DynaSAND

Syntactic Atlas of Dutch Dialects – Commentary – Volume I

Table of contents Volume I

3          Subject doubling and cliticization after ja and nee

3.1       Subject doubling

3.1.1     Introduction

            3.1.1.1  Doubling and tripling

            3.1.1.2  Doubling and tripling of personal subject pronoun

            3.1.1.3  Doubling of other nominal phrases

            3.1.1.4   Historical development

3.1.2     Discussion of the literature

3.1.3     Discussion of the maps

            3.1.3.1  Subject doubling first person singular

            3.1.3.2  Subject doubling first person plural

            3.1.3.3  Subject doubling second person singular

            3.1.3.4  Subject doubling second person plural

            3.1.3.5  Subject doubling third person singular masculine

            3.1.3.6  Subject doubling third person singular feminine

            3.1.3.7  Subject doubling third person singular neuter

            3.1.3.8  Subject doubling third person singular, doubling of lexical noun phrases

            3.1.3.9  Subject doubling third person plural

            3.1.3.10 Subject doubling; synthesis

3.2       Subject clitics following ja (‘yes’) and nee (‘no’)

3.2.1     Introduction

3.2.2     Discussion of the literature

3.2.3     Discussion of the maps

            3.2.3.1  Clitics following ja ‘yes’/ nee ‘no’, first person singular

            3.2.3.2  Clitics following ja ‘yes’/ nee ‘no’, second person singular and plural

            3.2.3.3  Clitics following ja ‘yes’/ nee ‘no', third person singular masculine

            3.2.3.4  Clitics following ja ‘yes’/ nee ‘no', third person singular feminine

            3.2.3.5  Clitics following ja ‘yes’/ nee ‘no’, third person singular neuter

            3.2.3.6  Clitics following ja ‘yes’/ nee ‘no’, first person plural

            3.2.3.7  Clitics following ja ‘yes’/ nee ‘no’, third person plural

            3.2.3.8  Generalised jaat

3.3       Literature on subject doubling and subject clitics following ja (‘yes’) and nee (‘no’)

3  Subject doubling and Subject clitics following ja (‘yes’) and nee (‘no’)

3.1  Subject doubling

3.1.1  Introduction

3.1.1.1  Doubling and tripling

Subject doubling is a phenomenon that neither occurs in Standard Dutch nor in the northern dialects. Yet, southwestern and western-central dialects can double the subject NP of the sentence by means of a subject pronoun which agrees with the subject NP in person, number and gender. The phenomenon is called subject doubling, as for example in (1).

(1)       a.         ‘k         Ga–ik              ook      mee.

                        I _weak   go–I_strong          too       with

                        'I’m  coming too.'

            b.         Ga–ze              zij       ook mee?

                        Goes–she_weak she_strong  too with?

                        'Is she coming too?'

            c.         Ik denk            da–je                gij                    ook mee kunt.

                        I think                         that–you_weak   you_strong           too with can

                        'I think you can come too'

            d.         Hij is groter     dan–ek             ik.

                        he is taller        than–I_weak        I_strong

                        'He is taller than me.'

As the examples show, doubling occurs in four syntactic constellations, viz. (a) in main clauses with regular order, (b) in sentences with inversion, (c) after the complementiser in sub-clauses, and (d) after the comparative complementisers als / of / dan in elliptic sentences.

            In type (a), i.e., main clauses with regular order, the subject cannot only be doubled but also tripled, as for example in (2).

(2)       a.         ‘k         Ga–'k              ik        ook mee.

                        I_weak     go–I_weak                       I_strong   too with

                        'I’m coming too.'        

            b.         Ze                    gaat–ze                        zij                   ook mee.

                        she_weak             goes–she_weak    she_strong            too with

                        'She’s coming too.'

Doubling and tripling are by far most frequent in sentences containing a personal pronoun as subject NP, as in the examples above. Section 3.1.1.2 deals with this type of construction more extensively. Constructions in which another pronominal subject (i.e. not a personal pronoun) or a lexical subject NP is doubled, are less widely spread; they are discussed in 3.1.1.3.

3.1.1.2  Doubling and tripling of personal subject pronoun

SAND lists a number of doubling and tripling constructions in sentences with a personal pronoun as subject, and the following is a survey of these constructions. The survey does not deal with differences in the geographic distribution of the various doubling constructions. These differences are discussed in 3.1.3. For clarity’s sake, table 1 does not reproduce the full range of pronominal variation observed: regionally different realisations are represented by one single, slightly standardised form, which may therefore stand for several variants. Thus, the first person plural wijlie, for example, represents variants such as wulle, wijle, wijder, mijnder, etc. Distinctions that the table does make are those between strong and weak forms and between subject forms and object forms. A detailed survey of pronominal variation is given in chapter 2. What is also lacking from the survey are examples of subject doubling after complementisers [p 42 t/m 54]; it is to be noted, however, that subject doubling in the sub-clause uses exactly the same combinations of pronouns as subject doubling in inversion.

                                        Table 1: doubling of personal subject pronouns

Doubling occurs in all combinations of person and number, whereas tripling is lacking in the second person singular and plural.

            There are two exceptions to the categorial status or word class of the personal pronoun. The first is that the strong neuter pronoun of the third person singular is borrowed from the paradigm of the demonstrative pronouns, as in (3).

(3)       a.         Dat                  is–hij               nu        al          veel     te         laat      [doubling]

                        that_strong           is–he_strong         now     already  much too       late

                        'It is already much too late now.'

            b.         Dat is–'t          het       nu        al         veel te laat                               [tripling]

                        that is–it          it          now     already much too late

                        'It is already much too late now.'

Secondly, apart from the doubling of the subject itself there is also doubling of the expletive element which in subjectless sentences functions as subject or which is combined with an indefinite subject NP. In some dialects, this element is adverbial, like its counterpart er in Standard Dutch, but in the southwestern dialects it occurs in certain types of sentences as the personal pronoun of the third person neuter singular.

(4)       a.         Er [adverbial]  wordt  hij        hier      elke avond       gedanst.

                        there_weak          is         he_strong here     every night      danced

                        'There is dancing here every night'

            b.         Komt   er [adverbial]               hij        daar     veel volk          naartoe?

                        comes  there                            he        there    many people   to that?

                        'Do many people go there?'

            c.         't [pronominal] Is–hij daar     geen woord      over     gezegd.

                        it                      is–he    there    no word           about   said

                        'No word was said about it.'

Subject doubling usually consists of a combination of weak and strong pronouns, as can be seen in columns 2 to 4 of Table 1. In this combination, the doubled subject is always weak; the doubling pronoun is strong and, in the case of tripling, it is preceded by a second weak form. Only one exception to this rule has been found, viz. the East-Flemish construction wij ga–me(n), in which the first pronoun is strong and the second weak, though the question can be raised whether the post-verbal clitic of this construction is not a conjugational affix rather than a pronoun. Postverbal –men not only combines with strong pronouns but with weak ones as well, and it is so frequently used in the dialects concerned that it should be considered as obligatory in some types of sentences. Both characteristics mentioned are rather associated with conjugational affixes than with pronouns. Setting aside this one deviation, which will be gone into in 3.1.3.2.3, we could formulate as a general hypothesis that in subject doubling a weak pronoun always precedes a strong pronoun.

            However, the first column of Table 1 shows that sentences with regular order can also have doubling with two strong pronouns, which seems to argue against the general hypothesis put forward above. Although the term ‘strong’ is appropriate here, since neither pronoun has the neutral vowel schwa, SAND oral sources show that one of the pronouns, notably the first, has a weaker realisation than the other. Consequently, along with Nuyts (who also perceives a prosodic difference between the two strong forms of the pronoun (Nuyts 1995:53)), we can put forward a more flexible formulation of the above generalisation, which would then account for all the cases observed: in subject doubling the second pronoun is always heavier and is always weightier than the first. As the following will make clear, the combination of two strong pronouns is not only restricted as far as its syntactic constellation is concerned (occurring in regular order only); its geographic distribution too is more limited than that of the combinations with both strong and weak forms.

3.1.1.3  Doubling of other noun phrases

It is not only personal pronouns that can be doubled: so can demonstrative and indefinite subject pronouns as well as lexical subject NPs, as is illustrated in examples (5) and (6). The geographic distribution of these doubling constructions is, however, much smaller than of those with a personal subject pronoun.

(5)       a.         De dien            is         hij/ het                         daar niet bij betrokken.

                        [demonstrative pronoun]

                        that one           is         he/it                 there not in involved

                        'He is not involved in that.'

            b.         Niemand          was      hij / het                       content.           [indefinite pronoun]

                        nobody            was      he/it                 satisfied

                        'Nobody was satisfied'

(6)       a.         Die man is       hij / het            hier al              twee keer geweest. [lexical NP]

                        that man is      he/it                 here already     two times been

                        'That man has been here two times already.'

            b.         Het weer          is         hij        daar dikwijls slecht.    [idem]

                        the weather     is         he        there often bad

                        'The weather is often bad there.'        

            c.         Is–t hij  Pol      onze afspraak vergeten?                     [lexical NP, proper name]

                        is–it he Pol      our appointment forgotten

                        'Has Paul forgotten our appointment?'

            d.         Anna    komt    hij / het            hier zelden.                  [idem]

                        Anna   comes  he/it                 here rarely

                        'Anna rarely comes here.'

In the three sentences of (6), the subject is a definite NP. Doubling occurs in sentences with indefinite subject NP’s as well, though not of the subject but of the expletive er or ‘t, as exemplified above in examples (4b) and (4c).

            The doubling constructions using non-personal subject pronouns and those with lexical NPs are largely ana­logous, both with regard to word order and to the categorial and morphological nature of the doubling pronoun. Their patterns deviate in several respects from doubling constructions with a personal pronoun in the standard subject position (Table 1). The following table presents a survey of the doubling constructions of demonstrative and indefinite pronouns and of definite lexical NPs, respectively, which are found in the SAND material and/or in the literature concerned.

                                    Table 2: doubling other pronouns and lexical NPs

On a number of points, the doubling constructions in Table 2 deviate from those with a personal subject pronoun of the third person in Table 1. The deviations bear on the relation between strong and weak forms and on agreement.

(i) Relation between strong and weak forms

As a rule the doubling pronouns in the constructions in Table 2 are strong. Since the doubled subject too is, by definition, a strong form, the generalisation suggested in section 3.1.1.2 (viz. that doubling constructions usually have a combination of strong and weak forms and that the first subject marker is always weaker than the second) is no longer applicable. Both subject markers are strong forms and there is no indication that one of the two forms is stronger or weaker than the other.

(ii) Agreement

Whereas personal subject pronouns are always doubled by pronouns with the same combination of person, number and gender, this rule of agreement does not always apply to doubling pronouns of other pronouns and of lexical NPs. There are two types of combinations which do not stick to the principle of gender agreement. The first is the combination in which a pronoun or a lexical NP with a male or female referent is doubled by a neuter personal pronoun of the third person singular, as for example in (5) and (6a,c,d) above. In the second type, subject NPs with a female or neuter referent are doubled by masculine pronouns, e.g., in (6b/d).

            Number agreement is absent in constructions in which a plural NP is doubled by the masculine or neuter personal pronoun of the third person singular instead of by the normal plural pronoun zij(lie). This is demonstrated in (7).

(7)       a.         De bakker en de slager           zijn      hij / het            al jaren             vrienden.

                        the baker and the butcher        are       he/it                 already years   friends

                        'The baker and the butcher have been friends for many years.'

• Frieda en Magda        zijn      hij / het            nergens te vinden.

Frieda and Magda       are       he/it                 nowhere to find

'Frieda and Magda are nowhere to be found.'

These remarks suggest that the pronouns of the third person singular, masculine and neuter, when used as doubling pronouns, lose their full meaning, their gender and number features being affected so much so that the pronouns involved can refer to all third-person referents, irrespective of whether these referents are male, female or neuter and singular or plural.

3.1.1.4  Historical development

The historical development of subject doubling has scarcely been studied, which bears no wonder, as the phenomenon is widely absent from Dutch historical texts. Absence from the sources does however not imply that the construction never existed in former times. Maybe it was characteristic of regional (informal) spoken language, a register of language of which written documents bear very little witness. The only known text corpus that does contain historic instances of doubling consists of a collection of legal reports from the city of Aalst in Oost-Vlaanderen, dating from the period between the beginning of the 15^th century and the end of the 17^th century. These reports were analysed by Vanacker in the framework of his study on the syntax of the Aalst dialect (Vanacker 1963). In his text collection, containing literal renderings of testimonies, Vanacker finds seventeen instances of soubject doubling, in fifteen of which a second person pronoun is doubled, and in the other two the first person singular pronoun. The earliest attestation, dating from 1496, runs as follows: “wil–de ghy zulck(e) zaken doen als…” (do–you_weak you_strong such things will do as… 'Do you want to do such things as…'; Vanacker 1963:320).

            In the literature on Dutch dialectology, hypotheses about the history of subject doubling are sparse indeed. There is only a brief speculative comment in De Meersman (1985); the occurrence of doubling with invariably a strong pronoun in post-verbal position, leads him to suggest that Dutch may once have had a VSO word order.

3.1.2  Discussion of the literature

3.1.2.1  Occurrences of subject doubling

The literature concerned mentions examples of subject doubling in quite a number of places and areas throughout the southern Dutch-speaking region. A modest map is given by Paarde­kooper (1952). De Vogelaer & Neuckermans (2002) too include a few maps. The various types of subject doubling are not all equally widely spread. West- and East-Flemish paradigms have virtually all the types of doubling and tripling mentioned in Table 1. Smessaert (1995:46, 1996:243/248) lists full paradigms for 'West-Flemish' in general, Haege­man (1992:60–61) for Lapscheure (West-Flemish), Vallaeys (1997:57–60) for Poperinge (West-Flemish), De Geest (1990:111) for Gent and De Meersman (1985:124) for Gijzegem (East-Flemish).

            By contrast, descriptions of dialects spoken in the Belgian provinces of Antwerpen and Brabant mention a limited number of types of doubling and tripling; Pauwels (1958:339) for Aarschot (Vlaams-Brabant), Swiggers (1987:164–65) for Leuven (Vlaams-Brabant), Nuyts (1995:45/52) for the city of Antwerpen and De Vriendt (2003:75) for Brussels (Vlaams-Brabant).

