Trumping Notions of Voice in Handel’s Sound

by Jordan Newman


When looking at a da capo aria in Handel’s Messiah, one is gazing upon a form that Charles Rosen informs us had already become "monotonously standard by the 1720’s."1  Perhaps at the height of its popularity, the da capo aria presents itself twice in Handel’s oratorio of 1741. In both instances, Handel generally appears to conform to the standard model.

The Trumpet Shall Sound in part III, however, sets itself apart in two formal respects: its opening ritornello is omitted upon repetition; and the ritornello itself is disproportionately large in comparison to the standard configuration - as exemplified in He Was Despised of part II. While neither of these elements is profound in isolation, when coupled with the dramatic narrative, this aria’s conventional form, which portrays "the almost unavoidable stereotype" 2  of da capo structure, offers a number of interesting formal relationships.

Da Capo

There existed a number of standard da capo aria forms in the 18th century, usually employing the use of ritornellos: instrumental passages acting as introductions, interludes, and/or closing sections to the vocal parts. The most common variety was what is now frequently referred to as "five-part da capo" featuring the following formal outline:

[Fig.1.a and 1.b from Julian Budden, "Aria" in The New Grove dictionary of music and musicians (2nd Ed., Vol.1), p.891.]

In thematic, textual, and tonal organization, The Trumpet Shall Sound (TTSS) observes this form meticulously, except for the omission of the initial ritornello upon the repetition of A. In fact, TTSS’s form verges on what Julian Budden calls "da capo al segno," one of the most popular variations of the five-part, where A is repeated after B, only from the entrance of the vocal solo 3 . The only difference between Budden and Handel’s models is that Budden indicates a "tutti", or ritornello section in the last part of B in da capo al segno, where clearly none exists in TTSS.

This is an important difference whose significance will become clearer below. Suffice it to say for the moment that TTSS lacks both the repeating ritornello of the standard da capo (as indicated by the purple dal segno box in fig.1.c) as well as the tail-end ritornello of the da capo al segno (note the lack of white "rit" box in 1.c), which points to a particular form in which Handel consciously omits of any kind of ritornello that might be expected in one particular area of the piece.

Another important point to glean from Budden is his articulation of the presence of five parts in the aria, implying that the ritornellos (which total six in the ideal model--see He Was Despised) are accessories to the solo vocal sections. Rosen verbalizes this when he remarks that "[w]e instinctively hear the solo sections of a concerto or song as forming a whole independent of the framing sections of the orchestra". 4 There is a general consensus among theorists, then, that ritornellos in da capo arias are mere "framing" devices whose purpose is to bolster the vocal passages--the real meat of the piece. This is where we start to grasp how TTSS eludes the formal pigeonhole.

Functional Differences: Two Soloists

While much of the ritornello material of TTSS can be interpreted in exactly this way, as ornamentation of vocal material, the reverse argument can be just as valid here. The musical importance given to the trumpet, the ritornello’s soloist, can elevate the ritornellos to the level of solo sections, with the bass sometimes occupying a subversive role in context. This is perhaps symbolized by, and informs, the unprecedented size of the opening ritornello: the trumpet’s song. The near-equal status given to voice and trumpet can confuse ritornello with concerto; repetitions and omissions throughout the aria further tip the scales of importance one way or another, melting the form down from its illusory stability into an ever-shifting counterpoint of connection and meaning. It is this relationship between voice and trumpet that completely explains a unique formal interpretation of this aria. I will proceed to sketch the how’s and why’s of this view, and I emphasize that what I indicate as my interpretation here attempts to capture what I actually hear in the music.

Charles Jennens, Handel’s librettist, decided to incorporate the only reference to a musical instrument in Messiah here in this aria. Handel’s job, it would seem, concerns keeping the instrument distant enough to understand the piece as a vocal solo for bass or baritone. However, the very nature of the trumpet in the text implicates it into the story; it moves the action of the verse; by its sound "the dead shall be raised incorruptible". Fittingly, the trumpet is treated subtly and not as simple textual paint or as a token program element.

