DynaSAND

Syntactic Atlas of Dutch Dialects – Commentary – Volume I

Table of contents Volume I

2  Subject pronouns

2.1        Introduction

2.1.1  Personal subject pronouns

2.1.1.1  General introduction

2.1.1.2  Strong pronouns

2.1.1.3  Weak pronouns

2.1.1.4  Pro-drop

2.1.2  Expletives: er and het

2.2       Discussion of the literature

2.2.1  Personal pronouns

2.2.2  Expletives

2.3        Discussion of the maps: personal pronouns

2.3.1  Subject pronouns first person singular

2.3.2  Subject pronouns second person singular

2.3.3  Subject pronouns third person singular masculine

2.3.4  Subject pronouns third person singular feminine

2.3.5  Subject pronouns first person plural

2.3.6  Subject pronouns second person plural

            2.3.7  Subject pronouns third person plural

            2.3.8  Overview maps on personal pronouns

2.4       Discussion of the maps: expletive er and het

            2.4.1  Expletive subject in regular order in existential clause

            2.4.2  Expletive subject in inversion in existential clause

            2.4.3  Expletive subject following complementiser in existential clause

            2.4.4  Expletive subject in inversion in impersonal construction

2.5       Literature on subject pronouns and expletives

2  SUBJECT PRONOUNS

2.1  Introduction

2.1.1  Personal subject pronouns

2.1.1.1  General introduction

Morphological variation within personal pronouns is a well-studied issue in Dutch dialect syntax. One of the causes of such morphological variation is the application of sound laws, whose effects are clearly visible in the pronominal paradigm. For instance, the Standard Dutch pronoun zij occurs in all dialects, but is sometimes pronounced as /zie/ or /zai/. Such phonological variation needs to be described by means of phonological dialect atlasses, not by a syntactic atlas such as SAND. However, there is morphological variation that cannot be explained as a result of sound laws, and which is indeed relevant to the study of Dutch dialect syntax. Some dialects have subject pronouns that elsewhere can only be used to refer to objects (1a–b). Other dialects have subject pronouns that are completely unknown to Standard Dutch and to most other dialects (1c–e).

(1)       a.         Hun     spreken           zo        raar.

                       Them speak              so        strangely

                       ‘They speak very stangely.’

            b.         Is–’m dood?

                       Is him  dead

                       ‘Is he dead?’

            c.         Zijlie    zijn      daar     nog nooit        geweest.

                       They   are       there    never               been

                       ‘They have never been there.’

            d.         Du       weetst daar     niets                van.

                       You     know   there    nothing           from

                       ‘You do not know anything about that.’

            e.         Ga–de                         mee     naar de markt?

                       Go you           with    to the market

                       ‘Are you going to the market with me?’

It is important to distinguish weak pronouns and strong ones when addressing morphological variation in subject pronouns (compare (1b,e) to (1a,c,d)). The distinction between weak and strong pronouns is relevant to Standard Dutch as well. In Standard Dutch, the use of weak or strong pronouns is a rather complex issue. The distribution of weak and strong forms is partly a matter of free variation, and partly a matter of prosodic, pragmatic and syntactic factors. One of the syntactic factors is the environment in which weak pronouns may occur. In Standard Dutch, weak pronouns can only occur in a few positions, such as immediately preceding or following finite verbs, or immediately following complementisers.

In Standard Dutch, weak pronouns are, in general, clear morphological variants of strong pronouns: weak pronouns, such as ze, have schwa as a vowel, whereas their strong counterparts, such as zij, have a strong vowel, which can be stressed. Thus, the relation between a weak and a strong pronoun is mostly a transparent one. There are two exceptions to this in Standard Dutch, where the relation between a weak and a strong pronoun is more complex and where a certain degree of asymmetry is observed in the pronominal paradigm. Firstly, in the second person plural only a strong pronoun (jullie) exists. As a weak pronoun, the pronoun je is used, which is the weak form of the second person singular, and hence triggers singular inflection on the verb. Secondly, the weak form of the third person masculine pronoun hij (i.e., ie) does not show schwa. In addition, this weak pronoun ie cannot be used in sentence-initial position, where the strong pronoun hij is the only option. In dialects, which show a larger degree of morphological variation, asymmetries are far more common than in Standard Dutch. For instance, all dialects have the weak form ze in third person plural, but some of them have the object pronoun hun or the dialectal pronoun zijlie as a strong counterpart (see (1a,c)). In those dialects, weak ze cannot be recognised as a merely ‘weakened’ form of the strong pronoun, and both pronouns are clearly not phonological variants.

1. (2)       weak vs. strong pronouns in third person plural

ze        is a phonological variant of               zij        (e.g., Standard Dutch)

ze        is not a phonological variant of         zijlie / hun

The distinction between weak and strong pronouns is a basic parameter which is equally important for the study of pronouns in Dutch dialects as are features like case, person and gender. Therefore, weak and strong pronouns are to be kept apart when discussing morphological variation in pronouns. Both strong and weak pronouns show morphological variation. Strong pronouns will be discussed in section 1.2, weak ones in section 1.3.

A rather special case of morphological variation in subject pronouns is observed in the second person singular, where the subject pronoun can be omitted in some dialects. This phenomenon is called ‘pro-drop’, and is exemplified in (3).

1.  
2. (3)       (Du)    komst moarn.

            (you)   come   tomorrow

Pro-drop is attested only in dialects with the pronoun du and the inflectional affix –st in second person singular. Pro-drop will be discussed in section 1.4.

2.1.1.2  Strong pronouns

Strong subject pronouns are shown on the following maps: map 38b (2sg.), 42a (3sg. masculine), 43c (3sg.feminine), 44a/b (1pl.), 46a/b (2pl.), and 47a/b (3pl.). Most of the morphological variation in strong subject pronouns is caused by three linguistic phenomena with ramifications throughout the entire pronominal paradigm. All three have been discussed extensively in the linguistic literature.

First, some dialects have at their disposal a plural pronoun that is originally a compound of a simplex personal pronoun and the noun lieden or lui. Although these ‘lieden-compounds’ are diachronically complex forms, in synchronic terms they must be analysed as simplex forms. Examples include the Standard Dutch pronoun jullie, and dialectal forms such as wullie, wulder, ... in first person plural; jelui, gullie, gulder, ... in second person plural; and zullie, zulder, ... in third person plural. Lieden-compounds are attested for first, second and third person plural, but they are notoriously widespread in second person plural and to a lesser extent in third person plural (similarly, lieden-compounds in Afrikaans only occur in second and third person plural). The existence of lieden-compounds has a large impact on the relation between strong and weak pronouns, since weak pronouns such as we and ze can easily be considered phonological variants of the strong pronouns wij and zij, but not as variants of, for instance, strong wullie and zullie. An overview of the lieden-compounds in the SAND corpus is provided on map 48b.

Secondly, a number of object forms have entered the subject pronoun paradigm (e.g., 3sg. feminine haar, 1pl. ons in Zeeland, 3pl. hun in Holland). The use of object forms in subject position is a well-known phenomenon in a number of western European languages such as French and English. In French and English, the phenomenon seems to affect strong pronouns rather than weak ones. An example from French could be the left-dislocated sentence Moi, je vais gagner; English examples include (dialectal) sentences such as We do not know, do us. All Dutch object pronouns occuring in subject position are strong pronouns as well, with one exception, namely the weak object pronoun for third person masculine singular ‘n / ‘m, which can be used in subject position in a number of southern Dutch dialects. Map 48a provides an overview of Dutch object pronouns in subject position.

A third phenomenon is the replacement of the Middle Dutch second person singular du by the original, second person plural pronoun ghi (and its formal variants). This replacement is complete in some areas where du has disappeared completely, and incomplete or has not even started in other areas where some dialects have only du or both du and another pronoun at their disposal as second person singular forms. The possible replacement of du by ghi or another pronoun is visible on maps 38b, 39a/b, 40a/b. Map 41b, which shows polite pronouns or ‘honorifics’, is instructive as well.

2.1.1.3  Weak pronouns

In sentences, adjacent constituents may affect each other’s phonetic form, by means of a number of phonetic processes (e.g., assimilation). Since weak pronouns only occur in the immediate proximity of finite verbs or complementisers, their form is easily influenced by the form of the finite verb or the complementiser, especially because combinations of these elements are highly frequent elements, in which ideosyncrasies easily develop. In the linguistic literature, the combination of a finite verb and a following pronoun has been labelled an ‘enclitic cluster’. Enclitic clusters, in which inflectional affixes are immediately followed by weak pronouns, often give rise to reanalyses affecting both the verbal and the pronominal morphology.

As it is sometimes hard to determine the precise form of the components that are present in a particular enclitic cluster, analysing an enclitic cluster may be problematic. In some cases, new pronouns have been formed, which are clearly influenced by the presence of an inflectional affix. The influence of the inflectional affix may be visible in two different ways: on the one hand, the pronoun can assimilate to the inflectional affix on the verb, such as 1pl. me, which is formed through assimilation of we to an infectional –n (compare 4a and b). On the other hand, sometimes the inflectional affix is analysed as a part of the pronoun. This is exemplified by means of (4d), in which the pronoun –tie is formed through fusion of an inflectional –t and the weak third person masculine pronoun ie.

(4)       a.         Gaan   we       dat       wel                  kunnen?

                       Go.1pl            we       that     affirm                        can.inf

                       ‘Can we do that?’

b.         Ga–me            dat       wel                  kunnen?

            go–we that                 affirm            can

            ‘Can we do that?’

            c.         Is         hij        al                     wakker?

                       is         he        already                        awake

                       ‘Is he already awake?’

d.         Is–tie   al                     wakker?

                       is–he   already                        awake

                       ‘Is he already awake?’

Sometimes, new pronouns that have been formed in an enclitic position do not merely follow the verb, but appear in other positions within the clause as well. They may occur following complementisers (5a) or preceding the verb (5b). This is not always the case, however: many dialects use the pronouns that have been formed in an enclitic position only in such an enclitic position. Consequently, in those dialects the selection of the subject pronoun in a particular sentence becomes partly a matter of word order: new pronouns originate following the verb, where pronouns are affected by the presence of an agreement marker on the verb. These are labelled ‘specifically enclitic pronouns’. In other syntactic environments, for instance when a weak pronoun precedes a verb, follows a complementiser or is attached to ja ‘yes’ or neen ‘no’, these specifically enclitic pronouns are usually not used, and the ‘original’ pronoun is kept. However, the linguistic literature provides examples of agreement on both complementisers and ja ‘yes’ and neen ‘no’ (see chapter 1 and 3.2), giving rise to possible reanalyses and assimilations in those syntactic environments as well. Hence it is not clear whether the form of the pronoun in proclitic position, following complementisers or attached to ja ‘yes’ or neen ‘no’ is necessarily the same in all dialects.

(5)       a.         As–me            niet      oppassen,       loopt   het       slecht af.

                       If–we              not      watch out,      run      it         badly   off

                       ‘If we do not watch out, it will end badly.’

b.         Me       gaan     moeten            oppassen.

                       We      go        must               watch out

                       ‘We will have to watch out.’

The rise of new weak pronouns in some syntactic environments may correlate with other linguistic phenomena. As has already been said, in some dialects, the pronouns that have originated in an enclitic position are found in other syntactic positions as well, sometimes causing the original pronoun to disappear. This gives rise to further complications: contrary to ‘ordinary’ weak pronouns (e.g., 1pl. we), weak pronouns that have been formed in an enclitic position (e.g., 1pl. me) can hardly be considered formal variants of strong pronouns (e.g., 1pl. wij). Hence the relation between strong and weak pronouns becomes less transparent.

The linguistic literature has payed quite a lot of attention to morphological variation in weak subject pronouns in general, and to enclitic clusters in particular. So far, the second person and the third person masculine singular, especially, have drawn a lot of attention (see 2.2.1). The relevant maps for weak subject pronouns are map 38a (1sg.), map 39a/b, 40a/b (2sg.), map 42b and 43a (3sg. masculine) and map 45a/b (1pl.).

2.1.1.4  Pro-drop

Dialects that have kept the second person singular pronoun du, have kept the agreement marker –st as well. Some of these dialects, especially those in the northeast of the Dutch language area (e.g., Frisian dialects among others), allow the pronoun du following the agreement marker –st to be not only reduced to weak de, but also to be dropped (reduction to zero), giving rise to a so-called pro-drop construction. Hence sentences occur in which no subject pronoun is present at all.

1. Pro-drop in Frisian

            a.    Komst           moarn?

                   come.2sg       tomorrow

                   ‘Are you coming tomorrow?’

            b.    Komst           moarn.                                 

                   come.2sg       tomorrow

                   ‘You are coming tomorrow’

            c.    …datst          moarn           komme   soeste.

                   ... that.2sg     tomorrow     come      should.2sg

                   ‘... that you should come tomorrow.’

Pro-drop does not occur in Standard Dutch or in any of the other West Germanic standard languages, and it is rare both in Dutch and in other West Germanic dialects. In addition, pro-drop is only attested in the second person singular. Some linguists have claimed that sentences such as (6) do not constitute valid examples of pro-drop, but rather show an extremely reduced pronunciation of the subject pronoun. Especially in sentences such as (6a,c), this interpretation does not seem far-fetched, since enclitic clusters in general tend to be pronounced as reduced forms. The fact that (6b) can have a declarative reading, however, proves that it is unmistakably an example of pro-drop. Given the declarative reading of (6b), there is no enclitic cluster present in the sentence since one would expect overt subjects to occupy the first, preverbal position in the sentence rather than to cliticise to the (postverbal) agreement marker –st. Hence pro-drop in (6b) cannot be attributed to phonological reduction in enclitic clusters. In some dialects, the agreement marker seems to be –ste rather than –st, as –ste occurs in a position where enclitic clusters never appear, namely following an inflected verb in a subclause (e.g., the verb soeste in (6c)). The existence of an agreement marker –ste in some dialects constitutes an argument to consider attestations of –ste to be agreement markers in stead of enclitic clusters. If one analyses –ste as an agreement marker, the number of possible instances of pro-drop increases dramatically. Besides clauses with inverted word order and an inflected verb on –st, and subclauses starting with an inflected complementiser on –st, all clauses with a verb or a complementiser followed by –ste (7a,b) may prove to be valid examples of pro-drop, at least in those dialects that have –ste on sentence-final verbs.

