjectival combinatorial patterns (e.g., González-Díaz 2010, this issue) or into intensification by (originally) bounded adjectives (e.g., Paradis 2000; Ghesquière 2010, 2014). This paper will explore, by means of a diachronic study, the intensification conveyed by the unbounded adjective good, looking into its degree modification and quantity modification uses.

Degree modification is understood here as the modification of the degree of gradable notions—quantitative or qualitative—as expressed by adjectives (1), nouns (2), adjectives and nouns (3), or adverbs (4). Quantity modification, which has only recently started to draw attention, refers to the modification of quantifying expressions, including both well-established quantifiers (e.g., very many) and grammaticalized or grammaticalizing quantifiers (e.g., “a great deal of money”), as in (5) and (6) respectively (cf. Ghesquière, Davidse, Njende & Breban 2016; Davidse, Ghesquière & Njende 2018; Davidse, Njende & Ghesquière forthcoming).

(1) She was an absolutely lovely person [. . .] (WBcannews, 2003)¹

(2) I had glandular fever in 1997, which was an absolute nightmare. (WBsunnow, 2000)

(3) What a stupid COW. (time.com/2961/the-20-best-worst-things-that-have-been-said-about-me-on-youtube, 3 August, 2020)

(4) The Absolute Very Best Restaurant in Bruges! (www.tripadvisor.com/ShowUserReviews-g188671-d2166198-r155750280-Le_Chef_Et_Moi-Bruges_West_Flanders_Province.html, 3 August, 2020)

(5) I suppose that may be true of some, but surely not of very many. (WBbrbooks, 1987)

(6) A girl could spend a whole lot of time feasting her eyes on his torso-hugging T-shirt. (WBsunnow, 2005)

Good, in addition to its original unbounded descriptive uses (7), has also developed degree modifying and quantity modifying uses. As a degree modifier, good heightens the degree of a gradable quality or quantity expressed by either another adjective or a noun, as in (8) and (9) respectively. As a degree modifier of adjectives, good combines with another unbounded adjective, expressing something like ‘very + approval’ (Adamson 2000:54; see also section 4.1). As a degree modifier of nouns, good functions as a booster, upscaling the degree of a gradable notion on an open-ended scale. The gradable notion modified can be either qualitative (e.g., “a good beating”) or quantitative (e.g., “a good quantity of honey”) (section 4.2). Finally, good also has quantity modifier uses, illustrated in (10). As in its degree modifier use, quantity modifier good is a booster: it upscales the quantity conveyed by an absolute quantifier on an open-ended scale (section 4.3).

(7) She had a good husband, a nice mansion and a brilliant lifestyle and she blew it all because she wanted to go on man-hunting trips. (WBsunnow, 1999)

(8) Venison is delicious if you have good strong teeth! (WBbrbooks, 1979)

(9) He’s never been knocked down before. I gave the guy a good beating for 12 rounds. (WBsunnow, 2002)

(10) It was a queer business, because usually, like a good many of those extremely introverted people, he can drink a good deal without its seeming to have any effect on him. (WBbrbooks, 1986)

Through detailed qualitative data analysis, this paper aims to provide a clear picture of the degree and quantity modifier uses of good and the conceptual and diachronic relationship between them, and as such contribute to a better general understanding of these types of adjective uses and how they differ from and resemble each other. In section 2, the phenomena of degree and quantity modification are further explored. Section 3 introduces the reasons for choosing good as an object of study and details the adopted methodology. The results of the case study on good are discussed in section 4, and section 5 reflects on the diachronic development of good, before turning to some concluding remarks in section 6.

2. Degree Modification and Quantity Modification

The view of degree modification and quantity modification held in this paper hinges crucially on the concept of “boundedness” (Paradis 2000, 2001; Kennedy & McNally 2005; Ghesquière & Davidse 2011; Ghesquière 2014). Boundedness concerns the construal of meaning as (not) having a clear boundary, as (not) being easily spatio-temporally demarcated. Nouns, verbs, and adjectives have all been discussed in the literature in terms of boundedness. For verbs, boundedness is typically linked with aspect or Aktionsart; for nouns it relates to the count/non-count distinction (e.g., Langacker 1987:58). However, for gradable nouns, boundedness can also be understood in much the same way as for gradable adjectives, referring to the open or closed nature of the evoked property scale. Gradable nouns (Paradis 2000:243) or degree nouns (Bolinger 1972), such as idiot, mess, and splendor, are inherently evaluative in nature, assigning a particular gradable quality to the noun phrase (NP) referent. As such, the propositional meaning of the NP an idiot is very similar to that of the NP an idiotic person: both NPs describe the referent as someone who behaves in an extremely stupid way.

The gradable qualities expressed by gradable nouns can, much like gradable adjectives, be modified in terms of degree, either by a preceding adjective or an element such as what or such. In the NP a big enthusiast in (11), for example, big does not assign a quality to the NP referent but rather heightens the degree to which the NP referent is an enthusiast or enthusiastic. Similarly, in (12), such indicates that the diagnosis was a relief to a great extent.

(11) “I’ve been a big enthusiast for nearly a decade [. . .] I’ve wanted to see this race for years.” (WBusnews, 2002)

(12) “I was given the all-clear,” he said. “It was such a relief.” (WBsunnow, 2001)

Building on Paradis’ (2000, 2001) work, Ghesquière and Davidse (2011) argued that gradable adjectives and nouns carry in them the possibility of being construed either as bounded or as unbounded.² Bounded properties are conceptualized as complementaries, i.e., properties that either apply or do not apply (e.g., dead - alive, true - false, complete - incomplete), while unbounded properties are properties inherently conceptualized as degrees (e.g., long - short, old - young).³ Whereas bounded adjectives evoke the extreme points or boundaries of a scale, unbounded adjectives locate a property as a region on an open-ended scale. The difference between the two types of gradable adjectives is drawn out by the type of degree modifiers with which they can be used.

