Neumes vs. notes

I've been reading Timothy McGee's 'The Sound of Medieval Song' to try to get some insight into the evidence for how a medieval musician would have 'decorated' (or added 'flowers' to) a melody. One point that emerges, which I had not really appreciated before, is that neumes - the earliest musical notation system, appearing from the 9th century - seem to have the purpose of notating musical gesture, expression or ornament rather than precise pitch or rhythm.

The general view on this is that the 'melody' (as we would think of it today, i.e. the pitch and duration of the 'notes') would have been known, or transmitted by ear, with the neumes providing a mnemonic aid. But it seems telling that it was felt to be of first importance to record how each word should be sung. The medieval treatises that try to describe their meaning show the substantial subtlety that could be conveyed (all following quotes are translations taken from McGee):

"The plica is the bending of a sound by a sound...one type ascends, the other descends" [Walter 129];  

"The plica is produced in the voice by the closing of the epiglottis with a finely controlled and closed vibrato of the voice" [Lambertus, 273];  

"The pes, growing with two small notes, desires to stretch upward; the high one liquefying, abandons what it represents. The clivis wishes always to descend with two small notes, and the note at its end denotes an indistinct sound. It precedes a pause, or lingers in the mouth of one who is pausing as if it had the status of a complete note" [Anon, summa musice, 540]

Quilisma is "a tremulous and ascending note" [Aurelian of Reome] 

 "Quilismas are so called by their resemblance, for quilos [moist] in Greek is 'soil' and mus, 'earth', as if the earth was moist from the reception of waters" [Walter, 95; McGee notes that this etymological origin is pure fabrication, but the intent appears to be a reference to the wavy marks left on a beach as the tide recedes]

And there are many more. As McGee concludes, these descriptions bring to mind Middle Eastern or Indian vocal styles, much more than the present 'classical' voice. 

An interesting exception

I've mentioned previously how all the evidence - particularly from pictures (or 'iconography' as the scholars say) - points towards medieval instrumentalists always playing from memory, and probably learning by ear.

So I was intrigued (while following up references to medieval ornament, which I plan to return to in future) to come across a chapter entitled "A sight reading vielle player from the thirteenth century" in this book. The picture discussed comes from a manuscript of the songs of Gautier de Coincy (Brussels, Biblitheque Royale de Belgique, MS 10747, fol. 3r). It does not appear to be online, nor is the picture viewable in the link to the book given above, so I have made my own little sketch from the book itself:

There seems no question the vielle player (probably intended to be Gautier de Coincy himself) is portrayed as reading music, resting on the same bench as he is sitting on, as he plays. The author of the chapter, John Haines, notes also that the artist seems to have made some effort to show the music accurately (I've tried to copy this accurately as well). There is a single stave with neumes but no words, with details like the uneven number of staves on the two pages suggesting the artist had an actual example they were working from.

One could argue for the "exception that proves the rule". Gautier was a composer, and produced books, and was a literate monk, so perhaps the artist wanted to emphasise that he was not a mere minstrel, and did so by this exceptional depiction of the book. Nevertheless the pose seems naturalistic enough to suggest it was not a complete invention. Certainly food for thought...

The flowers of musical science

So continuing on the theme of what to do in making a multi-instrumental arrangement of a monophonic medieval tune, the next obvious step is to think about finding some variation in playing of the tune. Here I mean variation other than 'just' changing the timbre, dynamics or tempo, but rather, playing something different from the given notes. This could include either simplification or elaboration of the tune, the latter what we would now call 'ornamentation'.

Looking for some information about what actual medieval evidence there is for how to go about this, I was reminded of the existence of the book "The sound of medieval song" by Timothy McGee, which I have come across before but not looked at closely. An omission I feel I should rectify before saying too much more on this topic! Based on the introduction, the book specifically sets out to survey the evidence for 'ornament' (understood broadly) in medieval sources. It also contains this nice quote from the early 14th century treatise of Petrus dictus Palma ociosa:

They speak of the flowers of mensural music when many sounds, or notes (which is the same thing), variously ornamented according to the quality of each one, are placed in due proportion against a single sound or a simple note of only the same duration. Although some say and affirm, however, that the flowers of musical science are numberless depending on the various methods of discant, and that it is not possible to have certainty about things numberless, for wishing for this reason to compose a technical treatise concerning such flowers.

The other commonly used term for ornament in medieval treatises (according to McGee) is 'colour', a term borrowed from rhetoric. Together the addition of 'colour' and 'flowers' seems to me to make an excellent mental image for the enhancement of a tune - like some lovely medieval manuscript...

Setting the tone

So you've decided to use several instruments together to play your monophonic song or dance. Perhaps one instrument will take the lead on the tune all the time, or perhaps this will switch around. So the next 'obvious' step for the experienced consort player is that you should 'set' the tune, i.e., write some 'harmony' lines for the other instruments to play. Right?

