Semiology VS Proportional Rhythm
Semiology VS Proportional Rhythm
By Christopher Jasper
Introduction: To treatise or not to treatise?
Within the field of Gregorian chant research, the pursuit of "historical accuracy" is an elusive goal and a slippery slope. As I argued in my previous post, absolute certainty regarding the historical performance practice of the ninth through eleventh centuries (and later) is an impossibility; the liturgical context of prayer, virtue and Ecclesiastical laws must always outrank the specificities of any musicological school. Furthermore, while the Holy See has declared an official rhythmic interpretation in the 1908 Vatican Gradual, it has nevertheless allowed for a plurality of other rhythmic interpretations, provided the integrity of the official melody is preserved.
For context, my own approach has been deeply shaped by a lifelong adherence to the semiological method of Dom Eugène Cardine which views the chant through the lens of verbal rhythm and the paleographic nuance of the earliest neumes. I was first trained by my choir director, Jeffrey Morse - himself a friend and pupil of the late Dr Mary Berry of Cambridge - after which I was trained directly at the Abbey of Solesmes under the tutelage of the late Dom Daniel Saulnier, a key figure in the post-Cardine era of Solesmes scholarship and a professor at the Pontifical Institute of Sacred Music. Over the last several years, my understanding of Gregorian Chant has further developed through the friendship of such luminaries as Monsignor Alberto Turco, former professor at the Pontifical Institutes of Sacred Music in Rome and Milan (someone to which even Dom Saulnier was indebted) and a renowned expert on Ambrosian, Old Roman and Gregorian modalities; and Don Giacomo Baroffio, former Dean of the PIMS in Rome, a tireless cataloger of Italian liturgical manuscripts and a central figure in the study of the "Nota Romana". Lastly, I have also lived life as a monk (postulant) in the Benedictine monastery - now abbey - of St Benedict in Norcia, Italy. Having spent my life singing and directing choirs and teaching according to these principles, I enter this examination as someone who is deeply rooted in the established semiological tradition.
When read with a strict and literal interpretation, there is a distinct dissonance between the explicit, mathematical instructions found in Carolingian musical treatises and the fluid, oratorical rhythm of modern semiology (and the Solesmes Method of Dom Mocquereau before it). At the beginning of the chant restoration movement at the Abbey of Solesmes in the mid 19th century, the prevailing consensus was to view the medieval treatises as theoretical abstractions, distinct from the practical ars of the cantor. Yet, many musicologists around the world, both past and present, firmly believe the primary medieval sources indicate that the performance practice of the "Golden Age" (c. 750–1050) was grounded in a strict proportional rhythm characterized by structural ratios between long and short notes.
But even if the proportionalist reconstruction is the most accurate historical option, it does not invalidate the aesthetic and spiritual beauty of the semiological method. Rather, it opens a dialectic gap. We are faced with two competing hermeneutics: one that prioritizes the paleographic nuance and inconsistencies of the manuscripts (Semiology), and one that prioritizes the theoretical descriptions of the era (Proportionalism) and seeks to reduce the great variety of neume shapes to a simple and strict durational value of 2:1, long vs short. The intent of this post is not to declare a victor, but to present [a fraction of] the conflicting evidence objectively, acknowledging that the "thread of tradition" may be far more frayed than either side cares to admit. Finally, I will present my own views in light of all which is presented here.
The Epistemological Conflict: "Metricians" vs. The Scribes
The fundamental divergence in chant scholarship lies in the weight assigned to medieval music treatises versus medieval music notation. As J.B. Goschl pointed out in his 2008 presentation 100 Years of the Vatican Gradual, the early figures of the Solesmes restoration, such as Dom Prosper Guéranger and Dom Paul Jausions, were explicitly skeptical toward medieval theoretical writings, fearing that they were influenced by musica figurata or were "inevitably mensuralist". Consequently, when they encountered treatises describing strict proportions, they dismissed authors such as Hucbald, the Enchiriadis treatises, Guido d’Arezzo, Aribo, et al as "metricians rather than Gregorian scholars," suggesting they forced classical prosody onto a music that was inherently free. Dom Mocquereau famously stated that he based his theory not on the "shifting sands" of the medieval authors—who, he claimed, "do not really know what they are talking about"—but on the "unshakable rock" of the paleographic evidence. Ironically, his own theory is often in outright contradiction to the unshakable rock of the medieval neumes he held so dearly.
