Het Nederlandse lied in de Gouden Eeuw

[p. 321]

Summary

This study sets forth the most important principles of the writing of contrafacta, i.e. new texts to extant melodies, in late 16th and 17th-century Dutch songs. The knowledge of these principles will help in reconstructing the melodies of these songs, in finding their sources and in determining their reception.

The choice of this subject is easily understood if one realizes that this enormous repertory consists almost entirely of contrafacta. This in itself is not amazing: in other countries there have also been songs to pre-existing melodies, like the English ballad songs and the German songs on Fliegende Blätter. The extraordinary qualities of the Dutch song culture are the high artistic level of many of the songs and the beauty of the song books in which they were printed, often containing staff notation and fine engravings. These seem to contradict the contrafactum-technique, which is associated with the ‘unofficial’ musical practice of common people.

In the Introduction, I have articulated the character of this song culture using dichotomies such as contrafactum-composition, sacred-secular, reading-singing, popular and elite culture, oral and written transmission, monophony-polyphony. Firstly, I have attempted to formulate an explanation for this contrafactum repertory, the texts of which often surpass in quality those of contemporary art songs like the French airs de cour and English ayres. The almost complete absence of original Dutch melodies is striking in this Golden Age of Dutch culture, which immediately follows the period of the great Netherlandish polyphonic school. An explanation may be found in religious and political circumstances. Calvinism, the official religion of the young Republic, did not allow professional music in the services. Moreover, the Republic did not have a strong central court. Thus no professional choruses or orchestras existed - professional musicians were restricted to town musicians, organists and music teachers. As a whole, the musical climate was conservative: the first music which can be called Baroque is Constantijn Huygens's Pathodia published as late as 1647 - in Paris. On the other hand, there was a strong literary movement. Great poets like Pieter Cornelisz Hooft (1581-1647), Gerbrand Adriaensz Bredero (1585-1618) and Joost van den Vondel (1587-1679) loved songs and almost naturally wrote contrafacta - otherwise, their texts would not have been sung. They were just three of the dozens of poets who did so, including Dirck Rafaelsz Camphuysen, Jacob Cats, Dirck Volckertsz Coornhert, Karel van Mander, Jacob Revius, Joannes Stalpart van der Wiele, Jan Janszoon Starter and many others. Also Adriaen Valerius, whose Nederlandtsche gedenckclanck (1626) did not become famous until in the late 19th century, worked in this tradition. The combination of a weak

[p. 322]

musical and a strong literary culture as well as the foreign example of airs and ayres may explain the popularity of the contrafactum technique, which is a continuation of a flourishing 16th-century song tradition in the Southern Netherlands, comprising e.g. the ‘Antwerps liedboek’ (1544) and the Souterliedekens (1540).

One cannot call the songs studied in this book ‘folk songs’, as is sometimes done, unless one uses this term in a very broad sense. However, neither do they belong exclusively to the upper classes. Some of the songs written in the upper literary circles became widely popular and poets like Hooft and Bredero sometimes wrote contrafacta on simple folk tunes, or on the psalm melodies everyone sang in Church. Moreover, the majority of the songs did not have literary aspirations at all and this became more true as the 17th century went on. In short, one cannot make a sharp distinction between elite and folk songs in this repertory. The texts were transmitted mostly written, the melodies mostly orally. Only a minority (about one quarter) of the song books contained musical notation. The usual practice was to write a tune indication (‘wijsaanduiding’) above the text, e.g. ‘Op de wijze van Wilhelmus van Nassau’ (‘To the tune of Wilhelmus van Nassau’).

Most of the tunes in this song repertory came from abroad, especially from France, England and Italy. Dance melodies, airs de cour, madrigals and balletti were popular. German influence, so important in the Middle Ages, had almost completely vanished. On the contrary, some Dutch song texts were translated into German.

The secular tunes were also used in the numerous religious song books, whether Protestant or Catholic; as a matter of fact, the authors of sacred songs stated that their contrafacta were meant to dispel the secular texts from the memory of the singers.