Bounded properties take closed scale degree modification, whereas unbounded properties take open scale degree modification. Closed scale degree modifiers assess the property described by comparing the actually designated degree to the boundary they imply, i.e., as either reaching or not reaching that boundary (e.g., “absolute rubbish”) or as going some way towards reaching it (e.g., “almost complete”). Open scale degree modifiers measure the actual degree of the property on a scale with some form of assumed measure units (e.g., “a terrible bore,” “very strange”). In other words, these properties are inherently conceptualized as degrees which are measured on a scale going up or down from a reference point. The difference is visualized in Figure 1. On the one hand, the bounded properties “empty” and “full,” for instance, can be conceptualized as the lower and upper boundaries of a scale. When a container is then said to be almost full, the upper boundary point “full” is close to being reached. The unbounded property “strange,” on the other hand, can be visualized as instantiating its own property scale of strangeness, with very strange occupying a region high on the open-ended scale and a little strange ranking much lower on the scale.

Figure 1.

Degree Modification

Based on Quirk, Greenbaum, Leech, and Svartvik (1985) and Kennedy and McNally (2005), it is possible to design the typology of open and closed scale degree modifiers in Figure 2. The main distinction is a structural one based on the specific structure of the scale involved, open versus closed (Paradis 2000, 2001; Kennedy & McNally 2005; Tribushinina 2008). Then, within these two main types, a general subclassification can be made, relying on Quirk, Greenbaum, Leech, and Svartvik’s (1985) typology. Open scale degree modifiers subsume upscalers or positively oriented intensifiers, “scal[ing] the meaning of the modified element upwards from an assumed norm” (e.g., very, highly), and downscalers or negatively oriented intensifiers, scaling it downwards (e.g., a bit, slightly) (Nevalainen & Rissanen 2002:361).

Figure 2.

Typology of Degree Modifiers

Closed scale degree modifiers can be further subdivided into totality modifiers and approximators (e.g., almost, nearly, about), indicating that the degree of a property coincides with the boundary of a scale or is close to reaching it respectively. Totality modifiers are taken to include both Quirk, Greenbaum, Leech, and Svartvik’s (1985) maximizers (e.g., complete/completely, absolute/absolutely, total/totally) and minimizers or “negative maximizers” (Quirk, Greenbaum, Leech & Svartvik 1985:596) (e.g., hardly, barely), which posit the property at the upper and lower boundaries of a closed scale respectively.

Compromisers (e.g., quite, rather, kind/sort of, more or less) cross-cut the open/closed scale distinction in that they can modify both bounded and unbounded properties, as in (13) and (14) respectively.⁴ In (13), more or less locates the degree of ability midway on a scale ranging from the boundary points “able” to “unable.” In (14), more or less locates the degree of happiness roughly in the middle region of an open scale.

(13) With the success of previous raids, we believe we are more or less able to carry out strikes around the clock. (WBsunnow, 2001)

(14) Though none of them deserves it, Beck allows her creatures a more or less happy ending [. . .] (WBtimes, 2002)

Ghesquière, Davidse, Njende, and Breban (2016), Davidse, Ghesquière, and Njende (2018), and Davidse, Njende, and Ghesquière (forthcoming) have used the main types of degree modification as the basis for setting up categories of quantity modification. Quantity modification, as in very few and almost all, is also determined by the boundedness construal of the modified element. In other words, boundedness is crucial not only to the construal of adjectives but also quantifiers, as the distinction between bounded and unbounded adjectives is conceptually similar to the distinction between relative and absolute quantifiers generally made in the literature (Ghesquière & Davidse 2011).⁵

Absolute quantifiers (e.g., many, a lot, several, three, ten) specify the size of a set/mass by offering a “direct description of magnitude” (Langacker 1991:83). As such, absolute quantification is clearly scalar (Radden & Dirven 2007), invoking an open scale of magnitude. As visualized in Figure 3, absolute quantifiers measure along a directional scale without an upper boundary. The most obvious absolute quantification scale is the cardinality scale, which ranges from zero to infinity with clearly defined measure points, zero, one, two, three, etc.

Figure 3.

Absolute Quantification

Relative quantifiers (e.g., all, none, half, most) do not directly indicate size or quantity. Rather, they quantify by comparing the predicated or actually designated mass or quantity (P) with a reference mass (RT) and assessing whether or not they (partially) coincide. In other words, they “estimate the distance between the boundary points, P and RT” (Langacker 1991:108, 2016:8), thus clearly evoking a closed quantification scale with the boundary being the reference mass (see Figure 4).

Figure 4.

Relative Quantification

Like unbounded properties, absolute quantifiers take open scale degree modifiers, which locate the property or quantity in a higher (e.g., “very many”) or lower (e.g., “hardly any”) region on a scale. Relative quantifiers, like bounded properties, take closed scale degree modifiers, emphasizing that the size of the actually predicated mass coincides with the (maximal) reference mass (i.e., the upper boundary) (e.g., “absolutely all”) or indicating how far it falls short (e.g., “nearly all”). Figure 5 summarizes the different types of quantity modification absolute and relative quantifiers can take. This is very similar to Figure 2, which did the same for unbounded and bounded adjectives. Absolute quantity modification can be both upscaling (e.g., “such a lot”) and downscaling (e.g., “hardly any”). Relative quantity modification encompasses totality modifiers (e.g., “absolutely all”), and approximators (e.g., “almost none”). Compromisers, as with bounded and unbounded properties, can modify both absolute and relative quantifiers, e.g., “rather many,” “more or less half of them.”

Figure 5.

