Master’s thesis Research Master Linguistics University of Groningen

(1)

Master’s thesis

Research Master Linguistics

University of Groningen

Bekje-af, knettertjegek en dolletjesgelukkig.

The use and development of intensifying diminutive compounds in

Dutch within the framework of Construction Morphology

Author: Caroline Elisabeth Jane Morris Student number:

Address: Phone number:

E-mail address: cej.morris [at] gmail.com

First supervisor: Prof. Dr. M. Norde Second supervisor: Prof. Dr. J. Hoeksema

(2)

Abstract

Bekje-af, knettertjegek en dolletjesgelukkig.

The use and development of intensifying diminutive compounds in Dutch within the framework of Construction Morphology

By Caroline Morris

Dutch adjectives are frequently intensified by forming adjectival compounds (peperduur, bloedheet, stampvol, hoognodig, superblij). These compounds (discussed as ‘elative compounds’ in Hoeksema (2012)) are an integral part of the expressive lexicon of standard Dutch (Fletcher 1980, Hoeksema 2012). A more marginal and informal type of compound, and the topic of this thesis, is the intensifying diminutive compound, where a diminutive suffix is inserted between the intensifier and the adjective (stikjeheet ‘suffocate-DIM hot’, doodjeseng, ‘dead-DIM scary’, haartjescherp

‘hair-DIM sharp’, broodjenuchter ‘bread-DIM sober’). This corpus-based analysis shows

that eight of the diminutive compounds included have a high frequency (over 500 attestations). Out of the 696 compounds analysed, 311 are attested in the diminutive form. Many of these have a low frequency, suggesting a high productivity for this type of word formation. The first examples of intensifying diminutive compounds date from around 1900 and similar constructions are also found in German, Frisian, Swedish and English. The bulk of the Dutch data is from web fora and blogs, but some examples are from more formal contexts. Usage is not restricted to certain regions, gender or age groups, although younger generations (13-30 years) do prefer intensifiers that can be combine with many adjectives (super-, knettertje-) to adjective-specific ones.

In this thesis intensifying diminutive compounds are placed within the framework of Construction Morphology (Booij 2010), which allows for a schematic notation of pairings of form and meaning. The high type productivity leads to the suggestion of a general schema for all intensifying diminutive compounds (see (1)), which stands in a paradigmatic relation (see (2)) with the general schema for intensifying compounds given by Booij & Hüning (forthc.:23).

(1) [xDIMAi]Aj<-> [very high degree of SEMi]j

(2) < [x Ai]Aj<-> [high degree of SEMi]j ≈ [x DIMAi]Aj<-> [very high degree of SEMi]j>

(3)

1. Introduction

A very frequent way of intensifying the meaning of a Dutch adjective is to use an adjectival compound. The right constituent of such a compound is an adjective. The left constituent is an intensifier, which indicates a high degree of the property expressed by the adjective. In such compounds the intensifier is often a noun, as in peperduur ’pepper expensive, very expensive’ or bloedheet ‘blood-hot, very hot’. It can also belong to a different word class (verb: stampvol ‘stamp full’, adjective: hoognodig ‘high necessary’, preposition: doornat ‘through wet, prefix: superblij ‘super happy’) or be an element which no longer occurs independently (starnakelzat ‘very drunk’). These intensifying compounds (discussed as ‘elative compounds’ in Hoeksema (2012)) have been around in Dutch at least since the Middle Ages, are covered in grammars and are a part of the expressive lexicon of standard Dutch (Hoeksema 2012:122).

A relatively new development, which is not mentioned in grammars of Dutch, is the practice of inserting a diminutive ending after the left constituent. According to both Hoeksema (2012:123) and Reker (1996:44) these forms have existed since the 1990s. Hoeksema (2012:123) treats them as informal alternatives to the base forms without a diminutive, while Reker (1996:44) refers to them as intensifying the meaning of the intensifying compounds further, schematically shown in (1).

(1) zwart gitzwart gitjezwart

black jet black jet-DIMblack

black very black even blacker

(5)

(2) zwart gitjezwart gitzwart black jet-DIMblack jet black

black very black even blacker

Apart from the mentions it receives in Hoeksema and Reker, no note appears to have been made of intensifying diminutive compounds in linguistic publications, nor has any historical research been carried out. The first aim of this paper is therefore to describe the properties and development of these compounds, focusing on diachronic development and distributional differences between groups of language users on the basis of the criteria age, region, gender and occupation. Adopting a constructional approach to language change (Booij 2010), I will carry out a corpus-based analysis. As not all intensifying compounds are found in combination with a diminutive and frequencies vary strongly, I will research the grammatical properties of the compounds, such as word class, allomorphy and the differences between adjective-specific and non-adjective-specific intensifiers. Any variation in age or frequency of the base forms - the intensifying compounds - will also be taken into account. Finally I will investigate for which of the three semantic interpretations of intensifying diminutive compounds (informal, strengthening and weakening) evidence can be found.

The theoretical preliminaries necessary for the interpretation of the results and the analysis that I will provide will be discussed in Chapter 2. The focus of this chapter will be on the form and use of adjectives and diminutives in Dutch, on intensifying compounds in Dutch and other Germanic languages and on the framework of construction morphology (Booij 2010) within which I intend to place the analysis in this thesis. The corpus investigation will be carried out with the help of Google and based on a list of intensifying compounds contributed by Jack Hoeksema. Information on the frequency and age of the base forms stem from the Corpora from the Web (NLCOW2012-00X) and Van der Sijs’ chronological dictionary (2002). Method and material with be explained further in Chapter 3. In Chapter 4 I will discuss the results and in Chapter 5 I will suggest a subschema for intensifying diminutive compounds following Booij’s approach to Construction Morphology. Finally in Chapter 6 I will summarise my findings and conclusions.

(6)

(7)

2. Theoretical preliminaries

2.1 Diminutives in Dutch

Diminutives are used to express the meaning ‘small’ and other meanings derived from this, often with evaluative connotations. These connotations can be depreciative, pejorative, appreciative, affectionate, approximative or relativisation (Bakema & Geeraerts 2000:1045, 1048). A combination of these connotations and the simply explicative meaning ’small’ is also possible. This thesis is concerned with the morphological formation of diminutives through a suffix, rather than the periphrastic formation as in a small house. Bakema & Geeraerts (2000:1045f) give examples from many different languages of how diminutives can be formed morphologically such as prefigation, suffixation, reduplication and submorphemic changes (that is systematic changes to a phoneme, as in Basque txerri ’small pig’ derived from zerri ’pig’). In Dutch, Frisian and German suffixation is used; the other Germanic languages do not have a morphological diminutive form. Dutch, German and Italian are named as examples of languages where diminutive suffixes are attached mostly to nouns and less frequently to words from other word classes. Diminutives are extremely frequent in Dutch, especially with nouns. When suffixed to count nouns, the derived form does not always refer to a small (or dear, or insignificant) version of the noun. A telefoontje is not a small telephone but a telephone conversation and a hartje is not only a small heart but can also refer to the centre of a city or the middle of winter (Bakema & Geeraerts 2000:1049). When suffixed to mass nouns diminutives can have a partitive meaning: ijs (‘ice’, ‘ice cream’) → ijsje (‘ice lolly’), bier (‘beer’) → biertje (‘bottle of beer, glass of beer’). More rarely the suffix can be combined with words from other word classes such as verbs (weetje ‘know-DIM’ ‘a piece of knowledge’, moetje ‘must-DIM’ ‘a shotgun marriage’) and adjectives

(nieuwtje ‘a small piece of news’, blondje ‘a blond’), and adverbs (toetje lit. ‘after-DIM’,

‘dessert’, stilletjes ‘quietly’). Diminutivising of prepositions is extremely rare (uitje ‘outing’, ommetje ‘walk’) (De Haas & Trommelen 1993:282f.).