This dismissal or skepticism of the medieval treatises is at the crux of the debate. Is it conceivable that the very men of the time who sang, taught and wrote down the chant would teach a doctrine of longs vs shorts in a strict 2:1 ratio while practicing a contradictory "nuanced" verbal rhythm in the choir stalls? To suggest they were ignorant of their own art seems a logical absurdity. On the other hand, is it possible that these theorists were using the only terminology available to them—that of Greek and Latin prosody—to describe a musical phenomenon that was analogous but not identical to a strict, metronomical quantitative meter?
The Case for Proportionality: Omne Melos More Metri
The literary evidence for a measured, proportional rhythm in the ninth and tenth centuries is undeniable and remarkably consistent in its terminology. The Commemoratio brevis de tonis et psalmis modulandis (c. 900), a treatise contemporaneous with the best St. Gall and Laon manuscripts, provides a strong case against an equalist or purely nuanced interpretation. The author explicitly links rhythmic precision to mensura (measure), stating that "every melody should be carefully measured in the manner of metre" (omne melos more metri diligenter mensurandum sit).
The text goes on: "In fact all the longs must be equally long, all the shorts of equal brevity... so that they be neither more nor less, but one always twice as long as the other (semper unum alterum duplo superet)". This instruction implies that "long" and "short" were viewed as fixed, interchangeable units of value—a binary opposition of 2:1—rather than the subtle, flexible values of a solemn speech rhythm.
Support for this metrical view is also found in the Scolica enchiriadis (9th century), where the master instructs the disciple to sing "rhythmically" (numerose canere). When the disciple asks for clarification, the master explains that one must observe where to use longer and shorter durations (morulae), beating the song "in the manner of metrical feet" (veluti metricis pedibus cantilena plaudatur). The treatise even allows for changes in tempo, but specifies that such changes must adhere to a strict mathematical proportion: "halve a long duration or double a shorter one" (duplo id feceris).
Guido d'Arezzo (early 11th century) reinforces this view in his famous treatise Micrologus. He instructs the reader that neumes should correspond to one another "either in the number of notes or in the proportion of their durations (ratione tenorum)," specifying ratios of 2:1 for individual notes and ratios of 3:1, 3:2, and 4:3 for neumes and phrases.
Berno of Reichenau (d. 1048), writing at the twilight of this tradition, reinforces this view. He admonishes singers to pay attention to the neumes to discern where the "proportional shorter duration" (ratae morulae breviores) is to be measured against the longer one, explicitly stating that the chant is composed of a "harmonious combination of short and long sounds" just as verse is constructed of metrical feet.
Aribo Scholasticus (c. 1070) lamented that the care for proportional singing had "died long ago, even buried" (jam dudum obiit, immo sepulta est).
These proportional concepts find deep roots in St. Augustine's six volume treatise De Musica, which states that the consonance of "single to double" (1:2) is "naturally so implanted in us... that not even the ignorant can fail to perceive it".
If one accepts these texts at face value, the conclusion that early chant was mensural (metered/proportional) seems inescapable.
The Counter-Argument: Ambiguity, Nuance, and the Galloping Horse
The proportionalist interpretation assumes that the medieval terminology aligns perfectly with modern mensural concepts. However, these texts are not necessarily so clear-cut and dry to interpret as one may think. A closer reading of the same treatises reveals ambiguities that can indeed support a more nuanced, non-mensural interpretation.
First, consider the musical example provided in the Scolica enchiriadis for the concept of singing “rhythmically” (numerose canere): the antiphon Ego sum via. The text describes the rhythm of this chant not by assigning a value to every note, but by stating: "Only the last [syllable] in [each of] the three phrases is long; the remaining [syllables] are short".
If read literally, this would mean one would (or could) beat out all the short notes with the precision of a metronome and the long notes would be exactly twice as long. But this is not necessarily the case. It is certainly possible that what is intended is that all the ‘short’ notes are sung with relative equality (not the precision of a metronome) while maintaining the accentuation of the text. The long notes at the ends of the phrases would then be sung essentially double in length. This interpretation aligns remarkably well with the "nuanced" rhythm of the Solesmes tradition (specifically the mora ultimae vocis of Pothier and Mocquereau, minus Mocquereau's other oddities), where the structural articulation is achieved by lengthening the final notes of grammatical units.
