|Part of a series on|
Dutch orthography uses the Latin alphabet according to a system which has evolved to suit the needs of the Dutch language. The spelling system is issued by government decree and is compulsory for all government documentation and educational establishments.
- 1 Alphabet
- 2 Sound to spelling correspondences
- 3 Loanwords
- 4 Vowel length
- 5 Final devoicing and the 't kofschip rule
- 6 Diacritics
- 7 Apostrophe
- 8 See also
- 9 External links
- 10 Bibliography
- 11 References
The modern Dutch alphabet consists of the 26 letters of the ISO basic Latin alphabet and is used for the Dutch language. Five (or six) letters are vowels and 21 (or 20) letters are consonants. The letter E is the most frequently used letter in the Dutch alphabet, usually representing a schwa sound. The least frequently used letters are Q and X.
|Letter||Letter name||Spelling alphabet|
- The NATO phonetic alphabet is also used, and sometimes the two are even mixed.
- Standard Dutch pronunciation guide by P.C. Paardekooper
- The letters ⟨q⟩, ⟨x⟩, and ⟨y⟩ occur mostly in words borrowed from other languages, but may also appear in words and names that reflect older spelling conventions. ⟨Q⟩ is almost always followed by ⟨u⟩ (that is, ⟨qu⟩), because nearly every word with a ⟨qu⟩ is borrowed from French or Latin.
- Normally, ⟨y⟩ is generally called /ɛɪ/. However, when used in common speech and/or the need arises to distinguish the letter from ⟨ij⟩, it is most often referred to as Griekse ij (sometimes written ⟨Griekse Y⟩ )('Greek Y'); i-grec, a French word having a similar meaning; or ypsilon.
- The digraph IJ behaves like a separate letter for capitalisation. In alphabetically ordered lists, IJ may intermix with Y (usual for telephone directories) or come between ii and ik (common in dictionaries). In Dutch primary education the (more common) digraph IJ often replaces the (less common) Y as the 25th letter of the alphabet.
Sound to spelling correspondences
Dutch uses the following letters and letter combinations. Note that for simplicity, dialectal variation and subphonemic distinctions are not always indicated. See Dutch phonology for more information.
The following list shows letters and combinations, along with their pronunciations, that are found in modern native or nativised vocabulary:
- Used only in the suffix -isch /iːs/ and its form -ische /iːsə/.
- In words like thans, thee and thuis.
- ⟨e⟩ and ⟨i⟩, when unstressed, are sometimes (but not always) pronounced /ə/.
- ⟨ij⟩ is normally /ɛi̯/. It is exceptionally /i(ː)/ in the word bijzonder, and /ə/ in the suffix -lijk.
The following additional letters and pronunciations appear in non-native vocabulary, or words using older, obsolete spellings.
- ⟨c⟩ before ⟨e⟩, ⟨i⟩, ⟨y⟩, is pronounced /s/ (or /tʃ/ in some loanwords from Italian), and /k/ otherwise.
- The cedilla is used to indicate pronunciation as /s/ when ⟨c⟩ is followed by either ⟨a⟩, ⟨o⟩, or ⟨u⟩.
- ⟨ch⟩ is pronounced /k/ in Italian loanwords, /ʃ/, /tʃ/ in loanwords from other sources.
- ⟨g⟩ may be pronounced /ʒ/ or /dʒ/ before ⟨e⟩, ⟨i⟩ or ⟨y⟩ in words of Romance or English origin.
- ⟨(i)ll⟩ is found in words from French or occasionally Spanish.
- ⟨j⟩ is pronounced /x/ in Spanish loanwords, as a postalveolar in loanwords from other sources.
- The letter ⟨ñ⟩ occurs only in a few Spanish loanwords.
- Used only in some proper names like Zutphen. Words of Greek origin are nowadays written with ⟨f⟩, but formerly with ⟨ph⟩.
- In the rare cases when ⟨q⟩ is not followed by ⟨u⟩, it is pronounced /k/. In those cases, ⟨q⟩ is most likely a result of transliteration from languages such as Arabic.