(7)       a.         Komste           moarn?                                  

come.2sg        tomorrow

‘Do you come tomorrow?’

b.         … datste         moarn             komme            soeste.

            ... that.2sg       tomorrow       come               should.2sg

            ‘... that you should come tomorrow.’

The agreement marker –st does not occur exclusively in the north and the east of the Netherlands, but in (parts of) Belgian and Dutch Limburg as well. In Limburg, however, there is no reason to assume the existence of pro-drop, since the agreement marker –s(t) is always followed by –e, and an agreement marker –ste in subclauses is not attested.

An overview of pro-drop in Dutch dialects is provided on map 41a.

2.1.2  Expletives: er and het

The subject position in Dutch is, of course, not always filled by a personal pronoun. In some cases, a seemingly meaningless element occupies the subject position, like er or het. In Standard Dutch, the element er is used in four different ways:

(8)       a.         Locative                                 Hij is er.

                                                                      he is there

                                                                      ‘He is there.’

            b.         Quantitative / partitive          Ik heb er twee.

                                                                      I have there two

                                                                      ‘I have two of them.’

c.         Prepositional                         Heb     je         er        over    nagedacht?

                                                          have    you     there    about   thought

                                                          ‘Have you thought about it?’

            d.         Expletive                                Er        zijn      vier      soorten           ‘er’.

                                                                      there    are       four     kinds   ‘there’

                                                                      ‘There are four kinds of ‘there’.’

                                                                       Gisteren     zaten          er     vier mannen    in het café.

                                                                       yesterday   sat     there         four men         in the pub.

                                                                       ‘Yesterday, four men were sitting in the pub.’

It is the use of er in (4d) that is relevant here: er is not the ‘real’ subject of the sentence, but it is clear that er occupies a position that is normally only occupied by subjects. It is not clear how the expletive use of er originated. The fact that expletive er can often be replaced by d’r ‘there’, could indicate that er, like d’r, is a weak variant of daar ‘there’. However, in contemporary Standard Dutch, there are quite a few examples of sentences in which er cannot be replaced by daar.

(9)       a.         Daar   blijkt   een ongeluk     te zijn gebeurd.

                       there    seems  an accident      to be    happened.

                       ‘An accident seems to have happened there.’

            b.         Er / *daar       blijkt   hier      een ongeluk     te zijn gebeurd.

                       there               seems  here     an accident      to be    happened

                       ‘An accident seems to have happened here.’

(9a,b) shows that daar ‘there’ is interpreted as a locative element in Standard Dutch. In some dialects, however, daar is used as an expletive, yielding the possibility of replacing er by daar in (9b), which is not possible in Standard Dutch. The linguistic literature provides more examples of variation. Some dialects in Limburg, for instance, are known to use et ‘it’ as an expletive element in sentence-initial position (10). Elsewhere in the sentence, the expletive is absent (11).

(10)                 Limburg:                     Et        is         nog      soep.

                       Standard Dutch:         Er        is         nog      soep.

                                                          it/there            is         still      soup

                                                          ‘There is still soup left.’

(11)                 Limburgs:     Is           nog    soep?

                        Standaard:     Is er           nog    soep?

                                             is  there still    soup

                                             ‘Is there still soup left?’

The situation in Limburg is comparable to the situation in German, where expletives are only used in sentence-initial position, and to the situation in older Dutch. The data in Limburg also suggest that the use of expletives correlates with the linguistic constraint that stipulates that the first position in the sentence be filled by any other constituent than the finite verb (i.e., the ‘Verb Second’-constraint). In the theoretical linguistic literature, the ‘Verb Second’-constraint is often assumed to correlate with another property of Dutch syntax, namely the fact that Dutch clauses must have overt subjects (12). In (12), the verb regenen ‘to rain’, which semantically does not need to have a subject, is preceded by the expletive it.

(12)     Het      regent dikwijls           in         Frankrijk.

            it         rains    often               in         France

            ‘It often rains in France.’

In some dialects, constructions such as (12) show morphological variation comparable to the variation that is encountered for expletive er, where d’r and daar are attested as dialectal variants. As an expletive in sentences such as (12), some dialects use strong dat ‘that’ instead of weak het ‘it’.

2.2  Discussion of the literature

2.2.1  Personal pronouns

The pronominal paradigm of Dutch, Frisian and their dialects is discussed extensively in the linguistic literature. For Standard Dutch, see the ANS-grammar (section 5.2), where both the form of the pronouns and the rules for the use of weak and strong pronouns are given. For Standard Frisian, see chapter 4 in Tiersma (1999:55–59). Howe (1996) places both Dutch and Frisian pronouns in a West-Germanic perspective, and pays some attention to dialectal variation as well. The most elaborate description of dialectal variation in pronouns is undoubtedly De Schutter (1989a), which contains maps depicting the morphological variation for a number of weak pronouns (the maps, however, do not contain Frisian data). Examples of pronominal paradigms in particular dialects can be found in abundance, in virtually every dialect monography (see the online SAND bibliography for an overview). Linguistic publications with pronominal paradigms for particular dialects include Haegeman (1992) (dialect of Lapscheure in West-Vlaanderen), Swiggers (1987) (dialect of Leuven in Vlaams-Brabant) and De Schutter (1992) (dialect of the city of Antwerpen). There are also numerous publications that deal with one or more pronouns. These will be summarised briefly below.

Hoeksema (2000) and Zwart (2002) deal with the strong first person singular form ikke.

The second person singular has been discussed extensively. The Standard-Dutch system differs from the one in Standard Frisian. Standard Dutch has strong jij and weak je, and a polite form u (ANS 5.2.4). Standard Frisian has du (weak form: de), and a polite pronoun jo, that triggers plural agreement on the verb (Tiersma 1999:55–57). The linguistic literature provides a rather detailed picture of the attested variation in the dialects. As early in the 1920s, the situation in the Dutch provinces Groningen and Drenthe was described (and mapped) by Kloeke (1920). Dialectal variation in the Belgian province of Limburg is described and mapped by Stevens (1949). The conditions determining the use of the different forms in Limburg are also discussed by Vossen (1963). Almost all attested pronouns have been subject to scrutiny in the linguistic literature, and descriptions often include geographic information as well: the geographic distribution of gij is discussed by Hol (1936). Since Verdenius (1923), second person pronouns with an initial /j/ are believed to derive from a palatalisation of enclitic –di, a formal variant of Middle Dutch ghi. Gysseling (1966), however, believes that these pronouns originated through fusion of ghi itself and the second person agreement marker –t. Devos (1989) provides geographic data for West-Vlaanderen and Oost-Vlaanderen, including maps, supporting Verdenius’ view. Both Willemyns (1979) and Vandekerckhove (1993) discuss second person pronouns in a dialect in West-Vlaanderen (the article by Willemyns includes maps). For discussion of the eastern vocalic pronouns ie and ij, see the previously mentioned Kloeke (1920) and Verdenius (1926), who explains these pronouns as deriving from ghi, like jij. De Wilde-Van Buul (1942) mentions occurences of ie in the city of Rotterdam. In Brabant dialects, a pronoun –de occurs in an enclitic position. Schuurmans (1975:67) analyses –de as a combination of an inflectional –t and the pronoun –e. De Schutter (1997:37–38) agrees. De Schutter & Taeldeman (1986:122) and Stroop (1987:127–131), however, provide an alternative account, analysing –de as a pronoun in its own right. Stroop believes –de derives from Middle Dutch du. A final pronoun, d’r, is dealt with by Van Den Berghe (1951), who explains the existence of d’r as the result of a contamination of Middle Franconian ir and Lower Franconian gi.

The Standard Dutch polite pronoun u occupies a special position in the literature on Dutch second person pronouns. At least nine different etymologies have been proposed (Kuijper 1971–1972). All nine consider u to be a former object pronoun or possessive form that has entered the subject pronoun paradigm. Some of these etymologies state that the use of u traces back to the abbreviation U.E. (= Uwe Edelheid ‘Your Honour’). See also Bennis (2004).

The second person singular is the only grammatical person allowing pro-drop, namely in some dialects in the northernmost provinces Friesland and Groningen. The relevant dialects behave more or less like Standard Frisian. The presence of pro-drop in Frisian is controversial: Van der Meer (1991), Howe (1996:184) and Tiersma (1999:56) among others provide an alternative account for the pro-drop analyses by Visser (1988) and De Haan (1994, 1997) among others, analysing the Standard Frisian examples of pro-drop as phonologically reduced, specifically enclitic pronouns.

According to Kooiman (1950) and Hol (1951), the third person masculine singular shows a lot of morphological variation. One dialectal form is the weak object pronoun ‘m, which can be used in subject position in some dialects. The use of ‘m as a subject is even mentioned by the ANS-grammar (which considers the form to be a regional variant). Other variants include a number of pronouns beginning with a dental stop (e.g., tie, t’m, t’n), which have been accounted for in different ways. Van den Toorn (1959) thinks these pronouns are originally combinations of an agreement-t and a weak pronoun (ie, ‘m, ‘n), which have been reanalysed in an enclitic position. The resemblance between the forms with an initial dental stop and demonstrative pronouns may have facilitated this reanalysis. According to De Wilde-Van Buul (1942), 3sg.masculine die and tie originated as a strategy to avoid homophony with the (dialectal) second person singular pronoun ie. Pauwels (1958:336–337) and Schuurmans (1975) among others take the dental stop in the former object pronouns t’n and t’m to be evidence for the statement that these forms actually derive from demonstratives rather than from personal pronouns. Van Loey (1954), Koelmans (1968) and Stroop (1987) argue against this opinion. Koelmans (1968) includes a map showing the geographic distribution of [t]’n. The strong third person masculine pronoun has received less attention in the linguistic literature than the weak one. The use of the strong object pronoun hem as a subject is mentioned by Karsten (1939) and recently in Devos & Vandeweghe (2002–2003). Swiggers (1990) signals the existence of the pronoun emmekes.

Except for the relevant sections in the previously mentioned books by De Schutter (1989a) and Howe (1996), third person feminine pronouns are by and large neglected in the linguistic literature as the morphological variation is quite limited. Most dialects have zij (or a phonological variant), like Standard Dutch; Frisian has two pronouns: hja and sy. Ryckeboer (1972) attests relics of the old demonstrative zoe in Frans-Vlaanderen. According to Weijnen (1966:299), some dialects use masculine or even neuter forms to refer to feminine antecedents (see Weijnen 1971:27 for a map).

Apart from minor differences in the appearance of the pronoun in enclitic clusters, there is almost no morphological variation in the third person neuter pronoun: all dialects have het (weak ‘t); the Frisian spelling is it. Hence, the third person singular neuter pronouns have not been an issue in the dialectological literature, and they will not be dealt with in this atlas.

With respect to the first person plural, some dialects in Frans Vlaanderen and in the Dutch province of Zeeland can use the strong object pronoun ons as a subject (see Paardekooper 1969 for a map). De Schutter (1994:118–119) and De Vogelaer & Neuckermans (2002:253–255) discuss the specifically enclitic weak pronoun m’n. The latter article contains a map.

The second person plural pronoun in Standard Dutch is jullie. There are different views about the origin of jullie. According to Kloeke (1941), jullie is formed out of the singular pronoun and the noun lui ‘men’, whereas Van Loey (1948) states that the added noun is lieden (‘men’). As compounds of an ‘old’ pronoun and lui or lieden are attested in the first and the third person plural as well, the discussion is also relevant for the first person plural and the third person plural pronouns. In some dialects, the simplex pronouns and lieden (‘men’) do not seem to occur, suggesting that the original compounds must be analysed as simplex forms from a synchronic point of view. In other dialects, agreement phenomena may be indicating that we are indeed still dealing with complex pronouns (Van Koppen 2005). The Standard Frisian pronoun is jimme (also: jim). Jim(me) has been claimed to be both a simplex form, the –me being an original affix marking dative case, and a complex one, the –me being a relic of the noun man ‘man’ (see Howe 1996:196–198 for discussion). There is some linguistic literature on dialectal pronouns. Van den Berghe (1951) describes and maps the second person plural pronouns in an enclitic position for Belgium only. Hol (1953) provides a map with the geographic distribution of the southern pronouns with an initial /g/ (e.g., gijlie). Willemyns (1979) describes the situation in the Belgian province of West-Vlaanderen and in Frans-Vlaanderen, suggesting that the western pronouns with initial /j/ are originally object pronouns (Willemyns 1979:183–184).

In most dialects, the third person plural pronoun is zij; sy or hja in Frisian. The use of object pronouns in subject position, such as hun and hullie, has drawn some attention in the literature, e.g., in Kooiman (1969) and De Rooij (1990), who provides two maps depicting the geographic distribution of hun and some other pronouns. The use of hun has also been studied from a sociolinguistic perspective (Van Hout 1996, 1999; Cornips 2000). De Schutter (1987) mentions the existence of zun, a contamination of zij/ze and hun.

Morphological variation in pronouns is attested in other languages than Dutch and Frisian. There too, the distinction between weak and strong pronouns seems to be an important one (see Zwicky 1977 among others), although the distinction is far from universal (Siewierska 1999). In pronominal paradigms, there are tendencies as to which grammatical persons are more likely to be referred to using the same pronoun (Cysouw 2003). In the linguistic literature, it is also noted that the rarity of pro-drop constructions in Dutch is exceptional (Dahl 1990, Siewierska 1999, Haspelmath 2001).

2.2.2  Expletives

The most common expletive is er ‘there’. There is a considerable amount of literature on the use of er. See the ANS-grammar (1997:464–489) for a description of er in Standard Dutch. The situation in Standard Frisian is dealt with in Hoekstra (1991). Theoretical issues and correlations with some linguistic phenomena are discussed in Bennis (1983, 1985, 1986). Koelmans (1972) signals some morphological variation: er alternates with d’r/t’r, and the distribution is at least partly a matter of phonology. The article contains diachronical data.

