Dutch. A linguistic history of Holland and Belgium

[p. 161]

18 Historical morphology

In the introduction to the historical phonology it was mentioned how phonologically conservative Dutch is and thus how useful it can be for shedding light on the phonological development of its sister languages, German and English, which are quite progressive in this regard. With the morphology, the reverse is the case. German is morphologically quite conservative, whereas Dutch (and English even more so) has been very progressive in this field. There has been a continual development since Common Germanic times from synthesis towards analysis in the grammar of all Germanic languages, but this analytical development has gone somewhat further in Dutch than in German. In laymen's terms, one would say that Dutch grammar has become simpler; more specifically this means that the case system, which affected nouns, adjectives and pronouns, has been simplified virtually into extinction, and the verbal system has also undergone simplification since Common Germanic times. Gender, still a problem in Dutch for the foreigner, has been reduced to two categories, whereas German still has three, as in Indo-European. Consequently the relatively conservative, or in laymen's terms, ‘difficult’ grammar of German can often shed light on the morphological development of Dutch. Gothic is used as much in comparative morphology as in comparative phonology, although even here Gothic too has lost several important grammatical categories that we know were common to all Indo-European languages e.g. eight cases. As the process of analysis or simplification has been an ongoing one, the wealth of Middle Dutch texts that have come down to us often preserve grammatical subtleties which have since virtually disappeared e.g. distinction between masculine and feminine nouns, a wider choice of adjectival ending depending on gender and case. On the other hand, the simplifications which are now the norm had already begun in the Middle Ages and Middle Dutch texts thus do not always reflect the situation as it must have been originally. Part reason for this is that the all important weakening of full vowels in unstressed syllables to schwa, which occurred towards the end of the OWLF period (see p. 126), caused the loss of distinction in gender, case and personal form of the verb etc. because such grammatical categories had been reflected in the endings. The weakening of unstressed vowels in endings contributed more than any other single factor to the analytical development of the grammar of Germanic languages; analogy and falling together of originally separate categories followed as a direct result of this phenomenon, which was itself a consequence of the main stress having become fixed on the first syllable of a word (certain prefixes and compounds excepted).

[p. 162]

Case

Case was originally reflected in the endings of nouns and adjectives, articles and pronouns. Nowadays, with the exception of certain standard expressions, only the pronouns preserve distinctions based on whether they are the subject or object of the sentence. The remnants of the cases will be dealt with under the relevant categories.

Gender

Gender in Dutch was originally determined by the stem declension^1. to Which nouns belonged. There were three genders: masculine, feminine and neuter. Gender was determined purely according to the form of the word, a result of its declension, and had little to do with any sexual distinction; thus there are forms today such as het wijf (the woman - pej.), het meisje (the girl).

In Middle Dutch the distinction in declensional endings between masculine and feminine nouns had already begun to break down. Feminine nouns commonly ended in e, as they still do in German, and thus masculine and neuter nouns ending in e could be confused with feminine ones; a change of gender could then take place. Further confusion resulted, for example, from the fact that historically feminine nouns often employed the genitive s, an ending which was originally reserved for masculine and neuter nouns of a particular declension; thus we have today (and already in Middle Dutch) a form such as 's nachts (< de nacht, a feminine noun) by analogy with 's morgens etc. (< de morgen, a masculine noun.)

Neuter nouns still exist as a separate category in Dutch, but common gender nouns, the name given to the combined category of formerly masculine and feminine nouns, only betray a distinction when they are substituted by pronouns, but even here hij/hem is generally used for all nouns with the exception of those that designate a feminine being or nouns that end in historically feminine endings such as -heid, -ing, -iek etc.

1. Broadly speaking a-stem nouns were all masc. or neuter, o-stem nouns were all feminine and i- and u-stems contained words of all genders but were chiefly masc. and fem.; n-stems, or weak nouns, were of all three genders.

Definite articles de/het

The Dutchman knows intuitively whether a given noun is of common gender or neuter and virtually never makes an error. The foreign student is dependent on learning the definite article with each new noun, the article usually being the only reliable indicator of the gender of a noun. De/het are the result of a long history of simplification in Dutch.

The definite articles in all Germanic languages are derived from the demonstratives, as nouns originally had no article at all (whether definite or indefinite) in

[p. 163]

Indo-European languages - compare Gothic, Latin and Russian. The following paradigm for the definite article/demonstrative existed in Middle Dutch, but even there it was no longer rigidly applied:

	M	F	N
N	die (de)^2.	die (de)	dat ('t)	die (de)
A	dien (den)	die (de)	dat ('t)	die (de)
G	des ('s)	der	des ('s)	der
D	dien (den)	der	dien (den)	dien (den)

The unstressed forms de and 't of the nominative have become the definite article in ABN, while the stressed forms of the nominative are nowadays used as the demonstratives - plat still knows die/dat as definite articles and Afrikaans also uses die with this meaning. The simplification that English underwent was in favour of ‘the’ (Dutch de) for the definite article and ‘that’ (Dutch dat) for the demonstrative; German has preserved a paradigm similar to the above one for Middle Dutch, but there the masculine and neuter dative forms still retain the longer sound m, whereas Middle Dutch had weakened it to n. In German the definite article is also commonly used as the demonstrative.

Middle Dutch het has developed from unemphatic dat. 't was originally a proclitic form of the article and still 't or [әt] is the normal pronunciation of het today. The spelling with h is by analogy with the pronoun het (it- see p. 172) with which it fell together in pronunciation. The pronunciation [hεt] is thus a spelling pronunciation and not historically correct.

In Middle Dutch die (and thus later de as the unstressed form) is also used for the masculine sing. acc. and dat., breaking down further the formal distinction between masculine and feminine nouns, an issue which was causing great confusion by the sixteenth century. This led in the seventeenth century to artificial attempts by writers to keep the distinction between masc. and fem. alive. For details on the compulsory writing of den for masculine nouns in the acc. and dat. right up to 1947 see p. 41. The very artificiality of den caused unnecessary problems and was inevitably destined to extinction. Den is still found in standard expressions e.g. op den duur (in the long run) and in the places names Den Haag/Den Bosch where it is a locative dative i.e. at the park/forest.

Other case forms of the definite article, which are often difficult or impossible to separate from those of the demonstrative, also occur in standard expressions or higher style e.g. de heer des huizes (the man of the house), deskundig (expert, lit. ‘of it knowing’), dientengevolge (as a result of that, < *dien te den gevolg), ten slotte (finally, < *te den slot - lit. at the end), ter zake (tot hernat ter in hand, < *te der zaak). These examples also illustrate that prepositions always governed the dative case in Dutch.

2. Even today southern dialects distinguish three genders where den, de and het are used as the definite article for masculine, feminine and neuter respectively. As the paradigm here illustrates, the use of den in this way is unhistorical.

[p. 164]

Nouns

Nouns were classified according to stem i.e. whether the stem ending contained an a, o, i, u or n. A noun was declined according to the stem declension to which it belonged. Because the stem endings were unstressed in Dutch, the distinguishing vowels of the endings were all weakened to ә, written e, by the Middle Dutch period. Therefore only comparison with Gothic or Old English or Old High German (because of the lack of OWLF texts) can determine for sure which declension any particular noun belonged to; the distinction is an historical one and in Dutch it borders on the artificial, even in Middle Dutch. The vowel stem nouns are known as strong nouns, and the consonant stem nouns (mainly n) are known as weak nouns, a terminology borrowed from J. Grimm and one which is also used, although with different connotations, to describe verbs.

a-stems^3.

