Reforms of Dutch orthography
Anneke Neijt
Afdeling Nederlands and Center of Language Studies, University of Nijmegen
August 8, 2000.
Published with minor amendments in D.Börchers, F. Kammerzell and S. Weninger (2001) Hieroglyphen, Alphabete, Schriftreformen, Lingua Aegyptia, Studia monographica 3, Göttingen 2001, p. 209-222.
Moght iemant segghen / tes te langhe ghebeidt / ghy comt te láte: het hadde wel goed gheweest / hadde déze maniere van spellen ouer viif. of x. honderd iaren beghin ghenómen. Andwoorde: Het es waar.
‘Would someone say / you waited too long / you are too late: it had been acceptable / had this way of spelling started 500 years or 1000 years ago. Answer: it is true.’
Joos Lambrecht, Néderlandsche spellijnghe, 1550
Present-day Dutch orthography has its roots in medieval times, when a variety of orthographic conventions were followed, partly due to dialectal differences in pronunciation. After the invention of printing, a standard version of Dutch orthography slowly evolved, with publications on orthography as landmarks in the history of this development. The oldest publication preserved dates back to 1550: Joos Lambrecht’s Néderlandsche spellijnghe (71 pages). This work was published at a time when standards for Dutch spelling had been formed already, witnessed by Lambrecht's captatio benevolentiae, see above. The enormous variation in use, however, forced the publisher Lambrecht to propose a spelling standard. From that time on until 1800, some fifty books or chapters have been published on the rules and principles to follow when writing Dutch. There remained variation in writing, which not always was due to variation in pronunciation. The newly translated bible of 1637, known as the Statenvertaling, is generally considered one of the publications that helped creating a standard version of Dutch. The orthographic practice followed in this book has been influential, as was the orthography by famous writers such as Vondel (1587-1679), but official recognition of one orthographic system only took place in the nineteenth century. This period of spelling regulation by the governments of Belgium and the Netherlands will be discussed in this paper, as well as the principles of Dutch orthography and the three most important issues of the reforms.
1 The period 1795-1997
After the foundation of the Bataafse Republiek (1795), the need for a standard variant of Dutch was more urgent then before, since the introduction of democracy concurred with political debate amongst representatives of different regions. The Bataafsche Maatschappij van Taal- en Dichtkunde, an organization of men of letters, was officially asked to further the development of a unified orthography. This Maatschappij organized a contest for the question “To what extent should Dutch orthography be ruled by the sounds of the language and by ease of pronunciation?” The contest was won by Matthijs Siegenbeek (1774-1854), professor of Dutch language and oratory in Leiden. Siegenbeek answered the question affirmatively: orthographic rules should be in accordance with euphonic considerations and ease of pronunciation. For instance: Dutch kelder ‘cellar’, etymologically related to Latin cella and cellarium, should be written with d since [d] is present in Dutch pronunciation and schielijk ‘quick’, derived from schier and the suffix -lijk, should not be written with r, since [r] is not present in pronunciation. His proposals were accepted by the government in 1804, and in 1805 Siegenbeek’s spelling dictionary appeared.
In Belgium, French was the official language. Flemish activists unsuccessfully pleaded affiliation with the Netherlands (cf. Willems 1824), until they succeeded in having their native language officially recognized in Belgium in 1873. In the mean time, De Vries and Te Winkel, professor and teacher of Dutch respectively, prepared the publication of a large Dutch dictionary, the Woordenboek der Nederlandsche Taal, by establishing a new set of spelling principles and a new spelling dictionary, differing in minor points from Siegenbeek’s system. The spelling regulations by De Vries and Te Winkel were first accepted by the Belgian government (1864) and later also by the Dutch government (1883). From this time onwards, the Netherlands and Belgium have a common orthography.
De Vries and Te Winkel’s system was heavily criticized, most notably by Kollewijn (1891), who founded a Spelling Simplification Association (cf. Daman 1941). Several official Spelling Committees were formed, but their proposals were turned down. Partial reforms took place in 1930, 1934 and 1936, affecting primarily the writing of case endings, until in 1946 and 1947 the Belgian and Dutch governments decided on a more drastic reform. The new version of the spelling dictionary appeared in 1954. This first phase of the De Vries and Te Winkel spelling period is characterized by fierce opponents (writers such as Multatuli and Couperus), many spelling reform committees, lots of activities by the Simplification Association and counteractivities of other associations. From a governmental point of view, it was a chaotic phase, with a prudent reform in several minor steps, one re-reforming a former reform, and finally a drastic reform. In the mean time, Dutch was written in lots of different ways, as some writers and linguists wrote according to the proposals by the Simplification Association, others wrote according to the partial reforms or followed their own system and still others continued to use De Vries and Te Winkel (Geerts, Van den Broeck and Verdoodt 1977).