There is also literature on (expletive) er in the dialects. Grange & Haegeman (1989) discuss expletive er in West-Vlaanderen. Their description includes data on morphological variation in expletives. De Rooij (1991) provides data on the use of er in the dialects, and a map. In some dialects morphological variants for er are used. Koelmans (1975) maps the distribution of het and daar, using RND data (as gathered by Blancquaert & Pée 1925–1982). His data make Vanacker (1978) conclude that expletive ‘t derives from daar, via intermediate variants such as d’r and t’r. The most elaborate treatment of morphological variation in expletives is De Schutter (1989b). De Schutter provides a map using the same data as Koelmans, but with much more detailed information. The map is compared to dialect data from the 19^th century (the so-called ‘Willems-corpus’, see Goossens, Taeldeman & Weijnen 1989 for more information), leading De Schutter to a conclusion opposite to Vanacker’s: expletive daar has probably never been used in the southwest of the Dutch language area. Instead, in the southwest, expletive d’r may derive either from partitive d’r, either from expletive ‘t and partitive er. Recently, morphological variation in expletives has probably been subject to dialect levelling (see Ryckeboer 1995 for data from West-Vlaanderen).

2.3  Discussion of the maps: personal pronouns

2.3.1  Subject pronouns first person singular (map 38a) (map in dynaSAND)

Unlike most other grammatical persons, the dialect pronouns of the first person singular do not vary greatly in form. All the relevant morphological variation can be shown on one single map. Map 38a shows the form of the pronoun following a complementiser, which is representative for all syntactic environments in which first person pronouns occur. The map contains data for two SAND test sentences, both translation tasks, namely the sentences Als ik zuinig leef, leef ik zoals mijn ouders willen (‘If I live economically, I live like my parents want me to’) and ‘k Geloof dat ik groter ben dan hij (‘I believe I’m bigger than he’).

Since the morphological variation is quite limited, displaying some phonological variation on the map does not make it harder to interpret. Therefore, map 38a does show some phonological variation, unlike the other maps in this chapter. The largest part of the Dutch language area uses ik as a first person pronoun (or the weak variant (e)k). The southeast has ich (or a weak variant). The border between these two areas matches the so-called Uerdinger line, which indicates that the distribution of ik and ich is completely determined by phonological parameters. Also, the SAND corpus does not contain any examples of weak pronouns that are not simple reductions of the full pronouns ik and ich. However, in some dialects in the southwest, a schwa may appear on the weak pronoun, resulting in a pronoun ke. Map 38a shows four attestations of ke, but the pronoun is probably more widespread, and is possibly known througout the Belgian province of West-Vlaanderen.

In addition, most dialects in the language area probably have a strong form ikke at their disposal which is mainly used in an isolated position and can be considered as a stronger form of ik. However, the geographic distribution of ikke and possible constraints on the syntactic environments in which this pronoun may occur, are not investigated by the SAND project.

2.3.2  Subject pronouns second person singular

2.3.2.1  Subject pronouns 2 singular, strong forms (map 38b) (map in dynaSAND)

Second person singular pronouns vary greatly. A wide range of different forms is attested, and the relation between weak and strong pronouns is often not transparent. One of the causes of this high degree of variation is the fact that the original pronoun for the second person singular, du, has been replaced by the former, second person plural pronoun in large parts of the Dutch language area. The attested forms thus derive from different pronouns.

The SAND corpus contains quite a few second person singular forms. The following test sentences are used for map 38b: the paradigm of gaan (‘to go’), Als je gezond leeft, leef je langer (‘If you live healthily, you will live longer’) and Ze gelooft dat jij eerder thuis bent dan ik (‘She believes that you will arrive home sooner than I’). All three are sentences which have been presented to the informants as translation tasks during the SAND interview. There are, however, some sampling points at which all the answers to these questions only contain weak pronouns. In such cases, the translations for the sentence Je gelooft zeker niet dat hij sterker is dan jij (‘Surely you do not believe that he is stronger than you’) have been used. Since some dialects use object pronouns rather that subject pronouns in comparative subclauses, this sentence too did not provide suitable answers for all sampling points. Not all pronouns that are found in the SAND corpus are mapped. When both du and another second person pronoun are attested in a certain dialect, the latter is considered a polite form, and only du is mapped. Possible alternatives to du occuring in these dialects are dealt with in section 2.3.2.4. 

At present, the original second person singular pronoun, du (also dou), has a limited geographic distribution: it is used in the southeast (Dutch Limburg), the east (Twente) and the north of the language area (in Friesland, Groningen and, rarely, in Drenthe). At some sampling points in Dutch Limburg and especially in the east of Belgian Limburg, the originally accusative pronoun dich is attested.

All other pronouns on map 38b derive from Middle Dutch ghi, the second person plural pronoun. The pronouns that derive from ghi show a lot of morphological variation, especially when it comes to the initial consonant. A large, southern area, including the dialect areas Vlaanderen and Brabant, has pronouns with an initial /g/ (gie, gij). Southeast to this area is a smaller area with pronouns with an initial /dj/ (e.g. djie, djij) comprising a number of places in Belgian Limburg and Walshoutem, in Vlaams-Brabant. Dialects in the east of the Netherlands (the provinces Gelderland and Overijssel) mostly use ie (and variants). The same form is attested systematically in Drenthe and, sporadically, in Friesland, Utrecht and Noord-Brabant. There is even one attestation in Dutch Limburg as well. The west of the Netherlands (Noord-Holland, Zuid-Holland, Zeeland, Utrecht) has a pronoun with an initial /j/ (jij, jie), like Standard Dutch, or an originally accusative form like jo, jou or joe. The wide distribution of jij and jie is probably at least partly caused by influence of the standard language. A very isolated example of jij/jie is found in Tongeren (Belgian Limburg). Object forms like jo, jou and joe are also attested in the northern provinces Friesland and Drenthe, and on the (former) island, Urk.

Some of these pronouns have been mapped before. The northern du-area has been described by Kloeke (1920), and seems not to have changed in size. This holds for the southern du-area in Limburg as well: map 38b shows that the current geographic distribution is comparable to the situation as described by Stevens (1949). The northern border of the gij-area, which has been mapped by Hol (1936) hasn’t changed either, although the standard Dutch pronoun jij seems to infiltrate the gij-area, since some informants for dialects in the gij-area make use of both gij and jij.

The morphological variation in the initial consonant of the second person singular pronouns, is traditionally explained by the assumption that all pronouns that do not display an initial /g/, have been formed through reanalysis of the form of the pronoun in an enclitic cluster (see 2.3.2.2). In dialects in Vlaanderen and Brabant, this reanalysis has not occured, since strong pronouns never appear in an enclitic cluster: (second person) strong pronouns and finite verbs in Brabant and Vlaanderen are usually separated by a pronominal clitic (see chapter 3, on subject doubling), that does indeed show the effects of reanalyses which have happened in enclitic clusters.

2.3.2.2  Subject pronouns 2 singular, weak forms

2.3.2.2.1  Subject pronouns 2 singular, weak forms in initial position(map 39a) (map in dynaSAND)

Weak second person singular pronouns show the same amount of variation as the strong ones. In large parts of the Dutch language area, more than one weak form is attested. In such cases, the choice for a particular pronoun is usually determined by the syntactic environment in which the pronoun is found. There are four relevant syntactic positions: sentence-initial (or proclitic) position (preceding the finite verb), and three enclitic positions, i.e., following (most) verbs with a stem that ends on a consonant (e.g., leven ‘to live’), following verbs with a stem ending in /d/, /t/ or a vowel (e.g., gaan ‘to go’) and following complementisers. In fact, the distinction between the two verb classes is more complex than suggested here: verbs with a stem on /n/, /r/, /l/, /s/ and /z/ sometimes behave like verbs with a stem on a consonant, and sometimes like verbs with a stem on a vowel. Possibly, there are regional differences in the behaviour of some of these ‘intermediate’ verbs.

Map 39a shows the pronouns that are found in sentence-initial position, as attested in the SAND sentences Je gelooft zeker niet dat hij sterker is dan jij (‘Surely you do not believe that he’s stronger than you’, a question asked during the fieldwork interviews) and Je weet wel dat je slim genoeg bent (‘You do know that you are smart enough’, a question asked during the telephone interviews).

Turning to the dialects that have kept the original Middle Dutch pronoun du, it becomes evident that, in the case of the southern du-dialects, most of them have de as a weak pronoun or, in the case of the eastern and northern du-dialects in Friesland, Groningen, Drenthe and Overijssel, the pronoun is simply omitted by using a so-called pro-drop construction (instances of ‘pro’ on the map). Both the attestations of de and pro-drop do not constitute a massive area on the map, since a lot of dialects seem to use only strong pronouns in second person singular. There are probably some constraints on the use of weak pronouns in second person singular in these dialects.

As for the dialects with second person singular pronouns that derive from Middle Dutch ghi, the attested weak, sentence-initial forms are very comparable to the attested strong forms. The geographic distribution of the weak forms is very different though: only dje is attested in more or less the same area as its strong counterparts djie and djij. By contrast, weak je has a much wider distribution than its strong counterparts jij and jie, since it is used in a large part of the gij-area as well: Frans-Vlaanderen and the north and west of West-Vlaanderen have weak je but strong gij. In the area with strong ie, special weak forms are altogether rare.

Comparing map 39a to the older map in De Schutter (1989a:80), no differences catch the eye. Willemyns (1979) and Devos (1986) provide a more detailed map for West-Vlaanderen than map 39a, but the data hardly differ. Devos does find some very western attestations of ge that are lacking on the SAND map.

As said in section 2.3.2.1, the form with an initial /g/ (ge) is, among the descendants of ghi, the only one that has not been formed in an enclitic position. Similarly, sentence-initial de is obviously a very ‘normal’ reduction of strong du. The areas in which ge and de are attested neighbour one another. Hence it is remarkable that at a rather small number of sampling points in the southwest of Limburg in between these two areas, an originally enclitic pronoun (je, dje) is attested.

2.3.2.2.2  Subject pronouns 2 singular, weak forms in inversion, finite verb leef ‘to live’ (map 39b) (map in dynaSAND)

Weak pronouns following verbs with a stem ending on a consonant, like leven (‘to live’), are found in the test sentence Als je gezond leeft, leef je langer (‘If you live healthily, you will live longer’), a sentence that had to be translated during the SAND interview. Some of the pro-drop examples come from another test sentence (translate: Nu je klaar bent, mag je gaan ‘Since you are ready, you can go’).

For some sampling points, no weak pronoun is found following a verb with a stem on a consonant. For instance, this seems to be the case in a number of locations in Frans-Vlaanderen and West-Vlaanderen. The reason for this is that the relevant test sentence can be translated in these dialects without making use of inversion.

In the du-areas, four different enclitic clusters occur: st + pro, [st]de (/ste/), [s]e and [t]se. There are several possible analyses for [st]de, [s]e and [t]se, and it is possible that a different analysis is required for different sampling points in which a certain form is attested. For instance, [st]de can be analysed as the agreement marker –st and the pronoun de, as the agreement marker –st and the pronoun e, as –s and –te, or simply as one single pronominal element –ste that is not preceded by an agreement marker. Similarly, [s]e can be analysed as the agreement marker –s followed by the pronoun e, or, as a whole, as the pronoun se, and [t]se can be seen as –ts followed by the pronoun se or e, or, again as a whole, as the pronoun tse. Whatever the correct analysis may be, the pronouns that are found in these enclitic clusters are typical examples of so-called ‘specifically enclitic pronouns’, pronouns that are only used in an enclitic position. This is particularly clear in the case of [s]e and [t]se, where the pronoun is either e or se, both of which are pronouns that never occur in sentence-initial position. The geographic distribution of the variants is as follows: pro-drop (pro) is found in the northern provinces Friesland, Groningen and Drenthe. In Belgian and Dutch Limburg, [s]e and [st]de are the dominant pronouns. [t]se is a rare variant, and has originated possibly as a combination of se and the Standard Dutch inflectional –t.

In the areas where du has disappeared, the following forms occur: ge, de, dje, je, ie and d’r. Among these pronouns, ge is the only one that has not been formed in an enclitic position. The geographic distribution of ge is remarkable: the pronoun only occurs in the SAND interview for the town of Tienen (Vlaams-Brabant). A lot of sampling points at which ge is the sentence-initial pronoun, by and large those in the provinces Noord-Brabant, Antwerpen, Vlaams-Brabant and Oost-Vlaanderen, use de, a specifically enclitic pronoun. The geographic distribution of dje following leven is more or less the same as in sentence-initial position, apart from one isolated instance in the town of Kortrijk (West-Vlaanderen). Like dje, the distribution of enclitic je hardly differs from its distribution in sentence-initial position, although two remarks must be made. First, the eastern border of the je-area in West-Vlaanderen is situated more to the east on map 39b than on map 39a. Secondly, various sampling points in the Dutch part of the je-area have a weak pronoun ie. Both je and ie never combine with an inflectional –t, unlike the homophonous third person masculine pronouns je and ie, ensuring non-ambiguous reference at all times. A final attested pronoun is d’r, which occurs in Belgian and Dutch Limburg.

Older maps showing second person singular pronouns following a verb with a stem on a consonant, are provided by Willemyns (1979) and Devos (1986), who deal with West-Vlaanderen pronouns following durven (‘dare’). Their data are more detailed than the SAND data, and confirm the observation that je is more widespread in an enclitic position than in sentence-inital position. Their data also corroborate the isolated instance of dje in Kortrijk (West-Vlaanderen), since both Willemyns and Devos find a small dje-area in West-Vlaanderen, in the immediate surroundings of Kortrijk. Both also show a number of examples of de in West-Vlaanderen, all found close to the border to Oost-Vlaanderen.

A striking feature of map 39b is the large number of specifically enclitic pronouns that occur, such as [s]e and [t]se in Limburg and de in Brabant (and Oost-Vlaanderen) dialects. One can add to this the West-Vlaanderen instance of dje and all attestations of enclitic je in dialects in eastern West-Vlaanderen that use ge in initial position. There is only one form on the map that is not formed in an enclitic position, namely ge, which is attested only once. The plethora of specifically enclitic pronouns testifies to the importance of enclitic clusters for the genesis of new pronouns.