- ⟨qu⟩ is normally pronounced /kʋ/, but as /k/ before ⟨e⟩ or ⟨i⟩ in loanwords from French or Spanish.
- ⟨s⟩ may be pronounced /z/ before a vowel in words of foreign origin.
- /ʃ/ is used mostly in words of German origin. It is also /s/ finally in older Dutch spellings, which are nowadays spelled with just ⟨s⟩
- /ʃ/ and /tʃ/ only occur as independent phonemes in foreign words. These sounds are approximated using the "native" spellings ⟨sj⟩ and ⟨tsj⟩ respectively.
- ⟨ti⟩ followed by a vowel is pronounced /(t)si/ in loanwords of Latin origin.
- ⟨th⟩ is /t/ in words of Greek origin. In English loanwords, it is most often approximated with /s/ or /z/ according to the English pronunciation, or speakers may attempt to pronounce dental fricatives as in English.
- ⟨x⟩ is sometimes pronounced /ɡz/ between vowels, usually in the southern dialects.
- ⟨z⟩ is /ts/ in words of German or Italian origin. In Italian words it may also be /dz/ between vowels.
- In diphthongs, ⟨ij⟩ and ⟨y⟩ are obsolete spellings for ⟨i⟩. They are found mostly in names.
- ⟨ae⟩ is an obsolete spelling for ⟨aa⟩. In Latin words it is treated identical to ⟨ee⟩.
- In words of French origin.
- ⟨oi⟩ is pronounced /ʋɑ/ or /ʋaː/ in French loanwords. In archaic Dutch spellings, mostly names, it may also be /oː/ (for example Oisterwijk or Helvoirt).
- ⟨oo⟩ is generally /u(ː)/ in English loanwords.
- ⟨ou⟩ is generally /u(ː)/ in loanwords from French or Greek.
- ⟨ui⟩ is pronounced /ʋi(ː)/ in loanwords from French. In English loanwords such as cruise, it is /uː/.
- ⟨y⟩ is usually treated identically to ⟨i⟩ or ⟨ie⟩. It is also an obsolete spelling variant of ⟨ij⟩.
Loanwords often, but not always, conserve their original spellings, e.g. cadeau /kaˈdoː/ 'gift' (from French). The Latin letters c, qu, x, and y (from Greek υ) are sometimes adapted to k, kw, ks, and i, sometimes not. The Greek letters φ and ῥ become f and r and not ph or rh, but θ mostly becomes th (except before consonant, after f or ch and at the end of words). The combinations -eon-, -ion-, -yon- in loanwords from French are written with a single n: e. g., mayonaise, except when a schwa follows, e. g. stationnement.
Vowel length is always indicated, but it is indicated in different ways, using an intricate system of single and double letters.
Old Dutch still possessed phonemic consonant length in addition to phonemic vowel length, with no correspondence between them. Thus, long vowels could appear in closed syllables, and short vowels could occur in open syllables. In the transition to early Middle Dutch, short vowels were lengthened when they stood in open syllables. Short vowels could now only occur in closed syllables. Consonants could still be long in pronunciation, and acted to close the preceding syllable. Therefore, any short vowel that was followed by a long consonant remained short.
The spelling system used by early Middle Dutch scribes accounted for this, by indicating the vowel length only when it was necessary (sometimes by writing a double vowel, but also in other ways). As the length was implicit in open syllables, it was not indicated there, and only a single vowel was written. Long consonants were indicated usually by writing the consonant letter double, which meant that a short vowel was always followed by at least two consonant letters, or by just one consonant at the end of a word.
Later in Middle Dutch, the distinction between short and long consonants started to disappear. This made it possible for short vowels to appear in open syllables once again. Because there was no longer a phonetic distinction between single and double consonants (they were both pronounced short now), Dutch writers started to use double consonants to indicate that the preceding vowel was short, even when the consonant had not been long in the past. This eventually led to the modern Dutch spelling system.
Checked and free vowels
Modern Dutch spelling still retains much of the details of the late Middle Dutch system. The distinction between checked vowels and free vowels is important in Dutch spelling. A checked vowel is one that is followed by a consonant in the same syllable (the syllable is closed), while a free vowel is one that is not (the syllable is open). This distinction can apply to pronunciation or spelling independently, although a syllable that is checked in pronunciation will always be checked in spelling as well (except in some unassimilated loanwords).