Typology of Quantity Modifiers

3. What Does It Mean to Be Good and How to Find Out?

To further our understanding of the conceptual similarity between degree and quantity modification, I examine closely the adjective good, an originally unbounded descriptive adjective that has developed both degree and quantity modifier uses. Earlier studies on elements that have a similar range of uses have focused on originally bounded adjectives (e.g., Ghesquière 2010 on whole) or adverbs (e.g., Breban & Davidse 2016 on very). Paradis (2000:253) even argues that only bounded (and extreme) adjectives are recruited for noun-intensification. She states that as intensification requires the abstract notion of either high degree or totality, unbounded adjectives “do not have much to offer in terms of intensification” (Paradis 2000:253). By looking into the intensifying potential of an unbounded adjective, good, this empirical diachronic study hopes to show the relevance of this type of adjective to the field of degree and to further elucidate the conceptual and diachronic relationship between degree and quantity modification.

Recently, some interest has started to develop in (sub)modifier uses of crosslinguistic counterparts of English good. Castroviejo and Gehrke (2017), for instance, study degree modifying uses of Catalan bon as in “un bon cop” ‘a good blow’ and “un bon ensurt” ‘a good shock.’ Also, the quantity modifier use of good is touched upon in Davidse, Njende, and Ghesquière (forthcoming), in their discussion of the types of quantity modification taken by (a) few and all, as in (15).

(15) Set a date a good few months in advance and give yourself deadlines (WBtimes, 2004)

The unbounded gradable adjective good is described in the OED as “the most general and most frequently used adjective of commendation in English,” conveying “the sense of being of a high (or at least satisfactory) quality” and “worthy of approval” (OED, s.v. good, adj., n., adv., and int.: A. adj.). In fact, in this use, the “meaning” of good is highly context-dependent, varying not only according to the type of referent being described but also the speaker, community, time, etc. This led Umbach (2015) to refer to the meaning of good as only “quasi-denotational” and Hare (1952) to claim that there is no property shared by all “good” things. What makes a car good (its speed, safety, price, reliability, eco-friendliness, etc.) is not the same as what makes a meal good (the ingredients, seasoning, or even the company it is eaten in), nor is it the same as what makes a photograph good (the light, composition, subject, storytelling, etc.). The general sense of commendation or approval seems to be the only consistent element in the semantico-pragmatics of the unbounded gradable adjective use of good.

Although good is expected to be predominantly a qualitative adjective of commendation, it will be informative to see in the data how frequent and how recent (or old) the degree and quantity modification uses are, as well as how the collocational patterning differs between these uses. To this end, an exhaustive data set for good, including all the spelling variants listed in the OED headword section, was extracted from the online OED quotations database. As searches in more traditional corpora of English quickly showed that there would not be enough relevant hits to make even the most careful of claims, I turned to the OED quotations database. Using the OED quotations database as a historical corpus is now quite generally accepted and relatively widespread (for appraisal see Hoffmann 2004; Rohdenburg 2013). For this study, the OED was particularly appealing because of its size (necessary as the degree and especially the quantity modifier uses are infrequent), its time span (covering all historical stages from Old English to Present-Day English), and its variation in terms of genre and register.

Despite its many advantages, there are of course a number of drawbacks to using the OED quotations database as a corpus. As one of the reviewers pointed out, the structure of the OED is such that an even temporal, writer, and generic coverage cannot be guaranteed. I am very much aware of the weaknesses of the OED quotations database—as are, for example, Hoffmann (2004) and Rohdenburg (2013)—but using it has allowed me to find much needed, useful information (albeit with a few hiccoughs) that the existent traditional corpora were unable to provide. For this study, the lack of sufficient context sometimes complicated the analysis and required additional research to find more context. Quite some noise had to be removed as well, as double or multiple identical hits (e.g., titles) and irrelevant hits (e.g., nominal and verbal uses, fixed combinations such as good night) were not uncommon. Predicative adjectival uses were not included in the analysis as the focus here is on degree and quantity modification uses, and these are restricted to prenominal position (a complete idiot, *the idiot is complete). Because the OED search function is not very elaborate, all predicative uses had to be sorted manually. The resulting, manually filtered data featuring prenominal uses of good were then analyzed in terms of syntactic features, semantico-pragmatic function, and collocational patterning.

In accordance with the structure of the corpora used here, the data were divided into subperiods as follows: Old English (OE) (to 1150), Middle English (ME) (1151-1500), Early Modern English (EmodE) (1501-1710), Late Modern English (LModE) (1711-1920), and Present-Day English (PDE) (1960-2005). All periods were further subdivided into 100- or 70-year subperiods, again following the structure of the consulted corpora. Table 1 gives an overview of the subperiods and the number of hits in the quotations database.

Table 1.

OED Data for Good

Period	Time	Number of hits	Period totals
Old English	850-950	46	121
	951-1050	75
	1051-1150	35
Middle English	1151-1250	62	4195
	1251-1350	151
	1351-1420	1479
	1421-1500	2468
Early Modern English	1501-1570	3832	18,422
	1571-1640	4732
	1641-1710	5858
Late Modern English	1711-1780	4372	23,752
	1781-1850	4681
	1851-1920	11,699
Present-Day English	1960-2005	13,473	13,473

4. Good as a Degree and Quantity Modifier

4.1. Good as a Degree Modifier of Adjectives

Degree modifiers of adjectives or adjective intensifiers are likely to be the most studied items in the field of degree. Typically, their study is focused on bleached or fully grammaticalized intensifiers, as in “very/rather/pretty nice,” which have lost most of their original meaning and have been reduced to their degree meaning. The original truth meaning of very, for instance, is no longer part of its adjective intensifier use. The same is true of the temporal and beauty meaning of rather and pretty. Dixon (1982), however, has drawn attention to another, non-bleached or at least less bleached type of adjective intensifier, which has received much less attention in the literature (but see Feist 2012; González-Díaz 2018). These “value adjectives,” as he labeled them, can be found cross-linguistically and modify some other adjective which is “placed immediately after the VALUE adjective” (Dixon 1982:25). Dixon (1982:26) illustrates this type of adjectival use by means of good, the item under scrutiny here. In constructions such as “a good strong box” and “a good new fast car,” good is said to no longer function as a mere descriptive modifier but as a degree modifier of the following descriptive adjective.⁶ In Dixon’s (1982) view, these NPs do not involve recursive modification (*there’s a fast car, the fast car is a new one, the new fast car is a good one) or independent modification (the car is fast, the car is new, the car is good). Rather the sequence is best interpreted as ‘a fast car which is new and therefore good.’ Similarly, the NP lovely old trees refers to trees which are old and therefore lovely. This ties in with Adamson’s (2000:54) observations for lovely and nice in sequences such as lovely long legs and nice big houses. Here, the adjectives lovely and nice have not fully bleached, having retained “some of the original affect,” but now roughly have the meaning of ‘very + approval.’ González-Díaz (2018:50) has used the term indirect modification to describe this modifying function, as in her view the adjective does not only submodify the following adjective but also still modifies the head noun. As such, they would seem to form an intermediate type of element, hovering between modifier and submodifier at the same time.

