The writings of Michael Praetorius, Marin Mersenne and James Talbot have become increasing known over the years in both their original forms, or when isolated sections have been used by modern writers to amplify points dealing with particular subjects. This wide knowledge and acceptance of their work makes it important to consider their true reliability and usefulness for the modern researcher.
These three writers stand, justifiably, as the leading authors on musical instruments in the seventeenth century, and, remarkably, represent three different countries, and also differing backgrounds.
My interest in these writers is concentrated on James Talbot, whose manuscript I am transcribing and editing into book form. I first became aware of the problems of which I shall speak when carrying out an extension of this work by trying to make drawings of some of the instruments he measures. The problems of Praetorius and Mersenne have been mentioned by other authors, and I have frequently found myself having to reject or reconsider aspects of their writings due to uncertainties when discussing aspects of organology with other people.
The earliest of the authors is Michael Praetorius. Born in Thuringia, in about 1571, he worked as a practicing musician, both Kapellmeister and organist, and was a prolific composer and writer about music. His book on musical instruments, De Organographia, was the second book of his Syntagma Musicum, or "Musical Encyclopedia", and was written in the German language, and published in 1618 and 1619, with his plates of musical instruments, the Sciagraphia, published two years after the first edition of the Organographia, in 1620. Praetorius died in 1621.
Marin Mersenne was a French mathematician whose book, or, more correctly, collection of seven books, Harmonie Universelle, was published in 1636. In scope it is similar to the work of Praetorius, but rather than a separate volume of scaled plates, it has a series of illustrations placed within the text. These illustrations vary greatly in quality, from rough woodcuts to well engraved representations.
The final author I shall consider is James Talbot. Like Praetorius and Mersenne, he was a man of many talents, and in some ways is the most suitable to write a balanced work on instruments. Praetorius and Mersenne both have particular biases in their writings, whereas Talbot - not a musician himself - at least of any note - considers and incorporates the information of others - leading musicians and makers of his day, as well as earlier writers, including, not surprisingly, both Praetorius and Mersenne. Clearly a man with an organised mind, he (usually) gives full references when quoting from these authors, as well as acknowledging any contemporary makers or musicians who have provided him with information. It should be emphasised that these writers are not the earliest to write treatises on musical instruments - the works of Argicola, Virdung, and even earlier works by Arnaut deal with musical instruments either in their entirety, or as a fair proportion. Those works are not, however, used by modern researchers in the same manner, and I have therefore chosen not to consider them here.
* * * * *
The books of both Praetorius and Mersenne have been published in facsimile and in English translation. Praetorius wrote De Organographia in his native German, rather than the Latin used in the first volume of the Syntagma Musicum. This was clearly done to allow the book a wider readership. Praetorius describes his work, to quote from David Crookes' English translation as, quote, "an elementary guide, to furnish an introduction, or, as it were, a foundation stone, on which other more excellent musicians might go on building the entire structure". Unquote. His audience is explicitly described as being "makers and players of organs and instruments...". Elsewhere in his introduction Praetorius provides a timely reminder that he, quote, "may have gone astray or written something absurd...". Unquote.
The purpose of his writing has therefore been clearly explained, and may crudely be defined in twentienth century terms as a general coffee table book, albeit one written by a knowledgeable practical musician and theorist; aimed at an informed readership. His indefatigable work in various fields explains the delay with the publication of the plates, which did not appear until shortly before his death, some two years after the first edition of De Organographia.