(8)

/r/) respectively (koninkje ‘king-DIM’, boompje ‘tree-DIM’). For monosyllabic words ending in

/r/ and words ending in a nasal consonant or /l/ preceded by a short vowel -etje is used (knalletje ‘bang -DIM’, spinnetje ‘spider-DIM’). The allomorph -je is used after obstruents

(graatje ‘fish bone-DIM’, aapje ‘monkey-DIM’) and -tje is used in all cases not covered above

(keitje ‘boulder-DIM’, hemeltje ‘heaven-DIM’) (1993:279). There are a small number of words

where two diminutive endings can be used. Two of these could be relevant for this thesis namely bloemetje and bloempje (in bloemetjezoet and bloempjezoet ‘flower-DIMsweet’) and

kipje and kippetje (in kipjelekker en kippetjelekker ‘chicken-DIM well’)1. The form of the

diminutive is not formed solely for phonological reasons. For instance, the diminutive forms of the three rhyming nouns kip ‘chicken’, lip ‘lip’ and schip ‘ship’ are all different: kip-etje, lip-je and scheep-je. There are other exceptions to the simplified rules explained here, but these are not relevant in the context of the intensifying compounds I will be investigating. For an exhaustive overview of diminutive formation see De Haas & Trommelen (1993: 279). In informal or regional use of Dutch the suffixes -ke and -ie are also found. These suffixes will not be part of the current corpus investigation. One of the questions this thesis aims to answer is whether intensifying diminutive compounds belong to the informal or formal register, therefore the focus will be on forms containing the standard form -tje (or its allomorphs). While -ke and -ie can be inserted into intensifying compounds (Cf. 2.2, 4.4.2 and 4.7.3), the use of these suffixes in itself implies some form of informal use.

While with evaluative use there may still be some connection to smallness, based for instance on the general metaphors ‘small is beautiful’ or ‘small is insignificant’ (Bakema & Geeraerts 2000: 1048), this is not always the case. Apart from the explicative and evaluative meanings, there are cases of conventionalised use of diminutive forms. To illustrate what context-dependent meanings diminutive forms may have, let us look at the Van Dale dictionary entries for mannetje ‘man-DIM’ and vrouwtje ‘woman-DIM' (Den Boon et al. 2005).

Twelve definitions are given for mannetje and seven for vrouwtje. Apart from the meanings ‘small man’ and ‘young boy’ mannetje can refer to a drawn stick man, a handy-man, to aliens (marsmannetjes) or to the male of an animal or plant species, in a compound or independently. It is also used in a number of idioms, such as zijn mannetje staan (lit. ‘stand one’s little man’, ‘hold one’s ground’), mannetjes maken (lit. ‘make little men’, ‘make excuses’). Mannetje can refer to non-living creatures, it can mean a certain type of electric

(9)

plug, a bracket for hanging up decorative plates or the hand on a set of kitchen scales. Additionally various technical meanings of mannetje from the fields of ship building, electrical engineering, metal work and rowing are given. Vrouwtje has a number of meanings corresponding to those for mannetje. Apart from ‘small woman’ and ‘young girl’ this diminutive form can refer to the female of a species and a female socket. In addition vrouwtje can be used to refer to a female dog owner (a male dog owner is baasje ‘boss--DIM’)

or to describe (normal sized) women with a connotation of either contempt or endearment. The use of vrouwtje(s) and mannetje(s) to refer to a female or male of a species, both in a compound and independently is discussed more extensively by Van der Wouden (2007:127). These forms are an especially good example of conventionalised use of the diminutive as they only express information about the animal’s gender, there is no connection to size or evaluation whatsoever. Vrouwtjesolifant, ‘female elephant’ and mannetjesolifant, ‘male elephant’, do not refer to small elephants, and they have a neutral meaning. The same goes for compounds referring to animals with vrouwtje or mannetje as their right constituent: gorillavrouwtje ‘female gorilla’, gorillamannetje ‘male gorilla’. This use of vrouwtje and mannetje is included in the Van Dale dictionary under definition 5 for vrouwtje and definition 3 voor mannetje (Den Boon et al. 2005).

In sum Dutch diminutives are very versatile. This suffix can be combined with many word classes but most frequently and freely with nouns (Bakema & Geeraerts 2000:1048) and the many meanings of vrouwtje and mannetje alone make clear that diminutive forms in Dutch are not only related to smallness, but have a wide range of connotations and conventionalised meanings.

2.2 Intensifying compounds

Intensifying compounds have been used in Dutch since the fourteenth century (Van der Sijs 2002) to express a high degree of the property expressed by the right constituent. The latter can belong to various word classes, but this thesis will look only at adjectival compounds.

(10)

The noun-adjective compounds with intensifying meaning can be roughly divided into two categories: those which are used metaphorically and those where only the intensifier meaning remains (De Haas & Trommelen 1993: 428). The first type can be paraphrased in similes such as loodzwaar ‘lead heavy, as heavy as lead’, harkmager ‘rake skinny, as skinny as a rake’. Some even have corresponding idiomatic similes (beresterk/sterk als een beer) (Hoeksema 2012:105). This type is prototypical for adjective-specific intensifiers (Fletcher 1980:446). Not all similes that are collapsed into compound adjectives have an intensifying meaning. For instance in cases where the adjective is a colour name, the compound can express both a certain degree (or shade, one would normally say) (3a) and a high degree (3b) of a colour.

(3a) mintgroen ‘mint green’, kastanjebruin ‘chestnut brown’, marineblauw ‘navy blue’ (3b) sneeuwwit ‘snow white, very white’, kanariegeel ‘canary yellow, very yellow’,

pikzwart ‘pitch black, very black’

That there is no one-to-one connection between similes and intensifying compounds is also shown by the fact that there are similes without corresponding compounds: sluw als een vos ‘sly as a fox’, *vossluw ‘fox-sly’ and lelijk als de nacht ‘ugly as the night’, *nachtlelijk ‘night-ugly’. Some intensifying diminutive compounds can themselves be paraphrased as a simile, usually if the intensifier corresponds to a count noun. Note that the diminutive can give a more specific meaning to the simile. For instance poesjeslief could be paraphrased in ‘as sweet as small cats’. Small cats are sweet and many people might find them sweeter than large cats. However, for compounds such as kaarsrecht ‘candle straight’ the diminutive is semantically irrelevant as small candle are not usually straighter than big candles.

The second category contains compounds which can only be used with an intensifying meaning ‘very A’. Paraphrasing is still possible, for instance for straatarm (‘street poor’): ‘so poor they live on the street’ or ‘as mad as something with horns, like a bull’ for horendol. But these paraphrases are not set phrases and sound far-fetched. For more productive intensifiers it soon becomes impossible to find a suitable paraphrase.

(11)

interpreted as 'wonder clean’). Here the intensifier is interpreted as having only the meaning ‘to a high degree’. These forms used by children have not entered the standard language, but other analogous forms have (Hoeksema 2012: 105f.). This is preceded by the bleaching of the original meaning, as is very clear for keihard ‘boulder hard’, which can also be used to mean ‘very loud’ and ‘very fast’. ‘Loud’ and ‘fast’ are other meanings of hard, but not characteristics of boulders. This is the first step in the development of originally adjective-specific intensifiers into productive ones which can be combined with virtually any adjective (Fletcher 1980:448f.). There is no clear boundary between those compounds where the meaning can be derived from its constituents (metaphorical) and those where it cannot (De Haas & Trommelen 1993: 426ff.). Even an adjective-specific intensifier such as peperduur ‘pepper expensive’ does not give an exact idea of expense. It is used metaphorically and the meaning ‘very expensive’ is not composed of the meaning of peper and duur.