Several layers of allusion occur in and around this aria concerning the trumpet: the text of the verse indicates the trumpet; the trumpet implies the meaning of the text; the preceding recitative predicts the trumpet and foreshadows its own motives; and the overall harmonic development corroborates all of these relationships. One begins to realize how all of the elements in TTSS, including and especially form, can somehow be connected to a web, at the center of which lies the trumpet. There is no better place to start exploring these connections than in the opening ritornello.

The Ritornello

Before analyzing the minutia of this ritornello, one should note its position within the context of the oratorio. At 28 measures, it is the longest instrumental passage in Messiah after the overture, and Pastoral Symphony (No. 13). There are larghetto ritornellos of comparable size in two other arias, I Know That My Redeemer Liveth and If God Be For Us, but in terms of bar numbers, TTSS comes out ahead. The weight of the trumpet begins to make an impression here.

The trumpet’s role and the manner in which Handel sets it up are vital to the flow and purpose of the opening and aria. The ritornello does not sound like an instrumental break from the action of the story, but instead it seems to have direction, with the trumpet unfolding the events. This is partly due to the forecasting in the preceding recitative, where the bass sings "Behold, I tell you a mystery; We shall not all sleep, but we shall all be chang'd in a moment, in a twinkling of an eye, at the last trumpet." As motivic collateral, the opening words of the recitative (mm.1-2) offer up the same arpeggiated ascent in D major that begins the ritornello of TTSS (mm.1-4).

The immediate appearance of the trumpet in the ritornello, then, is another link in the chain of events in Handel’s narrative. The listener would expect nothing else. The trumpet is even more obvious in the subsequent solo with continuo and then once more as it is joined by the orchestra and continuo to close off the ritornello theme before the voice comes in.

Already, before the voice entrance, we have a playing out of solo and tutti sections, complete half and full cadences, something that can be argued to be a small development (see b' in appendix), and the successful realization of the introductory text of the recitative. In short, if there were not a singer on stage clearing his throat, this theme could presumably stand on its own, since the trumpet really is the voice here, and this is its "concerto."

Let me point out a couple of other factors that support this idea. Considering the other instrumental passages in Messiah I just mentioned, note that the overture is obviously an instrumental introduction; the Pastoral Symphony is somewhat of a break midway through part I (pastoral, calm, relaxing, nature); I Know That My Redeemer Liveth begins part III, making its opening ritornello an overture to this final part; and If God Be For Us is the final aria, sandwiched in between two weighty choruses, including the very last number of the oratorio. Its lengthy instrumental part, then, has a reflective or simply contrasting role in its context, and also provides a needed pause. Given the nature of TTSS’s opening, none of these descriptions or narrative roles of instrumental passages applies. Here, the opening ritornello is part of the story, not a set up, not a break or a creation for pure contrast.

What is this story then? Ignoring the preparatory libretto for now, the musical narrative of the ritornello divides itself into three discernable parts. The first, a, is a strong eight-measure statement by the trumpet and orchestra that ends on a half cadence in the key of D major (mm.1-8). The harmonic progression is outlined in figure 2, a Schenkerian middleground with the notes and slurs of the urlinie in blue and the supporting bass in red (the significance of the colours in all Schenkerian graphs will remain the same throughout this paper.):

Upon initial hearing and analysis (it retroactively changes when placed in context with the following sections), this figure can be reduced to a middleground in which the first three measures act as initial arpeggiation, introducing the head tone 5ˆ in m.4, which leads down step-wise to 2ˆ at the half cadence in m.8. The following eight measures, b (mm.9-16), move into the dominant area, acting as prolongation, not of the 2ˆ, but of the 5ˆ, which is the tip-off that a’s function differs from original inspection. Measures 17-20 (b') form a brief development of motivic material from b, winding back to I, and leading to c (mm21-28), where the 5ˆ is picked up and resolved down stepwise to the 1ˆ by m.28. The entire theme, then, can be reduced to a statement of 5ˆ in a, prolongation in b, and then resolution through 4ˆ 3ˆ 2ˆ 1ˆ in c:

If this is really just an opening theme preparing the voice solo, then why is the actual opening vocal material from mm.29 not introduced? And why does it require 28 measures, with extended prolongation and such elaborate composition? For example, why not simply repeat a after b and substitute the half cadence with a perfect authentic cadence? Why include a development and then a third section, which is entirely new material and really a grandiose eight measure-long cadential passage—whose importance is confirmed by its employment in the end of the whole aria (mm.149-156)? If, on the other hand, the purpose of this ritornello is to establish a definitive voice for the trumpet, then how do we account for the subsequent entrance of the human voice, for which this aria is in fact composed? Now we dive deeper into the internals of Handel’s music.

