Grammaticalization, subjectification and leftward movement of English adjectives of difference in the noun phrase

Folia Linguistica 42/2 (2008), 259–306. ISSN 0165–4004, E-ISSN 1614–7308 © Mouton de Gruyter – Societas Linguistica Europaea

Grammaticalization, subjectification and leftward movement of English adjectives of difference in the noun phrase1 Tine Breban Research Foundation-Flanders – K.U. Leuven This paper deals with grammaticalization and subjectification processes affecting adjectives in the English noun phrase (henceforth NP). It investigates Adamson’s (2000) claim that grammaticalization in the NP is accompanied by movement of the grammaticalizing adjective to more leftward positions in the premodifying string. I will argue that this type of movement is implied in the concepts of grammaticalization (Traugott 1989, 1995) and subjectification (Langacker 1990, 1998, 1999) as they apply to the NP. I will then illustrate my argument with a semantic analysis of six adjectives of difference, different, distinct, divers(e), several, sundry and various, whose diachronic development is established on the basis of six corpus studies. Keywords: grammaticalization, subjectification, leftward movement, English noun phrase, adjective

1 I would like to acknowledge the financial support provided by the Research Council of K.U.Leuven, which awarded me a postdoctoral scholarship from January until September 2007 (PDM/06/077), the Research Foundation-Flanders, which funds my current postdoctoral research project (FWO-O6260), as well as the Interuniversity Attraction Poles Programme – Belgian State – Belgian Science Policy, project P6/44 Grammaticalization and (inter)subjectification, the Spanish Ministry of Education and Science (grant no. HUM2007–60706/FILO) and the European Regional Development Fund. Thanks are also due to two anonymous Folia Linguistica reviewers for valuable comments and suggestions.

Brought to you by | Tobb Ekonomi ve TeknolojiAuthenticated | 193.140.109.10

Download Date | 4/23/14 12:34 PM

Tine Breban

260

1. Introduction: grammaticalization, subjectification and leftward

movement

Recently in grammaticalization studies, increasing attention has been directed at grammaticalization in a previously neglected area, viz. the noun phrase (NP). There have been a small number of papers focusing on specific topics such as measure nouns (Brems 2003, 2004), type nouns (Tabor 1994; Denison 2002; De Smedt et al. 2007), intensifiers (Rissanen 1999; Paradis 2000; Lorenz 2002; Nevalainen & Rissanen 2002; Méndez-Naya 2003, 2006, 2008). Adamson (2000) takes a more general point of view and formulates one of the most promising hypotheses about grammaticalization in the NP. She correlates functional shifts involving an increase in subjectivity with leftward movement in the English NP.2 Adamson (2000) explicitly claims that leftward movement in the English NP is connected with processes of SUBJECTIFICATION as defined by Traugott (1989, 1995), i.e. the acquisition of meanings in which the speaker expresses a personal stance. However, some of the examples that she discusses are characterized by processes of grammaticalization without involving this kind of subjectification. These include the development of determiners. This paper focuses precisely on the grammaticalization and subjectification of the English adjectives of difference from descriptive elements into secondary determiners and quantifiers. Using Langacker’s (1990, 1998, 2002) theory of subjectification, it will be argued that the association of grammaticalization/subjectification of these elements with leftward movement can in fact be given a theoretical foundation, which has not been pointed out in the literature so far: it can be

2 Adamson (2000) notes that this diachronic claim accords with the synchronic observation by amongst others Quirk et al. (1972), Hetzron (1978), Dixon (1982), Vandelanotte (2002), who propose that the order of prenominal adjectives in the NP embodies a left-right continuum from more subjective uses to more objective ones. Within the sentence, the association between leftward movement and grammaticalization, as exhibited for example by discourse markers moving to first position, is an established line of thought (see the discussion in Adamson 2000: 39–42). Bybee (1985; see also Traugott forthcoming) mentions a similar association for elements expressing mood in the verb phrase.



Grammaticalization, subjectification and leftward movement of adjectives

261

explained by the way the functional organization of the NP is reflected in its structural organization in English.3 1.1. Grammaticalization, subjectification and leftward movement in

the English NP As argued in Langacker (1991), the English NP can be characterized as a combination of four semantic functions, viz. TYPE SPECIFICATION, INSTANTIATION, QUANTIFICATION and GROUNDING, which interact in the following way. The entity denoted by a NP is construed as an instance of a certain type, which is specified in terms of its size and its relation to the speech event and its participants. In canonically coded NPs (Langacker 1991: 59), type, quantity and grounding (the latter two both presupposing and realizing instantiation) are expressed by separate elements in the NP: type by head noun and modifiers, quantity and grounding by quantifiers and determiners. In the English NP, these elements typically occur in a certain order: determiners and quantifiers tend to precede modifiers, which in turn precede the head noun (see amongst others Halliday 1994[1985]; Bache 2000). Type specification is hence associated with the righthand side of the NP whereas instantiation is effected at the lefthand side.

From a semantic point of view, the functions of type specification and instantiation are associated with different kinds of conceptual operations. The type specification enables the addressee to envisage the type of thing that the speaker is talking about on the basis of the description of the entity-type. The instantiation (encompassing quantification and grounding) is concerned

3 The claim made here for the English NP can possibly be extended to other languages in which the NP displays LEFTWARD BRANCHING, i.e. the head noun of the NP is found at the right end of the NP with the different modifiers such as articles, quantifiers and adjectives preceding it. With regard to other languages in which the NP branches out to the right or in which leftward and rightward branching are combined, the type of movement discussed here as leftward movement for English could take the form of PERIPHERY MOVEMENT (see Traugott forthcoming). This more general claim accords with the cross-linguistic functional model of the NP proposed by Rijkhoff (2002) on the basis of the typological investigation of 52 languages, in which the head noun constitutes the functional centre and the different modifiers are arranged on hierarchical layers around it. As a result, Rijkhoff claims that the order of the different modifiers with respect to one another remains the same independent of their location with respect to the head noun.



Tine Breban

262

with the conceptualization of the specific entity denoted as corresponding to this type-description. This is an abstract cognitive process shared by all NPs and hence no longer tied to the lexically specific. As a result, the righthand side of the NP, which specifies the type, typically features lexical elements with descriptively specific semantics, while at the lefthand side, which is concerned with instantiation, elements with general grammatical semantics are found. For the same reason, processes of grammaticalization in the NP can be expected to involve a shift from right to left.

For some grammaticalization processes the structural shift to the left of the NP has further conceptual relevance, viz. those that can be captured by Langacker’s notion of subjectification. As analysed by Langacker (1990, 1998, 1999), subjectification is a construal-related process that is part of the conceptualization act.4 Subjectively construed relations display the following characteristics. Firstly, they typically invoke the ground, usually speaker and addressee, as reference point. Secondly, the relations, which become foregrounded in the process of subjectification, are inherent in directly accessible objective relations that constitute the original objects of the conceptualization. Thirdly, even though they are foregrounded, the subjective relations do not become the object of the conceptualization themselves.

In the NP, such construal-related subjective relations take part in the grounding of the NP. Subjectification processes, by consequence, involve the development of determiner and quantifier meanings on the basis of originally descriptive meanings conveying objective relations. Syntag-matically, these processes are again reflected in a move towards the left end of the NP. 1.2. A case study: English adjectives of difference In this paper, I will investigate the diachronic development of a specific group of adjectives, the English adjectives of difference, which strongly bears

4 This interpretation of subjectification differs considerably from that of Traugott, who defines the process as “the tendency to recruit lexical material for purposes of creating text and indicating attitudes in discourse situations” (Traugott 1995: 47) or, more recently, as a type of semantic change causing meanings “to become increasingly based in the SP(eaker)/W(riter)’s subjective belief state or attitude toward what is being said and how it is being said” (Traugott 2003a: 125).




263

out the relation between grammaticalization/subjectification and leftward movement. The adjectives in question, viz. different, distinct, divers(e), several, sundry and various, express ‘difference’ in a broad sense. As we will see, they cover three different uses: an attribute use (Halliday 1994[1985]: 184) expressing that two entities are unlike each other in some respect, which is illustrated in (1), a postdeterminer use (Halliday 1994[1985]: 183) as markers of plurality/distributivity, as illustrated by (2) and (3), and a quantifier use, illustrated in (4). (1) This Hieroglyphicall devise doth so affect Children (who are generally

forward to communicate what they know) that I have observed them to teach others, that could not so readily learn, to know all the letters in a few houres space, by asking them, what stands A. for? and so concerning other letters backwards and forwards, or as they best liked. Thus when a childe hath got the names of his letters, & their several shapes [‘their physically distinct shapes’] withall in a playing manner, he may be easily taught to distinguish them in the following leaf. (HC1500–1710)5

(2) When we came to the Town, two several Churches [‘two instances of churches’] strove to receive us; but having some Acquaintance with the Father of the one, and not with the other, we excused ourselves to the latter, and took up with our Friend. (HC1500–1710)

5 The historical corpus data marked HC and CLMET are taken from two historical databases, viz. the Helsinki Corpus and the Corpus of Late Modern English Texts. The diachronic part of the Helsinki corpus is a 1.5 million word corpus that spans the period 750 to 1710. It aims to offer a representative coverage of the written language from the Old, Middle and Early Modern (British) English periods, whereby attention has been paid to geographical dialect, type and register of writing and sociolinguistic variation (see Rissanen et al. 1993). The CLMET was recently compiled by Hendrik De Smet on the basis of texts drawn from the Project Gutenberg and the Oxford Text Archives (see De Smet 2005). It consists of almost ten million words and covers the period 1710–1920. The synchronic examples marked CB are extracted from the contemporary English COBUILD corpus, which is a 56 million word selection of the Bank of English that can be accessed via the Collins WordbanksOnline service, and are reproduced here with the kind permission of HarperCollins Publishers.



Tine Breban

264

(3) Maximus enforced his exhortations by a liberal donative, purified the camp by a solemn sacrifice of expiation, and then dismissed the legions to their several provinces [‘one legion to one province, another one to another province, etc.’], impressed, as he hoped, with a lively sense of gratitude and obedience. (CLMET1710–1780)

(4) Paul came back and loaded his ship again and managed to make a few trips without encountering the pirates. Thus he was able to make and keep some profits. Now he was able to purchase a decked vessel of twelve tons in which he made several successful commercial voyages [‘a number of successful commercial voyages’] to the Connecticut coast. (CB)

In Sections 2 and 3, I will argue that these uses are related through processes of grammaticalization and subjectification, leading in a first stage from the propositional attribute use to the grammatical postdeterminer use and, in a second stage, from the postdeterminer use in plural NPs to the more grammaticalized quantifier use. Section 2 looks at the development from a theoretical, synchronic point of view, whereas in Section 3 the hypothesized evolution is investigated on the basis of diachronic samples for the six adjectives. In Section 4, I discuss how each new stage in the grammatical-ization of these adjectives involves further subjectification and goes together with a move to a more leftward position in the NP.

In Breban (2002/2003, 2006a), it was proposed on the basis of synchronic and diachronic corpus analyses that this path of grammaticalization could be found with the adjective different (synchronic + diachronic) and its Dutch counterparts verschillend and verscheiden (synchronic only). In the literature, a number of other studies deal with this development, or more accurately, with the polysemy involving a descriptive sense and a plurality/distributivity reading of adjectives such as different. These studies can be roughly divided into two groups. Firstly, there are a number of articles focusing specifically on the distributive reading of different (and same) (Carlson 1987; Moltmann 1992), sometimes in comparison with its equivalents in German (Beck 2000) and/or French (Laca & Tasmowski 2001; Tovena & Van Peteghem 2002). I will return to this topic in Section 2.2, when I look in more detail at the plurality use of the English adjectives of difference. The second, more recent, group of articles have in common that they invoke diachronic development to explain the association of distinct senses of different with the distinct categories of adjective and determiner. Laca & Tasmowski (2003) deal with French différent(s), but the general thrust of the article can be extended to English different. They argue that it is necessary from a syntactic as well as a




265

semantic point of view to distinguish A(djective)-différent and D(eterminer)-différents. They explain the latter as having evolved from the former through a process of specialization (Laca & Tasmowski 2003: 160). According to Laca & Tasmowski, English different has not reached the status of determiner yet. Tovena & Van Peteghem (2006) specify that the transition from A-différent to D-différents involves an intermediate step in which différent can still be classified as adjective but has undergone considerable semantic bleaching and has come to express that the plural set referred to is composed of individual instances. Denison (2006) discusses the development of a few adjectives of this class, viz. divers(e), several, certain and various, on the basis of their lemma descriptions in the OED. He argues that the development of these adjectives shows that there is not only semantic but also morpho-syntactic gradience between the categories of adjective and determiner.