Some intensifying compounds where the left constituent corresponds to a verb can be paraphrased as ‘so A that it V’ or ‘so A that I V’ (De Haas & Trommelen 1993:433). For instance fonkelnieuw ‘sparkle new’, ‘so new that it sparkles’ or stikheet ‘suffocate hot’ ‘so hot that I am suffocating’. This is not the case for all verb-adjective compounds, but there is often a semantic link (vrieskoud ‘freeze cold’, kotsziek ‘throw up sick’).

Prepositions that can have an intensifying meaning are door- (doornat ‘through wet’, doorkoud ‘through cold’ and in- (intriest ‘in sad’, ingemeen ‘in mean’). Over- is a special case. In some compounds it behaves like the other intensifiers and expresses a high degree of the property of the adjective (overheerlijk ‘very delicious’), while in other compounds over-expresses too high a degree of a property. This is the case for overvol ‘too full’, overrijp ‘over ripe’ and other compounds (Hoeksema 2012:134).

Another type of intensifier is the intensifying prefix (aarts- ‘arch’, super- ‘super’). When these are used the formation is not strictly speaking a compound (Hoeksema 2012: 111), but as they are built up of intensifier and adjective, they will be treated as such, for the purpose of the current analysis.

(12)

dagenlang ‘days long’, nachtenlang ‘night long’, urenlang ‘hours long’; four are measures of distance: ellenlang ‘yards long’2_{, kilometerslang ‘kilometers long’, meterslang ‘metres long’,}

mijlenver ‘miles long’; three are other measures: vuistdik ‘fist thick’, duimendik ‘inches thick’, metershoog ‘metres high’. These are not regular measure constructions as they do not simply state a length of time or distance, but have a special degree reading and imply that it is felt to be an extreme length or distance.

Finally there is a type of intensifying compound which De Haas & Trommelen (1993:436) call ‘pseudo-compounds’. This is because the intensifier is not an independent lexeme. Examples from the list used for this corpus-based study are foeilelijk ‘very ugly’, morsdood ‘very dead’, starnakelzat ‘very drunk’ and tjokvol ‘very full’. These intensifiers are not productive.

The use of these intensifying compounds is an integral part of Dutch, not marginal as in English. A complete list of such compounds would be many times longer than that used for this analysis, which indicates the vitality of this type of intensification in Dutch. Although the same types of compound are found in other Germanic languages, Dutch makes the fullest use of them. In German intensifying compounds are also quite common (Oebel 2011). Some of these, like stocksteif (‘stick stiff’) and steinhart (‘stone hard’), correspond one to one to Dutch intensifying compounds (stokstijf and steenhard). Others such as schweineteuer ‘pig expensive’ and furztrocken ‘fart dry’ do not have a Dutch counterpart. Frisian, Swedish and English also have intensifying compounds. I will carry out small-scale searches to find out whether intensifying diminutive compounds are found in these other Germanic languages as well. This is in order to test Hoeksema’s claim that this type of word formation is restricted to Dutch (2012:123).

As is explained in Hoeksema (2001) and Booij & Hüning (forthc.:20), intensifiers are among the elements which can be reduplicated to further reinforce their emphatic meaning. This can be done regardless of the original word class of the intensifier, but it can only be done with intensifying compounds, not with the left constituent of other compounds (Hoeksema 2012:109). For this purpose either coordinated reduplication with en ‘and’ or the phrase maar dan ook ‘but then also’ can be used. Examples of the resulting forms are bloed-en bloedheet ‘blood and blood hot’ and stebloed-en- maar dan ook stebloed-enkoud ‘stone but thbloed-en also stone-cold’. Non-coordinated reduplication is also possible: steen, steenkoud water ‘stone,

(13)

stone-cold water’. Cappelle has also found cases of a coordinated use of two different intensifiers: tjok- en bomvol ‘chock and bomb full’ and snip maar dan ook echt snip en snotverkouden ‘snip but then also really snip and snot cold’ (Cappelle 2012a: 67). Diminutivised left constituents are also found in these types of formations (cf. 4.5).

2.3 Construction Morphology

Construction Grammar (CxG) is based around the idea that constructions form the basis of grammar and that they are symbolic pairings of form and meaning. There are various different approaches to CxG, but this is what they all have in common. This symbolic correspondence between form and meaning, and their different properties, is shown in Figure 2.1.

Figure 2.1 Symbolic structure of a construction (Croft 2001:18)

(14)

Construction type Traditional name Examples

Complex and (mostly) schematic syntax [SBJ be-TNS VERB-en by OBL] Complex and (mostly) specific idiom [pull-TNS NP-'s leg]

Complex but bound morphology [NOUN-s], [VERB-TNS] Atomic and schematic syntactic category [DEM], [ADJ]

Atomic and specific word/lexicon [this], [green] Table 2.2 The syntax–lexicon continuum

A construction may have holistic properties that are not the sum of the meaning of its constituents, as is the case for intensifying compounds. The meaning of the right constituent (the adjective) is conveyed, but the semantic property of ‘high degree’ cannot be derived from the left constituent. Construction Morphology assumes that conventionalised complex words are listed in the lexicon. It is also possible to specify and refer to paradigmatic relations between complex words.

(15)

Schematic notations

In this thesis I will use the same schematic notation for constructions as Booij does. The elements of form and meaning are both placed within square brackets and connected by a double arrow (↔) to show the symbolic link. An example of this notation are the schemas Booij gives for diminutives (4a-b). Paradigmatic relationships are expressed by placing two or more schemas within angle brackets (< >), which demarcate a constructional schema, and connecting them with a wavy equal sign (≈).

(4a) [[X]Yi–je]Nj↔ [SMALL [ENTITY RELATED TO SEMi]]j

(4b) [[X]Ni–je]Nj↔ [SMALL [SEMi]]j

Within the hierarchical lexicon (4b) is a subschema of (4a), because (4b) is more specific, applying only to diminutives derived from nouns. In these schemas Booij does not take into account the many other meanings of the diminutive besides ‘small’. These various meanings (see 2.1) can however also be expressed in constructions, for instance as shown in (5a-b).

(5a) [[X]Yi–je]Nj↔ [WORTHLESS [ENTITY RELATED TO SEMi]]j

(5b) [[X]Ni–je]Nj↔ [SWEET [SEMi]]j

(16)

number of adjectives which can fill the open position) is restricted and these do not form a constructional idiom of their own. However, they do have subschemas (6c) which fall under the general schema in (6a), because they still have the meaning [high degree of SEM] which cannot be deduced from the meaning of the compound’s constituents. Intensifiers which are unproductive and are only used in one or a small number of compounds, still fall under the general schema, but they do not create a constructional idiom of their own.

(6a) [x Ai]Aj ↔ [high degree of SEMi]

(6b) [bere Ai]Aj↔ [high degree of SEMi] ‘bear adj’

[kanker Ai]Aj ↔ [high degree of SEMi] ‘cancer adj’

[super Ai]Aj↔ [high degree of SEMi] ‘super adj’

(6c) [kaarsrecht]Aj ↔ [high degree of RECHT] ‘candle straight’

[tjokvol]Aj↔ [high degree of VOL] ‘chock full’

Placing the schemas and constructional idioms for each intensifying compound under one general schema allows us to group these left constituents with intensifying meaning together. Which in turn makes it possible to show, for instance, that it is this group of morphemes with intensifying meaning that can be reduplicated to reinforce that meaning This reduplication can be conveyed in a constructional schema as in (7) (Booij & Hüning forthc.:20).

