How can I make 12 tone and atonal melodies sound interesting?

Question

I have recently started learning about serialism (12 tone melody writing) and atonal music. When I attempted to write a short melody, it didn't sound quite pleasing to the ears. Yes, I know atonal music is different than the usual tonal harmonies we are used to, but the melody I wrote just seemed like random notes of the chromatic scale put to a rhythm.

How can I make 12 tone melodies and atonal melodies seem more interesting and pleasing to the ear? Are there any suggestions on writing a melody in this style of music?

Answer 1

It is important to note that Schoenberg didn't invent twelve-tone serial music specifically to get away from tonality or functional harmony. He had already been writing atonal music for more than a decade without using any kind of serialism. The problem he was trying to solve was that he struggled to give his atonal music unity, and he hoped to achieve that by keeping all 12 notes in constant circulation and by using a particular twelve-tone row as the underlying principle of the music. The twelve-tone technique has in fact more of a structural than a harmonic or melodic implication.

So the key to how to approach twelve-tone music lies in the techniques that Schoenberg, Webern and Berg had already been using in their earlier atonal music. They had developed a method of using small pitch-class sets as sources for motives. This didn't change; the twelve-tone row and its permutations provided a constant background of the same pitch classes, but the different versions of the row provided common pitch-class orderings, intervals and subsets, which could then be used as motivic source material.

As I mentioned in my answer to one of your previous questions, an important part of Schoenberg's, Webern's and Berg's twelve-tone music was their choice of twelve-tone rows. They didn't choose these randomly, but studied their properties, and used rows whose different permutations had shared elements that they could derive motivic material from. Below are a few examples:

Schoenberg: String Quartet No.4 op.37 (1936)

X:1
L:1/1
M:
K:C
%%score V1 V2
V:V1 clef=treble name="P0"
V:V2 clef=treble name="I5"
% 1
[V:V1] d ^c A _B F _E E c _A G ^F B
[V:V2] "_0"G "_1"_A "_2"c "_3"B "_4"E "_5"^F "_6"F "_7"A "_8"^c "_9"d "_10"_e "_11"_B

In this series there are four groups of three adjacent notes that form a pitch set with prime form [0,1,5]:

X:1
L:1/1
M:
K:C
%%score V1
V:V1 clef=treble
% 1
[V:V1] "_0"d "_1"^c "_2"A | "_2"A "_3"_B "_4"F | "_7"c "_8"_A "_9"G | "_9"G "_10"^F "_11"B |

This not only means that there are several instances of adjacent intervals 1 and 4 or 1 and 5, but also that many permutations will share groups of three notes; e.g. the I5 series-form shares the note group G-Ab-C with the prime form P0.

This is used by Schoenberg e.g. to link the first phrase (based on P0) played by the first violin, to the next phrase (based on I5) which is played by the second violin. Both times the note group G-Ab-C is played in the same register, and the note B which follows (almost) immediately is given a long duration; this makes it sound like the second phrase picks up where the first one left off.

Webern: Concerto for Nine Instruments op.24 (1934)

X:1
L:1/1
M:
K:C
%%score V1
V:V1 clef=treble name="P0"
% 1
[V:V1] "p0""_0"B "_1"_B "_2"d | "ri7""_3"_E "_4"G "_5"^F | "r6""_6"^G "_7"E "_8"F | "i1""_9"c "_10"^c "_11"A |

The series for this work, as often with Webern, is derived by taking a 3-note series, and then adding inversions or retrogrades of it to make up a full twelve-tone row. As a result of this, many series-forms contain the same groups of three notes (in order or reversed):

X:1
L:1/1
M:
K:C
%%score V1 V2 V3 V4
V:V1 clef=treble name="P0"
V:V2 clef=treble name="P6"
V:V3 clef=treble name="I1"
V:V4 clef=treble name="I7"
% 1
[V:V1] "p0"B _B d | "ri7"_E G ^F | "r6"^G E F | "i1"c ^c A |
[V:V2] "r6 (rev)"F E ^G | "i1 (rev)"A ^c c | "p0 (rev)"d _B B | "ri7 (rev)"^F G _E |
[V:V3] "i1"c ^c A | "r6"^G E F | "ri7"_E G ^F | "p0"B _B d |
[V:V4] "ri7 (rev)"^F G _E | "p0 (rev)"d _B B | "i1 (rev)"A ^c c | "r6 (rev)"F E ^G |

So even though the work uses many permutations of the complete twelve-tone row, the material is extremely limited, and the same intervals and note groups occur many times in different guises.