            Taking into account that new pronouns can easily originate in an enclitic position, the geographic distributions of je and dje prove to be very instructive. In eastern West-Vlaanderen, some dialects combine enclitic je with sentence-initial ge and also, with strong gij rather than jij. The opposite situation, in which a dialect has strong jij but not weak je, is unattested in the whole Dutch language area. This leads to the conclusion that West-Vlaanderen pronouns with an initial /j/ have been formed in an enclitic position. Later, the initial /j/ has been generalised to the sentence-initial pronouns and, elsewhere, to strong pronouns as well. This explanation is probably valid for the larger part of the Dutch language area: the linguistic literature describes for 16th century Holland a morphological variation that is similar to the variation found in eastern West-Vlaanderen today. The geographic distribution of dje is equally instructive. Dje occurs in two distinct areas, namely the town of Kortrijk (West-Vlaanderen) and the southwest of Belgian Limburg. Kortrijk, on the one hand, is situated in between the West-Vlaanderen je-area and Oost-Vlaanderen, which uses de. Southwest Limburg, on the other hand, is situated in between the Brabant part of the de-area, and Tongeren and Lauw (Belgian Limburg), which have je. These data support the well-known hypothesis that dje is historically an intermediate stage in a transition from de to je.

2.3.2.2.3  Subject pronouns 2 singular, weak forms in inversion, finite verb ga ‘to go’ (map 40a) (map in dynaSAND)

Map 40b shows the morphological variation in weak, second person singular pronouns following the verb gaan (‘to go’). The answers for the following questions are used: the paradigm of gaan (‘to go’), which was requested during the fieldwork sessions, and the translations for the sentence Als je gaat dan ga je (‘If you go, you go’), which were asked during the telephone interview.

For a rather large number of sampling points, no weak pronoun following gaan can be found in the SAND corpus. This is probably due to methodological reasons: to a large extent these forms are ellicited by running through the entire verbal paradigm, which may stimulate the informants to use strong pronouns in their answers rather than weak ones. It is, however, still clearly visible that the weak pronouns following gaan (‘to go’) are not completely the same as the ones following leven (‘to live’). In the du-areas, the differences are small: pro-drop (pro) occurs in the northern and eastern provinces of Friesland, Groningen, Drenthe and Overijssel only. In the same area, some isolated examples of [s]e and [st]de are found. The southern du-area (Limburg) has [s]e and [st]de. [t]se is not attested following gaan (‘to go’).

In the other areas, the geographic distribution of ge (only attested in the Vlaams-Brabant town Tienen) and d’r (attested near the border between Vlaams-Brabant and Belgian Limburg) especially resemble the respective distributions on map 39b. The geographic distribution of dje on map 40a is very similar to the one on maps 39a and 39b as well (dje occurs mainly in the southwest of Belgian Limburg). However, map 40a does not show an isolated example of dje in Kortrijk (West-Vlaanderen) like map 39b. Another clear difference is the absence of the pronoun ie following gaan (‘to go’). In the Dutch provinces Noord-Holland, Zuid-Holland, Utrecht and Gelderland, je is the only weak pronoun that is found. The same pronoun je occurs in Overijssel and Drenthe. In Belgium, the dominance of je in West-Vlaanderen is striking. The most remarkable differences between map 39b and map 40a can be found in the area that has de following leven (‘to live’). Following gaan (‘to go’), the very same area shows no less than four different weak pronouns: de, re, e and te. De is found in most places in Noord-Brabant, Antwerpen, the north of Vlaams-Brabant, the northeast of Oost-Vlaanderen and the area around Gent (Oost-Vlaanderen). Re is an allophone of de that is found at four sampling points. In the centre and in the west of Oost-Vlaanderen, e is the dominant form. Two places, namely Oudenaarde and Ronse (both in Oost-Vlaanderen), have te.

There are older maps showing weak, second person singular pronouns following verbs with a stem ending with /d/, /t/ or a vowel (De Schutter 1989a:19 and Devos 1986). De Schutter’s map depicts the morphological variation in the pronouns following weten (‘to know’), and distinguishes more variants than map 40a. No remarkable differences are observed, although the pronunciation of je in Frans-Vlaanderen and West-Vlaanderen seems to be slightly different following weten (‘to know’) than following gaan (‘to go’): following weten, the pronoun is sje rather than je. The very detailed map of West-Vlaanderen by Devos (1986) confirms the image of map 40a, and shows even clearer that je is more widespread following gaan (‘to go’) than following verbs with a stem ending on a consonant, like durven (‘dare’) and leven (‘to live’).

Among the weak, second person singular pronouns following gaan (‘to go’), a lot of specifically enclitic pronouns are attested, all exemplifying the possibility of new pronouns to originate in an enclitic position. These newly originated pronouns dominate the map: the only pronoun that certainly did not originate in an enclitic position, ge, is attested in only one single location.

2.3.2.2.4  Subject pronouns 2 singular, weak forms following complementiser (map 40b) (map in dynaSAND)

The form of the weak, second person singular pronoun in an enclitic position does not only depend on the preceding sound (cf. the distinction between a verb stem ending with a vowel, /d/ or /t/ and those stems ending with another consonant), but also on the word class of the preceding constituent. Dialects may use a different form when a certain pronoun is cliticised to the finite verb or to a complementiser. Second person pronouns following complementisers are found in the SAND sentences Als je gezond leeft, leef je langer (‘If you live healthily, you will live longer’) and Ze gelooft dat jij eerder thuis bent dan ik (‘She believes that you are home earlier than I’), both part of a translation task that was presented to the informants during the fieldwork interviews.

In the du-areas, differences between the form of the weak pronoun following a finite verb or a complementiser are virtually non-existent. The sampling points in the northern du-area in general have pro-drop, a single instance of [st]de in Roswinkel (Drenthe) being the exception. In the southern du-area, [st]de and [s]e are found; the enclitic cluster [t]se that occured following leven (‘to live’) is not attested.

In the other areas, the geographic distribution of je, dje and d’r hardly differ from their distribution in other enclitic positions: je is found in a large, western area, reaching from Frans-Vlaanderen in the south, over West-Vlaanderen, Zeeland and Zuid-Holland to the north of Noord-Holland, and also covering the central and eastern Dutch provinces, namely Utrecht, Gelderland, Overijssel and the larger part of Drenthe. In addition, some more isolated instances of je can be found outside this area, including three very isolated examples in the towns of Lauw and Tongeren, both in Belgian Limburg, and in Nuth, in Dutch Limburg. Dje and d’r are found in Belgian Limburg. Ie is not attested following complementisers (cf. following gaan ‘to go’). The most remarkable feature of map 40b is probably the large number of attestations of ge, a pronoun that is particularly rare in an enclitic position following finite verbs. Cliticised to complementisers, ge is attested in Oost-Vlaanderen, Vlaams-Brabant, Antwerpen, Noord-Brabant, in the west of Belgian Limburg and once in Dutch Limburg. Pronouns like de, e and te, which have a wide distribution when encliticised to finite verbs, are only found in Oost-Vlaanderen (de, e and te), in the north of Noord-Brabant and in one place in Belgian Limburg (de). Finally, the pronoun ke needs to be discussed. Ke is only attested following the complementiser als ‘if’, and occurs in the southeast of Oost-Vlaanderen and in Noord-Brabant, in Vorstenbosch. All sampling points at which [as]ke is found, are situated in areas in which Standard Dutch /sch/ is pronounced /sk/ when /s/ and /ch/ are in the same morpheme. This indicates that the combination of a complementiser and an enclitic pronoun in those dialects is no longer felt as a combination of two different elements by speakers of those dialects.

In a wa            y, map 40b resembles each of the other maps showing weak, second person singular pronouns. In the largest part of the Dutch language area, the pronouns that are encliticised to complementisers are the same pronouns as used in an enclitic position following finite verbs. This is most obvious in the du-areas (Friesland, Groningen, the north of Drenthe, Twente and a large part of both Dutch Limburg and Belgian Limburg) and in the east of West-Vlaanderen, since the specifically enclitic pronouns that are cliticised to the complementiser, by definition, do not occur in other positions (e.g., pro in the northern du-area (see 2.3.2.3), [s]e in Limburg, je in eastern West-Vlaanderen, where it does not occur in other syntactic environments). Other areas, including those dialects with je and those with dje following complementisers, do not make use of a specifically enclitic pronoun, but behave in a similar manner by using a pronoun that originally formed in an enclitic position (and has subsequently been generalised to other syntactic positions). There is one area in which important discrepancies between the form of the pronoun following finite verbs on the one hand, and its form following complementisers on the other, are the rule rather than the exception, namely a central, southern area including by and large the provinces Vlaams-Brabant, Antwerpen and Noord-Brabant (i.e., the area in which Brabant dialects are spoken). In this area, the dominant form following complementisers is ge, the obvious reduced form of the full pronoun gij which is used in sentence-initial position as well. Nevertheless, specifically enclitic pronouns or pronouns that originated in an enclitic position (such as de) are not completely absent in the Brabant dialects (especially the peripheral ones). Oost-Vlaanderen is a transitional area: most places in Oost-Vlaanderen can use both ge and a specifically enclitic pronoun following complementisers.

2.3.2.3  Subject pronouns, pro-drop 2 singular; overview (map 41a) (map in dynaSAND)

In the preceding sections, pro-drop constructions are discussed together with weak pronouns. This is a sensible way to proceed, since there are some similarities between the use of weak pronouns and the use of pro-drop: in both cases, the actual subject must be easily identifiable and cannot bear emphasis. Map 41a provides an overview of all the attestations of pro-drop that are dealt with in previous sections, and includes answers to a whole range of SAND sentences, such as the paradigm of gaan (‘to go’), translation sentences Als je gezond leeft, leef je langer (‘If you live healthily, you will live longer’), Ze gelooft dat jij eerder thuis bent dan ik (‘She believes that you are home earlier than I’), Je gelooft zeker niet dat hij sterker is dan jij (‘Surely you do not believe that he is stronger than you’), Het gebeurde toen je wegging (‘It happened when you left’), Ik weet waar je geboren bent (‘I know where you are born’) and Nu je klaar bent, mag je gaan (‘Since you are ready, you can go’), all taken from the fieldwork interviews, and translation sentences Je weet wel dat je slim genoeg bent (‘You know that you are smart enough’) and Als je gaat dan ga je (‘If you go, you go’) from the telephone interviews.

All instances of pro-drop in the SAND corpus come from the north and the northeast of the language area, all from dialects that have kept the pronoun du and the corresponding agreement marker –st. The phenomenon is distributed differently depending on the different types of sentences in which it occurs. The map shows that pro-drop in sentences with inverted word order (labelled ‘_V’ on the map) is the most widespread type. Pro-drop following complementisers (labelled ‘V_’ on the map) is attested almost as often. Pro-drop in sentences with regular word order – with pro in sentence-initial position – (labelled ‘C_’ on the map), is relatively rare. One can conclude from this that all dialects with pro-drop have pro-drop following verbs. At first sight, the map shows five sampling points without pro-drop in sentences with inverted word order, but with pro-drop in one of the other syntactic positions, which are therefore possible counterexamples to this claim: Visvliet, Jorwert and Kloosterburen only have pro-drop in an enclitic position following complementisers; Schiermonnikoog and Makkinga can only have pro-drop in sentence-initial position. However, Visvliet, Jorwert and Schiermonnikoog are not ‘real’ counterexamples because an agreement marker –ste appears on the sentence-final verb in embedded clauses in these dialects. Thus all verb forms like gaste can be analysed as the stem ga and the agreement marker –ste (instead of the stem ga, the agreement marker –st and the pronoun de), giving rise to an abundance of examples of pro-drop in sentences with inverted word order. It is also the case the the agreement marker, –ste, is attested at four other sampling points. There are, thus, only two examples of dialects without pro-drop in sentences with inverted word order, but with pro-drop in other syntactic positions, i.e., following complementisers (Kloosterburen) or in sentence-initial position (Makkinga).

2.3.2.4  Honorifics (map 41b) (map in dynaSAND)

The maps for the second person singular pronouns show that the original, Middle Dutch pronoun, du, at present only occurs in some peripheral areas in the north and the east of the Dutch language area. The maps suggest a sharp borderline between the areas that have kept the ‘original’ situation in the periphery, and areas in which du has been replaced by a morphological variant of Middle Dutch ghi. In reality, however, the situation is more complex. In the area in which du is retained, other pronouns referring to the second person singular are attested, namely polite pronouns or ‘honorifics’. In some of the relevant dialects, the honorific pronoun is used far more frequently than du, and is thus hardly recognisable as a pragmatically marked, honorific form. Honorifics may be second person plural pronouns or special pronouns, which are only used for addressing someone in a polite way. Honorifics were explicitly ellicited both during the fieldwork interviews, by means of the translation sentence U gelooft dat Lisa even mooi is als Anna (‘You believe that Lisa is as pretty as Anna’), and during the telephone interviews, with the translation sentence Als u vindt dat u gezond leeft, leeft u dan vooral zo verder (‘If you think you are living healthily, keep on living like that’). In the translations for the sentences Als je gezond leeft, leef je langer (‘If you live healthily, you will live longer’) and Je gelooft zeker niet dat hij sterker is dan jij (‘Surely you do not believe that he is stronger than you’), both asked during the fieldwork interviews, honorifics are found as well.

Most dialects that have kept du as a second person singular pronoun use an originally second person plural pronoun as a polite pronoun. In Friesland, the original object pronoun jo is dominant. Hindeloopen has jie. Jo is found outside Friesland as well, in the provinces Groningen and Drenthe, where, like in Overijssel, ie is also used. At some sampling points, only the weak form je is attested, and it remains unclear what the corresponding strong pronoun might be. In Belgian and Dutch Limburg, gij or gir is used, or the weak form ge. The south of Dutch Limburg has dir (in one place: ihr). The same area also exhibits one single instance of sie (in Vaals), a third person plural which can be used as an honorific (such as also occurs in German).