Mersenne, on the other hand, was a mathematician, although music is featured heavily in his writings. This mathematical background is prominent throughout Harmonie Universelle, occasionally to the detriment of the book when viewed merely as a study of musical instruments. It should be pointed out, however, that Mersenne did not intend it to be a simple instrument study. To quote from him, using the English translation by Roger Chapman, the books of instruments are, quote, "the fruit of the other parts of harmony, since they put into practice and reveal in a thousand ways what is most agreeable in it". Unquote. This indicates what Mersenne thought of as the most important element of his study - harmony - be it of music, tuning or whatever. This theoretical viewpoint is essentially the opposite of the practically minded intentions of Praetorius.
Unlike Praetorius and Mersenne, Talbot's work exists only in the form of notes, clearly a work in progress complete with numerous crossings-out and additions. His manuscript - 241 pages of which have writing on them - survives in the library of Christchurch, Oxford, catalogued as Music Manuscript number 1187.
Talbot left no clear indications of his purpose or eventual intentions, although I believe they can be clearly inferred.
As this work is incomplete, and unknown to most researchers in its real form, a brief description is perhaps in order. It currently survives in eleven sections, which have been numbered after the manuscript's arrival at Christchurch. The sections range from four sides, in the case of the section on the harpsichord, to over thirty, in the string and wind instrument parts. It is clear Talbot did not intend there to be eleven separate sections. Three of them - catalogued by the Library as numbers 5, 8 and 9 - form a single group which more-or-less alphabetically discuss each of the instruments in turn; and sections 2 and 3 both refer to the organ, including some duplication of information. Section 1 concerns the harpsichord, sections 4 and 7 give measurements taken from string and wind instruments respectively, section 10 gives the range of various instruments in tablature form, section 11 details the aulos, and section 6 lists the instruments discussed with the names of those who have provided information on each type.
As a work-in-progress, and by examining the method used by Talbot to compile it, there can be little doubt that it was intended as a book, probably in a format very similar to that of Praetorius. The lists of measurements for various instruments are probably intended as the basis for drawing illustrations, presumably to resemble those of Praetorius. Talbot probably commenced it in or around 1689, when he was appointed Regius Professor of Hebrew at Cambridge, and ended, unfinished, following his departure from Cambridge to become the resident Rector at the Parish of Spofforth, Yorkshire, in 1700.
This work is written much later than the books of Praetorius and Mersenne, and, as such, new instruments are listed. Taken together though, the three books give a wide-ranging look at the variety of instruments found across the years and countries of seventeenth century Europe.
* * * * *
Unlike paintings, which are discussed elsewhere in this symposium and can generally be judged on their merits, written treatises are usually ambiguous in certain details, even when accompanied by drawings. The problems need not be exclusively confined to the historical era - two cataloguers describing a modern instrument in a museum may come up with noticeably different measurements, often as a result of different measuring approaches and methods. Distortions found in different published versions of photographs can also lead to errors when interpreted. There can be few makers who have not been puzzled by measurements in catalogues or books and often find themselves seeking clarification over any discrepancies.
In considering the work of writers such as Praetorius, Mersenne or Talbot, one must firstly put them into a reasonable perspective. There should be no doubt that the correct approach is to consider these authors to be correct unless there are conflicting reasons to discount what they have written. This should be true of each statement, unless they can be shown to have reasoned wrongly in a previous relevant passage. Therefore, should an author make a statement about violins that can be shown to be wrong, a comment about viols should still be considered correct unless shown otherwise, but further comments on violins might also be wrong if they relate to the previous incorrect statement.