- Checked in neither: la-ten /ˈlaː.tə(n)/
- Checked in spelling only: lat-ten /ˈlɑ.tə(n)/
- Checked in both: lat /lɑt/, lat-je /ˈlɑt.jə/
A single vowel that is checked in neither is always long/tense, a vowel that is checked in both is always short/lax. The following table shows the pronunciation of the same three-letter sequence in different situations, with hyphens indicating the syllable divisions in the written form, and the IPA period to indicate them in the spoken form.
|Letter||Free in both||Checked in both|
|a||aː||ra-men /ˈraː.mə(n)/ ("windows, to estimate")||ɑ||ram /rɑm/ ("ram")||ram-pen /ˈrɑm.pə(n)/ ("disasters")|
|e||eː||te-len /ˈteː.lə(n)/ ("to cultivate")||ɛ||tel /tɛl/ ("count")||tel-den /ˈtɛl.də(n)/ ("counted")|
|i||i(ː)||Ti-ne /ˈti.nə/ (a name)||ɪ||tin /tɪn/ ("tin")||tin-ten /ˈtɪn.tə(n)/ ("tints")|
|o||oː||ko-per /ˈkoː.pər/ ("copper, buyer")||ɔ||kop /kɔp/ ("cup, head")||kop-te /ˈkɔp.tə/ ("headed [a ball]")|
|u||y(ː)||Lu-kas /ˈly.kɑs/ (a name)||ʏ||luk /lʏk/ ("succeed")||luk-te /ˈlʏk.tə/ ("succeeded")|
Free ⟨i⟩ is fairly rare, mostly confined to loanwords and names. As tense /y/ is rare except before /r/, free ⟨u⟩ is likewise rare except before ⟨r⟩.
The same rule applies to word-final vowels, which are always long because they are not followed by any consonant (but see below on ⟨e⟩). Short vowels not followed by any consonant do not normally exist in Dutch, and there is no normal way to indicate them in the spelling.
Double vowels and consonants
When a vowel is short/lax, but is free in pronunciation, the spelling is made checked by writing the following consonant doubled, so that the vowel is kept short according to the default rules. This has no effect on pronunciation as modern Dutch does not have long consonants.
- ram-men /ˈrɑ.mə(n)/ ("rams, to ram")
- tel-len /ˈtɛ.lə(n)/ ("to count")
- tin-nen /ˈtɪ.nə(n)/ ("made of tin")
- kop-pen /ˈkɔ.pə(n)/ ("cups, heads, to head [a ball]")
- luk-ken /ˈlʏ.kə(n)/ ("to succeed")
When a vowel is long/tense, but is still checked in pronunciation, it is necessarily checked in spelling as well. A change is thus needed to indicate the length. This is done by writing the vowel doubled. Doubled ⟨i⟩ does not occur.
- raam /raːm/ ("window"), raam-de /ˈraːm.də/ ("estimated")
- teel /teːl/ ("cultivate"), teel-de /ˈteːl.də/ ("cultivated")
- koop /koːp/ ("buy, sale"), koop-sel /ˈkoːp.səl/ ("something bought")
- Luuk /lyk/ (a name)
The letter e
A single ⟨e⟩ indicates not just short and long e, but is also used to indicate the neutral schwa sound /ə/ in unstressed syllables. Because the schwa is always short, ⟨e⟩ is never followed by a double consonant when it represents /ə/.
- ap-pe-len /ˈɑ.pə.lə(n)/ ("apples")
- ge-ko-men /ɣə.ˈkoː.mə(n)/ ("(has) come")
- kin-de-ren /ˈkɪn.də.rə(n)/ ("children")
A word-final long /eː/ is written ⟨ee⟩ (or ⟨é⟩ in some loanwords) as an exception to the normal rules. This means that a word-final single ⟨e⟩ will always represent a schwa.