As González-Díaz (2018:50) points out, determining whether good functions as a descriptive modifier, simply in coordination with the following descriptive modifier, or whether it is a degree modifier, submodifying the following adjective requires “qualitative analysis of the context.” Careful contextual analysis as well as analysis of the adjectival and nominal collocates proved crucial in deciding on the function of good. Non-bleached adjective intensifier uses of good were found in the data from the period 1251-1350 onwards (see Table 5). In all cases, the unbounded adjective good modified another unbounded adjective. Also, adjective degree modifier good showed a clear preference for specific collocational sets, including more objective, measurable adjectives expressing dimension/size (16), strength (17), speed, and temperature as well as more subjective, evaluative types of adjectives denoting honesty (18), simplicity (19), taste, etc. As Table 2 shows, dimension/size adjectives were by far the most common. From the LModE period onwards, the collocational patterning of adjective intensifier good becomes very diversified. Yet, it always combines with other adjectives denoting a property that is considered positive in that particular context, as the commendation meaning of good is still transparent in these uses.

(16) [. . .] bumbast his buttockes with a good long sticke [. . .] (OED, 1600)

(17) [. . .] a good strong Chopping-knife [. . .] (OED, 1626)

(18) I think him a good honest gentleman [. . .] (OED, 1585)

(19) This is a good simple game for two players. (OED, 1881)

Table 2.

Collocational Behavior of Good as a Degree Modifier of Adjectives

	1251-1350	1421-1500	1501-1570	1571-1640	1641-1710
Dimension/size (e.g., deep, great, high, long, thick, etc.)	1	6	5	6	30
Strength (e.g., robust, solid, strong, sturdy, etc.)			3	5	8
Speed (e.g., fast, round, speedy, etc.)				3	2
Temperature (e.g., fresh, hot, warm, etc.)			2	1	2
Simplicity (e.g., dumb, plain, simple, temperate, etc.)		1	4	4	1
Ability (e.g., able)		1	0	0	1
Honesty/reliability (e.g., honest, moral, true, trusty, etc.)			8	5	5
Taste (e.g., dry, fat, hearty, honey-sweet, etc.)			2	2	5
Beauty (e.g., handsome)				2	1
Knowledge (e.g., ingenious, learned, wise, etc.)					3
Privative (e.g., clean, pure, sure, etc.)				2	4
Other (e.g., lively, well-baked, bouncing, etc.)		3	4	7	35

4.2. Good as a Degree Modifier of Nouns

As a degree modifier of nouns, good always functions as an open-scale modifier, upscaling or boosting the property evoked by the gradable nouns and indicating that it ranges high on an open-ended scale. The nominal collocates of degree modifier good can be divided into two main categories, nouns conveying qualitative properties and nouns conveying quantitative or quantifiable properties.

4.2.1. Good as a Qualitative Degree Modifier

Qualitative degree modifier uses of good are rather marginal in the data and are attested only from 1350-1420 onwards (Table 5). Collocationally, they are largely restricted to violent or negative action nouns such as smite, blow, blast, knock, and cry as in (20)-(23). In these NPs, good does not function as a gradable adjective assigning praise to the NP referent but rather heightens an implied degree of force associated with these nouns. Syntactically, the degree modifier status of good is evident in that the adjective is no longer found graded or negated in this use (*a very good blow, *not a good blow), and, semantically, it is no longer an antonym of bad.

(20) Clappe hym on the backe a good stroke [. . .] (OED, 1530)

(21) [. . .] you shall dissolue it with good rubbing [. . .] (OED, 1530)

(22) After a good disciplyning with a yerde, they [sc. children] kepe right wel doctrine of their scole. (OED, 1532)

(23) ‘Go away, my dear, and have a good cry,’ said the wise old ladies of an earlier generation. (OED, 1925)

Uses such as these are never very frequent in the OED data, but they are attested cross-linguistically. In non-standard Flemish Dutch uses such as these of goed are not uncommon (24). Castroviejo and Gehrke (2017) mention them for Catalan bon (25),⁷ and Spanish (26) and Galician (27) also have similar uses (Méndez-Naya, personal communication, February 2019). It is particularly interesting to note that in all three Romance languages the adjectives precede the noun, which testifies to their degree rather than descriptive use (Méndez-Naya, personal communication, February 2019).

(24) Verouderde kotdeuren waar je een goede duw tegen geeft en meteen binnen bent. (www.veto.be/artikel/doe-je-kot-op-slot, 3 August, 2020)

‘Worn-out digs doors you give a good push and you’re instantly inside’

(25) [. . .] un bon maldecap, un bon ensurt, un bon cop [. . .]

