Proposal for a method of segmentation for the analysis of the lighter forms between the sixteenth and seventeenth centuries

Index
1. Aims and characteristics of the method
2. Segmentation of the poetic text
3. First level of segmentation of the musical text
4 Second level of segmentation of the musical text
5. An example of comparative analysis
6. The stylistic evolution of the Mantuan canzonetta
7. Conclusion
8. References

1. Aims and characteristics of the method
The relationship between a poetic text and its musical setting can be examined from many points of view. One can investigate, for example, the rhetorical-musical procedures intended to illustrate the images within the text, as has often been done in analysing the madrigal. Or one might examine the relation between the respective metric and rhythmic structures, in order to see how far the music follows the scansion of the text, or, conversely, the degree of its ‘barbaric’ approach. One might also verify the melodic distribution of the poetic text, or in other words, following a recent illuminating proposal, verify “how much ‘poetic material’ is used … to set a melodic phrase” [Dalmonte-Privitera 1996, 25]. And one can, naturally, imagine many other different approaches, all equally valid according to the aim of the investigation. The methodological approach proposed here is conceived expressly for the repertory of lighter forms (canzonettas and the like) which were produced between the end of the sixteenth and the beginning of the seventeenth centuries; it has as its aim the evaluation of the rapport between the metric and rhythmic arrangement of the poetic text and the mechanisms of its setting.
The first problem to be faced is that of establishing a correct segmentation, which presents, to my mind, a problem of ‘circularity’. It seems obvious that the preliminary notion to consider in establishing a correct segmentation would be that of the cadence, a notion which can be derived from the contemporary theorists, above all Zarlino [1588; ed. facs. 1965, part III]. But even a superifical examination of the repertory shows clearly that the exclusive identifying of points of articulation based on that concept contrasts openly with the competence of any listener: there are in fact many cases in which a sense of caesura is felt without the presence of a resolution (sixth-octave and similar) essential to the compositional definition of the cadence. On the other hand, the contrary problem also exists: that is to say, in some cases the compositional mechanisms of the cadence are found in places in which the text does not allow us in any way to identify a caesura (example 1, respectively a) “ri-den” and b) “te pen-“).

example 1a - example 1b

Given the outlined aims, we will try to find a method which allows us to segment the entire polyphonic workings of a canzonetta (and for now we will limit this to the repertory for three voices). Naturally, the methodology for identifying the criteria of segmentation takes into account the various results of research carried out in this field; keeping in mind, in any case, that these researches are directed for the greater part towards the analysis of melodic procedures, an analysis which, at least in this phase, goes beyond the scope of the present work. Recently, for example, two
scholars have proposed a group of six criteria for segmentation which have proved excellent instruments for the analysis of the distribution of the text in melodic phrases within the ambit of the canzonettas for six voices by Orazio Vecchi [Dalmonte-Privitera 1996, 25-33]; those criteria are based on the concept of repetition and on some elements of a purely musical nature (imitation, pauses, long notes, cadences).
Our aim, in any case, is different: we are trying to understand where the points of articulation of the polyphonic ensemble can be placed: that is, the points in correspondence with which it is possible to affirm that the group of participating voices has finished intoning a significant portion of the text. It’s therefore necessary to define specific criteria for segmentation: we will begin with those concerning the poetic text, while the segmentation of the music will require two distinct levels.


