Taal en Tongval. Jaargang 58

[p. 273]

Abstracts

Marc van oostendorp en Leonie Cornips
Het morfologische landschap:
het Polen van de Nederlandse dialectologie?

This introductory article provides an overview of the ‘morphological landscape’ of Dutch dialectology. It has been observed frequently that morphology is at the interface of syntax and phonology; the ‘danger’ of this position is that both phonologists and syntacticians will claim large parts of the territory of morphologists as theirs. We argue, on the other hand, that it is exactly this central place of morphology which makes it an excellent territory for variational studies.

Sjef Barbiers & Marjo van Koppen
Een plaats voor tijd in het middenveld van het Nederlands

In Dutch child language and in some dialects of South-Holland a subject pronoun may intervene between the verbal stem and the past tense suffix. On the basis of this phenomenon we argue that (i) Dutch has a Tense-position located in the middle field of the clause, in addition to the clause initial C-position and the clause final V-position; (ii) The past tense suffix is combined with the verbal stem in syntax, not in the lexicon; (iii) The clause final position of finite verbs in embedded clauses is the result of VP-Intraposition; (iv) Parametrized pied piping options explain the variation between Standard Dutch on the one hand and Dutch child language and Dutch dialects on the other.

[p. 274]

Jan Don & Marian Erkelens
Vorm en Categorie

Trommelen (1989) observed that the possible phonological form of underived verbs in Dutch is much narrower than the possible phonological form that underived nouns may take. This observation seems to be some kind of ‘morpheme structure condition’. In an experiment with Dutch adults that we report we find that this condition is more than just a redundancy rule that linguists can formulate but that does not have any psychological reality. We asked ourselves whether Dutch native speakers would be able to make use of this information in deciding whether a specific - nonce- word belongs to the class of verbs or nouns. In order to answer this question, and investigate the psychological reality of the morpheme structure condition, we translated Trommelen's observations into five factors that each are specific for underived nouns in Dutch: true poly-syllabicity, final long vowel, ‘more than superheavy’ rhyme, final schwa, final schwa + [m]. We designed two sets of nonce stems: the first set could only be classified as nouns, according to the five factors; the second set could be either nouns or verbs. The outcomes of our experiment show that Dutch adults indeed prefer the classification ‘noun’ for stems that were designed according these principles, while they show chance behavior on the other set of nonce stems. This implies that native speakers are sensitive to the morpheme structure condition that restricts the set of possible verbs in Dutch.

Ben Hermans
De tonologie van onbeklemtoonde lettergrepen in een Limburgs dialect

It is shown that in Limburg dialects there are three tonal patterns in verbal paradigms: some verbs have a stable level high tone, other verbs a stable falling tone and still others exhibit an alternation between these two tones. The verbs of the first type have an underlying high tone, those of the second type have an underlying low tone, and the verbs of the third type are underlyingly toneless. A quintessential constraint of the analysis requires that syllables in the weak position of a foot should not be too complex. This implies that audible unstressed

[p. 275]

syllables cannot sustain a tone; only empty unstressed syllables are able to do so. Nominal and adjectival inflectional paradigms are more complex. To explain this it is suggested that nominal and adjectival inflectional affixes, but not verbal inflectional affixes, can be parsed outside the foot domain. This allows them to sustain their own tone, even if their vowels are unstressed and audible.

Frans Hinskens met medewerking van Leendert Brouwer
Van Abbegeasterkeatting naar Zansterpijpje.
-ster in noordelijke aardrijkskundige namen.

This article deals with the suffix -ster, as it occurs in geographical names, appellatives and adjectives in the northern part of the Dutch language area. Attention will be paid to its etymology and geographical spread, as well as to aspects of the morphology and phonology of this suffix and its derivates. It is shown how -ster, like the closely related suffix -stra which used to derive family names, resulted from the historical grammaticalisation of an Old Frisian genitive plural form. Synchronically, geographical names derived with -ster appear to be interesting in several respects. -ster's morphological valency turns out to be prosodically conditioned, like in the case of the -stra suffix. Morphologically, these names, appellatives and adjectives hardly differ from ‘regular’ words, phonologically they do not appear to differ at all - again, just like the -stra formations. It can therefore be established that they are part of the web of words of Frisian and Dutch.

Jarich Hoekstra
Uwz âde Friez'ne tonge, de Halbertsma's jonges en andere genitiefconstructies in het Fries

In this article two somewhat marginal genitive constructions in Modern West Frisian are discussed, viz. the genitive plural construction and the genitive of surnames. Whereas the former is restricted to the written language and probably results from recultivation of an Early Modern West Frisian construction, the latter belongs to the spoken language. Both constructions have in common, that a (usually definite) article may precede the genitive, which is impossible in

[p. 276]

the normal prenominal genitive construction. After a description of both constructions, they are compared to other genitive constructions in Frisian and an analysis is proposed. Although the constructions share the possibility of having a definite article with the so-called genitive compounds in Frisian, a relic of the Old Frisian prenominal genitive construction, it is shown that they cannot be structurally identified with them. It is contended that the genitive plural construction and the genitive of surnames derive from the modern prenominal genitive or possessive construction and involve a reinterpretation and reanalysis of the genitive element as a modifier.

