Modal Glen Halls © All Rights Reserved

'Modal' is term with many inflections and uses both in the classical and jazz world. The word modal in the phrase 'modal ', which is transferable to both contexts, refers primarily to a kind of function distinct from other conventional functions of classical music, though oft found in folk musics. Specifically, the modal cadence is, like all resolutions in tonal music, essentially upper partials closing on lower partials, and refers to the sound of predominantly whole step motion as opposed to the 1/2 step motion characteristic of the other functions.

For example, in minor ( dorian?) "Upper Partials closing on lower partials" closing on d a

9 11 13 closing on 1 3 5

The modal cadence presents a unique contextual issue: a modal cadence often occurs in a 'modal style' in which a drone is prolonged but is rarely fully at rest. Modal cadences are appropriate and somewhat 'open- ended' structures in this context. Cadence is such a grey area term; does it mean an end, is it motion to rest, or is it simply motion ( instability )to less motion ( relative stability ) In many instances such as in drone based modal jazz, modal cadences are really suspensions and resolutions over a continuing pedal point.

Unlike conventional functions where dissonance is clearly resolved and in the expected direction, most modal cadences are simply voice leading illusions which play with these expectations. First, from an acoustic-functional point of view, there are only three notes capable of true modal function, in other words , where all three chord tones are upper partials, all resolution tones are lower partials, and all motion is by whole step. These are -7 , 6, and 2 in major., -7,6, and 4 in major. From this perspective there are really only Two definitive modal cadences: bVII(7) to I in Major, and ii to i in minor.

We then again enter the grey areas of harmony and terminology. The reason there is really nothing written about modal cadences is that it requires judgment calls, and as mentioned earlier the word modal has so many connotations, which will hopefully be addressed to some satisfaction later in this paper. Many would regard IImi to I in major keys as well as bVII to Imi in minor keys as 'modal cadences'.

Certainly, given the options of our other functional terms such as dominant and subdominant, modal is the best, most representative and distinctive description of the above progressions. This may seem trivial, and perhaps given the complexities and realities of actual jazz performance, it is- nonetheless we shall attempt to make some distinctions. In the Dmi to progression, even though with the F-E is a proper subdominant resolution, from a perceptual point of view the progression is equally characterized by the two whole step resolutions in the outer voices. What this implies is a) outer voices will always carry extra perceptual weight b) any kind of strict parallelism will stand out as another kind of accent, and c) the behavior of the majority of tones ( in this case two whole steps vs. one half step) also contributes, to what degree I am not sure at this point. Arguably, the addition of 1/2 step motion renders these progressions stronger than the true whole-step modal cadences.

To summarize the distinctions made to this point: 1) There are only two true modal cadences in which all three tones resolve by whole step and in which all tones are upper partials closing on lower partials. 2) If a majority of tones resolve by whole step, and if these tones tones are in the outervoices and move in parallel, this will also be termed a modal cadence. Now it gets interesting.

There are other progressions which ought to be termed modal cadences, and which are infrequent in classical music. One recalls the term used in renaissance counterpoint, the 'phrygian cadence'. What does this mean? The phrygian , in layman's terms, is a in which the third degree is taken as root.. (Note, in actual fact there are many phrygian modes- there will be a phrygian mode for each type of basic scale set, and it is defined by the presence of the flattest flat in that scale set five descending fifths on a 'circle of fifths' relative to the tonic.)

The textbook phrygian cadence in counterpoint involves a descending 1/2 step in the lower voice with an ascending step in the upper voice. ( In contrast to other cadences in which the intervals are reversed)

If the idea is extended into the harmonic context TWO phrygian cadences are suggested.