2. Segmentation of the poetic text
It is known that, following the definition of the poetic function offered by Roman Jakobson in 1958 [1966, 181-218], there have been numerous attempts to define metric theory following the principles of Chomskian linguistics. These attempts have been the subject of a number of criticisms that we can define as of empirical nature; in any case, the distinction between an abstract metric model and its concrete realisation is still accepted by many scholars today, by virtue of its capacity to explain some characteristics of versification. Particularly useful for our intent, is the division of the text proposed by Beltrami into three parts: model, realisation and execution, where by model is meant ‘a group of explicit rules’, by realisation, ‘the text, correct with regard to the model to which it refers’, and by execution ‘an individual act which can be repeated in a different way each time, taking its point of departure from the text and chosing to present each time just one of the possible forms’. [Beltrami 1994, § 26].
In the case of the musical setting of a text which is structured metrically, the realisation will thus be constituted by the text itself, in the form in which it is received in its preliminary state by the musician, while we can consider its intonation as the execution. There are obviously a number of ways of approaching analytically an execution: in particular, in the case of a polyphonic setting one can examine the intonation formulas of the verses made up from the single melodic lines, or one can verify the ways of distributing the poetic text within the range of the entire polyphonic texture. We will now examine this second approach.
To perceive the relation between the division of a text into metric units and its breakdown into musical units, we need to define the concept of textual segment : we can call a textual segment (T.S.) every unit of a poetic text which can be obtained through its decomposition into lines and hemistichs following the rules which make up the model. A textual segment will coincide with a line whenever the model does not allow a further decomposition of the same line; or a hemistich whenever that decomposition is theoretically possible.
The sample used for defining the method outlined here (canzonettas for three voices from the Mantuan circle of composers between 1589 and 1607) involves exclusively lines of eleven, seven or five syllables. The five-syllable line needs no particular consideration: it always gives rise to a single textual segment. For lines of eleven and seven syllables, we need to examine them with regard to the existence of internal caesura.
The standard eleven-syllable line can always be broken down into two hemistichs, determined by the caesura that follows either after an accent on the fourth (endecasillabo a minore) or on the sixth syllable (endecasillablo a maiore). With regard to the segmentation following the metric model, the fact that the caesura takes place after a shortened word (tronca) or a monosyllable (allowing a setting in two hemistichs, responding exactly to the requirements of the same model), rather than within a word which is piana or sdrucciola (imposing a caesura which is enjambante or “italiana”) [Beltrami 1994, § 59-60], is of little importance. Naturally, in the second of these cases the setting will tend to favour the limit of the word, avoiding the caesura that is required by the metric norm; but even this is of little importance, since, in segmenting a poetic text, one simply sees how far it follows the model, inasmuch as it is a “realisation” of itself. In the case of a musical setting of a line, it is clear that a metrical caesura will not be made at the expense of the integrity of a word. However, it is convenient to retain, even in this case, the possibility of identifying textual segments. Thus, when we speak about the final syllable of a textual segment, we mean the last syllable of the word after which, or within which, the metric caesura falls. In the case of a verse like the following, for example:

consumar la vedre- || te a poco a poco [Gastoldi 1592, Torna dolce il mio amore]

the segmentation gives this result:

1° consumar la vedrete
2° a poco a poco.

Naturally, it will be the analysis of the music which will tell us whether the composer has respected the subdivision of the line (writing out the synaloepha vedrete a and setting twelve syllables) or not; in this specific case Gastoldi sets twelve syllables in one voice only, and all considered one can state that the two textual segments are set, in this case, to a single polyphonic unit:

example 2

With regard to our sample, it can be said that there is little difficulty in segmenting the text following the standard model of the eleven-syllable line, not finding particular deviations from the three classical types, with accents on, respectively, the sixth (6-10), the fourth and eighth syllables (4-8-10), and on the fourth and seventh syllables (4-7-10, rare from Petrarch on, and present only sporadically in our sample). Some examples follow:

6-10
che tacend' ed aman- || do giung' a morte [Rossi 1589, Io mi sento morire];
Io mi vivea com' a- || quila mirando [Viadana 1594, Io mi vivea com’aquila];
e contro mi fia 'l ciel || ed ogni stella [Franzoni 1605, “Poss’io prima morire”].

4-8-10
deh torna pur || e rendi l'alm' e 'l core [Rossi 1589, Torna dolce il mio amore];
e da' tuoi lu- || mi vibra strali ognora [Gastoldi 1592, Per mirar lo splendore].

4-7-10
eccomi, toc- || cami, stringimi, baciami [Gastoldi 1592, Poiché o mio fido amante];
Credete for- || se di farmi morire [Gastoldi 1595, Credete forse di farmi morire].

Whichever of the three might occur, following the model the eleven-syllable line will always give rise to two textual segments.
Regarding the seven-syllable line, whenever it is possible to find an accent on the fourth syllable (and this happens quite frequently), we decide to break it down into two textual segments, for the reason that one creates a unit, in the first measure, identical to the first hemistich of an eleven-syllable a minore:

Occhi riden- || ti e vaghi
d'ogni mio mal || presaghi (ibid.). [Viadana 1594, Occhi ridenti e vaghi].