U, the honorific that is used in Standard Dutch, is a predominantly western form (cf. attestations in Noord-Holland and Zuid-Holland and Utrecht), which may indicate that the pronoun originates in that region. In addition, u is found in more peripheral areas: all Dutch provinces, with the exception of Dutch Limburg, show at least one instance of u. This rather wide but fragmentary distribution is probably due to the influence of Standard Dutch. U did not penetrate into Belgium: there is not a single attestation of u in Belgium.

2.3.3  Subject pronouns third person singular masculine

2.3.3.1  Subject pronouns 3 singular masculine, strong forms  (map 42a) (map in dynaSAND)

The third person masculine singular needs to be dealt with in a somewhat different way than the other grammatical persons. The main reason for this is that the alternation between pronouns with /ie/ and the ones with the diphtong /ij/ is not phonologically determined as it is for the other grammatical persons. Usually, the centre of the Dutch language area has the diphtong /ij/, whereas more peripheral areas have kept the older /ie/-pronunciation. However, a larger number of central dialects in which the diphtong /ij/ occurs (cf. the presence of a strong form hij), have kept the older pronoun ie as well. In these dialects, such as Standard Dutch, hij is the strong pronoun and ie the weak one. The fact that ie is kept, suggests a certain tendency for dialects to have two different pronouns at their disposal, both a strong one and a weak one.

The SAND sentence in which the most strong pronouns occur, is Zelfs hij kan dat niet oplossen (‘Even he cannot solve that’), a sentence from the telephone questionnaire. In addition, the translations are used for Als hij nog drie jaar leeft, leeft hij langer dan zijn vader (‘If he lives three more years, he will be living longer than his father’), ‘k Geloof dat ik groter ben dan hij (‘I believe that I am taller than he’) and Je gelooft zeker niet dat hij sterker is dan jij (‘Surely you do not believe that he is stronger than you’), three questions from the fieldword interview. During the fieldwork interviews, informants were also asked whether the sentence Hem is dood (‘Him is dead’), with a strong object pronoun as a subject, was grammatical in their dialect.

Most pronouns on map 42a are variants of the pronoun hij (hie, hijn). The other pronouns include the original object pronoun (hem), doubled forms (jij, a combination of ie and ij) and some demonstratives. Strong pronouns with /ie/ are found in the periphery. In the western part of the language area, the north and the west of West-Vlaanderen and Zeeland constitute a massive (h)ie-area; in the eastern part, attestations of (h)ie are found in Belgian Limburg, Gelderland, the east of Overijssel and Drenthe, in Friesland and in some places close to the former IJsselmeer. All of these (h)ie-attestations are found in areas in which West Germanic /î/ did not develop into a diphtong (see FAND III.1.1, vol. 2 (2000), p.123-145). In the southeast of the language area (the south of Belgian and Dutch Limburg), remnants of the old pronoun /iz/ are found, in the form of (h)er, (h)ir. The original object form (h)em, a strong subject pronoun, is found mainly in West-Vlaanderen. The hem-area seems to be a buffer between the area with hij on the one hand and the hie-area on the other. Around Gent, originally reduplicated forms are found (jij, jem), which are originally a combination of weak ie and strong (h)ij or (h)em. In some of these dialects, the third person singular inflectional –t is reanalysed as part of the pronoun (e.g., in tjij, tjem). In some Brabant dialects, not only hij, but also the demonstrative die is used as a strong subject pronoun (or dieë(n), a form showing agreement with a masculine antecedent).

2.3.3.2  Subject pronouns 3 singular masculine, weak forms

2.3.3.2.1  Subject pronouns 3 singular masculine, weak forms in initial position  (map 42b) (map in dynaSAND)

Sentence-initial weak pronouns for the third person masculine singular were ellicited during the SAND interviews with the translation sentence Hij gelooft dat Bart en Peter sterker zijn dan Geert en Jan (‘He believes that Bart and Peter are stronger than Geert and Jan’).

The Standard Dutch, weak pronoun for the third person masculine singular constitutes a somewhat special case since it can only be used in an enclitic position. In sentence-initial position, Standard Dutch uses the strong pronoun, hij. In most dialects, a similar situation is observed: the vast majority of the dialects in the Dutch language area can indeed only use a strong, third person masculine pronoun in sentence-initial position. Exceptional examples of weak pronouns in sentence initial position are found in the south of Dutch Limburg, where five sampling points have weak ‘r, and, especially, in Frans-Vlaanderen, West-Vlaanderen and the locations of Sint-Laureins and Oosteeklo in Oost-Vlaanderen. In the Vlaanderen area with weak forms, there is morphological variation. The attested variants are ‘n, ne, enne, je and e (pronounced è). Among these forms, je is obviously derived from the strong pronoun (h)ie. Weak e (pronounced è) is probably a vocalised form of an older reduced pronoun e (pronounced as schwa). ‘n, ne and enne are possibly object pronouns in origin, but the possibility of another origin must be left open, since the geographic distribution of the three forms does not correspond completely to the geographic distribution of the original, enclitic object pronoun ‘n (see the rest of 2.3.3.2).

Two other sampling points in Belgium, namely the towns of Tienen and Wolfsdonk in Vlaams-Brabant, do have a weak masculine pronoun in sentence-initial position that is clearly an original object pronoun, i.e., ‘m in Tienen and t’m in Wolfsdonk.

2.3.3.2.2  Subject pronouns 3 singular masculine, weak forms following complementiser   (map 43a) (map in dynaSAND)

Map 43a shows answers to two SAND questions from the fieldwork interviews, namely Als hij nog drie jaar leeft, leeft hij langer dan zijn vader’ (‘If he lives three more years, he will be living longer than his father’) and Je gelooft zeker niet dat hij sterker is dan jij (‘Surely you do not believe that he is stronger than you’).

Weak, masculine, enclitic pronouns are far less rare than weak masculine pronouns in sentence-initial position, but they too are not attested throughout the entire Dutch language area. The most widespread pronoun is ie. The enclitic pronoun ie is found especially in the centre of the language area (Noord-Holland and Zuid-Holland and Utrecht being the nucleus of this area), but is attested in all Belgian and Dutch provinces except Groningen. Completely weakened forms, containing schwa, are only found in the periphery. This includes e (pronounced schwa) and (h)’r in the east, and je at various places in the Dutch language area, in Oost-Vlaanderen, Noord-Brabant and Gelderland. In the oher areas with completely weakened forms, the original object pronouns ‘m (or t’m) and ‘n (or ne) occur. (t)’m is found mainly in Vlaams-Brabant and Antwerpen, and on the Zeeland and southern Dutch islands (in Scherpenisse and Ouddorp) and in Gelderland (Kilder) as well. (t)’n is attested in Frans-Vlaanderen (as ne), in West-Vlaanderen, in the south of Oost-Vlaanderen, and in one place in the Netherlands, namely in Roswinkel (Drenthe).

Very few weak pronouns are found in Oost-Vlaanderen. Some dialects in Oost-Vlaanderen have remnants of the pronoun ie in the originally reduplicated, strong forms (t)jij and (t)jem, which have been formed by a fusion of ie and (h)ij or (h)em. As a simplex pronoun however, ie is hardly found in Oost-Vlaanderen. Although, strictly speaking, they don’t have a formally weak pronoun, the dialects with strong (t)jij and (t)jem in Oost-Vlaanderen have, like most dialects, two different pronouns at their disposal, a ‘normal’ one (mostly hij), and a stronger one ((t)jij, (t)jem).

When compared to the older map by De Schutter (1989a:29), map 43a shows a remarkably wide distribution of ‘m. De Schutter’s map only distinguishes a small ‘m-area in the proximity of the city of Antwerpen, and sporadic attestations in the provinces Antwerpen and Vlaams-Brabant. Map 43a shows a larger ‘m-area around Antwerpen. In addition, the Zeeland attestations of ‘m are absent on De Schutter’s map. Koelmans (1968) provides a map with the distribution of (t)’n in the southern half of the Dutch language area, which casts some light over the issue. There are three (t)’n-areas on Koelmans’s map: a Zeeland area, an area covering Frans-Vlaanderen and West-Vlaanderen, and a Brabant area. The first two areas have a size comparable to their size on map 43a. The third area, however, is much larger on Koelmans’s map than on the SAND map, including almost the whole of Vlaams-Brabant (except the easternmost part) and the southeast of Oost-Vlaanderen. In a large part of this area, (t)’n seems to be replaced by an other original object pronoun, (t)’m, which, according to De Schutter’s (1989a) data, used to occur far less frequently in earlier times.

2.3.3.2.3  Weak enclitic subject pronouns 3 singular masculine with initial /t/ (map 43b) (map in dynaSAND)

In many dialects, specifically enclitic pronouns are attested that consist of an original pronoun (‘r, ie, ‘m, etc.), preceded by a /t/. These were probably formed through reanalysis of an enclitic cluster that originally included a third person singular, inflectional –t, and a pronoun. Since the SAND material does not contain enough data to determine for each dialect whether a particular dialect has a form at its disposal with or without such an initial /t/, or even both, map 43b only shows those dialects in which such a pronoun is likely to occur. Map 43b shows attestations of /t/ before an enclitic pronoun in syntactic environments in which no inflectional –t is expected, such as following a regularly formed preterite (e.g., leefdet’r ‘lived he’), following is (e.g., istie ‘is he’) and following zou (e.g., zout’m ‘should he’). The material for the following SAND sentences was used: Zou hij dat gedaan hebben gekund? / Zou hij dat gedaan gekund hebben? (‘Could he have done that?’), Is hem dood? (‘Is he (lit.: him) dead?’), and Vroeger leefde hij als een beest (‘He used to live like an animal’). The first two sentences are dialectal constructions that the informants were asked to judge, the last sentence was part of a translation exercise.

All attestations of /t/ as shown on map 43b, may be part of the pronoun. Many attestations of possible specifically enclitic pronouns with an initial /t/ are found, especially in Vlaanderen, along the Rhine, in the border region between Antwerpen and Vlaams-Brabant and in the south of Belgian Limburg and Dutch Limburg. However, since some of these attestations can also be attributed to the presence of an inflectional –t in the regularly formed preterite, in is (‘is’) or in zou (‘should’), we cannot be entirely sure about the status of the /t/.

2.3.4  Subject pronouns third person singular feminine  (map 43c) (map in dynaSAND)

Third person singular, feminine pronouns display less variation than their masculine counterparts. Only the form of the strong pronoun differs from time to time; the weak pronoun is uniformly ze. Strong feminine pronouns are found in the SAND corpus in a number of sentences that were part of the fieldwork interviews: the paradigm of gaan (‘to go’), the translation sentence Als zij zo gevaarlijk leeft, leeft ze niet lang meer (‘if she lives so dangerously, she won’t be living long anymore’), and the sentences Is haar dood? (‘is she (lit.: her) dead?’) and Met hij te gaan werken moest zij heel de dag thuis blijven (‘As he went out for work, she had to stay at home’). These last two sentences were both presented in their dialectal form for the informants to judge. In addition, feminine pronouns occur in the translations of Als ze gaat dan gaat ze (‘if she goes, she goes’), a translation sentence from the telephone questionnaire.

The vast majority of the sampling points have (a phonological variant of) zij. Two places in Zeeland, Ouddorp and Scherpenisse, have haar, originally an object pronoun. In the Belgian province of Antwerpen, the towns of Lier and Mechelen not only have zij, but also zeur, a rare and until now unattested pronoun that probably has its origin as a contamination of ze or zij on the one hand, and the object pronoun haar (pronounced as eur in the relevant dialects) on the other hand. Finally, in the locations Bray-Duinen and Moerbeke (Frans-Vlaanderen), the old demonstrative zoe is attested.

At five sampling points, the informants spontaneously use a demonstrative (e.g., die) to translate a third person singular, feminine pronoun. The relevant sampling points do not constitute a real area, but are situated in, or very close to, the Brabant dialect area: one attestation comes from Noord-Brabant, two come from the province of Antwerpen, and two from Belgian Limburg, close to the border with Vlaams-Brabant.

The linguistic literature also mentions the use of masculine or neuter pronouns to refer to females. The SAND corpus does contain a number of such cases, but they all concern weak pronouns, and these cases are by no means numerous enough to provide an accurate picture of the geographic distribution of the phenomenon. Therefore, the use of masculine and neuter forms to refer to females will not be discussed.

2.3.5  Subject pronouns first person plural

2.3.5.1  Subject pronouns 1 plural, strong forms

2.3.5.1.1  Subject pronouns 1 plural, strong forms, simplex forms (map 44a) (map in dynaSAND)

Strong forms for the first person plural are found in different test sentences in the fieldwork questionnaire: translation sentences Als we sober leven, leven we gelukkig (‘If we live soberly, we live happily’) and Ze geloven dat wij rijker zijn dan zij (‘They believe that we are richer than they’). From the telephone questionnaire, the translation sentence Mogen we wel weten dat wij ook gevraagd zijn? (‘Can we know that we are invited too?’) is used, as is the sentence Ons hebben daar niks mee te maken (‘We (lit.: us) have got nothing to do with that’), which the informants were asked to judge.

The larger part of the Dutch language area uses wij or a phonological variant (e.g., wie, vie), like Standard Dutch. This is characteristic of most sampling points in the Netherlands and in Belgian Limburg, and some Belgian locations near Antwerpen and Gent. The southeast of the language area has pronouns with a final /r/ (wir, vir). Some dialects use an original object pronoun, ons on the Zeeland and Zuid-Holland islands and noes in Frans-Vlaanderen. All other pronouns can be considered as originally having been a compound of (a form of) wij and (a form of) the noun lieden.

Such lieden-compounds are found in three distinct areas: first, a small area in Gelderland, including one place in Utrecht (Eemnes); secondly, another small area comprising a number of Zuid-Holland and Zeeland locations, including the Noord-Brabant location Steenbergen; thirdly, a large southern area containing Frans Vlaanderen, the whole of West-Vlaanderen, the largest parts of Oost-Vlaanderen, Antwerpen and Vlaams-Brabant, and some locations in Belgian Limburg.