The following comparison of the paradigms of Gothic dags and Middle Dutch dach (day) illustrate the degree to which Gothic preserves the original stem vowels and Dutch has weakened them to e.

	sing.	pl.		sing.	pl.
N	dags	dagôs	N	dach	daghe
A	dag	dagans	A	dach	daghe
G	dagis	dagê	G	daghes	daghe
D	daga	dagam	D	daghe	daghen

This paradigm of dag in Middle Dutch illustrates several interesting facets of the historical development of Dutch: the alternation of voiced and unvoiced fricative in the plural and singular respectively - the ‘rule of uniformity’ (see p. 150) in spelling was not yet applied. Wherever case endings are added, the root syllable is open and consequently the vowel has been lengthened (see p. 127) - compare NNL dag/dagen (day/days), vandaag (today, < van daghe dat.), daags tevoren (the day before, < daghes gen.).

In the above paradigm, the plural ending -e in all cases but the dative is the result of analogy. The current plural ending -en^4. for the originally masculine and neuter a-stems, as for other strong nouns, is derived in all probability from an analogy with weak nouns (see p. 167); it is already common before the end of the Middle Dutch period. The fact that the dat. pl. already ended in -en must also have helped the process.

The a-stem nouns schoen (shoe - Germ. Schuh) and teen (toe - Germ. Zehe), as

3. Because IE. ŏ > GMC. ă and IE. ă > GMC. ŏ, what are called a- and o-stems here are sometimes referred to as o- and a-stems, but then this is from the point of view of IE. philology.

4. The fact that large areas of Holland and Belgium do not pronounce the final n of the plural may be the result of this n having come by later analogy in the written language only, but it is more likely to be due to a later apocope of the n after it had been applied by analogy; after all, all words in -en, whether nouns, verbs or adjectives, drop the final n in speech.

[p. 165]

their English and German cognates illustrate, did not originally end in -n; this is a plural n which ceased to be felt as a plural ending. The current plural forms schoenen/tenen are in fact a doubling up of the plural ending.

The oscillation that often exists in Dutch between plurals in -en and -s is ultimately to be traced back to an alternative means of forming the plural of a-stems. The s plural is an Ingwaeonic form and thus its ubiquitous presence in English and total absence in German words. Even the much quoted West Flemish sentence from the OWLF period has both vogala (> vogelen) and nestas (> nests) where NNL happens to now have the reverse, vogels and nesten. Plurals in -s were, and still are in the dialects, more common along the coast from West Flanders to North Holland. It is difficult to formulate rules for the use of -s or -en in Modern Dutch, but the origin at least of the alternative in -s is most probably that described here (see p. 59).

Neuter a-stems in GMC and MNL were identical to masc. a-stems except in the nominative and accusative plural where many had no ending and were thus identical to the singular. Remnants of this situation are still to be found e.g. op de been (on one's feet), vijf jaar (5 years), drie pond (3 pounds) - compare also English ‘sheep’, ‘deer’, ‘three pound’.

Quite a large, but nevertheless finite group of common neuter nouns, not all necessarily a-stems by origin however, take a plural in -eren (in Middle Dutch and various dialects also -er or -ers) e.g. ei/eieren (eggs), kalf/kalveren (calves), kind/kinderen (children), lam/lammeren (lambs). This plural formation has but one solitary representative in English, ‘children’, but is much more widely found in German than in Dutch although both languages have extended the use of the ending beyond its origin. In German too it is chiefly a neuter ending, but several masculine nouns have adopted it as well - compare Dörfer/dorpen, Häuser/huizen, Länder/landen; in the above mentioned cases German also has Eier, Kälber, Kinder, Lammer.^5. The ending -en on Dutch words such as land, huis etc. is by analogy with the other large number of nouns ending in -en in the plural. Note the forms kleren (clothes, < klederen) and bladeren (leaves, pron. blaren); occasionally the -eren ending is used beside -en with a distinction in meaning e.g. been (leg, bone) - benen (legs)/beenderen (bones), blad (page, leaf) - bladen (pages)/bladeren (leaves).

5. Strictly speaking, the ultimate origin of the ending is to be found in a small group of words one can best call iz/oz stems (the former with Umlautsfaktor) because the combination -er (with r < z by rhotacism) is also found in the singular of such nouns - compare eierdop (egg-shell), runderlap (cut of beef); Latin gens/generis (nom./gen. sing.) illustrates the same phenomenon.

wa- and ja-stems

If the a-stem ending was preceded by a w or a j in Germanic, the masculine or neuter noun in question was said to be a wa- or ja-stem respectively. From the point of view of Modern Dutch this is only of importance in as far as some words that end in w have retained the w of the wa-stem ending e.g. schaduw (shadow), sneeuw (snow) and the ja-stem nouns, which contained an Umlautsfaktor, have been umlauted in Dutch e.g. heer (army - Goth. harjis), kunne (gender - kin, Goth. kuni); also the e ending of kunne, which is found in other originally neuter stems in Middle Dutch

[p. 166]

and nowadays still in certain dialects too (e.g. bed - bedde, net - nette) is to be traced back to the final i of the nominative case of ja-stems.

o-stems^6.

o-stems nouns were all feminine. The following two paradigms of o-stem nouns (daad - deed, ziel - soul) show the degree to which the declensional endings had already been watered down in Middle Dutch when compared to an o-stem noun in Gothic (giba - gift):

	Sing.	Pl.	Sing.	Pl.	Sing.	Pl.
N	daet	dâde	siele	siele(n)	giba	gibôs
A	daet	dâde	siele	siele(n)	giba	gibôs
G	daet, dâde^7.	dâde	siele(n)	siele(n)	gibôs	gibô
D	daet, dâde	dâden	siele(n)	siele(n)	gibái	gibôm

The ziel paradigm shows particularly clearly the degree to which analogy was applied to the nominal endings after the weakening to e. It is believed that the now ubiquitous plural ending -en in Dutch, originally the plural ending for weak nouns of all genders, probably first made its way into the realm of strong nouns (i.e. vowel stems) via the feminine o-stems, which usually ended in -e in both the singular and the plural (but -en in gen. and dat. pl.). Weak nouns also ended in -e in the nom. sing.; the application of n to the gen. and dat. sing. of ziel in the above paradigm is also in imitation of weak nouns (see p. 167). The ending -en was undoubtedly regarded as a clearer indication of the plural than simply an -e. In English, precisely the reverse occurred and -en is thus found in very few nouns (oxen, children) and the -s (< as) ending of the masc. a-stems was applied across the board. German still retains a diversity of plural formation similar to the original situation in Germanic.

Because both o-stems, which were all feminine, and fem. n-stems ended in e, e became generally recognised as a feminine ending, as it still is in German for the same reason. In Dutch, however, the -e was usually apocopated in the post Middle Dutch period but there are still several originally fem. words that end in -e e.g. beschermvrouwe (patron), gave (gift), hulde (homage), koude (cold), schande (shame), geen sprake van (no mention of).

