Executive verdict
Dutch developed from Low Franconian varieties of West Germanic in the Low Countries and neighbouring Lower Rhine region. Its origins cross the modern borders of the Netherlands, Belgium, France and Germany. Old Dutch comprised related dialects with boundaries sometimes difficult to distinguish from neighbouring languages. 1 [1]
The useful outline is Old Frankish and Old Low Franconian, then Old Dutch, Middle Dutch and modern Dutch. These labels describe overlapping developments, not successive languages invented on particular dates. A university overview places Middle Dutch approximately between 1150 and 1500 and the transition to modern Dutch in the sixteenth century. 2 [2]
Standard Dutch developed much later than Dutch itself. Nor did the Statenbijbel of 1637 simply create it: historical linguists Hans Beelen and Nicoline van der Sijs challenge that foundation story. 3 [3] The Meertens Institute also distinguishes the familiar seventeenth-century account from research placing a broadly established standard in the eighteenth century. 4 [4]
Our assessment is that the best explanation separates linguistic descent, surviving written evidence and social acceptance. A manuscript shows what somebody wrote. A grammar shows what somebody prescribed. Neither, by itself, tells us how an entire population spoke.
Where does Dutch come from, and from which country?
Dutch comes from the Germanic speech of the Low Countries, especially its Low Franconian varieties. Assigning its origin exclusively to the modern Netherlands misses both its southern history and its eastern connections. 5 [5]
Old Frankish is the starting point, with qualifications
The regional history by Ann Marynissen and Guy Janssens describes Old Low Franconian as the early medieval dialects of the Franks in the Netherlands and northern present-day Belgium. It distinguishes western and eastern varieties and locates the Wachtendonck Psalms in the region between Venlo and Krefeld. 5 [5]
That geography matters. Krefeld is on the German side of today's border, while the southern Netherlands and Flanders share much of the language's early history. The Dutch Language Institute, abbreviated INT, includes French Flanders and part of the German Lower Rhine in its description of the Old Dutch area. 1 [1]
Old Frankish is therefore a useful ancestral label, but the Franks' political reach does not provide a precise linguistic map. Our inference is that the question “which country invented Dutch?” asks medieval evidence to fit a modern national category.
There was also language contact within that area. Anthony Buccini's peer-reviewed study proposes that contact between Frankish and North Sea Germanic speakers helped shape Dutch in the seventh and eighth centuries. This is a scholarly explanation of its emergence, not a licence to describe Dutch as a calculated mixture with fixed proportions. 6 [6]
Why “Dutch” looks like “Deutsch”
The name can mislead English speakers. Nederlands means Dutch. Older names such as Diets and Duutsch referred to the language of the people, distinguishing local speech from prestigious written languages such as Latin. Onze Taal explains that English borrowed its word “Dutch” from those older Dutch names. 7 [7]
The resemblance to German Deutsch reflects a shared naming history. It does not establish that present-day Dutch descended from present-day German. Nor does an old document's use of a word resembling Duits (German in modern Dutch) automatically identify its language by modern categories. 7 [7]
A relative of German and English
Dutch belongs to the Germanic family, within Indo-European. Comparative linguists establish relationships through recurring sound correspondences, rather than resemblance between isolated words. Dutch water (water) corresponds to German Wasser, ik (I) to ich, appel (apple) to Apfel, and maken (to make) to machen. 8 [8]
These pairs illustrate effects of the High German consonant shift. The university account dates that development to the sixth to eighth centuries and notes that the Dutch area largely remained outside it, with some Limburgish exceptions. “Low” in Low Franconian concerns geography, not inferior linguistic status. 8 [8]
How old is Dutch, and what does Old Low Franconian mean?
Dutch emerged gradually in the early medieval period. Old Low Franconian and Old Dutch overlap substantially as labels, but the exact starting date depends on the evidence and definition a historian uses. 5 [5]9 [9]
Oudnederlands (Old Dutch) is the historical label. Oudnederfrankisch (Old Low Franconian) foregrounds its dialect classification and Frankish background. Neither term denotes a uniform national language with an official spelling.