Furthermore, the famous 15th chapter of Guido d’Arezzo’s Micrologus (c. 1026), often cited by proportionalists for its mention of "metrical feet," also contains imagery that could be interpreted against a strict proportional interpretation. Guido compares the flow of the phrase to a "galloping horse." He writes: "Towards the ends of phrases the notes should always be more widely spaced... like a galloping horse, so that they arrive at the pause, as it were, weary and heavily" (quasi gravi more lassae perveniant). Many modern scholars interpret this description of a gradual slackening (apodosis) as evidence of a flexibility of tempo that is the antithesis of the mechanical "tick-tock" of strict proportionalism. A horse does not slow down by switching gears from a quarter-note to a half-note; it slows organically.
A Rebuttal Regarding the Horse
But, even this interpretation of Guido’s metaphor is not without challenge. As Alasdair Codona has argued, the metaphor of the horse may actually support the proportionalist view if understood through the lens of equestrian mechanics. Codona says that if Guido simply wanted to describe a gradual deceleration (a ritardando), he could have used the example of a man running to a halt. A man has only two legs and his gait remains the same while the duration between steps increases. A horse, however, changes gait. When slowing from a gallop to a trot, the animal switches from a four-beat pattern to a two-beat pattern. Codona suggests this is a "perfect metaphor for a binary change from four short notes (four beats) changing to two long notes (two beats) within the same duration". If Codona’s reading is correct, Guido’s "galloping horse" describes a structural, proportional shift in the metrical foot—affirming the 2:1 ratio of the Scolica—rather than a gradual, unmeasured slowing.
Paleographic Realities: The Inconsistency of Signs
The debate is complicated further by the manuscript evidence. If the rhythm were truly based on a fixed 2:1 ratio, one would expect a high degree of consistency in the manuscript tradition. This consistency does tend to appear from the 13th century onward (the practice of equalism having been established), but for the 9th-11th centuries, comparative analysis often reveals a perplexing landscape.
Dom Eugène Cardine and his successors have demonstrated that the early neumes capture a wealth of agogic nuances that defy rigid quantification. For example, the manuscript Laon 239 uses distinct signs for long and short notes (the virga and uncinus vs. the punctum), which proportionalists cite as evidence in favor of their view. However, semiologists point out the wide variety of intermediate shapes and the inconsistency of their application. In the Introit Si iniquitates, comparative analysis of the oldest sources (Laon 239, Chartres 47, Einsiedeln 121) reveals contradictions. Where one manuscript places a mark of lengthening (e.g. an episema), another might omit it, or use a cursive (short) form, etc. If the episema indicated a strict doubling of duration, such discrepancies would imply that the fundamental rhythm of the chant changed from monastery to monastery—a conclusion that undermines the concept of a unified "Gregorian" tradition (but conversely supports the idea of regional variances). Semiologists argue that these signs represent "nuances of value"—indications of musical weight, phrasing, and textual articulation that cannot be reduced to mathematical ratios.
However, the proportionalist response to this dilemma is also worth considering. Scholars such as Jan Vollaerts have argued that the apparent inconsistencies are often superficial or indicative of a lack of necessity rather than a lack of rhythm. Vollaerts sought to demonstrate that the manuscripts of the "model group" (including Laon 239 and St. Gall 359) actually show a remarkable 95-99% agreement on fundamental long and short values when analyzed comprehensively rather than in isolation. The discrepancies often cited by semiologists may result from the fact that St. Gall scribes used the episema or the letter t as a warning against error in difficult passages, whereas the Laon scribes used structurally distinct neumes (the uncinus vs. the punctum) to denote length. Thus, because the St Gall family of notation is cursive and places a greater emphasis on the melodic contours, the absence of an episema in St. Gall does not prove a note is short if the corresponding Laon manuscript clearly indicates a long. Furthermore, recent comparative analyses suggest that contradictions often arise when comparing manuscripts from different chronological strata; the confusion evident in later 11th-century manuscripts may well reflect the very rhythmic decay Aribo lamented, while the 9th and early 10th-century sources present a greater unified rhythmic tradition.
The Historical Trajectory: Decay or Evolution?