The drastic reform of 1954 did not end the discussion on spelling issues, nor did it put an end to governmental action. One aspect of the reform was generally accepted: the decision to allow writing of case endings only where they are present in the spoken variant of the language. Another aspect turned out to be impracticable: the freedom of choice for many non-native words, such as for ‘holidays’ either vakantie or vacantie. Therefore the government decided shortly after the publication of the spelling dictionary that only one of the spelling variants was to be used in official documents. This variant, the so-called “preferred” spelling, was the variant put forward as most appropriate by the spelling dictionary of 1954; in a number of instances this was the dutchified form (vakantie, also allowed vacantie), but in the majority of the words, this was the non-dutchified form (vacant, also allowed vakant ‘vacant’; apotheek, also allowed apoteek ‘pharmacy’). Schools should use this preferred spelling as well, although systematic use of the other variant was not forbidden.
Here, a note on systematic use is in order. Observe that from a linguistic point of view neither of the variants can be called systematic, as the list of preferred forms includes many dutchified spelling forms and the list of allowed forms includes many non-native ones. Schools that used the dutchified versions of all words with two variants thus were breaking the law, but as spelling mistakes were of minor importance to the exam results, no real problems arose. The only less favorable outcome has been that the spelling remained variable for different groups of users and sometimes even within the written production of one user. Stability of a standard for writing was not achieved.
Ten years after the reform, a new spelling committee was asked to investigate the possibility of better rules for the spelling of non-native words, for the spelling of verbal inflection and for the use of linking elements in compounds. The Belgian and Dutch linguists of this committee published their proposals in 1967, which led to fierce discussions between proponents of a simplified spelling, proponents of an even more simplified spelling than the one proposed and opponents of any reform at all. A more detailed proposal was published by the committee in 1969, but the governments decided not to reform at all.
In order to establish cooperate spelling regulation in Belgium and the Netherlands (and to foster language politics in a unified Europe), the Nederlandse Taalunie was founded in 1980. After several years of investigations by spelling committees who advised both on the acceptability of a reform and the possibility of a more systematic spelling, the Taalunie decided on a reform in 1994, the precise contents of which were published in 1995 and became effective in 1996 (Rapport 1988, Spellingrapport 1994, Spellingbesluit 1994, Woordenlijst 1995). Compared to the former one, this latest reform is not drastic. The preferred variant became the offical spelling, except that for some 40 words the non preferred variant was chosen. Minor points are changed: a more precise formulation of where to write diaereses (disambiguating digraphs from sequences of monographs, cf. the name Baäl and the noun baal “bale, bag”) and where to write e and en for linking morphemes in compounds. A more detailed description of the problems solved and created by this final spelling reform is given in Neijt and Nunn (1997). As to the near future: no principled changes are foreseen, but a new spelling dictionary will appear in 2005.
Drastic or not, the spelling reforms of the last two centuries have not changed the principles according to which words are written. And even Siegenbeek, who was the first to formulate spelling principles by government order, based his principles on earlier orthographic standards. Present-day Dutch orthograpy thus has a long history, which began far before governmental interference.
2 Principles
The way Te Winkel (1863) phrased his orthographic principles, which are similar to those of Siegenbeek (1804), has become widely known in The Netherlands and Belgium. Amongst a larger number of principles the most prominent ones are:
Regel der Beschaafde Uitspraak (Principle of Received Pronunciation)
[…] geef door letterteekens al de bestanddeelen op, die in een woord gehoord worden, wanneer het door beschaafde lieden zuiver uitgesproken wordt […]
‘Represent by means of letter signs all constituting parts that are heard in a word, when it is pronounced purely by civilised people.’ (Te Winkel 1863, p. 8)
Regel der Gelijkvormigheid (Principle of Uniformity)
Geef, zooveel de uitspraak toelaat, aan een zelfde woord en aan ieder deel, waaruit het bestaat, steeds denzelfde vorm […]
‘When allowed by pronunciation, use the same form for one and the same word and each of its constituting parts.’ (Te Winkel 1863, p. 12)
Regel der Afleiding (Principle of Etymology)
Bij de keus der gelijkluidende letterteekens beslist de afleiding of de oudere vorm uit den tijd, toen de nu gelijk geworden klanken nog duidelijk onderscheiden konden worden.