The INT's Old Dutch Dictionary covers approximately 500 to 1200. Its corpus begins earlier, in 475, because the collection includes early material such as inscriptions. By contrast, the university overview places Dutch's beginnings around 700 and its Old Dutch phase before 1150. These are different research boundaries, rather than competing claims about a language's birthday. 9 [9]1 [1]2 [2]
| Stage or milestone | Dates used by the source | What the label tells us |
|---|---|---|
| Broad Old Dutch dictionary period | Approximately 500 to 1200 9 [9] | A practical boundary for collecting early vocabulary, including material only partly identifiable as Old Dutch |
| Narrower Old Dutch historical period | Approximately 700 to 1150 2 [2] | A periodization centred on distinctive early Dutch varieties |
| Middle Dutch | Approximately 1150 to 1500 in the university overview 2 [2] | A long period of regional varieties and increasingly abundant writing |
| Transition to modern Dutch | Sixteenth century 2 [2] | Changes in the language alongside efforts to develop a common written form |
| Wider establishment of a standard | Eighteenth century in the Meertens summary 4 [4] | Acceptance across more users, rather than the first attempt to prescribe rules |
A corpus boundary is not a spoken-language boundary. The INT says that distinguishing early Dutch from Old Saxon and Old Frisian can be difficult, and that its corpus follows a broad admission policy. 1 [1]
Its source inventory includes the Bergakker inscription at approximately 475, legal glosses, place names, religious texts and fragments. That gives researchers a varied evidence base, but it does not make every item equally secure or equally informative. 10 [10]
Our inference is that “about the early Middle Ages” is a more honest answer to Dutch's age than a precise anniversary. Historical labels help us organize gradual change. They should not imply that speakers suddenly stopped speaking Frankish and began speaking Dutch.
What do the Wachtendonck Psalms tell us about Old Dutch?
The Wachtendonck Psalms preserve substantial early Dutch vocabulary and grammatical forms in a religious translation. They are especially valuable because much other early evidence consists of isolated words, names or short fragments. 11 [11]
A Dutch translation between Latin lines
The INT dates the Psalms to the tenth century. It describes an interlinear translation of psalms and canticles, adapted from a High German, probably Central Franconian, predecessor. Some words and forms from that earlier text remain. 10 [10]
“Interlinear” means that the translation sits between the lines of another text. In this case, a reader could follow the Latin with help from the vernacular equivalents. The Freie Universität Berlin explains that the translation largely follows Latin word order, limiting what researchers can infer about ordinary Old Dutch sentence structure. 11 [11]
That is a major interpretive constraint. A text can be excellent evidence for vocabulary while providing distorted evidence for spontaneous speech. Our inference is that the Psalms are best treated as an early witness with a particular purpose, rather than a transcript of conversation.
The original manuscript is lost
The name comes from Arnold Wachtendonck, a later owner of the manuscript. H. K. J. Cowan's scholarly edition traces the surviving evidence through copies and the word lists of Justus Lipsius. Cowan cites a letter from Lipsius dated 14 January 1599 describing the manuscript; the manuscript itself subsequently disappeared. 12 [12]
This creates separate questions about when the underlying translation was composed, when it was copied and how reliably later witnesses preserve it. Cowan discusses disagreements between editors over readings and corrections. Modern readers consequently encounter an edited reconstruction of transmitted evidence. 12 [12]
The following comparison distinguishes dates of early language material from the dates of copies that preserve it.
| Evidence | Date or period reported | What survives and what it can show |
|---|---|---|
| Malberg glosses associated with the Lex Salica | Sixth-century legal text preserved in eighth- and ninth-century copies 11 [11] | Vernacular legal vocabulary within a Latin textual tradition |
| Wachtendonck Psalms | Tenth-century translation; fragmentary copies from the sixteenth and seventeenth centuries 10 [10] | Connected early religious language, with translation and copying constraints |
| Leiden Willeram | Approximately 1100 10 [10] | An adaptation written at Egmond, still retaining features of its High German source |
| Hebban olla vogala | Approximately 1100 13 [13]14 [14] | A short pen trial whose linguistic classification remains debated |
| Ghent lepers' hospital statutes | Approximately 14 November 1236 13 [13] | An early official Dutch text, evidence of vernacular administrative use |
The table is not a ranking of “the first Dutch”. A legal gloss, a religious translation, a verse and an administrative document answer different historical questions.
Why Hebban olla vogala is not Dutch's birth certificate
The famous opening Hebban olla vogala means roughly “all birds have”. The verse survives as a pen trial in an English monastic setting. The Flemish Canon explicitly says it is not the oldest surviving Dutch and discusses its proximity to Old English. 13 [13]
Literatuurgeschiedenis likewise presents the traditional Flemish account while acknowledging that the text might be classified as an English dialect. Its status as an emblem of early Dutch literature is clearer than any claim that it represents the first Dutch ever written. 14 [14]
Our verdict is to keep the verse in the story, but give the Psalms and less picturesque evidence their proper weight. An attractive literary origin story should not replace the broader linguistic record.