In this article, I will take further the line of thought established by the latter group of articles and my own earlier work. I will further develop the seminal argumentation presented there for the category shift from attribute to postdeterminer to quantifier manifested by a subset of the adjectives of difference in English. The diachronic and synchronic arguments that I will put forward are based on the systematic analysis of corpus examples of the six adjectives for subsequent periods ranging from Middle English (1500) to Present-day English.6 In general, I will propose two major amendments with respect to Laca & Tamowski (2003), Tovena & Van Peteghem (2006), and Denison (2006). Firstly, I interpret the diachronic evolution within the framework of grammaticalization (cf. Traugott 1988, 1989, 1995) and subjectification as defined by Langacker (1990, 1998, 1999), which provide the necessary tools for a better understanding of the processes involved. Secondly, I propose that it is necessary to distinguish different subtypes of premodifying adjectives in the NP, which especially in the article of Laca & Tasmowski are dealt with as one monolithic group. I argue that this sheds more light on the different stages involved in the actual transition from prototypical adjective to postdeterminer and quantifier. It is this successive development that I will show is reflected iconically in the leftward movement of the adjectives in the NP. The overall aim of this article, then, is to arrive at a comprehensive overview of the diachronic development and the different uses of adjectives of difference such as different, various and several. The

6 With the exception of sundry, none of the adjectives studied figured in the older sections of the Helsinki Corpus (750–1500).



Tine Breban

266

results of this investigation will be tied up with earlier research on other and different (Breban 2002/2003, 2006a; Breban & Davidse 2003) in order to arrive at a general profile of the diachronic evolutions going on in the semantic field of difference in English (Section 3.5).

2. The three uses of the adjectives of difference: a diachronic interpretation

As indicated in the introduction, several studies have recognized the diachronic dimension to the polysemy of adjectives of difference, but it has never been studied in detail. Articles so far have focused on the beginning and end of the process, i.e. on the attribute and the postdeterminer/ quantifier use, rather than on the process itself (e.g. Laca & Tasmowski 2003; Tovena & Van Peteghem 2006). Discussions of the whole path of development have remained rather general (Breban 2002/2003, 2006a; Denison 2006). This section will attempt to fill this gap and provide an in-depth discussion of the different steps in the diachronic development of these adjectives based on the corpus study of different, distinct, divers(e), various, several and sundry. 2.1. The attribute use of the adjectives of difference In a first set of examples, the adjectives of difference attribute the quality of being different or dissimilar, in the sense of not sharing (m)any qualitative features, to two (or more) entities being compared. (5) Unlike film or television, theatre is a one-off event. No re-takes no going

back and trying things again. Lights, sound, action – it all happens live in front of you! That’s why every night is different and why theatre has survived and flourished for literally thousands of years! (CB)

(6) Billy starred as a drug dealer in the hard-hitting Robert Carlyle drama Looking After Jo Jo. That was after he played the sleazy nightclub owner George Skelly in Taggart. Then there were two episodes of The Bill as two completely different characters. (CB)

As illustrated in (5) and (6), this qualitative meaning can be expressed by two different uses of the adjectives, viz. that of predicative adjective in a copular




267

clause and, the use that interests us here, the attribute use in the NP (cf. Halliday 1994[1985]: 184). When expressing this meaning, the adjectives are typically gradable, as in, for instance, two completely different characters (cf. Breban & Davidse 2003: 282–284; Laca & Tasmowski 2003: 156). Structurally, the attribute takes up the central position in the prenominal slot of the NP: before classifying adjectives, which derive a more detailed subtype from the general type denoted by the head noun (Halliday 1994[1985]; Bache 2000), and following determiners and quantifiers. In (7), for example, the attribute different is together with its submodifier very, located after the quantifier two but before the classifier male. (7) Toy Story is in essence a buddy movie, about the bond that develops

between two very different male types who, in this story, just happen to be toys who discover the value of true friendship. (CB)

The semantics of the quality-attribution sense of the adjectives of difference are propositional (Traugott 1982, 1988) in character: they give a lexical description of a quality, viz. unlikeness, observable in the world that is discussed.

What I suggest in this article (see also Breban 2006b) is that on the basis of the lexical meaning of unlikeness, the adjectives then develop a new, more grammatical, sense, viz. difference as non-identity between discourse referents. This notion of ‘non-identity of reference’ is implied in the concept of propositional unlikeness itself: if two entities, prototypically two instances of the same general type, are different in the sense of not sharing certain distinguishing qualitative features, they are ipso facto two separate entities rather than the same one. In examples such as (8), that secondary (implied) meaning is given greater prominence, while the meaning of qualitative difference is backgrounded. More concretely, in (8), the first interpretation of the NP different products is that of simply two different instances of self-tanning products, one for the face and another one for the body. But the following description alludes to the qualitative differences characterizing the products. (8) Self tans work best on an existing tan, no matter how light. On pale,

untanned skin they will add colour but even the best can still look a little fake: the effect is more natural under make-up. Ideally, use different products for face and body – those for the face contain less tanning ingredient and more moisturiser. (CB)



Tine Breban

268

In the grammaticalization literature, contexts such as these are referred to as BRIDGING CONTEXTS, i.e. contexts that trigger an inferential meaning that is more plausible than the source meaning of the utterance (cf. Heine 2002). The notion of bridging contexts that is used here is Evans & Wilkins’s (2000) slightly broader concept, which includes those contexts in which two distinct meanings are plausible, each of which is supported by elements from the context. Evans & Wilkins view bridging contexts as a stage preceding polysemy, and hence as an indicator of ongoing semantic change. In agreement with Diewald (2002: 117), I would like to stress the formal dimension to the notion of a bridging context. Besides being semantically plausible, the two interpretations also have to make sense formally within the existing structure. For the different adjective uses concerned here, this means that the two senses that semantically fit in with the context have to be available from a structural point of view as well. Concretely, this means that for an example to be a possible bridging context, it has to be possible for the adjective to fill the central position in the NP, associated with the attribute reading, as well as the postdeterminer position associated with the new meaning of non-identity of reference (cf. Halliday 1994[1985]; Bache 2000). 2.2. The postdeterminer use of the adjectives of difference In a second set of data, the meaning of ‘referential non-identity’ is the only possible interpretation of the adjectives of difference. Functionally, these data do not express an “attribute” meaning: the adjectives do not attribute a quality to the entity designated by the NP, but indicate non-identity of referents, e.g. (9). (9) I think when you analyse the fact that we won five Super Bowls and ‘ve

done it with different head coaches, different quarterbacks and different players and the only real constant has been team owner Ed debartolo, it says something about the organisation. (CB)

In this example, different means ‘not the same ones’, which contrasts with the notion of ‘the same team’ evoked in the context. When interpreted from a diachronic point of view, examples such as (9), in which only a post-determiner reading is possible, testify to the fact that the postdeterminer use has gained a separate status from the lexical attribute use. For some adjectives, e.g. several, the original meaning even disappears when the new




269

meaning has gained ground. In terms of Traugott’s (1982, 1988, 1995) concept of grammaticalization, the adjective has acquired a different type of meaning, viz. a TEXTUAL meaning concerned with the referential status of the designated instance (cf. Breban 2006b). This semantic shift not only involves grammaticalization, but also subjectification (Langacker 1990, 1998, 1999). The adjective has come to express an aspect of the conceptualization of the entity as one referent within a set of referents. As such, its function is similar to that of the determiner in the NP, which also provides information about the identifiability of the referents of the NP. Within a functional description of the NP based on Halliday (1994[1985]), the adjective can be said to function as a postdeterminer, or secondary determiner that provides additional information with regard to the identifiability of the referent(s) of the NP. As for its structural properties, the adjective takes up the typical position of the postdeterminer in the NP: right behind the determiner or quantifier (Halliday 1994[1985]; Bache 2000).

A more precise semantic analysis reveals that this postdeterminer meaning of non-identity is interpreted slightly differently depending on the specific formal environment created by the NP.7 The semantics of the adjectives of difference, whether they be (gradable) difference in the descriptive sense or non-identity, always involve a relational component: they crucially involve a relation between (at least) two entities. A first entity is said either to be dissimilar to a second one or to be another entity than a second one. Now, the discourse realization of the second entity involved in the comparison can have different formal relations with respect to the NP denoting the first entity: it can either be expressed or implied somewhere in the discourse preceding or following the NP referring to the first entity (as in 10) or it can be expressed by the NP itself (as in 11). In the former case, comparison can be said to be NP-external, while in the latter it is NP-internal. (10) On the basis of some 6,000 interviews, that took place in six

underdeveloped countries over a ten-year period, they find evidence of an “Overall Modernity” syndrome. “Modern Man” (who appears to be without modern woman) is truly modern when he has changed as an individual, and modernity is indicated by the presence of a distinct set of

7 This fits in with the current conception of grammaticalization as involving items within constructions rather than lexical items as such (e.g. Heine 2003; Traugott 2003b; Himmelmann 2004).



Tine Breban

270

attitudes, which may be summarized as follows: 1) A readiness for new experience and an openness to innovation; 2) An interest in things other than those of immediate relevance; … (CB)

(11) It is an attempt to find out if there are people, other than those who have tested positive, who have knowledge or (have) been involved in doping in China. We have asked to speak to the normal suspects, the various bodies within the Chinese swimming association as well as swimmers, coaches and officials who may have assisted these athletes. (CB)

The referential non-identity meaning of the adjectives discussed here interacts with these two formal realizations in the following way. In the case of NP-external comparison, the adjective expresses a relation of non-identity of the referent of the NP with respect to another instance of the same general type available elsewhere in the discourse. In this way, it construes a “phoric” relation of non-identity between the referent of the NP and the other entity (cf. Halliday & Hasan 1976): this use can be referred to as the PHORIC

POSTDETERMINER use (Breban & Davidse 2003). With NP-internal comparison, the adjective conveys the non-identity of the multiple referents of the NP, as in (11). More specifically, it indicates that the instantial set denoted by the NP is conceived as composed of individual instances rather than as a single set in its entirety.