            The borders between dialects with a 'Flemish' full paradigm and those with a 'Brabant' incomplete paradigm do not overlap exactly with the borders between the provinces. The west of Vlaams-Brabant mostly links up with Vlaanderen. Furthermore there are discrepancies between main clause and sub-clause in Brabant; De Schutter (1994:121–22) notes that doubling rarely or never occurs in the second person in sub-clauses, whereas doubling in the first person singular, though it does exist, is no longer to be considered as a case of doubling. The reason is that the form kik (<weak k + strong ik) in Brabant dialects can no longer be split up (De Schutter 1994:116). It seems to be possible, then, for subject doubling to give rise to new pronouns. Therefore Pauwels (1958) and Nuyts (1995), among others, analyse egij and egullie as pronouns of the second person singular and plural respectively, rather than as doubling constructions.

            In the Netherlands, examples of subject doubling have also been found, though to an even smaller extent than in Belgian Brabant. In the Dutch province of Noord-Brabant doubling only occurs with second person pronouns in inversion and after complementisers in sub-clauses (Schuurmans 1975:40 and 75, Stroop 1987:122). For the province of Zeeland, the Woordenboek der Zeeuwsche Dialecten (WZD 1974:371) mentions cases of doubling in regular order only and moreover only in the first person singular. This, however, conflicts with the findings of G. Will, who has examined the phenomenon in all towns and villages of Zeeuws-Vlaanderen, using the recordings of spontaneous conversations filed in the Dutch Language Department of Gent University, recordings made in the sixties of the 20th century. Will has found examples of doubling in all villages of Zeeuws-Vlaanderen except in the area around the town Axel (Will 2004:232–272). It is to be noted that doubling is rapidly disappearing in Zeeuws-Vlaanderen; nowhere are any full paradigms to be found anymore.

            Subject doubling also occurs in other Indo-european languages, such as French(-canadian) (Nasdadi 1995), some Rhaetoromanic dialects (Haiman 1991) and a number of Bavarian dialects (Bayer 1984). The word order within doubling constructions in all these languages does not correspond to the one found in Dutch dialects.

3.1.2.2  Generalisations and interpretations

Regarding the distribution of strong and weak forms in doubling constructions, Haegeman (1992:61) and De Geest (1990:111–12) have put forward the following generalisation: subject doubling always uses a weak pronoun (or two in tripling) and a strong one, and there are no doubling constructions with two strong pronouns or with a lexical NP. However, other researchers have found the two latter types. Constructions with two strong pronouns are described in the studies of Brabant dialects (Pauwels 1958, Swiggers 1987 and Nuyts 1995) and in Vallaeys (1997) for the West-Flemish dialect of Poperinge. Nuyts (1995:53) proposes the generalisation that the second pronoun is always weightier than the first. It is a generalisation which is based exclusively on data from the dialect of Antwerpen city, but it proves to work quite well for all other data as well, with one exception, mentioned above, viz. the East-Flemish wij ga–me(n).

            In their discussion on East-Flemish wij ga–me(n) De Vogelaer & Neuckermans (2002) point out that –me(n) shows a syntactic distribution proper to conjugational affixes rather than to pronouns, as it also occurs in sentences having a weak preverbal subject. The final –n that is sometimes attached to –me is analysed by De Schutter (1994:118–119) as a conjugational affix, resulting from incorporation of the pronoun in a combination of a verb or a complementiser and a conjugational –n. According to De Schutter, incorporations of this kind are not unusual; in East-Flemish they also appear in some past tense conjugations, and similar combinations are found in other areas such as Groningen and Zuid-Holland.

            Most studies focus on the synchronic conditions in which subject doubling occurs. For the most part, sentences with and without subject doubling alternate freely, so that the question arises of the motives that speakers have for deciding on a construction with subject doubling. Emphasis on the subject has often been put forward as the decisive factor (Vanacker 1948:47, Pauwels 1958:339, Schuurmans 1975:40, Willemyns 1979:191), but other pragmatic elements also seem to play a significant role, such as the information structure of the sentence, the speaker’s emotional involvement in what the sentence is about and his or her empathy with either the listener or the referent of the pronoun (Vandekerckhove 1993, Nuyts 1995, Vandeweghe 2000).

            Looking at the differences in geographic distribution between the various doubling constructions with different subject types, De Voge­laer & Neuckermans (2002:247–248) hypothesise that the reason why NPs are easily doubled stems from their categorial status (pronoun against noun and common noun against proper name) and from their features of person and number. They thus perceive a relation between the geographic distribution of a given NP-type and the rank taken by that NP-type on the universal ‘person hierarchy’ and the ‘NP-type hierarchy such as they are defined in language typology (e.g., Croft 1990:127). In the former hierarchy, the first and the second person rank higher than the third, in the latter hierarchy pronouns rank first, followed by nouns, among which proper names precede common names. The higher the rank of an NP-type on a hierarchic scale, the more widespread its occurrence in doubling constructions is supposed to be.

            The existence of subject doubling may indicate that the weak pronouns within doubling combinations are developing into inflectional affixes. The insight that formal reduction and semantic bleaching can make pronouns shift towards conjugational affixes dates from as early as 19^th-century linguistics (see Ariel 2000 for a historical overview), and was strongly supported in the 1970s by Givón (1976). Recent research on grammaticalisation of pronouns has been carried out by Siewierska (1999).

            For synchronic syntactic analyses of subject doubling in a generative framework, see        

Haegeman (1990), Zwart (1993) en Van Craenenbroeck & van Koppen (2002a,b).

3.1.3  Discussion of the maps

3.1.3.1  Subject doubling first person singular

3.1.3.1.1  Subject doubling 1 singular, overview  (map 52a) (map in dynaSAND)

Map 52a presents the distribution of subject doubling in its four possible constructions, as illustrated in the following four examples:

(8)       a.         ‘k         Mag (–ek         ik        dat weten.       [regular order]

                        I_weak     may (–I_weak )   I_strong    that know

                        'It is alright for me to know that'

            b.         Mag–(e)k                    ik         dat weten?       [inversion]

                        may–I_weak                    I_strong    that know

                        'Is it alright for me to know that?'

            c.         ... dat   (–e)k    ik        dat mag weten            [sub-clause after complementiser]

                        … that I_weak     I_strong    that may know

                        '… that it is alright for me to know that'

• Hij is groter     dan–ek             ik.                   [after a comparative complementiser]

he is taller        than I_weak         I_strong

'He is taller than me.'

In the SAND project, the construction was collected by means of six questions: three translation questions (9a–c) and three ‘this occurs’ questions (9d–f).

(9)       a.         Als ik zuinig leef,        leef ik zoals    mijn ouders willen.

                        if    I  frugally live,      live I    such as my parents want

                        'If I live frugally, I live the way my parents want me to live.'

            b.         'k Geloof         dat ik groter     ben      dan hij.

                        I   think           that I  taller     am       than he

                        'I think I am taller than him.'

c.         Ze        gelooft             dat       jij         eerder thuis    bent     dan ik.

she       believes           that      you      earlier  home   are       than I

                        'She thinks you are home earlier than me.'

            d.         Jan       en        ekik                 hebben dat gedaan.

                        Jan       and      I_weak  I_strong       have     that done

                        'Jan and I have done that.'

            e.         Mag     ekik                 daar     ook      van      proeven?

                        may     I_weak  I_strong       there    too       of         taste

                        'May I too have a taste of that?'

            f.          'k         Zal       (ek)ik               't          wel      krijgen.

                        I_weak     will      (I_weak) I_strong     it                      get

                        'You’ll see I will get it.'

Doubling constructions in regular order, on the one hand, and in inversion and after complementisers, on the other, by and large have the same geographic distribution. Overall, their area of distribution comprises Frans-Vlaanderen, the provinces of West- and Oost-Vlaanderen, the province of Antwerpen with two places in the west of Limburg (Lommel and Tessender­lo) and Vlaams-Brabant apart from its eastern border area. Nevertheless, there are small deviations in the areas taken by the two constructions. Strikingly, the eastern border area of Vlaams-Brabant does not have the construction in inversion and after a complementiser, though it does have it in regular order.

Doubling constructions after a comparative complementiser are distributed less widely than the other syntactic constellations. The construction does not occur in Frans-Vlaanderen and in most of West-Vlaanderen, and it is unknown in a small number of localities scattered over the provinces of Antwerpen and Vlaams-Brabant.

3.1.3.1.2  Subject doubling 1 singular, variation in regular order  (map 52b)

In regular order the subject pronoun may be repeated by more than one doubling pronoun. As the map shows, constructions with one duplicating form, e.g., 'k ben ik ('I am I'), are geographically confined to the north and the west of West-Vlaanderen and the southeastern tip of Vlaams-Brabant. In by far the largest part of the area where this construction is found, tripling is the general rule, e.g., as 'k zal–(e)k ik ('I will I I'). Quadrupling is found one time: 'k zal–k het–ek ik ('I will I it I I') (Mere, Oost-Vlaanderen).

            The question can be raised if triple constructions with the postverbal clusters 'k ik and ek ik as doubling pronouns may be said to still have three free pronouns. There are actually indications that such clusters are no longer felt as combinations of pronouns, but rather as one single pronoun. The latter analysis seems to be borne out by the fact that the cluster (e)k ik (I_weak I_strong ) is rarely split up by a weak object pronoun, as in the example in (10).

• 'k         Zal–k               het       ik         wel      doen.

                        I_weak     will–I_weak         it          I_strong    affirm do

                        'Let me do it.'

This order is acceptable in four localities only, viz. Sint-Laureins and Koewacht in Oost-Vlaanderen and Herne and Overijse in Vlaams-Brabant.

3.1.3.2  Subject doubling first person plural

3.1.3.2.1  Subject doubling 1 plural  (map 53a) (map in dynaSAND)

The constructions whose distribution is presented in map 53a are illustrated in the following examples:

(11)     a.         We /wij            mogen (–me)               wij       dat  weten.                  [regular order]

                        we_weak /we_strong            may     (–we_weak )        we_strong that  know

                        'It is alright for us to know that.'

            b.         Mogen –me     wij                   dat wel            weten?                         [inversion]

                        may     –we_weak  we_strong          that affirm     know

                        'Do you think it is alright for us to know that?'

            c.         ... da–me          wij       dat       mogen weten [sub-clause after complementiser]

                        …that–we_weak             we_strong that     may     know

                        '… that it is alright for us to know that'

            d.         Hij is groter     dan–me            wij.                  [after comparative complementiser]

                        he is taller        than–we_weak    we_strong

                        'He is taller than us.'

The SAND data reflected in the map are taken from the answers to four translation questions (12 a–d) and to two ‘this occurs’ questions (12 e/f).

(12)     a.         Als      we       sober   leven,   leven    we       gelukkig.

                        if          we       frugally live,    live      we       happily

                        'If we live frugally, we live happily.'

            b.         Ze geloven       dat       wij       rijker    zijn      dan      zij.

                        they believe     that      we       richer   are       than     they

                        'They think we are richer than them.'

            c.         We geloven      dat       jullie    niet      zo        slim     zijn      als wij.

                        we believe       that      you      not       so        clever   are       as we

                        'We think you are not as clever as we.'

            d.         Mogen we wel            weten dat       wij       ook      gevraagd          zijn?

                        may     we affirm       know   that      we       too       asked               been

                        'Is it really alright for us to know that we too have been asked?'

            e.         We       zijn      (me)                 wij                   daar     nog      nooit    geweest.

                        we_weak             are       (we_weak )          we_strong                         there                never   been

                        'We have never been there.'

            f.          We       weten wullie              d'r        niks     van.

                        we_weak             know   we_strong             there    nothing  about

                        'We know nothing about it.'

Subject doubling of the first person plural is much less frequent than in the first person singular. Today it occurs in just a few places in the eastern half of Vlaams-Brabant and it is absent in a number of test places in the province of Antwerpen. By contrast, the two villages in Limburg where doubling of the pronoun of the first person singular is found, are also found back on map 53a.

            For the first person plural, the difference in geographic distribution between main clauses with regular order, on the one hand, and in inversion and after a complementiser, on the other, is considerably greater than for the first person singular. In the former syntactic constellation, doubling is the general rule in the Flemish provinces and is also frequently applied in the province of Antwerpen and in the western half of Vlaams-Brabant; in the eastern half of Vlaams-Brabant the construction is rare. By contrast, the SAND data show that doubling in inverted sentences and after a complementiser in sub-clauses is geographically confined to the Flemish dialects, i.e., to Frans-Vlaanderen and to West- and Oost-Vlaanderen. Doubling after a comparative complementiser also has a more limited geographic range than with the first person singular, this construction being found in Oost-Vlaanderen only.

            In Frans-Vlaanderen, not a single occurrence of doubling in regular order is found, but it is worth noting that an alternative construction is used. It puts the strong pronoun (in its object form) in left dislocation and has it followed, after a brief intonation pause, by the weak form, as in (13).

(13)     Nuus,               me       zijn      daar     nooit    geweest.

            us_strong ,            we_weak             are       there    never   been

            'We have never been there.'

3.1.3.2.2  Subject doubling 1 plural, variation in regular order  (map 53b)

Map 53b sketches the distribution of doubling and tripling as well as the relation between weak and strong pronouns in doubling constructions. Conspicuously, the area in which tripling occurs has shrunk considerably when compared with the map above, its range covering only Oost-Vlaanderen and Hulst in Zeeuws-Vlaanderen.

            It is also worth pointing that doubling sometimes uses two strong pronouns, as in (14).

(14)     Wij                   zijn      wij       waarschijnlijk             al         te         laat.

            we_strong             are       we_strong              probab            ly         already too      late

            'We are probably too late already.'

This combination is found mainly in the eastern part of Oost-Vlaanderen: the densest concentration is in the southern half of the province, along the border with Vlaams-Brabant, to which a few places in the northeastern corner of Oost-Vlaanderen should be added. The construction furthermore also occurs in two far-apart places in Vlaams-Brabant and in two places in Limburg on the border with the province of Antwerpen. That this type of combination is absent from the province of Antwerpen comes as a surprise, since Nuyts (1995) did find it in the urban dialect of Antwerpen city.

3.1.3.2.3  Subject doubling 1 plural; correlation tripling, affix –me and incorporation (–men)  (map 53c)

Map 53c presents the geographic correlation between subject tripling and two phenomena discussed in chapter 2, viz. (a) the use of –me(n) as an affix of pronominal inflection (see 3.1.1.2) and (b) the use of clitic men, previously analysed as resulting from incorporation of an original inflectional affix –n and the weak pronoun me (see also chapter 2, par. 2.3.5.2).

(15)     a.         We       ga–me(n)         daar     ook      naartoe.

                        we       go–we              there    too       to

                        'We too are going there.'

            b.         Ga–men           daar     ook      naartoe?

                        go–we              there    too       to

                        'Are we too going  there?'