One might ask whether all song texts were meant to be sung. Some songs are designated ‘to be sung or read’, so that we may assume that there existed an alternative reading practice. D.R. Camphuysen stated as a principle that all texts from his Stichtelycke rymen (1624) could be sung as well as read. Consequently, he only wrote jambic and trocheic verses. In his opinion, the characteristic irregular song rhythms were not ‘readable’.

In some cases the editor of a song book mentioned the reading option for the reason that he did not have a tune indication. In other cases, the tune indication is simply left out. This brings us to another problem: how to recognize song texts as such if there is no tune indication? In such cases one can look after song characteristics like the strophic form, irregular metres and rhythms, refrains and text repeats. Also the title may indicate a sung text: ‘lied’, ‘deun’ and ‘zang’ (all synonymous of ‘song’) and expressions referring to dance movements like ‘dans’, ‘courante’ etc. Also ‘psalm’, ‘koor’ and its equivalent ‘rei’ (chorus) and similar expressions may indicate a sung performance. Around 1600, the main term, ‘lied’, still expressed without exception a sung text, i.e. a contrafactum. Later, Vondel used the term for strophic texts which were not written to pre-existing melodies. Not coincidentally, he was one of the first Dutch poets who tried to persuade composers to set his texts to music. This resulted in a number of polyphonic settings of Dutch texts in the 1640's, most of them by the Haarlem town musician Cornelis Padbrué.

The contrafact technique normally implies the absence of word painting: this belongs to the metier of the composer, not of the poet. However, some poems by Cats and Krul who follow certain qualities of their melodies show that this principle may be applied to contra-

[p. 323]

facta as well, though it was done very rarely. At a more general level, correspondence of text and melody is rather common. In spite of the religious contrafacta on secular tunes which seem to suggest that everything was possible in this respect, certain melodies are often used for special categories of song texts. E.g. melodies such as ‘Engelsche fortuin’ and ‘Lachrymae’ for texts of mourning, ‘Potshonderdduizend slapperment’ (from the English jig ‘Singing Simpkin’) for humorous songs, ‘Si c'est pour mon pucelage’ for satirical poems, etc. Conversely, one sees very few love songs to the tune of a political song such as ‘Wilhelmus van Nassouwe’ or to a psalm melody.

The main subject of this study is the relation of text and music in contrafacta. This relation is most visible in the strophic forms, to which Part Three has been devoted. This part constitues the most significant section of this work, since it presents a new system for research in this field. Here the possibilities of the strophic repertorium, ‘voetenbank’, is investigated. In Part One the textual relation between contrafacta is studied, in which ‘borrowing’ is the keyword. Part Two is mainly about the device of the ‘tune indications’.

In Part One a threefold basic pattern of borrowing is presented: a contrafactum can borrow only the melody of its model and nothing more (1), so-called ‘irregular borrowing’; or it may borrow the melody plus the strophic form (2), which is the most usual option; or it may borrow the melody as well as strophic and literary aspects of the model (3). Apart from these basic patterns there are special possibilities, e.g. literary borrowing without borrowing the melody and the strophic form, or other cases of borrowing from different models. The literary borrowing is the main subject of this Part.

Of course, literary borrowing is common in medieval religious contrafacta. Sometimes they take over the first lines of their secular models and continue independently, sometimes they take over as much as possible from their models and change only a few words to turn the love of a girl into the love of Maria or God. These medieval devices, which I call ‘initial’ and ‘continuous borrowing’, can still be found in the 17th century. Now, Renaissance poets applied the principals of imitatio and aemulatio to the older devices. Transformation becomes important: the imitation is better if the original subject is changed or if it is combined with other ideas. In general, there seems to be a shift from verbal borrowing towards the borrowing of only themes and ideas.

A special borrowing device is the ‘pendant’, if there is a symmetrical relation between model and contrafactum. For instance, in the original song a woman is maltreated by her husband, but in the contrafactum the wife wears the breeches. Or, the first song describes the joys of love and the second the sorrows when the lady rejects the lover. These pendants are often written strophe for strophe after their models. The pendant principle is often applied in political songs, e.g. from the Beggars against those siding with the Spaniards or the other way around, or from Calvinists against Catholics, Remonstrants, Mennonites etc. In love songs the pendant principle is used for responses, often strophe for strophe. Sometimes the strophes of the original song and the response can be sung alternately, as in a dialogue.