(7) <[x en x Ai]Aj↔ [very high degree of SEMi]j> ≈ <[x Ai]Aj↔ [high degree of SEMi]j>

(17)

3. Method and material

A list of 696 intensifying compoundscontributed by Jack Hoeksema (cf. also Hoeksema 2012) forms the basis for the corpus investigation. The compounds on this list are a representative selection of the large number of possible intensifying compounds in Dutch. The intensifying left constituents stem from all possible word classes and take different diminutive allomorphs. The compounds themselves are also of varying frequency, ranging from hapax legomena such as concertgebouwecht ‘concert hall real’ to very common words such as bomvol or gloednieuw.3_{All these linguistic properties of the intensifying compounds will be}

categorised, to determine later whether they influence the occurrence or the frequency of the corresponding intensifying diminutive form.

The age of the intensifying compounds, without a diminutive, were found with the help of Van der Sijs’ chronological dictionary (2002) while Corpora from the Web (COW. Schäfer & Bildhauer 2012) was used to determine their frequency.4 _{The Dutch part of this corpus}

(subcorpus NLCOW2012-00X) was made in 2012 and contains 2,366,453,439 tokens in 121,582,724 sentences and 1,594,241 documents. This large collection of online documents was created through crawling rather than through entering search engine queries. This makes the corpus less biased towards certain hosts, i.e. the documents originate from a larger number of different websites. When many results are from the same websites, this also results in linguistic bias (Schäfer & Bildhauer 2012:487f). This corpus would therefore also have been first choice to determine the frequency and use of intensifying diminutive compounds as well. However, these forms are not (yet?) frequent enough to occur in even the largest linguistically processed online corpora. The NLCOW2012-00X corpus only returns a handful of results for even the most frequent forms. The Corpus of Spoken Dutch (CGN) (Oostdijk:2000) does not provide enough data either.

Instead Google lends itself as the most suitable interface for finding examples of intensifying diminutive compounds. Apart from the volume of data available another advantage of Google is easy accessibility of a broader context and information about the language user. Disadvantages are that many duplicate results are returned, while at the same time not all examples are shown. The search engine is aimed at finding information

(18)

rather than determining frequencies and it tries to second-guess the ‘best’ results based on previous searches and pages visited. Nevertheless Google shows more results than any other interface; these shortcomings just need to be kept in mind when looking at very specific aspects of the usage and users of intensifying diminutive compounds.5

The 696 compounds were thus entered into the search engine with the intensifier in the diminutive form. For each word different possible forms were included in the search. The diminutive ending can be used in the plural and singular, adjectives have a form with and without a flectional -e and compounds can be written as one or two words. Not all adjectives have a form with a flectional -e, for instance knaloranje, which already has an –e. There are rules on the spelling of compounds in Dutch (whether they are written as one word, hyphenated or written separately), but they are not always adhered to. To apply these rules strictly to the searches would therefore mean excluding relevant data.6 The combined possibilities result in eight forms for each compound, illustrated below for spiegeltjeglad ‘mirror-DIMslippery’.

singular, no flectional -e, joined spiegeltjeglad singular, flectional -e, joined spiegeltjegladde plural, no flectional -e, joined spiegeltjesglad plural, flectional -e, joined spiegeltjesgladde singular, no flectional -e, un-joined spiegeltje glad singular, flectional -e, un-joined spiegeltje gladde plural, no flectional -e, un-joined spiegeltjes glad plural, flectional -e, un-joined spiegeltjes gladde

Furthermore there are two intensifiers which can be combined with more than one diminutive allomorph (bloemetjezoet and bloempjezoet for bloemzoet (‘flower sweet’), kipjelekker and kippetjelekker for kiplekker (‘chicken well’) and thirty cases with a connective (or plural) -s, where a diminutive ending may be added before or after the -s (stervensjedruk and sterventjedruk ‘die-DIMbusy’). For a small number of other intensifiers different spellings

are allowed, for instance hoorntjedol and horentjedol (‘horn-DIM crazy’). All these

5_{Another factor which may have some relevance, but could not practically be taken into account, is the}

frequency of the left constituent with a diminutive ending, outside of the compounds (i.e. if bommetje is more frequent than bikkeltje, is bommetjevol ‘bomb-DIMfull’ more frequent than bikkeltjehard ‘bone-DIMhard’?).

6_{In this thesis I will use the standard spelling (this usually entails writing the compound as one word), except}

(19)

alternatives resulted in a total of 5600 search queries, which were carried out between March and June 2013.

According to De Haas & Trommelen (1993: 428) intensifying compounds do not have comparative and superlative forms. Hoeksema (2012:132) has found that comparative and superlative forms are attested, but that these are extremely rare. Therefore practical considerations mean that these forms will not be part of the current corpus investigation.

(20)

The frequencies for each attested compound were noted as can be seen in Figure 3.1. These are not the raw frequencies given by Google. The results were manually checked for duplicates and false hits such as linguistic articles or sentences where an adjective follows a diminutive form without them forming a compound. For those intensifying diminutive compounds occurring more than thirty times a random selection of thirty sentences was analysed. For those which occurred thirty or fewer times all cases were analysed. To make this (quantitative and qualitative) analysis possible the sentences found were stored in a Microsoft Access file to keep the context of each occurrence accessible. Figure 3.2 is a screenshot of the database and gives an idea of the type of data collected. Metadata about the year of the occurrence, the type of use, and the characteristics of the language user (age, region and gender) have been added to this file.

(21)

4. Results

4.1. Frequencies

The results of the search queries for all 696 intensifying compounds can be found in table form in appendix 1. For each compound attested in the intensifying diminutive form, a maximum of thirty cases were analysed, amounting to a total number of 3726 sentences. These sentences, along with the relevant metadata can be found in the appendix 2. Around forty-five percent (311 out of 696) of the intensifying compounds on the list are attested in the data with the left constituent in the diminutive form. The eight most frequently occurring words are bloedjeheet ‘bloed-DIM hot’, bloedjemooi ‘blood-DIM beautiful’,

bloedjesnel ‘blood-DIM fast’, bommetjevol ‘bomb-DIM full’, gloedjenieuw ‘glow-DIM new’,

mudjevol ‘bushel-DIMfull’, poedeltjenaakt ‘poodle-DIMnaked’ and schathemeltjerijk ‘treasure

heaven-DIMrich’, all with over 500 hits.

Three of these have the intensifier bloed- as their left constituent. Although bloed is a mass noun, bloedje is an independent noun meaning ‘darlings, (poor) little things’, often used to refer to children. The familiarity of this form – even though it is used with a different meaning – may be the reason that many compounds with bloedje- are found, in fact all intensifying compounds with bloed- are attested at least once with a diminutive suffix as well. The English adjective bloody, which is used by some speakers of Dutch as well, may have played a role here. In fact on one forum someone responds to “Ik ben nu bloedjeserieus”(‘I am blood-DIMserious now’) with "Ik ben niet anders geweest dan bloody

serious." (‘I have been nothing other than bloody serious’)7_{. Thus to some language users at}

least, these forms are interchangeable. The frequency of compounds with bloedje- could itself have led to language users forming other new intensifying diminutive compounds. The highly frequent gloedjenieuw (over 500) as well as gloedjeheet (98) look like analogous formations.

As table 4.1 shows there are a small number of forms with a high frequency and many more forms with a lower frequency. A large number of the intensifying compounds occur with a diminutive only once (68 words) or twice (36 words) in the corpus, suggesting productivity is high and that usage is not restricted to a small number of lexicalised forms.