What was Middle Dutch, and how different was it?
Middle Dutch was a group of related medieval varieties with substantial regional variation. Reading it requires more than replacing old spellings with modern ones, because sounds, grammar and meanings also changed. 15 [15]16 [16]
Sound changes make the period boundary useful
A central development was the weakening of vowels in unstressed syllables. The scholarly history of Dutch describes Middle Dutch schwa, a weak vowel, as developing from Old Dutch short vowels in unaccented syllables. The contrast helps explain why endings in older texts can look unfamiliar even when the word itself is recognizable. 15 [15]
The scholarly history edited by M. C. van den Toorn and colleagues compares Old Dutch holto (wood) with Middle Dutch hout (wood), and golt (gold) with gout (gold). These examples show changes within inherited vocabulary, rather than replacement of one language by another. 15 [15]
Its detailed comparisons also warn against expecting a single identical sound system everywhere. Our inference is that the Old to Middle Dutch boundary is useful for describing clusters of changes, while individual developments have their own regional histories.
There was no single medieval standard
The university account distinguishes five broad groups: Flemish, Brabantian, Hollandic, Limburgish and eastern Middle Dutch. It connects the eastern group to an Old Saxon background, complicating any account in which every regional variety simply follows the same Frankish path. 16 [16]
Writers and copyists could introduce features from their own dialects or adapt a text for a wider readership. Local documents whose origin and date are known are therefore particularly useful for historical dialect research. Literary manuscripts can combine features from different places and periods. 16 [16]
The INT's Early Middle Dutch Dictionary covers 1200 to 1300 and draws on a corpus containing more than 1.6 million words. It includes administrative documents as well as literature, personal names and place names. That coverage provides much more than a catalogue of famous poems. 17 [17]
The corpus size should not be compared with an ordinary speaker's vocabulary. It measures surviving collected text, and its dictionary headwords also include names. Our inference is that the expanding written record improves the questions researchers can ask; it does not prove that spoken Dutch suddenly became richer.
Towns changed what Dutch was used for
The university history connects the growth of towns with vernacular laws, contracts and other administrative writing. Latin retained major roles in church and scholarship, while local language gained space in civic life. 18 [18]
This is a change of function as well as quantity. Written Dutch became useful for regulating property and obligations among people who did not necessarily know Latin. The Ghent statutes in the table illustrate that administrative dimension. 13 [13]
The grammar was different too
Hüning and Vogl describe Middle Dutch as having four grammatical cases: nominative, genitive, dative and accusative. Endings on nouns and related words helped mark their functions. As those endings weakened, prepositions and more fixed word order increasingly carried the information. 19 [19]
Their example contrasts des vaders huis (the father's house) with het huis van de vader (the house of the father). The forms overlapped during the change. Recognizing that history helps explain why older Dutch can look grammatically closer to German without being German. 19 [19]
Negation also changed. A. M. Duinhoven describes a development from ne (an old negative particle), through ne ... niet (a two-part negation), to niet (not) alone. He finds the single niet already alongside other patterns from the thirteenth century. 20 [20]
Pronunciation also differs from what modern spelling might suggest. Hüning and Vogl explain that medieval ij represented a long single vowel rather than today's diphthong. They caution that reconstructing the exact pronunciation from spelling and rhyme retains uncertainty. 19 [19]
Those overlapping forms are a useful corrective to tidy timelines. A label such as Middle Dutch summarizes an era; it does not mean every speaker used the same construction throughout it.
How did Dutch become a standard language?
Standardization grew from practical communication needs and the prestige of particular forms of Dutch. It involved selecting forms, recording rules, extending the language's uses and persuading people to accept a norm. 21 [21]22 [22]
The distinction between changes in the language and changes in its social position is useful here. The university's account of language change explains how political and cultural circumstances influence vocabulary and grammar, including French influence in the southern Netherlands. Borrowing does not itself change the language's Germanic ancestry. 21 [21]
Why the centre shifted towards Holland
The university history describes early spelling and dictionary work in the southern Netherlands, followed by a northward shift. It identifies the fall of Antwerp in 1585 and migration to the north as important in that change, while cautioning against assuming that southern migrants simply remade northern Dutch in their own image. 22 [22]
Our assessment is that prestige offers a better explanation than linguistic superiority. Forms associated with influential writers, officials and commercial centres had more opportunities to circulate. Their success does not prove that other dialects were defective.
