Contrapuntal patterns

Introduction

Roman numerals describe first and foremost vertical structure but we have already seen examples how the chord alteration syntax with parentheses () can be used to describe horizontal, melodic, contrapuntal phenomena such as a fourth suspension or a 7-6 bass figure. Here we are going to have a look how the DCML harmonic annotation standard can be used to systematically express contrapuntal phenomena such as sequences. Before we dive in, let’s quickly brush up the basic principle.

Annotate corelli_op01n01a.mscx from m. 9 until the end, leaving out the part in between for now.

First movement of Corelli op. 1/1, mm. 9-14, annotated

Analytical decisions

There are a couple of details that different annotators might solve differently. It goes without saying that the literal repetition in the music requires the exact same annotation.

• The beginning of each phrase includes the upbeat: { is on beat 1.5 (“one and”) rather than on 2 (repetition accordingly).

• If taken literally, the figures 6-5 in m. 10 translate to V6/V viio/V or, more plausibly, V6/V V65/V since we can assume notes to keep sounding. This has been condensed into one label here.

• Several solutions are possible for beat 4.5 of m. 10 depending on what you consider to be a proper or a passing chord. According to what you would write in a harmonic reduction other solutions could be I6 vi, I6 V7 vi, or I6 IV V7 vi.

• The resolution of the fourth suspension in m. 11 was put on beat 2.5 reflecting both the harmonic rhythm and the placement of the figure 3, and interpreting the belated E in Vl. 2 as rhythmic ornament.

Once more, the chain of 7-6 suspensions is expressed in the Fauxbourdon IV6 iii6 ii6 with (2) suspensions of the root. The principal means for expressing contrapuntal similarity is an analogously similar syntactical pattern in the annotations. Let’s look at a more tricky example now.

A more difficult case

Let’s now finalize corelli_op01n01a.mscx by going through three different solutions for mm. 7 & 8. The goal is once more to capture the 2-3 dissonance chain. Another question we need to answer in parallel is where we change the local key to vi.

Solution 1

First movement of Corelli op. 1/1, mm. 6-9, possible solution

This version uses the earliest possible moment to change the local key to vi and then translates the figured bass mechanically. The annotation as iv65 iio6 IIIM65 i6 ii%65 is technically correct and meets the criterion that the contrapuntal pattern should be reflected in a syntactical pattern; although it does so in a rather opaque manner (through the root pattern -2 +1 -2 + 1 steps and alternating inversions). However,

• the 2-3 suspensions are implicit in this annotation but could have been made explicit through the () syntax, which is part of its purpose;

• the first chord of m. 8 can be viewed as part of the cadential pattern in the sense that it could be viewed as a cadential six-four chord where the fourth is suspended by a fifth.

Let us look at the second solution which tries to address the second of the two points.

Solution 2

First movement of Corelli op. 1/1, mm. 6-9, possible solution

In the DCML harmony annotation standard, the cadential six-four is usually written as the suspension chord V(64), distinguishing it from the passing i64 or I64. But how would we express a suspension of the fourth which itself is a suspension? Remember that V(64) means that 5 and 3 are not present. What about V(+6)? Well, doesn’t really capture the fact that the fifth is dissonant, does it? So, V(65)? Similar, maybe reach for a lifehack (in German “trick 17”) and use the special symbol v (the opposite of ^) to display the direction of chord tone replacement, V(6v5)? Yes, that one comes closest but then we have the problem of V supposing a major third (C#), when the thorough bass actually does not mention it. v(6v5) or even V(6v5-3) would fix that but it is quite tangible that we have manoeuvred us into a corner here. So for the sake of legibility, let’s call this a tonic where the root D is replaced by the upper and the lower neighbour, namely by the C that is implied by the bass figure. Chord tone replacement from below can generally be displayed through the mentioned ^ sign (e.g. i(^2) when 2 replaces 3 rather than 1) but this is not necessary for (7) (nor for any raised number as in i(#4#2) for replaced 5 and 3). Hopefully, this tour de force through the various options is capable of making the slightly unwieldy solution i64(72) palatable. Compared to IIIM65, it brings with it the advantage that i64 translates to the same notes as the cadential six-four and, compared to V, it expresses the resolution of the suspension over the same root, i6, just with a different inversion.

The problem in this solution is, that the sequential pattern is obfuscated because of the interpretation of the last two chords of m. 7 “as one hears it” in the old key (although the annotator acknowledged the indeterminacy of the local key by giving the alternatives viio6-iio6/vi). So let’s try to fix this, shall we?

Solution 3

First movement of Corelli op. 1/1, mm. 6-9, possible solution

In this solution, the key change in m. 7 allows for expressing the sequence in a consistent and easy-to-read manner while keeping the characteristic V7 vi progression before. Also note the similarity to the annotation of the 7-6 chain (which is, of course, the inversion of 2-3) at the beginning of this section: In both cases, the main characteristic as a chain of suspensions has been explicitly expressed by annotating suspension and resolution over an identical root, and even in the frequent case that the bass progresses in the moment of resolution.