The latter postdeterminer function can be further clarified on the basis of Langacker’s analysis of plural nouns as a special type of mass nouns (Langacker 1991: 77). According to Langacker, conceptualization as a heterogeneous set is the distinguishing property setting plural nouns apart from mass nouns which denote a homogeneous set. As this aspect of heterogeneity remains unprofiled in a regular plural NP, the addition of a postdeterminer such as various in (11) allows the speaker to put this heterogeneity in profile. This type of postdeterminers is hence especially frequent in contexts in which the different instances composing the instantial set are in some way individually addressed: for example, contexts in which the separate instances are counted, e.g. (12), contexts that further identify and/or describe the instances separately in the following discourse, e.g. (13), or contexts in which the instances are involved in separate processes, with which they interact distributively (see below). (12) If you suspect that you may suffer from a food allergy, you should consult

a doctor. Many different ingredients are present in the food you buy, so it’s not always easy to know what you can or can’t eat. (CB)




271

(13) I began to realise that being a single mum has distinct advantages. I don’t need to worry about preparing proper dinners or whether there are enough clean shirts. I can do what I want when I want, dress for comfort rather than style and breastfeed all day without worrying about the psycho-sexual implications for my partner. And there are no rows about why the baby never sleeps. (CB)

There is however a further complication to this second postdeterminer

meaning: internal comparison is not restricted to plural NPs, where the interpretation of the adjective expressing non-identity is very straightforward, but there is also a small set of examples in which the NP itself is singular and the presence of multiple referents is only implied by the semantic context (cf. Carlson 1987; Laca & Tasmowski 2001). In examples such as (14), the idea of plurality is present in the context, which causes the singular NP to be interpreted as involving multiple referents. The context can be said to have a distributive effect on the singular NP. (14) As at Hidcote, the garden here is divided into a sequence of compartments, each with a distinct theme. (CB)

The distributive reading is also a possible variant of the basic plurality reading of plural NPs with internal comparison. Such a distributive reading becomes available when the context contains a plural element interacting with the meaning expressed by the postdeterminer, as in (15). (15) Five experts on different zones of the body are waiting to answer your queries in person on our Health Works Hotline. (CB)

This distributive reading has been the subject of a number of studies building and commenting on each other (Carlson 1987; Moltmann 1997; Beck 2000; Laca & Tasmowski 2001; Tovena & Van Peteghem 2002). The main aim of these studies was to determine which element in the semantic context creates the distributive effect and functions as licensing factor. For the purposes of this article, the exact characterization of the licensing elements is outside the focus of attention. However, it is necessary to react against one of the conceptions underlying this quest, viz. the assumption that these distributive uses constitute a separate category that can be dealt with without relating them to other postdeterminer uses. In my opinion, the distributive use is simply a variant of the general (internal) postdeterminer



Tine Breban

272

use as a plurality marker, which can be triggered by certain elements in the context.8

Summing up, the new postdeterminer meaning of the adjectives of difference covers three different MODULATIONS of the basic meaning of referential non-identity: the phoric meaning, the plurality meaning and its contextually invoked variant, the distributive meaning. The focus in the rest of this article with respect to the postdeterminer will be on the plurality meaning (including the distributivity meaning in plural NP contexts), although other postdeterminer senses will be included in the discussion where relevant. The selection of the six adjectives discussed here was based precisely on the fact that these adjectives typically express the plurality meaning when used as a postdeterminer. 2.3. The quantifier use of the adjectives of difference The third use of the adjectives of difference is restricted to the formal environment of plural NPs with internal comparison. It expresses unspecified absolute quantification and can be paraphrased as ‘more than one’. Present-day English several typically conveys this quantifier use, as in (16). As observed by Langacker (1991: 84), several refers to a schematic number ranging from three to about seven, as it must remain possible to simultaneously hold the entities in mind as individuals. (16) It is a light room with tall windows overlooking the patio and some far

from industrious builders on the roof of a neighbouring building. There is a wall of mirror-doored wardrobes, an old, leather-topped desk, a freestanding mirror and several chintz-covered chairs over which are flung large, wool, deep-coloured paisley shawls. (CB)

The quantifier use of the adjectives of difference as represented by several

seems to be very similar to the meaning of some, which refers to an unspecified amount larger than one. It also displays the same contextual

8 The close relation between the two interpretations is recognized by Tovena & Van Peteghem (2002), who group the two together under one heading, but they place them on an equal footing, while it is argued here that the plurality reading is the overarching interpretation, with the distributive reading a contextually determined subtype of the former.




273

restrictions as some, as both occur only in affirmative contexts.9 Furthermore, as will be illustrated below, several of the adjectives display the same alternation as some between the absolute quantifier meaning indicating an unspecified amount and a relative quantifier reading indicating a portion (of unspecified size) of a contextually specified reference mass (cf. Milsark 1977). Langacker (2004: 104) refers to the two uses of some as “sm”, i.e. unstressed some, and “some”.

Langacker (2004: 104, 107) analyses sm as a grounding predication belonging to the set of indefinite articles. In my opinion, several functions in a very similar way, with the distinction that because its quantifier meaning is somewhat more specific, it places more emphasis on the size of the instance denoted. From a more general point of view, some and several seem to be part of a continuum of increasing quantitative specificity with pure indefinite reference acting as basic value.10

9 It has to be noted that some as well as the adjectives of difference can be used in non-affirmative contexts when they are used as a relative rather than an absolute quantifier (see below). 10 From this perspective, the analysis proposed here tallies with Davidse’s (2004) analysis of absolute quantifiers. According to Davidse (2004), absolute quantifiers express the same grounding relation as indefinite determiners, viz. they “instruct the hearer to conceptualize instances as corresponding to the categorization provided by the type specification of the NP” (Davidse 2004: 217). In contrast to real indefinite determiners such as the indefinite article and the zero article, however, absolute quantifiers merely presuppose this correspondence relation between type specification and instantiation, in order to be able to convey the size of the instantiation (Davidse 2004: 211).

An additional benefit of this analysis of absolute quantifiers, including several, is that it offers an explanation for the occurrence of these elements in definite NPs such as (i). (i) A translation company can find the several different translators needed for

the different subjects. (http://www.antlertranslation.com/html/faqs.html# We%20have%20people%20in%20the%20office)

Because the indefinite grounding relation is only implied in the act of absolute quantification, it can be superseded by an element expressing definite identification. In such NPs, rather than functioning as primary determiner, the absolute quantifier fulfils the role of secondary determiner which adds further identifying or quantifying information (see Bache 2000: 241–242).



Tine Breban

274

Structurally, the quantifier uses of the adjectives of difference usually occur in the same position as some, i.e. the first position in the NP, as in (17).11 (17) If you are considering help of this kind we advise you to approach several

different consultancies. Make sure you know exactly what services each is offering and what it will cost. In this way you can decide which organisation is most likely to meet your particular needs, and you can ensure that you get value for money. (CB)

The diachronic hypothesis I propose then is that the quantifier use constitutes a further step in the grammaticalization process, which takes the plurality-marking postdeterminer use in plural NPs as input. Concretely, it seems likely that these adjectives developed from denoting that the NP refers to a set of distinct individual instances to actually measuring the size of the set as a not exactly specified number of instances that can still be “held in mind as individuals” (Langacker 1991: 84; cf. the discussion of the development of divers(e), several and various in Denison 2006). 2.4. Diachronic hypothesis about the adjectives of difference Taking a dynamic perspective, I would like to propose that the English adjectives of difference display LAYERING (Hopper 1991: 22) in the sense of coexisting synchronic uses reflecting historical changes such as grammaticalization. In Present-day English, we find both original and grammaticalized uses of adjectives of difference, i.e. attribute versus postdeterminer and quantifier uses respectively. The postdeterminer and quantifier meanings that the adjectives can acquire are more grammatical than the attribute meanings from which they develop. While the qualitative attribute meaning conveys propositional content, the postdeterminer is no longer concerned with the description of entities. Rather, it marks the referential status of the instances being designated, either as distinct referents within the instantial set (which can be “distributed” over various instances of the same process), or as instances related to other discourse referents of the

11 In the few examples in which the quantifier use is combined with definite reference, the adjective of difference takes up the position directly following the definite determiner, but preceding any other postdeterminers; cf. (i) in footnote 10 above.




275

same general type. The quantifier use has delexicalized further and has shifted to the fully grammatical category of determiners/quantifiers. From a semantic point of view, increasing subjectification is the driving force behind the evolution of the adjectives of difference. Each of the new meanings is characterized by increased subjectification with respect to the preceding meanings. The postdeterminer meaning either profiles the internal organization of the instantiation, its plurality, or its relation to other discourse referents, backgrounding the original meaning of qualitative difference. However, although this use is more subjective in the sense that it is involved in the conceptualization of the referent of the NP (cf. Langacker 1990, 1991: 89–95), it has not yet attained the full subjective status of grounding predication. That is to say, they do not effect the actual grounding of the entity denoted by the NP. As discussed in Davidse et al. (2008) these postdeterminer uses are only SECONDARY determiners. They are comparable to secondary auxiliaries such as be going to, which require additional grounding themselves and can take other elements than the ground as reference point (Langacker 2002: 23). In the same way, the postdeterminers talked about here have to be grounded by another primary determiner or quantifier. In contrast to the secondary auxiliaries, however, they cannot take a reference point other than the ground.12

The second semantic shift turns the adjective into an absolute quantifier, functionally similar to some. The adjective now constitutes a grounding predication in its own right, instructing the addressee to establish relations of correspondence between the instances denoted by the NP and the type it describes in order to be able to “count” them.

In conclusion to this theoretical discussion, I would like to make a few remarks on the formal side of the grammaticalization process manifested by the adjectives of difference. In the current literature on grammaticalization it is an issue of debate whether one can talk about grammaticalization as a semantic process only, without any distinctive accompanying formal characteristics, or whether these formal characteristics, such as reanalysis, are

12 Davidse et al. (2008) argue that the inability of these postdeterminers to take a reference point different from the ground is due to their particular semantics of referential non-identity, which is naturally closely related to the grounding notion of identification. They show that there is no such restriction for the postdeterminer uses of other adjectives such as old, regular and necessary, which often take distinct reference points from the ground.



Tine Breban

276

a necessary characteristic of any real process of grammaticalization (e.g. Haspelmath 1998; Campbell 2001; Fanego 2004; Lehmann 2004). Without entering into this discussion, I would like to point out that the different developments discussed here show a pairing of semantic and formal shifts.

Firstly, the final step in the process of grammaticalization clearly involves a process of reanalysis, viz. the category shift from adjective to quantifier. For the shift from attribute to postdeterminer, the formal reflexes of the development are less obvious as both are from a strict categorial point of view adjectives. But, as I have shown in the discussion above, the category of prenominal adjective is not a monolithic block, but consists of certain subcategories that are formally distinguished by their different positions in the prenominal slot as well as by their different syntactic behaviour (i.e. the predicative and gradable alternations of the attribute). In this respect, the postdeterminer is also formally distinct from the attribute. This approach to categories is indebted to the main claim made by Radical Construction Grammar, viz. that constructions rather than categories are the basic, primitive units of syntactic representation and that categories are in effect derived from constructions (Croft 2001: 45–47).

The development of the adjectives discussed here displays another formal reflex, a type of formal specialization, viz. a reduction of the formal contexts or constructions that the adjective can occur in. For the attribute uses, there are no limitations with respect to the NPs that they can be part of. But the postdeterminer data show that these uses are, with a few odd exceptions, limited to plural NPs. As discussed above, both possible interpretations of the postdeterminer in singular NPs, viz. the phoric and the distributive reading, are only minor uses. With regard to the quantifier uses, the type of NP that these uses are found in, is again more restricted: excepting a small number of examples containing several,13 they occur solely in indefinite plural NPs. In fact, the transition to the quantifier stage is characterized by an increase in postdeterminer uses in indefinite NPs (cf. Breban 2006a on

13 Although the quantifier use of several is overall restricted to indefinite NPs (cf. Lyons 1999: 36), there are a few examples of the adjective in definite NPs in the corpus material. As these are all later examples (in the corpus they are limited to the Present-day English data), it seems reasonable to consider the use of quantifier several in a definite NP context as a new development driven by analogy with other cardinal quantifiers such as the cardinal numbers and many. The development is hence part of the process of PARADIGMATICIZATION of several, which is, as noted by Lehmann (1985), a possible formal reflex of grammaticalization.




277

different). A possible explanation for this contextual restriction lies in the meaning of the quantifier use of these adjectives. The unspecified nature of the quantification expressed naturally links up with indefinite identification.

3. Case study of six adjectives expressing difference

The aim of this section is to discuss in some detail how the general development suggested in the previous section is instantiated by the individual adjectives that are part of the corpus study. As the adjectives appear to have grammaticalized at different paces and to different degrees, the main focus will be on which meanings the adjectives developed at which periods in history and which proportions the different meanings take up in the overall number of examples of the adjectives in the different periods.