A common written language also had to work across regional audiences. The university account links increasing trade, political contact and printing with the desire for forms that travelled beyond a writer's locality. 18 [18]
Rules appearing is not the same as rules spreading
The first Dutch grammar identified in the university account is the Twe-spraack vande Nederduitsche letterkunst (Dialogue on Dutch grammar), published in 1584. Its significance lies in treating the vernacular itself as something that could be described and regulated. 22 [22]
But a prescription does not demonstrate general obedience. In research reported by Leiden University in April 2025, Eline Lismont compared grammars and schoolbooks with administrative texts, printed works, diaries and travel accounts. Edited print responded differently from writing for personal use. 23 [23]
Leiden reports that rules reached broader sections of the population in the eighteenth century. It also describes different outcomes for different rules: replacing ae with aa spread more readily than some grammatical prescriptions. 23 [23]
This suggests that standardization should be measured through use as well as publication. A history built only from celebrated grammars risks mistaking the ambitions of language authorities for the habits of their contemporaries.
For the same reason, written standardization, spoken standardization and official spelling need separate timelines. A person could write conventional Dutch while speaking a regional dialect. A government could authorize a spelling without immediately creating a shared accent.
Did the Statenbijbel create modern Dutch in 1637?
The Statenbijbel was an important religious and cultural publication, but the claim that it created standard Dutch is too strong. Its publication date is secure; its supposed role as the language's foundation is disputed. 3 [3]24 [24]
The Statenbijbel (States Bible) appeared in 1637. Statenvertaling (States Translation) names the translation associated with it. The Dutch Canon emphasizes its lasting familiarity among believers and credits it with expressions in everyday Dutch. 24 [24]
Beelen and Van der Sijs reach a more restrictive conclusion about linguistic causation. They report no influence on Dutch spelling and very little on grammar, while warning that many allegedly introduced biblical expressions predate the translation. 3 [3]
These accounts should not be silently blended. One foregrounds cultural reach and familiar language; the other tests the stronger claim that the translation supplied innovations adopted by standard Dutch.
The analytical test is straightforward: to demonstrate innovation, an expression must be absent from earlier evidence, present in the translation and subsequently adopted through it. Mere familiarity with a Bible expression cannot establish that chain.
Our inference is that the translation's visibility helps explain why it became a convenient landmark. A printed work has a date, authors and institutional sponsors. Distributed changes in usage are harder to narrate and less easily attached to a single monument.
The Meertens Institute's current overview accordingly places the traditional seventeenth-century narrative beside research arguing for an established standard in the eighteenth century. Leiden's evidence about wider uptake helps explain why those dates need not describe the same stage. 4 [4]23 [23]
The careful formulation is therefore that standard Dutch developed over time, with religious publication among its contexts. We could not verify a defensible percentage of modern Dutch attributable specifically to the Statenvertaling.
When did Dutch get official spelling and shared institutions?
Dutch acquired an official spelling long after it had become a written language. Government rules and shared institutions formalized conventions that had developed through a much longer history. 25 [25]26 [26]
The Nationaal Archief identifies Matthijs Siegenbeek's spelling, introduced in 1804 at government request, as the first official arrangement. Onze Taal independently gives the same milestone. 25 [25]26 [26]
The archive also cautions against calling everything before that date spelling anarchy. Medieval writers already shared substantial conventions despite variation. “No official spelling” and “no recognizable writing practices” are different propositions. 25 [25]
| Institutional milestone | Date | What changed |
|---|---|---|
| Government-authorized Siegenbeek spelling | 1804 25 [25]26 [26] | An official uniform spelling was introduced |
| De Vries and Te Winkel spelling project | 1863 26 [26] | Work on a major dictionary prompted a new spelling framework |
| First Groene Boekje (Green Booklet) | 1954 26 [26] | Joint Dutch and Flemish work produced a shared word list |
| Treaty establishing the Taalunie | 9 September 1980 27 [27] | Belgium and the Netherlands established cooperation around their shared language |
Spelling rules standardize written forms. They cannot by themselves make everyone pronounce a vowel alike, remove local vocabulary or guarantee agreement over grammar. Our inference is that official spelling is evidence of institutional standardization, not proof that everyday Dutch became uniform.