Again, this is not a bad instinct per se, but contains many potential pitfalls. I'll talk about those in a future post (précis: think counterpoint rather than harmony)  but first want to discuss some other options that should perhaps be considered first, both as being simpler, and also often more justifiable. So assuming one instrument at any time is going to be carrying the tune, what could the other instrumentalists do?

(As in previous posts, the following is an attempt to summarise what we have learnt over a long period of playing, and listening to others play, this music. For some it might seem very obvious, but thinking back, it was not so obvious to us when we started out, so I hope it may be useful to others!)

1. Drone on the tonic (if in doubt, this is almost certainly the last note of the tune). Nearly all medieval tunes stay in one mode and sound good with a drone; indeed as I have discussed before, instruments with an in-built drone (e.g. hurdy-gurdy and bagpipe) were popular, and many others were probably played, by default, with a sounding drone (e.g. fiddle and harp). It makes an instantly recognisable sound that differs from renaissance and later harmony. If you are not used to it, it is a good way to educate the ear into enjoying 'tension' and 'resolution' (e.g. from a second interval to unison or seventh to octave).
2. Make the drone rhythmic. This is a great way to add some life and variety. If you are feeling a need for a drum, try this first. It is also a useful way to gain practice in being able to vary emphasis and attack in phrasing (however your instrument allows) without having to worry about the notes. For plucked strings (and actual drums for that matter) don't forget damping as a way to emphasise or change the pulse.
3. Add a drone and a fifth. Generally a tune that works with a drone will also sound good with the fifth added sometimes: either together with the drone if the instrument allows; or changing to the fifth for some phrases; or using the tonic and fifth in a  rhythmic pattern.
4. Add a 'movable' drone. Okay, by definition a drone is the same note throughout the piece. So this is a step towards 'setting' the tune. Listening to the tune with a drone, there will often be sections of a bar or longer that stay in 'tension'. Changing the drone to a neighbouring note of the tonic for those sections will usually work. I sometimes call this the 'home and away' approach, as propounded by the fantastic harpist Bill Taylor, who links it with the intriguing 'binary measures' found in the Robert ap Huw harp manuscript. For a more scholarly discussion of how this might be justified by existing medieval polyphonic sources see this article by Margo Schulter. 
5. Simplify or elaborate the tune... see my next post!

Three final points about drones: all of the above are possible for recorder players if you have one in your ensemble; all are good places to start if accompanying a singer; and I'd also recommend them as an approach to 15th century dance music as well as medieval music.

The many instrument solution

So what to do when faced with a relatively simple, single line tune, and the intent to perform it instrumentally? This is something we do quite a lot, so here is an attempt to set down some of the thoughts emerging from our experience and research...

[Note that if there is an option to sing, e.g. it is a cantiga with words, rather than an estampie without, this is always worth considering, even if none of the group are 'singers'. Just attempting to sing it will often make clear details of emphasis and phrasing, related to the text, that can help the instrumental performance. But for now I will assume the performance itself will be purely instrumental].

One thing often seen and heard is the 'play every time through on a different instrument' solution. This seems an obvious way to add some variety and contrast (though it can become monotonously predictable if there is always a strict alternation, and the same happens on the next piece, and the next...). It's a good starting point, but in my opinion, has some pitfalls:

• One that is my current bug-bear is when players keep putting down and picking up new instruments - something I think it highly unlikely that medieval instrumentalists ever did, unless in very unusual circumstances. The risk of coming in with an instrument that is out of tune seems reason enough alone to be wary. It is even more annoying when it is a recording and the instruments vary from lutes to shawms with the impossible volume balance being evened out in the editing suite. If you want to show off your multi-instrumentalist skills, save the new instrument for another piece! Similarly, I would like to think we are past the stage where the primary purpose of period concerts was for the audience to goggle at the biggest possible variety of unusual instruments - that's getting dangerously near to the old 'drums and fun' territory.
• For many instruments, it is not necessary to stop and start playing them to produce dynamic or other contrasts which allow the 'lead' to effectively switch from one instrument to another. And learning how to create those contrasts on your instrument is going to be a useful skill. Perhaps most problematic here are instruments such as recorders that have a more limited dynamic range - sometimes stopping and starting may be necessary to allow another instrument to show through. Of course, not using a recorder at all is probably the more medieval solution, which brings me to...
• Using the wrong set of instruments. Recorders or similar fipple pipes may have existed in the middle ages, but recorder consorts are a renaissance phenomenon. The medieval 'ensembles' depicted playing are nearly always instruments with contrasting timbres (e.g. harp and fiddle). Having said that, there are of course the famous pictures of musicians from the cantigas manuscript that show many pairs of players of identical instruments. I wouldn't want to dismiss this as just some artistic conceit, so it deserves consideration as a plausible instrumentation.

So even if you have a nice, small ensemble of instruments, with contrasting timbres, and variable dynamics (but in a similar dynamic range), playing 'just the notes' several times through is not likely to be satisfying for performers or audience. I'll talk about 'what next' in the next post.