Although the main part of this section focuses on the development of the six adjectives in non-phoric plural NPs, the development of less frequent grammaticalized uses such as the phoric use and the distributive use in singular NPs will be discussed briefly in Section 3.4. In the first subsections, the adjectives will be dealt with in terms of their degree of grammaticalization. More specifically, Section 3.1 deals with the two adjectives that have run the full course of grammaticalization, several and sundry. The discussion primarily focuses on several, because the use of sundry is restricted mainly to the idiomatic phrase all and sundry in Present-day English. Section 3.2 discusses the development of various, different and distinct, which show a lesser degree of grammaticalization. Diverse/divers will be dealt with in a separate section (3.3) as their development is complicated by spelling issues. In Section 3.5, I will discuss the way in which these corpus findings corroborate the grammaticalization-cum-subjectification hypothesis and I will formulate some suggestions as to how they can be interpreted in the light of the diachronic evolutions that characterize the field of difference in general.

The database that was used in this corpus study was compiled in the following way. Its diachronic part consists of five sets of historical data comprising 100 examples (if that many examples were available) for each of the adjectives. The first two sets, covering the periods 1250–1500 and 1500–1710, were extracted from the Helsinki Corpus and the other three sets, covering the periods 1710–1780, 1780–1850, 1850–1920, were taken from the CLMET (Corpus of Late Modern English Texts). In addition, the database



Tine Breban

278

includes a synchronic set of 200 examples extracted from the COBUILD corpus, which consists of data from 1990 onwards.

Before turning to the actual analysis, I would like to briefly make some cautionary remarks with regard to the quantitative findings that are presented here as evidence for the grammaticalization hypothesis. The results are of course to a certain extent biased by the corpus material itself: the available material is limited both in amount – often only few instances of a certain adjective were found for a particular period – and in genre, as it is restricted to written material that has survived the passing of time. I would like to point out that the quantitative analysis of the data is limited to the discussion of the relative frequencies of the different uses and invokes the basic hypothesis that grammaticalization can be observed in the increase of grammaticalized uses over subsequent periods (cf. Bybee 2003). Bybee (2003: 203–205) actually suggests that grammaticalization goes together with a double increase in frequency, viz. an increase in TYPE FREQUENCY, which deals with the spread over the type of contexts that the grammaticalizing item/construction can occur in, as well as in TOKEN FREQUENCY. In this paper, I will mainly be concerned with the latter type of frequency. Hoffmann (2004, 2006), who in his study of complex prepositions addresses the problems posed by the diachronic corpus investigation of low-frequency grammaticalizing items, has suggested that the quantitative limitations of the material available can be counterbalanced if the token frequency is measured against the frequency of the concept in all its different realizations. This proposal will be partially taken up here. I will not compare the frequency of a particular use of one adjective with the general frequency of this use, but I will compare it with the relative frequency of the same use for the other adjectives. 3.1. Several and sundry Within the set of six adjectives, two adjectives have grammaticalized completely, viz. sundry and several. They are, as shown in Tables 1 and 2, from respectively 1780 and 1850 onwards used only in the grammaticalized functions of postdeterminer and quantifier.




279

SUNDRY size of

sample pred attr attr-pd pd pd-quant quant distrib

sing pd phoric pd

1250–1500 15 100%

0 3 20%

0 7 46.7%

1 6.7%

2 13.3%

0 2 13.3%

1500–1710 47 100%

1 2.1%

3 6.4%

0 14 29.8%

10 21.3%

17 36.2%

1 2.1%

1 2.1%

1710–1780 15 100%

0 0 1 6.7%

5 33.3%

4 26.7%

5 33.3%

0 0

1780–1850 29 100%

0 0 0 3 10.4%

7 24.1%

19 65.5%

0 0

1850–1920 22 100%

0 0 0 5 22.7%

7 31.8%

10 45.5%

0 0

1990– 72 100%

0 0 0 9 12.5%

6 8.3%

55 76.4%

2 2.8%

0

Table 1. Diachronic overview of the different uses of sundry14

SEVERAL size of sample

pred attr attr-pd pd pd-quant quant distrib sing pd

phoric pd

1250–1500 0

0 0 0 0 0 0 0 0

1500–1710 100 100%

1 1%

8 8%

3 3%

40 40%

9 9%

37 37%

2 2%

0

1710–1780 100 100%

0 1 1%

2 2%

32 32%

10 10%

55 55%

0 0

1780–1850 100 100%

0 0 1 1%

9 9%

9 9%

81 81%

0 0

1850–1920 100 100%

0 0 0 7 7%

11 11%

82 82%

0 0

1990– 200 100%

0 0 0 10 5%

30 15%

160 80%

0 0

Table 2. Diachronic overview of the different uses of several In earlier periods however, they were used lexically, both as attributes in the NP and as predicative adjectives, as illustrated by (18)–(19) and (20)–(21) respectively. (18) This Hieroglyphicall devise doth so affect Children (who are generally

forward to communicate what they know) that I have observed them to teach others, that could not so readily learn, to know all the letters in a few houres space, by asking them, what stands A. for? and so concerning other letters backwards and forwards, or as they best liked. Thus when a childe

14 Abbreviations: PRED = predicative adjective; ATTR = attribute use; PD = postdeterminer use; DISTRIB SING PD = distributive singular postdeterminer; PHORIC

PD = phoric postdeterminer; QUANT = quantifier use.



Tine Breban

280

hath got the names of his letters, & their several shapes withall in a playing manner, he may be easily taught to distinguish them in the following leaf. (HC1500–1710)

(19) 1614 W. B. Philos. Banquet (ed. 2) 113: The sundryest kindes of extremities. (OED s.v. sundry, a. 4, obs.)

(20) … and clense with your hand both the lidde and inward sides of the churne, and hauing put all together, you shall couer the churne againe, and then with easie stroakes round, and not to the bottome, gather the butter together into one intire lumpe and body, leauing no peeces thereof seuerall or vnioyned. (HC1500–1710)

(21) For Prouidence includith all, whither they be diuers or infinite, but Desteny deuideth euery thing according to her motion, distributing it to place, to forme, & tyme: that this deuiding of temporall order joyned to the diuine pleasure may be made Prouidence, But that joyning, being seuerd & deuided into tymes, that is Fate. Which tho they be sondry, yet they depend one of an other. (HC1500–1710)

These two adjectives also clearly illustrate the further grammaticalization

from postdeterminer, e.g. (22) and (23), to quantifier, e.g. (24) and (25). The quantifier use was available for both adjectives from the earliest data onwards. At first, it was proportionally less frequent than the postdeterminer use. But as the figures for several show, it further increased and has become predominant in the present day. (22) The Psalmist very elegantly expresseth to us the several gradations by

which men at last come to this horrid degree of impiety; …Men are usually first corrupted by bad counsel and company, which is called “walking in the counsel of the ungodly”; next they habituate themselves to their vicious practices, which is … (HC1500–1710)

(23) This hous of Seynt Victour is in Paris, to whech I trowe þei longe. We haue in our libraries many sundry bookes þat to chanones of þat hous mad; on of hem hite Hewe, þe oþir hite Richard, notabel clerkis þei wer and men of holy lyf. (HC1250–1500)

(24) Carps and Loaches are observed to Breed several months in one year, which Pikes and most other fish do not. And this is partly proved by tame and wild “Rabbets”, as also by some “Ducks”, which will lay eggs nine of the twelve months, and yet there be other “Ducks” that lay no longer than about one month. (HC1500–1710)




281

(25) And besides, it is very unlike, that I of all Men woulde confederate in such a matter against the Lieutenant of the “Tower”, whose Daughter my Brother hath married, and his House and mine alyed together by Mariage sundry times within these few Yeres. (HC1500–1710)

The adjective several not only displays the highest proportion of quantifier

uses, it also shows the widest array of different quantifier uses. Similar to prototypical absolute quantifiers such as many and the cardinal numbers, several can be used in contexts where its scope is restricted to the following quantifier, as illustrated in (26).15 It has been used in this way from the earliest data onwards, but throughout the data this particular use has become more and more frequent. In the earliest examples (1500–1780), the following numeral had to be plural, but from 1780 onwards the following numeral is, as in the same combination with ordinary absolute quantifiers, singular rather than plural, as illustrated in (27). (26) In ye afternoone there came severall hundreds of his people: to ye meetinge where I was. (HC1500–1710) (27) We have provided accommodation now for several thousand of the most

helplessly broken-down men in London, criminals many of them, mendicants, tramps, those who are among the filth and offscouring of all things; but such is the influence that is established by the meeting and the moral ascendancy of our officers themselves, that we have never had a fight on the premises, and very seldom do we ever hear an oath or an obscene word. (CLMET1850–1920)

Secondly, from 1710 onwards, the corpus contains a few examples in

which several is used in combination with definite identification. The examples in the CLMET data all instantiate one specific combination, viz. the next several N, as illustrated by (28). In the COBUILD data there is no longer such a restriction, as shown by (29). However, the data combining several with definite identification seem to be subject to a stylistic restriction. The distribution of the pattern in the COBUILD corpus reveals that the combination occurs predominantly in written texts and higher registers.

15 In Breban (2002/2003), a similar use is discussed for the Dutch adjective verschillend which similar to several has grammaticalized furthest of all adjectives in the semantic field.



Tine Breban

282

(28) It is interesting thus to find the once domesticated cattle breaking into three colours, of which some one colour would in all probability ultimately prevail over the others, if the herd were left undisturbed for the next several centuries. (CLMET1780–1850)

(29) In the several years immediately after a divorce, children become more defiant, more negative, more aggressive or depressed or angry. If they are of school age, their school performance typically drops for at least a while. (CB)

Finally, several has not only a prototypical absolute use, but it can also be

used as a relative quantifier in certain contexts. As proposed by Milsark (1977), all absolute quantifiers can in fact be used relatively, i.e. implying the existence of a larger set of instances from which the quantified set is drawn. From the oldest data onwards, several could express relative quantification when complemented by a genitive or an of-phrase specifying the full set of instances that several takes a portion of, e.g. (30). In the later data, there are also a small number of examples, such as (31), in which several seems to express relative quantification without the explicit expression of the full set. (30) And this “Phaenomenon”, as I have elsewhere more largely shewn,

proceeds from a propriety which belongs to all kinds of fluid Bodies more or less, and is caused by the Incongruity of the Ambient and included Fluid, which so acts and modulates each other, that they acquire, as neer as is possible, a “sperical” or “globular” form, which propriety and several of the “Phaenomena” that proceed from it, I have more fully explicated in the sixth Observation. (HC1500–1710)

(31) … the torch-bearers moved quickly on. “We are nearing the sea,” said, in a calm voice, the person at their head. “Liberty and wealth to each slave who survives this day! Courage! I tell you that the gods themselves have assured me of deliverance. On!” Redly and steadily the torches flashed full on the eyes of Glaucus and Ione, who lay trembling and exhausted on his bosom. Several slaves were bearing, by the light, panniers and coffers, heavily laden … (CLMET1780–1850)

Milsark (1977) suggests that in such examples as (31) the relative reading is triggered by placing additional stress on the quantifier.16

16 It has to be remarked that as the corpus data do not contain information on stress, the presence and importance of this factor has to be taken as an intuitive judgment only.




283

Finally, it has to be noted that the quantitative picture presented in the data for sundry is less neat than that of several. This is due to the decrease in overall use of sundry which is gradually disappearing from the active vocabulary and which is largely restricted in Present-day English to the expression all and sundry. From a qualitative angle, the quantifier sundry is very similar in its different possibilities to several. Although it cannot occur in definite NPs, it was used in the construction sundry of as a relative quantifier and in the restricted scope construction with a numeral in the older data. But because of its generally decreasing use, these special uses are no longer found in the Modern and Present-day English data. 3.2. Various, different and distinct The adjectives that will be discussed in this section show a lesser degree of grammaticalization. In contrast to several and sundry, they can still be used in their lexical sense as attribute and predicative adjective in Present-day English, although they have grammaticalized to a considerable extent. I will start the discussion with various and then move on to different and distinct.