A central point of contention is the historical trajectory of chant rhythm. It is widely acknowledged that by the eleventh and twelfth centuries, the performance of chant had become "equalist" (or cantus planus), where all notes were of roughly equal value. Proportionalists view this as a corruption or "decay," citing Aribo (c. 1070), who laments that "in olden times great care was observed... to compose and sing proportionally. But this idea has already been dead for a long time—even buried" (iam dudum obiit, immo sepulta est).
However, opponents of the strict proportionalist view question the narrative of a "catastrophic loss" of tradition. They argue it is implausible that a universal rhythmic system based on strict 2:1 ratios could vanish so completely that scribes continued to write the pitches with accuracy while suffering a collective amnesia regarding the fundamental rhythm. They suggest that the "rhythmic" signs of the ninth and tenth centuries were perhaps always indicative of subtle nuances—oral traditions of delivery—that naturally evolved or faded without representing a rupture in the essence of the music.
But a plausible historical solution for this loss does exist, countering the "implausible amnesia" argument. The introduction and spread of organum (polyphony) provides a compelling explanation for the equalization of chant rhythm. In Hucbald's 9th century treatise De harmonica institutione, he explicitly states that organum (two part singing) requires a slow tempo (morositas), noting that in such performance "it is hardly possible to maintain proper rhythmic proportion" between long and short notes. As polyphony grew in complexity and importance during the 12th century onward, the cantus firmus (the chant melody) was necessarily slowed down to accommodate the upper voices, eventually fossilizing into the equal-note cantus planus. Furthermore, the shift from adiastematic neumes (more rhythmically precise and varied) to diastematic staff notation and square notes created a technological bottleneck; the new notation perfected pitch transmission at the expense of rhythmic precision. As proportional rhythm advocate Patrick Williams points out, we can see historical proof that rhythmic traditions can indeed decay and equalize as shown in the history of the Genevan Psalter: melodies like the "Old Hundredth" were originally composed with lively, syncopated long and short rhythms, but within a few centuries were universally sung in equal, heavy notes (isometric rhythm) in congregations across Europe. If this could happen to printed music in the modern era, it is certainly plausible for oral tradition in the Middle Ages.
Beyond Pitch and Duration: The Manifestation of Sound
While the fierce debates between proportionalists and semiologists regarding the durational values of the notes are essential, a too narrow focus on rhythm and melody risks reducing the chant to a mere abstraction of time and pitch. If we stop here, ignoring the other elements of vocal music—its timbre, dynamics, tempo, ornaments, and the soul (the Christian faith and prayer of the chant) —we risk reconstructing a skeleton without flesh. The historical sources indicate that the aesthetic of the "Golden Age" was characterized not by the soft, ethereal smoothness often associated with the nineteenth-century revival, but by a vocal production that physically engaged the whole body, was timbrally rich, and ornamented.
The Vox Virilis and the Aesthetic of Intensity
The prevailing modern aesthetic of Gregorian chant, heavily influenced by the Romantic ideal of "sacred" sound as something set apart from the secular through softness and "angelic" purity, faces a stark challenge from medieval descriptions of the voice. The Cistercian reforms of the twelfth century explicitly demanded a vocal quality described as vox viriliter ("a manly voice")... for men, of course. The Instituta generalis capituli admonishes singers not to imitate the "lasciviousness of minstrels" with "feminine" or shrill voices, but to preserve a "seriousness" (gravitatem) through a "manly" sound. St. Bernard of Clairvaux echoes this, calling for a voice that is "round, virile, lively and succinct" (rotunda, virili, viva et succincta voce). If they insist that men sing with a strong and manly voice (presumably what we refer to colloquially as “chest voice” and with some gusto) and to not sing with a “feminine” voice, specifically denouncing the “false voice” (most likely referring to “falsetto”), this would seem to suggest that women in nunneries did sing in their head voice as most women do with chant today. Likewise, following quickly on the heels of the Cistercians, the Dominican order also advocated for a strong, virile voice.