[p. 324]

In normal dialogues the borrowing principle has two particular objects, the dialogue as such and the pattern of the speeches. I have developed a graphic presentation in order to compare these patterns in different songs to the same melody. Some melodies seem to attract dialogues, or rather: some dialogue songs have often been imitated. In two of them, Hooft's ‘Ach Amaryllis!’ to the tune of the folksong ‘Boerinneken’, and a dialogue by B. Ruissenberch to the English echo-tune of ‘Malle Symen’ (Malsims), the dialogue pattern plays an important role. In another, the jig melody ‘Potshonderdduizend slapperment’, it does not - here it is rather the association with the sung comedies which influences the poets to choose this melody for comic dialogues.

It is often possible to draw trees of contrafacta borrowing from each other. In the case of ‘Ach Amaryllis!’ and ‘Malle Symen’ at least three ‘generations’ can be discerned. Also Hooft's popular songs ‘Vluchtige nimph’ and ‘Windeken daer het bosch af drilt’ - itself an imitation of ‘Vluchtige nimph’ - have an interesting tree of followers. The number of imitations form a reliable indication of the popularity of a song, more than the number of cases in which the song is mentioned in tune indications. The latter sometimes appear to be dissimulative because they do not mention the immediate model but the model of the model - which is mentioned in the immediate model!

Part Two is dedicated to the tune indication, the usual substitute for musical notation in song books. Firstly, it is regarded as one of the main information categories or ‘fields’ of a song. The tune indication is the musical part of the inscription of a song. There is also a textual part, the title, which may indicate the type of song, the subject, the occasion or the singing personage. Another important field is the text incipit, of course, the most individual though not always unique field. Song books usually have incipit registers, like other poetry books. Only a few song books, in which the music plays an important role, have a register of tunes.

Though in the Middle Ages tune indications were known, e.g. in French troubadour songs, the first tune indications in Dutch songs date from as late as about 1500, e.g. in MS Berlin 190. These early tune indications belong to religious contrafacta. In secular songs, tune indications are used from about the middle of the 16th century.

Some attention is paid to the syntax of tune indications. The most common formulas in the 17th century are ‘Op de wijze van’, ‘Op de stemme van’, ‘Op de voys (van)’ and ‘Op de toon’, which all mean the same: ‘To the melody of’. Of these, ‘stemme’ may be a neologism, coined by D.V. Coornhert. Some tune indications do not mention the song they refer to by its name. Among these, I distinguished implicit (when the first line tells us which melody is meant), reflexive (‘Alst begint’ - ‘Like it begins’) and open (tune name left out) tune indications as well as indications which refer to the music notation which accompanies the song, and tune indications which refer to the song before. Sometimes two or more melodies are mentioned. These double, triple etc. tune indications pose a problem for the musicologist since it is not always clear whether the tune names refer to the same melody or to different melodies.

Special attention is paid to additional singing instructions following the tune indication proper. These additional instructions are rather rare, but they suggest certain problems of the contrafactum practice. Some of the instructions tell the singer to repeat particular

[p. 325]

phrases, or to leave out such a repetition. This may indicate that the editor did not know the original melody and chose one which fitted only imperfectly. In other cases one is told to leave out the second half of the melody or to start halfway. This is seen only with Genevan psalm melodies. In the earliest French psalms rhymed by Marot and Bèze, followed in the Netherlands by Petrus Datheen, many of the psalm texts end with an incomplete strophe: the congregation simply stopped halfway through the melody, which was designed for this purpose. This practice is echoed in some of the contrafacta to these tunes. Another unexpected device is the asynchronous use of strophes. E.g. the singer may decide to take four or eight lines for a strophe, depending on the melody he chooses.

Part Two concludes with a short discussion of the classical heuristics of Dutch contrafacta. The most important entries are the text incipit and the tune indication, which should be checked in all possible combinations. The most important classical card index of Dutch songs is found in the P.J. Meertens-Instituut in Amsterdam.