Par. 3.1.1.2 argues that the clitic element –me(n), in sentences such as 15(a), has two characteristics which make it behave as an inflectional affix rather than as a pronoun. On the one hand, it easily combines with sentence-initial pronouns, on the other it occurs in a number of (East-Flemish) dialects with virtually all verbs in regular order, so that it appears to be an obligatory element. The fact that this –me(n) is found almost exclusively in dialects which also have tripling, suggests a causal relationship between the two phenomena. Examples of tripling are found slightly further south than those of –me(n), which may indicate that the former phenomenon is older than the latter. Looked at from that perspective, it is fair to assume that the historical origin of the pronominal affix lies in tripling constructions and that its use was in due course extended to sentences without subject tripling: we ga–me(n) wij > we ga–me(n). In this evolution a diminished productivity of subject doubling and tripling may have played a role. If the affix –me(n) finds its origin in the decline of tripling, this also explains the use of two strong markers of person in sentences such as (14), displaying a combination not covered by the generalisation hypothesis formulated above, according to which the first pronoun is always stronger than the second one.

            The –men form, as either a conjugational affix or a clitic pronoun, is densely concentrated in Oost-Vlaanderen, but it is sporadically found in the province of Zeeland and it even occurs in three places in the west of Vlaams-Brabant. The nucleus area of – men is Oost-Vlaanderen, where both inflectional –me and tripling are highly productive, suggesting a causal link between the use of –men on the one hand and both inflectional –me and tripling on the other. This analysis implies that the instances of –men in Zeeland and Brabant are to be considered as relicts of earlier tripling constructions. For Zeeuws-Vlaanderen anyway, Will’s study (based on dialogue fragments from the sixties of the 20th century) affirms that tripling and especially the use of

–me(n) occur more frequently there than the SAND data suggest. Will (2004:259) notes the following example of tripling in Eede, in the west of Zeeuws-Vlaanderen:

(16)     Want   me       zij(n)–me         wèdder           ... geëvacueerd,           natuurlijk.

            for       we_weak             were–we_weak    we_strong                         … evacuated               of course

            'For we were evacuated, of course.'

For –me(n) as an affix, Will (2004:258–59) quotes an example from, among others, Sas van Gent.

(17)     We       zij(n)–me         in Barcelona    geweest.

            we_weak             are–we_weak       in Barcelona    been

            'We have been to Barcelona.'

In addition to Sas van Gent, Will finds examples of the affix –me(n) in Sint Jansteen, Heikant and Stoppeldijk, which suggests that the construction is still fairly common in the east of Zeeuws-Vlaanderen.

3.1.3.3  Subject doubling second person singular

3.1.3.3.1  Subject doubling 2 singular  (map 54a) (map in dynaSAND)

Map 54a shows the distribution of the doubling constructions with the 2nd person singular in the four syntactic constellations, as exemplified in (18):

(18)     a.         Je/ge/gij                                   moogt gij                    dat       niet weten.

                        [regular order]

                        you_weak /you_strong                     may     you_strong           that      not know

                        'You mustn’t know that.'

            b.         Moog–de / moog–je / moog–e             gij                    dat       wel      weten? [in­ver­sion]

                        may–you_weak                                       you_strong           that      affirm know

                        'Is it alright for you to know that?'

            c.         ... da–de / da–je / da–e gij        dat       moogt weten

                        [sub-clause after complementiser]

                        … that–you_weak          you_strong that    may know

                        '… that it is alright for you to know that'

• Hij is   groter   a(l)s–ge            / a(l)s–e         gij.

he is     taller    as–you_weak                              you_strong

'He is taller than you.'

Unlike for the first person (singular and plural), no examples of tripling have been found for the second person. The map is based on the five questions of the SAND questionnaire, among which three translation questions (19a-c) and two ‘this occurs’ questions (19d/e).

(19)     a.         Als      je         gezond             leeft,    leef      je         langer.

                        if          you      healthily          live      live      you      longer

                        'If you live healthily, you live longer.'

            b.         Ze        gelooft             dat       jij         eerder thuis    bent     dan      ik.

                        she       believes that    you      sooner  home   are       than     I

                        'She thinks you are home sooner than me.'

            c.         Je         gelooft             zeker   niet      dat       hij        sterker             is         dan      jij?

                        you      believe surely  not       that      he        stronger is       than     you

                        'You don’t believe he is stronger than you, do you?' 

            d.         Weet    je         gij        al                     dat  je              gij        ook      naar     het feest          mag      komen?

                        know   you_weak you_strong already         that you_weak    you_strong too     to         the party         may     come

                        'Have you already been told that you too are invited to the party?'

            e.         Ge                   weet    gij        d'r        niks     van.

                        you_weak            know   you_strong there  nothing            about

                        'You know nothing about it.'

Doubling in the second person singular is distributed much more widely than in all other combinations of person and number, whose distribution is, roughly speaking, confined to the Flemish dialects and to the dialects spoken in Belgian Brabant. However, as map 54a shows, the phenomenon has also been traced in the large eastern half of the Dutch province of Noord-Brabant, in two places on the southwestern border of the province, in one test place in the Dutch province of Limburg (Meterik) and once in the province of Gelderland (Geldermalsen). The SAND data put Zeeuws-Vlaanderen outside of the area where the construction occurs, with the exception of Hulst. Judging by Will’s data, the construction used at one time to be stronger in Zeeuws-Vlaanderen, since Will (2004:254–257) finds examples of doubling in 16 of the 39 places in the area he examined.

            The distribution of each of the constructions is quite different. The widest distribution is observed for doubling in inverted sentences, which occurs everywhere throughout the area where doubling is found. A striking finding is that the geographic parallelism between the appearance of doubling in inversion and in sub-clauses after complementisers, which exists for the other personal pronouns, does not hold good for the second person singular. Only in Vlaanderen do we see a massive block of subject doubling after complementisers, viz. in an area which is today defined by the SAND data as comprising no more than the two Flemish provinces but which until some time ago used to include part of Zeeuws-Vlaanderen, at least according to Will’s data (2004:256). Outside of Vlaanderen, the presence of the phenomenon is rather diffuse and sporadic: there are a few isolated examples along the northern edge of the Brabant dialect area, near the big rivers, two in the southwest of Noord-Brabant province and three in the transition area between the (south) Brabant dialect and the Belgian Limburg dialect. All these occurrences may well be relicts of an earlier general spread of the construction over an area comprising Brabant and the west of Limburg.

            In regular order, the examples of subject doubling with the second person singular cover roughly the same area as the cases with the other personal pronouns, notably Vlaanderen, to which Will’s data (2004:255) add Zeeuws-Vlaanderen, and the Brabant provinces of Belgium. Outside of this area, there is only one single place in the Noord-Brabant province (Ossendrecht).

            Finally, doubling after a comparative complementiser proves to be virtually confined to East-Flemish and to a few neighbouring dialects, as is the case with the other grammatical persons. In addition, one isolated example occurs in Limburg (Lummen), viz. (groter) azzegij ((taller) than you).

            The Limburg form –egij is a particularly instructive one: Limburg, unlike some parts of Oost-Vlaanderen is, shows no examples of a weak clitic pronoun –e, so that it is easy to no longer perceive the form –egij as a combination of two different pronouns. Consequently, the form –egij from Lummen should probably be analysed as a unit in its own right, just like the cluster –ekik of the first person singular in a number of dialects.

3.1.3.3.2  Subject doubling 2 singular, variation in regular order  (map 54b)

Although second person singular doubling constructions in regular order as a rule have a strong initial pronoun and a weak postverbal one, there are occurrences of two strong pronouns, just like with the first person plural. This shown in (20).

(20)     Gij       moogt gij        dat       niet weten.

            you_strong may   you_strong that    not know

            'You mustn’t know that.'

Occurrences of this construction can be found in approximately the same areas in which the combination wij … wij is heard, viz. Oost-Vlaanderen, the south of Brabant and the west of Limburg. Still, the frequency of these occurrences is not as high as in the case of the first person plural: though again most examples are situated in Oost-Vlaanderen, this can hardly be considered to be a geographic concentration in the east and the south of the province.

3.1.3.4  Subject doubling second person plural

3.1.3.4.1  Subject doubling 2 plural  (map 55a) (map in dynaSAND)

Map 55a traces the distribution of doubling constructions with the second person plural in the four syntactic constellations exemplified in (21).

(21)     a.         Ge / je / gijlie   zult      gullie / gijder   dat       wel      krijgen. [regular order]

                        you_weak / you_strong        will      you_strong that    affirm get

                        'You will no doubt get it.'

            b.         Moog–de / –je / –e      gullie / gijder   dat       wel      weten? [in­ver­sion]

                        may–you_weak               you_strong           that      affirm know

                        'Is it alright for you to know that?'

            c.         ... da–de / da–je           gullie / gijder   dat       moogt weten                                                  [sub-clause after complementiser]

                        … that–you_weak          you_strong           that      may     know

                        '… that it is alright for you to know that'

            d.         Hij       is         groter   a(l)s–ge           gijlie. [after a comparative complementiser]

                        he        is         taller    as–you_weak      you_strong

                        'He is taller than you.'

The map is based on SAND data produced by four translation questions (22 a–d) and two “this occurs” questions (22e/f).

(22)     a.         Als      jullie    zo        gevaarlijk         leven,   dan      leven    jullie    nooit    zo                                lang      als        ik.

                        if          you      so        dangerously     live      then     live      you      never   so                                long     as         I

                        'If you live so dangerously, you will never live as long as me.'         

            b.         We       geloven            dat       jullie    niet      zo slim            zijn      als wij.

                        we       believe             that      you      not       so clever          are       as we

                        'We think you are not as clever as we.'

            c.         Jullie    geloven            zeker   niet dat            zij        armer   zijn dan jullie.                                      you      believe             certainly not that            they     poorer are than you

                        'You don’t believe they are poorer than you, do you.'

            d.         Als      jullie    gaan     dan      gaan     jullie.

                        if          you      go        then     go        you

                        'If you go, then you go.'

            e.         Weet    je         gullie    al                     dat       je         gullie    ook      naar     het feest          mag      komen?

                        know   you_weak you_strong already         that      you_weak you_strong          too       to                                 the party may come

                        'Do you already know that you too can come to the party?'

            f.          Ge       weet    gullie    d'r        niks     van.

                        you_weak know  you_strong there  nothing            about

                        'You know nothing about it.'

The geographic distribution of subject doubling in the second person plural is nearly identical to that of the second person singular. The geographic range of the construction is also almost the same, and that also goes for distributional differences between the syntactic constellations. Once more, the distribution is largest for inverted sentences, a constellation which covers the whole of Vlaanderen, Brabant and the neighbouring western part of Limburg. In the other constellations, it is only in Vlaanderen that we find a dense concentration (Herne in the south of Brabant joining up with Vlaanderen). Furthermore, doubling in regular order frequently occurs in the province of Antwerpen, more rarely in Vlaams-Brabant and not at all to the north of the Belgian-Dutch frontier. With regard to doubling after a complementiser in sub-clauses, the concentration remains heavy in Vlaanderen, but outside that area the frequency has shrunk even further. Examples are rare and isolated: one in the northwest of the province of Antwerpen (Kapellen), two in Vlaams-Brabant (Bou­tersem and Zoutleeuw), one in Belgian Limburg (Houthalen), and one very isolated occurrence in the Dutch province of Noord-Brabant (Geldermalsen), which is actually the only one north of the frontier.

3.1.3.4.2  Subject doubling 2 plural, variation in regular order  (map 55b)

With the second person plural, subject doubling in regular order is less prone to variation than with most other grammatical persons. The SAND data contain no examples of tripling, not even with plural second person pronouns. However, we do find variation in the various types of pronouns which can be combined. A number of dialects can only combine a weak and a strong pronoun, whereas others can have two strong pronouns in a subject doubling construction, as for example in (23).

(23)     Gijlie   kunt     gijlie    daar     niets    aan       doen.

            you_strong can     you_strong there  nothing about do

            'You can do nothing about it.'

The examples of subject doubling with two strong pronouns are widely scattered over the provinces of Antwerpen, Vlaams-Brabant and Oost-Vlaanderen; they tend to be concentrated in the east of Oost-Vlaanderen.

3.1.3.5  Subject doubling third person singular masculine

3.1.3.5.1  Subject doubling 3 singular masculine  (map 56a) (map in dynaSAND)

Map 56a presents the distribution of doubling constructions with masculine pronouns of the third person singular in the four syntactic constellations examined. Examples are given in (24).

(24)     a.         Je / ie / hij          zal     hij / hem          dat       wel      krijgen. [regular order]

                        he_weak / he_strong will      he/him_strong       that      affirm get

                        'He will get it, I’m sure.'

            b.         Mag–ie            hij/hem            dat       wel      weten?                         [inversion]

                        may–he_weak     he/him_strong       that      affirm know

                        'Is it alright for him to know that?'

            c.         ... dat–ie          hij        dat       mag      weten   [sub-clause after complementiser]

                        …that–he_weak he_strong that      may     know

                        '…that it is alright for him to know that'

            d.         Zij       is         groter   dan–ie              hij.     [after comparative complementiser]

                        she       is         taller    than–he_weak     he_strong

                        'She is taller than him.'

It is worth noting that more variants of this construction have been observed than the examples illustrate. Sections 3.1.3.5.2 and 3.1.3.5.3 deal with this variation in greater detail.

            The data in the map are based on six questions from the SAND questionnaire, viz. three translation questions (25a–c) and three ‘this occurs’ questions (25d–f).

(25)     a.         Als      hij        nog      drie jaar           leeft,    leeft     hij        langer   dan      zijn vader.

                        if          he        still      three years      lives     lives     he        longer  than

                        his father

                        'If he lives another three years, he will have lived longer than his father.'

                        b.         'k         Geloof             dat       ik         groter   ben      dan      hij.

                                    I           believe that      I           taller    am       than     he

                                    'I believe that I am taller than him.'

                        c.         Je         gelooft             zeker   niet      dat       hij        sterker             is         dan jij?

                                    you      believe             not       that      he        stronger is       than you

                                    'You don’t believe he is stronger than you, do you?'

                        d.         Ik         denk    dat–ie              hij        morgen            ook      komt.

                                    I           think    that–he_weak      he_strong tomorrow        also      comes

                                    'I think he is coming too tomorrow.'

                        e.         Hij       kan      hij        d'r        ook      niks     aan       doen.

                                    he_strong can       he_strong there    also      nothing            about   do

                                    'He can do nothing about it either.'

                        f.          Hij       weet    hij        d'r        niks     van.

                                    he_strong knows  he_weak there    nothing            about

                                    'He knows nothing about it.'

It is in regular order that subject doubling has the widest distribution, a situation which is not different from the other pronouns of the third person and from the first person plural. Apart from Vlaanderen, the construction is also concentrated in the western half of  the province of Antwerpen and in Vlaams-Brabant, and it extends to the east with some ocasional occurrences in a few places in the west of Limburg. Map 56a has no separate symbol for occurrences of tripling in regular order, but map 56b does.

