Music of an Earlier Time: in respect of before and after

Andrew Lawrence-King offers an online practical workshop that explores how Time might have felt to 17th-century musicians, hosted by Amherst Early Music Festival on Saturday June 5th, 2021: further details here.

More than a century after Einstein’s annus mirabilis of 1905, in which he published four paradigm-shifting papers on Quantum Theory, Brownian Motion, Mass/Energy Equivalence and Relativity, most of us still consider the implications of Relativistic Time to be paradoxical.

We live our everyday lives and make music within a pre-1900 understanding of Time ‘like an ever-rolling stream’. This is Newton’s concept of Absolute Time, first published in his 1687 Principia, and bitterly contested amongst mathematicians for decades afterwards. Amongst 18th-century musicians and the general public, acceptance of Newtonian ideas must have lagged even further behind, just as for most of us today, Hawking’s Brief History of Time (1988) remains counter-intuitive.

In this sense, the works of Josquin, Dowland & Monteverdi, even Lully & Purcell, are ‘music of an earlier Time’. Until Newton, Time was defined by Aristotle’s Physics (4th cent. BC) as ‘a number of change/motion in respect of before and after’.

At this participatory workshop, everyone is sure to have a Good Time. Andrew leads some mind-games and musical practice, exploring how it might feel to make pre-Newtonian music. What can we discover, not just quantitively, but about the quality of Aristotelean Time? Dare we follow Phaeton’s example, and seize the reins of Apollo’s time-chariot? What about ‘unmeasured’ genres? What really counts in Early Music?

Open to singers and all instrumentalists, relevant to all repertoires up to at least 1800. Instruments are optional, but please bring with you an open mind and a free hand (or foot)!

Looking for a Good Time?

This article is a personal summary and commentary on the Colloquium presented online on Sunday May 2nd 2021 by Aapo Häkkinen and Domen Marinčič and hosted under the aegis of the Helsinki Baroque Orchestra in collaboration with the Slovenian Academy of Sciences and Arts – more details, and a video of the live event, here. Where I have reproduced citations given by presenters, I ask the reader to bear in mind that I was taking hasty notes from a live talk, and to consult the video and the original sources for authoritative details.

The event’s subtitle Tempo Rubato – Use, Flexibility and Modification of Time deserves further comment (see below), appearing to threaten a presentation of only one side of a debate which – like discussions of Vibrato in Early Music – all too often features campaigning for fixed personal opinions, rather than investigation of historical evidence. But as various speakers gave their papers, there was ample consideration of temporal structures, and if anything was missing, it was investigation of how rhythm might be ‘malleable’ (to use a word that emerged during the event).

In general, questions of What and When were examined carefully, and Jed Wentz gave an impeccably concise and impressively persuasive account of ‘How to do Affekt‘ in the mid-18th century: otherwise, questions of how to apply the rich information provided were left for another occasion.

One such future occasion might be my presentation on Music of an Earlier Time for Amherst Early Music, which will offer participatory exercises exploring how the period philosophy of Time can be applied to practical music-making, using historical terminology, conceptual frameworks and embodied practices: Saturday June 5th 2021, read more here.

The event image was an 1851 illustration of the motto Tempus fugit (Time flies). The metaphor dates back at least to Classical Antiquity, and is cited in the opening phrase of the first ‘baroque opera’, Cavalieri’s Anima e Corpo (1600) read more here. The context of this motto, in Virgil’s Georgics (29 BC), Cavalieri’s drama, and in general (including this 19th-century illustration) is memento mori, a reminder that our life-time is short, with the implied challenge to use time well.

Other images from earlier periods address the specific question of the relation of Time and Music (significantly, Movement is usually – always? – also featured). One of my favourites is Poussin’s Dance to the Music of Time (c1635), rich with iconographical symbolism: you’ll notice Time playing his music on Earth, but don’t miss Apollo’s Time-Chariot in the sky above.

I also have a favourite early 20th-century image which would seem to express most aptly the ‘malleability’ of Tempo Rubato in this period, Dali’s The Persistence of Memory (1931). And indeed the first paper at this Colloquium examined the strong focus on metronomic precision in this very period, but with little mention of the strong advocacy of Tempo Rubato in precisely the same period.



We should take great care with the term Rubato. Paderewski’s views on Tempo Rubato were first published in 1909: the term is also strongly associated with Chopin (1810-1849). In a very quick online search, the earliest dictionary reference I found was 1883.

The standard work on the subject is Hudson Stolen Time. Hudson (from page 43) declares Tosi (1723) to the be the first to use the term, in the phrase rubamento di tempo, and in the context of the aria patetica (passionate aria): Tosi specifies that this happens ‘exactly on the true motion of the bass’ (as translated by Galliard in 1743). Hudson cites Roger North’s terminology of “breaking and yet keeping time”, found in several sources, the earliest being an untitled MS c1695, shortly after Tosi’s visit to London.

It is clear that these citations c1700 refer to occasional freedom for a soloist to anticipate or (more usually) delay, whilst the bass continues steadily. Galliard’s mid-18th-century footnote draws attention to Tosi’s repeated insistence on regard for, and strictness of Time, and to the ‘singular’ [rare, unusual, isolated] application of ‘stealing the time’: again we read that “the bass goes an exactly regular pace” and that the soloist “returns to exactness, to be guided by the bass”.

The bass goes an exactly regular pace (Tosi/Galliard 1743)

Hudson also cites Quantz Versuch (1752) illustrating eine Art vom Tempo rubato, again with anticipations and delays to the solo flute, whilst the continuo-bass remains steady.

Froberger’s c1710 marking a discretion and the notation of preludes non mesurées show that some music was indeed unmeasured, and more work is needed to explore how such music would have been realised, for example by careful examination of sources that combine specific note-values with (seemingly contradictory) indications of being ‘un-measured’. See my take on Senza misura in baroque music, here (scroll down the article until you see the Cuisenaire Rods!)

Even with an improved understanding of how to play unmeasured music, the case has certainly not been proven that such an approach should be applied to measured music. Indeed, the quality of being ‘measured’ is the essential defining quality of most early music. See Time: the Soul of Music, here.


Before 1800, the concept of ‘using Time’ seems to be found exclusively in the context of how one makes best use of one’s lifetime, rather than in music-making. Thus Herrick’s (1648) Gather ye Rose-buds while ye may / Old Time is still a-flying concludes …. use your time for love and marriage, following the motto carpe diem. In the musical context, Time is characterised as measuring music, indeed for Zacconi (1592) the terms tempo, misura, battuta and tatto (Tactus) are synonomous.

But the word ‘use’ also had a particular meaning in the context of 17th-century Arts. ‘Art’ itself was defined as a collection of rules, a set of organising principles. What we nowadays mean by ‘art’ – those mysterious, ineffable beauties that transcend everyday experience is renaissance ‘Science’. In this period terminology, ‘Use’ is the nitty-gritty of what was actually done in practice – we might almost think of it as ‘technique’. In general, period sources tell us more about period ‘art’ – the rules – than about period ‘use’ – how to do it. More about Science, Art, Use here.

In this particular sense of ‘techniques related to Time’, the ‘use of Time’ is a fascinating topic for historical investigation: we should be careful not to equate this with any period assumption that Time was a commodity available for musicians to use as they chose to.

Historical discourse rarely (if ever?) characterises Time as ‘flexible’, before the period Paderewski, Dali and Bergson, whose interlinking of psychological Time and Freewill dominated the philosophy of culture in the early 20th century. Read more here.


Although ‘modification’ of Time has come to be an accepted phrase in modern-day discussions of Metre in Music (see for example George Houle’s essential book (1987) here), this is not historical terminology nor a period concept of the relationship between humans and Time. There is no doubt that something of this nature was practised – as Domen Marinčič showed in his presentation – but the period phraseology was of “guiding” or even “driving” Time: the Italian word guidare is also used for driving a chariot. We can catch a glimpse of the period concept when we consider the myth of Phaeton, who seized the reins of Apollo’s time-chariot, but was unable to control it and crashed spectactularly. Early Music welcomes careful drivers…

Careful consideration of terminology is vital, if we are to avoid imposing modern-day assumptions when we glibly apply modern-day vocabulary to earlier periods; and if we wish to understand how the rules of period ‘Art’ were embedded in historical philosophies, in order to appreciate how those old rules felt ‘natural’ to musicians back then.