Anke Jongkind & Pieter van Reenen
De je-werkwoorden in de Friese dialecten van het G(oeman)TRP

Frisian has two classes of weak verbs. The first of the two classes is much alike the class of weak verbs in Dutch, the second has a typical Frisian character. This second class of verbs is also called the class of je-verbs, because of its infinitive on -je: klop-je ‘to knock’. However, the class of je-verbs and the second class of weak verbs do not fall completely together. For instance, many je-verbs have in the past tense characteristics of strong verbs. In this study we propose a framework in which the differences between je-verbs and second class of weak verbs can been made explicit on the basis of features of 31 je-verbs in 36 Frisian dialects of the Goeman-Taeldeman-van Reenen-Project (GTRP). Analysis of these dialects shows that from a morphological point of view they hardly differ, in contrast to individual verbs which exhibit considerable differences, especially in the past tense.

Gertjan Postma
Uitwendige syntaxis van voegwoordvervoeging en werkwoordelijke inversievormen

From the late middle-Ages onward, most variants of Dutch develop two new properties: double present tense paradigms (direct versus inverted) and inflected complementizers. There are two approaches that relate the two phenomena.

[p. 277]

Goeman (1980) accounts for inverted presents and complementizer inflection in terms of a encliticization from the nominal domain, i.e. the structural subject position. Zwart (1993) develops a formal analysis that casts the opposition in terms of two distinct inflections in the extended verbal domain (C and I-inflection). The diachronic and geographic distribution of both phenomena do not co-occur, however. In this paper, we make a contrastive formal analysis of Goeman's and Zwart's approach. Subsequently are discussed: non-modification of the subject, the ban on the port-manteau effects in the Comp-inflection, correlation with asymmetric V2 and with number neutralisation in the pronominal system, diachronic correlation between decay of negative comp-inflection and rise of pronominal comp-inflection. It is concluded that the origin of comp inflection must be a combined extraction from both the nominal domain and the extended verbal domain.

Ariane van Santen en Nicoline van der Sijs
Een historisch-morfologische database

In this article we want to put forward a proposal for the way in which a historical-morphological handbook of present-day standard Dutch may be compiled. Such a handbook can only be compiled on the basis of a data base, and consequently we will go into the set-up of the data base required. For the compilation of the data base we will follow a four-step plan. As for the time being we will restrict ourselves to derivations, the first step consists in making an inventory of the affixes to be dealt with. Subsequently the derivations with these affixes will have to be collected and described, resulting in a so-called derivation base. On the basis of and in connection with this derivation base an affix base will be compiled. The information provided by this affix base will finally lead to a historical-morphological description of each affix.

[p. 278]

Georges de Schutter
Nouns in /r/ and their Plurals in the Southern Dialects of Dutch

Pluralisation in Dutch is an unstable part of the grammar. Most plurals are derived from stems by means of one of two different endings, <en> and <s>; in the eastern part of the region there is a third ‘current’ way: plural forms without an ending. The rules and tendencies underlying the choice between the alternatives are often conflicting, they have numerous exceptions, and are prone to change. Most dialects of Dutch, including the majority of southern ones, share thse general characteristic with the Standard language. Dialectal variation is studied here in three classes of nouns with monosyllabic stems:

-	those ending in /r/, going back to originally monosyllabic stems (e.g. boer(en) ‘farmer(s)’) (a)
-	those ending in /r/ only in the plural (e.g. blad-blaar(en) ‘leaf-leaves’) (b)
-	those ending in /r/, going back to historical disyllabic stems (e.g. broer(s) ‘brothers’) (c)

In those dialects in which <en> and <s> compete, the tendencies are often different in monosyllabic and disyllabic words; we can thus expect that there is a clash in the cases (b) and (c): the choice may be determined either by the ‘old’ disyllabic status, or by the more recently acquired phonotactic make up.

The table (in English) in the article p. xx confronts the three categories of stems in /r/ with other phonotactic formations (in the case of a//1), and with disyllabic forms that did not suffer reduction (in the cases of b//2 and c//3).

The table illustrates that Dutch and (most) central southern dialects display the same characteristics in the three parallel pair categories. This, however, is not true with b//2 in the southeast, nor with either b//2 or c//3 in the southwest. In these cases the reduced words seem to adjust to a//1, the contemporary formal matches of the reduced stems.

A number of other phenomena also point at this adjustment tendency; most prominent among these is the development of new singular forms ending in /r/ with category b nouns (e.g. blaar < blaar+en, to replace traditional blad ‘leaf’, or aar < aar+en ‘Du. ei(eren), E. egg(s)’).