Lacking an arrangement

So it has been a while, but as there still seems to be a few people out there viewing the pages, it seems worth trying to revive this blog...

A musical colleague (sadly no longer resident here) is visiting over the holiday season, and  has prompted some discussion of the issue of 'arranging' medieval tunes. More specifically, her experience in forming a new medieval music group in her new (old) home has brought forward a common issue. How do early music instrumental enthusiasts - who are perhaps most often introduced to the genre by playing in recorder consorts - react when faced with a typical monophonic song from the middle ages?

For example, a great medieval resource are the cantigas from the thirteenth century collection of Alfonso the Wise. But the manuscript (or transcription) provides a single vocal line. What should an instrumental ensemble do with this?

Option 1: Play the notes. This tends to be the instinctive reaction of the 'modern' medieval musician with a classical training background. Indeed, they often pride themselves on the ability to play the notes accurately on first sight; and even seem to think it a waste of time to play them again if they have been played accurately the first time. For most cantigas, this provides a rather minimal challenge, and does not take long.

There is little evidence in iconography that medieval instrumentalists ever played from music, so it seems unlikely that this is an authentic approach. But even if learned by ear, would an instrumentalist (or ensemble) simply have played the tune 'as written' (or in unison) perhaps a few times through? If not, what would they have done?

Some options to be discussed in future posts...



 

Roast and boiled

So, we attempted the version of Danse de Cleves suggested by Robert Mullally in his new book, and remain unconvinced. It is not completely implausible for the musicians to speed up and the dancers to slow down enough to make it work; but seemed to significantly reduce the enjoyment of both of this otherwise lovely dance and tune. I'd be very interested to hear other experiences.

Actually, looking more closely at the argument in Sachs, he (unlike Mullally) does not attempt to reconcile his theory that all basse danse (and Italian bassadanza) should be in 12/8 (effectively moving in 4/4) with the Brussels notation for Danse de Cleves, but instead argues that it (and the other dances with mensural notation in the manuscript) are not, in fact, basse danse at all.

This would potentially undercut what I had planned as my next argument that the 6/4 tempo is the more viable interpretation for the basse danse breve. This is based on the Brussels manuscript dance 'Roti boully joyeulx', which appears (from name and instructions) to be related to 'Rostiboli Gioioso' (attributed to the mid-15th century dance master Domenico, though with steps only recorded in the slightly later 'Ebreo' manuscripts) which in turn shares clear structural similarity with 'Gioioso in Tre' (i.e. 'for three') for which Ebreo claims choreographic credit, and which in one Ebreo manuscript has music (reproduced as example 4 in this article by Barbara Sparti). To close the loop, the relationship of the music for 'Gioioso' and 'Roti boully' is pretty clear.

This dance appeared to be a 15th century European-wide hit, also appearing as 'Rostibin' in a German dance instruction manual, and with variants on the name appearing in multiple sources, including a late 15th century poem in Scots as 'Rusty Bully'.

In all versions it appears to be a 'balli', that is, a dance with mixed tempo. The Italian 'Rostiboli' and 'Gioioso' appear to start in bassadanza (although the tempo is not explicitly given, the step sequence is highly typical) and then change to saltarello followed by piva. The Burgundian 'Roti boully' starts in 'pas de braban' (= saltarello?), has a second section that seems faster (=piva?) and then explicitly says "then follows the basse dance" with the notation changing to (mostly) breves, with a step (e.g. 'd' for double) written above each.

So focussing just on this part of the music, there is an almost perfect match of the breve sequence in 'Roti boully' to each bar of the explicitly written out 6 beat rhythm in 'Gioioso'. In fact reference to 'Gioioso' nicely solves some of the oddities in 'Roti boully', such as missing steps above some of the breves, where faster movement through the preceding notes to match 'Gioioso' (as is notated specifically in the final section of 'Roti boully') would have provided the right number of notes to steps. It seems abundantly clear from these (and other) Italian choreographies that a bassadanza double corresponds to a 6/4 bar.

So, in transcribing this section of 'Roti boully' as 4/4 (as he does) Mullally would have to argue that the rhythm of basse danse has changed in this highly similar, and more or less contemporary, dance, or ignore the relationship to 'Rostiboli' and 'Gioioso' altogether. In fact, the latter is what he does. His extensive note on this dance (pp62-65) mentions a whole set of other literary references (such as the Scottish 'Rusty bully') but says nothing at all about the Italian sources. There is a short and somewhat unclear discussion of the general relations between the Burgundian basse dance and Italian bassa danza and ballo on pages 20-21, but the specific relationship between 'Roti boully' and 'Rostiboli' is not mentioned.

So maybe this is a ridiculously obscure point to get hung up on, but as someone who greatly loves the 'queen of tempi' it matters a lot to know if I've been doing it wrong. And (full disclosure) I'll even admit that I have presented a 4/4 basse danse in the past, on our first CD (I'll let the arranger responsible identify himself, if he wants to admit to it). But I'd be really sorry to see that catch on as a result of this recent publication.