Final e could have other origins, however (either a ja-stem, i-stem or n-stem) and this formal resemblance often led to a change of gender i.e. to feminine gender e.g. ellende (misery), kudde (flock), kunne (gender), oorlog (war < MNL oorloghe). These were all originally neuter nouns and are now of common gender. Although hart (heart) and oor (ear) are still neuter nouns, for example, in the expressions van ganser harte (with all one's heart) and ter ore komen (to reach one's ear) show them being treated as feminine nouns. Such confusion about inflection and gender was already common in Middle Dutch and was the beginning of the simplified grammar of Dutch. The loss of case and gender in English must have followed a similar path.

6. See footnote 3 on p. 164.

7. Dâde in the gen. was already due to an early analogy with the dative, but in the Middle Dutch period daet, originally only the nom. and acc. form, was being applied by analogy in all cases.

[p. 167]

jo-stems

Just as there were ja-stems beside the a-stems, so there were originally also jo-stems beside the o-stems; this is only relevant from a modern point of view in as far as it explains the umlauted and sometimes geminated forms (better preserved in German) of some feminine words e.g. hel (hell, < helle - Germ. Hölle), brug (bridge, < brugghe - Germ Brücke).

i- and u-stems

Nouns of these two declensions were nearly all masculine or feminine. The u-stem declension was never large and even in Primitive West Germanic they began to join the i-stems. As final i and final u both became e in Middle Dutch anyway, the falling together of these declensions was helped even more. Some Middle Dutch nouns that end in e are thus historically i- or u-stems e.g. sone (son, NNL zoon), beke (brook, NNL beek); also stede (town, NNL stad) where the e < i acted as an Umlautsfaktor. Stede is still found as an ending in place names. The German plural formation in -̈ is also to be traced back to i-stems.

By the i and u endings weakening to -e, and thus falling together in form with feminine o-stems, the stage was set for firstly the feminine u- and i-stems to follow the fem. o-stems, and then for the masc. u- and i-stems, which didn't differ formally from the feminines, to follow suit. As the fem. o-stems were meanwhile asimilating with the weak nouns of all genders, particularly as far as the adoption of the -en plural ending is concerned, the masc. and fem. u- and i-stems also followed suit in this regard. The history of the spread of the -en plural ending in Dutch is a classic example of the far-reaching effects analogy can have on the grammar of a language.

The n-stems or weak nouns

Although strictly speaking the term weak nouns is applied to all consonant stems (i.e. also dr- and nd/nt-stems), the number of non n-stem consonant stems is so small that the term is usually understood to denote n-stems only and the two terms have thus become synonymous.

Weak nouns were chiefly masculine and feminine but a few very common neuter nouns were also weak, namely hart (heart), oog (eye) and oor (ear). Compare the following paradigms of a masc. and fem. weak noun in Middle Dutch and Gothic:

haan (rooster - masc.)
MNL:		sing.	pl.	GOTHIC:	sing.	pl.
	N	hâne	hânen		hana	hanans
	A	hâne	hânen		hanan	hanans
	G	hânen	hânen		hanins	hananê
	D	hâne	hânen		hanin	hanam
tong (tongue - fem.)
MNL:				GOTHIC:
	N	tonge	tongen		tuggô^8.	tuggôns
	A	tonge	tongen		tuggôn	tuggôns
	G	tonge(n)	tongen		tuggôns	tuggônô
	D	tonge(n)	longen		tuggôn	tuggôm

8. Under the influence of Greek orthography Goth. writes ƞg as gg whereas in Dutch ƞg > ƞ.

[p. 168]

The above Middle Dutch paradigms already show analogy in the singular where the usual ending is -e; in German, masc. weak nouns still have -en in the acc., gen. and dat. singular. Interesting relicts of n inflection in the singular in Dutch are the place names 's-Gravenhage (= at the park of the count) and 's-Hertogenbosch (= the forest of the duke) from the nouns graaf (MNL grave) and hertog (MNL hertoghe).

Comparison of the above paradigm of MNL tonge with that of MNL siele (p. 166) shows little formal difference between the two, and in fact they fell together completely at that early date. Separate paradigms for the two, as given here, only have an historic validity even for MNL.

A few masc. words that end in -e in NNL owe this -e to the fact that they were originally weak nouns that did not lose the ending by apocope e.g. getuige (witness), postbode (postman); jongen (boy, < jonghe - Germ. Junge) has adopted an n in the nom. by analogy with the other cases; other nouns that formerly had -en in the acc., gen. and dat. singular now behave like all other nouns but can be compared with their German cognates that still add -en in those cases e.g. graaf/Graf (count), heer/Herr (gentleman), hertog/Herzog (duke), mens/Mensch (person), vorst/Fürst (prince).

The spread of the -en plural ending from weak nouns to all other classes has been dealt with under strong nouns (see p. 167).

Adjectives

Adjectival inflection, of which the meagre remnants in NNL are the presence or lack of a final e, has a long complicated prehistory in Germanic and Indo-European. Suffice it to say that the weak and strong declensions of the adjective, which are still alive and well in German today, were still in existence in Middle Dutch, although the two had begun to fall together and simplify even then.

Historically the weak declensional ending had been used in definite contexts i.e. after de/'t, die/dat, deze/dit and possessives, whereas the strong endings had been used in indefinite contexts i.e. after een/geen, elk/ieder etc. - compare Germ. der gute Mann, ein guter Mann. Because of the confusion between the two originally separate declensions of the adjective - partially due to the adjectival ending often being borrowed from the preceding article (e.g. terzelfdertijd - at the same time) - it is not possible to give complete weak and strong paradigms even for Middle Dutch as one can for NHG. The use of strong and weak endings in Middle Dutch, and thus also in the many standard expressions still used today that contain archaic adjectival endings, was already not entirely in accordance with the historical distinction. The following paradigm represents the situation in Middle Dutch:

[p. 169]

	M	F	N	PL
N	goede, goel	goede, goet	goede, goet*	goede
A	goeden	goede, goet	goede, goet	goede
G	goets	goeder	goets	goeder
D	goeden	goeder	goeden	goeden

Where alternative forms are given for the masc. and neuter, the distinction is due to the weak form existing side by side the strong form (marked *); in the feminine nom. and acc. both weak and strong declensions originally had -e, but in the strong declension it was apocopated. The -en endings can be of either strong or weak origin.

The rules for adjectival inflection in NNL (see p. 60) are based on a very much simplified version of the above paradigm i.e. the absence of any ending on the adjective before a sing. neut. noun in an indefinite context (e.g. een goed huis) is ultimately to be traced back to the strong declension of the adjective, whereas the presence of an -e ending (e.g. het goede huis) is based on the weak declension. Otherwise the endings of the above paradigm are found in countless expressions e.g. goedenavond (good evening, masc. acc.), blootshoofds (with a bare head, neut. gen.), tegelijkertijd (at the same time, fem. dat.), ter elfder ure (at theeleventh hour, fem. dat.). As previously mentioned under nouns, final e was usually apocopated in Dutch, but the distinction between the e ending and no ending, which actually serves no practical function, has remained; this is probably because of the former occurrence of -en (pron. ә) as an adjectival ending i.e. ә < en possibly helped preserve ә < e.

In the seventeenth and eighteenth centuries men of letters attempted to artificially keep such case endings alive. The last vestige of such attempts was the recommended -en ending on adjectives (and articles) for sing. masc. nouns in the acc. and dat. (e.g. van den goeden man), which was not abolished from the written language until 1947 (see p. 41).

The frequent omission of the e ending, particularly before masculine agents (e.g. een goed man), has nothing to do with strong/weak endings; it is purely stylistically (and occasionally semantically) determined. Afrikaans also now inflects chiefly according to sound and number of syllables, not according to grammar.