This insistence on virilitas suggests a full-bodied, chest-connected vocal production that contrasts sharply with the sotto voce (soft voice) aesthetic often associated with monastic choirs such as Solesmes and Fontgombault which have had significant influence on choirs all around the globe. The seemingly rougher, more grounded vocal production finds corroboration in the descriptions of the Frankish reception of Roman chant. John the Deacon (c. 875) and Adémar de Chabannes (c. 1025) describe the Franks’ inability to execute the subtle ornamentations of the Roman school; the Franks were described as having "thunderous voices" (vocum suarum tonitruis) and throats that "cracked" rather than projected the delicate inflections of the Romans and sounded like "wheels breaking on the rocks". While this was a critique of the Franks, it reveals that the ideal Roman sound involved a sophisticated manipulation of timbre and articulation that required greater control—a control that the "barbaric" loudness of the Franks could not replicate. The Roman sound was often referred to as the Vox Vinnola ("vined voice") expressing the agility of the voice like vine or a curled lock of hair.
Modern interpreters such as Marcel Pérès have argued that the loss of this "energy in the phonation" and the deliberate dismissal of ornamentation in the 20th century were based on a "sophism" that equated a regular beat and material vocal presence with a lack of spirituality. Pérès suggests that looking to surviving oral traditions (such as Corsican or Eastern rites) might bring us closer to the vox viriliter vocal aesthetic of the early medieval period as an alternative to the refined aesthetic of the 19th-century conservatory.
On a personal note, while I generally do not like what I would call the excessive amount of ornamentation used by Ensemble Organum or Damien Poisblaud's ensemble Les Chantres du Thoronet, I do very much enjoy their vox viriliter.
The Tremula and the "Violent" Ornament
If the vocal quality was more robust than previously imagined, the execution of ornamentation may have been as well. The Solesmes tradition has typically interpreted neumes such as the quilisma and oriscus as light, passing nuances or subtle preparations for the following note. However, the theoretical evidence points toward a much more explicit and perhaps percussive execution.
The Quilisma (often associated with the vox tremula) is described by Aurelian of Réôme (c. 850) as a "tremulous and rising sound". While Solesmes theorists have argued that "tremulous" implies a timid lightness (and therefore more likely to be sung as portamento or glissando), Aribo (c. 1070) explicitly links the tremula to a fluctuation of pitch or intensity, stating that "the same notes are produced with alternating strong and weak pulses of the voice, as if trembling" (quasi tremendo). This supports the view that the quilisma functioned as something a trill or a distinct vocal undulation rather than a mere portamento. Even the instructions of the 1908 Vatican Gradual (the current “official” rhythm) give the direction to sing the quilisma in this manner.
A similar description is given for the oriscus. Walter Odington (c. 1300) describes the Pes Quassus (a version of the pes/podatus which begins with an oriscus) as a "turbulent note sung with a trembling voice and much shaking" (voce tremula et multum mota), explicitly defining the motion as "violent" (violentus motus). This description is difficult to reconcile with a "smooth" or "ethereal" performance style. It suggests that these neumes represented specific, agile vocal maneuvers—likely microtonal inflections or glottal agitations—that imparted a physical energy to the melodic line. Other descriptions seem to imply that the oriscus was like a mordent: an ornament involving a rapid alternation between the principal note and an adjacent note—either above or below—before returning to the principal note. There are also certain letters in certain MSS which stand for words like fragore (crashing/breaking). One can only imagine what effect that was to intend!
Furthermore, the thirteenth-century treatise of the Dominican priest Jerome of Moravia provides a detailed taxonomy of ornaments, including the flos harmonicus (a rapid vibration or trill) and the reverberatio, which implies a repercussive or striking quality (two centuries earlier, Aribo described singing the bi/tristrophae as a repercussion of sounds, "like the swift beating of a hand"). But because Jerome and Odington wrote nearly 300 years after the original rhythm began to decline, their descriptions of certain specific vocal ornamentations could be interpreted as novelties, sprung from the nascent figured music and polyphony then developing. Others may view their descriptions as a continuity of a vocal technique that was likely even more prevalent in the oral tradition of the Golden Age before the full fossilization of the repertoire into equalist notation.
And while the Dominican chant tradition is a beautiful one, it nevertheless has a separate and distinct performance practice from the “Gregorian” tradition (as do the Cistercian, Sarum, etc) which came before it during the Golden Age. Even the Dominican and Cistercian traditions are not entirely free from musicological debates of performance practice.