2.3.5.1.2  Subject pronouns 1 plural, strong forms, complex forms (map 44b) (map in dynaSAND)

Map 44a shows a number of complex pronouns, the so-called lieden-compounds. The lieden-compounds display morphological variation, as shown on map 44b. The form with the widest distribution is wijlie (variants include wullie, wijle, etc.), which is found in the Netherlands and in the Belgian provinces of Antwerpen and Vlaams-Brabant. A rather special form, wijlui, is attested in Beekbergen, in the province of Gelderland. In the border zone between Vlaams-Brabant and Belgian Limburg, variants with a final /s/ are found (e.g., wijlies). Other variants are wulder (also: wielder), attested in Oost-Vlaanderen and in the east of West-Vlaanderen; wieder (also: wodder, wudder), attested in the rest of West-Vlaanderen, Frans-Vlaanderen and Zeeland Vlaanderen; wiender (or wunder), attested in four locations in Oost-Vlaanderen and West-Vlaanderen; miender and wijr. Miender is a rare and special variant of wiender, which is only found in the two towns Brugge and Roeselare in West-Vlaanderen. The initial /m/ is explained by the assumption that miender is formed in an enclitic position through assimilation of /w/ to the inflectional /n/ of the verb, a phenomenon usually affecting weak pronouns only. Finally, wijr (or wuir), which is attested in Oost-Vlaanderen, resembles the pronouns with a final /r/ that are attested in Limburg (see map 44a). In Oost-Vlaanderen, however, the final /r/ is the result of a formal reduction of the second part of a lieden-compound.

2.3.5.2  Subject pronouns 1 plural, weak forms

2.3.5.2.1  Subject pronouns 1 plural, weak forms in initial position  (map 45a) (map in dynaSAND)

Weak, first person plural pronouns in sentence-initial position are found in two SAND sentences, namely the translation sentence We geloven dat jullie niet zo slim zijn als wij (‘We believe that you are not as smart as we’), and We zijn (me) wij daar nog nooit geweest (‘We have (lit. we have we) never been there’), a dialectal sentence which had to be judged by the informants, and subsequently translated.

            It is remarkable that, for a rather large number of sampling points, no weak, first person plural pronoun in sentence-initial position is found in the SAND corpus. For some locations, only strong pronouns are found. The tendency to use strong pronouns rather than weak ones, seems to be especially strong in the east of the Dutch language area, probably indicating that the distinction between weak and strong pronouns is far less systematic in the east than in the west. Whether this also means that dialects exist without weak pronouns, must be left open at this point.

            In the locations that do have a weak pronoun at their disposal, four forms occur: we (cf. Standard Dutch), ve, v’r (with a final /r/), and me, a pronoun formed in an enclitic position. Among these pronouns, we has the widest geographic distribution, since it occurs throughout the language area, the exception being Frans-Vlaanderen. Ve, a merely phonological variant of we, is found in two places in Limburg. The weak form v’r is found in four sampling points in the south of Limburg that also have a strong pronoun with a final /r/ at their disposal. Finally, me is used in a subset of the dialects with a lieden-compound as a full pronoun, more precisely, in most sampling points in West-Vlaanderen and Oost-Vlaanderen, but also in Antwerpen, Vlaams-Brabant and in two locations in Zeeland.

            According to the older map by De Schutter (1989a:90), weak pronouns are indeed rare in the east of the language area. When compared to De Schutter’s map, map 45a shows a very limited distribution of me. De Schutter, using RND material, finds me to be the dominant form in the provinces Antwerpen, Vlaams-Brabant, Oost-Vlaanderen and even Zeeland, where the SAND material only lists isolated instances of me. In West-Vlaanderen, me is virtually the only pronoun visible on De Schutter’s map. These data indicate that me is losing ground, both in Belgium and in the Netherlands.

2.3.5.2.2  Subject pronouns 1 plural, weak forms following a complementiser  (map 45b) (map in dynaSAND)

The geographic distribution of weak, first person plural pronouns in enclitic positions differs from the distribution of the proclitic forms. There are, however, no significant differences between the enclitic position following complementisers and the enclitic position following verbs. Since the SAND corpus contains more pronouns following complementisers, these forms are mapped. Three test sentences are used: the translation sentences Als we sober leven, leven we gelukkig (‘If we live soberly, we live healthily’) and Ze geloven dat wij rijker zijn dan zij (‘They believe that we are richer than they’), both from the fieldwork questionnaire, and the translation sentence Mogen we wel weten dat wij ook gevraagd zijn? (‘Can we know that we are invited too?’), from the telephone questionnaire.

The attested pronouns are the following: we, ve, v’r and me (the same forms that are used in proclitic position), and m’n. The geographic distribution displays some minor differences. We, again, is found all over the Dutch language area, including the eastern part in which proclitic weak pronouns are rare. Ve is found more often in an enclitic position than in proclitic position (nine attestations vs. two). V’r is found at four sampling points, which are not the same locations as the ones that have v’r on map 45a, but which are found in the same area of southern Limburg. Me is found much more often in an enclitic position than sentence-initially, in a large area, comprising Frans-Vlaanderen, West-Vlaanderen and Oost-Vlaanderen, Antwerpen, Vlaams-Brabant, Zeeland and Zuid-Holland. In addition, isolated instances of me occur in Moerdijk (Noord-Brabant) and Vriezenveen (Overijssel). The wide distribution of me in an enclitic position is not surprising, since me has been formed in an enclitic position, through assimilation of we to the inflectional –n on the verb (or the complementiser). Finally, the pronoun m’n is a remarkable fusion of me and an inflectional –n that follows the pronoun. It is found in the border zone between Oost-Vlaanderen, West-Vlaanderen and Zeeland on the one hand, and Vlaams-Brabant, Antwerpen, Noord-Brabant and Zuid-Holland on the other hand. M’n occurs quite frequently, especially in Oost-Vlaanderen.

When compared to the older map by De Schutter (1989a:36), one must conclude again that me is losing ground in favour of we, although the tendency is less clear than for proclitic we. The rare attestations of me that De Schutter finds in Noord-Brabant, have disappeared on map 45b. Furthermore, on De Schutter’s map, me is de dominant form in Antwerpen and Vlaams-Brabant, whereas map 45b shows me and we in equal proportions in these provinces.

2.3.6  Subject pronouns second person plural

2.3.6.1  Subject pronouns 2 plural, strong forms, simplex forms  (map 46a) (map in dynaSAND)

Many second person, plural pronouns are attested in the SAND corpus. From the fieldwork questionnaire, the following test sentences are used: the paradigm of gaan (‘to go’), and the translations to Als jullie zo gevaarlijk leven, dan leven jullie nooit zo lang als ik (‘If you live so dangerously, you will never live as long as me’), We geloven dat jullie niet zo slim zijn als wij (‘We believe that you are not as smart as we’) and Jullie geloven zeker niet dat zij armer zijn dan jullie (‘Surely you do not believe that they are poorer than you’). From the telephone questionnaire, the translations to Als jullie gaan dan gaan jullie (‘If you go, you go’) are used.

Many different pronouns are found, and the variation is caused by different parameters. Some dialects use a simplex form rather than one that was originally complex (see map 46a). These dialects include the dialects in which du has not been replaced by the original second person plural pronouns gij, jo or their morphological variants, as spoken in parts of Groningen and Limburg. In Limburg, gij, gir and dir are the dominant forms; two locations have djij, one has jir. In the northeast, simplex forms are relatively rare: jo (or joe) is found in some places, and two locations have ie, but the originally complex Standard Dutch pronoun jullie has penetrated the area. The initial sound of these pronouns varies considerably, probably as a result of adaptations of certain pronouns in the enclitic position. Map 46a also shows the pronoun jimme (also: jim), a pronoun that is used almost exclusively in Frisian dialects, and which has been analysed in the linguistic literature, both as a simplex and as a complex form.

2.3.6.2  Subject pronouns 2 plural, strong forms, complex forms  (map 46b) (map in dynaSAND)

The variable, initial sound of the simplex forms is also found for the originally complex forms, the so-called lieden-compounds, which are shown again on map 46b. The pronouns with an initial /j/ have the widest distribution, mainly because of the wide distribution of jullie: jullie is found in the whole of the Netherlands, with the exception of some parts of Friesland, Limburg and Zeeland. Especially in more peripheral dialects, jullie often alternates with other pronouns, probably indicating that jullie has entered these dialects as a simplex form through the influence of either central dialects or Standard Dutch. Other pronouns with an initial /j/ are judder, julder and junder, which are found in Zeeland, Frans-Vlaanderen and the west of West-Vlaanderen, and jelui and joel, which are found in the jullie-area. The southern half of Belgian Limburg and the Vlaams-Brabant town of Walshoutem, have pronouns with an initial /dj/ (e.g., djijlies). In the eastern provinces of Gelderland and Overijssel, originally complex pronouns with an initial vowel are found (e.g., iele). Finally, most Vlaanderen and Brabant dialects, as spoken in West-Vlaanderen and Oost-Vlaanderen, the town of Hulst in Zeeland Vlaanderen, Vlaams-Brabant, Antwerpen, Noord-Brabant and the towns of Geldermalsen, Druten, Groesbeek and Kilder in Gelderland, have  pronouns at their disposal with an initial /g/. Since the initial sound of the lieden-compounds in some dialects differs from the initial sound of the strong second person singular pronoun, these lieden-compounds may have been formed with weak, second person pronouns or with object forms of the second person singular rather than with full pronouns.

Some of the geographic variation is caused by the morphology of the second part of the lieden-compounds, namely the noun lieden or lui. These compounds fall apart into three categories: firstly, the forms with a final /r/, which are predominantly found in Vlaanderen; secondly, the pronouns ending on /ie/ and /e(n)/ that are found in Vlaams-Brabant, Antwerpen and in the Netherlands; thirdly, the eastern Dutch forms ending on –lui, -lu, etc. The first group, the forms ending on /r/ (found in Vlaanderen), probably derives from the genitive plural of lieden, namely lieder. The ‘Flemish’ final /r/, as found in Vlaanderen, should not be confused with the final /r/ as attested in Limburg, which is probably a remnant of the older, Germanic final /z/ of the pronoun rather than a part of a genitive lieder. The pronoun gij-lieder has developed into forms such as gulder and gielder (especially in Oost-Vlaanderen), and into julder (in Zeeland and Frans-Vlaanderen). In some cases, the /l/ is dropped, leading to gieder (in West-Vlaanderen), or replaced by /n/ (e.g., in gunder, giender). In the southeast of Oost-Vlaanderen, the intervocalic /d/ is replaced by /j/, which, in turn, is easily deleted, leading to second person plural pronouns like guir. The second group, the pronouns ending on /ie/, /e/ and /en/, derive from a compound with lieden, which has been severely reduced in form. This group also includes the pronouns ending on /s/ (e.g., gijlies, gijlings, in which the –s is probably a plural-s), as attested in Limburg. The third group of pronouns, which are attested in the east of the Netherlands, seem to be derived from a compound with lui as the second part (e.g., ielui, jelui, etc.).

Finally, a number of rare forms that are not mapped must be mentioned, since their inclusion would make the map uninterpretable. For instance, the informants in Nieuw-Scheemda (Groningen) not only use jullie and jo, but also zij as a second person plural pronoun. In Marken (Noord-Holland), besides jullie, speakers also use the pronouns jammekaar and jelawe. Zoeterwoude (Zuid-Holland) has junie. Beesel (Dutch Limburg) not only has geul, but also gir. Elsewhere in Limburg, ihr (or uhr) is found, in addition to dir (in Heerlen, Nieuwenhagen, Kerkrade and Vaals). Lauw, in Belgian Limburg, has djelles, but most of the time djij is used to refer to second person plural subjects.

2.3.7  Subject pronouns third person plural

2.3.7.1  Subject pronouns 3 plural, strong forms, simplex forms  (map 47a) (map in dynaSAND)

Strong, second person plural pronouns are found in the following SAND sentences: the paradigm of gaan (‘to go’) and the translations to Als ze voor hun werk leven, leven ze niet voor hun kinderen (‘If they live for their jobs, they do not live for their children’) and Jullie geloven niet dat zij armer zijn dan jullie (‘You do not believe that they are poorer than you’), all asked during the fieldwork interview. In addition, the following sentences were used from the telephone interviews: the translations to Als ze gaan dan gaan ze (‘If they go they go’) and the answers to Hun/hullie hebben daar niks mee te maken (‘They (lit.: them) have got nothing to do with that’), a sentence that the informants had to judge in its dialectal form.

Map 47a shows all originally simplex forms. The most widely attested variant is zij (or variants, like zie), which occurs frequently throughout Netherlands and in the Belgian provinces of Limburg, Antwerpen and Oost-Vlaanderen. Also widely attested, is the original object pronoun hun, which is frequently used in Noord-Holland and Zuid-Holland, Utrecht, Gelderland and Noord-Brabant. Isolated instances of hun are found in Zeeland (Goes, Hulst), Groningen (Onstwedde), Drenthe (Een, Emmen, Coevorden) and Overijssel (Sibculo). Both zij and hun are used in lieden-compounds as well (see below). Less frequently attested pronouns are the contamination form zun (pronounced /zyn/, and found in Antwerpen and Lier, in the province of Antwerpen), the Frisian hja (only found in Ferwert), and ie, which is found as a third person plural pronoun in Onstwedde and Warffum in Groningen, and in Diever in Drenthe. At some sampling points in the Belgian provinces of Antwerpen, Vlaams-Brabant and Limburg, the third person plural pronoun is spontaneously translated using the demonstrative die.