In the other cases, not being able to identify any caesura able to refer in any way to the metric model, the seven-syllable line will give rise to a single textual segment:

Tanto è bella costei [Gastoldi 1592, Tanto è bella costei].

Finally, it is necessary to emphasise that it is impossible to avoid a minimum of arbitrariness in the segmentation of lines, and that reflects always on the segmentation of the musical text: for our purpose, however, we maintain that following the norms outlined in this paper will minimalise the variants and will not compromise the results of the analysis.


3. First level of segmentation of the musical text
We talk about a first level of segmentation of the musical text to indicate that this action is not intended to break down a composition into musical phrases, but only to see how the poetic text is distributed throughout the polyphonic ensemble, considering how the text itself is broken into significant portions. These portions can be made up either of textual segments (that is, lines and hemistichs, following the definition given in the previous section), or of words or groups of words that the composer has in some way isolated in the course of the setting; or again, it can involve portions of text which are larger than a textual segment. It is evident that a composer can follow one of three ways of proceeding: he can distribute the text in such a way as to create a reciprocal correspondence between textual segments and musical setting (example 3); he can construct a musical discourse which comprises a larger number of elements than that of the textual segments (example 4); or he can encompass in a single musical phrase more than one textual segment (example 5). This last possibility will be particularly frequent in those cases in which is set an eleven-syllable line with a caesura enjambante followed by synaloepha, whenever the composer does not effectively intend to set twelve syllables. With this present section we propose to offer criteria which will allow us to define these possibilities in strict terms.

examples 3 - 4 - 5

The preliminary condition which will allow us to affirm that a polyphonic composition sets a textual segment, or whatever portion of text, is that the concluding syllable of that portion is given simultaneously in the various voices (as we will see, not necessarily in all of them). It is therefore necessary to define strictly the concept of simultaneousness. Simultaneousness can be considered present in one of the following cases:
1. When the active voices effectively pronounce the same syllable at the same moment (example 6, t’amo e tuoi).
2. When one or more voices pronounce a syllable during the prolongation of the same syllable by one or other voices (example 7, ben).
3. When one or more voices are inactive (or have a pause) and one or more voices pronounce the same syllable that the inactive voices have pronounced before the pause (example 8, morirò).

examples 6 - 7 - 8

It is important to underline the fact that simultaneousness, as such, is not a reciprocal relationship: a syllable can be pronounced in one voice simultaneously with the pronunciation of a number of syllables by another voice. Furthermore, it is necessary to emphasise that the final syllable of a word must be considered not pronounced when it is in synaloepha with the following syllable.
That said, we can now give the criteria for segmentation which will allow us to outline the portions of musical text that we will call polyphonic segments. We do not aim to give here criteria which are valid for the whole of the polyphonic repertory; the validity of these criteria is only relevant, for the reasons given previously, to the canzonettas for three voices which make up the sample chosen for the purpose of this enquiry.

CRITERIA 1.
A polyphonic segment concludes when the condition of simultaneousness is met in all voices for the last syllable of a textual segment, or when it is met in at least two voices for the final syllable of a line (example 6, tuoi, final syllable of hemistich; example 9, ancora, final syllable of a line: every exposition of the line by the two upper voices is a polyphonic segment).

example 9

CRITERIA 2.
A segment concludes when the condition of simultaneousness is met in all voices for the last syllable of a word, if the length of the same word, from its last stressed syllable (if the word is piana or sdrucciola) or from its last unstressed syllable (if the word is tronca), is, at least in one voice, larger than half a measure if the measure is binary, and of the same duration or larger than a measure if the measure is ternary. If the word is monosyllabic it is to be considered the end of a shortened word (tronca) and one counts it together with the unstressed syllable immediately preceding. The monosyllabic opening of a line is instead an autonomous word, and must on its own meet the conditions of duration [example 10a, empia, 10b, quando e pietà, 10c, dunque, but not fuggir, 10d, crudel; 10e, fai; 10f, se: the entrance of Canto I marks the end of the first polyphonic segment].

examples 10 a-b, 10 c, 10 d-e, 10 f

CRITERIA 3.
A segment concludes when the condition of simultaneousness is met in all voices for the last syllable of a word if, when the same syllable is pronounced, a pause follows in all the voices involved.