The Taalunie anniversary account describes a further development: recognition of Dutch as a language with several centres. It acknowledges differences within standard Dutch in the Netherlands, Belgium, Suriname and the Caribbean. 27 [27]
This gives the historical story a different endpoint from “everyone eventually speaks the same Dutch”. Institutions maintain shared resources while allowing legitimate geographical variation.
Is Flemish the same language as Dutch, and where does Afrikaans fit?
Belgian standard Dutch and Netherlands standard Dutch are varieties of the same language. Afrikaans has a different historical relationship: it developed from Dutch and became a separate language. Frisian developed alongside Dutch as another West Germanic language. 28 [28]2 [2]
“Flemish” is a broad everyday label that can obscure distinctions between standard Belgian Dutch, regional dialects and informal speech. For learners, identifying the variety and context is more informative than relying on the label alone.
Taaladvies.net defines standard Dutch through its usability in public life, including administration, education and media. It explicitly allows geographical and stylistic variation, rather than treating the standard as one permanently fixed form. 28 [28]
Belgian Dutch has its own accepted forms
The Flemish government's language advice gives confituur (jam) in Belgium and jam (jam) in the Netherlands as standard counterparts. It also distinguishes accepted Belgian vocabulary from forms used in informal speech that do not qualify as standard. 29 [29]
That advice links present attitudes to a history of language cultivation. During the twentieth century, Flemish campaigns often promoted northern forms and rejected local alternatives. The official account describes the subsequent move towards acknowledging distinct Belgian and Netherlands centres. 29 [29]
Our assessment is that this history explains why some learners receive contradictory corrections. A tutor may be applying a national standard, a formal-register preference or an older ideal of conformity. Those are different grounds for advice.
Afrikaans is a descendant, rather than a preserved medieval Dutch
Afrikaans developed from Dutch in the Cape's colonial contact setting. Research by Ronel van Oort and Wannie Carstens emphasizes interactions between European settlers, indigenous Khoi communities and enslaved people. The language's development belongs to that unequal multilingual society, not exclusively to settlers' speech. 30 [30]
Afrikaans can therefore illuminate Dutch history, but it should not be treated as an unchanged recording of an earlier stage. Nor should similarity imply automatic mutual comprehension. Our inference is that descent supplies a historical connection; understanding a present-day speaker remains a practical skill.
What this means for learners
Use history to make modern Dutch more intelligible, while keeping your practice anchored in the variety you need today.
Ask for a geographical and situational label when corrected. “Standard in Belgium”, “formal written Dutch” and “local dialect” give you usable information. “That is not Dutch” often leaves the real issue unexplained. The official advice distinguishes variation within the standard from other registers. 28 [28]29 [29]
Treat familiar-looking medieval words cautiously. The university account notes that varen (to travel historically, now usually to travel by water) once had a broader meaning. Recognizing a word's spelling is not enough to recover its older sense. 21 [21]
Read an annotated passage before a whole medieval work. Ask which manuscript the edition follows, whether spelling has been modernized and whether a translation accompanies it. Different editorial choices can change the apparent distance between historical and current Dutch.
Use historical dictionaries for historical questions. The Old Dutch Dictionary supplies modern Dutch translations of citations, while the Early Middle Dutch Dictionary includes information on attestations, variants and source texts. They help you investigate evidence rather than guess from resemblance. 9 [9]17 [17]
Separate ancestry from imitation. The Germanic relationship can help an English or German speaker notice patterns, but related words and structures still need checking in current Dutch. Learning a relative does not remove the need to learn the language you intend to use.
If you want help selecting a current learning routine, see our guide to the best way to learn Dutch. For guided work on pronunciation, regional usage or reading, compare online Dutch tutors and ask about those specific interests.
What we could not verify
We could not identify a single year when Dutch began. The dictionaries and historical surveys use different period boundaries, and surviving evidence does not mark the beginning of speech.
We could not verify a universally accepted classification of Hebban olla vogala. The scholarly and educational sources explicitly preserve debate about Dutch and English features. 13 [13]14 [14]
We could not inspect the lost original Wachtendonck manuscript. Claims about it depend on surviving copies, word lists and critical editions. 12 [12]
We found no defensible numerical division of Dutch into Frankish, Saxon, Frisian or other ancestral contributions. Contact histories do not justify percentages.
We also could not quantify how much of present-day Dutch derives specifically from the Statenvertaling. Cultural prominence, familiarity and demonstrated linguistic innovation require different evidence.