Strictly speaking, the F to Emi progression would be termed a strong subdominant progression, not a modal cadence. Dmi to Emi , however, would be termed a phrygian modal cadence. Why? Traditional theory leaves off at just this point, just where it gets confusing and less definitive. First, it is reasonable that a 'modal' cadence might also reflect the colour (tonal possibilities ) unique to a given mode, if possible. Secondly, if we consider the Dmi to Emi example, upper partials are not closing to the expected lower partials. This is the voice leading illusion mentioned earlier, and we will now see that there are many other examples. In other words, in E mi , or if you will in the case of a basically minor tertian stack which takes E as its root, partials 7 9 and 11 ( or D mi) ought to close on partials 1 and 3. ( root doubled- The behavior of a -7 partial will be dealt with elsewhere. From an acoustic perspective and in terms of conventional resolution it ought to close on either 5,1 and 3, or 6 , 1, and 3, or 1 , 1, and 3. In all cases the listener would have the impression that the dissonance is resolved as expected. ) Here 7, 9, and 11 resolve to 1 3 and 5. The 9th and the 11th will typically resolve down to the root and third respectively. Significantly, we also note that in a phrygian mode the 9th is flat. This b2 subdominant will strongly suggest resolution to the root, but instead it moves away by whole step. So, we do have upper partials closing on lower partials, but working against this feeling of closure is the impression of deceptive resolution in at least two voices. This is a unique effect and as mentioned at the outset it is largely responsible for the impression of open-ended closure. This also is a characteristic of many modal cadences. Make no mistake, harmony is complex and we ought not to look for simple and tidy solutions. If the twentieth century has taught us anything, it is that all sounds, and especially the sounds of tonal harmony, may exhibit complexity - the suggestion of more than one impulse or meaning simultaneously.

As to the issue of modal cadences reflecting the character of a given scale mode, this is true to a degree. Just as certain scales and modes present new possibilities for 1/2 step resolution, or in other words present new and often unique subdominant, dominant, or mixed functions, it should follow that new possibilities for whole step motion or 'modal function' might also present themselves. So, before looking at modal types individually and offering more qualifications, let us review the situation.

Upper to lower partial motion, in parallel ( an issue still requiring more discussion) will be in one of these four forms:

* This is yet another contingency. Lower partial triads starting from 3 will be termed an 'alternate' resolution and triads from 6 will be termed'ïdeceptive'. Both are slightly weaker or 'less closed' than resolutions to 1,3, 5, naturally.

Let us remind ourselves of the tones we are talking about. In, for example, C major:

Bb to C

D to C

D to E

A to

A to B

F to G

F# to E And in C minor:

F to Eb

Db to Eb

Ab to Bb.

OK. I realize it's awkward and cumbersome, but we must bear in mind that in addition to the parameters previously mentioned, some modes may exhibit up to three modal cadences. The same cadence may occur in more than one mode. For the sake of terminology we have to make more distinctions, weak and unfounded though they may seem. In terms of the number of subdominants we are all familiar with the order of the modes: Lydian, Ionian, Mixolydian, Dorian, Aeolian, Phrygian, Locrian, . . Mode 8 , mode 9, mode 10, and mode 11. The first instance of a particular chord and cadence type will be given priority. If more than one cadence is possible within a given scale , priority will be given to the resolution to the lowest partials. We will likely end up with 'primary' and 'secondary' cadences

I promise that the list will follow shortly, but there is one more nuance one should be aware of- the modal sus chord. It was suggested earlier that a great many modal cadences in the context of real music and especially modal jazz, are really suspension. In other words the root of the first term of the cadence is not the root of the chord. Somewhere in a lower octave the drone of the tonic is perceived. However, why not play the tonic directly with the first term of the cadence? This is an interesting quality, more ambiguous, which is termed a modal sus chord.

For example, given the Lydian cadence D to C in C major, ( 9 +11 13 to 1 3 5 ) place the D triad directly over the C and you have a C Lydian Sus chord.

True, it is a secondary dominant II7 in 3rd inversion, but given a drone context in which the lydian scale has been established, or perhaps with the root doubled, it would not be perceived this way.

Given all of the rationale above, the following is a 'list' of likely modal cadences corresponding to the traditional scale modes.

Lydian - the sound of F# falling to E instead of rising to G is distinctive.

Note, V to Vimi Ami is available, but not unique to this mode.