            The examples of doubling in inverted sentences and in sub-clauses after a complementiser, constructions which overlap geographically speaking, cluster together in a compact Flemish area, with two extensions into the south west of Vlaams-Brabant. Yet, strikingly, quite a number of East-Flemish dialects have no doubling after verbs and after complementisers, a feature with which they deviate from the above maps. This is illustrated in the dialects of the Land of Waas in the northeast of Oost-Vlaanderen, in a few places in the northwest of Meetjesland and for one place in the south of the province, between the rivers Schelde and Dender. It is probably not a coincidence that in each case the dialect involved has no separate weak pronoun for the masculine third person singular (see chapter 2.3.3). Still, doubling in these two constellations is also absent from the dialect of Brugge in West-Vlaanderen, even though that dialect does have such a weak form.

            It is doubtful if certain doublets, such as East-Flemish (t)jij and eastern West-Flemish (t)jem, are still generally felt to be combinations of two separate pronouns, in spite of the etymologising spelling of these forms used in the examples. Just like the form (e)kik for the first person singular, (t)jij (< (t)ie–ij) and (t)jem (<(t)ie–em) tend to become generalised and to occupy other positions, witness the example in (26).

(26)     Zelfs    (t)jij / (t)jem     kan      dat       niet      oplossen.

            even     he / him_strong     can       that      not       solve

            'Even he cannot solve that.'

The distribution of such fossilised clusters after the particle zelfs is mapped in 56b and 56c.

            There is a dense concentration of doubling constructions after a comparative complementiser in the south of Oost-Vlaanderen, around Gent and further eastward along the river Schelde. In the southwest of Brabant two places join up with the East-Flemish concentration. Finally, the phenomenon also makes a few rare appearances in West-Vlaanderen.

3.1.3.5.2  Subject doubling 3 singular masculine, variation in regular order  (map 56b)

The map shows the distribution of the various combinations of sentence-initial and post-verbal pronouns in regular order sentences. It gives a picture, more in particular, of the distribution of weak and strong pronouns in these combinations, making no distinction between forms which are originally subjects and originally objects. That distinction is made for doubling with a post-verbal cluster consisting of two pronouns (i.e., tripling).

            With regard to the relation weak-strong forms, four types of combinations occur:

(27)     a.         sentence-initial weak pronoun + post-verbal strong pronoun, e.g., e(n) kan hem / je kan hij;

            b.         sentence-initial ie + post-verbal strong pronoun taking a different form than ie, e.g. ie kan hem. Unlike the other weak pronouns, the initial pronoun does not have a reduced vowel, though it is phonetically weaker than the doubling pronoun;

            c.         sentence-initial strong pronoun + identical postverbal strong pronoun, e.g., hij kan hij;

            d.         sentence-initial strong pronoun + a different postverbal strong pronoun, e.g., hem kan hij, as heard in Tienen, in the southeast of Vlaams-Brabant.

A comparison with doubling constructions of other personal pronouns reveals a striking difference, viz. the widespread occurrence of combinations with two strong pronouns. The strong pronouns involved are always identical except in one place (Tienen). The combination is clearly concentrated in an area consisting of the east of Oost-Vlaanderen, the provinces of Antwerpen and Vlaams-Brabant and a number of places along the western border with Limburg. In most of the dialects concerned (in Oost-Vlaanderen, Brabant and the west of Limburg) weak masculine personal pronouns are anyway most unusual in sentence-initial position and, indeed, are sometimes completely absent. Outside of the area outlined, the construction is sporadically found in the west of Oost-Vlaanderen, in the neighbouring northern area of West-Vlaanderen and even once in Frans-Vlaanderen.

            The 'usual' combination in subject-doubling constructions in regular order (viz. a weak pronoun in sentence-initial position and a strong post-verbal doubling pronoun) is heavily concentrated in the western half of West-Vlaanderen, with two places in Frans-Vlaanderen joining up with the Belgian province. The east of West-Vlaanderen also has quite a number of occurrences but there the construction often alternates with the combination of preverbal ie and strong postverbal hij or hem. The initial ie is actually a perceptibly reduced one: it is phonetically situated somewhere in between the clear (closed) [i] and the schwa. This type of combination puts the east of West-Vlaanderen in an intermediate position in between the western construction with a weak initial pronoun and a strong postverbal one and the East-Flemish variant with two strong forms.

            East of the region with preverbal ie, in the western half of Oost-Vlaanderen, we find an area where subject tripling occurs frequently. The bulk of the examples combine a preverbal hij with a postverbal cluster –jij, which consists of a weak form ie and a strong pronoun hij. It is only in the West-Flemish Oostveld that we find a weak form before the verb, viz. je. The postverbal cluster –jij is probably felt to be one single indivisible pronoun (see 3.1.3.5.3), comparable with (e)kik in the first person singular.

            A final remark concerns one location, the East-Flemish Moerzeke, where we meet with constructions such as hij kan–tje, combining a strong initial pronoun with a postverbal clitic. This clitic may be an affix, comparable with the second element of the doubling construction wij ga–me(n) in the first person plural (see 3.1.3.2.3).

3.1.3.5.3  Subject doubling 3 singular masculine, variation in inversion  (map 56c)

Doubling clusters of subject pronouns of the masculine third person singular display a large measure of morphological variation. This is partly explained by the fact that in quite a number of dialects the pronominal subject for the masculine third person singular takes the form of original object pronouns. Looking solely at the original morphosyntactic status of the clustered pronouns (in other words, not taking the weak/strong difference into account), we can distinguish four different combinations:

(28)     a.         subject form + subject form: (t)–ie hij, e.g., Zal–(t)–ie hij ?

            b.         subject form + object form: (t)–ie hem, e.g., Zal–(t)–ie hem ?

            c.         object form + subject form: t–(e)n hij, e.g., Zal–t–(e)n hij ?

            d.         o­b­j­e­c­t­ f­orm + o­b­j­e­c­t ­f­o­rm: t–(e)n hem, e.g., Zal–t(e)n– hem ?

In labelling the first pronoun in (c) en (d) as an object form, some caution must be exercised. After all, it is not certain that the –en form originated from an object form in all the dialects where it occurs. This doubt arises first of all with regard to the west of West-Vlaanderen, where the preverbal en in particular may also have grown from a reduction of subject hij (pronounced as ie), to which an –n was added as an intermediate sound if the word following the pronoun began with a vowel. The possibility cannot be ruled out that such an initial pronoun, the result of reanalysis, has extended its range of action to the postverbal position (see 2.3.3.2.1).

            The combination of postverbal object –en – assuming that it is in fact an original object form – with subject –hij as a doubling pronoun is widespread in the large western half of West-Vlaanderen and along the coast, and emerges again along the eastern edge of the south of Oost-Vlaanderen and in two places in southwestern Brabant. Doubling clusters found in between these two areas of concentration always have a reduction of the weak ie as the first pronoun. Virtually without any exception, it is doubled with the strong subject form hij in Oost-Vlaanderen; in West-Vlaanderen predominantly with the object form hem. These clusters are presumably no longer perceived as combinations of two pronouns, but have rather been reanalyzed, whether or not along with the inflectional affix –t, as one single pronominal form (t)jij and (t)jem, respectively. A separate indication that reanalysis may have taken place is that the forms (t)jij and (t)jem also prove to occur in cases where there is no doubling, as for example in (26) in 3.1.3.5.1.

            A final observation concerns the transitional north-south oriented area between the t–(e)n–hij type and the (t–)ie–hem type, where the western t–(e)n is doubled with the eastern West-Flemish strong form hem.

3.1.3.6  Subject doubling third person singular feminine

3.1.3.6.1  Subject doubling 3 singular feminine  (map 57a) (map in dynaSAND)

Doubling constructions with feminine pronouns for the third person singular are exemplified in three syntactic constellations, as illustrated in (30). The SAND questionnaire does not include a question eliciting doubling after a comparative complementiser.

(29)     a.         Ze / zij                                     zal                   zij        dat       wel      krijgen. [regular order]

                        she_weak  / she_strong         will                  she_strong that     affirm get

                        'She will get it, no doubt.'

            b.         Mag–ze                       zij       dat       wel      weten?                         [inversion]

                        may–she_weak                she_strong that     affirm know

                        'Is it alright for her to know that?'

            c.         ... da–ze           zij        dat       mag      weten       [sub-clause after complementiser]

                        … that–she_weak           she_strong that     may     know

                        '… that it is alright for her to know that'

The map is based on SAND data obtained by means of one translation question (30a) and three "this occurs" questions (30b–d).

(30)     a.         Als      zij        zo        gevaarlijk         leeft,    leeft     ze        niet      lang      meer.

                        if          she       so        dangerously     lives,    lives     she       not       long

                        anymore

                        'If she lives so dangerously, she won’t live long anymore.'

                        b.         Met     hij        te         gaan werken    moest (ze)                  zij        heel de dag       thuis blij­ven.

                                    with     he        to         go working      had to  (she_weak)          she_strong all day home stay

                                    'As he took up a job, she had to stay at home all day.'

                        c.         Ze        leeft (ze)          zij        op water en brood       deze week.

                                    she_weak lives (she_weak )            she_strong on water and bread     this week

                                    'She is living on bread and water rations this week.'

                        d.         Ze        weet    zij        d'r        niks     van.

                                    she_weak knows she_strong there   nothing            about

                                    'She knows nothing about it.'

The distribution patterns in map 57a are the same as those in a number of maps above. Once again, doubling in regular order sentences covers the largest area: it consists of a continuous concentration in Vlaanderen, a somewhat less compact distribution in the province of Antwerpen and Belgian Brabant, and finally two occurrences on the western border of Belgian Limburg. There are no differences in distribution for doubling in inverted sentences and after a complementiser; in these two constellations doubling occurs only in Vlaanderen, in Hulst in Zeeuws-Vlaanderen and in Herne in Vlaams-Brabant.

3.1.3.6.2  Subject doubling 3 singular feminine, variation in regular order  (map 57b)

As far as the use of weak and strong forms is concerned, the map shows a nearly general spread of combinations with a weak initial pronoun and a strong one as doubling pronoun. Only in one place does the combination consist of two strong pronouns, viz. Boutersem in Vlaams-Brabant.

            Examples of tripling too have been found, again in Oost-Vlaanderen only and at that few and far between in the northern half of the province. It remains unclear if the post-verbal cluster ze–zij of these examples is is to be analysed as a combination of two pronouns instead of as one single (clitic) pronoun or as a combination of a strong pronoun with an affix –ze. The possibility for clitic pronouns to develop into affixes has already been brought up in the discussion of plural first person pronouns, more particularly with regard to the analysis of the element –me in combina­tions such as we ga–me wij. Whatever the underlying reasons for this evolution may be, the scarcity of tripling occurrences suggests that the formation of a double pronoun ze–zij, and/or the development of an af­fix –ze (if development there is) has progressed much less far than with a number of other combinations of person and number.

3.1.3.7  Subject doubling third person singular neuter  (map 58a) (map in dynaSAND)

This section deals with two different doubling constructions in which pronouns of the third person play a role, notably (i) combinations in which both the 'regular' subject and the doubling pronoun(s) are neuter third person pronouns, and (ii) constructions in sentences in which the place of the regular subject is filled by an expletive element (an equivalent of the expletive er in Standard Dutch), which is then doubled by one or more neuter pronouns of the third person singular.

            The two constructions are discussed together because they are structurally very similar and because they have largely overlapping geographic distributions. Both are considered by most respondents to be archaic. The following are examples of the two types in three of the four syntactic constellations. The fourth syntactic constellation, after a comparative complementiser, does not produce any examples of doubling with neuter third person pronouns.

(i) Pronoun subject + doubling pronoun(s)

The pronoun subject can be empty as in (31a) and (31b), or referential as in (31c).

(31)     a.         't          Is hij                nooit    anders              geweest.                      [regular order]

                        it_weak    is he_strong          never   different          been

                        ’t          is–(t)–ie                       hij        nooit    anders   geweest.

                         it_weak   is–(it_weak)–it_weak          he_strong never   different been

                        ’t          Is–t      het        nooit   anders geweest.

                        it_weak    is–it_weak it_strong              never different been             

                        'It has never been different'

            b.         Is–t      hij        alweer aan het regenen ?                                              [inversion]

                        is–it_weak  he_strong again raining

                        Is–t–ie             hij        alweer aan het regenen?          

                        is–it_weak –it_weak  he_strong again raining

                        Is–t      het       alweer aan het regenen ?

                        is–it_weakit_strong again    raining

            c.         ... da–t             hij        nooit zal lukken                                              [subordinated clause]

                        ... that–it_weak     he_strong never will succeed

                        ... da-t              (-ie)     hij        nooit    zal       lukken

                        ... that–it_weak   (he_weak )he_strong            never   will      succeed

                        ... da–t het       nooit zal lukken

                        ... that–it_weak   it_strong   never will succeed

                        ‘... that it will never succeed.’

The SAND data include examples of doublings such as those above in the three syntactic constellations, but truly systematic eliciting of the construction was done only in main clauses with regular order and with inversion. To this end two "this occurs” questions were used in the inquiry by telephone.

(32)     a.         't          Is         hij        / (–t) ie           hij        al         lang geleden.

                        it_weak    is         he_strong / (it_weak)  he_weak he_strong            already            long ago

                        'It’s already long ago.'

            b.         Is–t                  hij        /–t        het       al         lang geleden?

                        is–it_weak                 he_strong /it_weak  it_strong   already            long ago

                        'Is it long ago already?'

(ii) Expletive element + doubling pronoun(s)

Here a preliminary remark should be made. The expletive er of Standard Dutch, when used in the dialects which double this element (notably the Flemish dialects), can be represented by the adverbs er and d'r as well as by the neuter third person pronoun 't. The pronoun 't when used as an expletive cannot be considered the subject of the sentence because there is no agreement with the conjugated verb. See sentence (33b).