The Voice(s)

When the bass sings, "The trumpet shall sound" (mm.29-31), the ascending arpeggio is a condensation of the ritornello's trumpet opening (mm.1-3). There is a very close connection, then, between the two "voices;" the linkage of the trumpet and bass openings becomes quite clear:

As if to point out the association, Handel actually restates mm.1-2 in the accompaniment while the bass is still on the word "sound" in mm.30-31. The trumpet cuts out right away before "and" (m.32) to interrupt three measures later (m.35-36) imitating—two octaves higher—the voice solo’s enunciation of the head tone on the word "raised," and adding a dotted rhythm. From mm.37-40, trumpet and bass perform a duet with continuo accompaniment; the voice acts as a bass line to the trumpet’s already familiar theme from the ritornello, ending in a half cadence replicating the corresponding four measures of the opening ritornello (mm.5-8). This half-cadence is then re-replicated immediately by the continuo alone in mm.41-44, marking the end of the initial part of the theme in I. There are several observations to make here.

Despite the difference of melodic material between the bass’s theme and the trumpet’s in the ritornello, it is close enough that the identical harmonic layout is all that is needed to fully establish that this is all music based on a of the ritornello. In other words, if one is focusing on confirming the relationship between the two thematic entities, there is no problem. The trumpet part in the opening A theme, in fact, is a clone of the material it has in a of the ritornello, with slight rhythmic variation and interruptions by the voice. However, I have just foreshadowed what my ears perceive as the focal issue here. In the previous paragraph, I called mm.35-36 an interruption of the bass’s music by the trumpet, but I also just called the bass part an interruption of the otherwise-complete trumpet line. My deliberate sloppiness is meant to demonstrate the ambiguity of which voice is the central force. While musically, trumpet and bass weave a beautiful, unified tapestry, each being complete by itself but coming together coherently, the confusion of who owns the product is a source of tension. The following chart outlines some of the attributes that might qualify which of the voices might be the "real" voice in this theme:



This exercise illustrates the complexity of the issue of "voice" as introduced in the opening A material, a sapling which continues to blossom in the rest of the piece. Since mm.41-44 provide a separate ending to this section that excludes both voices, we can ultimately note how unified the two really are, even in absence. But if I were to proclaim a "leader" for the initial A theme, then I would probably choose the bass, since the listener has already had 28 measures of trumpet, and because by the early 1700’s we had come to expect an oratorical aria to feature a human voice with text. Rosen and his stereotype also win here, since we can now see the opening ritornello as part of the frame for the bass’s material—the "real" stuff. By no means, however, is the issue closed.

Take, for example, the chain of events: 1) recitative sings about impending trumpet, 2) said trumpet plays, 3) bass enters singing that the trumpet will play. There is a time discrepancy here, which in the grand scheme of things paints the entire bass entrance as an afterthought. Perhaps the trumpet "shall sound," but really, it already did. The bass’s text is describing what already happened in the ritornello—"happened" implying that it was an "event." Essentially, the ritornello is the dramatic event, and the bass is providing a review of it, despite my conclusion that the bass solo sounds more important. If this is true, then we should be able to trace the progression of two levels of form in TTSS, one traditional-musical (harmonic/melodic) and one dramatic (narrative), which might or might not correspond with each other’s functions.