These criteria have been chosen above all to guarantee a segmentation which will not contrast with the perception of the listener. The first criteria in itself does not require particular comment. A clarification is useful, however, with regard to the conclusion of the line, for which simultaneousness in two voices is considered sufficient: this precaution has been introduced to avoid considering as a single segment the successive setting of two lines, when, at the conclusion of the first, one voice is given as “sfasata” or out of time. And this is just because generally such sfasature do not make themselves felt as a true lack of conclusion (one can see what would happen, without this precaution, in example 11, face). In particular, with regard to example 9, already cited, we can see that, like other similar examples, it is characterised by the homorhythmic proceeding in short note values of the two upper voices, which brings about the triple execution of the whole line while the lower voice intones the line once only, in long note values. In this case, the tendency to neutralise the lower voice is evident, since it becomes the harmonic support (a kind of basso continuo) of the two sopranos. This seems to confirm the necessity of leaving the third voice out of consideration when making the segmentation, at least in correspondence with the end of the line.
There remains something to be said with regard to the second criteria (for the third, see note 12), that we can call “criteria of duration”. Its role is to allow for the identification of procedures of isolation, for rhetorical purposes, of portions of the text which do not coincide with textual segments, including single words. The examples of this given below demonstrate sufficiently, it seems to me, its raison d’être. With regard to the monosyllable, it is clear that is carries out, as a rule, the role of a final syllable, as long as it is preceded by other syllables belonging to the same line.

example 11

Before applying the criteria outlined here to the analysis of pieces, a further explanation is necessary. In the case of polystrophic canzonettas, the analysis is carried out on the setting of the first strophe; it is known that the application of the same musical formula to a series of strophes will always necessitate adaptation. In defining the method we have not been concerned with this problem, since, above all in the presence of lines of different length, “the composer … concerns himself with the full realisation of the rhythm of the first strophe alone, and in consequence pays less attention to the rhythm of the others” [Assenza 1997, 142].


4 Second level of segmentation of the musical text
The criteria outlined so far allow us, as we will see, to verify the distribution of the poetic text within the range of the polyphonic complex of single works; they are necessary for the decomposition of the works into musical phrases, and yet they are not sufficient. To be able to speak of a musical phrase for a polyphonic segment it must be concluded by a musical proceeding responding to the contrapuntal criteria that define the concept of cadence. More generally speaking, we can say that the segmentation of a musical work becomes pertinent only if it presupposes the acquisition of procedures which, within the ambit of the repertory to which the work belongs, are perceived as equally conclusive “signs”. In our specific case, it is useful to have recourse to the cadential formulas in two voices, that we will from now on call clausulas, on the basis of the definitions proposed by Zarlino in the third part of the Istituzioni harmoniche.
The most important clausula is that in which the two voices proceed from an interval of a major sixth to the interval of an octave (or from a minor third to a unison). This can happen either by ascending semitone and falling tone, or by descending semitone and rising tone (examples 12 a and b):

example 12

The second case, as is known, is the so-called Phrygian cadence. For our aims, there need not be other determining factors for the Phrygian cadence, given its relative sporadicity. In the first case, however, it is necessary to distinguish between a “simple” clausula and a “diminished” clausula, according to whether the sound constituting the sixth (or the lower third) is introduced together with the other, or whether it is preceded by a delay (example 12 c).
But Zarlino also defines another form of clausula:

Besides these kinds of cadence, there is another which may finish on an octave or unison; the which one makes, when one places the second figures of the lower part and that of the higher part distant from each other by a ditone, making the lower part descend by the jump of a fifth, or ascend by a fourth, and the higher part ascend by a degree … [Zarlino 1965, 251; example 12 d]