VARIOUS size of sample

pred attr attr-pd pd pd-quant quant

distrib sing pd

phoric pd

1250–1500 0

0 0 0 0 0 0 0 0

1500–1710 1 100%

0 1 100%

0 0 0 0 0 0

1710–1780 100 100%

9 9%

20 20%

13 13%

45 45%

8 8%

5 5%

0 0

1780–1850 100 100%

4 4%

13 13%

8 8%

42 42%

19 19%

12 12%

0 2 2%

1850–1920 100 100%

7 7%

11 11%

5 5%

54 54%

11 11%

12 12%

0 0

1990– 200 100%

1 0.5%

4 2%

1 0.5%

108 54%

28 14%

58 29%

0 0

Table 3. Diachronic overview of the different uses of various

As is clear from Table 3, the development of various shows a gradual decrease of lexical uses in favour of grammaticalized uses. The only example of various in the Helsinki Corpus is an attribute use dating from the period 1640–1710, reproduced as (32); but the CLMET data contain a considerable amount of grammaticalized, especially postdeterminer, uses, throughout the different periods, e.g. (33).



Tine Breban

284

(32) My aim being to discover the old Art of teaching Schoole, and how it may be improved in every part suteable to the years and capacities of such children as are now commonly taught; I shall first begin my discourse concerning a petty-Schoole, & here or else where I shall not busie my self or Reader about what a childe of an extraordinary towardliness, and having a teacher at home, may attain unto, and in how short a space, but onely shew how a multitude of various wits may be taught all together with abundance of profit and delight to every one, which is the proper and main work of our ordinary Schooles. (HC1500–1710)

(33) Claude Lorraine, on the contrary, was convinced that taking nature as he found it seldom produced beauty. His pictures are a composition of the various draughts which he has previously made from various beautiful scenes and prospects. (CLMET1710–1780)

The postdeterminer uses increase a little over time to cover more than half of the instances of various for the period after 1850, filling the vacuum left by several and sundry, which had shifted away from postdeterminer to quantifier function at the end of the 18th and the first half of the 19th century. Like several and sundry, various has also developed the more grammaticalized quantifier use, illustrated in (34), which is attested from around 1700 onwards. This quantifier use slowly increases in the CLMET data, but only really breaks through in the COBUILD data where it covers one third of the whole sample. (34) In men, we various ruling passions find; In women, two almost divide the

kind: Those, only fixed they first or last obey – The love of pleasure, and the love of sway. (CLMET1710–1780)

With different and distinct, the situation is less clear-cut.




285

DIFFERENT size of

sample pred attr attr-

pd pd pd-

quant quant distrib

sing pd phoric pd

1250–1500 0

0 0 0 0 0 0 0 0

1500–1710 13 100%

5 38.4%

3 23.1%

0 2 15.4%

0 0 1 7.7%

2 15.4%

1710–1780 100 100%

12 12%

18 18%

9 9%

51 51%

1 1%

0 2 2%

7 7%

1780–1850 100 100%

29 29%

17 17%

12 12%

29 29%

0 1 1%

2 2%

10 10%

1850–1920 100 100%

33 33%

25 25%

2 2%

27 27%

1 1%

0 2 2%

10 10%

1990– 200 100%

57 28.5%

47 23.5%

12 6%

62 31%

5 2.5%

2 1%

1 0.5%

14 7%

Table 4. Diachronic overview of the different uses of different

DISTINCT size of sample

pred attr attr-pd pd pd-quant

quant distrib sing pd

phoric pd

1250–1500 1 100%

1 100%

0 0 0 0 0 0 0

1500–1710 9 100%

7 77.8%

2 22.2%

0 0 0 0 0 0

1710–1780 7517 100%

23 30.7%

21 28%

3 4%

15 20%

0 0 10 13.3%

3 4%

1780–1850 63 100%

31 49.2%

16 25.4%

2 3.2%

10 15.9%

0 0 0 4 6.3%

1850–1920 51 100%

10 19.6%

11 21.6%

6 11.8%

18 35.3%

0 0 2 3.9%

4 7.8%

1990– 119 100%

31 26%

4118 34.5%

7 5.9%

31 26%

0 0 4 3.4%

5 4.2%

Table 5. Diachronic overview of the different uses of distinct Both adjectives start off from a stage with lexical uses only. For distinct this is apparent from the figures in Table 5. For different, by contrast, the first period in the Helsinki Corpus for which data are available (1570–1640) contains one predicative

17 Examples that consist of distinct in its new lexical meaning ‘clear’, which are found in the corpus from 1710 onwards (consisting of 22, 35 and 46 examples for the periods 1710–1780, 1780–1850, 1850–1920 respectively and 67 examples for the Present-day English data), are left out of consideration so as not to distort the quantitative data on the specific process of grammaticalization discussed in the present article. 18 The large number of attribute examples in the COBUILD data is due to the many examples dealing with the Quebec situation in Canada; e.g. the NP a distinct society makes up 16 out of the 41 attribute examples.



Tine Breban

286

example as well as one postdeterminer example. However, as discussed in Breban (2006a), all the examples dating from before 1570 quoted in the OED and the MED are lexical (both predicative and attribute) uses, which suggests that different did in fact go through a first phase with lexical uses only. The following examples illustrate the early lexical uses of distinct and different. (35) The sayde Mayster Harry right shortely weywardly and angerly answered

seyyng that the sayde Mayer and Citeseyns shold right noght have a do with the sayde Bysshoppis tenantis with yn the cite and suburb of the same, ne the sayde tenantis with the sayde Mayer and Citeseyns, but only be ceparat and distyncte awey fro the sayde cite, … (HC1250–1500)

(36) Who when he lookes out of the glasse of his hye prouidence, knowith what for ech man is best. And that he knowes is best, that he gyues him. And this is the greate miracle of destenyes order, when it is treated by a skyllfull person, at which the ignorant woonder. And that I may somwhat touche what mans Reason may comprehend of Godes depth, in that mater that thou supposest to be most just, & keeps greatest equalitie, it seemes all be different from him that knowith what Prouidence is. (HC1500–1710)

(37) c1449 Pecock Repr. 438: Petir … was heed in a dyuers and different maner fro ech other Apostle. (MED s.v. different, adj.)

The data for the later periods do not show a continuation of the grammaticalization pattern: rather than a decrease in lexical uses and an increase in postdeterminer uses, the data, in particular for the period 1780–1850, seem to suggest strengthening of the lexical use at the expense of the grammaticalized uses. Several factors may have played a role in this development. For example, the rise of a new lexical meaning for distinct, viz. ‘clear’, illustrated in (38), from 1710 onwards, may have caused stronger entrenchment of the lexical (predicative) use of the adjective. (38) But one of the prisoners’ wives, who was present, and came in with her,

gave us a more distinct account: she informed us that as my wife, my daughter, and herself, were taking a walk together on the great road a little way out of the village, a post-chaise and pair drove up to them and instantly stopt. (CLMET1710–1780)

Furthermore, both adjectives are in competition with various, which is gaining ground as a postdeterminer and later as a quantifier, and which may have blocked the further grammaticalization of different and distinct.




287

Similarly, it was argued in Breban (2006a) that, at roughly the same time, the phoric use of different, illustrated in (39), in which the adjective conveys a relation of referential non-identity between the instance denoted by the NP, e.g. a different place, and a second instance of the same type that is available in the discourse context, e.g. where the men were standing, lost out to other. With the exception of a very restricted set of contexts (see below), other became the usual marker of this phoric relation of non-identity. (39) The women were standing off at a different place from where the men were

stationed, and when the firing began, those of the men who escaped the first onslaught went in one direction up the ravine, and then the women, who were bunched together at another place, went entirely in a different direction through an open field. (CB)

Another factor could be that there are no other adjectives available to “renew” the lexical semantics of qualitative difference, as was the case when sundry and several moved on to their grammaticalized uses and the Romance adjectives various, different and distinct were borrowed to provide these descriptive meanings.

All this seems to suggest that the grammaticalization of different and distinct was halted in the 1780–1850 data, in the sense that it reached an equilibrium that has been maintained up to the present day. This hypothesis is further confirmed by the virtual absence of the next stage of the grammaticalization development, the quantifier use, from the data. The data for distinct show no evidence of this development at all, and although there are a few examples allowing a quantifier reading of different, e.g. (40), it has remained rather infrequent in my corpus data for the later periods. (40) A criminal was branded, during my stay here, for the third offence; but

the relief he received made him declare that the judge was one of the best men in the world. I sent this wretch a trifle, at different times, to take with him into slavery. (CLMET1780–1850)

This equilibrium not only applies to the proportion of lexical versus

grammaticalized uses of the two adjectives, it also goes together with a stabilization of the distribution of the adjectives of difference over different functions and contexts. Several and the disappearing sundry mainly occupy the function of quantifier, while various was up until 1920 the most frequent postdeterminer. In the COBUILD data, various is starting to establish itself as



Tine Breban

288

a quantifier, occurring especially in bridging contexts allowing either a postdeterminer or a quantifier interpretation, while several is developing more and more characteristics typical of absolute quantifiers and sundry is dropping out of use altogether. Distinct and different are still mainly used with a fully lexical value: for distinct this includes both the original lexical sense of qualitative difference and the new one that can be paraphrased as ‘clear’. However, distinct and different are also commonly used as postdeterminers, especially in the following contexts. Distinct is often found in NPs containing a cardinal quantifier, as in (41), whereas different often expresses distributivity, e.g. (42), and is rather frequent in definite NPs, e.g. (43). Throughout history, these three contexts were successively occupied by several and various. In the Present-day English data, different also often comes to be used in combination with a cardinal quantifier, as in (44), which is in turn at the expense of the postdeterminer uses of distinct. (41) McCauley said the new shuttle radar pictures supported a historical

picture that he and several colleagues had been developing for more than a decade. He said during the past 40 million years, three distinct systems had etched themselves into north Africa. (CB)

(42) “So I’ll say what I think but I’d really have a bit more confidence if I’d got my friends because I just feel better with my friends.” “Your friends aren’t always going to be there. You might go off to different countries and you’ve got to think by yourself when you get to our age …” (CB)

(43) The members of the executive or administration of a business or organization: They will not necessarily be the owners of the business, but will be selected by the owners to be responsible for the different functions of the organization. (CB)

(44) The cards, from Futera, will put you on the wave of popularity rugby union is riding after the excitement of the World and Bledisloe Cups. There are 110 different cards in the series featuring past and present rugby greats including Tim Horan, Jason Little and Campese Ken Catchpole, Roger Gould and Andrew Slack. (CB)

In addition to these grammaticalized uses in the formal environment of NP-internal comparison, both adjectives also have minor uses as phoric postdeterminer in NPs with external comparison and as distributive postdeterminer in singular NPs with internal comparison (see Section 3.4).




289

3.3. Divers/diverse Although the adjective divers(e) has been going through the same general process of grammaticalization as the other adjectives, the process is far from straightforward in the data due to the relation between the two spelling variants divers and diverse. Summarizing from the entries for divers and diverse in the OED, the distribution of the two variants is as follows. Before 1700, the two were interchangeable and could be used lexically as well as in the grammaticalized postdeterminer and quantifier use. From 1700 onwards, the variant diverse was mainly restricted to lexical uses, while divers was limited to the grammaticalized uses. The corpus data investigated here, however, show that the relation between the two forms is more complicated than this.