In the case of adjectival inflection, we thus see Dutch very much occupying a medial position between German on the one hand, where two completely separate paradigms for weak and strong inflection exist, and English on the other hand, where all endings have been dispensed with - compare, however, ‘the olden days’.

Comparative and superlative of the adjective:

The comparative in -er and superlative in -st have, as in German and English, not changed substantially since the earliest records. (For the inclusion of d after r in comparatives, see p. 159)

Numerals

Originally in Germanic the numerals 1-12 were inflected with endings similar to

[p. 170]

those for the adjective (see p. 169), according to gender and case. The only numeral for which there is still a full paradigm of endings in Middle Dutch is een, which, in its unstressed form (әn or 'n), had developed into the indefinite article, as in the other Germanic languages. Remnants of the earlier inflection of een, and less frequently of the other numerals, are still found in the literary language and standard expressions e.g. het leed ener moeder (the sorrow of a mother, fem. gen.), tweeërlei (of two kinds, gen. pl.), met z'n tweeën (two of them, dat. pl., < tween plus analogy with the nominal plural ending -en).

Sometimes the form of the cardinal numeral now used is by origin one of the inflected forms e.g. twee (neut.) but twintig, drie (fem.).

The word geen (no, not a/any- Germ. kein) is found in Middle Dutch as negheen (<* nih + aina not one) and is inflected like een.

For the phonology of vijf (five) see p. 146, for zes (six) p. 151 and acht (eight) p. 153. In negen (nine - Germ. neun, Goth. nium, Lat. novem) we see the change of w>g - compare zag/saw, Eng. drag/draw. Tien (ten - Germ. zehn, Goth. taíhun, Lat. decem) shows, at least in Gothic, h<k via the First German Sound Shift plus later contraction in Dutch, English and German.

The numerals elf/twaalf (eleven/twelve - Goth. áinlif/twalif) were compound words, as comparison with the Gothic shows, where the meaning was something like ‘(ten and) one left’ etc.

Dutch drie/dertien/dertig show the same metathesis of r that occurs in English three/thirteen/thirty - compare Germ. drei/dreizehn/dreiβig.

For the alternation of veertien/veertig with vier see p. 138.

In Middle Dutch the forms viftien/viftich also occur with a shortening of the long vowel in vijf as in English five/fifteen/fifty; these alternative forms also often show the regular Low Franconian shift of ft>cht in Middle Dutch e.g. vichtien/vichtich (see p. 154).

The ending -tig (Goth. tigus) for numerals 20-90 contrasts with palatalised g in Eng. twenty etc. In West Germanic the word hund (ten) was used as a prefix for 70, 80 and 90 - it is still found as hund in Old English texts and as ant in Old Saxon; thus the Middle Dutch forms tseventich, tachtich, tnegentich and, by analogy, also tsestich. Today the former presence of the prefix is still evidenced in tachtig, where the initial vowel facilitated retention, and in the voiceless pronunciation of the initial consonant in zestig and zeventig and, by analogy, also in veertig and vijftig; eastern dialects also still know the form tnegentig. The original division between the two forms -tig and hund- being at 70 reflected an older counting system based on a ‘greater hundred’ (120).

Personal Pronouns

The paradigms for the personal pronouns in Middle Dutch are as follows:

		1st person	2nd person	3rd person (masc.)	3rd person (fem.)	3rd person (neut.)
	N	ic	du	hî	sî	het
Sing.	A	mî	dî	hem	haer	het
	G	mijns	dijns	sijns	haer	(sijns)
	D	mî	dî	hem	haer	hem

[p. 171]

						all genders
	N	wî	ghî			sî
Plural	A	ons	u			hem, hen
	G	onser	uwer			haer
	D	ons	u			hem, hen

Comparison with the forms of the pronouns in Modern Dutch on p. 61 will show considerable differences between the situation in Middle Dutch and that in ABN today.^9. The greatest change has been in the forms of the second person singular and plural, the forms of direct address. The changes that have occurred in Dutch since the Middle Ages are similar to those which have taken place in English over the same period; the following early Modern English forms can thus be useful for comparison:

	singular	plural
N	thou	ye
A	thee	you
G	thine	your
D	thee	you

The plural form ghi was at the same time the polite form of address, both sing. and plural. Du and its related case forms ceased to be used in the sixteenth century^10. and were replaced by the plural forms gij/u. North of the rivers along the coast, however, the alternative palatal forms jij/jou existed. This situation is still reflected in the use of gij and jij in the Low Countries today: in the south, gij is the usual form of address (both sing. and plural), whereas in the north, gij is felt to be an antiquated form with biblical connotations because of its use in the seventeenth century State Translation of the Bible^11. (see p. 108). But as jij stood beside gij in the post Middle Dutch period as both a singular and plural form, both polite and familiar, a need was felt for a new plural form and a new polite form; thus the appearance of jullie (<je + lie[den] = you people^12.) as the familar plural form and u as a polite form in both the singular and plural. U as a subject pronoun is ultimately derived from a seventeenth century form of address, Uwe Edelheid (Your Honour), which was abbreviated in writing to Uwe Edt, U(w)e Ed., U Ed. or U.E., which abbreviation was pronounced uwé and was finally shortened, by a shift of stress to the first syllable, to u. The existence of u as an original object form of gij^13. must have assisted the final shift to u, which was already current in the seventeenth century. U still betrays its third person origins by its ability to be accompanied by a third

9. The unemphatic forms given on p. 61 were also existence in Middle Dutch - see p. 176.

10. Still found in certain dialects, however.

11. Compare the connotations of ‘thou’ in English.

12. Compare the popular form ‘youse’ in Eng. which is apparently also the result of a need to distinguish between the singular and the plural. The dialects also know the forms wullie, gullie, hullie (Afr. hulle) and zullie.

13. Such influence is not without precedent - compare the Eng. subject pron. you (dat/acc. of ye) and the frequent use of hun (instead of zij) as a subject pronoun in colloquial Dutch today; in Swedish too de (nom.) has been replaced by dat. dom in the spoken language.

[p. 172]

person form of the verb (e.g. u is/heeft) and the reflexive pronoun zich; the second person forms u bent/hebt and reflexive pronoun u are also current, however.

In the above Middle Dutch paradigm it will be noticed that the acc. and dat. form of all persons are identical, as they are in English too, but unlike German - compare mich/mir, dich/dir etc. The lack of final r (< Gmc. s/z via rhotacism) in the pronouns is an Ingwaeonic phenomenon - compare he/hij - Germ. er, me/mij - Germ. mir. Historically one would thus expect *mik/mi, *dik/di in Middle Dutch but the dative forms have been adopted in the accusative for all persons - a simplification of the paradigm in other words.

The presence of so many pronouns with initial h is also an Ingwaeonic trait - compare he/hij - Germ. er, hun/hem - Germ. ihnen/ihn, her/haar - Germ. ihr. Comparison of Dutch het with Eng. it and Germ. es shows that the initial h in this case is not etymological; it has been added by analogy with the other personal pronouns in h (see het meaning ‘the’ on p. 163).

The object forms of the plural have changed somewhat since the above situation in Middle Dutch. The forms in -m are clearly originally dative forms which have since been weakened to hen; but beside hen we now also have hun, which is historically purely a regional phonetic variant of hen, with rounding of the vowel. The current distinction made between hen (acc. or direct object) and hun (dat. or indirect object) in the written language, where the spoken language only knows hun, has been artificially imposed by grammarians (see p. 108).