This type of cadence is not frequent in two part counterpoint : «such ascending and descending leaps are more suitable to the lowest voice of a composition for more voices…» [Zarlino 1965, 252]
In our case, it is evident that the model in question constitutes the most frequent cadential formula. Since we are dealing with compositions for three voices we will not make any distinction between the model 6th-8th (or its inversion 3rd-unison) and the model 3rd-8th, but limit ourselves to distinguishing the cases of the simple clausula from those of the diminished clausula (that can appear either with the suspension 7-6 or that of the suspension 4-3). The leap of an ascending fourth or descending fifth of the lower part corresponds to the clausula basizans examined by Meyer [1988, 92-94]. In our repertory for three voices it can be accompanied – or not – by the clausula tenorizans (the movement descending by conjunct degree of the Zarlinian clausula for two voices seen previously), but this results as bearing little weight in the conclusive “force” of the cadence.
To these models it is necessary to add other kinds of clausula, which together constitute those that Zarlino defined as “imperfect” cadences, or that belong to those procedures that he describes as “evading the cadence”. Amongst these various models, it would seem that the selection investigated here requires a specific definition for the formulas that make up the inversion of the 3rd-8th and the 6th-8th progressions, or the 8th-3rd and 8th-6th progressions, in which the third or the sixth have necessarily to be major: we will define these as suspended clausula (example 12 e).
All the other kinds of clausula will be classed together as weak. For their identification, the following is valid: in the lack of formula corresponding to the simple, diminished, Phrygian and suspended clausula, we will in any case acknowledge the end of a phrase when the conditions of duration outlined in criteria 2 of the first level of segmentation of the musical text are satisfied in at least two voices. I will summarise. A musical phrase concludes when, in correspondence with the conclusion of a polyphonic segment, one finds a simple, or diminished, or Phrygian or suspended clausula; or when, always at the conclusion of a polyphonic segment, the conditions of duration posed by criteria 2 of the first level of segmentation of the musical text are satisfied in at least two voices; these conditions are also necessary in correspondence with the conclusion of a textual segment (or when criteria 1 of the first level of segmentation is satisfied). This formulation of the criteria of second level of segmentation is necessary, once again, to prevent the results of analysis contrasting with commonsense. Even in the absence of principal clausula, it is clear that the convergence of at least two voices at the conclusion of a polyphonic segment is perceived as the sign of a conclusion of phrase, as long as its duration is sufficiently ample.
Let us examine some examples:
Example 9 presents the case of a single musical phrase, articulated in three polyphonic segments (criteria 1 of first level), and ending with a simple clausula. Example 6 proposes an example of diminished clausula (t’amo) and another of a Phrygian clausula (occhi tuoi, where the simultaneousness of the three voices at the end of a hemistich had already indicated at first level the end of a polyphonic segment). Example 10 a) presents a case of weak clausula where, independently of the fact that one is at the end of a textual segment or not, it is necessary to apply in any case a criteria of duration to identify also a conclusion of phrase. It will be evident that in this case Zarlino’s definition is fully satisfied: the words, in fact, have not “finished perfectly their sentence”. Example 13, lastly, shows a case of suspended clausula.

example 13

To reassume: in order to be able to speak of musical phrases it is necessary, but not sufficient, that a polyphonic segment be identified as a preliminary condition. To define which of the polyphonic segments constitute musical phrases we need to proceed to segmentation at second level.


5. An example of comparative analysis
We propose now the comparative analysis of two works taken from the Mantuan repertory, example 14 being by Salamone Rossi and example 15 by Lodovico Viadana. These two compositions have a unique common characteristic: both are laid out following the ternary form AA B CC.

example 14

Beginning with the work by Rossi, the segmentation of the poetic text (remembering that we are considering only the first strophe) gives the following pattern (the symbols H and L indicate, respectlively, the conclusion of a hemistich and an entire line):

Vattene pur da me
H
cruda lontano
L
che nel mio pett’ ognor
H
ti serbo e godo
L
anzi ch’il dio
H
d’Amore
L
vuol che fuggendo
H
mi nodrisch’ il core
L