Synthesis
Therefore, a reconstruction of the "authentic" performance practice (as much as is possible) must go beyond the simple debate of proportional vs. free rhythm and melodic restitutions. We must also consider the possibility that the chant of the ninth and tenth centuries was characterized by a vox virilis—a resonant, projective voice capable of "violent" ornamentation (violentus motus), a vox vinnola - a supple and agile voice not necessarily at odds with a vox virilis; distinct repercussions and, possibly, measurable rhythmic proportions. If we strip the chant of these corporeal elements in favor of a disembodied "spiritual" smoothness, we may well be creating a modern aesthetic object that, while beautiful in its own right, fails to capture the vigorous reality of the medieval cantor. The ancient desert fathers and the entire monastic tradition have always held a strong affinity to praying not just with the mind and voice, but with the whole body.
Conclusion
Where does this leave the modern interpreter? We are presented with a proportionalist theory that aligns perfectly with the literal reading of the treatises and accepts more ornamentations but struggles to account for paleographic variety and inconsistencies. On the other hand, we have a semiological hypothesis that respects the graphic diversity of the manuscripts but tends to lack many of the older ornamentations and requires us to read the explicit mathematical instructions of the theorists as mere metaphors or borrowings from grammar.
If the medieval performance practice was indeed proportional, then Cardine’s Semiology is necessarily flawed because it is based on the premise of rhythm derived from the text. If, however, the treatises were pedagogical abstractions relying on classical vocabulary to describe a fluid practice, then the strict reconstruction of 2:1 ratios is a modern understanding and imposition of mathematical order on a practice that was understood more freely. (As an extra note, there is also an additional, in-between theory often referred to as the “Semi-Proportionalist” advocated by figures like Dirk Van Kampen)
In my own opinion, I currently believe that the medieval treatises and the oratorical rhythm of semiology can be understood in harmony with each other. If I were to describe what, in my mind, exemplifies the combination of these two and reflects what seems more likely to be 9th century performance practice- since I have not yet made any recordings reflecting this practice, though I do intend to - I might describe it as a combination of the following (generally speaking):
A subtle, general evenness of the flow of the text which would agree with the “nuanced” interpretation of shorts as short and longs as long. This would not necessarily detract from well observed accentuation and phrasing of the text.
Semiology of Dom Cardine and Modality of Dom Jean Claire/Alberto Turco (with some modifications). In other words, greater attention to which notes are long and which are not, which notes are structural vs ornamental and hence which are to have greater emphasis, etc.
Virile vocal qualities (for men), something along the lines of Ensemble Organum or Les Chantres du Thoronet (not their rhythm or excessive ornamentation).
Introit: Resurrrexi Ensemble Organum
Vir erat - Chantres du Thoronet - Grégorien
For women, I very much like the sound of Les Peregrines and the nuns of Le Barroux.
Agnus Dei 9 (IX) - fêtes et solennités mariales
Ornamentation: treating the Quilisma (tremula) as a quick shaking note, the oriscus as a mordent and bi/trivirgae and bi/tristrophae with repercussions. As regards repercussions (and liquescents), I have been observing those for over 20 years already. As a possible example of the quilisma being sung in the manner already mentioned, below is a recording of St Hildegard of Bingen’s O Frondens Virga sung by Chanticleer. You can hear the quilisma at timestamps 0:38, 0:44, 0:56, 1:06 and throughout.
CHANTICLEER: O Frondens Virga by Hildegard von Bingen (1098-1179)
Ultimately, however, we must remember that "historical accuracy"—while a noble intellectual pursuit—is not the primary end of the liturgy. We will never know with certainty how a monk of St. Gall in the year 900 (or 1400) sang the Introit. What matters most is that the chant serves its function as prayer. As I stated at the outset, technical precision must always be subordinated to prayer, humility, charity, and the laws of the Church. Whether one sings with the free rhythm of Semiology or the measured tread of the proportionalists, if the singing does not arise from a humble heart and lead to the glorification of God, it is but "sounding brass or a tinkling cymbal" (1 Cor 13:1). I will continue to teach the oratorical free rhythm and semiology of Dom Cardine (with additional notes like the ornamentations and vox viriliter), as it provides a unified language for the Church of today and deeply respects the primacy of the text. However, I will also continue my studies of the medieval treatises as well as the proportional interpretations of Vollaerts, Murray, Van Biezen, etc. out of personal interest.
No matter which interpretation we ascribe to, may we all do so with humility, charity, respect for others and the greater glory of God. Amen.