In discussions of Historical Performance Practices related to Musical Time, there is also a need to distinguish clearly between two – interlinked – questions: tempo as the ‘speed’ of music; tempo as the regularity or otherwise of rhythm at any given speed. In both these aspects, tempo is closely related to Affekt. And underlying all of this, but not addressed in this Colloquium, is the question of Time itself, since Science, Philosophy and general perceptions have changed significantly over the centuries that separate Aristotle, Newton, Einstein and Hawking.

Alexander Bonus on ‘metronomic’ Tempo

This paper was concerned with the use of the metronome to establish rhymthic regularity, not with questions of ascertaining musical speed. Although Maelzel’s metronome was patented in 1815, and Loulié’s chronometre was describle in 1696, AB made the point well that the use of machines to train musicians to play in ‘metronomic rhythm’ became prominent only in the early 20th century. He did not address the prominence of discussions of Tempo Rubato in this very same period: surely these two phenomena are closely interlinked.

AB’s message seemed rather to be that ‘metronomic’ playing is undesirable, but is a phenomenon of the 20th century. It is hard to disagree with those points, although I readily confess that I greatly appreciate the excitement and emotional power of late 20th-century rock music (but my favourite vintage pre-dates the routine use of click-tracks in popular music, and this may well be crucially significant).

AB was greatly concerned that “the belief that tempo is defined by clock technology” is “so central… to performance… even to the reading of notation”. I would agree that this is largely a 20th-century phenomenon, as is the contermporary (over-) reaction against any kind of rhythmic regularity. Both are a feature of modern-day Early Music, and that is regrettable in both instances.

Citing Hofmann in 1905 “keeping of absolutely strict time is thoroughly unmusical and deadlike”, AB also gave highly negative spin to such phrases as “the inexorable beat”, and “the beat is the steady pulse”, even to the whole notion of musicians being beholden to “external tempo”, to rhythm “not depending on the will”,

Here AB went too far, in trying to limit regularity to the 20th-century alone. The inexorable character of Time itself is expressed by Virgil (the tempus that fugit – flies – does so inreparabile, unrecoverably), and the steadiness of the renaissance Tactus beat is strongly characterised in many period sources, for example Zacconi.

Zacconi’s person who administers the Tactus’ would create an ‘external tempo’ for all other members of an ensemble. And the ancient concept of the Music of the Spheres, still current in the 17th century, implies that human ‘free will’ is a lower priority than the divine perfection of heavenly music, which we should imitate in our earthly performances. According to the religious views of the time, “free will” is in general a concept fraught with dangers.

The concept of the Music of the Spheres also connects good music-making to physical, spiritual and moral health. AB noted that in 1895 regular rhythm “equates to good, healthy behaviour”. No doubt, Dowland would approve.

AB also mentioned a crucial distinction: in 1889 “by accurate rhythm is not meant metronomic accuracy”.

Here is the gateway towards a much more productive approach than mere trash-talk about metronomes. In what way was Zacconi’s and Dowland’s measure ‘equal’, and precisely where was there room for what we would nowadays call ‘freedom’?

Julia Dokter on German organ music c1700

Julia Dokter’s presentation outlined the central conclusions of her book, published in the last few days, Tempo & Tactus in the German Baroque, here. Her approach was an exhaustive survey of theoretical sources, applied to case-studies of various musical compositions, all within a specific genre. She was most properly cautious about extending specifics from this particular repertoire and genre to other countries, periods or genres. Nevertheless, the concepts she introduced seem to reflect fundamental practices related to musical time in this period. Notably, her results were strikingly parallel to those presented in Aapo Häkkinen’s paper, addressing another repertoire and with different methodology.

JD focussed on Tempo transitions, informed by three types of period notation: time-signatures, note-values and tempo-words.

Looking at Baroque time-signatures, as they evolved from renaissance mensuration marks, she first cited Michael Praetorious in the early 17th century, explaining two notational systems for duple metre: Motets in C/ time, counted as two semibreves; and Madrigals in C time, counted as two minims. These two notations are NOT proportional: the minim-beat in C is neither twice as fast, nor the same as the semibreve-beat in C/. Rather, the difference is ‘about one and a half’: we should bear in mind that the period concept of a ‘half’ is not necessarily as strictly 50%, but rather more loosely as some part less than the whole and more than nothing. So we beat C/ somewhat slower (and that beat represents semibreves), and C somewhat faster (and this beat represents minims). This is similar to what we read in Zacconi, who warns that the Tactus-beater should not mistake his mensuration marks and give the beat at the ‘other speed’, as this would probably crash the entire ensemble.

JD explained that this is consistent with a general principle in Baroque practice, that time signatures denominated in smaller note-values (i.e. 3/8 compared to 3/4) have a slower Tactus, so that the small note-values go faster, but not twice as fast. Her later examples extended this principle to a general principle that passages in very small note-values would be assumed to require a slower Tactus – in order to be playable at all!

JD applied another early 17th-century practice, the triple-time proportions of Sesquialtera (slow), Tripla (medium fast) and Sestupla (very fast) to her case-studies in high Baroque organ-music.

Her conclusion is that there were effectively two systems of notation (and execution) of duple Tactus, each with its three associated triple-time Proportions. Similarly, each faster Proportion might be only somewhat faster, not necessarily twice as fast as the slower Proportion.

So whether in duple or triple, smaller denominations of time-signatures and smaller note-values in what JD calls the ‘surface activity’ both suggest a slower Tactus. The result is faster surface activity, but not so much as twice as fast.

This is a principle we see at work as early as Monteverdi’s 1610 Vespers, where at Et exultavit in the Magnificat there is the marking va sonato tardo, perche li doi tenori cantano di semicroma (will be played slowly, because the two tenors are singing semiquavers).

Nevertheless, as became clear in Domen Marinčič’s presentation (below), this period principle is contra-indicated by other early 17th-century indications also observed by Julia Dokter in the later repertoire, that changes to faster surface activity may require a faster Tactus, to heighten the contrast.

JD noted that the “demise of the Proportional system” described by Kirnberger in 1776 can already be seen in the music of J.S. Bach c1740. As C becomes a quadruple metre, with four crotchet beats (whereas c1600 it was a duple metre with two minim beats, see above), the old system of proportions collapses. I would add that the emergence of fashionable French dances, many of them in triple metre but with subtly different speeds, rhythmic structures and subjective affekts, also contributed to the slipping of the gears of the old Proportional system.

Nevertheless, JD proposed strict proportions for JSB’s Eb Major “St Anne” Organ fugue (the associated Prelude has passages in alla Francese ‘overture’ style), with the constant beat transferring from semibreve to dotted minim to dotted crotchet.

I would add to this, citing Carissimi’s comment in Ars cantandi published in German translation in many editions around 1700, about the affektive quality of proportions:

The triple-metres all agree with regard to quantity, division and proportion, as is easily understood, but in the slow or fast quality, known by the Italians as tempo and by the French as mouvement, they are utterly different.

Carissimi Ars Cantandi (1692 etc)

Here Carissimi contrasts the quantitative – “mathematical” – elements of beat (constant), note-values (consistent between different triple mensurations), proportion (simple ratios of the underlying duple metre) with the affektive quality that results. Although the duration of any given note-value (e.g. a minim) is the same in Sesquialtera, Tripla or Sestupla, the contrasts in harmonic motion and surface activity create very different feelings for triple metres of three semibreves, three minims, or twice-three crotchets. And this qualitative, affektive element is ‘known by the Italians as tempo and by the French as mouvement‘. And in German, Bewegung.