[p. 279]

Jose Tummers, Dirk Speelman & Dirk Geeraerts
Geografische differentiatie van de adjectivische buigingsalternantie bij neutra in het gesproken Standaardnederlands

In this contribution we will analyze to what extent the inflectional variation of attributive adjectives in spoken standard Dutch is determined by geographical parameters. In particular, we will consider the geographical constraints on the alternation between inflected and uninflected adjectives in a definite NP with a singular neuter head noun, as illustrated by the distinction between het aardig-e kind (‘the nice-inflected child’) and het aardig-∅ kind (‘the nice-zero child’). The following questions will be tackled:

Does a statistical analysis using the Corpus Gesproken Nederlands (‘Corpus Spoken Dutch’) corroborate the hypothesis that adjectival inflection is different in both national varieties of Dutch, viz. Belgian Dutch and Netherlandic Dutch?

Is it possible to refine this distinction between both national varieties in more fine-grainded, regional distinctions in accordance with the existing dialectological state-of-the-art on adjectival inflection?

Gunther De Vogelaer - Magda Devos - Johan van der Auwera
Voegwoordvervoeging: morfologisch of syntactisch verschijnsel?

It is a well-known fact that in a number of Dutch dialects, complementizers agree in person and number with the subject of the embedded clause. Traditionally, complementizer agreement has been conceived of either as a morphological phenomenon or as a syntactic one. Morphological accounts of complementizer agreement consider the agreement marker on the complementizer to be part of an enclitic pronoun; syntactic accounts consider the agreement marker to be a fully functional inflectional affix. Both types of accounts suffer from important drawbacks, however. In this paper, we try to reconcile morphological and syntactic accounts of complementizer agreement, by attributing them a different place in the diachrony of the phenomenon. Whereas complementizer agreement has probably developed as a morphological phenomenon, it has become a syntactically relevant phenomenon in later stages.

[p. 280]

Sera de Vriendt
Mannelijke persoonsnamen in het Brussels Vlaams

Standard Dutch has two types of derived nouns for designing masculine persons, both involving suffixation: the suffix is either indigenous (i.e. Germanic) or a loan (usually of Romance origin). The first type is also found in the Brussels Flemish dialect, as in nomina agentis ending in - er (bekker, from bakke, Standard bakker) and names for inhabitants ending in - eir (Brusseleir, Standard Brusselaar). The second type also occurs in this dialect. However, on the basis of the form and the pronunciation of the base and the suffix, most of these nouns are to be considered as direct loans from French rather than as derivations within the dialect itself (e.g. violonist, Standard violist, French violoniste, from, respectively, vioul, viool and violon).

Derived personal nouns allow us to clearly demonstrate the impact of French on the lexicon of the Brussels Flemish dialect.

Le néerlandais possède deux types de noms dérivés pour désigner des personnes du sexe masculin. Dans les deux cas, le nom est composé d'un radical et d'un suffixe, mais le suffixe peut être soit néerlandais, donc d'origine germanique, soit un suffixe d'emprunt, généralement d'origine romane. Le premier type existe également dans le dialecte flamand de Bruxelles, par exemple dans des noms d'agent en - er (bekker, du verbe bakke, néerlandais bakker) ou des noms d'habitants d'une ville ou d'un village en - eir (Brusseleir, néerlandais Brusselaar). Le second type existe également. Toutefois, au vu de la forme et de la prononciation du radical et/ou du suffixe, la plupart de ces noms doivent être considérés comme des emprunts au français et non comme des formes dérivées par un procédé propre au dialecte. Ainsi la forme dialectale violonist ne peut pas être construite à partir du substantif vioul, mais est empruntée au français violoniste.

Les noms de personnes dérivés constituent un exemple frappant de l'influence du français sur le lexique du dialecte bruxellois.

[p. 281]

Chris De Wulf & Johan Taeldeman
T-deletie in de Nederlandse dialecten: een globaal overzicht

In this article we try to supplement and to confront Goeman's findings on -t-deletion in most of the northern Dutch dialects (see Goeman 1999) with the situation in the southern half of the Dutch/Flemish dialect area. We found out that the most fundamental distinction is to be made between ‘word-internal’ -t-deletion (-t before a pause) and ‘word-external’ -t-deletion (when there is also a right-hand condition).

In both cases phonological and grammatical conditions come into conflict and this results (amongst other things) in a different hierarchy of lexical -t-deletion (e.g. in vuist ‘fist’) and suffixal -t-deletion (e.g. hij vist ‘he is fishing’) in different dialect areas.

The major differences between ‘word-internal’ and ‘word-external’ -t-deletion can be summarised as follows:

(1)	Dialects with ‘word-internal’ -t-deletion display ‘word-external’ -t-deletion as well, whereas dialects with ‘word-external- deletion do not necessarily delete -t in ‘word-internal’ position.
(2)	There is a stronger impact of phonological factors on ‘word-internal’ -t-deletion than on ‘word-external’ deletion.
(3)	The grammatical status of -t, however, has a stronger impact on ‘word-external’ -t-deletion than on ‘word-internal’ deletion.
(4)	The effect of the left-hand phonological context varies: a. ‘word-internal’: the more sonorant the left-hand segment, the less -t-deletion; b. ‘word-external’: vice versa.
(5)	A strong consonant cluster increases the likelihood of -t-deletion in ‘word-internal’ position. Under ‘word-external’ conditions a strong consonant cluster favours -t-reduction but decreases the chance of total deletion: in this case the surrounding consonants seem to ‘protect’ word-final -t from total loss.