(33)     a.         D'r / t              is         hij /      is–(t)–ie hij /                            is –(t) het         alweer een boom omgewaaid.                                      [regular order]

                        there_weak /it_weak is        he_strong /is (it_weak ) he_weak  he_strong /      is it_weak  it_strong             again

                        a tree   blown down

                        'Again a tree has been blown down.'

            b.         Zijn     d'r hij              /zijn d'r–(t)–ie hij                               /zijn d'r–(t) het

                                                hier      bomen omgewaaid?                 [inversion]

                        Are      there_weak  he_strong/are there_weak  (it_weak ) he_weak  he_strong            /are there_weak  (it_weak)  it_strong               here     trees blown down

                        'Have any trees been blown down here?'

            c.         ... dat–d'r hij /            dat–d'r–(t)–ie hij /                              dat–d'r–(t) het             hier   

                                                dikwijls ingebroken wordt                  [subordinated clause]

                        …that–there_weak  he_strong /that there_weak  (it_weak ) he_weak  he_strong          that there_weak  (it_weak ) it_strong here     often    broken into      is

                        '… that houses often are broken into here'

Like the expletive er in Standard Dutch, its dialectal equivalents emerge in sentences with an indefinite subject NP (examples a and b) as well as in subjectless passive sentences (example c). In the case of the former type, the question can be raised if it is the expletive element that is doubled or rather the lexical subject. After all, lexical subjects can also be doubled in sentences without an expletive element, as will be shown in the following section. However, subject-doubling sentences of that type are clearly more restricted geographically speaking (see map 58b): if the expletive element is omitted, doubling becomes ungrammatical in quite a number of dialects. The geographic discrepancy between sentences with and without expletive element suggests that it is in fact the expletive element which is doubled in the former type.

            All the doubling constructions illustrated in (33) can be found in the SAND data, but very few of them were elicited systematically. The various questionnaires were not concerned with subjectless passive sentences and sub-clauses, but they did systematically examine doubling constructions in regular order sentences and inverted sentences with an expletive element and an indefinite subject. The telephone inquiry included two ‘this occurs’ questions.

(34)     a.         Heeft   d'r hij /                                     d'r t het                                    hier      een man           gewoond          of         een vrouw ?

                        has       there_weak  he_strong  /       there_weak it_weak it_strong                      here     a man               lived                or         a woman

                        'Did a man live here or a woman?'

            b.         Heeft   d'r  hij /                        d'er ie hij                      niemand           mijn paraplu    gezien ?

                        has       there_weak  he_strong /        there_weak  he_weak  he_strong nobody         my umbrella seen

                        'Has nobody seen my umbrella?'

The distribution of the doubling constructions discussed in (i) and (ii) above, in regular order and in inverted sentences, is shown on map 58a. As it proved to be hard to elicit the two constructions from respondents, the map also takes account of occurrences in spontaneous conversations.

            The geographic distribution of doubling constructions with/of neuter third person pronouns is more restricted than in all cases discussed so far. The nucleus of the area in which they occur lies in the west of Oost-Vlaanderen and in the east of West-Vlaanderen. In Oost-Vlaanderen, the concentration bulges out eastward along the river Schelde from Gent as far as Dendermonde. There are a few sporadic occurrences in the southeast of Oost-Vlaanderen, but in the western half of West-Vlaanderen and in Frans-Vlaanderen the construction is unknown.

            The distribution shows some striking differences depending on the sentence type. In regular order doubling of a neuter third person pronoun is found in all the test places of the area where the construction occurs, except two places, both in the south of West-Vlaanderen. The other combinations are absent in a string of places along the edge of the area: the three locations between the rivers Schelde and Dender in the south of Oost-Vlaanderen, in Lokeren in the northeast of the province and in Torhout in West-Vlaanderen. Doubling of expletive elements is frequent in the northwest of Oost-Vlaanderen and in the western half of West-Vlaanderen; it is also found in two places along the Schelde east of Gent. The distribution in regular order is very similar to that in inversion; it is only in two places in the southwest of Oost-Vlaanderen that the combination occurs in inverted sentences only. Finally, doubling of the neuter third person in inverted sentences is the least usual construction, being found only in the northwest of Oost-Vlaanderen and around Gent, with one isolated occurrence in West-Vlaanderen (Kortrijk).

            The fact that every type of doubling of or with neuter third person pronouns is absent in the peripheral west, where the oldest forms that are still alive, suggests that doubling of and with third person pronouns is younger than subject doubling of the other personal pronouns.

3.1.3.8  Subject doubling third person singular, doubling of lexical noun phrases  (map 58b) (map in dynaSAND)

Not only pronominal subjects but also non-pronominal subjects can be doubled with a pronoun. These constructions are found in regular order sentences as well as in inverted sentences and after a complementiser, as in (35).

(35)     a.         Die man          heeft    hij        daar     niets    mee te maken. [regular order]

                        that man          has       he_strong there    nothing            with to do

                        'That man has nothing to do with it.'

            b.         Heeft   hij        die man           daar     iets      mee te maken? [inversion]

                        has       he_strong that man          there    something with to do

                        'Has that man something to do with it?'

            c.         ... dat   hij        die man           daar     niets    mee te maken heeft [after complementiser]  

                        …that  he_strong that man          there    nothing with to do      has

                        '… that that man has nothing to do with it'

Doubling of non-pronominal subjects proves to be rather unstable in most test places: it is hard to elicit and some respondents’ judgments about the (un)grammaticality of the instances they are offered sometimes contradict their own spontaneous linguistic usage. Respondents sometimes reject doubling though they do use it in spontaneous conversation, so much so that we have nothing but spontaneous occurrences for a number of places. We are forced to exclude doubling in sub-clauses from the discussion; the data presented in SAND are too scarce and too fragmented. The map is based on answers to a large number of ‘this occurs’ questions: (36a–e) are regular order sentences, (36f–l) are inverted sentences.

(36)     a.         Marie heeft    zij        daar     niets    mee te maken.

                        Marie  has       she_strong there   nothing            with to do

                        'Marie has nothing to do with it.'

            b.         Pol       heeft    hij        daar     niets    mee te maken.

                        Pol       has       he_strong there    nothing            with to do

                        'Pol has nothing to do with it.'

            c.         Marie en Pol   hebben zij        daar     niets    mee te maken

                        Marie and Pol have     they_strong there nothing            with to do

                        'Marie and Pol have nothing to do with it.'

            d.         Dat kind          heeft    hij        daar     niets    mee te maken.

                        that child         has       he_strong there    nothing with to do

                        'The child has nothing to do with it.'

            e.         Niemand          heeft    hij        daar     iets                  mee te maken.

                        nobody                        has       he_strong there    something        with to do

                        'Nobody has anything to do with it.'

            f.          Is         hij        die man            nu        al         getrouwd?

                        is         he_strong that man          now     already            married

                        'Has that man married already?'

            g.         Woont             hij        die man            nog      altijd    in jouw straat?

                        lives     he_strong that man          still      always in your street

                        'Does that man still live on your street?'

            h.         Is         hij        Pol       hier      geweest?

                        is         he_strong Pol       here     been

                        'Has Pol been here?'

            i.          Hoe     heeft    hij        Pol       dat       opgelost?

                        how     has       he_strong Pol       that      solved

                        'How did Pol solve that?'

            j.          Is         zij        Marie hier      al         geweest?

                        is         she_strong Marie here     already            been

                        'Has Marie already been here?'

            k.         Is         hij        dat onweer      hier      al         gepasseerd?

                        is         he_strong that storm       here     already            passed

                        'Has that storm already passed here?'

            l.          Zijn     zij        Pol en Marie   hier      al         geweest?

                        are       they_strong Pol and Marie  here  already            been

                        'Have Pol and Marie already been here?'

As can be seen on map 58b, this type of subject doubling is confined to a very small area; it is typical of the northwestern quarter of Oost-Vlaanderen, including the region between the rivers Leie and Schelde. In regular order sentences some examples of the construction are found reaching out to the rest of Oost-Vlaanderen and the east of West-Vlaanderen. There are no differences in the distribution of doubling after a verb and after a complementiser, but slight differences do occur depending on gender and number of the subject and on whether the subject is a common noun or a proper name. Masculine and feminine nouns are doubled more often than neuter nouns, singular nouns more often than plural nouns, definite NPs more often than indefinite NPs and, finally, proper names more often than common nouns. Differences of this kind, however, can hardly ever be traced on dialect maps; it is rather a question of the frequency with which a given phenomenon occurs. Also, the SAND network is somewhat too wide-meshed to be able to detect the many differences in the geographic distribution of doubling instances of the various types of lexical NPs.

            A special feature of doubling of lexical NPs in regular order is the following: the weightiest subject marker, in other words the non-pronominal subject, comes in preverbal position. This is a departure from the usual pattern of pronominal doubling, in which the sentence-initial pronoun assumes a weak form or at least a form with less weight than the doubling pronoun. The fact that lexical NPs do not observe the rule may indicate that doubling in sentences with this type of subject has not evolved organically but was introduced through analogy with pronominal doubling constructions. This is actually borne out by the distribution shown in map 58b: the area of concentration lies entirely inside the region where pronominal subject doubling occurs most frequently in all grammatical persons and in all syntactic constellations, viz. the (north)west of Oost-Vlaanderen and the area around Gent. The frequent and diversified use of the construction with pronominal subjects may have triggered an extension to constellations with non-pronominal subjects.

            Similar to what we find for pronominal subjects, the doubling of lexical NPs involves strong pronoun forms or clusters with a strong element. The morphological shape of the doubling pronoun is subject to a great deal of variation, as is shown in (37).

(37)     a.         Pol       kan      hij        /           kan (–t)–ie hij                          /kan hem /        kan (–t) het                              dat niet. [masculine singular]

                        Pol       can–     he_strong /            can (–it_weak ) he_weak  he_strong     /can him_strong /  can (–it_weak ) it_strong                  that      not

                        'Pol cannot do that.'

            b.         Marie kan zij /            kan–ze zij                   /kan (–t)–ie hij                         /kan (–t) het                 dat       niet. [feminine singular]

                        Marie  can she_strong /    can-she_weak  she_strong    /can (-it_weak ) he_weak  he_weak     /can (–it_weak ) it_strong     that      not

                        'Marie cannot do that.'

            c.         Dat kindje       kan hij             /kan (–t)–ie hij                         /kan (–t) het     dat niet. [neuter singular]

                        that child         can–he_strong      /can (–it_weak ) he_weak  he_strong    /can (–it_weak) it_strong that not

                        'That child cannot do that.'

            d.         Marie en Pol   kennen zij        /kennen (–t)–ie hij                   /kennen (–t) het

                        daar     niets van. [plural]

                        Marie and Pol know they_strong  /know (–it_weak  he_weak  he_strong  /        know (–it_weak ) it_strong   there   nothing about

                        'Marie and Pol know nothing about it.'

The examples clearly show that pronouns which were originally masculine and neuter forms can be used to double lexical subject nouns of the three genders as well as plurals. The meaning of the pronouns in question appears to have faded somewhat, resulting in the loss of specific gender and number features, so much so that these pronouns are (or can be) coreferential with all third person subjects.

            The discrepancies of person and number which can be observed between the subject and its doubling pronoun are an additional confirmation of the earlier hypothesis according to which the doubling of lexical subjects finds its origin in the extension (for pragmatic reasons) of the phenomenon outside its prototypical pronominal constellation.

3.1.3.9  Subject doubling third person plural

3.1.3.9.1  Subject doubling 3 plural  (map 59a) (map in dynaSAND)

The following examples (38) illustrate subject doubling of plural pronouns of the third person in the four syntactic constellations.

(38)     a.         Ze        zullen zij(lie)   dat       wel      krijgen. [regular order]

                        they_weak will they_strong             that      affirm get

                        'They will no doubt get it.'

            b.         Mogen–ze       zij(lie) dat       wel      weten?             [inversion]

                        may–they_weak they_strong that   affirm know

                        'Is it alright for them to know that?'

            c.         ... da–ze           zij(lie) dat       mogen weten. [subordinated clause]

                        …that–they_weak they_strong that            may     know

                        '… that it is alright for them to know that'

            d.         Hij       is         groter   dan–ze                         zij(lie).             [after a comparative complementiser]

                        he        is         taller    than–they_weak they_strong

                        'He is taller than them.'

The map reflects data from four SAND questions:

(39)     a.         Als      ze        voor hun werk             leven,   leven    ze        niet      voor hun kinderen.

                        if          they     for their job     live      live      they     not       for their children

                        'If they live for their work, they don’t live for their children.'

            b.         Ze        geloven            dat       wij       rijker    zijn      dan      zij.

                        they     believe             that      we       richer   are       than     they

                        'They believe that we are richer than them.'

            c.         Jullie    geloven            niet      dat       zij        armer   zijn      dan      jullie.

                        you      believe             not       that      they     poorer are       than     you

                        'You don’t believe that they are poorer than you.'

            d.         Ze        weten zullie   d'r        niks     van.

                        they     know   they     there    nothing            about

                        'They know nothing about it.'

The distribution patterns of the various doubling constructions mapped in 59a are largely analogous to those of the other third person pronouns. Once again, Vlaanderen is the nucleus, and doubling in the constellations (a) to (c) occurs without any exception. And again, we find that the only doubling construction that occurs outside of Vlaanderen is that in regular order, once more with the exception of Herne in the southwest of Brabant, a place which again proves to echo the situation of Vlaanderen. The distribution of doubling in regular order sentences eastward of the Flemish dialect area again extends to the west of Limburg, but this time the number of occurrences is much smaller than in the maps discussed earlier: in the provinces of Vlaams-Brabant and Antwerpen, there are four examples each, and there are two on the western edge of Belgian Limburg. In the constellation after a comparative complementiser, doubling is again found exclusively in Oost-Vlaanderen.

3.1.3.9.2  Subject doubling 3 plural, variation in regular order  (map 59b)

In the same way that we find morphological variation in doubling constructions with other persons, we find it for the third person plural in regular order sentences, in that a number of dialects can use two strong pronouns whereas other dialects only combine a weak and a strong pronoun. The SAND data do not show tripling for the third person plural.

            Similar to the situation of the other personal pronouns, the examples of subject doubling with two strong pronouns do not constitute a compact area: we find them scattered over the province of Antwerpen, Vlaams-Brabant and Oost-Vlaanderen, with a conspicuous concentration in the east of Oost-Vlaanderen.

3.1.3.10  Subject doubling: synthesis  (map 60a)

Map 60a presents a general overview of the differences in the distribution of subject doubling according to sentence type (regular order, inversion, sub-clause after complementiser), person and number of the subject and, specifically for the third person, the kind of subject NP (pronoun or lexical NP).

3.1.3.10.1  Regular order against inversion and order of the sub-clause

As far as sentence type is concerned, a clear-cut difference emerges between doubling in regular order sentences, on the one hand, and in inverted sentences and sub-clauses after a complementiser, on the other. The latter two constellations differ between themselves on one point only, viz. in the second person, and together they cover a vaster area than doubling in regular order. In the former construction, i.e., in regular order sentences, subject doubling occurs almost exclusively to the south of the Belgian-Dutch frontier, in a compact area comprising the two Flemish provinces and Frans-Vlaanderen as well as the two Belgium Brabant provinces (Vlaams-Brabant and the province of Antwerpen), and finally the (north)western periphery of Belgian Limburg. This area shows hardly any differences in constructions between the various grammatical persons. The only two places in the Netherlands where doubling in regular order occurs – Hulst and Ossendrecht – are on or near the frontier.