So we are considering not only speed of beat, and the surface activity within that beat, but also the affektive quality that results. Read more on Quality Time here.

Depending on context, we may choose (or be encouraged by tempo words) to exaggerate contrasts by over-doing Proportional changes (see DM below). Or we may need (with the support of tempo words) to reduce the contrast as calculated mathematically, whilst (presumably) still acheiving an effective contrast of Affekt.

In the conference, JD expressed the opinion that – in the later repertoires she studied – tempo words that appear to suggest heightened contrast merely warn against a more common practice of automatically reducing contrast. DM gave clear evidence – mostly in earlier repertoires – that tempo words exaggerate the expected contrasts. I would suggest that we have evidence to show that proportions could be precise, lessened or heightened in various contexts, and that tempo words help us judge which way to go, with affektive contrasts as the end-goal, and playability as an inevitable limitation.

Domen Marinčič on variations in Tactus speed

Domen Marinčič summarised his excellent paper on Now faster, now slower, citing sources that incontravertibly show changes of Tactus speed, starting with Vicentino (1555) changing measure according to the text. He suggested that this practice of changing the Tactus was only one of many options, such as “agogic freedom”, “rubato”, “managing rhythm expressively”, but he did not cite evidence for or discuss in detail these other options.

Whilst Mersenne (1636) gives a default tempo of around one beat per second (i.e. a one-metre pendulum), DM reminded us that Mersenne also considered the pendulum an ineffective tool, since so many different lengths would be required for the slightly different speeds that are required. Nevertheless, DM’s suggestion that it is left up to performers to choose their own tempo is unsatisfactory: the period discourse asssumes that there is a correct tempo, and the performers’ job is to find it, not choose their own. Of course, we don’t always have enough, and clear enough, information to find the tempo for sure, but nevertheless, that is what we are supposed to try to do!

I would add here that Frescobaldi gives us a practical method for finding the correct tempo, a method that is quite different from the mathematical calculations and abstract musical analysis that we tend to use nowadays. Frescobaldi’s instructions are to play the music through (in some default tempo, presumably considering a standard Tactus and the apparent surface activity): as one listens, one will understand what Affekt the music has. And for how to proceed from Affekt to execution, see Jed Wentz’s paper (below). More on Frescobaldi Rules OK?, here.

Frescobaldi also gives us vital information and essential practical advice. The information is that even ‘difficult’ music with changes of Tactus is facilitated by using a Tactus beat. This contra-indicates any assumption of general rubato in this repertoire, replacing it with highly specific instructions for when and how to change the beat. The practical advice is therefore to study by physically beating Tactus with the hand, and play keyboards etc whilst physically beating Tactus with the foot. Tactus is not just a theoretical concept, it is an embodied practice.

DM cited Glareanus increasing the speed (not necessarily in strict mathematical proportion) by changing mensuration mark. Banchieri beats both C and C/ with a minim-beat, but at different speeds. Other sources change the note-value associated with the Tactus beat (e.g. Zacconi, who also changes the beat-speed accordingly). Praetorius uses a variety of time-signatures to indicate different tempi. An Entrée in L’Amour Malade has exceptionally many changes of time signature, and therefore, tempo.

DM pointed out exampes in very well-known repertoires where even highly respected modern editions have ‘rationalised’ or suppressed differences in time signatures that would seem to indicate tempo contrasts: between successive Minuets and Bourées in J.S. Bach’s Cello suites, and the Minuets in the first Brandenburg Concerto. This idea was echoed, from a very different approach, in Jed Wentz’s paper.

I urge readers to consult DM’s published article here.


In discussion amongst the presenters, AB reiterated his central points, that the intense application of the metronome to music education, the mimicking of the ticking metronome as the model of rhythm, and the concept that a mechanical standard should be followed, are all 20th-century phenomena.

JD added a fascinating observation from her comparisons of sources of J.S. Bach’s works, that the composer seems to have changed his priority from trying to notate speed, to trying to notate Affekt. This fits well with Carissimi’s ideas of durational Quantity and affecktive Quality (see above).

JD also suggested that strict mathematical proportions might be just the outline structure and the theoretical basis: “in practice, it becomes more malleable”. In the sense that the proportional change itself might be slightly greater or less than the mathematical ratio, this suggestion is thoroughly supported by period evidence, including many citations presented during this event.

A written comment by an online listener expressed disappointment at so much talk of notation and structure, opining that all this had been heard before from Willem Retze Talsma in 1980, and interested to hear about “freedom, departing from those “absolute” tempi… that is the accelerando and ritardando from the basic tempo, gradually”.

Certainly, all the evidence heard during this Colloquium fitted excellently with the notion of well-structured Tempo Giusto, though with different quantitative speeds according to mensuration marks; and with systems of Proportional relationships for triple metre, but with the possibility of ‘tweaking’ those mathematical ratios one way or another in particular circumstances. No evidence was presented at this event for any general “freedom”, nor for gradual changes of accelerando/ritardando. Indeed Frescobaldi clearly states that changes of Tactus are executed by suspending the Tactus momentarily in the air, and then starting the new movement resolutely. Based on all the evidence I have seen, my coaching mnemonic for ensembles and students is “use the gear-shift, not the accelerator/brake”.

I had the opportunity to meet, hear and talk with Talsma in the early 1980s, and this was my first encounter with the concept of Tempo Giusto. Of course, the ‘double-beat metronome’ theory for Beethoven etc has by now been totally exploded more here, but my research findings utterly support the fundamental concept of a (more-or-less) fixed speed (but ‘fixed’ subjectively, not with any kind of clock) in mensuration mark C [although this changes during the 17th-century from a duple to a quadruple measure, see above]. Indeed, Beethoven himself comments on this concept, wishing to be free from it (and thus confirming its strong presence until then).

But, in spite of the remarks of the online listener, the application of Tempo Giusto nowadays differs sharply from Talsma’s version in the 1980s, in that we measure the ‘correct tempo’ with a slow Tactus, avoiding the ‘sewing-machine’ effect of Talsma’s measuring of small note-values. During recent decades, there was even an idea that counting in ever-larger note-values might be better and better (still supported by Robert Hill amongst others). Roger Mathew Grant’s excellent book on Beating TIme and Measuring Music (2014) shows that measuring (by the Tactus hand) was done at a particular note-value (c1600 the minim in C, and the semibreve in C/; c1700 the crotchet in C).

Inja Stanovic‘s paper on the technologies of early recording, though valuable and interesting, seemed to me to belong to another occasion. Of course, the recording industry has had a most powerful effect on modern-day Early Music, supporting it immensely, especially with the arrrival of the CD in the 1980s. But my personal experience is that the technologies of the late 20th century had less influence on performer choices than did record producers. Almost invariably, young HIP ensembles making their first recordings were supervised by more senior ‘classical’ producers, and the process was dominated by seeking to control tuning, vertical unanimity of rhythm, and the avoidance of any surprises. We used to joke that our task was to play until something woke up the producer and he called “Cut”.

By the time a new generation of producers with Early Music experience emerged, the expectations of record companies, the listening public, and even of performing musicians, had been firmly set in a certain path. Seriously, we can well consider how today’s Early Music might have turned out, if all those thousands of CDs had been commanded by jazz producers, who might have prioritised groove and swing over vertical unanimity, drama and emotion over bland smoothness.

One of the presenters (AB?) cited Roger North’s remark that chronometers are very ‘whimmish’, that there is nothing better than a roll of paper in the [human] hand. Daniel Friderici (editor of the 1625 print of the Finnish Piae Cantiones, more here, and recent Finnish recording here) was also cited “some beat time like a clock, and this is an error”. All this encourages us to investigate precisely how the practice of Tactus-beating differed from clockwork, given the overwhelming weight of evidence that the character of the Tactus was steady, equal, unchanging etc.