‘a good worry a good shock a good blow’ (Castroviejo & Gehrke 2017, slide 16/45)

(26) [. . .] y necesitan que les den de cuando en cuando una buena paliza para meterlos en cintura nuevamente [. . .] (CREA)

‘and they need to get a good beating from time to time to bring them into line again’

(27) que o pillen, carafio. Entre todos poderanlle zorregar unha boa tunda (TILG)

‘hope they catch him, carafio [expletive]. Among all of them they will be able to give him a good beating’

4.2.2. Good as a Quantitative Degree Modifier

Unlike the qualitative degree modifier uses which are attested only from ME onwards, quantitative degree modifier good is already found in the OE data from 925 onwards. In NPs such as a good handful, a good mile away, and a good while, good as a quantitative degree modifier indicates that the quantity or size referred to by the noun is “substantial” (Bolinger 1972:37) and exceeds the norm/expectations, thus ranging high on an open ended scale. This “excess” is always experienced positively, confirming the idiom “the bigger the better.”

Collocationally, good is much less restricted as a quantitative degree modifier than as a qualitative one. Table 3 shows the frequencies of occurrence for the different collocational sets of nouns.

Table 3.

Collocational Behavior of Good as a Degree Modifier of Quantitative Nouns

	Size(e.g., measure, quantity, liter)	Time(e.g., space, while)	Money(e.g., money, pittance, sum)	Speed(e.g., pace, speed)	Distance(e.g., fare, journey, way)	Other(e.g., demand, whang)
850-950	2
951-1050	2	2
1051-1150	1		2
1151-1250	1
1251-1350	3	1	1	3	1
1351-1420	8	1	1	3	1
1421-1500	14	3	8	4	2	3
1501-1570	17	19	5	1	6	6
1571-1640	44	19	6		8	1
1641-1710	94	40	8		19	8
1711-1780	50	22	3		10	9
1781-1850	24	8	3		3	9
1851-1920	70	9	10		12	1
1960-2005	53	4	2		1	3

The most frequent collocational categories attested in the data include: size nouns (28 and 29); nouns indicating distance (30 and 31), time (32 and 33), and speed (e.g., pace, speed, scour; 34); and nouns referring to an amount of money (e.g., money, sum, meed; 35). Notably, good can modify the degree of both more abstract and concrete indications of size (number, quantity, measure versus handful, spoonful, liter), distance (distance, way, journey, walk versus meter, mile, yard), or time (while, space versus hour, day, year). For all collocational sets, with the exception of the size nouns, the concrete quantity or size nouns postdate the abstract ones.

(28) Tak marigolde leues & wasche hem & do þer to a good quantyte of hony [. . .] (OED, 1450)

‘Take marigold leaves and wash them and add to that a good quantity of honey’

(29) Genim giðcornes leafa gode handfulle. (OED, 950)

‘Take a good handful of spurge-laurel leaves’

(30) Stand or syt a good waye of from the fyre [. . .] (OED, 1565)

(31) We crossed a rough plain to the left for a good mile. (OED, 1892)

(32) These thinges were done a good space after. (OED, 1569)

(33) I worked for a good hour. (OED, 1994)

(34) He fled away gud speid. (OED, 1512)

(35) Therunto had we nede to haue a good summe of money [. . .] (OED, 1548)

Considering that good is originally an unbounded gradable adjective of commendation one expects the degree modifier uses of the adjective to also be unbounded or open scale. As hypothesized by Paradis (2000, 2001, 2008), the shift from descriptive to noun-intensifying meaning can be conceptualized in terms of a foregrounding mechanism: the boundedness configurations or scalar construal of the lexical uses are foregrounded at the expense of their lexical content. In other words, the unbounded descriptive meanings will be backgrounded, while the open-ended scalar construal will be foregrounded. In (28), for instance, the quantity of honey is measured on an open-ended quantity scale and located in a region high on that scale. Similarly, in (32), good indicates that things were done at a later time than expected, placing the period referred to in a region high on an open-ended temporal scale.

The concrete quantitative nouns do not themselves indicate ranges on open-ended scales. Rather they can be thought of as measure points on open-ended scales. What good then does is to extend that point into a region slightly higher on the scale than the original measure point, but not reaching the next measure point on the scale. A good handful of leaves, for instance, is slightly more than one handful, but less than two handfuls. If one walks a good mile, one actually walks a bit farther than a mile, but not as far as two miles. The exact distance walked is unclear (1.2 miles, 1.4 miles, etc.). What is certain is that the next measure point on the scale has not been reached. This use paves the way for the true quantity modifier uses discussed in section 4.3.

4.3. Good as a Quantity Modifier

As a quantity modifier good modifies in terms of degree the quantity designated not by a noun but by an absolute quantifier. As discussed in 4.2.2, the restriction to absolute quantifiers is not surprising. Relative quantifiers can be conceptualized as working on closed scales, indicating whether the quantity referred to coincides with the end point of a closed scale (or reference mass) or by how much it falls short. Absolute quantifiers, in contrast, can be thought of as indicating a quantity or size on an open-ended quantity scale, either as a region on a scale (e.g., many, a deal [of], a few) or as a distinct measure point (e.g., cardinal numbers). In Langacker’s (2016) terminology, these absolute quantifiers evoke “continuous” and “quantized” open-ended quantification scales respectively. Good, being an originally unbounded descriptive adjective, has developed open-scale modifier uses, naturally co-occurring with other unbounded or open scale items. With absolute quantifiers, both continuous and quantized, good then measures a quantity on a scale up from an implied norm or reference point on an open measurement scale (Figure 6).

Figure 6.

Quantity Modifier Good

The quantity modifier use of good is found in the OED quotations database from the ME period 1420-1500 onwards (Table 5). Collocationally, quantity modifier good is quite restricted, combining only with the following absolute quantifiers: cardinal numbers (36), many (37), a few (38), plenty (39), and the grammaticalized/grammaticalizing quantifiers a bit (of) (40) and a deal (of) (see Table 4). A good deal (of) was by far the most frequent combination attested in the data, modifying adjectives (41) and nouns (42) as well as occurring independently (43). It has now become so frequent and entrenched as a quantifier unit that it is listed as a separate entry (Macmillan, s.v. a good deal) or at least as a fixed collocation (Collins, OED, s.v. deal) in dictionaries, meaning ‘a lot’ or ‘a considerable amount of.’ In other words, a good deal (of) might be on its way to becoming a new independent (grammaticalized) absolute quantifier.