The application of the criteria of first level of segmentation of the musical text allows us to identify the articulation of the polyphonic segments at the following points: measure 8 (simultaneousness at the end of the line); measures 12 (idem), 15 (idem); 18 (idem) and 23 (idem), end of piece. Passing to the second level of segmentation, we can see that, in this specific case, every polyphonic segment identified at first level also constitutes a musical phrase: each of them, in fact, conclude with a diminished clausula.

example 15 (attenzione sono 2 pagine)

Before drawing conclusions from the above, let us compare this with what happens in the work by Viadana, observing the segmentation of the text:

O quante volte
H
o quante
L
esser bramai
H
di bella donn’ amante
L
pensand’ ov’ è
E
beltà
L
di ritrovar
H
pietà
L
o come vaneggiai
L
donna bella crudel
H
fu sempre mai
L

The results of the musical segmentation at first level show: measure 2 (simultaneousness at the end of the line; we should remember that the final syllable –te of the first hemistich is not considered pronounced by reason of the synaloepha); meas. 3 bramai (end of hemistich); 7 (end of line). Given the rigorous homorhythm, the four textual segments that follow (meas. 7-10, pensando-pietà) constitute the same number of polyphonic segments; meas. 12 (end of line), 13 (idem); 14 (idem); 15 (bella (by reason of criteria 2 first level); 16, crudel (end of hemistich); 17 (simultaneousness of two voices out of three at the end of line), and 19, conclusion. In all, we can number 14 polyphonic segments.
Passing now to the second level of segmentation, we can immediately observe that the musical phrases are less in number than the polyphonic segments. In fact, it is possible to identify the conclusion of phrases only at the following points: measures 2 (criteria of duration); 3 (suspended clausula); 7 (diminished clausula, end of section); 10 (Phrygian clausula, end of section); 14 (criteria of duration); 15 (idem); 16 (simple clausula); and 19 (diminished clausula, end of piece). We have in all eight musical phrases.
A comparative analysis of these two works brings to light some interesting facts. In the first place, we should emphasise the different relationship between textual segments and polyphonic segments which can be seen in these two works. In the canzonetta by Rossi, in comparison to the eight textual segments we have only five polyphonic segments: the relation between T.S./P.S. is therefore greater than one. In other words, the composer reveals in this piece a tendency towards continuous composition: contrapuntal writing takes over large sections of the text which number more than one segment. To the contrary, in the work by Viadana, for the most part homorhythmic, the relation of T.S./P.S. is inferior to one (11/14); this indicates a tendency to adapt the musical language to the segmentation of the text, and, indeed, through the isolation of single words and their repetition, to the multiplication of the polyphonic segments. Even more interesting is the relation between T.S./P.S. if considered within the single formal sections. While Rossi maintains constantly this relation in favour of textual segments (in every section, always one more than the polyphonic segments), Viadana has a more varied approach (A: 4/3; B: 4/4, C: 3/7). The proliferation of the polyphonic segments in the third section, in any case, carries the effect of an artificial rhetoric entirely extraneous to the courtly language of the work by Rossi. The difference of style, on the other hand, is obvious, given the neutralisation of the role of the bass voice which one finds in the measures 16-18 of the canzonetta by Viadana, and which the formulation of the first criteria of the first level has allowed us to underline.
The two works in consideration also show important differences with regard to the use of repetition. If we consider his use of individual voices, we can see that Rossi is not sparing in his use of repetition: on the level of the polyphonic ensemble, however, these repetitions do not have a “segmenting” effect, contrary to that which happens in the work by Viadana.
Finally, the relation between polyphonic segments and musical phrases seems significant. If there is a total correspondence in the piece by Rossi, in that of Viadana we can observe a desire to articulate clearly the polyphonic intonation of the lines, even leaving aside the conclusive signals given by the cadences.
The comparative anaylsis given here demonstrates clearly the results of the investigation carried out during the entire survey. We will summarise.