(22)

Number of hits for

each compound Compounds

Over 500 8

100-500 26

10-100 94

Under 10 183

Table 4.1 Frequency breakdown for intensifying diminutive compounds

In the rest of this chapter I will look first at the diachronic development of the intensifying diminutive compounds and the relevance of their base forms’ age and frequency. Then I will discuss the results for the linguistic factors, concerning both the properties of the intensifying compounds themselves, and those of their diminutive form, as well as some specific cases of reduplication and combination. Furthermore I will discuss the weakening versus strengthening meaning, and informal versus formal use of intensifying diminutive compounds. I will look at the results of the demographic factors and finally I will give some evidence for the existence of intensifying diminutive compounds in other languages than Dutch.

4.2 Diachronic development

A year can be given for all but 94 of the 3728 results. With this information we can establish how long intensifying diminutive compounds have been in existence and whether any diachronic change has taken place during this period.

Hoeksema (2012:123) and Reker (1996:44) both state that this type of compound first reared its head in the early 1990s. However, with the help of the digitalised historical archives of the national library of the Netherlands8 and the Digital Library of Dutch9, it is possible to predate this estimation by a considerable margin. The oldest attestation I have found is muisjesstil in a newspaper from 1883. Muisje(s)stil is the most frequent word in the corpus in the 19thand 20thcenturies and could be based on (or at least reinforced by) the simile zo stil als muisjes (‘as quiet as little mice’) rather than the base form muisstil.10There may therefore be some doubt as to whether muisjesstil (‘mice-DIM quiet’) is genuine

evidence for the existence of the intensifying diminutive compound at this time. There is

8_{Available online at}_{http://kranten.kb.nl/}

(23)

however other evidence from around the turn of the 20th _{century. In 1889 potjedicht}

(‘pot-DIMclose’) was used in a translation of Dickens’ Notes from America to describe the (misty)

conditions at sea. The original says “It was decidedly close” and an earlier translation from 1842 uses potdigt. The forms poesjeslief ‘cat-DIM sweet’ and vuurtjesrood ‘fire-DIM red’ are

attested in novels from 1896 and 1901 respectively. As far as poesjeslief ‘cat-DIM sweet’ is

concerned, the comparison lief als poesjes ‘as sweet as small cats’ may have played a more important role than a base form poeslief. Also, poesje lief is attested as a pet name11_{as early}

as 1737.12 The intensifiers in the majority of the 19th and 20thcentury material are nouns, almost all of them count nouns. This makes it harder to draw a distinction between those which are based on intensifying compounds and those which could be based on a simile. Despite this two of these early attestations (potjedicht 1889 and vuurtjesrood 1901) are clear examples of intensifying diminutive compounds: the mist that Dickens writes about is not as close as a small pot, nor does ‘as red as small fires’ make any sense, as a small fire is just as red as a big one. There are almost certainly more examples from this time, in undigitalised publications or with base forms which were not included in the list. That this is the case is exemplified by the chance encounter I had with the words staafslank and staafjesslank ‘rod thin’ in a novel published in 1906.13_{Due to the high level of arbitrariness in the digitalisation}

of publications and in the selection of intensifying compounds included in this research, it is not necessary to give potjedicht and vuurtjesrood a special status.

Throughout the rest of the 20th _{century more intensifying diminutive compounds are}

found, with new words frequently appearing (type: 29, tokens: 74). The data from the 20th

century stems from very diverse and fascinating sources. There are a number of hits from various newspapers, usually from letters to the editor, columns or the children’s pages; though once muisje-stil is used in a report about a woman being robbed. In a trade union publication from 1944 torentjehoog is used in an article written in the slang of a Rotterdam dock worker, while in a more formal article by the same author in the same paper torenhoog is used instead. Intensifying diminutive compounds are also found in novels, songs and poems. There are twenty-one hits from poems and song lyrics in the entire corpus (1883-2013) and in these cases the metre will have influenced the choice for the diminutive forms

11_{A term of endearment, rather than a name given to pets.}

(24)

over their base forms as well as any other considerations. The most formal source from this period would be the war diaries of O.C.A. van Lidth de Jeude, Minister of War in Queen Wilhelmina’s government in exile during the Second World War. He uses poesjelief twice to describe Queen Wilhelmina.

Around 1950 the form bloedje-link was used in an advertising slogan by De Kuyper, a Dutch distillers. Photo A shows one of the company’s crates and was taken around 1950 during the Dutch military campaign in Indonesia (see (8)). Bloedje-link is also used in the war diaries of a soldier who fought in Indonesia at the same time (9).

Photo A

De Kuyper’s advertisement with use of bloedjelink (8) De Kuyper: zelfs waar ‘t bloedje-link is!

De Kuyper: even where it blood-DIM risky is.

‘De Kuyper: even where it’s very risky’14

(9) Dat is altijd een gevaarlijk karwei en bloedje link That is always a dangerous job and blood-DIMrisky

‘That is always a dangerous job and very risky.’ Kloote, L. 194815

Finally the transcripts of meetings of the Dutch Lower House of Parliament provide some data from 1924 and 1973. Example (10), from the archive of parliamentary documents (1814-1995), was uttered by an MP in August 1973.

(10) Voor planmatige beleidsdenkertjes is dat om hoorntjesdol van te worden For planning policy thinkers-DIM is that to ‘horns-DIMcrazy’ of to become

‘This drives systematic policymakers crazy’

http://resourcessgd.kb.nl/SGD/19721973/FULLTEXT/SGD_19721973_tekst_0003342.xml

14_{The photograph taken between 1946 and 1950 is in the photo database of the Dutch national archives}

http://proxy.handle.net/10648/af0307fc-d0b4-102d-bcf8-003048976d84

(25)

After 1990 there is a sudden increase in data: almost half of the 20th_{century data (35 out of}

75 attestations) is from the 1990s. This would fit with Hoeksema’s (2012:123) and Reker’s (1996:44) observations that this phenomenon became popular in the 1990s. However, the sudden increase in my data is at least in part down to the invention of internet. Online forums first crop up in the corpus in the late nineties; (sports) clubs start uploading their magazines and leaflets, and newspapers start publishing articles online. From earlier periods this kind of material is only available through digitalised databases, and therefore less frequent. The sudden increase in data may be just that: an increase in data which cannot necessarily be equated with a sudden increase in use.

Much more data is available from the 21stcentury: 3556 tokens for the first 13.5 years of the 21stcentury compared to 74 tokens for the entire previous century. Table 4.2 shows the number of occurrences per year.

Table 4.2 Occurrences of intensifying diminutive compounds in the 21st_{century by year}

A first glance at this table might lead to the conclusion that intensifying diminutive compounds have become steadily more common (with a slight glitch in 2009) since the beginning of the 21st _{century. However, the amount of data on the internet has also got}

larger since the beginning of the 21st_{century. Consequently I think it is only safe to say that}

this phenomenon has not got less frequent; usage appears to be quite stable.

4.3 Age and frequency of intensifying compounds

The chronological dictionary used to determine the age of the intensifying compounds (Van der Sijs: 2002) only includes 99 of the most common compounds. Despite this apparent lack of data, the information for these 99 compounds is of interest. Even based on this

16_{Up until June 2013}

Year Hits Year Hits

(26)

admittedly small amount of data there is little evidence for a relationship between the existence or frequency of intensifying diminutive compounds and the age of the base form. The most frequently attested intensifying diminutive compounds (400+ attestations) have base forms of strongly varying ages, from the year 1617 for propvol ‘stuff full’ and 1889 for poedelnaakt ‘poodle naked’ to the year 1989 for bloedmooi ‘blood beautiful’. And among those intensifying compounds that are not attested with an inserted diminutive at all we find steenoud ‘stone old’ (1350), bliksemsnel ‘lightning fast’ (1804) and loeisterk ‘blare strong’ (1984). It is also not the case that the intensifying diminutive compounds which are attested in the 19thand 20thcenturies have older base forms than those which are not found until the 21stcentury. All in all it appears that the distribution of compounds of different ages over the list is so arbitrary that the addition of chronological information for more compounds could not create a more structured pattern relating frequency to age.