Musical and Dramatic Form

Taking the traditional musical model, the next events of the aria conform properly. The half cadence in m.44 provides the necessary pivot chord for the new key of A major (D:V=A:I). The bass sings yet the same opening material featured throughout—the arpeggiated ascent which was introduced in the recitative, expanded upon in a of the ritornello, and condensed in the bass’s entrance. For registral reasons, however, the voice must start on the 5ˆ (E) and ascend an octave, emphasizing the dominant note rather than the tonic (mm.44-46). Therefore "sound" (m.46) and "raised" (m.50) are the same note this time, instead of the first being the tonic and the second being the dominant as in the preceding material. By m.52, the thematic pattern becomes clear and familiar: each of the textual phrases "the trumpet shall sound," "and the dead shall be raised," and "(and the dead shall be raised) incorruptible" takes its own musical material based on music from the ritornello. Considering the bass’s material as derivation from the ritornello, mm.29-58 can be charted thusly:

Example 5.

Key Solo Ritornello
29-32 The trumpet shall sound 1-3
33-36 and the dead shall be raised 3-4
37-40 and the dead shall be raised  incorruptible 5-8
41-44 (instrumental) 5-8
45-48 The trumpet shall sound 1-3 (3-4? Inverted arpeggio)
49-51 and the dead shall be raised  3-4
52-54 incorruptible 7-8
55-58 incorruptible 7-8

All of the bass material so far comes from a of the ritornello. There are two sections here, divided by the change in key. The symmetry is apparent between the two sections in the textual and musical material, where we are given one version in I, and one in V, which I have called a' and a'' in the A section (see appendix). By the time the bass comes in with "and we shall be changed," (m.59) there has been a temporary movement to I in mm.55-58, which the voice picks up and twists into a developmental pattern leading to a perfect authentic cadence back in the dominant key of A major in m.68. It does this initially by using a motive from b' of the ritornello, namely the descending leap of a fourth from the last beat of m.58 (rit, m.16) to the first beat of m.59 (rit, m.17) in the voice, followed by the same figure in the accompaniment in the next measure. This sparks a continuous stream of eighth-notes in the voice that last until m.66, where the cadential passage begins; all the while, accompaniment outlines a series of climbing thirds that eventually introduces tonicizations of A major (m.65) and leads to the re-establishment of the dominant.

In the ritornello, the function of b' was to bring the music back home from the dominant key. Here, the bass is also shifting the key, but this time from I to V. The word "changed," which is sung over top (or under) all of this, is appropriate, since that is what is happening to the key of the material and to the function of the borrowed motive. The following ritornello enforces V, at first (mm.69-73) with material from the original ritornello’s b, which is the exact function b had originally, and then from mm.74-78 with a one-measure figure repeated four times, a sort of codetta.

All of this makes perfect formal sense, except that, again, we must choose whose b' is the real one: the ritornello's or the solo's. In other words, is b' in the opening ritornello simply a foreshadowing device for the introduction of b'' in the solo section, or is b'' but a development of the stable exposition of b'? To take sonata form, whose terminology I have already borrowed, one may note how the development process is rarely if ever foreshadowed in the exposition, since the development is supposed to be a section that plays with our familiarity of stable, familiar themes. For all intents and purposes, b'' does indeed develop the ritornello material, taking as its canvas the unstable section—the one that is already a development of previous material to begin with. This means that even the most unoriginal section (in that its material is derived) of the ritornello acts as expository material to be developed by the solo, marking the ritornello with an air of completeness and stability that the solo lacks. The fact that the first instance of real development in the solo uses, not the solo, but material "exposed" as a development in the ritornello, supports this idea. The solo cannot even think up its own development; it must use the ritornello as a guide. In any case, we are forced, once again, to blow up an ambiguous bubble—this time around b', since its main function is confused between actively developing the past, and passively sitting like raw material for the future.