7-6 suspensions versus seventh chords

One could argue that Solution 1 is better because it shows the continuation of the sequential logic up to the ii%65 chord. The argument points to a fundamental problem of the semantic difference between

• the historically older seventh in the sense of a patiens which “wants to” resolve downwards into the sixth over the same agens (and, thinking in modern terms, over the identical root a third or a fifth below the agens), and

• the “emancipated” seventh as a fourth chord tone, which “wants to” resolve downwards as well, but does so a priori into the consonant chord tone pertaining to a different root (preferably a fifth below).

Hint

Some think that the main difference between the two is the absence or presence of a fifth but the distinction does not hold because many 7-6 suspensions come as 75-6 suspensions.

The two resulting viewpoints are demonstrated most clearly above. Solution 1 shows that a horizontal 2-3 suspension chain (as well as its inversion 7-6) can be implicit in an annotation pattern that focuses on verticalities, as is most obvious, for example, in all inversions of the ii7 V7 I progression. Solution 3, through it’s use of (), by default is the more horizontally oriented one, and therefore it can be more sensitive to and expressive of the voice-leading tendencies of particular tones. Leaving the annotation pattern through the use of ii%65 can be justified twofold:

• at this moment, the sequential pattern is taken over by a wide-spread cadential pattern;

• the logical continuation #viio64(72) would obfuscate the chord’s pre-dominant function (the legal heir of the historically more accurate antepenultima function).

Note

These subtle semantic differences are part of the reason why we need human annotators.

Annotate mm. 233-252 of beethoven_n03_1.mscx taking contrapuntal patterns into account.

First movement of Beethoven’s Piano Sonata no. 3, mm. 234-240, annotations

At this point of the tutorial, this first section hopefully did not present you with any problems. There is no global or local key in the beginning to show we are dealing with an excerpt (in C major).

First movement of Beethoven’s Piano Sonata no. 3, mm. 241-245, annotations after review

In this segment you see labels in three different colours reflecting the result of a review.

black:

labels left untouched by the reviewer;

red:

labels removed by the reviewer;

green:

labels inserted by the reviewer.

Please pause and ponder for a moment what the reasoning behind each of the two versions might have been and how they relate to yours.

The annotator (in red) had learned their partimento well and defended their interpretation as a stereotypical realization of a basso che scende legato. The reviewer, on the other hand, wanted to highlight the continuum with the following Fauxbourdon (mm. 243-5). Since both are legitimate views, they settled on integrating both perspectives:

First movement of Beethoven’s Piano Sonata no. 3, mm. 241-245, annotations reflecting expert consensus

Note that an inconsistency in the annotator’s version had to be fixed in the process: They had to opt for either [IV(2), iii(2), ii(2)] or [vi2, V2, IVM2], not a mix of both.

And, for matters of completeness, here comes the rest:

First movement of Beethoven’s Piano Sonata no. 3, mm. 246-252, annotated

Analytic conventions

The above-mentioned decision problem arises in many cases where voice-leading phenomena are to be taken into account, for example the “5-6 movement”. In the examples we have seen so far, this has been consistently encoded in the form i VI6, I vi6, IV ii6, iv iio6 etc. which stand firm as conventional annotation patterns. In cases where annotators find it more reflective of their interpretation to write i i(6), iv iv(6) etc. they may, of course, do that but need to be prepared for reviewers asking for justification.

In the previous assignment on Beethoven we have seen an example where a historically informed approach based on contemporary partimento practice lead to a notation that used proper seventh chords (vi2 ii6 V2 I6 IVM2 viio6) rather than highlighting the chain of suspensions (ii6(42) ii6 I6(42) I6 viio6(42) viio6). This may be justified once more with the notoriety of the falling fifths sequence. It forms a renowned recurring musical pattern and, vice versa, there is an analytical convention to group all its different manifestations under the analytical pattern of “roots descending by fifths”. The overarching sequence of falling seconds is, of course, always implied, as are the suspensions as soon as seventh chords are involved. And last but not least, as a weaker justification we can ask ourselves which of the two mentioned chord representations is closer to what we have in mind when realizing a basso che scende legato on a keyboard instrument ourselves.

Another convention that often leads to questions concerns typical opening patterns such as i ii%2 #viio7 i or I ii2 V65 I because they can also be interpreted as actually describing a suspension in the bass voice, expressed as i #viio7(2) #viio7 i or I V65(42) V65 I. Although the latter two may definitely be legitimate interpretations in some cases, we generally stick to the aforementioned convention. [Otherwise, why would we be teaching students all inversions of I ii7 V7 I? ;⚫︎) ]

Now let’s have a look at the :doc:`remaining features<features>` of the DCML standard in conjunction with some special cases.