- jee /jeː/ (expression of woe), je /jə/ ("you")
- mee /meː/ ("along, with"), me /mə/ ("me")
- wee /ʋeː/ ("contraction of the womb"), we /ʋə/ ("we")
Because the position of the stress in a polysyllabic word is not indicated in the spelling, this can lead to ambiguity. Some pairs of words are spelled the same, but ⟨e⟩ represents either stressed /ɛ/ or /eː/ or unstressed /ə/ depending on how the stress is placed.
- be-de-len /ˈbeː.də.lə(n)/ ("to beg") or /bə.ˈdeː.lə(n)/ ("to impart with, to grant")
- ver-gaan-de /ˈvɛr.ˌɣaːn.də/ ("far-going, far-reaching") or /vər.ˈɣaːn.də/ ("perishing")
The length of a vowel generally does not change in the pronunciation of different forms of a word. However, in different forms of a word, a syllable may alternate between checked and free depending on the syllable that follows. The spelling rules nonetheless follow the simplest representation, writing double letters only when necessary. Consequently, some forms of the same word may be written with single letters while others are written with double letters. This commonly occurs between the singular and plural of a noun, or between the infinitive and conjugated forms of verbs. Examples of these alternations are shown below. Note that there are no examples with /i/, because free ⟨i⟩ does not occur in words of native origin.
|When free||When checked||Short/lax
|When checked||When free|
|aː||laten /ˈlaːtə(n)/ ("to let")||laat /laːt/ ("(I) let")||ɑ||lat /lɑt/ ("lat")||latten /ˈlɑtə(n)/ ("lats")|
|eː||leken /ˈleːkə(n)/ ("appeared", plural)||leek /leːk/ ("appeared", singular)||ɛ||lek /lɛk/ ("(I) leak")||lekken /ˈlɛkə(n)/ ("to leak")|
|–||–||–||ɪ||til /tɪl/ ("(I) lift")||tillen /ˈtɪlə(n)/ ("to lift")|
|oː||bonen /ˈboːnə(n)/ ("beans")||boon /boːn/ ("bean")||ɔ||bon /bɔn/ ("ticket")||bonnen /ˈbɔnə(n)/ ("tickets")|
|y(ː)||muren /ˈmyːrə(n)/ ("walls")||muur /myːr/ ("wall")||ʏ||mus /mʏs/ ("sparrow")||mussen /ˈmʏsə(n)/ ("sparrows")|
There are some irregular nouns that change their vowel from short/lax in the singular to long/tense in the plural. In these nouns, the spelling does not alternate between single and double letters. However, the sound /ɪ/ becomes /eː/ in the plural in such nouns, not /iː/, and this is reflected in the spelling. Examples are:
- dag /dɑx/ ("day"), da-gen /ˈdaː.ɣə(n)/ ("days")
- weg /ʋɛx/ ("road, way"), we-gen /ˈʋeː.ɣə(n)/ ("roads, ways")
- schip /sxɪp/ ("ship"), sche-pen /ˈsxeː.pə(n)/ ("ships")
- lot /lɔt/ ("lottery ticket"), lo-ten /ˈloː.tə(n)/ ("lottery tickets")
As a rule, the simplest representation is always chosen. A double vowel is never written in an open syllable, and a double consonant is never written word-finally or when next to another consonant. A double vowel is also never followed by a double consonant, as this could be simplified by writing them both single. There are a few exceptions to this, however.
The past tense of verbs may have a double vowel followed by a double consonant to distinguish these forms from the present tense.
- ha-ten ("hate"), haat-ten ("hated"), both /ˈɦaː.tə(n)/
- ra-den ("guess"), raad-den ("guessed"), both /ˈraː.də(n)/
Compounds should be read as if each word were spelled separately, and may therefore appear to violate the normal rules. This may sometimes cause confusion if the word is not known to be a compound.
- dag-ar-bei-der /ˈdɑx.ˌɑr.bɛi.dər/ or more fluently /ˈdɑ.ˌɣɑr.bɛi.dər/ ("day labourer"), a compound of dag ("day") + arbeider ("labourer"). So it is not divided as *da-gar-bei-der */ˈdaː.ˌɣɑr.bɛi.dər/. If it weren't a compound, it would be written *dag-gar-bei-der to keep the first "a" short.