The gen. forms (originally ‘of me’, ‘of you’ etc.) in the above paradigm are the forerunners of the current possessive pronouns. The forms ons (our) and uw (your) have been shortened and a possessive form jouw has developed beside jij/jou. The use of haar for both ‘her’ and ‘their’ (compare Germ. ihr-her, ihr-their) has since been replaced in the plural by hun, the dative pronoun, but in very formal written style haar is still used for the (frequently feminine) plural e.g. de prinsessen en haar echtgenoten (the princesses and their spouses). The unemphatic form of possessive hun, d'r, is a remnant of the situation as it was in former times, however.

The unemphatic forms of the pronoun (see p. 61) have also existed since the earliest times and are commonly found as enclitic and proclitic forms in Middle Dutch e.g. hoordi < hoorde gi or < hoorde gij; hi nam et (he took it) > hi namet > hi naemt, where the proclitic pronoun places the short vowel of the verb in an open syllable and causes lengthening. English too knows unemphatic pronouns but does not write them e.g. ya (< you), ee (< he).

The third person sing. and plural reflexive pronoun zich is of German origin, as the final gutteral fricative indicates. Dutch formerly (and even now in dialects and Afrikaans) used the object form of the third person pronouns as reflexives, as we still do in English e.g. he washes himself/she washes herself/they wash themselves - compare Afr. hy was hom/sy was haar/hulle was hulle. Zich did not become common in Holland until the sixteenth century, under the influence of Reformation literature from Germany. Its incorporation into the State Translation of the Bible in the seventeenth century assured it of a place in ABN. However, in plat one still very commonly hears the following forms, an indication that zich is not an indigenous word: z'n eigen (lit. his own)/d'r eigen (lit. her own, their own).

[p. 173]

Demonstrative and relative pronouns

As the definite article developed from the unstressed forms of the demonstrative pronoun, the original paradigm for the demonstratives is on p. 163. Certain adverbs and conjunctions (actually also adverbial phrases) still retain case forms of die as they appear in that paradigm e.g. sindsdien (since, adv. < since that), indien (if, conj. < in that [case]); also met dien verstande (on the understanding that). The genitive form diens also still alternates with zijn as a possessive pronoun in certain contexts e.g. de gouverneur en diens echtgenote (the governor and his wife).

Originally in Germanic there was no particular relative pronoun, as is still often the case in English and the Scandinavian languages e.g. The man (whom, that) I saw yesterday. Dutch, like German, also gave the function of relative pronoun to the demonstrative pronoun (die, dat) but unlike German uses the interrogative pronoun (wie) when a preposition is involved. The common use of waar as a relative instead of wie in combination with prepositions and the rather plat (also Afrikaans) use of wat as a general relative regardless of antecedent, also show the utilisation of originally interrogative forms; the variety of possibilities of relative pronouns, also well illustrated in English (who[m], that, which, what, -), is a result of the absence of specifically relative forms in Old Germanic. The use of welk as a relative pronoun is probably in imitation of Latin and it is a form which is still only found in formal written style.

The initial w of interrogative forms (e.g. wat-what, waar-where, wanneer-when, welk-which, wie-who) is derived from Germanic forms in hw (< IE. kw via the First German Sound Shift). They are common to all Germanic languages. Middle Dutch knows a paradigm for wie, both as an interrogative and a relative pronoun, which is similar to that for die (p. 163) by which it had been greatly influenced. This accounts for the rather formal genitive forms wiens (masc.) and wier (fem. and pl.- very formal) which are occasionally used in the written language but are more commonly replaced by the periphrastic forms wie z'n/d'r or van wie in speech e.g. Wiens jas is dat? (Whose coat is that? - interr.), De man wiens jas... (The man whose coat...) - compare German Wessen Mantel ist das? Der Mann, dessen Mantel....

The loss of case in Dutch has led to the more frequent occurrence of analytical forms such as waarvan (whose), aan wie (to whom < former dat. wien). The use of van in the genitive is particularly striking in Dutch when compared with English and above all German: het huis van de man (the man's house - Germ. das Haus des Mannes). Constructions like the English are not unknown in Dutch (e.g. mijn vaders auto - my father's car) but they are not nearly as common (see p. 58).

Verbs

Regular and irregular verbs are known as weak and strong verbs respectively in Germanic languages, terms (invented by J. Grimm) which are also applied to nouns and adjectives, but where the concepts which they indicate are of course quite different. There is no distinction between weak and strong verbs in the present tense; only in the imperfect and in the formation of the past participle is the

[p. 174]

difference between the two evident. The dental suffix of the imperfect and past participle of weak verbs is characteristic of all Germanic languages.

Originally in Germanic there were but two tenses - the present tense, which was also used to express the future, and the preterite (now called imperfect), which expressed all actions in the past. However new compound tenses, the perfect and the pluperfect, which are formed from the verbs ‘to be’ and ‘to have’ plus the past participle (previously used only adjectively and substantively) had already begun to develop simultaneously in both Romance and Germanic languages in the old period i.e. prior to 1100.

By the Middle Dutch period, Dutch thus had the following tenses: present tense, imperfect tense, perfect tense, pluperfect tense and a future tense formed with the auxiliary zullen (Eng. shall) plus the infinitive. There were also three moods - the indicative, the imperative and the subjunctive (both past and present) The subjunctive in Dutch has undergone a similar fate to that in English since the Middle Ages and is now only found in standard expressions and occasionally in literary style. The development of the verbal system in Dutch and English has been parallel; only in use of the tenses (see p. 65) have the two languages diverged to any great extent. German, on the other hand, has preserved both the present and past subjunctive.

Present tense

In the present tense, where there was and is no distinction between a weak and a strong verb, a verb was conjugated in Middle Dutch as follows:

	woenen^14. (to live)	nemen (to take)
1	ic woene	ic neme
2	du woens (< woenes)	du neems (< nemes)
3	hi woent (< woenet)	hi neemt (< nemet)
1	wi woenen	wi nemen
2	ghi woent (< woenet)	ghi neemt (< nemet)
3	si woenen	si nemen

The present subjunctive differed from the above only in the third person singular, which was identical to the first person singular.^15.

Comparison of the above conjugation, which stays close to the situation as preserved in Gothic, with that on page 63 reveals the following changes since Middle Dutch times:

The -e ending of the first person singular has been apocopated, chiefly a post Middle Dutch development; it still occurs occasionally in standard expressions e.g. Verzoeke... (I kindly request that...) Verblijve... (I remain...)

The second person singular, like the third person singular and second person plural, usually showed apocope of the vowel in the final syllable in Middle Dutch, but the older longer form is still found in many Middle Dutch texts. The ending of

14. This spelling was common in Middle Dutch for long o.

15. This third person present subjunctive ending is still found in formal imperatives e.g. lang leve de koningin (long live the queen); men neme één ei (take one egg, in recipes).

[p. 175]

the second person singular is now identical to the third person but there has been a change of pronoun since the above situation existed (see p. 171); since the substitution of du with jij (< gij), the ending-t, a second person plural ending, is now used in the singular. In Middle Dutch the ending -st often accompanies du, as in German, the origin of this ending being the inverted form of the verb and subject i.e. nemes - du > nemestu > nemest - du > du nemest. Inversion of jij with the verb ending in t led to assimilation of the t in early Modern Dutch e.g. jij neemt but neem je?