2.3.7.2  Subject pronouns 3 plural, strong forms, complex forms  (map 47b) (map in dynaSAND)

As has already been mentioned, both zij and hun occur as the first element of lieden-compounds. The lieden-compounds display a lot of morphological variation, which is shown on map 47b. Lieden-compounds with zij occur predominantly in, and just outside, a large area comprising Frans-Vlaanderen, West-Vlaanderen and Oost-Vlaanderen, Zeeland Vlaanderen, Vlaams-Brabant, Antwerpen and Noord-Brabant, but isolated instances are found far outside this area as well, including instances in Friesland, Groningen and Drenthe. The morphological variation is comparable to the variation attested for the first and second person plural: the southwest has variants on –der and –nder (e.g., zieder, zunder), and, more to the east, -lder (e.g., zulder). The southeast of Oost-Vlaanderen has zijr. In the Brabant dialects (as spoken in Noord-Brabant, Antwerpen, Vlaams-Brabant) and at some sampling points in the immediate surroundings of the Brabant area, zijlie is used (elsewhere zijle(n)), a pronoun which is also found in more northern locations in Noord-Holland, Utrecht, Gelderland, Drenthe and Overijssel. The Frisian location, Lemmer, has zijen. Some places in the border area between Vlaams-Brabant and Belgian Limburg have zijlies or zijlings. Lieden-compounds with hun (e.g., hullie, hunnie) are found more often than hun itself, in an area with a size comparable to the hun-area, with the Dutch Randstad (the urbanised area in the west of the Netherlands) as the nucleus, and reaching from Noord-Holland and Zuid-Holland to Utrecht, Noord-Brabant and Gelderland. Here too, isolated instances are found outside the core area, like in Emmen and Zwartemeer in Drenthe, in Haaksbergen (Overijssel; form: ului), in Stavenisse (Zeeland; form: ulder) and on the former island Urk (Flevoland; form: eurlui).

2.3.8  Overview maps on personal pronouns

2.3.8.1  Subject pronouns with accusative form  (map 48a)

On a number of the maps that were discussed earlier in this chapter, the use of object pronouns as subjects was observed. Map 48a provides an overview of the phenomenon, using information from a whole range of test sentences that were used to question the form of the pronouns of all grammatical persons (see the relevant sections for an overview). Object pronouns in subject position are attested in second person singular, (for both the ordinary modes of address and the honorific ones), third person singular masculine and feminine, first person plural, second person plural and third person plural. All information that is used for the maps presented above, is reused here, with the exception of the data for the second person plural, because it is not clear whether some of the most frequently used second person plural pronouns are original object forms or not. These include jullie, junder, jim(me), etc. For the same reason, the attestations of the polite pronoun u have also been excluded from te map.

            The use of object forms in subject position is well-attested in Dutch dialects. The majority of the sampling points exhibit at least one object pronoun that can be used as a subject. In addition, the phenomenon is also geographically widespread: except for some parts of Oost-Vlaanderen, Dutch Limburg and Belgian Limburg and some regions in the east and the northeast of the Netherlands, the phenomenon is found all over the Dutch language area. The geographic distribution depends to a large extent on the grammatical person of the subject. For instance, hun (‘them’) is only found in subject position in the Netherlands (especially in the centre of the language area), whereas the subject pronoun hem (‘him’) is exclusively Belgian (attested in West-Vlaanderen, Oost-Vlaanderen and Antwerpen). Ons (or noes ‘us’) is used in Frans-Vlaanderen and Zeeland; haar (‘her’) only in Zeeland. Two different second person singular pronouns are used as subjects, i.e., jo (jou, joe ‘you’), in the north of the Netherlands and in Zuid-Holland, and dich (‘you’), in Limburg.

The pronouns ‘n and ‘m constitute a somewhat special case, since they are the only weak, object pronouns found in subject position. Their geographic distribution is remarkable: they are found in an area containing Frans-Vlaanderen, almost the whole of West-Vlaanderen and Oost-Vlaanderen, Vlaams-Brabant, Antwerpen and Zeeland, and in some sampling points just outside this area. In addition, two very isolated instances strike the eye, in Kilder (Gelderland) and Roswinkel (Drenthe).

There is only one area where more than two object pronouns have penetrated into the subject pronoun paradigm, namely the Zuid-Holland and Zeeland islands. Since these islands are situated in the periphery of both the hun-area and the ‘n/’m-area, it is not very likely that the phenomenon has spread from the islands onwards. Rather, the concentration of object forms in subject function seems to be the result of a coincidental overlap of different areas with one or more object pronouns in subject position. This has facilitated a number of younger, very local developments.

2.3.8.2  Subject pronouns –lieden; overview  (map 48b)

For the first, the second and the third person plural, both simplex and originally complex pronouns are attested, the so-called lieden-compounds, which consist of an ‘original’ pronoun and (a variant of) the noun lieden or lui. The geographic distribution of these lieden-compounds varies for the different grammatical persons in which lieden-compounds are found. Like map 48a, map 48b shows data that are used for other maps, namely 44a/b, 46a/b, and 47a/b. For an overview of the relevant test sentences, see the discussion of these maps.

The differences in geographic distribution are quite large. In the second person plural, lieden-compounds have the widest distribution: they occur in the entire language area, with the exception of large parts of Belgian Limburg, Dutch Limburg, and Groningen. In the third person plural, lieden-compounds occur less frequently: they are rather rare in Groningen, Drenthe and Overijssel, and the distributional pattern is less dense than for the second person plural in the rest of the language area too. Finally, in the first person plural, lieden-compounds are relatively rare, and are almost only attested in a area containing Frans Vlaanderen, West-Vlaanderen and Oost-Vlaanderen, Antwerpen, Vlaams-Brabant, Zeeland and the Zuid-Holland islands, and in a smaller area spanning the northeast of Gelderland and the north of Utrecht.

Two factors, both stimulating a wider distribution of the lieden-compounds, can be mentioned: support of the standard language on the one hand, and a disambiguating function vis-à-vis another pronoun on the other. In the second person plural, both stimulating factors are at work: the wide geographic distribution of the lieden-compounds in the second person plural can be attributed for a good deal to the wide distribution of jullie, the Standard Dutch pronoun. In addition, lieden-compounds also ensure a non-ambiguous reference to the second person plural. This is because simplex forms such as jij, djij or gij, which are originally second person plural pronouns, at present mainly serve as second person singular pronouns. In the third person plural, Standard Dutch does not have lieden-compounds. However, 3pl. lieden-compounds do have a disambiguating function, since they distinguish the third person plural from the third person singular feminine. Finally, in the first person singular, neither of the stimulating factors is present. Pronouns like wullie, wulder and wieder are not present in Standard Dutch, and the simplex pronoun wij is unambiguous in all dialects.

2.4  Discussion of the maps: expletive er and het

2.4.1  Expletive subject in regular order in existential clause (map 49a)

Expletives display similar morphological variation as personal subject pronouns. On the one hand, dialects differ in whether they use a strong expletive (daar) or a weak one (het / er / d’r or t’r). On the other hand, the linguistic literature suggests that some variants are formed in an enclitic position (er, t’r), while others, like d’r, are merely reduced variants of strong daar. However, alternative hypotheses exist for the origin of each of these expletives. The form of the weak, sentence-inital expletive was ellicited during the fieldwork interviews, using the test sentence Er wil niemand (niet) dansen (‘No one wants to dance’). The geographic distribution of the less frequently used expletives, het and daar, was also an issue during the telephone interviews. The informants were asked whether the sentences Het / Daar zat een inbreker in deze kast (‘There was a burglar sitting in this closet’) were grammatical in their dialects. The use of the locative constituent in deze kast (‘in this cupboard’) excludes a locative interpretation of daar (‘there’).

The most frequently used expletive is d’r (‘there’), which is found all over the Dutch language area. The form used in Standard Dutch is er (‘there’). Its distribution is quite limited, as it is only attested at 77 sampling points. The er-attestations are found in all provinces of the Netherlands, but the form is rare in Friesland. In Belgium, only isolated instances of er are found. Both in the Flemish dialect area (in West-Vlaanderen and Oost-Vlaanderen) and in Dutch Limburg, the expletive het (‘it’) is used. The fourth weak form, t’r, is quite rare and seems to occur in dialects in a transitional zone in Oost-Vlaanderen, between the het-area (in Vlaanderen) and the d’r-area (in Brabant). Strong daar occurs frequently, especially in the eastern half of the Dutch language area. However, spontaneous instances of daar, attested during the fieldwork interviews, are much more rare and almost absent outside Belgian Limburg, Vlaams-Brabant and Antwerpen. This is the same area in which daar is sometimes the only expletive that is attested, whereas most sampling points with daar outside this area also have another expletive at their disposal. Finally, the expletive can be dropped in one place in Groningen, i.e., Nieuw-Scheemda.

Map 49a can be compared to the older maps in Koelmans (1975), De Schutter (1989b) and Ryckeboer (1995). A comparison to Koelmans’ map would not be very instructive, though, since De Schutter deals with the same data in a much more detailed way, distinguishing exactly the same variants as map 49a. Both De Schutter’s map and map 49a show a similar image: on both maps, the d’r-area exceeds the other areas in size, including ’t-areas in the southeast and the southwest and a daar-area in the southeast. On both maps, there are attestations of Standard Dutch er (especially in the Netherlands), and of zero-expletives, in the north. The geographic distribution of daar differs slightly: De Schutter does not find instances of daar in Noord-Holland, but does do so in Zeeland, while the SAND corpus provides exactly the opposite picture. Also, the westernmost attestations of d’r are situated much more to the west on map 49a than on De Schutter’s map (cf. Ryckeboer 1995).

2.4.2  Expletive subject in inversion in existential clause  (map 49b)

Map 49b shows the variation in the form of expletives in sentences with inverted word order, which is investigated by means of the following telephone question: Gisteren zat ø / het / er /daar een inbreker in deze kast (‘Yesterday there was a burglar sitting in the closet’). Informants were asked explicitly which of the following expletives could be inserted in the sentences: ø , het / ‘t, er (or d’r / t’r) and daar. Het / ’t did not occur. In sentences with inverted word order, the difference between er, d’r and t’r is no longer visible, because of the presence of a –t on the verb zat (‘sat’), so map 49b only distinguishes zero, the weak forms er, d’r and t’r, and strong daar.

Map 49b shows that er (d’r / t’r) is used in the entire language area. Especially remarkable is the homogenous behaviour of West-Vlaanderen, an area in which het is used in sentences with regular word order. Daar occurs virtually throughout the language area, with Frans-Vlaanderen, West-Vlaanderen and the larger part of Oost-Vlaanderen as the exceptions. The density of the daar-attestations is particularly high in the east and the southeast. In addition, the expletive can be dropped at many sampling points. This is even more prevalent in the Netherlands than in Belgium.

2.4.3  Expletive subject following complementiser in existential clause  (map 50a)

Expletives can occur in sentence-initial position, following verbs, and following complementisers. Map 50a shows the form of the expletive following of (‘if’), in the following sentence from the telephone questionnaire ’t Is net of ø / er / het / daar een hond in deze kast zit (‘It looks as if a dog is sitting in the closet’), a sentence from the telephone questionnaire. Four possibilities were presented to the informants, who had to select the ones that were possible in their dialects. Het did not occur. Here too, er, t’r and d’r could not be distinguished, since the complementiser carries an inflectional –t in some dialects, or is followed – or replaced – by dat (‘that’) in others.

Map 50a is almost identical to map 49b. Er (d’r / t’r) is attested in the entire Dutch language area. In fact, in West-Vlaanderen and Oost-Vlaanderen, er (d’r / t’r) is the only possibility, with the exception of two eastern Oost-Vlaanderen sampling points. Outside West-Vlaanderen and Oost-Vlaanderen, daar occurs, but it needs to be mentioned that spontaneous attestations of daar outside Belgian Limburg and Vlaams-Brabant are rare: the answers to the translation sentence Het lijkt wel of er een man in de tuin staat (‘It seems as if a man is standing in the garden’) contain only 14 instances of daar, 11 of which are found in Belgian Limburg and the neighbouring part of Vlaams-Brabant. The same holds for the zero-attestations: the expletive can be dropped in a large part of the language area, but spontaneous translations without an expletive are often found in a smaller area: more than half of the 40 zero-translations of the expletive in Het lijkt wel of er een man in de tuin staat (‘It seems as if a man is standing in the garden’) come from either Belgian Limburg or Dutch Limburg.

When compared to map 49b, the number of zero-attestations on map 50a is rather low.

2.4.4  Expletive subject in inversion in impersonal construction  (map 50b)

In Standard Dutch, impersonal constructions use the expletive het (‘it’). However, in some dialects there is considerable morphological variation with respect to the expletive. In fact, the variation that is found seems to correspond to the variation between weak expletives like er and d’r which alternate with strong daar (‘there’). In some dialects, the position of het (‘it’) in clauses that, semantically, do not require subjects, is filled by dat (‘that’), originally a demonstrative. Informants were explicitly asked whether they could ask for the exact time using the test sentence Hoe laat is dat? (‘How late is it (lit. ‘that’)?’), a sentence that was part of the telephone questionnaire. The construction seems to be attested almost exclusively in Belgium. Apart from that observation, there is no clear geographic distribution: the attestations seem to be distributed almost randomly over the map.

2.5  Literature on subject pronouns and expletives

Aalberse, S. (2004). 'Waar bestu bleven? De verdwijning van het pronomen 'du' in een

taalvergelijkend perspectief.' Nederlandse Taalkunde 9. 231–252.

ANS = Haeseryn, W. e.a. (1997). Algemene Nederlandse Spraakkunst. Groningen:

Wolters-Noordhoff.

Bennis, H. (1983). 'De PRO-drop-parameter en subjektloze zinnen in het Nederlands.'

Spektator 12. 409–427.

—        (1985). 'Het-raising.' In H. Bennis & F. Beukema (eds.) Linguistics in the Netherlands. Dordrecht: Foris. 11–20.

—        (1986). Gaps and Dummies. Dordrecht: Foris.

—        (2004). 'Pronoms de la deuxième personne en néerlandais: Contrastes en forme et en interpretation.' Franco-British Studies 33/34. 10–22.

Blancquaert E. & W. Pée (eds.) (1925–1982). Reeks Nederlands Dialect-atlassen (RND).

Antwerpen: De Sikkel.

Bosch, P. (1983). Agreement and anaphora. London: Academic Press.