To touch more on the idea of "lack," there is more that attests to ritornello supremacy here in A of the solo. In example 6, notice the general confusion and disarray of the thematic and harmonic layout. The voice and continuo each seem to be climbing thirds, but their registral shifts are staggered, creating a play of inward, parallel, and outward movement while the individual and comparative intervals remain the same for and between the two lines. This is an outgrowth of the section’s initial canon-like entries of voice and accompaniment, whose downward leaps of a fourth are staggered by one measure (mm. 59-60). Determining the relationship between and direction for each line becomes a messy affair because of this continual shifting. Harmonically, we are not sure where to hear the return of the key of A major, if it is indeed a "return." After all, the brief visit to D major in mm.55-58 after the establishment of A major is permeated by G sharps, implying a looming A major in the unsatisfactory envelope of D major. b' begins on a ii chord (e) of D major, not generating a particularly strong tonic identity to set off the developmental pattern. The ominous G sharps return in mm.61-62, on tonic harmonies, casting doubt yet again on the idea of D major as tonic, but not quite establishing A major either. This establishment is only suggested at the V chord of A major in m.65, and then completed in the short four-beat cadence in mm.67-68. If the ritornello’s importance is not yet clear, it will be so shortly.

I labelled the ensuing ritornello bd (see appendix), because it comprises material from the initial ritornello’s b followed by new "codetta" material. However, in hearing the onset of this ritornello in m.69, one cannot help but notice the omission of the first two measures of the original b, so that instead of expressing an F-sharp resolution to E in the melody (which would correspond with mm.10-11), this ritornello first flaunts the originally-subsequent G sharp to A (from corresponding mm.11-12), and then continues as the original b does, repeating the initial two-phrase structure. To make sense of this odd phrasing where the second ritornello only repeats 75% of the first, we have to look back at this disarray in the previous section.

I trace a gradual movement from the beginning of b'' of general disorder towards coherence: the above mentioned thematic and harmonic ambiguity gives way to the cadence, where voice and continuo are finally in unison, one key is definitely established, and then the ritornello stabilizes everything by reinforcing A major with trumpet and orchestra. However, by starting on the second half of the b phrase, this ritornello links itself indelibly to the preceding solo section by tricking the ear into thinking that the entire b'', or perhaps the whole solo section, was an expanded first-two-measures of the ritornello, picked up by the ritornello-proper at m.69. Measures 69-70, in fact, sound more to me like a comment on the cadence that just occurred instead of the beginning of something else, problematizing the boundary of solo and ritornello, despite the fact that voice cuts out and orchestra pipes in. One could possibly hear the last two measures of b'' and the first two measures of the ritornello as one continuous phrase. The ritornello, then, is not simply filler or reinforcement; it can be interpreted as the main event, as the solo folds itself ever so subtly into the ritornello’s narrative, not the other way around.

From m.79, the bass’s original material comes back in the key of D major, marking the beginning of A'. This time, "the trumpet shall sound," with its usual rising arpeggio, is repeated, but the first "sound" touts the 5, and the second "sound" reaches the 1, which marks a reversal of the 1:5 order heard at the top. In fact, if we take the emphasized scale degrees of all of the bass entries with this text, we get a general outline of 1-5-1:

"sound" "raised"
Aa 1------5
a' 5------5
A' 5------1

The reversal sounds final; after all, we are now back in the home key. The accompaniment, including trumpet, supports the feeling of finality by majestically repeating the opening ritornello’s a (mm.84-90), but continuing to reach up two octaves above the D from m.84. Furthermore, the vi harmony that complements every utterance of the word "dead" in all previous bass entries, is replaced here by a IV, removing the original tension created by the minor-mode harmony. With this air of conclusiveness, it is a little bit of a surprise, then, that Handel continues with 50 more measures of A' material before introducing the closing ritornello. However, this can be expected, since Rosen notes, "[i]n Handel’s arias, [A'] tends to be a very free development of the thematic pattern of [A] until almost the end of Handel’s life" 5 .  Chalk it up to Handel’s personal style, then; more than restatement is likely in this section.

Although the development that occurs in A' deserves due attention in all of its elements, it is the motivic mixture that most concerns this essay. By glancing at the appendix, the assortment of colours and letters is an indication of the thematic complexity within this section. Unlike A, which uses parts of a and b' exclusively and separately, A' features almost schizophrenic combinations of derivative material, mostly from the ritornello’s b, b' and c, in developmental vertical musical blocks. The busiest section is undoubtedly mm.129-140, where at least three motives occur simultaneously with all instruments stirred in the mix, representing an overall synthesis of events and ideas.