6. The stylistic evolution of the Mantuan canzonetta
The first relevant fact that can be derived from the survey is that usually the number of polyphonic segments is greater than that of the textual segments: this is a confermation of the “scandito” or declaimed style of the canzonetta which was indeed only to be expected. With regard to this we can see, however, a significant division between the publications which belong to the first phase of the time-span under consideration (Rossi 1589 and Gastoldi 1592), and the later collections. In Rossi, the works which show a negative difference between the P.S. and the T.S. in at least one macroform section amount to thirteen, equal to about 75% of the entire collection, and in ten cases (about 58%) the overall relation between T.S. and P.S. is greater than one. In Gastoldi’s first collection the cases of negative difference between P.S. and T.S. in at least one macroform section are fifteen, and constitute, again, 75% of the examples analysed, while the pieces in which the comprehensive relation between T.S. and P.S. is greater than one amount to thirteen (65%). In Gastoldi’s second collection there are only five cases in which the relation between T.S. and P.S. is greater than one (about 25% of the examples analysed); but the pieces in which the difference between P.S. and T.S. is negative, at least in one macroform section, are thirteen (about 75% of the nineteen pieces analysed).
If we now consider Viadana’s collection and that of Franzoni, we can see that they contain almost exclusively pieces in which the relation between T.S. and P.S. is less than one. The presence of works which show a negative difference of P.S. –T.S. in at least one macroform section is also definitely less notable: seven cases in Viadana (39% of the eighteen pieces analysed) and fifteen in Franzoni’s first collection, where they constitute approximately 53% of the twenty eight pieces analysed. We can thus consider Viadana’s collection to be the turning point of the survey in relation to the musical distribution of the poetic text, a turning point which chronologically separates, moreover, the two collections by Gastoldi, the second of which can be considered a “border-line case”.
If, in fact, it is true that in Rossi and in Gastoldi we find sometimes “a rapid and decisive declamation” which is here and there “ softened by giving way to vocal flourishes” [Assenza 1997, 189], thus recalling the style of Marenzio, it is equally true that these collections (that of Rossi, and the first of Gastoldi), show a contrapuntal texture which is sometimes dense, so much so to notably affect the segmentation of the poetic text in a continual discourse. On the other hand this would seem to be an element deriving from late 16th century humanist restraint, a kind of ennobling of the minor forms, rather than a innovation as such.
That restraint is completely absent in Viadana’s collection. Besides having no difficulty in accommodating the more “humble” characteristics of the villanella tradition (parallel fifths are to be found in these works, even if disguised), Viadana is the most radical composer in the canzonetta style from amongst the Mantuan composers of the lighter forms: the most careful not to submerge more than one textual segment in polyphonic segments. With respect to his collection, even the chronologically most advanced of the group examined, the Fioretti by Amante Franzoni, are less audacious: in his first book the works that have at least one section which is little segmented constitute no more than half, as has been seen, even if the overall picture conferms in any case the
predominance of polyphonic segments over textual segments. In particular Franzoni’s works are
notable for their marginal use of the basso continuo, which is really the basso seguente of a voice which always has the same rhythm as the others. From that point of view, the works of Viadana which neutralise the vocal role of the bass (see again example 15) are much more significant; the same Gastoldi, in the second last work of his second collection, takes a further step in that direction (example 16: the three polyphonic segments are also three musical phrases).

example 16

To reassume: if during the period 1589-1607 (the duration of the duchy of Vincenzo Gonzaga) one finds within the canzonetta a progressive move away from madrigalistic writing in favour of a more articulated pronunciation of the text, we have to acknowledge that Viadana’s collection occupies an unusual and stylistically advanced position. To find something more advanced, we must take into account those pieces which abandon classical texts in favour of models inspired by the school of Chiabrera.


7. Conclusion
The results which have come to light through the application of the method proposed here confirm, on a rather more rigorous level, that which we can gather from an intuitive reading of the repertoire. It might therefore seem redundant. However, the fact that the analysis of this Mantuan sample sustains, historically, that which we might have supposed, seems to confirm the validity of this method; the which, as well, produces less obvious conclusions, such as the particular nature of Viadana’s collection.
However, beyond the results that such a method can offer in its restricted formulation applied to a specific repertory, we believe that the general principle behind the present work is important: that is to say, the hypothesis of a double level of segmentation for the vocal polyphonic repertory. It is obvious that, in defining the two levels of segmentation, one must in the first place consider the pertinence of the criteria in relation to the characteristics of the repertory examined. It remains clear that to proceed to a preliminary segmentation which is not based on the concept of the cadence allows us to avoid two contradictory problems which are equally troublesome: on the one hand, that of ascribing segmental values to passages which, while corresponding to the compositional definition of the cadence, do not derive from that the sense of a textual caesura; on the other, that of not identifying as musical phrases those polyphonic portions which, although not concluded by cadential formulas, constitute in themselves as many expositions of textual portions. In the case of the vocal repertoire of the sixteenth century this will be of advantage in the analysis of tonal movement: establishing how orthodox a cadential pattern is with regard to the mode, presupposes that it is clear which, effectively, are the cadences of the work.