(36) The first is good ij fote & half or thre Feet depe [. . .] (OED, c1440)

(37) A good many intelligent people..obviously admired Stevenson. (OED, 1952)

(38) I have sold a good few Bramberries. (OED, 1864)

(39) Here are also good plenty of large Soals, taken in Troul-nets. (OED, 1697)

(40) [. . .] a good bit of spice to give the critlings a flavour. (OED, 1851)

(41) The Archers of England shuld not be only a great deale ranker, and mo then they be: but also a good deale bygger and stronger. (OED, 1545)

(42) Paules Churche..cost me a good deale of money, the leading thereof. (OED, 1563)

(43) These bookes, I would haue him read now, a good deale at euery lecture. (OED, 1567)

Table 4.

Collocational Behavior of Quantity Modifier Good

	Cardinal number	Many	A few	A bit (of)	A deal (of)	Plenty	Other	Total
1420-1500	1							1
1501-1570	2	3			3			8
1571-1640	1	4						5
1641-1710	1	3			6	1		11
1711-1780	3	7			66	1	1	78
1781-1850	5	16		2	112			135
1851-1920	10	117	6	6	367		1	507
1960-2005	18	55	6	5	283			367

5. The Diachrony of Prenominal Good

The data have provided evidence that prenominal good has developed a variety of non-descriptive uses, of which the quantity-related ones are clearly predominant in all periods. Already from 950 onwards, good engages in quantitative degree modification, as in a good while/mile/speed, upscaling the quantity or size evoked by the head noun of the NP on an open-ended scale. Near the end of the ME period good also starts to be used as a quantity modifier in the OED quotations, upscaling the quantities denoted by absolute quantifiers, as in a good few/deal/two. These uses, mainly because of the combination a good deal (of), rise steadily in frequency, peaking in the mid-nineteenth century (Table 5).

Table 5.

Non-descriptive Prenominal Good in the OED Quotations

	Adjective intensifier		Qualitative degree modifier		Quantitative degree modifier		Quantity modifier
	N	per 1000	N	per 1000	N	per 1000	N	per 1000	Hits
850-950					2	43.5			46
951-1050					4	53.3			75
1051-1150					3	85.7			35
1151-1250					1	16.1			62
1251-1350	1	6.6			9	59.6			151
1351-1420			1	0.7	14	9.5			1479
1421-1500	11	4.5	5	2.0	34	13.8	1	0.4	2468
1501-1570	28	7.3	10	2.6	54	14.1	8	2.1	3832
1571-1640	44	9.3	8	1.7	78	16.5	5	1.1	4732
1641-1710	97	16.6	5	0.9	169	28.8	11	1.9	5858
1711-1780	42	9.6	11	2.5	94	21.5	78	17.8	4372
1781-1850	32	6.8	28	6.0	47	10.0	135	28.8	4681
1851-1920	135	11.5	19	1.6	102	8.7	507	43.3	11,699
1960-2005	161	11.9	25	1.9	63	4.7	367	27.2	13,473

The qualitative non-descriptive uses are much less common in the data. Qualitative adjective intensifier uses, as in a good sharp knife, are found only after 1250, and it is not until after 1350 that good is able to convey qualitative degree modification. Both uses are also quite restricted in terms of collocational behavior. Good can only modify the degree of positive evaluative adjectives and violent or negative action nouns such as blow, beating, and cry.

The question arises as to how these different uses developed. Other studies on adjectives that have developed both degree and quantity modification uses have shown that the quantity modification uses typically develop from the degree modification ones. For whole, for instance, Ghesquière (2010, 2014) reconstructed a pathway from closed-scale degree modification (44) to open-scale degree modification (45) and, finally, to open-scale quantity modification (46). The open-scale degree modification conveyed by whole cannot be predicted on the basis of its inherent semantics, but developed through collocational extension and reclustering. The shift from open-scale degree modification of quantitative nouns to quantity modification of absolute, open-scale grammaticalized quantifiers took place in binominal constructions of the type “whole N₁ of N₂,” in which N₁ is a size noun in the broad sense and the entire construction serves to emphasize that there is a very large quantity of N₂, as in (47).

(44) [. . .] and that no man, of any nation, ever beheld her that did not fall in love with her; and that she had all the slaves perpetually at her feet; and the whole country resounded with the fame of Clemene [. . .] (PPCEME, Behn, 1684; Ghesquière 2014:166)

(45) [. . .] the German FLOH with his knife and fork, insatiate, not rising from table; whole swarms from all the Russias, and Asiatic hordes unnumbered - all these were there [. . .] (CLMETEV, Kinglake, 1844; Ghesquière 2014:159)

(46) He is speaking up for a whole lot of people who feel that the American dream has passed them by. (CBtimes; Ghesquière 2014:171)

(47) There’s now a whole bunch of studies from different cities that show the same thing. (Brems 2003:299, cited in Ghesquière 2014:162)

For very, Breban and Davidse (2016) argued that the emergence of the quantity modification use in the late sixteenth century is the last major functional shift in the item’s development, as was the case for whole. As a submodifier of quantifiers, “very changes the quantity indicated, augmenting it in examples such as very many/much and diminishing it in cases like very few/little” (Breban & Davidse 2016:242), as in (48) and (49) respectively. Breban and Davidse (2016:242) found a shift from qualitative degree modification to quantity modification having taken place, enabled “by the conceptual analogy between ranges of qualities, measured on qualitative scales [e.g., clean in [50] below], and ranges of quantities, measured on the non-specific quantitative scales derived from the cardinality scale” (Langacker 1991:82-83).