‘The choice of one out of different letters for similar sounds is based on the derivation or older form which was in use when sounds that are similar now, could still be distinguished clearly.’ (Te Winkel 1863, p. 14)
Regel der Analogie (Principle of Analogy)
[…] de woorden wier spelling noch door de uitspraak, noch door de gelijkvormigheid, noch door de afleiding wordt bepaald, worden op dezelfde wijze geschreven als andere, wier spelling met zekerheid bekend is en die oogenschijnlijk op overeenkomstige wijze gevormd zijn.
‘Words for which the spelling cannot be decided by means of pronunciation, uniformity, or etymology, are written in the same way as words of the same structure for which the spelling is known for sure.’ (Te Winkel 1863, p. 15).
Dutch orthography thus obeys both the Phonological Principle (cf. the first of these four principles) and the Morphological Principle (the second and fourth one), but the former takes precendence over the latter, because of the proviso “when allowed by pronunciation”. The interpretation of this proviso is not crystal clear however, as shown by the following examples:
phonologically inconsistent, morphologically consistent:
[t] – [d] in hond – honden ‘dog – dogs’
[n] – [m] in onaardig – onprettig ‘not nice – unpleasant’
phonologically consistent, morphologically inconsistent:
[] – [o] in apostel – apostolisch ‘apostle – apostolic’
[d] – [t] in stemde – lekte ‘voted – leaked’
The four rules involved are Auslautverhärtung, Nasal Assimilation, Vowel Reduction in unstressed final syllables and Voice Assimilation in past tense suffixes. Here, it may be that the different interpretation of the proviso relates to the different status of the phonological rules involved, such that for instance Auslautverhärtung and Nasal Assimilation are exceptionless rules, part of the grammar of modern Dutch, whereas the examples of Vowel Reduction and Voice Assimilation are relicts of older stages of Dutch. Further study may clarify this point.
For Te Winkel, the Etymological Principle applied to words with a changed pronunciation, such as wij – wei (‘we – meadow’), which in older stages of Dutch were pronounced [wi] and [w], but nowadays share the same diphtong []. In recent discussions, the principle is interpreted much wider, such that it applies to loan words as well, and hence influences the choice of <c> in electie (‘election’), based on the word’s Latin origin (eligo – electum) and <k> in elektriciteit (‘electricity’), based on the word’s Greek origin (ήλεκτρου).
The Principle of Analogy can be illustrated by the spelling landt (verb stem land plus third person singular t, ‘lands’) analogous to fietst (verb stem fiets, plus third person singular t, ‘bikes’) or the use of a linking s in compounds such as stationsstraat (‘station street’), with an s added because of a comparable compound stationsweg (‘station road’).
Graphotaxis and sub-lexicons
The four principles presented above are by no means a complete account of Dutch orthography, for which also a number of graphotactic rules has been developed and a strategy for incorporating loan words (Nunn 1998). Graphotactic rules apply to letter sequences after morphemes have been concatenated. These are also known as autonomous rules, since they apply to letter strings irrespective of the phonology. Examples of such rules are vowel letter degemination in open syllables and consonant letter degemination at the end of syllables, cf.:
maan – manen (not maanen) ‘moon – moons’
ik eet – hij eet (not eett) ‘I eat – he eats’
In the use of these rules Dutch differs from German, as witnessed by the fact that words with the same etymology and the same pronunciation receive a different spelling (Van Megen and Neijt 1998):
double vowel letter single vowel letter
Du. toon, sjaal Ge. Ton, Schal ‘tone, shawl’
single consonant letter double consonant letter
Du. kan, vol Ge. kann, voll ‘can, full’
Notice that because of these rules, Dutch orthography is morphologically less consistent than German:
Du. toon – tonen Ge. Ton – Tonen ‘tone – tones’
Du. vol – volle Ge. voll – volle ‘full – full+suffix’