Jed Wentz on ‘the Art of Acting’ (1753)

This was an inspiring and well-structured presentation, summarising Aaron Hill’s instructions to mid-18th-century actors on how to acheive the appropriate body, facial and vocal expression for a particular Affekt. JW began with Kirnberger’s re-iteration of the doctrine of ‘moving the Passions’ i.e. that motion and emotion are connected (in German, Bewegung and Gemüthsbewegung], with Bewegung as ‘what the French call mouvement‘ i.e. musical Tempo (see my remarks on Carissimi, above): “and the composer must properly hit on this Movement, according to the nature of the feeling” – Die Kunst Part II page 106 – and my thanks for JW for his exemplary citations during a speedy online session). “This is a study that lies outside the music.”, a study which “the composer shares with the poet and orator”.

Hidden in these citations is a vital point: whilst both composers and performers must employ the art of Rhetoric, they each have different responsibilities. The determining of Tempo (in response to the Affekt of the text) is the responsibility of the composer, who notates it as precisely as the period systems allow (and though more precise indications by chronometres were available, it seems they were not wanted): the performer’s responsibility is to understand the composer’s notation and follow it. We read in Quantz that the performer should also be like an orator, and Quantz’s highly detailed instructions on how to do this do not suggest altering the notated tempo, or any kind of general rubato, but rather explain how to structure musical time with a ‘pulse’ around 80 bpm.

JW cited Coeffeteau’s requirement in A Table of Human Passions (1621) page 17 that there should be ‘perceptible changes to the body and voice” of the person feeling the emotion. JW then showed the methodology of theatrical director Aaron Hill, who also produced Handel’s Rinaldo in 1711, as published in the first (posthumous) edition of his Art. The actor should not attempt to ‘imitate a passion’ (by speaking his lines) until “fancy has conceived so strong an image or idea of it…. as to move the same impressive springs within his mind”. Imagination must conceive a strong idea, which (by the action of Energetic Spirits of Passion transmitting from the brain to the body) impresses its form on the muscles of the face; instantly the same impression is felt in the muscles of the body; and the those muscles (whether ‘braced’ or ‘slack’) transmit their own sensation to the sound of the voice and the disposition of the gesture.

Extending the ancient doctrine of the Four Humours, Hill categorises 10 Dramatic Passions (and Love, the ancient Quintessential, can be mixed with any of these): Joy, Grief, Fear, Anger, Pity, Scorn, Hatred, Jealousy, Wonder, Love. JW referred to Hill’s concept of the “quality of the eye”, and the sequence of this technique: reflecting on the idea in the mind, feeling it idea in the body; a physical response of the eyes and nerves; only then should the actor speak.

In this way, the actor avoids the danger of “overleaped distinctions” – missing emotional contrasts. And on stage, these “beautiful and pensive pausing places will appear to be the natural attitudes of thinking”. Without the application of this technique, the audience will remain unmoved.

JW concludes that pauses in performance are therefore essential.

I would comment that Hill’s methodology contrasts sharply with the usual operating procedure in most modern-day HIP productions of ‘early opera’. Usually the focus is on teaching hand, and perhaps body, movements, with the danger that these – however beautiful – strike the audience as being ‘stylised’ and passionless, not genuine expressions of emotion. But Hill’s concepts are ancient, based on Quintilian’s theory of “visions” and the doctrine of Enargeia. Rather than manipulating the voice or the body directly, the first step is to create an imagined Vision of what is described in the text.

In first rehearsals of a new play, actors often struggle to ‘change gear’ quickly enough. But good coaching, effective private practice and sufficient rehearsal should empower an actor to make strong changes of Affekt as quickly as needed. Indeed, many sources on Early Opera emphasise how powerful an effect such sudden strong contrasts have on the audience, see Cavalieri for example. It requires careful judgement to decide how much ‘beautiful and pensive pausing’ to allow in performance.

And in theatrical music (or indeed any passionate musical performance), that careful judgement has already been exercised by the composer, and the appropriate amount of pause has been notated. Samuel Pepys praised Henry Lawes for his precise notation in musical rhythm of ‘every pointing comma’. Monteverdi varies how each speech starts: with the continuo directly, before the continuo, shortly after, after a longer pause. Cavalieri notates the space for affektive changes during the silences at the end of each phrase [last notes notated long are conventionally sung short, see Doni – giving time for reflection, gesture etc within the regular Tactus].

I would argue that since the composer has already notated the appropriate Movement for the emotion at hand (as described by Kirnberger, also in 1600 by Peri and by Pepys in the late 17th-century), the performer’s task is as Hill requires, to create the response in his body and voice before singing, yet to do so within the dramatic timing carefully notated by the composer. Otherwise, we risk spoiling the pauses and continuations carefully notated by the composer: think of a waiter enthusiastically adding salt to potatoes that were already salted to perfection by the chef! More on Pavans and Potatoes here.

But see also JW’s discussion on Mattheson, below.

Aapo Häkkinen on 18th-century tempo relationships

It was most interesting to note that AH reached very similar conclusions to Julia Dokter, albeit in somewhat different repertoire and with an utterly different, yet properly thorough, methodology. His approach was to examine large-scale works and construct – not a pyramid of tempi, based on the slowest tempo – but what he called an ‘hour-glass’ of tempi, centred on a fundamental Tempo Giusto in C-time in the area of 60 to 80 bpm.

From this starting point, the denominators of time-signatures indicate for example that 3/8 is faster than 3/4. And then Tempo words modify (to a lesser extent) the broad indication given by the time-signature. Both these principles are well accepted in modern-day musicology, and the speed-order of the Tempo words is not significantly in doubt. And during the course of say a Handel opera, there are so many movements, each carefully marked with time-signature with or without additional tempo word, that we end up with a large number of tempi in a well-defined order.

If we seek the central speed of Tempo Giusto, and avoid impossible extremes of fast or slow, yet create an appreciable distinction (at least a few bpm) between each and every tempo, there is, as AH put it “very little leeway in choosing tempi if one takes all the tempo words into account”.

And his findings indicated sets of tempi related by proportions, just as JD found by her, rather different, investigation.

For the application of this methodology – ordering the tempi of a large-scale work, and hence determining a fairly precise tempo for every movement – to Handel’s Orlando see here.

In their parallel, but independent, investigations, JD and AH implicitly relied upon two essential period principles, which have guided all serious study in this area, but which many performers are reluctant to accept. Firstly, the historical role of performers was not to choose their own tempo, but to find the correct tempo, which the composer’s notation was intended to convey. And secondly, two movements from the same large-scale work, or two pieces from the same repertoire, that have the same indications of tempo (mensuration marks or time signatures, level of activity i.e. characteristic note-values, time words, dance type etc) are intended to have the same tempo, as near as humanly possible.

One can make a lot of progress in any well-defined repertoire, by looking for as many pieces as possible with the same indications, and finding the range of tempi in which all of them work. As AH put it, if you have enough data, there is usually very little “leeway”. It is possible to find the correct tempo, if we take the trouble to look hard enough, rather than just inventing our own.


The Colloquium’s halfway point was marked with a musical performance from Domen and Aapo, before each presenter gave a second talk. In this segment there was also discussion between the various presenters, and some questions posted by online listeners were answered.

Jed Wentz again

JW warmed to his theme, emphasising the embodied experience of affektive performance. “The Actor feels the Affket in his body”.

JW looked at Mattheson’s discussion of Affekts. I note that Mattheson, as with Hill and other 18th-century sources, goes beyond and even contradicts the 17th-century categorisation into Four Humours. As JW reported, Mattheson describes Joy as a spreading out of our animal spirits (an outward, sanguine humour – ALK), whereas Sorrow is a contraction (ALK – inward, Melancholy); Love is based on a scattering of the spirits (outward, sanguine – ALK).

JW turned to Mattheson’s analysis of Hope, famously applied to an innocent little Courante. Hope is an elevation of feeling, whereas Desperation a complete collapse of the same [outward, warm sanguine humour, inward cold phlegmatic – ALK]. These Affekts can be very naturally represented with sounds, above all when the other factors, especially Zeitmasse (the amount of Time, a different word for a different shade of meaning of Tempo – ALK) play their part.