(48) The said pest come in the towne of kirkcaldie that thair deit (‘died’) verrie mony. (OED, a.1578, Lindsay; Breban & Davidse 2016:242)

(49) Batha, whereof now there remaine but very few ruines. (OED, 1600, Pory; Breban & Davidse 2016:242)

(50) [. . .] put it into a very cleane boule of wood [. . .] (PPCEME, 1570-1640, Markham; Breban & Davidse 2016:239)

For good, like for very and whole, a path can plausibly be posited from descriptive modification to degree modification to quantity modification. Semantically, this path is characterized by subjectification, and, syntactically, there is a category shift from modification to submodification.

This study has focused on prenominal uses of good, but the majority of the instantiations of the adjective good in the OED quotations database are descriptive modifier uses—prenominal ones (e.g., good reviews, OED 1997) but mostly predicative ones (as in 51).

(51) Lime juice is very good for you in this climate, (OED, 1955)

This descriptive modifier use of good, conveying a general meaning of commendation, is the predominant one throughout all periods covered in the OED and it is also the earliest one attested. It is from this unbounded evaluative descriptive use the open scale degree modifier uses most likely developed, first with nouns inherently associated with quantity or size, later with more qualitative nouns. The idea that “bigger is better” certainly goes some way in explaining this shift from description to degree modification. A quantity or size that is “good” is typically going to be substantial. If you are looking for a house, for example, you want the rooms in the house to be good quality (e.g., bright, clean, etc.), but also of a good, substantial size (52).

(52) Wanted. A House, to Rent or Purchase, with three good rooms on a floor. (OED, 1789)

This “substantial” degree modifier reading then gradually widened collocationally to both more abstract and more concrete quantity nouns. From 1420 onwards, good could modify the quantitative notions conveyed not only by nouns, but also by true quantifiers, functioning no longer as a degree modifier but as a quantity modifier. Crucially, the scalar construal of good remained consistent in all its uses—descriptive, degree, and quantity modification. All of these uses are unbounded or open scale uses. Descriptive modifier good attributes an unbounded gradable, positive evaluative property to a NP referent that can be modified by open scale modifiers, positing the property higher or lower on an open-ended scale, e.g., “his English is (not) very good.” As a degree modifier, good no longer attributes a quality to the NP referent, but rather modifies a property associated with the NP referent in terms of degree. As a quantity modifier, good actually no longer engages in modification but in submodification, as it modifies another modifier or quantifier. As such, there has been a categorial shift from modifier to submodifier. Similarly, as a modifier of adjectives (“a good strong box”), good submodifies the descriptive modifier it precedes.

There is no evidence to suggest that the development of the modifier of adjective use of good is immediately linked to the development of the other uses. The meaning of good in this use is still lexical to some extent and is very much determined by the following adjective which gives the reason as to why the NP referent deserves to be attributed the property “good.”

6. Conclusion

The purpose of this paper was to come to a better understanding of the degree and quantity modification uses of prenominal good by looking into data ranging from OE to PDE. Being ‘good’ can be paraphrased as being ‘worthy of approval.’ This unbounded quality meaning has given rise to open scale degree modification uses, both with nouns and adjectives. As a submodifier of adjectives, as in “a good strong box,” good has partly retained its abstract approval meaning, but it is backgrounded somewhat in favor of a degree reading so that good can now be paraphrased as ‘very + approval’ (Adamson 2000:54). In its degree modifier use with nouns, as in “a good distance,” the commendation meaning has bleached and the degree reading is now prominent, with good being roughly equivalent to ‘substantial, remarkable in size.’ The degree reading is also present with qualitative, violent action nouns, as in “a good blow,” in which good stresses the violence or severity of the action. With both qualitative and quantitative nouns as well as with adjectives, good places the degree of the modified property high on an open-ended scale. Similarly, when good modifies quantifiers, it locates the quantities evoked high on an open-ended scale. As such, the boundedness or scalar construal of good remains consistent in all of its uses (cf. Paradis 2000).

This study has confirmed the developmental pathway attested for whole (Ghesquière 2010, 2014) and very (Breban & Davidse 2016) leading from degree modification to quantity modification. Further studies will have to confirm whether this pathway holds for other unbounded adjectives as well.

Footnotes

Acknowledgements

I would like to thank the journal editors and the reviewer for their valuable comments. Special thanks go to the guest editors who organized a wonderful workshop at ICEHL 20 and carefully guided us through the publication process. I would also like to express my sincere gratitude to Ngum Meyuhnsi Njende for her help in collecting and sorting the corpus data. Any remaining errors and shortcomings are of course my own.

Declaration of Conflicting Interests

The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author received no financial support for the research, authorship, and/or publication of this article.

Notes

Author Biography

Lobke Ghesquière is Associate Professor of English in the Faculty of Translation and Interpretation at the University of Mons and research fellow in the research unit Functional and Cognitive Linguistics at the University of Leuven. She has published on grammaticalization and (inter)subjectification phenomena in the English noun phrase, and her current research is mainly concerned with degree modification and focusing in the noun phrase.

References

CREA = Real Academia Española: Banco de datos. Corpus de referencia del español actual. http://www.rae.es (February, 2019).

OED = Oxford English dictionary. https://www.oed.com (June, 2018).

TILG = Santamarina, Antón (dir.), Ernesto González Seoane & María Álvarez de la Granja. Tesouro informatizado da lingua galega (Versión 4.1). Santiago de Compostela: Instituto da Lingua Galega. http://ilg.usc.es/TILG/ (February, 2019).

WB = Collins WordbanksOnline Corpus. https://wordbanks.harpercollins.co.uk (June, 2018).

Adamson

Sylvia

. 2000. A lovely little example: Word order options and category shift in the premodifying string. In Fischer

Olga

Rosenbach

Anette

Stein

Dieter

(eds.), Pathways of change: Grammaticalization in English, 39-66. Amsterdam: John Benjamins.

Bolinger

Dwight

. 1972. Degree word s. The Hague: Mouton.

Breban

Tine

Davidse

Kristin

. 2016. The history of very: The directionality of functional shift and (inter)subjectification. English Language and Linguistics 20(2). 221-249.