            In inverted sentences and in sub-clauses, by contrast, the bulk of the Dutch province of Noord-Brabant (more precisely some two thirds of the province, to the east) also falls within the area where the construction is found, and there are even some examples much further north, in the region of the big rivers. One occurrence is found in Dutch Limburg.

            However, the expansion of doubling in inverted sentences north of the Belgian-Dutch frontier exclusively involves second person pronouns. Contrary to subject doubling in regular order, subject doubling in inverted sentences shows marked differences in geographic distribution depending on person and number of the subject which is doubled.

3.1.10.2  Distribution by person and number of the subject

Map (60a) reveals major geographic differences with regard to the paradigmatic density of subject doubling in inversion, i.e., the sum (calculated for each location) of the subject types – the various personal pronouns and the lexical subjects of the third person – which are (or can be) doubled. The spread over the paradigm is largest in the Flemish provinces, where as a rule (nearly) complete doubling paradigms are listed – 'complete' meaning: all pronouns except the third person neuter. In order to keep the map from becoming too complex, the third person masculine singular has not been included. The geographic distribution of subject doubling for the first person plural, the third person singular and the third person plural is identical, and only slightly different from that for the third person masculine singular. The reason for this small difference is probably that a number of dialects do not have a weak pronoun for the third person masculine singular.

            In the dialects of Belgian Brabant, subject doubling looms less large, though its occurrence is more than sporadic, doubling showing up in the first person singular and the second person singular and plural. In the test places of Noord-Brabant and Limburg, it is only second person pronouns that can be doubled.

            Even in Vlaanderen, with the greatest concentration of occurrences, the picture is not homogeneous: we find maximum density in a funnel-shaped area in Oost-Vlaanderen, with its narrow base in the dialects around Gent and widening westward as far as the provincial border of West-Vlaanderen, with one occurrence beyond that border (Oostkerke). In this nuclear area, it is not only pronouns that can be doubled but lexical NPs as well.

            Around this East-Flemish funnel, we find a large number of test places in which doubling of all personal pronouns occurs but not of lexical NPs. This picture is fairly general in the eastern half of West-Vlaanderen and in Oost-Vlaanderen along the Schelde south and east of Gent. These secondary nuclei therefore comprise the west of West-Vlaanderen, the Land of Waas in the northeast of Oost-Vlaanderen, the Dender region in the southeast of Oost-Vlaanderen and the southern border of the province; in these secondary nuclei the density is lower. To the west, the decrease in density is most pronounced in the periphery, viz. in Frans-Vlaanderen. To the east, the Land of Waas and to a lesser extent the Dender region represent a transition to Brabant, where even fewer doubling constructions are found.

            It should be mentioned, finally, that subject doubling in inverted sentences and after complementisers usually shows the same distribution. But there is one exception to this generalisation: the second person. Both in the second person singular and the second person plural, subject doubling in inversion is general and widespread in the dialects of Brabant. However, after complementisers it is only in Vlaanderen that the phenomenon is general, since the examples in Brabant are few, being found along the northern border and in the southeast. This remarkable geographic distribution suggests that doubling after complementisers used at one time to be general in the dialects of Brabant as well, but has lost ground in more recent times.

            As a matter of fact, it is not only the sporadic Brabant distribution of doubling in the second person after complementisers which makes it likely that subject doubling is on the wane. The fact that occurrences of doubling in regular order often do not constitute a compact area, again mainly in Brabant, is also an indicator of decline. Moreover, in the literature we also find data which make the decline of subject doubling in Brabant a plausible hypothesis. De Vriendt (2003:75), for example, lists a number of doubling constructions in the dialect of Brussels (Vlaams-Brabant) which cannot be found back in the SAND data. Actually, not only in Brabant does it look as if doubling is giving ground. A comparison of the SAND data with the data for Zeeuws-Vlaanderen (Will 2004) reveals that subject doubling in Zeeuws-Vlaanderen has today become considerably less productive than fifty years ago. Much of the present-day geographic distribution of subject doubling can therefore be viewed in the context of the decline of the phenomenon in a number of regions.

3.2  Subject clitics following ja (‘yes’) and nee (‘no’)

3.2.1  Introduction

In some dialects, the interjections ja (‘yes’) and nee (‘no’) can be followed by a subject clitic referring to the subject of the previous sentence, in most cases a statement or a question to which an answer is provided. (40) is an example of a conversation in which a clitic is attached to ja (‘yes’). Examples such as (40) do not occur in Standard Dutch.

(40)     QUESTION:   Heb     je         al         gegeten?

                                    have     you      already            eaten

                                    ‘Have you had dinner already?’

            ANSWER:      Ja–k     (of: neen–k)

                                    yes–I

                                    ‘Yes’

In (40), the clitic ’k (‘I’) is coreferential to the subject je (‘you’) in the question. In the dialects in which the phenomenon is attested, there is variation as to the particular pronouns that can be attached to ja (‘yes’) and nee (‘no’). In some dialects, the phenomenon occurs for all grammatical persons. For instance, the dialect of Poelkapelle in West-Vlaanderen provides a complete paradigm (41).

(41)     sg.        1          jaa–k    (‘yes–I’)

                        2          jaa–je   (‘yes–you’)

                        3m.      jaa–ie   (‘yes–he’)

                          f.        jaa–s    (‘yes–she’)

                          n.       jaa–t     (‘yes–it’)

            pl.        1          jaa–w   (‘yes–we’)

                        2          jaa–je   (‘yes–you’)

                        3          jaa–s    (‘yes–they’)

There are also dialects in which only a restricted number of clitics can be attached to ja (‘yes’) and nee (‘no’). Therefore, an accurate geographic description of the phenomenon must include maps for all the grammatical persons.

            Dialects not only differ as to the completeness of the paradigm of ja (‘yes’) and nee (‘no’). There is also morphological variation within the pronouns that are used in the construction. The clitics in (41) look like obvious reduced forms of the weak pronouns ’k, je, ie, ze, ’t, we, je and ze (‘I, you, he, she, it, we, you and they’ respectively), which are attested in Standard Dutch as well. But in many cases, it is not immediately clear what might be the nature of a certain pronoun following ja (‘yes’) or nee (‘no’). Consider for instance (42), in which a number of possible first person plural pronouns are given.

(42)     QUESTION:   Waren jullie                op tijd thuis?

                                    were    you (pl.)          in time home

                                    ‘Did you get home in time?’

            ANSWER 1:    Jaa–w

            ANSWER 2:    Jaa–m

            ANSWER 3:    Jaa–m’n

                                    yes–we

                                    ‘Yes’

In (42), several morphological variants of the weak, first person plural pronoun are used, as described in 2.3.5. In chapter 2, it was mentioned that in some dialects, the choice of whether to use a particular pronoun depends on its syntactic position: in proclitic position (preceding the finite verb), different pronouns are used than in an enclitic position. In addition, in an enclitic position to the complementiser, other pronouns may be used than in an enclitic position to finite verbs. The data in (42), however, show that the form of the pronoun following ja (‘yes’) or nee (‘no’) cannot be predicted straightforwardly on the basis of the data that were described in chapter 2. All forms in (42) come from dialects in Vlaanderen. In the dialects in Vlaanderen, 1pl. we is found almost exclusively in sentence-initial position, whereas 1pl. m’n is attested only in an enclitic position. Apparently, some dialects use a sentence-initial pronoun following ja (‘yes’) and nee (‘no’) (cf. jaa–w), and others an enclitic pronoun (cf. jaa–m’n).

Furthermore, there are dialects in which an agreement marker is found between ja (‘yes’) and nee (‘no’) on the one hand, and the clitic on the other hand, as in (43).

(43)     QUESTION:   Hebben            ze        al         gegeten?

                                    have                 they     already            eaten

                                    ‘Have they had dinner already?’

            ANSWER:      Jaa–n–s

                                    yes–3pl–they

                                    ‘Yes’

In the linguistic literature, (43) is described as an instance of ‘agreement of ja (‘yes’) and nee (‘no’)’. The agreement marker following ja / nee is not used in all dialects that show clitics following ja / nee. Furthermore, the geographic distribution of agreement of ja (‘yes’) and nee (‘no’) cannot be derived from the distribution of any particular agreement marker. Hence this phenomenon too must be mapped in its own right.

3.2.2  Discussion of the literature

There is not much linguistic literature dealing with clitics following ja (‘yes’) and nee (‘no’). Some dialect grammars or dialect dictionaries from the areas in which the phenomenon is found, provide a paradigm of ja (‘yes’) or nee (‘no’). These paradigms are often incomplete. The form that is attested most often, is 3sg.neuter jaa–t. Paardekooper (1993) and Smessaert (1995) discuss the phenomenon extensively. Furthermore, clitics following ja (‘yes’) and nee (‘no’) are mentioned by Devos (1986), De Vogelaer (2003) and Van Craenenbroeck (2004).

Paardekooper (1993) discusses the morphological variation within pronouns following ja (‘yes’) and nee (‘no’), providing three maps that also allow a reconstruction of the recent diachronic developments. One map is based on the so-called ‘Willems-corpus’, which dates from the 19th century (see Goossens, Taeldeman & Weijnen 1989); a second one is based on the RND dialect atlases (Blancquaert & Pée 1925–1982); and a third one shows Paardekooper’s own data and data from several dialect monographies. Paardekooper assumes that the phenomenon is disappearing gradually, and that 1sg. jaa–k and 3sg.neuter jaa–t are able to ‘survive’ longer than the other forms of the paradigm of ja (‘yes’). Smessaert (1995) discusses a paradigm from West-Vlaanderen.

Devos (1986:176–177) only deals with second person pronouns. The geographic distribution of jaaj and jaag in West-Vlaanderen leads her to conclude that the construction has an elliptic origin: the geographic distribution of the pronouns je and ge (‘you’) following ja (‘yes’) correlates with the distribution of sentence-initial pronouns rather than with the distribution of enclitic pronouns, indicating that the clitics following ja (‘yes’) are originally sentence-initial pronouns. Van Craenenbroeck (2004:223–255) assumes a correlation between clitics following ja (‘yes’) and nee (‘no’) and one specific elliptic construction: short do-replies, a dialectal construction resembling the English use of to do in short answers (e.g., Hij kan niet komen. Ja hij doet. ‘He can’t come. Yes he does.’). The correlation is illustrated with a map, which indeed shows a comparable geographic distribution for the two phenomena. Finally, De Vogelaer (2003:196–199) considers the combination of n and s in the third person plural form jaans not to be a combination of an agreement marker n and a clitic s, but analyses forms such as jaans as combinations of ja and one single pronominal element ns that has been formed in an enclitic position.

Both the possibility to attach subject clitics to response particles such as ja (‘yes’) and nee (‘no’), and agreement of response particles, seem to be unattested in the languages of the world: neither of both phenomena are described in typological overviews such as Corbett (1991, 2000) and Siewierska (2004).

3.2.3  Discussion of the maps

3.2.3.1  Clitics and agreement following ja ‘yes’/ nee ‘no’, first person singular

3.2.3.1.1  Clitics and agreement following ja ‘yes’/ nee ‘no’, 1 singular  (map 61a) (map in dynaSAND)

The possibility of clitics following ja (‘yes’) or nee (‘no’) was explicitly explored during the fieldwork interviews. Informants were asked whether they could react to the utterance Wil je nog koffie? (‘Do you want some more coffee?’) using a form like jaa–k (‘Yes–I’). In addition, a translation was asked.

The map shows that the phenomenon occurs in Frans-Vlaanderen, West-Vlaanderen and Oost-Vlaanderen, at one sampling point in Zeeland Vlaanderen and in the southern periphery of Vlaams-Brabant and, in addition, at a number of very isolated sampling points. In some locations, an /n/ is found between ja (‘yes’) and the clitic: in the town of Veurne in West-Vlaanderen (attested form: jaank), in Arendonk in the province of Antwerpen, and in the town of Huizen in Noord-Holland (form: janik). Especially in Veurne, this /n/ is likely to be an agreement marker. In Arendonk and Huizen, the /n/ may very well be an intervocalic linking–n (see 3.2.3.1.2).

At some sampling points outside the jaak-area, the form jaat is found as a possible answer to questions such as Wil je nog koffie? (‘Do you want some more coffee?’). In these dialects, first person singular jaak is no longer used, and only relics of the phenomenon occur, in the form of the third person singular neuter form jaat, which can be used as an emphatic form of ja (‘yes’) in all answers, whatever the grammatical person of the subject of the previous utterance.

When compared to the map in Paardekooper (1993:144), the striking decline of the construction is visible: the geographic distribution on map 61a is much more limited than on the map that Paardekooper draws on the basis of the Willems-corpus, a collection of data gathered around 1885. At this time, forms like jaak were attested throughout the southern half of the language area, although even then, there were several dialects in Brabant and Limburg in which they were not used. For the northern half of the Dutch language area, no older data are available.

3.2.3.1.2  Correlation agreement affixes, 1 singular  (map 61b)

Map 61a shows some possible instances of a first person singular agreement marker following ja (‘yes’). It is instructive to compare these possible instances of an agreement–n following ja (‘yes’) to the geographic distribution of –n following other inflected constituents, such as verbs and, in some dialects, complementisers. Map 61b shows all attestations of an inflectional –n following ja (‘yes’), complementisers such as a(ls) or da(t) (labelled ‘C-nasaal’ on the map), and the verb ga (labelled ‘V-nasaal’ on the map;          [note, however, that the distribution of –n following verbs with a stem on a vowel, such as ga (infinitive gaan) is not representative for all verbs]).

An inflectional –n is found following verbs with a stem ending on a vowel (like gaan ‘to go’) at most sampling points in Friesland, Noord-Holland, Antwerpen, Vlaams-Brabant and Belgian and Dutch Limburg. In addition, in Frans-Vlaanderen, West-Vlaanderen and Oost-Vlaanderen, Zeeland Vlaanderen, Zuid-Holland, Noord-Brabant and Groningen, examples of an inflectional –n occur. Following complementisers, the use of an inflectional –n seems to be found exclusively in Vlaanderen, with instances in Frans-Vlaanderen, West-Vlaanderen and Oost-Vlaanderen, and Zeeland Vlaanderen, all in places that also have an –n following the verb gaan (‘to go’). Thus, an inflectional –n following complementisers seems to occur (almost) exclusively in dialects that also have an –n following verbs. The reverse, however, does not seem to be the case: there are dialects with an inflectional –n on 3pl. ga, which do not have an agreement–n following complementisers.