The results for the frequency of intensifying compounds in COW show some overlap with the frequency lists for intensifying diminutive compounds based on Google search queries. The lists (see appendix 1) do not correspond one to one and some compounds are highly frequent with a diminutive but much less frequent without, or vice versa. By way of example ijzertjesterk ‘iron-DIM strong’ and brandjenieuwsgierig ‘burn-DIM curious’ are both found

three times with Google, while in the NLCOW2012-00X corpus ijzersterk it occurs 4464 times and brandnieuwsgierig only once. Higher up the list of frequencies, the differences are not as extreme, but there are still some. Haarscherp ‘hair sharp’ is found 176 times with a diminutive and 1412 times without, while nokvol ‘ridge full’ is found 173 times with a diminutive and only 28 times without. When the frequency of the base form is high, this does not necessarily mean intensifying diminutive forms will be found. However, the other way around there is a clear trend. When a base form does not occur in the NLCOW2012-00X corpus - and thus can be said to have a lower frequency overall – it is very unlikely that any intensifying diminutive compounds will have been found for this word. This may seem obvious, but it confirms that the intensifying compounds are indeed the base forms and that the diminutive versions are not formed on their own. One exception to this trend is aapjesgeil ‘monkey-DIM horny’, which is attested four times with a diminutive, but not

(27)

no alternative where the same corpus can be used for both forms: the diminutive forms are too infrequent to be found in COW and the base forms are too frequent for Google to give anything more than an unreliable estimate.

The overlap between the two frequency lists by no means explains all the variation in frequency among intensifying diminutive compounds. The frequency of the base form could be relevant in another way as well though. Each individual language user is different when it comes to developing schemas, because each individual has different lexical knowledge and makes different generalisations (Booij 2012:89). It is possible that language users coin or use intensifying diminutive compounds for base forms they are very familiar with, because they are frequent in their own environment. This hypothesis can, however, not be tested by means of a corpus investigation.

4.4 Linguistic factors 4.4.1. Word class

Word class Attested inDIM

form Not attested inDIMform

Total Relative number

attested Noun 216 235 451 48% Measure noun construction 6 10 16 38% Verb 50 53 103 49% Adjective 6 25 31 24% Preposition 1 15 16 6% Prefix 25 50 75 33% Other 2 2 4 50% Total 311 385 696 47%

Table 4.3 Frequencies of intensifying diminutive compounds by word class

Nouns

(28)

normally do not occur with a diminutive, such as bloed ‘blood’, gras ‘grass’, water ‘water’, ijzer ‘iron’ and dood ‘death’, are attested in diminutive intensifying compounds. This points towards the use of intensifying diminutive compounds not being linked to regular usage of the diminutive. Measure noun constructions have been treated as a separate category. Six out of sixteen cases were attested in the diminutive form, confirming that measure nouns behave in the same way as other intensifying compounds.

Adjectives/Adverbs

Intensifying diminutive compounds with two adjectival constituents are rarer than noun-adjective compounds. Six out of 31 cases are attested and for all but one of them the left constituent is dolletje(s) (‘crazy, fun’), which in its plural diminutive form is an existing adverb meaning wonderful. According to Van Dale (2012: dolletjes) it is meisjestaal ‘girls’ language’, but in my data its use as an intensifier is equally frequent among men and women of different ages. The only other adjective found in the diminutive form is klaartjewakker ‘clear-DIMawake’. Both klaar- and dol- can be seen as borderline cases; they could be either

intensifiers or part of adverb-adjective combinations (Hoeksema 2012: 118).

Verbs

Of the 103 verb-adjective compounds fifty occur with a diminutive ending on the left constituent. In eleven of these cases the verb stem is knetter- or kneiter- (see 4.7.3). Besides these eleven there are a notable number of compounds with the right constituent nat ‘wet’: drijfjenat ‘drip-DIM wet’, druipjenat ‘drip-DIM wet’, kletsnat ‘splash-DIM wet’, kliedernat

‘mess-DIMwet’, zeikjenat ‘piss-DIM wet’. Other verbal left constituents occurring more than

once are stikje- ‘suffocate-DIM’ (4 times), kotsje- ‘sick-DIM’, kraakje- ‘crack-DIM’, loeitje ‘blare-DIM’, smoortje- ‘suffocate-DIM’ and stamp(ens)je- ‘stamp-DIM’ (all 3 times) and

kakeltje-‘cackle-DIM’, piepje- ‘squeak-DIM’, snoeitje- ‘clip-DIM’ and sterventjes- ‘die-DIM’ (all twice).

There are also many compounds with these same verb stems as their left constituent which are not attested with a diminutive (*loeitjestrak ‘blare-DIMtight’, *stikjeleuk ‘suffocate-DIM

(29)

Prepositions

No results were found for doornat ‘through wet’ or any of the eleven compounds with in-. Of the four compounds with the preposition over as their left constituent, only overtjevol17 ‘over-DIM full’ is attested. Prepositions with an intensifying meaning can be reinforced

through reduplication: door-en-door-nat, in-en-intriest (cf 2.2 and 4.5). Prepositions cannot occur with diminutives in any other context, making it unlikely to find them in the context of intensifying diminutive compounds. Why overtjevol is attested is not clear. Perhaps it has to do with the alternative meaning of over-, which expresses too high a degree of a property (Hoeksema 2012:134, De Haas & Trommelen 1993:435). This is the case for overvol ‘too full’ and this exceptional meaning may be connected to it occurring with a diminutive, in contrast to the other prepositions. Alternatively this form may have been coined under influence of diminutive use of the intensifier German über- ‘over’. Übertje(s)- appears to be more frequent than overtje(s)-. See ‘prefixes’ below for more on über-.

Prefixes

Of the 75 compounds with a prefix as their left constituent, 25 can be found in diminutive form. In all but two of these cases the prefix in question is supertje-. Along with the other forms classified here as prefixes, super (and thus also supertje) may be viewed by some as an adverb of degree. It is no surprise that supertje- is the most frequent within this group, as research has shown that super- is as likely to express a high degree of a property of an adjective as erg ‘very’ (Hoeksema 2012:134).

The two exceptions where supertje- is not used are aartsjemoeilijk, used twice by the same person and übertjecool which occurs once. The prefix über- is a loan from German, which is also used with an intensifying meaning in Hungarian (Majtényi 2012). It was probably borrowed into Dutch via (American) English, rather than directly from German (Hoeksema 2012:103, Booij & Hüning forthc.:20). German über- is the equivalent of Dutch over-, the only Dutch preposition which was attested with a diminutive ending in the data. Although übertjecool is only attested once, übertje can be found with other adjectives which were not on the list, such as vet ‘cool’, schattig ‘cute’ and druk ‘busy’.

Other prefixes used as intensifiers, all of them of Latin origin are hyper-, turbo-, ultra- and mega-. None of the words on the list which contain one of these prefixes are found with an

(30)

inserted diminutive, but megaatjes ‘mega-DIM’ is attested – from 2006 onwards – with other

adjectives such as duur ‘expensive’, druk ‘busy’, populair ‘popular’ and slecht ‘bad’.

Other

Despite there being no independent words tjok and mors, let alone diminutive forms tjokje and morsje, both tjokjevol (18 times) and morsjedood (15 times) are attested. Starnakeltjezat and foeitjelelijk, however, are not. Perhaps tjok- and mors- are reinterpreted as nouns or verbs (cf. Dutch verb morsen ‘to spill’) or formed analogously with other intensifying diminutive compounds with vol ‘full’ and dood ‘dead’.