DIVERS size of sample



phoric pd

1250–1500 42 100%

2 4.8%

3 7.1%

0 22 52.4%

4 9.5%

10 23.8%

0 1 2.4%

1500–1710 83 100%

9 10.8%

1 1.2%

2 2.4%

18 21.7%

18 21.7%

30 36.2%

1 1.2%

4 4.8%

1710–1780 17 100%

0 0 0 4 23.5%

3 17.7%

10 58.8%

0 0

1780–1850 20 100%

0 0 0 6 30%

4 20%

10 50%

0 0

1850–1920 5 100%

0 0 0 3 60%

1 20%

1 20%

0 0

1990– 3 100%

0 2 67.7%

0 0 0 1 33.3%

0 0

Table 6. Diachronic overview of the different uses of divers

DIVERSE size of sample



phoric pd

1250–1500 58 100%

4 6.9%

4 6.9%

3 5.2%

27 46.5%

12 20.7%

8 13.8%

0 0

1500–1710 17 100%

1 5.9%

0 0 3 17.6%

2 11.8%

11 64.7%

0 0

1710–1780 0

0 0 0 0 0 0 0 0

1780–1850 3 100%

0 3 100%

0 0 0 0 0 0

1850–1920 21 100%

8 38.1%

10 47.6%

1 4.8%

2 9.5%

0 0 0 0

1990– 200 100%

57 28.5%

102 51%

10 5%

28 14%

3 1.5%

0 0 0

Table 7. Diachronic overview of the different uses of diverse



Tine Breban

290

In the Helsinki data, the two indeed behave rather similarly, but even here there are a number of differences. Overall the data indicate, as shown in Tables 6 and 7, loss of the lexical uses, illustrated in (45) and (46), and an increase in grammaticalized uses, more specifically, a further development from predominantly postdeterminer, (47) and (48), to quantifier use, (49) and (50). (45) The .vij. day ys fortunat to begynne alle werkys vp-on; that persone that ys

born that day schuld be dysposyd to be sotel off wytt and dyuerse off condycionnys and chongabyl, and dysposyd to lyfe longe;… (HC1250–1500)

(46) The nombre of þe intestynes is 6. And þogh þay be alle contynue, neuerþelatter þei haue dyuers schappes and offices, by þe whiche þai beeþ departed, þat is to say, 3 smale and as many grete, whos cathologe is put of Galien, “De Vtilitate”, capitulo. (HC1250–1500)

(47) And God woot that in alle these langages and in many moo han these conclusions ben suffisantly lerned and taught, and yit be diverse reules; right as diverse pathes leden diverse folk the righte way to Rome. (HC1250–1500)

(48) In the mene tyme many strange and woundyrfulle bylle were sete in dyvers placys, sum at the kyngys owne chambyr doore at Westemyster, in hys palysse, and sum at the halle dore at Westemyster, ande sum at Poulys chyrche dore, and in many othyr dyvers placys of London. (HC1250–1500)

(49) Also, Madam, yff it lyke you, I have bene with my brodyr, John Betson, ffor money, and be my trouth I can none have off hym, he hath shewid me that my Mayster your husbaund and ye owe hym ffor dyverse wynes, Summa xij. li. vj. s. viij. d., as by the parcells herin closed more clerelyar it appereth, the which parcelles my brodyr saith that þei be trew. (HC1250–1500)

(50) Apon this mene he stiked faste, and thoghte hit was resonable and ever asked of me divers tymes what y wolde seye therto, all as y conceve to tempte me, and to concente to a mene, &c. … (HC1250–1500)

However, this path is restricted to the Helsinki data for diverse, which even disappears from the data in the period 1710–1780. The development is more gradual for divers and continues up until 1850.

From 1850 onwards, divers starts to disappear from the language, whereas diverse resurfaces in the period 1780–1850 and is then, indeed, restricted to lexical uses only, e.g. (51).




291

(51) With one word on his style as a prose-writer this short paper must close. The most diverse opinions have been uttered on the subject. The critics trip up each other with charming independency. (CLMET1780–1850)

From 1850 onwards, diverse seems to go down the path of grammaticaliz-ation again and ends up expressing the grammaticalized postdeterminer meaning, illustrated in (52), in over 15% of the COBUILD data, though it has not (yet) developed a clear quantifier use. (52) The story begins immediately after the First World War when in nineteen-

nineteen the council decided to build a new civic centre which would bring together all the diverse arms of the administration in a single area. (CB)

As the figures in Table 6 show, the variant divers has basically disappeared in the COBUILD data. 3.4. Other grammaticalized uses of the six adjectives The previous sections described the main process of grammaticalization manifested by the six adjectives investigated in this article, which takes place in plural NPs expressing internal comparison. But, as indicated in the tables, the adjectives also went through parallel processes of grammaticalization in two other, less frequent, formal environments, which I will very briefly turn to in this section. 3.4.1. Grammaticalization in singular NPs expressing internal comparison At a certain stage in their history, all adjectives except various have developed a non-phoric grammaticalized postdeterminer use in a singular NP. This use is, however, always limited to a small portion of the data. From an evolutionary point of view, the presence of this singular postdeterminer use coincides with the stage in which the grammaticalized postdeterminer is the main use of the adjective in plural NPs, and it disappears as soon as the grammaticalized quantifier use takes over.

For several, sundry and divers, for example, the only examples of this singular postdeterminer use, illustrated by (53), (54) and (55) respectively,



Tine Breban

292

are found in the period 1500–1710, when the quantifier use had not yet become the main one.19 (53) For to euery seuerall place, yea to euerye little moment of the earth in an

oblique Spheare, belongeth his proper Horizon and seuerall altitude of the Pole, whereby it appeareth that the Horizons are infinite and without number. (HC1500–1710)

(54) The creation of all thinges, & the disposing of mutable Natures, & what euer by any meane is mooued, getes the cause, order, & forme of Godes mynde, stabilitie. And this sett in the top of her Purenes, appoyntes a sondry manner for ech action: which order, when it is beheld in the very cleerenes of diuine vnderstanding, is named “Prouidence”. (HC1500–1710)

(55) Dost thou thinke that the father of al things hath taken from any outwarde thinge the same soueraygne good, whereof it is sayde he is ful, or els thynkest thou that he hath it naturallye of hymselfe? As thoughe thou shouldyst thinke that God hymselfe and the blessednes of God be of dyuers distyncte substaunce and not vnite all in one or of one onely substaunce? (HC1500–1710)

Similarly, the two adjectives that have never developed a real quantifier use, different and distinct, can still be used in this singular context in Present-day English, as shown in (56) and (57). (56) RIDE FROM HOTEL TO HOTEL: Each night is a different hotel, different

village. (CB) (57) The forces of production refer to the organizational and technological

means applied to the process of production and, as such, they serve both to characterize the process of production under any specific mode of

19 Table 1 indicates two examples of this type of use for sundry in the COBUILD data, but these do not fit in with the regular pattern described in this section. The two examples referred to, one of which is included here as illustration, do not have a distributive meaning. Rather they express simple plurality but with a singular uncount noun as head instead of the normal plural count noun. (ii) I replaced the book and stood in the middle of the room, gazing at the

sundry video equipment, the computers and printers, the endless file cabinets, and the ubiquitous testimonies to Wald himself. The room seemed to ring with his presence. (CB)




293

production (MOP) and to create the conditions for change from one MOP to another. … In other words, the Marxist vision of history was informed by a succession of MOPs in time, each of which constituted a distinct historical stage. (CB)

Interestingly, this particular postdeterminer use seems to have been a very

productive pattern for distinct (taking up 40% of the non-phoric postdeterminer data) in the period 1710–1780, when the grammaticalization of the adjective had not yet been halted. (58) shows an example from this period. (58) Matter, therefore, is not a mode but a substance, and each part of matter

is not a distinct mode, but a distinct substance. (CLMET1500–1710) In the subsequent periods the grammaticalization of distinct came to a halt and the singular distributive pattern became much less frequent and even disappeared from the data in the period 1780–1850. In Present-day English, the pattern remains available for some of the adjectives, viz. distinct and different, but is clearly not very frequent. 3.4.2. Grammaticalization in NPs expressing external comparison The second minor grammaticalization pattern is the development of a phoric postdeterminer use, similar to other. This use originates in NPs expressing external rather than internal comparison. Again the availability of this minor postdeterminer use coincides with the phase in which the quantifier use of the adjective in a plural NP has not yet become established as the predominant one.

For sundry and divers, a few phoric postdeterminer uses are attested before 1710, e.g. (59) and (60), and the 1500–1710 data for several contain one potential (cata)phoric postdeterminer example, viz. (61). (59) For as muche as the almykanteras in thin Astrelabie ben compowned by

two and two, where as somme almykanteras in sondry astrelabies be compowned by 1 and 1, or elles by 2 and 2, it is necessarie to thy lernyng to teche the first to knowe and worke with thin oune instrument. (HC1250–1500)



Tine Breban

294

(60) But to leaue this new Leach craft, with thier doting inuentions, I wil here

speake of diuers remedies very respectiue & appropriate, for the cure of the before named infirmity, as a president and example for young practizers of Chirurgery to follow. (HC1500–1710)

(61) The word “Mogoll”, in their language is as much as to say, the great white King; for he is a white man and of the Race of the Tartares. He is King of many Kingdomes, and writeth himselfe in his stile, “Patteshaw Shelham Shogh”, that is, the King of all the great coynes. For there is a seuerall coyne at “Lahore”, another at “Bramport”, another at “Surrot”, another at “Cambaia”, another at “Sabbarton”, and another at “Awgru” … (HC1500–1710)

Similarly, the corpus included a few examples of phoric grammaticalized

various dating from 1780–1850, i.e. before the main breakthrough of the quantifier use attested in the COBUILD data. One of these is reproduced here. (62) Some persons, I know, estimate happiness by fine houses, gardens, and

parks; others by pictures, horses, money, and various things wholly remote from their own species; but when I wish to ascertain the real felicity of any rational man, I always inquire WHOM HE HAS TO LOVE. (CLMET1780–1850)

For distinct and different, finally, this postdeterminer use is again still

present in the Present-day English data, as (63) and (64) illustrate, but it never really gained a strong foothold, probably because its development was blocked by the predominance of other, which is restricted to this grammaticalization pattern. (63) The Indian economist S. Lall, in a 1975 paper, took the idea of

dependence to mean underdevelopment, no more and no less, and from that angle went on to argue that since some highly industrialized economies, by definition non-dependent, also were dependent in crucial ways, then either they were not developed or dependence said and explained nothing about underdevelopment except “to pick off some salient features of modern capitalism as it affects some less developed countries and put them into a distinct category of dependence”. (CB)




295

(64) You know when you go to other countries they’re proud of their local accent and things. It’s surprising though just what bits and bobs y just you hear around about you know and er like you say older people especially have a lot more sort of “twang” words as we call ‘em that you can er go back to you know. And if you don’t understand it well, can you think of anything say in the house that you would have a different name for? (CB)

3.5. Discussion: grammaticalization and subjectification in the field of difference In Section 3, I have confronted the grammaticalization hypothesis from lexical attribute to individualizing postdeterminer to quantifier with historical data material of six adjectives of difference that have one or both of these grammaticalized uses in current English, viz. different, distinct, divers(e), several, sundry, and various. The main conclusion that can be drawn from this investigation is that the data confirm the suggested grammaticalization and subjectification path. Firstly, the general grammaticalization hypothesis is supported by the data in two ways. On the one hand, there are adjectives such as several and sundry which display the loss of lexical uses, i.e. their data contain a first stage with both lexical and grammaticalized uses and a later one with grammaticalized uses only. On the other hand, the data for various, distinct and different show the acquisition of grammaticalized uses in that they consist of early data with lexical uses only and a subsequent stage in which both lexical and grammaticalized uses are present. Secondly, this diachronic study lends support to the specific two-step grammaticalization path from attribute over individualizing postdeterminer to quantifier. Concretely, the data confirm that the quantifier use constitutes a later development. The data for several, sundry and various (as well as the early data for divers(e)) manifest a decrease of postdeterminer uses and a corresponding increase of quantifier uses. However, the historical evolution of distinct and different makes clear that the development of adjectives of difference along this path of grammaticalization is not the straightforward completion of the subsequent grammaticalization processes. Even though their data show the onsets of the grammaticalization path, viz. the development of the individualizing postdeterminer use, they also reveal that the acquisition of this grammaticalized use did not cause the adjectives to lose their lexical semantics nor did it result in the further grammatical-ization of these adjectives into quantifiers.