Mattheson shows how the Affekt might change to Desire in certain phrases of the Courante, which as JW pointed out, might suggest a pause for transition and/or a different tempo for the new Affekt. JW was also properly cautious with this suggestion, since it contradicts the instructions for dance-music found in many period sources. I would also mention that Mattheson’s switch to Desire implies a gross change to a Choleric Humour, that earlier sources would not regard as consistent with (Sanguine) Hope.

One possible approach that might square these circles is to follow JW’s advice and apply the historical technique for creating Affektive contrasts. Modern-day performers tend to make an intellectual decision to change the tempo, “because there is a change of Affekt”. But the historical practice was to feel the Affekt in the body, and allow changes of timbre, tempo etc to happen as a consequence. I would translate this as “you try to keep steady measure, and you genuinely believe you are doing so; but the changes of Affekt you experience create a change of tempo, as measured by a dispassionate observer (or indeed, a metronome, that most dispassionate observer of all!”

Sources cited by JW are very firm that conventional tactics (e.g. changing tempo) alone will have little emotional effect on the listeners. The essential first step is for the performer to change their own affektive state, and this is what moves the passions of the listeners. My comment is that if this goes well, both performer and audience will feel that the tempo was the constant, it was their affektive state that changed.

JW continued with various citations: one can form an emotional [Sinnliche] idea of all the emotions [Regungen] and form one’s inventions to it – this was directed to composers.

Dealing with Sorrow, much more than with the other emotions, anyone [ALK, this is addressed to composers, but could well be apposite for performers also] who would represent sorrow in sound must feel and experience it himself; otherwise all the so called loci topici [musical clichés] are useless. I would read this as a warning against the kind of Rhetorical Studies that focus on finding and naming those clichés, as if this alone will make the performance more communicative for listeners. Very few courses on Rhetoric spend time teaching students to imagine and feel within the body each of the Four Humours: though I consider this essential fundamental training in Historical Performance.

For example, coaching Continuo-players (on theorbos, lutes, harps etc) to respond to text, I show how obvious cues from the text can be realised with simple changes to instrumental timbre (corresponding to Hill’s “braced” or “slack” muscles!): nearer the bridge/soundboard (more gritty) or further up the string (sweeter); relaxed or tensed fingers etc. But the more significant technique is just to create mental visions of the text as it goes by, as if creating a video-film to the sung text as a script, and allow those mental visions to change the physical aspect of your fingers, so that the sound of the instrument changes as a result. This is hard to specify in technical detail, but has a stronger effect for listeners.

JW cited Diderot Memoires (1748) page 192 as opposing the use of clock-like tempo devices for anything more than a few bars to establish the tempo. After that the player should continue alone: “nothing more than the pleasure of the harmony suspends him”. I’m reminded of Frescobaldi’s advice: if you want to know how a piece of music feels, than just play it (see above).

Domen Marinčič on Tempo words

DM referred to Milan El Maestro (1536) as an early example of tempo words that modify the effect of the musical notation. This is in the context of a particular style of fantasia, that contrasts harmonies in long notes – consonancias – with fast passagework – redobles. As DM mentioned, vihuela sources contain a lot of information on tempo, and Milan gives a specific tempo – in words – for each of his fantasias. More on Milan here. More on the 16th-cent Spanish Art of Time here.

DM cited sources stating that purely mathematical proportions fail to observe decorum, text, or harmony.

Decorum is a technical term of Rhetoric, the requirement that every detail be consistent with the Rhetorical purpose – ALK.

In some English 17th-century sources, ‘soft’ is linked to ‘drag’. In 1619, Praetorius links ‘piano’ to slowly. A 1613 source asks for certain passages to be softer and faster. Türk (1789) asks for certain passages to be softer and slower.

Much more in DM’s published article.

Julia Dokter on Tempo words

JD had a slightly different take on the effect of Tempo words. “Tempo words either reiterate or modify information otherwise communicated” i.e. by changes of time-signature and/or note values.

This idea, that Tempo words might merely reiterate what the musical notation has already told us, is controversial. DM considers that a tempo word that ‘goes the same way’ as a change of notation does not simply reiterate, but rather intensifies the change. I am inclined to agree with DM, as I see this usage going back all the way to Milan 1536, where the wording is unambiguously about changing the Tactus to exaggerate the change in note-values.

It could be interesting to look for examples of a single work with proportional changes, some with modifying tempo-words, others without such words, to see whether proportional changes were always ‘tweaked’, or might sometimes be left plain and ‘mathematical’.

And perhaps this is the moment for me to add that the sung text itself can be full of “tempo-modifying words”. It would indeed lack decorum, to sing ‘Drop, drop slow tears’ in a ‘default tempo’ un-modified by slow, lacrimose affekt. My take on this is the “LY” principle: how do you sing any given text? Take the emotionally significant word and add “LY”. So we sing “Drop, drop slow tears” not necessarily softly, but certainly slowly and tearfully. We sing “Awake sweet love” not necessarily louder, but certainly wakefully, sweetly and lovingly. And so on.

Conventional dynamics, mp, mf, piano and forte are hopelessly gross and unrefined – no wonder they are little used in 17th-century music. But the sung text provides highly specific performance instructions. And – as reported by JW – treatises on the Art of Acting tell us how to put those instructions to work, by applying techniques of Vision and Enargeia (the emotional power of detailed description).

JD emphasised that in passages of Stylus Phantasticus, the Tactus Tempo is drastically slower, whilst note-values are much shorter.

I questioned JD’s assumption that Sytlus Phantasticus should be performed with ‘malleable tempo’. She mentioned Mattheson’s characterisation of the style as full of all kinds of surprises and changes, including temporal effects. But surely – these are notated already. Nobody is proposing to create additional harmonic surprises by treating the notated pitches as ‘malleable’…

And I think this comparison of notated pitch, and notated rhythm, is most useful. There is a 20th-century tendency to treate notated pitches seriously, whereas tempo and rhythm are the performer’s free choice. Rather like Autobahn driving: we respect the one-way signs, but choose our own speed, unless the Authenticity Police are present. 🙂

Seriously, we now understand that the written pitches can be changed historically (history of A) systematically (transposition according to chiavette) or creatively (divisions, according to style rules and historical models). I would suggest that tempo and rhythm are notated to a similar extent, and that any changes a performer introduces should be historical (the changing speed of Tempo Giusto etc over the centuries), and within one repertoire either systematic (as JD, AH and DM all showed) or – if creative – should follow style rules and historical models (as JD and DM are investigating). There is no ‘freedom’ for rhythm, any more than there is for pitch: just a lot of historical information to be understood and applied.

JD’s other argument for ‘malleability’ in Stylus Phantasticus was subjective, and none the worse for that, based on her rich experience of this repertoire. When you play this stuff, some adjustments seem necessary, to make sense of the wierd music. I’m sure she is right. But I suspect that those adjustments can be made within a steady Tactus – there is plenty of space to do this, since the note-values are so very small and the Tactus beat (crotchet, presumably) so very slow. Indeed, with such very slow Tactus, and so much surface activity, one’s perception/control of the Tactus diminishes.

DM noted a sequence of markings adagio – a battuta, which might imply ‘malleability’ in the adagio. I would be inclined to take this literally, that the singer would not beat time with the hand during the adagio, and would start again – for beating time was the standard practice – afterwards. I would link this to the prohibition on beating time in theatrical music (since it distracts from the stage action, and from believing that the onstage character is ‘real’), which is gradually extended to passionate solo songs in general. And it’s also practical – whilst you are singing small note-values and/or affektive ornaments etc, you don’t want to be beating a super-slow Tactus with your hand, it’s physically inconvenient and distracting for everyone.