The long vowel in all persons of a verb like nemen above is the result of it having originally been followed by a second syllable which opened the preceding syllable to cause lengthening. Even once the vowel of some endings was syncopated, the analogy with other persons preserved a long vowel throughout - compare German ich nehme (long), du nimmst (short, with Umlaut of e < OHG nimis, OWLF *nemis).

Although jullie has nowadays replaced gij as the second person plural in ABN and can also take a -t ending, it is more usual for it to take -en by analogy with the first and second persons plural.

We see here a unitary plural form where the ending of the first and third persons, which fell together in pre-history, is applied to the second person. In the Saxon based dialects of the east of Holland and northern Germany, however, it is the dental ending of the second person which has been applied analogously to the other two persons. This is also the case in Old English and Old Frisian.

5 Due to its origin, u (see p. 171), which did not exist in Middle Dutch, can take either a second or a third person singular ending, but as both these persons now end in -t, this duality is not usually obvious; the verbs hebben (to have) and zijn (to be) are the only ones where the second and third persons differ (see p. 182).

The imperfect and the past participle of weak verbs^16.

The origin of the dental suffix of the preterite (now imperfect) in Germanic has been the subject of much debate among philologists; it would go beyond the aim of this book to delve into this problem. Suffice it to say here that there is considerable evidence (especially in Gothic) to suggest that the ending is derived from the verb ‘to do’.

The imperfect of a weak verb in Middle Dutch was as follows:

1	ic hoorde	(< ede)
2	du hoordes	(< edes)
3	hi hoorde	(< ede)
4	wi hoorden	(< eden)
5	ghi hoordet	(< edet)
6	si hoorden	(< eden)

16. In Germanic there were originally four classes of weak verbs (as in Gothic) which were distinguished by the vowels in the endings. With the weakening of vowels in unstressed syllables, this distinction was doomed to extinction and eventually all four groups fell together. This was already the situation by the Middle Dutch period where the endings -de or -ede were applied willy-nilly to any weak verb, whereas historically -ede was the ending of the first class of weak verbs, the so-called jan-verbs.

[p. 176]

The imperfect indicative and subjunctive of weak verbs had fallen together in Middle Dutch.

The current alternation of -te(n) and -de(n) endings employed in the formation of the past tense of weak verbs is the result of assimilation of the initial consonant of the ending (i.e. -de < -ede) to the final sound of the stem of the verb. Wherever the medial vowel was preserved in Middle Dutch, although forms without it are already more common by then, assimilation of d to t was prevented. Nevertheless, Middle Dutch spelling is still too erratic to show the regularity modern spelling does with regard to these endings (see p. 63).

In those areas where gij is still used, the ending is as in the above paradigm. Otherwise, as du and its verbal forms have died out and jullie has followed the other plural persons, there is now a complete falling together of the three persons of the singular to -de or -te and of the three persons of the plural to -den or -ten. In practice, of course, all six endings are pronounced the same in that large area of the Low Countries where final n after ә has been apocopated. We witness here a de facto simplification of verbal endings which is analogous to the situation in English - compare I/you/he/we/they worked.

The dental ending of the past participle of weak verbs:

Historically the t/d ending of the past participle is also derived from an original d, but with the same distinction having been applied as for the imperfect. In this instance, however, as final d is always devoiced to t in Dutch, the distinction in spelling is not heard unless the past participle is inflected (i.e. when used adjectively or substantively) e.g. gekocht (bought)/geverfd (painted) - both pronounced t, but het geverfde huis (the painted house), pronounced d.

The prefix ge-

The unstressed prefix ge-, which in both Dutch and German precedes all past participles whether of weak or strong verbs^17., was formerly known in English too, but English, along with Frisian, has dispensed with it^18.; it is still heard in some English dialects in palatalised form as ye-. It occurs in Gothic as ga- but is not yet associated with past participles there; it is by origin a prefix which designated completion (i.e. perfectiveness of an action in general and thus also some infinitives could and still do begin with ge- e.g. geloven (to believe), gelukken (to succeed), geraken (to attain). As past participles by definition stress the perfectiveness of an action, the prefix became very much identified with them in West Germanic and was eventually applied to all - for some strong verbs it became a means of distinguishing between the infinitive and the past participle e.g. vallen (to fall) - gevallen (fallen). Nevertheless there were several verbs which were apparently already felt to be perfective in meaning whose past participles still did not take ge- in Middle Dutch e.g. bracht (brought), komen (come), vonden (found), worden (become). From the fifteenth century these verbs too began to adopt ge-, as did

17. Except when such verbs already begin with an unstressed prefix i.e. be-, her-, ont-, ver- etc. or are inseparable verbs.

18. Loss of the prefix is also found in several Dutch dialects, notably in West Friesland, the north and the east. Other areas know an ә prefix.

[p. 177]

verbs of foreign origin - compare Dutch gereserveerd, gestudeerd etc. and German reserviert, studiert.

The verbs blijven (to remain, < *belijven), blussen (to extinguish, < *belussen - compare Germ. löschen), and vreten (to eat, < *vereten) contain a contracted unstressed prefix which is no longer felt to be such and thus can take ge- in the past participle i.e. gebleven, geblust, gevreten. Compare also Dutch geloven (to believe) - past participle geloofd, German glauben - past participle geglaubt.

Verbs which are often regarded as ‘irregular’ by the student of the language but which are in fact historically weak verbs are brengen/bracht/gebracht (to bring), denken/dacht/gedacht (to think) dunken/docht/ - (to seem); kopen/kocht/gekocht (to buy), plegen/placht/ - (to be used to), zoeken/zocht/gezocht (to seek). Comparison with the English past forms ‘brought/thought/sought’ illustrates that the concept is not foreign to English either. The group is also smaller in German than in Dutch. The past tense of the first three verbs shows the results of a prehistoric compensatory lenghtening (i.e. loss of n plus lengthening of the preceding vowel) with a later shortening of the vowel before the cluster -cht; in English the vowel is still long, unlike Dutch. The shift from final *-kta > -chta, which must have occurred before the loss of the nasal, is thus also prehistoric and common to all West Germanic languages (and Gothic). The latter three verbs have also undergone this shift and lost the final e, which is still present in German (brachte, dachte), by apocope^19.; it was apparently considered superfluous as there are other clear signs that these are past tense forms, which is not the case with other weak verbs. Kopen (kocht) shows the regular Low Franconian shift of ft > cht (see p. 153); ft < pt is comparable to the cht < kt in the other verbs in this group i.e. a shift from voiceless stop to voiceless spirant (Auslautsverschärfung). Brengen/bracht shows alternation of vowels due to Ablaut but the vowel in dacht is the result of Rückumlaut i.e. the e in denken (<*þankjan) goes back to an original a which has been retained in the past, where there was no Umlautsfaktor following.

19. Apocope of final e is characteristic of many aspects of the historical morphology of Dutch and contrasts distinctly with German - see the 1st person sing. of the present tense of verbs (p. 174), loss of -e by feminine nouns (p. 166) and weak nouns (p. 167).