Tine Breban

296

Now that all different uses of the adjectives several, sundry, various, different, distinct and divers(e) have been looked at from both a synchronic and a diachronic perspective, I will situate their developments in the context of other diachronic semantic developments within the field of difference. For the six adjectives in the present corpus, their main grammaticalization path involves the shift from lexical attribute to postdeterminer marking plurality to quantifier. The other two paths of grammaticalization and subjectification, viz. the development of a distributive postdeterminer use in singular NPs with internal comparison and of the phoric postdeterminer use in NPs with external comparison, are represented by only few examples. But, as shown in previous work (Breban 2002/2003, 2006a; Breban & Davidse 2003), the phoric use is usually realized by other adjectives belonging to or relating to the field of difference.

The adjective that is the most frequent marker of phoric non-identity is other, which in Present-day English has fully grammaticalized from a qualitative lexical adjective with attribute and predicative uses into a phoric postdeterminer (and classifier) establishing phoric relations of non-identity in the discourse. Other adjectives that function as phoric postdeterminers similar to other are adjectives of order (e.g. additional, further and next), adjectives of similarity (e.g. similar and comparable) and even, in special contexts, adjectives of identity in combination with an indefinite determiner (e.g. (an) identical in English, eenzelfde in Dutch).

The field of difference is thus characterized by two major paths of grammaticalization/subjectification, the development of phoric markers in NPs with external comparison and of plurality markers in NPs with internal comparison. As the previous discussion implies, the two patterns tend to be mutually exclusive in the sense that adjectives either specialize in one or the other direction. Other for example is limited to the phoric pattern, while several is restricted to the second path. The other adjectives, different, distinct, divers(e), sundry and various, had/have phoric as well as individualizing postdeterminer uses, but they clearly show a skewed distribution in which the individualizing uses largely outnumber the phoric ones. The overall make-up of the Present-day English field of difference can be represented in the following way.




297

different/ distinct

additional/further/next similar/comparable (an) identical Figure 1. The field of difference in Present-day English

The historical evolution leading to this current distribution seems to be determined by two processes: RENEWAL, i.e. the process whereby existing meanings take on new forms (Hopper & Traugott 2003[1993]: 122, based on Meillet’s 1915–1916 concept of renouvellement), and COMPETITION. The first concept, renewal, can be held responsible for the disappearance of sundry and diverse, which had completed the two-step grammaticalization and subjectification process to the quantifier use when several reached the second stage and replaced them as the most common quantifier. As several became especially associated with the quantifier function, the individualizing postdeterminer function which it fulfilled earlier became the main use of various. The most recent similar development seems to be that various is becoming increasingly used as a quantifier, which leaves the individualizing postdeterminer use to be expressed by the adjectives that started grammaticalizing the latest, different and distinct.

The second factor, competition between the different adjectives, is observable in the striving towards complementary distribution of the different adjectives within the field of difference. This tendency can explain what happened to the development of different and distinct from the end of the eighteenth century onwards. The grammaticalization process of these adjectives was stunted by other adjectives of difference that already fulfilled

lexical attribute + predicate

phoric postdeterminer other

diverse postdeterminer

marking plurality/distributivity

in plural NP

in sg NP

various/[divers]

quantifier several/[sundry]



Tine Breban

298

the more grammaticalized functions. Concretely, they appear to have been stopped from further grammaticalizing into quantifiers by various and several and from fully developing as frequent phoric markers by other.

The interaction of these two processes results in an overall development that appears to consist of a succession of “semi-stable situations” (cf. Brems 2003 on the different degrees of grammaticalization of the size noun constructions heap(s) of, pile(s) of and bunch(es) of) in which the distribution of the adjectives over the different uses appears to have reached an equilibrium, i.e. in which competition is neutralized. But these situations are constantly being pressured by the natural forces of grammaticalization and the tendency towards renewal.

4. Testing the leftward movement hypothesis

In Sections 2 and 3, I have both synchronically and diachronically developed the claim that the adjectives of difference have been going through a double process of grammaticalization and subjectification from a lexical attribute use to the grammaticalized postdeterminer use and in a second stage from the postdeterminer to a quantifier use. As briefly noted in Section 2, these three functions have been associated with three typical positions in the English NP (see Halliday 1994[1985]; Bache 2000). Attributes occupy the central prenominal position in the NP in between postdeterminers and classifiers (Halliday 1994[1985]: 184–185). Postdeterminers are found in “post”-determiner position preceding any attributive adjectives. As quantifiers the adjectives occupy the first position of the NP when reference is indefinite and the position following the definite article when the NP has definite identification. As a result, the suggested grammaticalization-cum-subjectification path can be expected to be structurally reflected in a double move to the left, viz. from the central attribute position in the prenominal zone to postdeterminer position and, when the adjectives develop the quantifier meaning, from postdeterminer position to determiner position, and in this way to provide a further illustration of Adamson’s (2000) leftward movement hypothesis. As the following discussion will show, the different positions are confirmed in the diachronic data for especially sundry and several.

Firstly, the leftward movement hypothesis stipulates that in current English these adjectives, several and sundry, which have grammaticalized




299

completely, can no longer occur in the central (attribute) position, but have to occur to the left of it. Concretely, in the COBUILD data this is reflected in the absence of (1) examples in which these adjectives occupy a position to the right of other attributes and (2) examples in which they are coordinated with another attribute. Examples (65) and (66) convey that these two patterns were available to several and sundry in older stages of English, when they could still be used as attributes. (65) 1596 Spenser State Irel. Wks. (Globe) 681/2: All men should marke theyr

cattell with an open severall marke upon theyr flanckes. (OED s.v. several, a., adv., and n. A.I.1.c, obs.)

(66) 1532 (?a1400) RRose (Thynne) 1437: Ful gaye was al the grounde ... With many a fresshe and sondrie floure. (MED s.v. sondri, adj. 1.b)

Secondly, notwithstanding the fact that neither of the adjectives has fully

shifted to quantifier use only, the shift from postdeterminer to quantifier is in the same way apparent from the absence of certain orderings that were available in older stages of the language. More specifically, there are no examples in the COBUILD corpus in which several or sundry occupy the position following postdeterminer other or in which they are found to the right of another quantifier. As examples (67)–(69) illustrate, both patterns are attested in the available diachronic material. (67) c1470 Henry Wallace I.29: Elrisle..Auchinbothe, and othir syndry place.

(OED s.v. sundry, a. 3, obs.) (68) 1570–1576 Lambarde Peramb. Kent (1826) 198: The third Brooke ... being

crossed in the way by seven other sundry bridges. (OED s.v. sundry, a. 3.b, obs.)

(69) When we came to the Town, two several Churches strove to receive us; but having some Acquaintance with the Father of the one, and not with the other, we excused ourselves to the latter, and took up with our Friend. (HC1640–1710)

In the corpus data, the combination quantifier + several is present in the first two available samples, 1500–1710 and 1710–1780. When used as quantifier, the adjectives themselves fill the position that is in example (69) occupied by two. The combination other + several/sundry, illustrated with two examples from the OED, viz. (67) and (68) was not attested in the diachronic database.



Tine Breban

300

In sum, the data for several and sundry show the loss of their most rightward, central, position and a move to the determiner zone. In a second stage, the subjectification to quantifier use is mirrored in the loss of the adjectives’ ability to occur to the right of other postdeterminers such as other or other absolute quantifiers. Instead, they always occupy the leftmost position of the determiner string in indefinite NPs themselves.

5. Conclusion

I would like to conclude this article by drawing attention to two important points that were made in the article with respect to grammaticalization in general. Firstly, the article has provided additional evidence for Adamson’s (2000) hypothesis that the association between leftward-movement and grammaticalization can also be observed at the level of constituent structure in the NP using real language data. More specifically, I have argued that the association can be legitimized conceptually for those particular instances of grammaticalization in the NP that can more specifically be characterized as instances of subjectification as defined by Langacker. Because subjectification in this sense precisely refers to the development of grammatical uses related to quantification and grounding, it by extension involves a move to the instantiation end of the NP, i.e. the lefthand side. It can be expected that those elements exhibiting subjectification in other types of phrases such as the verb phrase display a similar leftward movement in English. However, future studies will have to be conducted to substantiate this hypothesis.

A second issue that merits further research is the need to look at grammaticalization/subjectification within specific formal contexts or constructions. The present discussion has shown that in different constructions a similar semantic process can lead to different meanings, which open different possibilities for further development. This diversification can trigger specialization of different semantically similar adjectives for one of these meanings, hence taking different paths from their original synonyms. The present article provides of course only one specific instance of the interaction between semantic grammaticalization/ subjectification and the formal processes that it goes together with, but I believe it marks the topic as a very interesting and fruitful one for further research.




301

References Adamson, Sylvia. 2000. A lovely little example: word order options and category shift

in the premodifying string. In Olga Fischer, Anette Rosenbach & Dieter Stein, eds. Pathways of change: grammaticalization in English. Amsterdam & Philadelphia: John Benjamins, 39–66.

Bache, Carl. 2000. Essentials of mastering English: a concise grammar. Berlin & New York: Mouton de Gruyter.

Beck, Sigrid. 2000. The semantics of different: comparison operator and relational adjective. Linguistics and Philosophy 23: 101–139.

Breban, Tine. 2002/2003. The grammaticalization of the adjectives of identity and difference in English and Dutch. Languages in Contrast 4: 167–201.

Breban, Tine. 2006a. The grammaticalization of the English adjectives of comparison: a diachronic case study. In Roberta Facchinetti & Matti Rissanen, eds. Corpus-based studies of diachronic English. Bern: Peter Lang, 253–288.

Breban, Tine. 2006b. Grammaticalization and subjectification of the English adjectives of general comparison. In Angeliki Athanasiadou, Costas Canakis & Bert Cornillie, eds. Subjectification: various paths to subjectivity. Berlin & New York: Mouton de Gruyter, 241–278.

Breban, Tine & Kristin Davidse. 2003. Adjectives of comparison: the grammaticalization of their attribute uses into postdeterminer and classifier uses. Folia Linguistica 37: 269–317.

Brems, Lieselotte. 2003. Measure noun constructions: an instance of semantically-driven grammaticalization. International Journal of Corpus Linguistics 8: 283–312.

Brems, Lieselotte. 2004. Measure noun constructions: degrees of grammaticalization and delexicalization. In Karin Aijmer & Bengt Altenberg, eds. Advances in corpus linguistics. Papers from the 23rd International Conference on English Language Research on Computerized Corpora (ICAME 22) Göteborg 22–26 May 2002. Amsterdam: Rodopi, 249–265.

Bybee, Joan L. 1985. Morphology: a study of the relation between meaning and form. Amsterdam & Philadelphia: John Benjamins.

Bybee, Joan L. 2003. Mechanisms of change in grammaticalization: the role of frequency. In Joseph & Janda, eds. 602–623.

Campbell, Lyle. 2001. What’s wrong with grammaticalization? Language Sciences 23: 113–161.

Carlson, Greg N. 1987. Same and different: some consequences for syntax and semantics. Linguistics and Philosophy 10: 531–565.

Croft, William. 2001. Radical Construction Grammar: syntactic theory in typological perspective. Oxford: Oxford University Press.