Castroviejo

Elena

Gehrke

Berit

. 2017. “Good” as an evaluative intensifier. Paper presented at GLOW—Workshop Compositionality at the Interfaces, 14 March, Leiden.

Claridge

Claudia

Kytö

Merja

. 2021. Degree and related phenomena in the history of English: Evidence of usage and pathways of change. Journal of English Linguistics 49(1). 3-17.

10.

Collins = Collins dictionary. https://www.collinsdictionary.com (August, 2018).

11.

Davidse

Kristin

Ghesquière

Lobke

Njende

Ngum Meyuhnsi

. 2018. Pursuing the analogy between quality and quantity modification. Paper presented at the BAAHE 2018 Conference on Intensity, 30 November, University of Mons.

12.

Davidse

Kristin

Njende

Ngum Meyuhnsi

Ghesquière

Lobke

. Forthcoming. A Cognitive Grammar approach to the modification of absolute and relative quantifiers. Review of Cognitive Linguistics.

13.

Dixon

Robert

. 1982. Where have all the adjectives gone? And other essays in semantics and syntax. Berlin: Mouton de Gruyter.

14.

Feist

Jim

. 2012. Premodifiers in English. Cambridge: Cambridge University Press.

15.

Ghesquière

Lobke

. 2010. On the subjectification and intersubjectification paths followed by the adjectives of completeness. In Davidse

Kristin

Vandelanotte

Lieven

Cuyckens

Hubert

(eds.), Subjectification, intersubjectification, and grammaticalization, 277-314. Berlin: Mouton de Gruyter.

16.

Ghesquière

Lobke

. 2014. The directionality of (inter)subjectification in the English noun phrase: Pathways of change. Berlin: Mouton de Gruyter.

17.

Ghesquière

Lobke

. 2017. Intensification and focusing: The case of pure(ly) and mere(ly). In Napoli

Maria

Ravetto

Miriam

(eds.), Exploring intensification: Synchronic, diachronic and cross-linguistic perspectives, 33-53. Amsterdam: John Benjamins.

18.

Ghesquière

Lobke

Davidse

Kristin

. 2011. The development of intensification scales in noun-intensifying uses of adjectives: Sources, paths and mechanisms of change. English Language and Linguistics 15(2). 251-277.

19.

Ghesquière

Lobke

Davidse

Kristin

Njende

Ngum Meyuhnsi

Breban

Tine

. 2016. Degree and quantity modification: The diachronic development of very and whole. Paper presented at the 4th conference of the International Society for the Linguistics of English (ISLE 4), 18-21 September, Poznań.

20.

González-Díaz

Victorina

. 2010. Iconicity and subjectivisation in the NP: The case of little. In Conradie

C. Jac

Johl

Ronél

Beukes

Marthinus

Fischer

Olga

Ljungberg

Christina

(eds.), Signergy, 319-345. Amsterdam: John Benjamins.

21.

González-Díaz

Victorina

. 2018. Adjective stacking in Early Modern English: Some stylistic considerations. In Leung

Alex Ho-Cheong

van der Wurff

Wim

(eds.), The noun phrase in English: Past and present, 47-75. Amsterdam: John Benjamins.

22.

Hare

R. M

. 1952. The language of morals. Oxford: Oxford University Press.

23.

Hoffmann

Sebastian

. 2004. Using the OED quotations database as a corpus: A linguistic appraisal. ICAME Journal 28. 17-29.

24.

Kennedy

Christopher

McNally

Louise

. 2005. Scale structure, degree modification, and the semantics of gradable predicates. Language 81(2). 345-381.

25.

Langacker

Ronald W.

1987. Nouns and verbs. Language 63(1). 53-94.

26.

Langacker

Ronald W.

1991. Foundations of Cognitive Grammar, vol. 2, Descriptive application. Stanford: Stanford University Press.

27.

Langacker

Ronald W.

2016. Nominal grounding and English quantifiers. Cognitive Linguistic Studies 3(1). 1-31.

28.

Macmillan = Macmillan dictionary. https://www.macmillandictionary.com (August, 2018).

29.

Nevalainen

Terttu

Rissanen

Matti

. 2002. Fairly pretty or pretty fair? On the development and grammaticalization of English downtoners. Language Sciences 24(3-4). 359-380.

30.

Paradis

Carita

. 2000. Reinforcing adjectives: A cognitive semantic perspective on grammaticalization. In Bermúdez-Otero

Ricardo

Denison

David

Hogg

Richard M.

(eds.), Generative theory and corpus studies: A dialogue from 10 ICEHL, 233-258. Berlin: Mouton de Gruyter.

31.

Paradis

Carita

. 2001. Adjectives and boundedness. Cognitive Linguistics 12(1). 47-65.

32.

Paradis

Carita

. 2008. Configurations, construals and change: Expressions of DEGREE. English Language and Linguistics 12(2). 317-343.

33.

Paradis

Carita

Willners

Caroline

. 2006. Antonymy and negation: The boundedness hypothesis. Journal of Pragmatics 38(7). 1051-1080.

34.

Quirk

Randolph

Greenbaum

Sidney

Leech

Geoffrey

Svartvik

Jan

. 1985. A comprehensive grammar of the English language. London: Longman.

35.

Radden

Günter

Dirven

René

. 2007. Cognitive English grammar. Amsterdam: John Benjamins.

36.

Rohdenburg

Günter

. 2013. Using the OED quotations database as a diachronic corpus. In Krug

Manfred

Schlüter

Julia

(eds.), Research methods in language variation and change, 136-157. Cambridge: Cambridge University Press.

37.

Tribushinina

Elena

. 2008. Cognitive reference points: Semantics beyond the prototypes in adjectives of space and colour. Utrecht: LOT.

38.

Umbach

Carla

. 2015. Evaluative propositions and subjective judgments. In Meier

Cécile

van Wijnbergen-Huitink

Janneke

(eds.), Subjective meaning: Alternatives to relativism, 127-168. Berlin: Mouton de Gruyter.