The imperfect and past participle of strong verbs

The degree to which Dutch has preserved the seven Ablaut series known to all Germanic languages is an unusually conservative aspect of its morphology. But even English, in most respects even more morphologically progressive than Dutch, has preserved the original situation quite well. The Ablaut series as they are in Dutch today are on p. 64 and can be compared with the following situation as it was in Middle Dutch:

[p. 178]

1		grîpen	ic greep	wi grēpen	gegrēpen
2	a	bûgen	booch	bōgen	gebōgen
	b	gieten	goot	gōten	gegōten
3	a	binden	bant^20.	bonden	gebonden
	b	werpen	warp^20.	worpen^20.	geworpen
4		nēmen	nam	nâmen	genōmen
5		gēven	gaf	gâven	gegēven
6		vāren	voer	voeren	gevāren
7	a	vallen	viel	vielen	gevallen
	b	hangen	hinc	hingen	gehangen

The only significant changes have been in series 3 where the singular of verbs in group a have adopted the vowel of the plural and past participle. In German the reverse has occurred in the preterite - compare band/banden/gebunden. In addition, those verbs in group b whose stem ends on l or r + consonant, have since adopted the vocalism of group 7 in the preterite e.g. helpen (to help) - hielp, sterven (to die) - stierf, werpen (to throw) - wierp; compare German helfen - half, sterben - starb, werfen - warf.

The alternation of a short vowel and a long vowel in the singular and plural of verbs in series 4 and 5 is an example of the phonological conservatism of Dutch; this is the original Germanic situation whereas German shows analogy with the plural in these instances, as both the singular and the plural contain a long vowel - compare nahm/nahmen, gab/gaben.^21.

The seven Ablaut series show fewer deviations in Dutch than in German, where, for example, group 1 has divided into two sub-sections: German steigen/stieg/gestiegen, schneiden/schnitt/geschnitten and Dutch stijgen/steeg/gestegen, snijden/sneed/gesneden; group 2 in German has verbs with both long and short vowels depending on the consonant that follows: biegen/bog/gebogen^21. (long), riechen/roch/gerochen^21. (short) versus Dutch buigen/boog/gebogen, ruiken/rook/geroken^22.

Strong verbs were conjugated as follows in the past tense in Middle Dutch: nemen (to take)

	indicative	subjunctive
1	ic nam	nâme
2	du names, < naems	nâmes, < naems
3	hi nam	nâme
1	wi nâmen	nâmen
2	ghi nâmet, < naemt	nâmet, < naemt
3	si nâmen	nâmen

20. These forms are those that differ from the situation as it is today.

20. These forms are those that differ from the situation as it is today.

20. These forms are those that differ from the situation as it is today.

21. German spelling does not reflect as regularly as Dutch spelling which vowels are long and which are short.

21. German spelling does not reflect as regularly as Dutch spelling which vowels are long and which are short.

21. German spelling does not reflect as regularly as Dutch spelling which vowels are long and which are short.

22. The Second German Sound Shift created in many cases a double consonant sound in German which locked the preceding vowel into a closed syllable and prevented lengthening, unlike in Dutch, where lengthening in open syllables could take place unimpeded.

[p. 179]

The s ending of the second person sing., which is also found in Middle High German, is not historical - Gothic has namt - but is probably by analogy with the present tense and the subjunctive. Nowadays analogy has reduced the indicative to two forms nam/namen. In areas where gij is still used, its verb still ends in -t and takes the vowel of the plural i.e. in Ablaut series where the plural differs from the singular e.g. gij waart (you were) but jij was.

The imperfect indicative and the imperfect subjunctive of weak verbs had fallen together in Middle Dutch, as in MHG and ME. The imperfect subjunctive of strong verbs, which has now fallen together with the imperfect indicative in Dutch and English, still preserved different forms in the first and third persons singular in Middle Dutch, however - compare als het ware (as it were).

The falling together of the above paradigms which we see taking place here, did not occur in German where Umlaut distinguishes between the two e.g. ich nahm/nähme, ich aβ/äβe^23. etc. Dutch and English, in as far there is any need to distinguish between the two forms, now use a periphrastic construction (see p. 67) i.e. a further analytical development that these two share that German doesn't.

23. The e ending of the imperfect subjunctive was originally an i (i.e. an Umlautsfaktor)and was weakened to e in unstressed syllables.

Strong verbs changing class or becoming weak i.e. mixed verbs

Although Dutch, German and English all still preserve certain patterns among their strong verbs which can be traced back directly to the seven Ablaut series inherited from Old Germanic, in all three languages a degree of interchange between the classes of the strong verb and between strong and weak verbs has occurred. The instances of a strong verb joining another strong class^24. with which it shares one or other phonetic similarity are not as numerous or as obvious, from the point of view of the student, as those of strong verbs that have become (usually only partially) weak. Comparison of such words in related languages can reveal former situations which analogy has since disguised. The English verb ‘to laugh’, for example, is a weak verb, as it is in German (lachen/lachte/gelacht); only when one is confronted with Modern Dutch lachen/lachte/gelachen does one realise that Eng. ‘to laugh’ and German lachen were originally also strong. Lachen, as its past participle in Dutch indicates, was formerly a group 6 verb and it occurs in Middle Dutch as such i.e. lachen/loech - loeghen/ghelachen.

It is group 6 in particular that has shown a tendency to turn weak in the imperfect but the phenomenon is not exclusive to this group:

bakken	(to bake)	bakte	gebakken	(group 6)
laden	(to load)	laadde	geladen	(group 6)
malen	(to grind)	maalde	gemalen	(group 6)
wreken	(to avenge)	wreekte	gewroken	(group 4)
brouwen	(to brew)	brouwde	gebrouwen	(group 2)
weven	(to weave)	weefde	geweven	(group 5)

24. For example: wegen (to weigh, group 5) joined group 2; steken (to sting, group 5) got a new past part., gestoken, by analogy with breken (to break) etc. in group 4. (see also p. 178).

[p. 180]

The term often given to these verbs is mixed verbs. They are the product of various analogies such as we hear children and foreigners and even ourselves applying daily to English verbs e.g. snuck < to sneak, shat < to shit; striked < to strike.

What we now call group 7 is, from an Indo-European and Old Germanic point of view, a separate class altogether from groups 1-6. It contained the so-called reduplicating verbs which formed their past with or without Ablaut of the root vowel of the verb plus the addition of a reduplicating prefix. However, as such reduplicated forms are not recorded in any of the West Germanic languages^25., not even in the oldest period, tradition has assigned them the status of group 7 in the Ablaut series as they now form their past tense by the application of Ablaut in the same way as other strong verbs. That they have at one time had a special status is possibly reflected in the fact that many of these group 7 verbs are now mixed verbs or have become totally weak:

heten	(to be called)	heette	geheten
spannen	(to stretch)	spande	gespannen
stoten	(to push)	stootte	gestoten
vouwen	(to fold)	vouwde	gevouwen
vloeken	(to swear)	vloekte	gevloekt
zaaien	(to sow)	zaaide	gezaaid
waaien	(to blow)	waaide	gewaaid
		also woei (< group 6)

25. The past tense of doen (to do) is a possible exception (see p. 182).

Modal verbs or preterite-presents

The term modal verb, which all readers should be acquainted with, to denote the verbs kunnen (to be able), mogen (to be allowed to), moeten (to have to) and zullen (will), is of little validity or utility when one looks at the rather irregular conjugations of such verbs from an historical point of view. In terms of historical linguistics these verbs belong to the so-called preterite-presents, a class to which several other verbs also did or do belong. A preterite-present, a concept known to other Indo-European languages as well as to all other Germanic languages, is a verb which in form is a past tense (kan, mag - compare nam-took, lag-lay) but in meaning is a present (ik kan-I can, am able; ik mag - I may, am allowed to). The origin of such verbs lies in a time when there was but one form of the past tense, the preterite, which thus also rendered the perfect tense i.e. ik kan - lit. I have come to know (preterite) therefore I can (present).