Alexander Bonus mentioned the boom in sales of pocket-watches in the time of Roger North. There is a far bigger story here of the circa 1800 glorification of machines, musical machines, dolls etc that moved by mechanical means, and the imitation of natural and human movement by machines. The admiration of the semitone mechanism of the late 18th-century pedal harp, harpe organisée is part of this story. Such machines were prized because they successfully imitated the perfection of the Clockwork of the Heavens. This is an uncomfortable topic for the anti-metronome brigade, as is the desire of earlier philosophers to make astronomy more regular than it really is. The wish that planetary orbits be circular blocked scientific advance until Kepler established ellipses beyond doubt, and Newton provided a mathematical model for this.

Just as with the Vibrato debate, we cannot hide behind over-simplistic black-and-white positions. Historical Tempo was both regular and irregular – we have to understand how this worked in each repertoire, and we are unlikely to find a ‘one size fits all’ solution. But just as Jed Wentz made the case for an embodied approach to Affekt, I would suggest that we can only begin to understand Tempo if we embody it as they did back then, with the physical movement of the Tactus Hand. If we try to solve problems only by abstract thinking, we are certainly going to ‘overleap distinctions’…

AB cited Brower (1929) advocating a “metronome in one’s head”. I’m not so appalled by this: but what I want to have in my head is a vision/memory of a Tactus Hand, with visions of the changing, text-based Affekts projected onto it!

Descartes comments on the particular significance of ‘first part of the measure’ were cited. Good stuff, and let’s also keep in mind the influence of French dance, Lully’s down-bow on the down-beat etc on how time felt for his contemporaries. We cannot generalise 18th-century concepts of the hierarchy of the bar back into the early 17th-century, when most music was unbarred anyway.

Nevertheless, we do need to seek an (embodied) understanding of how time felt for musicians of the past. This Colloquium made a valuable contribution to advancing such understanding, and the organisers and contributers should be warmly thanked for their work.

Tactus is a hands-on experience!

Sherlock Holmes and the Wedding Dance: Tactus & Proportions in Monteverdi’s ‘Lasciate i monti’

Sherlock Holmes

“We must look for consistency. Where there is a want of it we must suspect deception.”


In a fine article, Dancing at a Wedding (Early Music magazine 2008 here) Virginia Christy Lamothe offers “some thoughts on performance issues in Monteverdi’s ‘Lasciate i monti’ (Orfeo, 1607) from her particular viewpoint as a musicologist and dancer. In a chapter for the multi-authored book edited by Timothy Watkins, Performance Practice: Issues and Approaches (Ann Arbor, 2009) she applies her expertise to Monteverdi’s Ballo delle Ingrate.

Lamothe examines these staged dances in the light of Roger Bowers’ theories of highly conservative, ‘medieval’ notation of Proportions (Bowers on Proportions in Orfeo here). Her own contribution is to place these dances in the context of period manuals for social dancing, Caroso Il Ballarino (1581) here  and Negri Le Gratie d’Amore (1600) and Nuovi Inventioni di Balli (1604) here.

There is certainly much to be learnt from Lamothe’s approach, to which we should add information specifically about theatrical dancing, in particular the Preface to Cavalieri Anima e Corpo (1600) here  and the anonymous c1630 treatise Il Corago (edited by Fabbri & Pompilio here). Both recommend that a ballo should be choreographed by an expert dancing-master, both mention specific dance-steps, continenze, gagliarda, canario that we see also in choregraphies for social dancing.

Il corago

Theatrical Dances

Cavalieri identifies slow steps such as riverenze, continenze and other passi gravi which are performed by the whole company con gravita. He suggests that singers might dance with instruments in their hands, and reminds us of his famous all-singing, all-dancing finale to the 1589 Florentine Intermedi, the Ballo del Gran Duca. But he distinguishes these slow, walking steps from the kind of ballo saltato (jumping steps) with gagliarda, canario, capriole (capers) etc, that would be performed by dance soloists (either two or four, depending on the size of the stage) without singing.

Listen to a duet version of Cavalieri’s Ballo del Gran Duca with Andrew Lawrence-King & Xavier Diaz Latorre here:


Monteverdi’s Balletto (Orfeo)


Lasciate i Monti scoring


The stage directions in the original print of Monteverdi’s Orfeo describe Lasciate i Monti as a balletto (diminuitive), and indeed it is small-scale compared to Cavalieri’s 1589 spectacular. It is shorter than many choregraphies for social dances. But nevertheless, the ensemble is larger than for Monteverdi’s Tirsi & Clori, designated a ballo.  Here, in Orfeo, we have 5 voices and 13 instruments. I will write about the instrumental ensemble in another post.

This balletto has three sections, in three time-signatures: C, 3/2, and 6/4.

Lasciate C

C: Lasciate i monti


Lasciate 3.2

3/2 Qui miri il Sole


Lasciate 6.4

6/4 Ritornello


Lamothe equates these with three movements commonly encountered in social dances: a opening section of riverenze, continenze and other passi gravi; a central gagliarda; a final salterello (jumping dance). The dotted rhythms of the final section of Lasciate i monti strongly suggest one particularly vigorous dance, the canario.


I agree with these identifications of the first and third sections. But I am not convinced that the central section is a gagliarda. It’s not a problem that the first phrase starts with an upbeat qui | miri il sole – Lamothe’s gagliarda example from Caroso does this.

Caroso gagliarda

Caroso gagliarda


But Monteverdi chose to make every phrase start this way, even when the word-accents suggest something different vos- | tre carole. In contrast, galliard phrases normally start on the first beat of the bar, even if the second note takes the word accent.

My thoughts are winged incipit

Dowland Galliard: My thoughts are winged with hopes, my hopes with love,


As Lamothe writes, “Specific types of dances followed particular rhythmic patterns.”. “Each galliard step requires six counts”, including gettati in aria, throwing yourself into the air, on the fourth count.

But the phrasing of qui miri il sole consistently contradicts the phrasing of the galliard step. If Monteverdi had wanted a galliard here, he could have written it like this.

Qui miri galliard

How ‘Qui miri il Sole’ could have been a galliard


Not everything in 3/2 is a galliard. Many of Negri and Caroso’s social dances include movements in 3/2 that are not galliards. Cavalieri tells us that theatrical galliards are danced without singing.  I conclude that Monteverdi’s sung 3/2 movement, Qui miri il sole, is not a galliard.

Monteverdi’s balletto and Cavalieri’s ballo

As we try to identify the three sections of Monteverdi’s 1607 balletto, there is an excellent match with Cavalieri’s 1600 ballo, the grand finale of Anima e Corpo, This is a larger-scale piece, with six strophes and many singers (Cavalieri asks the whole cast, whether on or off-stage, to sing). But the essential structure is the same: C (sung, major mode); 3/2 (sung, minor mode); jumping dance (major mode, instrumental). The texts of both link the movement of celestial bodies to earthly dancing. Cavalieri’s central movement also begins before the beat, in this case with two upbeats – it too, is not a galliard.

Cavalieri’s galliard comes as the jumping dance at the end of the first strophe, with perfectly regular phrases beginning on the beat, and obvious galliard rhythm. And – just as we read in his Preface – after the second strophe, the final dance is varied: now it’s a canario, with dotted rhythms in 6/4, and beginning (like Monteverdi’s canario) after the downbeat. The even-numbered strophes of Cavalieri’s ballo exactly match Monteverdi’s balletto:

  1. C (sung, major mode, could be danced by singers, riverenze, continenze etc) This section is con gravita.
  2. 3/2 (sung, minor mode, starts with upbeat, could be danced by singers but not with jumping steps, definitely not a galliard)
  3. 6/4 (instrumental, major mode, starts after the downbeat, canario, vigorous dance for the dancing-masters)

Cavalieri’s ballo has six strophes, so this particular sequence of three movements comes three times. Monteverdi’s balletto seems also to come three times: the sequence is two iterations of the balletto , some dialogue, and a final reprise of the balletto.


