Changing the local key

The syntax

At the beginning of the tutorial, we saw that the first label in every piece defines a global major or minor key as a note name followed by a dot and a chord label, e.g. C.I for the first tonic chord in a C major (lydian, mixolydian) piece. What is implicit in this label is that the local key in the beginning is the key of I. If the C major piece started, say, with a slow introduction in A minor, the first label would be C.vi.i, indicating the local key of vi, thus as scale degree relative to the global key.

Remember

The local key is always expressed relatively to the global key. Consequently, the distances between the local tonics of the keys of III, VI, and VII from the global tonic are major intervals in a major piece and minor intervals in a minor piece.

To change the local key in the course of the piece it is enough to begin the label with the corresponding scale degree followed by a dot and a chord label relative to the new local key. The question often is when to introduce the key change. For example, the following progressions are equivalent, the first label simply indicating the local key of I:

  1. I.I viio6 I6 viio/V V V7/V V

  2. I.I viio6 I6 viio/V I/V V.V7 I

  3. I.I viio6 I6 V.viio I V7 I

  4. I.I V.viio6/IV IV6 viio I V7 I

  • In (1), the local key does not change at all, expressing a mere tonicization of V in a context that continues in the key of I.

  • (3) interprets the first three chords in the ‘old’ key and from the moment the leading tone of V is introduced, it assumes a change of key.

  • (4) even performs the change to the key of V two chords earlier, interpreting the ‘old’ tonic as the ‘new’ IV chord. Coming from the key of I, one would have to justify what evokes the impression of key change at this early point.

  • (2) expresses that the key of V is first tonized and then becomes the new local key. We would probably have a hard time coming up with a real-word example where this interpretation would be justified.

Note

By the way, there is no real difference between notating viio/V V (1) and viio/V I/V (2), except that the latter facilitates reading the chord progression as viio I. This is strongly encouraged in cases where the tonicization includes a cadence, e.g. I.I ii6/V V7/V I/V}.

Common mistake

Since the root of viio/V has an accidental, novice annotators are often tempted to write #viio/V. This is incorrect because the leading tone is scale degree 7 in major, not #7. These are all correct:

  • I.viio/V

  • i.viio/V

  • I.#viio/v

  • i.#viio/v

Try it out!

Get back to corelli_op01n01a.mscx and continue your annotations up to beat 1 of m. 6. Where do you change the local key?

Here are three possibilities, is yours among them? What don’t you like about the others?

Option 1: m. 5
First movement of Corelli op. 1/1, mm. 3-6, key change in m. 5

First movement of Corelli op. 1/1, mm. 3-6, key change in m. 5

This version stays in the key of I as long as possible and does not reflect the signal B in m. 4 in any way.

Option 2: m. 4
First movement of Corelli op. 1/1, mm. 3-6, key change in m. 4

First movement of Corelli op. 1/1, mm. 3-6, key change in m. 4

This version takes the signal B into account by changing the local key immediately after. It is therefore more expressive than the previous one.

Option 3: m. 3
First movement of Corelli op. 1/1, mm. 3-6, key change in m. 3

First movement of Corelli op. 1/1, mm. 3-6, key change in m. 3

Here, the annotator decided to interpret the whole passage in the ‘new’ key of V, but, being uncomfortable with the resulting minor v in m. 4, decided to add the alternative interpretation as a borrowing from the ‘old’ key of I, as in the two versions above. Note that in the new key of V, the old key of I is located on scale degree IV, hence the expression ii6/IV. The primary solution v6 was probably kept to highlight the 7-6 consecutive with its stepwise descending roots. This version certainly the least elegant one and probably hard to justify.

Taking all arguments together, Option 2 is the winner.

Change of key or tonicization?

But what if we take the following two measures into account, mm. 6 & 7?

First movement of Corelli op. 1/1, mm. 5-7, annotated

First movement of Corelli op. 1/1, mm. 5-7, IV(9) rather than IV(2) because Bb is present in the bass and therefore the placed chord tone is 8, not 1.

The music is clearly back in the key of I. Doesn’t that make a fourth option more likely where the local key does not change to the key of V at all and m. 5 is simply considered as a tonicization (V/V iv/V ii65/V V/V I/V)? Yes, that is possible and once again it is the annotator’s decision what they want to express. Let us look at the piece’s tonal plan when sticking to this decision and annotating both the cadence to C in m. 6 and the cadence to d in m. 9 as temporary tonicizations (applied chords, in more general terms):

The blue line shows the local key which remains in the key of I in this version. Red lines show keys that are tonicized using applied chords (i.e. labels including a slash), and the green lines show where the tonic of the temporarily tonicized key is present. Whereas this kind of interpretation might be sensible when looking at a longer piece or, for example, the whole trio sonata, considering the short form alone calls for an analysis that reflects the two cadences to other keys, resulting in this tonal plan:

Going forward, let’s have a look at how to encode contrapuntal patterns such as sequences.