- een-en-twin-tig /ˈeː.nən.ˌtʋɪn.təx/ ("twenty-one"), a compound of een ("one") + en ("and") + twintig ("twenty"). If it weren't a compound, it would be written *e-nen-twin-tig, to avoid having a double vowel at the end of a syllable.
- mee-doen /ˈmeː.dun/ ("to participate"), a compound of mee ("along (with)") + doen ("do"). If it weren't a compound, it would be written *me-doen, to avoid having a double vowel at the end of a syllable. The word mee itself has a double vowel because of the exception with final -e, as noted above.
Final devoicing and the 't kofschip rule
Final devoicing is not indicated in Dutch spelling; words are usually spelled according to the historically original consonant. Therefore, a word may be written with a letter for a voiced consonant at the end of a word, but pronounced with a voiceless consonant in that position:
- heb /ɦɛp/ "(I) have", but hebben /ˈɦɛbə(n)/ "to have"
- paard /paːrt/ "horse", but paarden /ˈpaːrdə(n)/ "horses"
- leg /lɛx/ "(I) lay", but leggen /ˈlɛɣə(n)/ "to lay"
Weak verbs form their past tense and past participle by addition of a dental, ⟨d⟩ or ⟨t⟩ depending on the voicing of the preceding consonant(s) (see Assimilation (linguistics)). However, because final consonants are always devoiced, there is no difference in pronunciation between these in the participle. Nonetheless, in accordance with the above rules, the orthography operates as though this devoicing did not take place; the consonant that is written in the past participle matches that of the other past tense forms where it is not word-final. To help memorise when to write ⟨d⟩ and when ⟨t⟩, Dutch students are taught the rule "'t kofschip is met thee beladen" ("the merchant ship is loaded with tea"): if the verb stem in the infinitive ends with one of the consonants of "'t kofschip" (-t, -k, -f, -s, -ch or -p), the past tense dental is a -t-; otherwise it is a -d-. However, the rule also applies to loanwords ending in -c, -q or -x, as these are also voiceless.
|Dutch||Meaning||Dutch sentence||English corresponding sentence|
|werken||to work||ik werkte||I worked|
|krabben||to scratch||ik krabde||I scratched|
⟨v⟩ and ⟨z⟩
The letters ⟨v⟩ and ⟨z⟩ are somewhat special:
- They are permitted only at the start of a syllable in native words, not at the end.
- For historical reasons, they are never preceded by a short/lax vowel, so they never occur doubled.
- When the sounds /v/ and /z/ occur at the end of a syllable, they are written ⟨f⟩ and ⟨s⟩ respectively.
In this case, therefore, final devoicing is reflected in the spelling:
- blijven /ˈblɛivə(n)/ ("to stay") → blijf /blɛif/ "(I) stay"
- huizen /ˈɦœyzə(n)/ "houses" → huis /ɦœys/ "house"
However, ⟨f⟩ and ⟨s⟩ are written also at the end of a syllable that is not final. In this case, the pronunciation remains voiced, even though the spelling shows a voiceless consonant. This is most common in the past tense forms of weak verbs:
- leven /ˈleːvə(n)/ ("to live") → leefde /ˈleːvdə/ "(I) lived"
- blozen /ˈbloːzə(n)/ ("to blush") → bloosde /ˈbloːzdə/ "(I) blushed"
Compare this to verbs in which the final consonant is underlyingly voiceless. Here, the dental assimilation rule calls for the ending -te, which gives away the voicelessness of the previous sound, even where the spelling of that sound itself does not:
- blaffen /ˈblɑfə(n)/ ("to bark") → blafte /ˈblɑftə/ "(I) barked"
- ruisen /ˈrœysə(n)/ ("to rustle, to hiss") → ruiste /ˈrœystə/ "(I) rustled"
Some modern loanwords and new coinages do not follow these rules. However, these tend to not follow spelling rules more generally. For example buzzen ("to page (call on a pager)") → buzz ("(I) page"), buzzde ("(I) paged").