Tine Breban

302

Davidse, Kristin. 2004. The interaction of identification and quantification in English determiners. In Michel Achard & Suzanne Kemmer, eds. Language, culture and mind. Stanford, CA: CSLI Publications, 507–533.

Davidse, Kristin, Tine Breban & An Van Linden. 2008. The development of secondary deictic meanings by adjectives in the English NP. English Language and Linguistics 12.3.

Denison, David. 2002. History of the sort of construction family. Paper presented at ICCG2 (Second International Conference on Construction Grammar), Helsinki, September 6th–8th. [Available at www.llc.manchester.ac.uk/SubjectAreas/ LinguisticsEnglishLanguage/AcademicStaff/DavidDenison/]

Denison, David. 2006. Category change and gradience in the determiner system. In Ans van Kemenade & Bettelou Los, eds. The handbook of the history of English. Oxford: Blackwell, 279–304.

De Smedt, Liesbeth, Lieselotte Brems & Kristin Davidse. 2007. NP-internal functions and extended uses of the ‘type’ nouns kind, sort, and type: towards a comprehensive, corpus-based description. In Roberta Facchinetti, ed. Corpus linguistics 25 years on. Amsterdam: Rodopi, 227–257.

De Smet, Hendrik. 2005. A corpus of Late Modern English texts. ICAME Journal 29: 69–82.

Diewald, Gabriele. 2002. A model for relevant types of contexts in grammatical-ization. In Wischer & Diewald, eds. 103–120.

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

Evans, Nicholas & David Wilkins. 2000. In the mind’s ear: the semantic extensions of perception verbs in Australian languages. Language 76: 546–592.

Fanego, Teresa. 2004. On reanalysis and actualization in syntactic change: the rise and development of English verbal gerunds. Diachronica 21: 5–55.

Halliday, Michael A.K. 1994[1985]. An introduction to Functional Grammar. London: Edward Arnold.

Halliday, Michael A.K. & Ruqaiya Hasan. 1976. Cohesion in English. London: Longman.

Haspelmath, Martin. 1998. Does grammaticalization need reanalysis? Studies in Language 22: 315–351.

Heine, Bernd. 2002. On the role of context in grammaticalization. In Wischer & Diewald, eds. 83–101.

Heine, Bernd. 2003. Grammaticalization. In Joseph & Janda, eds. 575–601. Hetzron, Robert. 1978. On the relative order of adjectives. In Hansjakob Seiler, ed.

Language universals: papers from the Conference held at Gummersbach/Cologne, Germany, October 3–8, 1976. Tübingen: Gunter Narr, 165–184.




303

Himmelmann, Nikolaus P. 2004. Lexicalization and grammaticization: opposite or orthogonal? In Walter Bisang, Nikolaus P. Himmelmann & Björn Wiemer, eds. What makes grammaticalization? A look from its fringes and its components. Berlin & New York: Mouton de Gruyter, 21–42.

Hoffmann, Sebastian. 2004. Are low-frequency complex prepositions grammaticalized? In Hans Lindquist & Christian Mair, eds. Corpus approaches to grammaticalization in English. Amsterdam & Philadelphia: John Benjamins, 171–210.

Hoffmann, Sebastian. 2006. Grammaticalization and English complex prepositions: a corpus-based study. London & New York: Routledge.

Hopper, Paul J. 1991. On some principles of grammaticization. In Elizabeth Closs Traugott & Bernd Heine, eds. Approaches to grammaticalization. Volume I: Focus on theoretical and methodological issues. Amsterdam & Philadelphia: John Benjamins, 17–35.

Hopper, Paul J. & Elizabeth Closs Traugott. 2003[1993]. Grammaticalization. Cambridge: Cambridge University Press.

Joseph, Brian D. & Richard D. Janda, eds. The handbook of historical linguistics. Oxford: Blackwell.

Laca, Brenda & Liliane Tasmowski. 2001. Distributivité et interprétations dépendantes des expressions d’identité. In Georges Kleiber, Brenda Laca & Liliane Tasmowski, eds. Typologie des groupes nominaux. Rennes: Presses Universitaires de Rennes, 143–166.

Laca, Brenda & Liliane Tasmowski. 2003. From non-identity to plurality: French différent as an adjective and as a determiner. In Josep Quer, Jan Schroten, Mauro Scorretti, Petra Sleeman & Els Verheugd, eds. Romance languages and linguistic theory 2001. Selected papers from ‘Going Romance’, Amsterdam, 6–8 December 2001. Amsterdam & Philadelphia: John Benjamins, 155–176.

Langacker, Ronald W. 1990. Subjectification. Cognitive Linguistics 1: 5–38. Langacker, Ronald W. 1991. Foundations of Cognitive Grammar. Volume II:

Descriptive application. Stanford, CA: Stanford University Press. Langacker, Ronald W. 1998. On subjectification and grammaticization. In Jean-

Pierre Koenig, ed. Discourse and cognition: bridging the gap. Stanford, CA: CSLI Publications, 71–89.

Langacker, Ronald W. 1999. Losing control: grammaticization, subjectification, and transparency. In Andreas Blank & Peter Koch, eds. Historical semantics and cognition. Berlin & New York: Mouton de Gruyter, 147–175.

Langacker, Ronald W. 2002. Deixis and subjectivity. In Frank Brisard, ed. Grounding: the epistemic footing of deixis and reference. Berlin & New York: Mouton de Gruyter, 1–28.

Langacker, Ronald W. 2004. Remarks on nominal grounding. Functions of Language 11: 77–113.



Tine Breban

304

Lehmann, Christian. 1985. Grammaticalization: synchronic variation and diachronic change. Lingua e Stile 20: 303–318.

Lehmann, Christian. 2004. Theory and method in grammaticalization. In Gabriele Diewald, ed. Grammatikalisierung. [Special issue of Zeitschrift für Germanistische Linguistik 32.2: 152–187].

Lorenz, Gunter. 2002. Really worthwhile or not really significant?: a corpus-based approach to delexicalization and grammaticalization of intensifiers in Modern English. In Wischer & Diewald, eds. 143–161.

Lyons, Christopher. 1999. Definiteness. Cambridge: Cambridge University Press. MED = Kurath, Hans, Sherman M. Kuhn, John Reidy & Robert E. Lewis, eds. 1952–

2001. Middle English Dictionary. Ann Arbor, MI: University of Michigan Press. Meillet, Antoine. 1915–1916. Le renouvellement des conjunctions. Annuaire de

l’École Pratique des Hautes Études. [Reprinted in Antoine Meillet. 1958. Linguistique historique et linguistique générale. Paris: Champion, 159–174.]

Méndez-Naya, Belén. 2003. On intensifiers and grammaticalization: the case of swiþe. English Studies 84: 372–391.

Méndez-Naya, Belén. 2006. Adjunct, modifier, discourse marker: on the various functions of right in the history of English. Folia Linguistica Historica 27: 141–169.

Méndez-Naya, Belén, ed. 2008. English intensifiers. [Special issue of English Language and Linguistics 12.2: 213–394].

Milsark, Gary. 1977. Toward an explanation of certain peculiarities of the existential construction in English. Linguistic Analysis 3: 1–29.

Moltmann, Friederike. 1997. Reciprocals and same/different: towards a semantic analysis. Linguistics and Philosophy 15: 411–462.

Nevalainen, Terttu & Matti Rissanen. 2002. Fairly pretty or pretty fair? On the development and grammaticalization of English downtoners. Language Sciences 24: 359–380.

OED = Simpson, John A., ed. 1989. Oxford English Dictionary (2nd edition) on CD-ROM Version 3.1.1. Oxford: Oxford University Press.

Paradis, Carita. 2000. Reinforcing adjectives: A cognitive semantic perspective on grammaticalisation. In Ricardo Bermúdez-Otero, David Denison, Richard M. Hogg & C. B. McCully, eds. Generative theory and corpus linguistics. A dialogue from 10 ICEHL. Berlin & New York: Mouton de Gruyter, 233–258.

Quirk, Randolph, Sidney Greenbaum, Geoffrey Leech & Jan Svartvik. 1972. A grammar of Contemporary English. London: Longman.

Rijkhoff, Jan. 2002. The noun phrase. Oxford: Oxford University Press. Rissanen, Matti. 1999. On the adverbialization of RATHER: surfing for historical

data. In Hilde Hasselgård & Signe Oksefjell, eds. Out of corpora: studies in honour of Stig Johansson. Amsterdam & Atlanta, GA: Rodopi, 49–59.




305

Rissanen, Matti, Merja Kytö & Minna Palander-Collin, eds. 1993. Early English in the computer age: explorations through the Helsinki Corpus. Berlin & New York: Mouton de Gruyter.

Tabor, Whitney. 1994. The gradual development of degree modifier sort of and kind of: a corpus proximity model. In Katherine Beals, Gina Cooke, David Kathman, Sotaro Kita, Karl-Erik McCullough & David Testen, eds. Papers from the 29th Regional Meeting of the Chicago Linguistic Society. Vol. I: 451–465. Chicago: Chicago Linguistic Society.

Tovena, Lucia M. & Marleen Van Peteghem. 2002. Différent vs autre et l’opposition réciproque vs comparatif. Lingvisticae Investigationes 25: 149–170.

Tovena, Lucia M. & Marleen Van Peteghem. 2006. La place de différents dans le syntagme nominal et les déterminants. In Francis Corblin, Sylvie Ferrando & Lucien Kupferman, eds. Indéfini et prédication. Paris: Presses de l’Université Paris-Sorbonne, 449–460.

Traugott, Elizabeth Closs. 1982. From propositional to textual and expressive meanings: some semantic-pragmatic aspects of grammaticalization. In Winfred P. Lehmann & Yakov Malkiel, eds. Perspectives on historical linguistics. Amsterdam & Philadelphia: John Benjamins, 245–271.

Traugott, Elizabeth Closs. 1988. Pragmatic strengthening and grammaticalization. In Shelley Axmaker, Annie Jaiser & Helen Singmaster, eds. Proceedings of the 14th Annual Meeting of the Berkeley Linguistics Society. General Session and Parasession on Grammaticalization. Berkeley, CA: Berkeley Linguistics Society, 406–416.

Traugott, Elizabeth Closs. 1989. On the rise of epistemic meanings in English: an example of subjectification in semantic change. Language 65: 31–55.

Traugott, Elizabeth Closs. 1995. Subjectification in grammaticalisation. In Dieter Stein & Susan Wright, eds. Subjectivity and subjectivisation: linguistic perspectives. Cambridge: Cambridge University Press, 31–54.

Traugott, Elizabeth Closs. 2003a. From subjectification to intersubjectification. In Raymond Hickey, ed. Motives for language change. Cambridge: Cambridge University Press, 124–139.

Traugott, Elizabeth Closs. 2003b. Constructions in grammaticalization. In Joseph & Janda, eds. Oxford: Blackwell, 624–647.

Traugott, Elizabeth Closs. Forthcoming. From ideational to interpersonal: a reassessment. In Kristin Davidse, Lieven Vandelanotte & Hubert Cuyckens, eds. Subjectification, intersubjectification and grammaticalization. Berlin & New York: Mouton de Gruyter.

Vandelanotte, Lieven. 2002. Prenominal adjectives in English: structures and ordering. Folia Linguistica 36: 219–259.

Wischer, Ilse & Gabriele Diewald, eds. 2002. New reflections on grammaticalization. Amsterdam & Philadelphia: John Benjamins.



Tine Breban

306

Author’s address Faculty of Arts, Department of Linguistics Katholieke Universiteit Leuven Blijde-Inkomststraat 21 PO box 3308

received: 19 March 2007

B–3000 Leuven, Belgium e-mail: [email protected]

revision invited: 2 May 2007 revised version accepted: 23 September 2007



Documents

Grammaticalization, subjectification and leftward movement of English adjectives of difference in the noun phrase