Bresnan, J. (2001). 'The Emergence of the Unmarked Pronoun.' In G. Legendre, J. Grimshaw

& S. Vikner (eds.) Optimality-Theoretic Syntax. Cambridge, MA: MIT Press. 113–142. (http://www-lfg.stanford.edu/bresnan/download.html)

Cornips, L. (2000). 'Over inherente normen binnen de eigen groep. Hun doen allemaal van

            die rare woorden, weet je.' Taal en Tongval 52. 47–61.

Cysouw, M. (2003). The paradigmatic structure of person marking. Oxford: Oxford University Press.

Dahl, Ö. (1990). 'Standard Average European as an exotic language.' In J. Bechert, G. Bernini & C. Buridant (eds.) Toward a typology of European languages. Berlin/New York: Mouton de Gruyter. 3–8.

De Haan, G.J. (1994). 'Inflection and Cliticisation in Frisian, –sto, –ste, –st.' North Western

 European Language Evolution (NOWELE) 23. 75–90.

—        (1997). 'Voegwoordcongruentie in het Fries.' In Hoekstra, Eric & Caroline Smits (eds.) Vervoegde voegwoorden. Amsterdam: P.J. Meertens-instituut. 50–67.

De Rooij, J. (1990). 'Over hun en hen, en hun. Vorm en functie van de niet-gereduceerde

voorwerpsvormen van het persoonlijk voornaamwoord in de derde persoon meervoud, in standaardtaal, ouder Nederlands en dialect.' Taal en Tongval 42. 107–147.

—        (1991). 'Regionale variatie in het gebruik van er II en III.' Taal en Tongval 43. 18–46, 113–136.

De Schutter, G. (1987). 'Morfologische categorieën van 'persoonspronomina' in de

Nederlandse dialecten.' Taal en Tongval Themanummer 1: Morfologie. 43–54.

—        (1989a). Pronominale clitica in de Nederlandse dialecten. Antwerpen: Antwerp Papers in Linguistics.

—        (1989b). 'Presentatieve constructies in het Nederlands. Een vergelijking van RND en materiaal-Willems.' In J. Goossens, J. Taeldeman & A.A. Weijnen (eds.) Taal en Tongval Themanummer 2: 100 jaar enquête Willems. 83–94.

—        (1992). 'Persoonlijke voornaamwoorden in het Antwerps.' In H. Bennis & J.W. De Vries (eds.) De binnenbouw van het Nederlands: een bundel artikelen voor Piet Paardekooper. Dordrecht: ICG publications. 343–352.

—        (1994). 'Voegwoordflectie en pronominale clitisering waarin Vlaams en Brabants bijna elkaars tegengestelden zijn.' Taal en Tongval 46. 108–131.

—        (1997). 'Incorporatie-in-C in de Vlaamse en Brabantse dialecten.' In E. Hoekstra & C. Smits (eds.) Vervoegde voegwoorden. Amsterdam: P.J. Meertens-instituut. 31–49.

De Schutter, G. & J. Taeldeman (1986). 'Assimilatie van stem in de zuidelijke Nederlandse

dialekten.' In M. Devos & J. Taeldeman (eds.) Vruchten van zijn akker: Opstellen van (oud)medewerkers en oud-studenten voor Prof. V.F. Vanacker. Gent: Seminarie voor Nederlandse Taalkunde en Vlaamse Dialectologie. 91–133.

De Vogelaer, G. & A. Neuckermans (2002). 'Subject doubling in Dutch: a dialect

phenomenon in cross-linguistic perspective.' Sprachtypologie und Universalienforschung (STUF) 55. 234–258.

Devos, M. (1986). 'Het persoonlijk voornaamwoord 2e pers. enk. in het Westvlaams:

geografie en historiek.' In M. Devos & J. Taeldeman (eds.) Vruchten van zijn akker: Opstellen van (oud-) medewerkers en oud-studenten voor Prof. dr. V.F. Vanacker. Gent: Seminarie voor Nederlandse Taalkunde en Vlaamse Dialectologie. 167–191.

Devos, M. & W. Vandeweghe (2002–2003). 'Pronominale substitutie na voorzetsel in het

West-Vlaams.' Taal en Tongval Themanummer 15/16: Dialectsyntaxis in bloei. 209–236.

De Wilde-Van Buul, G. (1942–1943). 'Het enclitisch pronomen personale van de tweede en

derde persoon singularis in het Rotterdams.' Tijdschrift voor Nederlandse Taal- en Letterkunde 62. 290–301.

FAND = J. Goossens, J. Taeldeman & G. Verleyen (1998: deel I; 2000: deel II + III; in

voorbereiding: deel IV). Fonologische Atlas van de Nederlandse Dialecten. Gent: Koninklijke Academie voor Nederlandse Taal- en Letterkunde.

Geluykens, R. (1992). From discourse process to grammatical construction: on left-

dislocation in English. Amsterdam & Philadelphia: John Benjamins.

Givón, T. (1976). 'Topic, pronoun, and grammatical agreement.' In C.N. Li (ed.) Subject and

Topic. New York: Academic Press. 149–188.

Goossens, J. (2000). 'Onslieden en ons leden.' In S. Gillis, J. Nuyts & J. Taeldeman (eds.)

Met taal om de tuin geleid: een bundel opstellen voor Georges De Schutter ter gelegenheid van zijn pre-emeritaat. Antwerpen: UIA. 187–191.

Goossens, J., J. Taeldeman & A.A. Weijnen (eds.) (1989). Taal en Tongval themanummer 2:

Honderd jaar enquête Willems.

Grange, C. & Haegeman, L. (1989). 'Subordinate clauses: adjuncts or arguments: The status

of 'het' in Dutch.' In D. Jaspers et al. (eds.) Sentential complementation and the lexicon. Dordrecht/Providence: Foris. 155–171.

Gysseling, M. (1966). 'Het persoonlijk voornaamwoord 2e persoon, vooral in de 13de eeuw.'

Verslagen en Mededelingen van de Koninklijke Vlaamse Academie voor Nederlandse Taal en Letterkunde 1966. 195–212.

Haegeman, L. (1992). Theory and description in generative syntax: A case study in West

            Flemish. Cambridge: Cambridge University Press.

Haspelmath, M. (2001). 'The European linguistic area: Standard Average European.' In M. Haspelmath et al. (eds.) Language Typology and Language Universals/ Sprachtypologie und sprachliche Universalien/ La typologie des langues et les universaux linguistiques. Ein internationales Handbuch / An International Handbook Vol 2. Berlin/New York: de Gruyter. 1492–1510.

Hoekstra, J. (1991). 'Expletive der and resumptive pro in Frisian.' Leuvensche Bijdragen 80.

61–80.

Hol, A.R. (1936). 'De noordgrens van het pronomen gij.' Tijdschrift voor Nederlandse taal-

en letterkunde 55. 225–240.

—        (1951). 'Hij.' Nieuwe Taalgids 44. 179–181.

—        (1953). 'Dialect-grenzen in Midden-Gelderland.' Taal en Tongval 5. 71–92.

Howe, S. (1996). The personal pronouns in the Germanic languages: a study of personal

pronoun morphology and change in the Germanic languages from the first records to the present day. Berlin: de Gruyter.

Karsten, G. (1939). 'Hem en Hun als onderwerp.' Nieuwe Taalgids 33. 369–372.

Kern, J.H. (1911). 'Is de beleefdheidsvorm U een verbastering van U.E.?' Nieuwe Taalgids 5.

            121–133.

Kloeke, G.G. (1920). 'De aanspreekvormen in de dialecten onzer Noordelijke provinciën.'

Tijdschrift voor Nederlandse Taal- en Letterkunde 39. 238–273.

—        (1941). 'Over 'jullie' en enige andere pronomina.' Nieuwe Taalgids 35. 161–170.

—        (1948). 'Uit de voorgeschiedenis van het beleefde pronomen 'u'.' In Th.C.Van Stockum, H.W.J. Kroes & D.J.C. Zeeman (eds.) Verzamelde opstellen, geschreven door oud-leerlingen van prof. dr. J.H. Scholt, aangeboden door vrienden, leerlingen en oud-leerlingen ter gelegenheid van zijn aftreden als hoogleraar aan de UVA. Amsterdam: Universiteit van Amsterdam. 15–21.

Koelmans, L. (1968). 'Over het enclitisch pronomen ­–en.' Taal en Tongval 20. 17–24.

—        (1972). 'Over Nederlands (d)er, vroeger en nu I.' Nieuwe Taalgids 65. 191–198.

—        (1975). 'Over het adverbium (d)er in de Nederlandse Dialecten.' In R. Jansen-Sieben, S. de Vriendt & R. Willemyns (eds.) Spel van Zinnen. Album A. van Loey. Bruxelles: editions de l'université de Bruxelles. 177–187.

Kooiman, K. (1950). 'Hij.' Nieuwe Taalgids 43. 324–333.

—        (1969). 'Hun = Zij.' Nieuwe Taalgids 62. 116–120.

Kuijper, J. (1971–1972). 'U.' Spektator 1. 198–201.

Paardekooper, P.C. (1969). 'Een Fransvlaams-Zeeuws parallel: ons 'wij'.' Taal en Tongval

21. 211–215.

Pauwels, J.L. (1958). Dialect van Aarschot en omstreken. Tongeren: Belgisch

Interuniversitair Centrum voor Neerlandistiek.

Ryckeboer, H. (1972). 'Zoe, persoonlijk voornaamwoord derde persoon vrouwelijk

enkelvoud, een uitstervend relikt.' Taal en Tongval 24. 67–70.

—        (1995). ' 't Viel een neusdoek brachten mijn zatte': dialectverlies in West- en Oost-

Vlaanderen.' In R. Belemans &  H.H.A. van de Wijngaard (eds.) Het Dialectenboek 3: Dialect in beweging: 100 jaar na de enquêtes van Willems en Aardrijkskundig Genootschap. Waalre: Stichting Nederlandse Dialecten. 237–252.

Schuurmans, N.J. (1975). Verbindingen met specifiek enclitische pronomina in het

Westbrabants. Themanummer van de Mededelingen van de Nijmeegse centrale voor dialect- en naamkunde. Nijmegen: Nijmeegse centrale voor dialect- en naamkunde.

Siewierska, A. (1999). 'From anaphoric pronoun to grammatical agreement marker: why objects don’t make it.' Folia Linguistica 33. 225–251.

Stevens, A. (1949). 'Pronominale isomorfen in Belgisch-Limburg I.' Taal en Tongval 1.

132–154.

Stroop, J. (1987). 'Enclitische verschijnselen in het Westbrabants.' Taal en Tongval 39.

121–140.

Swiggers, P. (1987). 'Voornaamwoorden met onderwerpsfunctie in Brabantse dialecten.'

Leuvense bijdragen 76. 159–170.

—        (1990). 'Een dimunitiefpronomen: emekes.' Taal en Tongval 42. 159–161.

Tiersma, P.M. (1999). Frisian Reference Grammar. Leeuwarden: Fryske Akademy.

Vanacker, V.F. (1978). 'Het of der.' Nieuwe Taalgids 71. 616–621.

Vandekerckhove, R. (1993). 'De subjectvorm van het pronomen van de 2e p. ev. in de

            Westvlaamse dialecten.' Taal en Tongval 45. 173–183.

Van den Berghe, G. (1951). 'Het subject pronomen van de 2^e persoon pluralis, enclise.' Taal

en Tongval 3. 109–116.

Van den Toorn, M.C. (1959). 'De herkomst van het enklitisch pronomen ie, resp. die/tie.'

Nieuwe Taalgids 52. 85–90.

—        (1977). 'De problematiek van de Nederlandse aanspreekvormen.' Nieuwe Taalgids 70.

520–540.

Van der Meer, G. (1991). 'The 'conjugation' of subclause introducers: Frisian –st.' North

Western European Language Evolution (NOWELE) 17. 63–84.

Van Hout, R. (1996). 'Waar kwamen hun onze taal binnen?' In R. van Hout & J. Kruijsen

            (eds.) Taalvariaties. Toonzettingen en modulaties op een thema. Dordrecht: Foris.

143–156.

—        (1999). 'Waarom veroveren 'hun' onze taal? Sociale en taalkundige verklaringen voor de opkomst van een subjectpronomen.' Toegepaste taalwetenschap in artikelen 62–2: Thema's en trends in de sociolinguïstiek 3. 75–86.

Van Koppen, M. (2005). One probe, two goals. Aspects of Agreement in Dutch dialects.

         Diss. Leiden University. LOT Dissertations.

Van Loey, A. (1948). Middelnederlandse spraakkunst I, Vormleer. Groningen: J.B. Wolters.

—        (1954). 'Het pronomen 'ten'.' Taal en Tongval 6. 171–174.

Van Riemsdijk, H. (ed.) (1999). Clitics in the Languages of Europe. Berlin & New York:

Mouton de Gruyter.

Verdenius, A.A. (1923). 'De ontwikkelingsgang der Hollandse voornaamworden je en jij.'

Tijdschrift voor Nederlandse Taal- en Letterkunde 43. 109–116.

—        (1926). 'Over de aanspreekvorm ie (i–j) in onze Oostelijke provinciën.' Tijdschrift

voor Nederlandse Taal- en Letterkunde 45. 23–32.

Visser, W. (1988). 'In pear klitisearringsferskynsels yn it Frysk.' In S. Dyk & G. de Haan

(eds.) Wurdfoarried & Wurdgrammatika. Ljouwert: Fryske Akademy.

Vossen, A.F. (1963). 'Grammatisch geslacht en aanspreekvormen.' Taal en Tongval 15.

63–64.

Wales, K. (1996). Personal pronouns in present-day English. Cambridge: Cambridge

 University Press.

Weijnen, A.A. (1966). Nederlandse dialectkunde. Assen: Van Gorcum.

—        (1971). Schets van de geschiedenis van de Nederlandse syntaxis. Assen: Van Gorcum.

Willemyns, R. (1979). 'Vorm en functie van de aanspreekvormen in West- en Frans

Vlaanderen.' Handelingen van de Koninklijke Commissie voor Toponymie en Dialectologie 53. 171–193.

Zwicky, A. (1977). On Clitics. Bloomington: Indiana University Linguistics Club.