Dutch uses acute accent to mark stress, and diaeresis (trema) to disambiguate diphthongs/triphthongs. Occasionally, other diacritics are used in loanwords. Accents are not necessarily placed on capital letters (e.g., the word Eén at the beginning of a sentence), unless the word is entirely written in capitals.
Acute accents may be used to emphasize a word in a phrase. It can be put on the vowel in the stressed syllable of the emphasized word. If the vowel is written as a digraph, an acute accent is put on both parts of the digraph. Although this rule includes ij, the acute accent on the j is frequently omitted (resulting in íj instead of íj́), as putting an acute accent on a j is problematic in most word processing software. If the vowel is written as a trigraph (or more), the accent is put on the first two vowel letters.
Stress on a single short vowel (not a digraph or trigraph) is occasionally marked with a grave accent: Kàn jij dat? (equivalent to the example below), wèl; even though this is technically incorrect.
Additionally, the acute accent may also be used to mark different meanings of various words, including een/één (a(n)/one), voor/vóór (for/before), vóórkomen/voorkómen (to occur / to prevent), and vérstrekkend/verstrékkend (far-reaching/issuing), as shown in the examples below.
|Dat was háár ijsje.||That was her ice cream.|
|Ik wil het nú!||I want it now!|
|Dat is héél mooi.||That is very nice.|
|Kán jij dat?||Can you (are you able to) do that?|
|Tóé nou!||Come on!|
|Die fiets is niet óúd, hij is níéuw!||That bike is not old, it is new!|
|Hij heeft een boek.||He has a book.|
|Hij heeft één boek.||He has one book.|
|Ik zal voor jou opstaan.||I will get up for you.|
|Ik zal vóór jou opstaan.||I will get up before you.|
A diaeresis is used when a combination of vowel letters may be mistaken for a digraph, or interpreted in more than one way: "egoïstisch" (egoistic), "sympathieën" (sympathies, preferences), "reëel" (realistic), "zeeën" (seas). On a line break which separates the vowels but keeps parts of a digraph together, the diaeresis becomes redundant, and isn't written: ego-/istisch, sympathie-/en, re-/eel, zee-/en.
Diacritics in loanwords
Besides being used to mark stress, as described above, acute accents are also used in many loanwords (mainly from French) such as logé (overnight guest), coupé (train compartment), oké (okay), and café.
As in English, an apostrophe is used to mark omission of a part of word or several words:
|zo'n||(zo een)||such a|
|'s avonds||(des avonds)||in the evening|
|'s winters||(des winters)||in the winter|
|'s Gravenhage||(des Gravenhage)||The Hague|
The letter immediately following a word-initial apostrophe is never capitalized: "'s Avonds is zij nooit thuis." (She is never at home in the evening.)
- Dutch braille
- History of Dutch orthography
- IJ (digraph)
- Matthijs Siegenbeek
- Nederlandse Taalunie
- English spelling
- Grand Dictation of the Dutch Language
- De grondbeginselen der Nederlandsche spelling. Ontwerp der spelling voor het aanstaande Nederlandsch Woordenboek (1863) by L.A. te Winkel
- De grondbeginselen der Nederlandsche spelling. Regeling der spelling voor het woordenboek der Nederlandsche taal (1873) by L.A. te Winkel and M. de Vries
- De woordenlijst der Nederlandse taal online (2005) by the Dutch Language Union (Taalunie)
- De witte spelling (2006) by the Society "Onze Taal"
- "New Spelling" of the Dutch language
- Phonetic/spelling alphabets for various languages
- Vincent van Heuven, Spelling en Lezen. Hoe Tragisch Zijn de Werkwoordsvormen?, Assen: Van Gorcum, 1978.
- Rob Naborn, De Spelling-Siegenbeek (1804), Doctoraalscriptie, Vrije Universiteit, Amsterdam, 1985.
- Marijke van der Wal, Geschiedenis van het Nederlands, Utrecht: Het Spectrum, 1994.
- Nicoline van der Sijs, Taal als mensenwerk. Het ontstaan van het ABN, Den Haag: Sdu Uitgevers, 2004.