Ionian= the 'modal' cadence Dmi to C , so common in Gospel and folk music.

Mixolydian One instance of the true or pure modal cadence described at the outset.

Dorian= This mode has more possibilities than any other.

Aeolian IVmi to bIII is best example. IImib5 is also a possibility, but only if it resolves in parallel to the relative major, bIII, in other words, with the rising voice-led illusion

Phrygian Two excellent examples here. We don't often encounter a bVIImi to Imi progressions, but according to our criteria it is the strongest and most definitive phrygian modal cadence. The second example, bII to bIII has the stronger deceptive effect and produces a weaker close.

Locrian At the locrian and beyond there are , arguably, no modal cadences fitting the above criteria. There are some interested sounds, however, which are close to modal cadences.

The question is whether bVI, containing to minor tonic tones, arguably, is a resting chord- a deceptive tonic, or whether it would be perceived as another moving chord, a proper minor subdominant. Perhaps the point is moot in that utility of modal cadences is that resolutions remain open-ended, often following a soprano melody in an ongoing parallel line.

OK. This is a familiar and very striking sound. Perhaps at this point we are really looking at resources of the diminished scale. This chord is yielded by the so-called subdominant diminished scale, I.e. the diminshed scale from b2.

The bV to bVI progression above, however, is not part of a diminished scale. It is perhaps the proper Locrian Cadence.

Some other questions to be answered.

It might be asked, given our criteria, must all modal cadences display strict parallel motion?

In the purest sense, no. At the beginning it was mentioned that bVII to I and IImi to I mi are really the only pure modal cadences, I.e., in which all tones move by whole tone and all motion is from upper to lower partials. Correct me if my arithmetic is false, but the only way to achieve a modal cadence with contrary motion is to limit the number of tones to 2. For example:

Remember, we are trying to resolve to 1,3, 5, and to a lesser extent 7. If we add a third tone, Bb, we really have an inversion of a parallel motion.

By leaving out an inner voice we have defaulted to a more surface oriented mode of perception. In other words, with strict parallel motion the ear ( brain ) implicitly understands that tertian stacks, or partials , are being manipulated. It is a less subtle, less intellectual mode of perception. If this is compromised, we are now forced to consider changes in density and the illusion of polyphony. These are great effects, but we are no longer dealing directly with per se. Remember also that most modal cadences involve a kind of voiceleading illusion where two voices do not resolve to the expected lower partial but are more- or-less 'forced' up to the unexpected lower partial. Faith in this illusion is only possible if parallelism is maintained.

Must all modal cadences be in root position?

This issue is the same as for conventional function: cadence on an inversion is less closed, it is a weaker close, or looking at it the other way, it is a more buoyant close- it keeps things open. From the perspective of modal function, however, all properties remain the same as for root position closes. The issue is really contrary motion again. So long as both terms of the cadence are in the same inversion the best sense of modal cadence is maintained. If inversions alternate an impression of independence of voices may emerge, which, even though from an acoustic perspective there may be whole step resolution to lower partials, dilutes the overall effect.

Must the outer voices always form a perfect fifth and the P4, M6 if we allow inversion. In other words, why do we seem to avoid the diminished fifth?

Well known among orchestrators, certain intervals will seem to fuse. The lower tone seems to support the upper; they bind. 8ves, P5th, and major and minor 6ths and 3rds are the best intervals for this. The P4th and minor 7th also fuse, but with a little less certainty. And, for 21st century ears the major 7th is the boundary. With this interval one hears three things: the top, the bottom, and the intervallic quality, which will be regarded as consonant, I.e. accepted as stable. The diminished fifth at present remains just beyond this 'acceptable-as-stable' frontier. It gives an impression of two independent voices needing resolution in contrary motion.