Two of the three sampling points with –n following ja also have –n following verbs: Veurne (West-Vlaanderen) and Arendonk (Antwerpen). In addition, Veurne has complementiser agreement using the agreement marker –n. In the town of Huizen (Noord-Holland), an –n is only found following ja, but Huizen is situated close to an area that has gaan, with –n, as the first person singular form of gaan (‘to go’). Hence it is possible that the form janik is a relic of a formerly present inflectional–n. But it is also possible that the /n/ in janik must be explained as an intervocalic linking–n.

3.2.3.2 Clitics and agreement following ja ‘yes’/ nee ‘no’, second person singular and plural

3.2.3.2.1  Clitics following ja ‘yes’/ nee ‘no’, 2 singular and plural  (map 62a)

Most dialects do not have a specific weak pronoun or clitic at their disposal for the second person plural, but use second person singular pronouns instead. This holds for all dialects in which clitics can be attached to ja (‘yes’) and nee (‘no’), so both second person singular and plural pronouns can be dealt with in one single paragraph (see 2.3.2.2). Map 62a and 62b show all instances of second person clitics following ja. In all dialects with clitics following ja (‘yes’) in the first person singular, the third person singular feminine or neuter, or the third person plural, the corresponding form for the second person singular and plural was asked. The informants had to provide an answer to the following questions: Mag ik een koekje? Ja... (‘Can I have a cookie? Yes-...’) and Mogen wij een koekje? Ja... (‘Can we have a cookie? Yes-...’).

The geographic distribution of second person clitics following ja (‘yes’) differs slightly from the distribution for the first person singular, but is roughly the same as the distribution for most other grammatical persons: the phenomenon is found in Frans-Vlaanderen, in West-Vlaanderen and in Oost-Vlaanderen. As to the morphological variation, the area seems to be divided into a western and an eastern part: jaaje is the western form found in Frans Vlaanderen and in the west and the north of West-Vlaanderen. Jaag is the eastern form, which is used in Oost-Vlaanderen and in the east and the southeast of West-Vlaanderen, but which is also found further to the west, in the jaaje-area. Three southern sampling points in the border zone between Oost-Vlaanderen and West-Vlaanderen have jaagt.

Jaaje, showing a final schwa, occurs more frequently than jaaj (without final schwa). Only in Brugge, Kooigem (West-Vlaanderen) and Waals-Kappel (Frans-Vlaanderen), jaaj is the only possibility; in Poelkapelle, Wulvergem (West-Vlaanderen) and Merckeghem (Frans Vlaanderen), both jaaj and jaaje are found.

Like the other maps, map 62a shows a smaller construction area than the map by Paardekooper (1993:144), which describes the situation around 1885. Devos (1986:177) provides a map for West-Vlaanderen, based on data gathered in the 1980s, showing an almost identical situation as that shown on map 62a.

3.2.3.2.2  Correlation pronouns with initial /j/, 2 singular and plural  (map 62b)

The morphological variation in the second person pronouns that can be attached to ja (‘yes’) correlates with the considerable morphological variation in second person pronouns in general, as discussed in chapter 2. A lot of variation is found especially in the southwest of the Dutch language area: in most southwestern dialects, weak and strong pronouns cannot be considered phonological variants, and must be kept apart for discussion. In addition to this basic distinction between weak and strong pronouns, different weak pronouns are used in different syntactic environments. In chapter 2, four syntactic environments were distinguished: (i) the sentence-initial (or proclitic) position, (ii) the enclitic position to verbs with a stem ending with a vowel (e.g., gaan ‘to go’), (iii) encliticised to verbs with a stem ending with a consonant (e.g., leven ‘to live’), and (iv) encliticised to complementisers. This makes five different pronouns (four weak ones and a strong one), to which a sixth one can be added, namely the pronoun following ja ‘yes’. The pronouns with an initial /j/, originally enclitic forms, are particularly interesting (see also chapter 2). In the southwest of the language area, it is common for dialects to have both pronouns with an initial /j/ and other forms. Map 62b shows all pronouns with an initial /j/ in the relevant area.

Pronouns with an initial /j/ have their widest geographic distribution in an enclitic position following gaan (‘to go’): in this position, they are the dominant form in Frans-Vlaanderen, West-Vlaanderen and Zeeland. In addition, they occur in the southeast of Oost-Vlaanderen and in the town of Ronse (Oost-Vlaanderen). Following verbs with a stem ending with a consonant, like leven (‘to live’), and following complementisers (labelled ‘C+-je’ on the map), the distribution of the pronouns with initial /j/ is less wide. For both syntactic environments, an almost identical geographic distribution is found. Some sampling points only have a j-pronoun (je) in one of these two environments, but there are not enough sampling points in the relevant area to decide whether the j-pronouns indeed have a different geographic distribution in both syntactic positions. It is, however, perfectly clear on map 62b that je is significantly more widespread in an enclitic position than in a proclitic position: proclitic je is not attested in the eastern periphery of West-Vlaanderen, where enclitic je is rather common. Following ja (‘yes’), the geographic distribution of the initial /j/ is even more limited: jaaj and jaaje are only found in the north and the west of West-Vlaanderen, and in Frans-Vlaanderen. Finally, in strong pronouns, an initial /j/ is rare in the southwest: strong pronouns with an initial /j/ are only found in the Netherlands.

Map 62b shows a (small) discrepancy in the distribution of the initial /j/ in proclitic position and following ja (‘yes’). This is unexpected, since pronouns following ja (‘yes’) are assumed to be originally proclitic pronouns in an elliptic construction in which the verb is dropped.

3.2.3.3  Clitics and agreement following ja ‘yes’/ nee ‘no’, third person singular masculine

3.2.3.3.1  Clitics following ja ‘yes’/ nee ‘no’, 3 singular masculine  (map 63a)

In the dialects with a weak pronoun following ja (‘yes’) in the first person singular, the third person singular feminine or neuter, or the third person plural, the other forms of the paradigm of ja (‘yes’) were asked during the fieldwork interviews. The relevant test sentence was Komt hij morgen ook op je verjaardag? Ja... (‘Is he coming to your birthday party tomorrow? Yes-...’). Weak, third person singular masculine pronouns following ja (‘yes’) are, as for most grammatical persons, attested in Frans-Vlaanderen, West-Vlaanderen and Oost-Vlaanderen. In the west of the construction area (Frans-Vlaanderen, the west of West-Vlaanderen) and in the east (the east of Oost-Vlaanderen), the form jaan (also: jane) is used. In between these two jaan-areas, in the east of West-Vlaanderen and the west of Oost-Vlaanderen, the dominant form is jaaj (also: jaje), an obvious phonetic reduction of jaie. A small number of sampling points in the surroundings of Gent have jaij. Gent itself, has both jaaj and jajij, a form unique to Gent.

The geographic distribution of some of these forms is striking. For instance, in a remarkably large number of sampling points the clitics ‘n (e.g., in jaan) and ie (e.g., in jaaj) are attached to ja (‘yes’). This is unexpected, since the clitics that follow ja are assumed to originally be sentence-initial forms, and sentence-initial clitics for the third person singular masculine are rare outside West-Vlaanderen (see 2.3.3.2). The use of ij and jij following ja (‘yes’) is remarkable for another reason: unlike all other pronouns that can be attached to ja (‘yes’), ij and jij are not weak pronouns (see 2.3.3). There is a very clear geographic correlation between the use of ij and jij following ja (‘yes’) and the use of the same pronouns in another construction, namely subject doubling in sentences with a nominal subject (see 3.1.3.8).

Finally, it needs to be remarked that the use of clitics following ja (‘yes’) used to be more widespread in the past for the third person singular masculine as well (cf. Paardekooper 1993:144).

3.2.3.3.2  Correlation pronouns ‘n, ‘m, ‘r and ie, 3 singular masculine  (map 63b)

The morphological variation in the pronouns following ja (‘yes’) is similar to the variation in other syntactic environments, like the sentence-initial position (labelled ‘‘r/‘n/‘m V_fin’ on the map), and the enclitic position following verbs and complementisers (labelled ‘C/V_fin ‘r/‘n/‘m/ie’ on the map). Map 63b shows the distribution of the weak pronouns ‘n, ‘m, ‘r and ie in the southern half of the language area. In addition, the attestations of strong ie in sentence-initial position are shown as well, since the distribution of sentence-initial ie is likely to correlate with the distribution of ie following ja (‘yes’). For an overview of the dialects with strong and weak ie, see 2.3.3.1 and 2.3.3.2.

Discrepancies between the enclitic position and other syntactic environments are the rule in the southern part of the Dutch language area: the four pronouns, ‘n, ‘m, ‘r and ie are all attested more widely in an enclitic position than in other positions. Sentence-initial attestations of ‘m and ‘r are notoriously rare. Sentence-initial ‘n and ie are not rare, and are found especially in West-Vlaanderen and the west of Oost-Vlaanderen. However, one has to take into account that most instances of ie come from the area in which ie is the strong pronoun. Both for ie and for ‘n, the distribution following ja (‘yes’) does not correspond completely to the distribution in sentence-initial position. For instance, an area near Gent (Oost-Vlaanderen) has ie following ja (‘yes’), but neither sentence-initial ie, nor enclitic ie. A number of sampling points in western West-Vlaanderen display the opposite behaviour (ie is found only in sentence-initial position). For ‘n, the most striking discrepancy is found in the east of Oost-Vlaanderen, where some sampling points have ‘n following ja (‘yes’), but no sentence-initial ‘n, nor, in some cases, enclitic ‘n, causing ‘n to be a pronoun which is exclusively found following ja (‘yes’).

Concerning the use of clitics following ja (‘yes’), some instances of ie and ‘n following ja (‘yes’) in Oost-Vlaanderen are situated outside the area in which the same pronouns are found in proclitic or even enclitic position. These are undoubtedly the most remarkable discrepancies. Some of the other discrepancies that are visible on map 63b, are discussed in chapter 2 (2.3.3).

3.2.3.4  Clitics following ja ‘yes’/ nee ‘no’, 3 singular feminine  (map 64a) (map in dynaSAND)

During the fieldwork interviews, all informants were asked whether they could answer to the question Gaat ze dansen? (‘Does she go to dance?’) using the expression jaas (‘yes-she’). In addition, a translation of jaas was asked. Map 64a shows the results.

Once more, the area in which the construction is attested is the same as for most other grammatical persons, and includes Frans-Vlaanderen, West-Vlaanderen and Oost-Vlaanderen. The third person singular feminine shows less morphological variation than the other grammatical persons: most sampling points have jaas. The town of Lovendegem (Oost-Vlaanderen) has jaats. In five places, Oostkerke, Kooigem and Kortrijk in West-Vlaanderen and Sint-Laureins and Lovendegem in Oost-Vlaanderen, jase is found, a form that may be known to other dialects as well. However, the possibility of using jase was not an issue during the fieldwork interviews, so it is not possible to provide an accurate picture of its geographic distribution. In some places, the form jaat occurs. Jaat was originally a neuter form that has been generalised to other grammatical persons (see 3.2.3.8).

Here too, it is clear that third person singular, feminine clitics following ja (‘yes’) used to be far more widespread in the past (cf. the map in Paardekooper 1993:144).

3.2.3.5  Clitics following ja ‘yes’/ nee ‘no’, 3 singular neuter  (map 64b) (map in dynaSAND)

The possibility of using third person singular neuter clitics following ja (‘yes’) was explored at all sampling points, during the fieldwork interviews. The informants were asked whether they could answer to the question Staat het huis te koop? (‘Is the house for sale?’) using a form like jaat (‘yes-it’). In the dialects in which clitics occured following ja (‘yes’), a translation of the answer was asked.

The phenomenon is found in the usual construction area, comprising Frans-Vlaanderen, West-Vlaanderen and Oost-Vlaanderen, but also far outside this area, with instances in the province of Antwerpen and in the south of both Belgian Limburg and Dutch Limburg. There is hardly any morphological variation: all dialects have jaat. In four places (Oostkerke and Moorsele in West-Vlaanderen, and Deinze and Eeklo in Oost-Vlaanderen), the informants spontaneously used the emphatic form jate as well, which is probably used elsewhere too.

The geographic distribution of clitics following ja (‘yes’) is much wider in the third person singular neuter than in the other grammatical persons, but does not even come close to the 19th century distribution as described by Paardekooper (1993:144), who observes the phenomenon in the entire area for which data are available, namely the southern half of the Dutch language area (the area south of the Rhine) and a few places to the north of the Rhine. However, Paardekooper’s map too, shows a very fragmented distribution of the phenomenon in all provinces but Frans-Vlaanderen, West-Vlaanderen and Oost-Vlaanderen. Also, Paardekooper does not distinguish the ‘normal’ instances of jaat, as shown on map 64b, from the so-called ‘generalised’ ones, in which the ’t in jaat does not refer to a neuter, third person subject (see 3.2.3.8). Since both types of jaat are distributed differently, Paardekooper’s map should also be compared to a map that shows both types of jaat, like map 66c.

3.2.3.6 Clitics and agreement following ja ‘yes’/ nee ‘no’, first person plural

3.2.3.6.1  Clitics following ja ‘yes’/ nee ‘no’, 1 plural  (map 65a)

In the first person plural too, clitics are found following ja (‘yes’) in some dialects. Map 65a shows data that are gathered in the dialects in which clitics were found following ja (‘yes’) in the first person singular, the third person singular feminine or neuter, or the third person plural: in these dialects, the question was asked whether first person clitics could be attached to ja (‘yes’) as well and, if yes, what would the form of the clitic be. The test sentence was Zijn jullie tevreden met het cadeau? Ja... (‘Are you pleased with the gift? Yes-...’).

This map shows a picture comparable to the other maps: clitics following ja (‘yes’) are found in the southwest, more precisely in Frans-Vlaanderen and the Belgian provinces of Oost-Vlaanderen and West-Vlaanderen. Three different morphological variants are attested (see 2.3.5 for discussion of each of these forms): jaam is the dominant form both in Oost-Vlaanderen and in West-Vlaanderen. Two locations in the southeast of Oost-Vlaanderen (Aalst and Ninove) and one in Vlaams-Brabant (Herne) have jam’n. Finally, the form jaaw is found, in West-Vlaanderen. Some dialects have both jaam and jaaw. Other dialects, including the ones in Frans-Vlaanderen, only have jaaw at their disposal.

At four sampling points, three in West-Vlaanderen (Oostende, Nieuwpoort and Menen) and one in Oost-Vlaanderen (Bevere), the form jame is found, next to jaam. Jame is an emphatic variant of jaam, obtained by adding a schwa to jaam. Presumably, jame has a wider distribution than these four places in which a spontaneous attestation is found.