8. References

Salomone Rossi (1589), Il primo libro delle canzonette a tre voci, Amadino, Venezia (ed. mod. Complete Works vol. VI, CMM 100/6, a cura di D. Harrán, America Institute of Musicology, 1995).

Giovanni Giacomo Gastoldi (1592), Canzonette a tre voci con un balletto nel fine, Amadino, Venezia.

Lodovico Viadana (1594), Canzonette a tre voci…, Amadino, Venezia (ed. mod. a cura di G. Vecchi, A.M.I.S., Milano, 1965).

Giovanni Giacomo Gastoldi (1595), Canzonette… libro secondo, Osanna, Mantova.

Amante Franzoni (1605), I nuovi fioretti musicali… col suo basso generale per il clavicimbalo…, Amadino, Venezia.

Amante Franzoni (1607), Il secondo libro delli fioretti musicali…, Amadino, Venezia.

*****


Assenza C. (1997), La canzonetta dal 1570 al 1615, LIM, Lucca
.
Baroni M.- Dalmonte R.- Jacoboni C. (1999), Le regole della musica, EDT, Torino.

Beltrami P.G. (1994), La metrica italiana, Il Mulino, Bologna.

Cardamone D.G. (1981), The “canzone villanesca alla napolitana” and related forms. 1537-1570, UMI, Ann Arbor.

Cornulier B. de (1982), Théorie du vers. Rimbaud, Verlaine, Mallarmé, Seuil, Paris.

Dalmonte R.- Privitera M. (1996), Gitene, canzonette. Studio e trascrizione delle canzonette a sei voci d’Horatio Vecchi (1587), Olschki, Firenze.

Dürr W. (1994), “Imitar la poesia. Luca Marenzio ed il madrigale Amatemi ben mio”, in Colzani A.-Luppi A.- Padoan M. (cur.), Il madrigale oltre il madrigale. Dal Barocco al Novecento: destino di una forma e problemi di analisi, A.M.I.S., Como, 227-254.

Halle M.- Keyser S. J. (1980), “Metrica”, in Enciclopedia Einaudi, 9, Einaudi, Torino, 254-284.

Jakobson R. (1966), Saggi di linguistica generale, a cura di L. Heilmann, Feltrinelli, Milano.

La Via S. (1997), “«Natura delle cadenze» e «Natura contraria delli modi»”, Il Saggiatore Musicale, 4/1, 5-51.

Mangani M. (1996-97), Per un’analisi delle forme minori tra Cinque e Seicento. Indagine su un campione mantovano, Università degli Studi di Bologna.

Meier B. (1988), The modes of classical vocal polyphony, The Broude Brothers, New York, trad. E. S. Beebe.

Ruwet N. (1983a), "Nota sulle duplicazioni nell'opera di Claude Debussy", in Linguaggio, musica, poesia, Einaudi, Torino, pp. 55-85 (ed. orig. Revue belge de Musicologie, 1962, 16, pp. 57-70).

Ruwet N. (1983b), "Metodi di analisi in musicologia", in Linguaggio, musica, poesia, Einaudi, Torino, pp. 86-119 (ed. orig. Revue belge de Musicologie, 1962, 20, pp. 65-90).

Schwind E.- Polth M. (1996), voce "Klausel und Kadenz", in Musik in Geschichte und Gegenwart., 2a ed., Bärenreiter, Kassel, 1996, vol. 5, pp. 255-282.

Zarlino G. (1558), Istitutioni harmoniche, ed. facs. The Broude Brothers, New York, 1965.

 

< previous | index | next >