On the other hand, it would appear to be unwise to consider historical documents to carry greater weight than other evidence. Should an author's writings suggest something that appears to be contradicted by a surviving instrument, logically either both scenarios could be correct, or the writing could be wrong. Arguably, our interpretation of the instrument could be incorrect, although anyone arguing this should have compelling reasons - aside from a historical document - to support the accusation.
Generally, the problems of interpretation found in the writings of Praetorius, Mersenne and Talbot are ones of ambiguities, rather than simple errors. I shall like to give an analysis of some of these.
At one extreme there are instruments which simply cannot work according to their descriptions and illustrations. Often the greater amount of information amplifies the problem. An example of this is the small English Cittern, or Klein Englisch Zitterlein, to use Praetorius' terminology. Both the description and illustration show problems. In his chapter on the Cittern, Praetorius describes the instrument as tuned an octave above the four-course lute of former times, and defines the string pitches as, from top to bottom, g2, d2, a1 or bb1, and f2. These tunings are repeated in his section on compasses. However, in his section on the lute, it is described as being tuned an octave and a fourth lower, ie. d1, a, f, and c, from top to bottom.
Logically, one must assume that as Praetorius wrote the tuning from g twice this is correct, however, the illustration, found on plate 16 shows such a tuning could never work. The instrument has a measured string length of 340mm. An iron string, needing to reach a pitch of g2 at a pitch of A466 can be no longer than about 212mm, about two-thirds of the illustrated length. It has been proposed that a super-strong steel wire was used. This is plausible, although unsupported by evidence from surviving instruments or wire samples. This treble wire would still not allow the instrument to work as described by Praetorius, as he specifies brass wire was also used. Brass wire 340mm long can safely reach a playing pitch no higher than g# at modern pitch, or g at 466, rather than the desired a and bb that Praetorius quotes. Simply put, there are too many errors to consider this instrument as reliable. It could be that the instrument has been drawn at the wrong scale, or that Praetorius gave the wrong tuning, and a good case can be made for either scenario.
A related problem is found on plate 17, which shows three metal-strung plucked instruments, and the Italian Lyra Gamba. All three metal-strung instruments have treble string lengths too long for iron strings at Praetorius' pitch levels. The possibility that super-strong wire was used exists, although is considerably weakened by the evidence from the English Cittern. More likely is a mistake, probably by the engraver who could either have got the scale wrong or made an error in proportion when transferring the illustrations. It is not a unique example of this occurring in the Sciagraphia. The different relationship between the string length and pitch of the three plucked instruments must also bring into question the scale of the Lyra Gamba in the same illustration.
Other plates show similar problems. Plate 20, with the illustrations of the viols shows the instruments with notably different scales - the tenor viol has a proportional string length some 15 percent - the equivalent of two and a half semitones - longer than the bass viol. If the illustration is correct it is clear that no ordered proportional set has been shown.
Plate 21 gives perhaps the best indication of some of the problems involved in analysing Praetorius' work. The plate illustrates the major instruments of the violin family, showing two pochettes, described as being an octave higher; a descant violin pitched a fourth high, a standard violin, a tenor violin, and a bass violin. If we assume that the violin is tuned near to string breaking point, albeit with a suitable safety margin, the other instruments should be in an acceptable proportion. This table analyses the various illustrations on the plate :-