Part Three is concerned with a hitherto scarcely used search field, the strophic form. I developed a description system using the number of accents per verse, the rhyme scheme and the gender of rhymes as basic parameters. A song from P.C. Hooft's play Baeto (1617) serves as example:

O eegae waerdt, wat lust u 't hart
Soo seere tot den rouw te terghen?
En wilt met duldeloose smart
'T gemoedt soo veel niet langher verghen.
Uw' borst van alle der quellingen stoet
Opruimen doet.

.	.	.	.	.	.
4A	4b	4A	4b	4C	2C

The dots above the ciphers indicate that the first syllable of each verse does not have an accent.

I have compiled a database with similar strophic schemes for about 5,700 Dutch songs from the first half of the 17th century. It is called ‘voetenbank’. In this database, now in the P.J. Meertens-Instituut, one can easily look up the forms. The main application is for songs which have no tune indication, and for songs the tune indication of which cannot be solved. E.g. Hooft did not supply a tune indication for the song just quoted, like he never indicated songs in his plays. The voetenbank contains two songs with identical strophic forms, both by Hooft, with the tune indication ‘Cessez mortels the souspirez’. Hooft's song from Baeto appears to fit to this air de cour by Pierre Guédron. Thus, we can rest assured we have found its original melody.

This method adds an important dimension to contrafactum research, but it has two particular problems. The first is that one strophic form may serve several melodies, the second that one melody may be used for several strophic forms. The first problem implies that we always have to look for additional evidence to distinguish the original melody from doubles. Similarity of strophic forms itself is only significant if the forms are very charac-

[p. 326]

teristic. The second problem is due to what I have called ‘strophic variation’. This takes two forms: transformations with and without metrical consequences. In the first case the melody must be changed, in the second it remains unaffected. Strophic variation may concern all strophic parameters, e.g. the number of accents per verse, the number of syllables per verse, rhyme scheme including rhyme gender, number of lines per strophe etc. When one knows the possibilities of strophic variation one can easily recognize variations and also deliberately look for them.

Special attention is paid to the strophic parameter which is the most difficult to establish, the number of accents. I have formulated a system of metrics for Dutch song in the 16th and 17th centuries, i.e. in a transitory period from medieval tonic verse to the syllabo-tonic verse of the Renaissance. In tonic verse the number of accents per verse is constant but the number of syllables may vary, in syllabo-tonic verse also the number of syllables per verse is constant. This period includes also the purely syllabic verse which Datheen and others produced in the 1560's in imitation of French prosody. Here the accents of the text do not always correspond with the accents of the melody. Though at the beginning of the 17th century the syllabo-tonic system became generally accepted in Dutch poetry, poets like Bredero and Hooft still used the tonic principle for songs, particularly when the melody was in triple time. Such rhythms can be regarded as characteristic for songs. My metrics, in which stressed syllables are separated by either one or two weak syllables, are designed to produce the same strophic scheme for one melody, whatever verse system is used.

The last theoretical problem studied in this book is the correspondence of text and music. It is exemplified by means of the rhyme scheme, which corresponds to the structure of the melody. Two ‘rhyme rules’ have been formulated, which establish this relationship for musical repetitions. Such repetitions urge the poet to use the same rhymes. Nevertheless there is an interesting exception to this first general rule. Inversedly, if a rhyme has been used in a repeated section of the melody, it cannot be used outside that section. This rule also has very clear exceptions.

Part Three is concluded by a test of the strophic heuristics described above. In 1979 Fred Matter published his authoritative edition of the melodies belonging to Bredero's Groot liedboeck (1622). He found 62 melodies for the 177 contrafacta in this songbook, some of which are sung to the same tune. 22 melodies could not be found. Using the strophic method I could find 9 of these 22 ‘lost’ melodies. In five other cases I found interesting clues such as models and imitations, which however did not lead to a melody. Thus in more than half of the cases the strophic method supplied information which could not be found with the classical method.

As this test shows, the strophic method is not only useful for tracing melodies but can also help in reconstructing literary relations. Thus the method is not only useful for musicologists, but also for literary historians who are interested in the genesis and reception of songs.