One of the challenges in reconstructing choreographies of this period is figuring out the precise repeat schemes. Lamothe problematises the repeats of the various sections of Monteverdi’s balletto. The music of the C section is written out twice, with a new text the second time;  the 3/2 has the music written out only once, with two strophes of text underlayed; the 6/4 is written out twice, but when the whole ballo is reprised a few minutes later, the repeat of the final section is indicated by repeat marks.

Whilst period choreographies show many different repeat schemes for different dances, the ABC ABC form usually adopted for modern performances of  Lasciate is not typical of Negri and Caroso’s social dances. However, we’ve seen that Monteverdi’s balletto is closer to Cavalieri’s staged ballo than to social dances.

I agree with Lamothe that the question of repeats is problematic. There is much to be said for her suggestion of ABB*C.   (B* means the B music with a new text). The effect for the listener is logical, since Monteverdi’s music has writtten-out repeats for A and C sections, but not for B. So the listener (unaware of how the music was notated) would hear an effect of AA*BB*CC. (The C section is untexted, of course). If one agrees with this logic, then the reprise of the ballo (a few minutes later) could be taken the same way.

I recommend this solution. But I do not rule out (as Lamothe does) the ABC AB*C …. ABC that we hear in many performances today. Her example of a social dance by Caroso is an arrangement of Cavalieri’s Ballo del Gran Duca (1589). Tellingly, the original theatrical ballo and the social dance arrangement have different repeat schemes. Theatrical and social dances were related, but not identical.

Galliard tempi

Lamothe draws attention to particular features of the Galliard movement in social dances, in particular that it is often specifically marked as gagliarda. She suggests that this implies the tempo of the galliard could be specific to that dance (i.e. not in proportion), or variable. Certainly it is true that the physical demands of this jumping dance require that the dancing master has just the tempo he needs. In The Harp Consort, our standard operating procedure for galliards (when performed as a single item, i.e. not as part of a larger ballo ) is to allow the dancer to set the tempo.

Contrary to what one might at first suppose, a slow galliard is physically more demanding than a fast one. At a slow tempo, the dancer must remain in the air longer – this allows more time for capriole, leg-beats in the air – which inevitably requires a higher jump. So slow galliards are more ornamented (just as with musical ornamentation in madrigali passegiati, where a slower Tactus may be taken). And slower, ornamental galliards with their extreme physicality were particularly for men.

Lamothe cites Yvonne Kendall, a scholar of early dance and music, who “reminds us that … the rhythm associated with a particular dance may not be tied to a particular metre signature.” Monteverdi notates galliard sometimes in 3/2 minims (which I take to be tripla, i.e.   three minims to one tactus at about MM60). This is a fast tempo, and we find it for example in Ballo delle Ingrate, which was danced not by singers, nor by the dancing-masters, but by the aristocratic ladies of the court.

Ballo delle Ingrate gagliarda

Ballo delle Ingrate gagliarda


But in the ballo ‘De la bellezza le dovute lodi’ the galliard is in 3/2 semibreves (which I take to be sesquialtera, i.e. three semibreves to two tactus beats). This is a slow tempo, and is set to the words Ben fallo Alcide il forte referring to that mythical strongman, Hercules.

De la bellezza galliard (bass)

De la bellezza galliard (bass)


The agreement of fast/slow tempi with the physicality of fast/slow galliards and with texts associated with women/men seems to support my interpretation of proportions, in which the note-values (minims or semibreves) give us the information we need, to choose between tripla and sesquialtera proportions.

De la bellezza galliard (cantus 1)

De la bellezza galliard (cantus 1)


By the way,  in the cantus primus part for ‘De la bellezza’, there is no time-signature for that same galliard at Ben fallo Alcide. According to Roger Bowers’ theory, this would be a fatal error, a singer would not be able to interpret the notation. According to my theory, the absence of a time-signature is a minor problem; the time signature doesn’t tell the singer anything they didn’t already know.

"What do you make of it, Holmes?" "It is obviously an attempt to convey secret information."

“What do you make of it, Holmes?” “It is obviously an attempt to convey secret information.” “But what is the use of a cipher message without the cipher?”


Since dance movements in 3/2 may or may not be galliards, I suggest that the marking gagliarda in Caroso and Negri indicates not so much the tempo of the movement but rather the particular rhythmic pattern, the characteristic  ‘groove’ of this specific dance. Lamothe again: “specific types of dances followed particular rhythmic patterns.”

We see the same idea with her example from Caroso’s social dance, where the final jumping dance is specifically identified as a salterello. The fast tempo and dotted rhythms might otherwise have suggested a canario (although there are other rhythmic clues in Caroso’s salterello that would contradict an initial diagnosis of canario). 

Caroso saltarello

Caroso saltarello

Who were the dancers?

We should move away from the conventions of Grand Opera in the Romantic period, away from the modern categories of soloists, chorus and corps de ballet. The first ‘operas’ were performed with small casts, who doubled several roles. Soloists often sang as the chorus, too. Chorus members, even soloists might dance.

Lamothe takes into account Tim Carter’s calculations of a company of 9 or 10 singers, a typical size for early ‘opera’, that could have performed Orfeo. But there is strong argument that the first performance in 1607 involved only 8 singers, SSATTBB plus Francesco Rasi as Orfeo. Either way, the singers for the balletto will be mostly, if not entirely, one-to-a-part.

It is possible that singers might sing one-to-a-part and simultaneously dance. This was done by the three female soloists in the 1589 Ballo del Gran Duca – they even played instruments too! But castrato Giovanni Gualberto Magli, who sang three roles (plus chorus parts) in Orfeo, arrived in Mantua eight days before the performance, knowing only the Prologue, and much concerned about ‘so many notes’ that he had to learn. I find it implausible that he would have learnt a dance too, although period commentators report that he performed well when it came to the show.

Certainly, singers could not have sung Qui miri il Sole one-to-a-part and simultaneously danced a galliard, with its combination of three low jumps and one high jump every two seconds. (Try it for yourself!) At most, they could have done some passi gravi during this section. This supports my argument that this section is not a galliard. 

In Mantua in 1607, female courtiers were not allowed to appear on stage – that is why castrato Magli had to be brought in from Pisa. I assume that this prohibition would apply to dancing just as much as to singing. The courtly ladies who danced the Ballo delle Ingrate in 1608 (that year saw a change of policy, which also allowed La Florinda to sing the title role in Arianna), could not have represented the ninfe who are invited to leave the mountains and dance the balletto in 1607.

The most likely explanation is that the dancers were the court dancing-masters. On the model of social dances, they could have danced all three sections. Or ,on the model of Cavalieri’s theatrical ballo, the singers could have done simple movements (riverenze, continenze etc) in the slow C movement, and other passi gravi  in the sung 3/2 (but not a jumping galliard). The instrumental Ritornello then featured a jumping dance, almost certainly a canario, performed by the (male) court dancing masters. Maybe four, or perhaps just two of them, since the performing space for Orfeo was small, and there are already 13 musicians and 5 singers on stage.

One of the court castrati, Issachino Massarini, was also a dancing-master (though he was elderly by 1607). Lamothe speculates that this singer might have sung the role of Euridice, and also danced the balletto. In this interpretation, Rasi as Orfeo would also have to dance the balletto. I haven’t found anything about Rasi’s dancing skills (he was a singer, theorbo player and composer), but I’m sceptical that he would find dancing a vigorous canario the ideal preparation for his great song, Rosa del Ciel, which comes about half a minute afterwards, even if he was not needed for the chorus parts of the balletto. I also find it hard to believe that old Issachino would have managed the sequence of ballo choruses, galliard dancing, and Euridice’s recitative, followed by more chorus-singing, another galliard and the chorus Vieni Imeneo, all in the space of a few minutes.

Nevertheless, Lamothe’s suggestion reminds us that a male dancing master might have danced this balletto as a female character, just as the castrati singers sang female roles.