4.4.2. Allomorphs Diminutive

allomorph Attested inintensifying diminutive compounds Not attested in intensifying diminutive compounds Total Relative number attested tje 129 143 272 47% je 159 183 342 44% etje 15 26 41 27% je/tje 6 24 3018 20% pje 1 7 8 13% etje/je 1 0 1 100% pje/etje 0 1 1 0% kje 0 1 1 0% Total 311 385 696 45%

Table 4.4 Frequency of different diminutive allomorphs in intensifying diminutive compounds

Of the compounds where the left constituent takes the allomorph –je or –tje almost half actually occur as intensifying diminutive compounds. Though the allomorph –etje may seem to occur less in such compounds, these figures are strongly influenced by the eleven compounds with in- ‘in-‘ (indroevig ‘in sad’, inslecht ‘in bad’) and seven with stom- ‘stupid’ (stomdronken ‘stupid drunk’, stomverbaasd ‘stupid surprised’). Neither of these intensifiers has a nominal origin, and as we have seen prepositions and adjectives are far less likely to take a diminutive in the first place. Taking this into account, the frequencies for je, tje and -etje are similar. The allomorphs -kje and -pje on the other hand are very uncommon in

18_{All thirty of these compounds have a connective or plural -s. -je could have been added both before and after}

(31)

intensifying diminutive compounds. Perhaps these allomorphs are more strongly marked for diminutive meaning, which could deter language users, especially for a form such as boomlang > *boompjelang (tree tall) where a literal interpretation of the diminutive form makes no sense as an intensifier (‘as tall as a small tree’). Bloempje also has a more literal meaning of ‘small flower, while bloemetje is more conventionalised and refers to a bunch of flowers.

Sometimes the standard suffix -tje and the more informal –ie are both used in intensifying diminutive compounds within the same phrase or sentence. The examples in (11) show a combination of two intensifiers (cf.4.5. Type e). In addition proppievol ‘stuff-DIM full’ and

kleddertjesnat ‘soak-DIMwet’ are used in one sentence, as are spiksplintertjenieuw ‘splinter-DIMnew’ and poepietrots ‘poo-DIMproud’. This shows that in some cases both suffixes have

the same level of (in)formality.

(11a) spekkie spiegeltje glad

bacon-DIMmirror-DIMslippery

‘very very slippery’ (4061)19

(11b) zeikie kleddertje nat pee-DIMsoak-DIMwet

‘very very wet’ (2928)

4.4.3 Vowel clash

All Dutch diminutive allomorphs end in a vowel. When the right constituent of an intensifying compound starts with a vowel, this leads to a vowel clash with the -e of the diminutive suffix. Out of a total of 367 sentences where this is the case only thirteen cases are found where the vowel clash is not avoided in some way: stokjeoud, doodjeop, strontjeirritant (all occur once) and bloedjeirritant (occurs 10 times). In the data we find three strategies applied to avoid vowel clash and thus increase readability of the word. The first strategy is adding -s to the diminutive, which can be interpreted either as a plural ending or as a connective -s. Language users also hyphenate the compound, according to Dutch spelling rules or write the two constituents of the word separately. Van Goethem and Hiligsmann (2012:3) note that regular intensifying prefixoids which end in a vowel, such as reuze- (lit. ‘giant’) are sometimes written separately to avoid vowel clash and improve

(32)

readability. This is also applicable to the diminutive endings on intensifiers, but it should be noted as well that hyphens, a connective -s and spaces are also used when the right constituent starts with a consonant. A fourth strategy to avoid vowel clash is to not use an intensifying diminutive compound in the first place. Overall it is evident that adjectival right constituents which start with a vowel are less likely to be preceded by a diminutive than those starting with a consonant, i.e. compounds where the right constituent starts with a vowel are less likely to be attested as intensifying diminutive compounds (29%, compared to 46% for those starting with consonants).

4.4.4 Number

The diminutive suffixes inserted into intensifying compounds can occur in both plural and singular form, that is, with and without the ending -s. In 23 percent of the intensifying diminutive compounds the plural ending is used. These 881 cases often involve animals (aapjestrots ‘monkey-DIM proud’, beertjessterk ‘bear-DIM strong’, hondjesbrutaal ‘dog-DIM

rude’, kippetjeslekker ‘chicken-DIMwell’) or count nouns (prinsjesheerlijk ‘prince-DIM lovely’,

puntjesgaaf ‘point-DIM great’, torentjeshoog ‘tower-DIM high’). The connective -e in base

forms such as apetrots and beresterk may play a role here, having been misinterpreted as the Dutch plural ending -en (often with a silent -n). In other cases an -s is used although the intensifier is not a countable object or even a noun at all (supertjes snel ‘super-DIM fast’,

stikjes heet ‘suffocate-DIMhot’). Context does not play a role either; it is for instance not the

case that a predicate describing a plural subject or object gets the plural suffix (Zij is bloedjes mooi ‘She is blood-DIMbeautiful’). As there is no connection to a plural meaning, apart from

a possible misinterpretation of connective -e, this is apparently a purely formal use of -s, perhaps more related to the connective -s, than to the plural ending. Most noun-noun compounds, where the left constituent is a diminutive, but not an intensifier, are ill-formed without the -s (bloemetjesjurk, *bloemetjejurk ‘flower-DIM dress’, stoeltjeslift, *stoeltjelift

‘chair-DIM lift’) even when no plural meaning is possible (sprookjeshuwelijk,

*sprookjehuwelijk, ‘fairytale-DIM wedding’) (Van der Wouden 2007: 130). This could well

(33)

4.4.5 Types of use

The large majority of intensifying diminutive compounds are used predicatively. Slightly over ten percent of the cases in the data are attributive (385 of 3726 cases). Nominalised and absolute use are less common; they occur 28 times and 21 times respectively. All types of use that are possible for bare adjectives and intensifying compounds are possible for intensifying diminutive compounds. Examples of each use are given in (12 -16).

Predicative use

(12a) OMG ze is echt potje doof OMG she is really pot-DIMdeaf

‘OMG, she is really completely deaf’ (2079)

(12b) Ik word zo doodjeziek van dit gezing I become so dead-DIMsick of this singing

‘I am sick and tired of this singing’ (149)

Attributive use

(13a) een knalletje roze My Little Pony a bang-DIMpink My Little Pony

‘a bright pink My Little Pony’ (312)

(13b) een roetje zwarte uitlaat a soot-DIMblack exhaust

‘a very black exhaust’ (494)

Nominalised use

(14a) Uiteindelijk hebben we een hageltje nieuwe besteld in Amerika Eventually have we a hail-DIMnew ordered in America

‘Eventually we ordered a brand new one in America.’ (4322) (14b) In het pikjedonker kwamen we aan op de luchthaven.