String length

Highest note

Pitch, relative to violin (in cents)

Str. L. proportion (in cents)



a2 or b2 , e3

+500 or +700 ,+ 1200

+ 516



a2 or b2 , e3

+500 or +700 ,+ 1200

+ 173




+ 500

+ 484









- 700

- 316




- 1900

- 1540

            Errors can be seen firstly with Praetorius' text compared to the plate. In his list of tunings he lists the three-string instrument, possibly a rebec, as having a highest string tuned to a or b, whereas the plate describes the instrument, assuming it is the same one, as being an octave above the violin's top string, ie. tuned to e. In any case, as the table illustrates, the measured string length in proportion to the violin, as seen in the last column, will only allow the rebec shaped instrument to reach a note of a, and the boat-shaped pochette can only reach a tone higher that the violin, rather than the required minimum of a fourth as specified in the text - much less the octave as the plate suggests. The other instruments are all acceptably scaled in relation to the violin.
In the work of James Talbot there are problems of ambiguity surrounding many of the measured instruments. An example which is applicable to many of the stringed instruments can be seen with the violin. Talbot lists his measurements from the top to the bottom of the instrument. Firstly he measures from the top of the scroll to the nut, and then from the nut to the lower end of the fingerboard, a measurement that he has questioned in the text. It is not stated which - if either - of the measurements includes the nut itself. This has implications for the string length, and consequently the potential pitch. A modern nut is typically about a quarter inch wide at the bottom, meaning Talbot's violin could have a string length of either twelve and three quarters or thirteen inches. Talbot gives a cumulative total, but a study of his manuscript shows that this measurement is a simple addition of his various figures, rather than a direct measurement.
Likewise, there are other measurements that are ambiguous. It is uncertain how the body length and widths were measured - they could be either over the arch, or direct line. And his measurement of the sides, or Rimms, as he calls them, is one and a half inches, which presumably, though not stated at all, includes the thickness of the back and soundboard at the edge.
One may speculate that the intention would have been for Talbot to give the measurements to the engraver who used them as the basis for the illustrations in the final publication, presumably being checked by Talbot as drawings before being engraved or cut for publication. Thus, these ambiguities that exist in the manuscript would not occur in the final copy. The illustrations would, of course, be no more accurate than the quoted dimensions, and would be one step further removed from the original instruments themselves.
One may also speculate that Praetorius worked in a similar manner, either by taking direct measurements from an instrument which were passed on to the artists, or possibly by the artists working directly from the instruments using some form on proportion grid. Various reversed illustrations show that there was at least one intermediate step in between the instrument and its published likeness. Of the two possibilities, direct measurement would be far more accurate. It should be pointed out that the scale used by Praetorius - often a ratio around 1 to 10 - means that a small error in any of the illustrations can mean large errors when calculating dimensions from the plates. Certainly errors in the illustrations of this nature were made, as any examination of the rulers make obvious. The use of perspective, which can be clearly seen in the illustrations of rectangular keyboard instruments further complicates any meaningful use of the plates as truly accurate representations of the depicted instruments.
Although I have personally faced the problems of trying to understand the manuscript of James Talbot, I believe his information is probably the most useful to a modern researcher. His measurements, which admittedly can and do contain errors of a scribal nature, are in a form that generally allows an understanding of an instrument, with probably less dimensional inaccuracy than Praetorius' illustrations or Mersenne's generally crude drawings. The textual information of all three authors, on the other hand, is generally of great value. Despite their different backgrounds and intentions, all give good details of the instruments they describe. Mersenne's mathematical background, and emphasis on harmony of various sorts, often results in tables which superficially provide much information that further consideration shows to be incorrect, but his book often gives information that is not considered by the other writers.
Praetorius is perhaps the best equipped to write about instruments, given his practical background as a musician. However, perhaps the most reliable source may be Talbot. Talbot's information came from various musicians and makers of instruments. He managed what can only be described as a huge diplomatic success in obtaining information about organs from arch-enemies Father Smith and Renatus Harris. His work appears to be honest, and, surviving as it does in manuscript, complete with various crossings out and additions, gives an insight into his research methods.
* * * * *
The specific uses of these works are generally obvious. Any instrument that is described can be assumed to have existed, although all three authors mention and describe instruments which did not still exist at their epoch. They can amplify evidence found in surviving instruments, and in some cases - the cornemuse of Praetorius and polyphone of Talbot being particular examples, they can provide detailed evidence about instruments types which no longer survive, or with instruments such as the geigenwerck survive as single examples in a form quite dissimilar to those illustrated.
The writings can also give invaluable information concerning the history of particular instruments - the arrival of the bassoon in England, or the leading makers at a particular time. Occasional other comments can provide fertile leads to follow - how usable various types of metal were for trumpet making and their tonal characteristics according to William Bull, the leading English maker of the time; or where particular materials came from - Talbot specifically says that catline gut strings originated in Catalonia, for example. The illustrations serve a purpose similar to that of a photograph today - they show the visual characteristics of an instrument, and, not withstanding what I have previously said, give an indication of the dimensional proportions from instrument to instrument. And again, they can provide interesting clues that are otherwise known - for example the tapering sides of the orpharion illustrated by Praetorius show a similarity to the cittern family, unlike the essentially parallel sides that are found in the surviving instruments. Other features such as tangent rails on clavichords are illustrated by Praetorius, many years before the first surviving examples appear, and his rectangular virginal is illustrated with split accidental keys - unique in a northern-type instrument.
For these and other reasons, the various documentary sources are all invaluable for our research into seventeenth century instruments. However, as I have shown, the details must always be treated with extreme caution, as the errors and ambiguities found in all three of the works make many detailed conclusions made on the basis of the texts and drawings open to question. Perhaps our attitude should be to study closely, but not microscopically, as I am sure that all three authors would be a little surprised, though no doubt delighted, to find their works studied as we enter the twenty first century.

Edinburgh, July 2000.