Another possibility is that some of the 13 on-stage instrumentalists could have danced and played – more plausibly in the passi gravi of the first two movements, than in the canario. But since Stephen Player of The Harp Consort has often played guitar whilst simultaneously dancing a canario, perhaps there was a dancing-master or two amongst Monteverdi’s players. Perhaps not the bass-violinist, harpsichordists or the harpist, but what about the three chitarrone-players? We do have period images of dancers with lutes and long-necked instruments.

Eighteenth-century Engraving of Commedia dell'arte Actors on Stage


Riciulina and Metzetin



Part of Lamothe’s argument is that Proportions might not apply to dancers. Dancers (and many conductors) are sometimes reluctant to struggle with the intellectual and practical challenges of Proportions, but period philosophy and historical musicology combine to emphasise that they are essential.

Ha, ha, keep time! How sour sweet music is / When time is broke and no proportion kept!

William Shakespeare Richard II (c1595) Act V scene v

To support the argument for inconsistency, Lamothe cites musicologist Jeffrey Kurtzman, who is certainly correct to draw attention to the complexity of evidence from period theorists. Working within the scholastic method, theorists had to reconcile the new practices of the 17th century – Monteverdi’s seconda prattica,  the novelty of continuo-notation and (I argue) a simplified use of proportional notation – with the conservative traditions of previous generations. As George Houle observes in Meter in Music 1600-1800, changing usages were fundamentally incompatible with older notation systems, and the theorists’ attempts to unite the irreconcilable were confused, confusing, and in vain.

I do not see that inconsistent realisations could work, in a period when there was often no rehearsal at all. And I do not agree with Bowers’ position, that musicians of Monteverdi’s generation were still putting into practice highly conservative, ‘medieval’ systems – Bowers’ theories are too complicated to work in real time, when sight-reading. Yes, something of the old ways was conserved in the notation, and period theorists struggled to make sense of this. But the actual Use had moved on, and (as just continuo notation replaced intabulated accompaniments) the practice (as opposed to the theory) of Proportions had been simplified, made more practical.

We can be sure that practising musicians found pragmatic solutions that worked, that provided consistent and unambiguous realisations of proportions for ensembles sight-reading together. For, unlike opera, most 17th-century ensemble performances had very little or no rehearsal. I argue:

Practical musicians must have found workable, consistent solutions.

You can examine my theory of proportions here.

Holmes theories

“The theories which I have expressed there, and which appear to you to be so chimerical, are really extremely practical — so practical that I depend upon them for my bread and cheese.”


In addition to arguing that period musicians realised Proportions consistently, I also reject the suggestion that 17th-century dancers had a different way of thinking. Such a divergence may happen today, but the familiar concept of the Renaissance Man reminds us that historically there was an aesthetic unity across many artistic disciplines, guided by high philosophy, i.e. period Science.

In Mantua, Massarini danced and sang; Rasi composed, sang and played theorbo; Monteverdi sang, composed and played the viol. Elsewhere we know of dancer/swordsmen, swordsman/vihuelists, and many dancing violinists, and continuo-playing singers.

It’s clear that there must have been a consistent approach across all these disciplines. We must therefore synthesise all the available information.

It is a capital mistake to theorize before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts.

“It is a capital mistake to theorize before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts.”


I suggest that the missing links in Lamothe’s and Bowers’ arguments about Lasciate i Monti are the links: the link from the previous speech, the link to the next speech. Their suggested metronome mark of approximately 50/60 as  the Tactus Aequalis is certainly highly plausible. And Bowers agrees that the notation implies the same Tactus for the whole opera:

There thus appears to be every indication of the prevalence of a single tempo for the entirety of these pages, the fundamental durations of the unblackened semibreve, minim and semiminim remaining constant throughout.

Houle goes further, showing  considerable evidence of the same Tactus for the entire repertory, within this period.

But this ‘single tempo for the entirety’ must be minim = approx 60, not semibreve.

The confusion here is one that we find in period sources too, between Tactus (which strictly means the complete down-up movement) and Semi-Tactus (which is the proper name for the down-movement alone, or the up-movement alone). Many sources use the word Tactus for both, which leads to some theoretical confusion.

But in practice, there is no confusion. Just try the recitative Ma tu gentil cantor  that follows the first part of the balletto,  

Minim = 60? Maybe. Semibreve = 60? Impossible!

Minim = 60? Maybe.
Semibreve = 60? Impossible!


or Euridice’s speech Io non diro qual sia before the reprise,

Minim = 60? Maybe. Semibreve = 60? Impossible!

Minim = 60? Maybe.
Semibreve = 60? Impossible!


or the chorus afterwards, Vieni Imeneo.

Minim = 60? Maybe. Semibreve = 60? Impossible!

Minim = 60? Maybe.
Semibreve = 60? Impossible!


All work well at around minim = 60, none could be imagined at semibreve = 60!


Chain Missing Link Question

When considering Proportions, Tactus supplies the missing link


So, according to the well-agreed principle that the whole work has a unified Tactus, around MM 60,  I’m convinced that the C section of  Lasciate goes at approximately minim = 60. This is slower than many modern performances, but corresponds to the period description of this kind of movement as con gravita, with passi gravi.

Combining Tactus & Proportions

And here is perhaps the most powerful tool for the scholar of Proportions. If we combine the period principle of Proportions with the period principle of Tactus, we can eliminate many theoretical options as impossible. Starting with minim ~ 60 for Lasciate i monti (unfamiliar to our modern ears but certainly possible), the big question is what proportional relationship to apply at Qui miri il sole.

Eliminating the impossible I – Sesquialtera

If we follow Bowers’ theory and apply Sesquialtera to Qui miri il Sole, we would have dotted semibreve (three minims) = 30, i.e. minim = 90. I find this too slow for Qui miri, but not necessarily impossible.  (Bowers himself arrives at a faster tempo, by applying Sesquialtera, but starting impossibly fast in the previous C section). But what about the Ritornello after the balletto reprise?

Minim = 90? Impossible!

Minim = 90? Impossible!


Or (killer example) the final Moresca at minim = 90? Impossible!

Moresca from Orfeo

Minim = 90? Impossible!


Eliminating the impossible II – Sextupla

Alternatively, if we apply the very fast (Sextupla) proportion to 3/2 Qui miri il Sole, we would get dotted semibreve (three minims) = 120..  I find this too fast, but not necessarily impossible. But now the 6/4 canario would be dotted minim (three crotchets) = 240. Impossible!

Dotted minim = 240? Impossible!

Dotted minim = 240? Impossible!


What remains?

My practical theory of proportions (here again) indicates Tripla, dotted semibreve = 60 for Qui miri il Sole. (This would work as a fast galliard tempo, but I don’t think this is relevant, since the movement is not a galliard.) Then the canario comes out at dotted minim = 120 (i.e. the same dotted semibreve = 60 as in the previous movement). This (as Latrophe agrees) is a good canario tempo.

As Sherlock Holmes said:

When you have eliminated the impossible, whatever remains, however improbable, must be the truth.

[Sir Arthur Conan Doyle, The Sign of the Four (1890), page 111]

Livanov as Holmes

Holmes never said, “Elementary, my dear Watson”


I find this balletto an interesting and significant test-case, for which my theory of Proportions provides a plausible solution.

But I’m certainly not so arrogant as to insist that anyone else should accept my theory. And I’m just having some fun by quoting Conan Doyle, even if investigating period performance practice can be something like Holmes’ Detective Science. However, I do suggest that you might apply my methods to test your own theory of Proportions. A good theory should be consistent, unambiguous, and easily applied for quick decisions in the real world of practical music-making.

So, as Sherlock Holmes said on the next page of The Sign of the Four:

You know my methods, apply them!


Please join me on Facebook and visit our website . Further details of original sources are on the website, click on “New Priorities in Historically Informed Performance”

Opera, orchestra, vocal & ensemble director and early harpist, Andrew Lawrence-King is director of The Harp Consort and of Il Corago, and Senior Visiting Research Fellow at the Australian Research Council Centre for the History of Emotions.