This map can be compared to two older maps, both by Paardekooper. When compared to Paardekooper’s (1993:144) map depicting the situation around 1885, the construction area has shrunk considerably: in the 19th century, complete paradigms of ja (‘yes’) were found far outside the current area showing the phenomenon (which is Vlaanderen), although Vlaanderen was the only area in which almost all dialects exhibited the phenomenon. When compared to the map based on the more recent RND-material (Paardekooper 1993:145), the differences are not that large: Paardekooper finds first person plural clitics following ja (‘yes’) in Frans-Vlaanderen and West-Vlaanderen, in the southern half of Oost-Vlaanderen and in the west of Vlaams-Brabant. This is roughly the same area that has the construction on map 65a. In addition, Paardekooper also finds instances of the phenomenon in three isolated locations in the Netherlands (two in Zuid-Holland and one in Utrecht).

3.2.3.6.2  Correlation pronouns with initial /m/, 1 plural  (map 65b)

The morphological variation of the first person plural pronouns following ja (‘yes’) is comparable to the general morphological variation in first person plural pronouns, as described in 2.3.5. Map 65b shows the distribution of all first person plural pronouns with an initial /m/, such as the weak pronouns me and m’n and the strong pronoun miender.

The initial /m/ has its widest distribution in weak, enclitic pronouns (labelled ‘C/V_fin me’ on the map): enclitic me occurs frequently in Frans-Vlaanderen, West-Vlaanderen and Oost-Vlaanderen, Zeeland and in the larger parts of Antwerpen and Vlaams-Brabant; enclitic m’n (C/V_finm’n) occurs in the east of Oost-Vlaanderen and Zeeland. In sentence-initial position, me (me V_fin) is frequently found in Frans-Vlaanderen, West-Vlaanderen and the west of Oost-Vlaanderen, but only sporadically in more eastern and northern areas. Sentence-initial m’n is not attested. Following ja (‘yes’), the initial /m/ is only found in West-Vlaanderen and in the south of Oost-Vlaanderen, with the form jaam (or jame), and in the towns of Aalst, Ninove (Oost-Vlaanderen) and Herne (Vlaams-Brabant) (jam’n). There are only two places with a strong pronoun with an initial /m/, namely Brugge and Roeselare, both in West-Vlaanderen.

The geographic distribution of all these pronouns leads to the conclusion that all dialects in which an initial /m/ is found in one ore more first person plural pronouns, have enclitic me. This is evidence for the hypothesis that the enclitic forms with an initial /m/ are older than the other first person plural pronouns with a /m/ (see also 2.3.5.2). This also shows that the initial /m/ indeed originates in an enclitic position. Both following ja (‘yes’) and in sentence-initial position, the initial /m/ is less widespread, but the precise geographic distribution differs in both cases. There are both dialects with only m(e) following ja (‘yes’) and with only proclitic me. However, the ones with m(e) following ja (‘yes’) outnumber the other ones, and, in addition, these constitute an area, in the southeast of Oost-Vlaanderen. This difference is unexpected, since the linguistic literature suggests that the pronouns following ja (‘yes’) were originally proclitic pronouns.

3.2.3.7  Clitics and agreement following ja ‘yes’/ nee ‘no’, third person plural

3.2.3.7.1  Clitics and agreement following ja ‘yes’/ nee ‘no’, 3 plural  (map 66a) (map in dynaSAND)

Map 66a shows the different forms of ja (‘yes’) that were given as a reply to the question Hebben ze al gegeten? (‘Have they had dinner already?’). It was not only asked whether a clitic could be attached to ja (‘yes’), but also whether an agreement marker could appear between ja (‘yes’) and the clitic.

The area in which clitics can be attached to ja (‘yes’) is the same for the third person plural as for most other grammatical persons, and includes Frans-Vlaanderen, West-Vlaanderen and Oost-Vlaanderen. The morphological variation is quite limited. In some dialects, especially in the west and the north of West-Vlaanderen, and in and around Gent (in Oost-Vlaanderen), an inflectional –n is observed (in jaans); all other dialects have jaas. Two sampling points in West-Vlaanderen, Moorsele and Kortrijk, not only have jaas, but also the emphatic form jase.

Paardekooper (1993:144) shows that the phenomenon used to be more widespread in the 19th century. He also mentions dialects with agreement markers on ja (‘yes’) in West-Vlaanderen and Oost-Vlaanderen, and, unlike map 66a, in one location in Vlaams-Brabant, two locations in Noord-Holland, and in one location on the Wadden island of Texel (Noord-Holland).

3.2.3.7.2  Correlation agreement affixes, 3 plural  (map 66b)

An agreement–n is not only found following ja (‘yes’), but also following verbs and complementisers. Map 66b provides an overview. Following verbs (gaan ‘to go’ in this particular case; ‘V-nasaal’ on the map), the inflectional –n is found in almost the entire Dutch language area, the exceptions being a small area in Limburg (with –nt), and two bigger ones: one in the east of the Netherlands (with –t), and one in Friesland (with –ne). Following complementisers (‘C-nasaal’ on the map), the agreement –n is found in the southwest, the majority of the attestations coming from Frans-Vlaanderen, West-Vlaanderen, Oost-Vlaanderen or Zeeland. Two more isolated instances are found north of this area, in Lopik (Utrecht) and Strijen (Zuid-Holland). All sampling points with complementiser agreement on –n are situated in areas with 3pl. –n following gaan (‘to go’). This holds for the attestations of jaans as well, which are found exclusively in Frans-Vlaanderen, West-Vlaanderen and Oost-Vlaanderen. In addition, all sampling points that have jaans, also have complementiser agreement on –n in the third person plural. Hence, the presence of an agreement-n following gaan (‘to go’) in a particular dialect seems to be a necessary condition for the presence of an inflectional –n following complementisers and ja (‘yes’).

3.2.3.8  Generalised jaat  (map 66c)

Clitics following ja (‘yes’) are mainly found in the Vlaanderen dialects. Only in Frans-Vlaanderen, West-Vlaanderen and Oost-Vlaanderen, are (more or less) complete paradigms of ja (‘yes’) attested, in which (nearly) all clitics can attach to ja (‘yes’). One form in particular is far more widespread in Dutch dialects though, namely jaat (‘yes–it’). Some dialects, both in Vlaanderen and elsewhere, use the form jaat (‘yes–it’) in sentences in which no reference is made to a third person neuter subject. Hence, jaat can be used in a more general way. In the dialects that know this ‘generalised’ use of jaat, there are no restrictions on the subject of the sentences to which jaat is the response: generalised jaat occurs irrespective of the grammatical person of the subject in the preceding utterance, and must be considered merely as an emphatic form of ja (‘yes’). Map 66c shows all of the instances of jaat, both as a third person singular, neuter form and as a generalised form (see also map 61a, 64a, 66a). Some attestations of generalised jaat were shown on map 61a, 64a and 66a, since they were given in reply to the test sentences that were used as a data source for these maps. Others have been found elsewhere in the SAND corpus, often in the spontaneous conversations between the informants that were recorded immediately before and after the fieldwork interviews.

As has already been said, jaat as a third person singular, neuter form is only found in the southwest of the Dutch language area. The attestations of generalised jaat are found in the periphery, or even far outside the usual construction area, although there are a few exceptional instances of generalised jaat in Vlaanderen dialects as well. Map 66c shows a clear correlation between generalised jaat and the loss of complete paradigms of ja (‘yes’) in certain dialects. There is also probably a correlation between generalised jaat and the frequency of clitics following ja (‘yes’): in the west of the dialect area of Vlaanderen, where complete paradigms of ja (‘yes’) are common, attestations of ja (‘yes’) and nee (‘no’) without a clitic are rather rare (cf. Paardekooper 1993:157, who mentions the existence of an area in West-Vlaanderen in which clitics obligatorily follow ja ‘yes’ and nee ‘no’ in all answers, even in the late 20th century). More to the east, clitics following ja (‘yes’) and nee (‘no’) become more rare, and precisely in these eastern dialects, generalised jaat is found as a stronger, emphatic form of ja.

The (relative) rareness of ja (‘yes’) and nee (‘no’) without an attached clitic is observed in the SAND data for Oost-Vlaanderen and West-Vlaanderen. There are fewer instances of ja, and its phonological variants such as jou, jau, joa and jao, irrespective of the function of the form, in West-Vlaanderen and Oost-Vlaanderen than for the other Belgian provinces, Antwerpen, Vlaams-Brabant, and Limburg. The average number of attestations of ja per sampling point is 19,15 (West-Vlaanderen) and 21,56 (Oost-Vlaanderen) in the west, and 40,78 (Antwerpen), 54,07 (Vlaams-Brabant) and 46,26 (Belgian Limburg) in the east: the simplex form ja is obviously more rare in the west than in the east.

Despite its wide geographic distribution, jaat too seems to be losing ground. Paardekooper’s (1993:144) map shows instances of jaat in a larger area, and with a higher density.

3.3  Literature on subject doubling and subject clitics following ja (‘yes’) and nee (‘no’)

Ariel, M. (2000). ‘The development of person agreement markers: from pronouns to higher

accessibility markers.’ In M.Barlow & S. Kemmer (eds.) Usage-based Models of Language. Stanford: CSLI Publications. 197–260.

Bayer, J. (1984). ‘COMP in Bavarian syntax.’ The Linguistic Review 3. 209-274.

Blancquaert E. & W. Pée (1925–1982) (red.). Reeks Nederlands Dialect-atlassen (RND). Antwerpen: De Sikkel.

Corbett, G. (1991). Gender. Cambridge: Cambridge University Press.

—      (2000). Number. Cambridge: Cambridge University Press.

Croft, W. (1990). Typology and universals. Cambridge: Cambridge University Press.

De Geest, W. (1990). 'Universele grammatica op de Gentse toer: de relevantie van

            clitisering met doubletten voor het Principes-en-Parameters-Model van de Generatieve Grammatica.' Taal en Tongval Themanummer 3: Dialectsyntaxis. 108–124.

De Meersman, A. (1985). 'Een VSO-maneuver in de Zuidnederlandse dialecten.' In H.

            Ryckeboer, J. Taeldeman & V.F. Vanacker (eds.) Hulde-album Prof. dr. Marcel Hoebeke. Gent: RUG Seminarie voor Nederlandse taalkunde. 123–131.

De Schutter, G. (1994). 'Voegwoordflectie en pronominale clitisering waarin Vlaams en

            Brabants bijna elkaars tegengestelden zijn.' Taal en Tongval 46, 108–131.

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

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

De Vogelaer, G. (2003). 'Person marking in Dutch dialects.' In B. Kortmann (ed.) Dialectology meets typology. Berlin: Mouton De Gruyter. 181–210.

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.

De Vriendt, S. (2003). Grammatica van het Brussels. Gent: Koninklijke Academie voor

            Nederlandse Taal- en Letterkunde en Brussel: Academie van het Brussels.

Givón, T. (1976). ‘Topic, pronoun, and grammatical agreement.’ In C.N. Li (ed.) Subject and Topic. New York: Academic Press. 149-188.

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

          jaar enquête Willems.

Haegeman, L. (1990). ‘Subject pronouns and subject clitics in West-Flemish.’ The

            Linguistic Review 7. 333-363.

—        (1992). Theory and description in generative syntax: A case study in West

            Flemish. Cambridge: Cambridge University Press.

Haiman, J. (1991). ‘From V/2 to subject clitics: Evidence from Northern Italian.’ In E. Traugott & B. Heine (eds.) Approaches to grammaticalization, Vol. 2. Amsterdam: John Benjamins. 135-157.

Nadasdi, T. (1995). ‘Subject NP doubling, matching, and minority French.’ Language Variation

            and Change 7. 1-14.

Nuyts, J. (1995). 'Subjectpronomina en dubbele pronominale constructies in het Antwerps.'

            Taal en Tongval 47. 43–58.

Paardekooper, P.C. (1952). 'Syntagma's van het type ge moet gij + het/er. ' In L.G.J.Verberne,

& A. Weijnen (eds.) Land van mijn hart: Brabantse feestbundel voor Mgr. Prof. dr. Th. J. A. J. Goossens. Tilburg: Henri Bergmans. 64–69.

—        (1993). 'Jaak/Neenik enz.' Tabu 23. 143–170.

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

            Interuniversitair Centrum voor Neerlandistiek.

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

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

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

—        (2004). Person. Cambridge: Cambridge University Press.

Smessaert, H. (1995). 'Morfosyntaxis van het Westvlaamse bè–jaa–k–gie.' Tabu 25. 45–60.

—        (1996). 'Pronominal cliticization in West Flemish.' In E. Schiller, E. Steinberg &

B. Need (eds.) Autolexical Theory: Ideas and Methods. Berlin/New York: Mouton de Gruyter. 241–289.

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.

Vallaeys, A. (1997). Spraakkunst van het Poperings. Poperinge: Stad Poperinge.

Vanacker, V.F. (1948). Syntaxis van het Aalsters dialect. Werken uitgegeven door de

Koninklijke Commissie voor Toponymie en Dialectologie 4. Tongeren: George Michiels.

—        (1963). Syntaxis van de gesproken taal te Aalst en in het Land van Aalst in de XVde, de XVIde en

            de XVIIde eeuw. Brussel: Belgisch Interuniversitair Centrum voor Neerlandistiek.

Van Craenenbroeck, J. (2004). Ellipsis in Dutch dialects. Diss. Universiteit Leiden. Utrecht: LOT.

Van Craenenbroeck, J. & M. van Koppen (2002a). ‘Subject doubling in Dutch dialects.’ In M. van Koppen et al. (eds.) Proceedings of Console IX. 55–64.

—      (2002b). ‘Pronominal doubling and the structure of the left periphery in southern Dutch.’ In S. Barbiers et al. (eds.) Syntactic Microvariation. Meertens Institute Electronic Publications in Linguistics 2.

Vandekerckhove, R. (1993). 'De subjectvorm van het pronomen van de 2e persoon enkelvoud

            in de Westvlaamse dialecten. Taal en Tongval 45. 173–183.

Vandeweghe, W. (2000). 'Pronominale reduplicatie: sporen in het AN.' In V. De Tier, M.

Devos & Jacques Van Keymeulen (eds.) Nochtans was scherp van zin. Huldealbum Hugo Ryckeboer. Gent: Vakgroep Nederlandse Taalkunde RUG. 439–443.

Van Riemsdijk, H. (1999). Eurotyp 5: Clitics in the Languages of Europe. Empirical

            Approaches to Language Typology. Berlin: Mouton de Gruyter.

Will, G. (2004). Zeeuws of Vlaams? (Morfo-)syntactische verschijnselen in de dialecten van

             Zeeuws-Vlaanderen. Oosterhout: eigen beheer.

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

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

WZD = Ghijsen, H.C.M. (1974). Woordenboek der Zeeuwse dialecten [derde druk]. Den

       Haag: Van Goor & Zonen.

Zwart, J.-W. (1993). Dutch Syntax: A minimalist approach. Diss. Universiteit Groningen.