In the pitch-DIMdark arrived wePARTat the airport

‘We arrived at the airport in the pitch-black (darkness)’ (3256) Absolute use

(15a) stampjesvol!

stamp-DIMfull (1935)

(15b) vedertje licht! feather-DIMlight

‘very light’ (674)

(34)

(16a) supertje strak uitgevoerd super-DIMtight performed

‘very well performed’ (4228)

(16b) door er gewoon knoepertjes hard aan te trekken by there just knoeper-DIMhard on to pull

‘by just pulling very hard’ (4158)

4.5 Reduplication and combination

Among the intensifying diminutive compounds where a diminutive suffix is inserted into a compound there are a number of more unusual forms, where the intensifier is reduplicated or combined with another intensifier. These are comparable to the forms found for normal intensifying compounds, mentioned in the background section. There are around eighty cases which fall into eight subcategories: reduplication (a,b); coordinated reduplication (c); reduplication without diminutive suffix (d); combination of two intensifiers (e,f); coordinated combination of two intensifiers (g); combination of two intensifiers, first without diminutive suffix (h); coordinated combination of two intensifiers, first without diminutive suffix (i); reduplication with the phrase maar dan ook (j,k) and finally a residual category (l,m,n).

a pisje pisje pisjelink ‘pis-DIMpis-DIMpis-DIMrisky’

b supertje supertje tevreden ‘super-DIMsuper-DIMcontent’

c graatje en graatjemager ‘bone-DIMand bone-DIMskinny’

d knetter en knettertje gestoord ‘crackle and crackle-DIMmad’

e bloedje snikje heet ‘blood-DIMgasp-DIMhot’

f gloedjes haartjes scherp ‘glow-DIMhair-DIMsharp’

g snotje en snipje verkouden ‘snot-DIMand snip-DIMcold’

h kei vedertje licht ‘boulder feather-DIMlight’

i tjok en propjevol ‘chock and stuff-DIMfull’

j spuug maar dan ook spuugje lelijk ‘spit really spit-DIMugly’

k laddertje maar dan ook werkelijk laddertjezat ‘ladder-DIMreally ladder-DIMdrunk’

l propje-, echt propjevol! ‘stuff-DIM, really stuff-DIMfull’

m bloed-, ja, bloedje-rood… ‘blood-, yes, blood-DIMred’

n gloedje de fonkel spiksplinter hageltje nieuwe ‘glow-DIMsparkle splinter hail-DIMnew’

In the German data there is also one case that fits into the residual category: mega-tunten-tucken-stöckchenschwul ‘mega-queers-poofs-stick-DIM-gay’. These examples show firstly a

(35)

not uncommon (Kreidler 2000:959). Secondly it shows that there is always space – and an apparent need – for yet another intensifying step.

4.6 Meaning

This apparent need for more intensification brings us to the question of the strengthening or weakening meaning of intensifying diminutive compounds. It is hard to determine for each attestation whether the language user is expressing a higher or lower level of intensity than that of the unaltered intensifying compound. However there is some evidence that these forms are not semantically identical to their base forms. The diminutive form is used almost as a superlative in some cases, for instance in (17) which follows on from two uses of brandschoon in the same post. In (18) supertje also appears to be superlative.

(17) Maar het is brandje schoon. But it is fire-DIMclean

‘But it is spotless’ (1639)

(18) Super, super, supertje mooi! Super, super super-DIMnice!

‘Super, super, really super nice!’ (3897)

The four examples below (19-22) show more cases where the diminutive form expresses a high level of intensity, although for (19) and (20) it is not definite that bekaf and strontvervelend would not have conveyed the same meaning.

(19) Ben nu alleen bekje af, echt heel moe

Am now though ‘worn out’, really very tired (28) (20) heeeeeeel moe dus heel erg strontjevervelend

very tired so very very poo-DIMannoying (2719)

(21) bloed-, ja, bloedje-rood…

blood-, yes, blood-DIMred (3942)

(22) propje-, echt propjevol

(36)

Further evidence is that being schathemeltjerijk ‘treasure heaven-DIM rich’ is used as the

extreme opposite of starving to death (for full context see Appendix 2:5071) and a very ill child with a high temperature is described as gloedjeheet ‘glow-DIMhot’ (Appendix 2:4541).

This last example also shows that language users are not just playing with language when they use this form, but that it can be used in a serious context.

There is no evidence that the insertion of -tje ‘weakens’ the intensifying meaning of the compound. A strengthening intensifying meaning for this type of word formation implies a strong demotivation of the diminutive suffix, but this is not unique. As was pointed out in the background section (cf. 2.1), there are other cases in Dutch where diminutives are used without any semantic link to the original meaning ‘small’. Whether a stronger intensifier meaning is more widely intended than in the examples above could perhaps be further investigated in an experiment-based analysis to see whether there is a consensus in the interpretation of the level of intensity.

The bulk of the examples are from forums, weblogs or social media (such as Twitter and Facebook). This constitutes informal language use, verging on spoken language. Other data which is not from these online media also has an informal nature. The kind of sources concerned are children’s pages or columns in newspapers, children’s books, advertisements and dialogue in novels. There are also 20 attestations in (mostly unpublished) poems, where the poet has some artistic licence when it comes to creative language use and in which the diminutive ending might come in handy to make a compound fit the metre. Intensifying diminutive compounds are found to a small extent in more formal sources or situations: the acknowledgements in a PhD thesis, the narrative of novels, debates in parliament and newspaper articles in De Telegraaf, HP/De Tijd, Trouw and De Volkskrant.20 _{These more}

formal situations are very few in number and for many of them a sense of informality could be argued. This type of word formation thus appears to be restricted mainly to informal use. I would like to suggest that this restriction does not mean that diminutive endings only create an informal variant of the intensifying compounds, but that the added, more emphatic meaning plays a role. If these forms have stronger intensifying meaning, as I have argued above, they may be used mainly in informal language, because such strong emotional intensification is not called for in formal usage. In addition this type of word

(37)

formation is very marginal when it comes to frequency, which is another reason it would not occur in formal more standardised use.

4.7 Demographic factors 4.7.1 Gender

The gender of the language user is known in 87 percent of cases, from the names or online profiles of the users. Of the total number of 3726 sentences, 1540 are written by women and 1685 by men. Despite the 501 cases where the gender of the language user is unknown, it is clear that intensifying diminutive compounds are used by both men and woman in equal measure. Any differences between men and women are lexical differences to do with the context and the right constituent of the compound. Men will often use words such as aaltjeglad ‘eel-DIMsmooth’ or hageltjenieuw ‘hail -DIMnew’ to describe their cars or gadgets,

while women use knalletjerood ‘bang-DIMred’ for clothes, piepjeklein ‘squeak-DIMsmall’ for

babies and graatjemager ‘fishbone-DIM skinny’ for horses, dogs and celebrities. It is also

women who tend to find things doodjeseng ‘dead-DIM scary’ or be aapjestrots ‘monkey-DIM

proud’ of each other, and their friends and children are either supertjelief ‘super-DIMsweet’

or strontjevervelend ‘shit-DIM annoying’. Intensifying diminutive compounds with the right

constituent hard ‘hard, loud, fast’ are usually used by men (bikkeltjehard, spijkertjehard, bloedjehard, knettertjehard). I could go on, but these examples show that the right constituent is more relevant than the use of the diminutive when it comes to the differences between the language use of men and women. So while language users of different genders have different interests and worries, they do not use different morphological elements to express them.

4.7.2 Age

(38)

the construction in the 1990s, and the largest group of known users from that time is between 31 and 45 years old.

Table 4.5 Number of language users per age group

The youngest language user in the data is a little girl of around five years old, who wants an ice cream and is quoted by her mother as having said (23). There are two more cases of very young children, quotes by their (grand)parents, another five-year-old (knettertjesgek ‘crackle-DIM mad’) and a 7-year-old (stokjestijf ‘stick-DIM stiff’). Another ten children aged

between ten and twelve used intensifying diminutive compounds online themselves. This shows that this type of word formation also occurs among the youngest age group, who are underrepresented on the internet. If young children are familiar with it, it is quite likely that they make their own new formations, which make no sense as similes, just as they make new intensifying compounds (Hoeksema 2012: 104; Vooys 1916). The oldest user of the construction is a 73-year-old man from Limburg who writes (24). The eldest age group (65+) is also underrepresented online, which makes it all the more noteworthy that data from this group can be found. There can be no doubt that this form is used in private or real-life communication as well as online.

(23) Ik heb het zoooooo warm, echt bloedjewarm I have it soooooo warm, really blood-DIMwarm

‘I’m so hot, really very hot’ (4869)

(24) Glim Glim en spiegeltjesglad Shine shine and mirror-DIMsmooth

‘Shiny, shiny and very smooth’ (3187)

Age group Language users