The preterite-presents, like the imperfect of strong verbs (actually preterites), know only one form for all persons in the singular and one for the plural:

present:
kunnen	mogen	zullen	moeten
ik kan	mag	zal	moet
jij/u kan/kunt^26.	mag	zal/zult^26.	moet
hij kan	mag	zal	moet
wij kunnen	mogen	zullen	moeten
jullie kunnen	mogen	zullen	moeten
zij kunnen	mogen	zullen	moeten

26. Younger analogical forms with the plural gij to which jij is related.

26. Younger analogical forms with the plural gij to which jij is related.

[p. 181]

imperfect:
kon (< konde)/konden	mocht/mochten	zou (< zoude)/zouden	moest/moesten
past participle:
gekund	gemoogd, gemogen	gezuld	gemoeten

The infinitives have been formed by analogy with the plural forms of the present according to which Ablaut series the verb is derived from.^27. Similarly, as the present tense forms are preterites in origin, new analogous imperfects had to be invented; they are all weak. The past participles are a good example of the options that were at the disposal of these verbs for forming their past tenses - kunnen and zullen have formed past participles by analogy with weak verbs, moeten by analogy with strong verbs while mogen has exploited both possibilities; nowadays the strong form of mogen is the more common, however.

Weten (to know - compare ‘to wit/I wot’) is also a preterite-present by origin i.e. ‘I have seen’ thus ‘I know’ (compare Lat. vidi - I have seen). For this reason it also has irregular past forms: wist^28./geweten - compare German wissen/ich weiβ/wuβte/gewuβt.

Willen (to want to):

This verb, although usually regarded as a modal verb, has a different origin from the other modals. It is not a preterite-present but, as comparison with Gothic clearly shows, a preterite subjunctive by origin where the meaning must have been something like ‘I would like’ therefore ‘I want’ - compare German ich möchte:

ik wil	wij willen
jij wil/wilt	jullie willen
hij wil	zij willen

imperfect: wou (< woude < *wolde), wilde/wilden
past part: gewild

The imperfect forms wilde(n) and the past part. gewild are by analogy with weak verbs, as is wou, but here we see an Ablaut variant common in Hollands and competing with the Flemish form wilde; the plural wouden (pron. wouwen) is never written and is regarded as plat Hollands. In English too, forms with i and o alternate e.g. will, won't, would.

It should be noted that Dutch retained the verb zullen (Eng. shall) as the auxiliary for forming the future tense, whereas English, after letting ‘shall’ and ‘will’ compete for some time, has finally opted for the latter to form the future tense.

27. The forms of the preterite-presents enable us to identify which Ablaut series they originally belonged to e.g. weten (1), kunnen (3), zullen (4), mogen (5), moeten (6).

28. With apocope of the final e (see p. 177). The i of the past tense alternates with the e of the infinitive because short i lengthened to long e in open syllables - compare schip (ship)/schepen (ships).

[p. 182]

The monosyllabic verbs

The six monosyllabic verbs known to Dutch are of various origins. Doen, staan and gaan are original root verbs. Doen (to do) with its past tense deed/deden is possibly the only surviving example of a reduplicating verb (see p. 180). Staan (to stand) and gaan (to go) contrast with German stehen and gehen, where the h is purely orthographical, and they too are usually pronounced as one syllable. West Germanic knew forms *gangan and *standan. The past tense of staan, stond/stonden, as well as the English infinitive, are clearly derived from this alternative form. The past tense of gaan, ging/gingen, is the result of an analogy between *gangen and hangen (to hang) and vangen (to catch) which are class 7 verbs; in Middle Dutch (and Old English) they too occur as the younger contracted forms haen and vaen (< vāen <*vāXen<*vanXen). In plat, the analogical form sting (= stond) is heard as well as gong (i.e. ging).

The verbs slaan (to hit - Germ. schlagen) and zien (to see - Germ. sehen) are not root verbs by origin but later contractions of forms with medial [X] - compare Gothic slahan, saíhvan.

Zijn

The final monosyllabic verb, ‘to be’, shows great irregularity of form as it has in fact inherited its constituent parts from various verbs - compare ‘I go/I went’. In simple terms there are forms derived from b-roots (Eng. be), s-roots (Lat. essere) and the verb wezen (Eng. was/were), a strong verb belonging to group 5. The present tense in Middle Dutch was as follows:

ic bem/ben	wi sijn
du best/bist	ghi sijt
hi es/is	si sijn

In Modern Dutch, du and its form of the verb have died out, but jij has preserved the b-root in jij bent by analogy with the first person. Jullie has fallen into line with the other plural forms so that we now have only one form in the plural. In plat, the analogous form benne is sometimes heard throughout the plural.

The verb wezen, which also exists as an alternative infinitive to zijn in Modern Dutch, has provided zijn with its past tense was/waren (also the imperfect subjunctive form ware) and its weak past part. geweest. The strong past part. gewezen, which German uses, now exists in Dutch only as an adjective e.g. de gewezen burgemeester (the ex-lord mayor). Dutch was/waren and English was/were show the alternation of s and r (< z) as a result of grammatical change; in German war/waren, the consonant of the plural has been adopted by analogy into the singular. The alternative infinitive wezen also provides the imperative - wees braaf (be good) - but in Middle Dutch the b-root (Eng. be) and s-root (Germ. sei) forms also existed. The present subjunctive form zij is now only found in certain standard expressions e.g. God zij dank (God be thanked).

Hebben

The verb ‘to have’ is historically a weak verb but it now shows irregularities which are peculiar to it. To begin with, only hebben, along with zijn, has a different ending

[p. 183]

in the present tense for the second and third person singular i.e. jij hebt, hij heeft and thus u hebt or heeft (see p. 171).^29. Heeft occurs in Middle Dutch as hevet, where v could alternate regularly with b in intervocalic position, and the vowel of the ending is preserved, which thus lenghtened the root vowel in an open syllable. The modern form is a contraction of the Middle Dutch form. Jij has preserved the second person plural ending indigenous to gij, which it has replaced. Dutch hebben contrasts with German haben (also zeggen/sagen - to say) as it joined the first class of weak verbs, which contained an Umlautsfaktor in the ending, in the pre-Middle Dutch period. In the past tense and past part. the original a has been preserved and the b has been assimilated i.e. had (older hadde)/gehad - compare Germ. hatte (also with assimilation)/gehabt. Such alternation of vowels between the present and past tense forms is found in many more verbs in German e.g. kennen (to know) - kannte/gekannt; rennen (to run) - rannte/gerannt^30. - compare Dutch kennen - kende / gekend; rennen - rende/gerend.

29. Plat Hollands uses heb in the third person singular.

30. This is due to a phenomenon known as Rückumlaut (see p. 177).

Bibliography

LOEY, A. van Schönfeld's historische grammatica van het Nederlands. Thieme, Zutphen, 1959 (see p. 160).

BREE, C. van Leerboek voor de historische grammatica van het Nederlands. Wolters-Noordhoff, Groningen, 1977.