The closing ritornello comes as expected, but with the entire b section omitted, and therefore without prolongation of the dominant, and with greater tonal stability. Both the textural and motivic "schizophrenia" of A' and its powerful, secure answer in the ritornello are, of course, a set up for a contrasting B section, whose trumpet-less and orchestra-less texture is minimal and whose material is new but immediately coherent.

The B Section
The new lines are "for this corruptible must put on incorruption, and this mortal must put on immortality." In the key of b minor, the first line is stated twice, and features a half cadence in the first utterance, and a perfect authentic cadence in the second. The second line is also stated twice, half cadencing the first time in the key of f minor, and by the end of the section making a full cadence in the same key. The first line statements (appendix, B: f) comprise the same musical material, except for some ornamentation and the different cadence the second time. The second verse statements (appendix, B: g), while appearing different, can be seen to be the retrograde of each other, even at the most superficial layer where the first descends step-wise, and the second climbs.

This music for the first proclamation of "immortality" (mm.184-192) can actually be interpreted as a kind of inversion of the material from the word "changed" in mm.123-128, providing the linkage between this otherwise completely distinct section, and A.

The literal implications of the two words’ comparative meanings are paralleled in the inverted music. "Immortality" implies a consistency, and stability, whereas "changed" is obviously the opposite. On the level of irony, it is interesting how the bass can be purporting immortality/stability in the B section, which is ultimately the section created for contrast, out of the need for "change" from the A.

Along those same lines, the clarity and tranquillity of this section marks it as a break from the action of the A, the irony being: break, tranquility, repose—these are generally the terms I used to describe the instrumental passages of Messiah, providing a break from the story and action. This B section does the exact same thing, but it is far from instrumental, in fact, it is the most overtly vocal part of the piece, being scored for just voice and continuo. In contrast to He Was Despised where the mood takes on something of adrenaline and the texture and rhythm pick up force at the B, we see the opposite here, which associates the voice with a bit of triviality here. Perhaps the most plausible argument for equal importance between the two soloists comes next, where without the privilege of the opening trumpet-solo ritornello, the music returns to A immediately after B.

Omission = Egalitarianism and Revolution

Returning to A, we begin with the bass solo in m.29. Remember, when the bass first came in at the beginning of the piece, I concluded it was the winner of attention and importance. However, after B, when we begin with precisely this same bass entrance, it is all about trumpet and orchestra, since they have been absent for the last few minutes in the entire B. The bass’s text actually makes sense now, since when the trumpet enters immediately following the assertion that "the trumpet shall sound," we are listening to it, and not thinking about the bass, and we are without the baggage of the long trumpet ritornello we heard the first time around. Because of the omission, the trumpet is elevated within the very music the bass owned, and it now occupies a central role. This is an irony that would be lost if the piece were a normal da capo, reverting back to the opening ritornello, since of course trumpet would be the focus there—but a marginal one, unconnected with the "real" sung music that follows; the formal relationship between the parts would remain the same if it were to conform.

By changing the larger structure in this aria, then, Handel gives an ironic sort of synthesis that replaces importance on some of the very music we have already heard, not by altering the musical material, but by changing our perception of it through dramatic form. We can trace this idea from the very beginning, for example, the re-conceptualization of the ritornello’s a, and the cropping of the b in the second ritornello. If all of this re-evaluation seems hard to swallow, perhaps the final point to make is that The Trumpet Shall Sound is ultimately a messianic song about being "changed." Bass and trumpet are the necessary vessels, allowing Handel to transmit this message.


  Charles Rosen, Sonata Forms (New York: W.W. Norton, 1980) p.30. (Back)
2.  Ibid., p.30. (Back)
3.  Julian Budden, “Aria” in The New Grove dictionary of music and musicians, edited by Stanley Sadie; executive editor, John Tyrell (2nd Ed., Vol.1), p.892.  (Back)
4.  Rosen, p.39.  (Back)
5.  Ibid., p.40.   (Back)


© Copyright 2001 by Jordan Newman.