One has the impression both the Ab and the F have not resolved as expected. This suggests that contrary resolution by whole step of SUBDOMINANT may really the key to the modal cadence effect. Yes, this is a valid modal cadence, perhaps Aeolian 3 , but it doesn't seem quite right. When we speak of modal harmony, recalling the works of Debussy and the pandiatonicism of Milhaud, we remember that the intent was really to escape functional harmony- to work against it. It was an emancipation of melody and line, chords moving parallel beneath soprano lines, with little regard to barlines and stressed and unstressed beats. The level of focus in the music had shifted to stratification and textural complement. One's main impulse for using a modal cadence is , in a sense, to excite the entire scale simultaneously, to blur the distinctions of diatonic permutation. The focus is then shifted to line and the totality of the scale mode , and these qualities work against some other line and some other mode- stratification. The best modal progressions are those which sound similar. In Renaissance counterpoint the idea was to avoid accent- accent by dissonance, accent by leap, accent by improperly placed direction change. Modal cadences, particularly in the droned, modal jazz style, is in many respect a return to that ideal- Line is paramount, and momentum is to be maintained- with as much transparency as possible. This has been a lot of work, merely to suggest that the diminished 5th draws attention to itself on account of the tritone. This in itself creates an accent and works against line and transparency.

Finally, why are we restricting it to three tones? What happens when we add a fourth, fifth, or sixth tone? (Yes, it was correct to point out that I seemed to be conveniently ignoring this problem by restricting the discussion to triads.) The answer, I believe, has to do with intent. Is the intent to create a moving chord of a certain strength , a resting chord of a certain strength, or perhaps a chord which is both. One example of this type of chord is the common suspension- for example , a major triad in which the third is suspended as the fourth, a chord which has both stable and unstable elements.

It seems a trivial example, but conceptually it is huge. Another example of this type of complex chord, meaning it may reflect more than one meaning at the same time, is the overloaded tertian stack.

Common practice harmony, by the way, cannot answer this question. The addition of the upper partials to the chordal stack, not as unresolved part motion but as chord tones, begins with Debussy, as do most great innovations of the 20th century. The practice was adopted and continued, however, in the jazz tradition. It is the question of complexity again, that difficult to define region of tonal harmony where we do not wish to tread, where chords have contradictory tendencies. It is arguable than when a chord is interpretable perceptually, both resting or moving -stable and unstable, it is neutralized, freed, easily moved. As an illustration, consider heavy backdrops for theatrical productions. Some are very heavy and difficult to move, so to compensate an ingenious method is employed whereby pulleys with approximately equal counter-weight are attached. As a result, a small tug will lift an enormous set. Likewise a chord which contains both resting and moving tendencies is ideal for the parallel accompaniment of melodic line- the classic modal context.

Consider the following elaboration's of the Ionian modal cadence Dmi to C

Cadence a) Dmi to C is clear. Upper partials closing on lower partials, relative to the drone. Cadence b) Dmi7 to C(7) is less clear. Relative to the drone the C is not an upper partial. It may in fact reinforce the drone. The Dmi7 might be considered an inversion of C ionian sus (Dmi/C) . As we move through c) and d) the situation becomes even more grey. In both cases the first term contain lower members which are dissonant upper partials with respect to the drone and upper members which are the lower partials relative to the drone. At the second term it is reverse. The lower members are the proper resting tones, but now the upper component is unstable. It has become a question of subjective aural hierarchy, specifically this: Do we always justify the elements of a chord vertically and literally, or are even the upper elements of a chordal stack at liberty to suggest dissonance with respect to the drone. Can they function separately? In the second term of d) the C(7) 9 11 chord, so to speak, is it both a C triad and a B diminished triad? Yes- certainly the Bo does not have the dissonance of a stand-alone Bo triad, but it is not negligible either. This dissonance is enough to suggest that another chord should follow, and if there is a perceivable melodic line to follow it seems all the more appropriate.

Ultimately theory cannot define the behaviour of these chords. The individual musician must rehearse these progressions and establish a kind of internal aural memory of the physiological effect , the relative resting and moving tendencies of these sonorities. Only then can he or she use them for expressive purposes. The previous pages imply an academic predictability for this aspect of modal harmony, which cannot approach actual practice.