Berg: Violin Concerto (1935)

X:1
L:1/1
M:
K:C
%%score V1
V:V1 clef=treble name="P0"
% 1
[V:V1] "_0"G "_1"_B "_2"d "_3"^F "_4"A "_5"c "_6"E "_7"^G "_8"B "_9"^c "_10"_E "_11"F

This series contains four triads, their tonic a fifth apart like the open strings of a violin, and they appear as triads at several points in the work.

X:1
L:1/1
M:
K:C
%%score V1
V:V1 clef=treble
% 1
[V:V1] "Gm""_0"G "_1"_B "_2"d | "D""_2"d "_3"^F "_4"A | "Am""_4"A "_5"c "_6"E | "E""_6"E "_7"^G "_8"B |

Berg was not averse to tonal elements in his music, unlike Schoenberg and Webern, and especially the Violin Concerto often blurs the line between tonal and atonal, and even contains a Bach chorale as a literal quote.

---

As you can see, choosing the twelve-tone row for its combinatorial properties, and exploiting the possibilities of certain transpositions of the inverted and retrograde form, make it possible to create many connections between small motives in the work, and avoid what you describe as sounding like "random notes from the chromatic scale".

This also shows that the idea of all twelve tones being equally important shouldn't be taken at face value. The way Schoenberg, Webern and Berg use shared elements between different series-forms to generate motives creates temporary hierarchies between the notes, and a sense of foreground and background, of melodic motif and accompaniment, much like moving to the next chord in a tonal work changes the hierarchy between tonic and chord-tones and non-chord-tones.

---

Example

If you are asked to write a short phrase, or an unaccompanied melody, it is of course difficult to apply many of the techniques used by Schoenberg, Webern and Berg, and turn it into something interesting. However, the way Webern often derives twelve-tone series from shorter 3, 4 or 6-note groups, creates a microcosm in which you can use the repetition of intervals to link the parts of the series together. Consider this example:

X:1
L:1/1
M: 
K:C
%%score V1
V:V1 clef=treble name="P0"
% 1
[V:V1] "^p0""_0"C "_1"E "_2"^F | "_3"_B "_4"d "_5"_A | "^i7""_6"g "_7"_e "_8"^c | "_9"A "_10"F "_11"B |

This series has four groups of three notes, and each of these pitch-class sets has prime form [0,2,6]. The first two groups combined and the last two groups combined form whole-tone scales, which are a transposed inversion of each other.

When turning the series into a melody, you can emphasize the shared elements between the different parts and make it sound far from random, e.g.:

X: 1
M: 6/8
Q: 1/8=110 "slowly"
L: 1/8
K: C
!p! C2 (E^F3) | _BD.!>!_A A2 (A | G2) (_E^C3) | =A=F.!>!=B B.!>!bb |

The use of the whole-tone scale provides the listener with a link to more familiar pre-twelve-tone music, and the (slightly dramatic) downward half-tone shift into another whole-tone scale brings the symmetry in pitch to the fore. The rhythm clearly sets the groups of three notes apart, and repeating the note Ab marks the end of the first whole-tone phrase (and helps to set apart the half-tone step to G), and then repeating the rhythm links the two phrases together. (I added the octave jump just for effect, and as a promise of what comes next.)

Answer 2

There are almost 10 million distinct possible 12-tone rows, and essentially infinitely many different ways of using them. I could carefully construct a row so that it uses only a small number of different intervals, or so that it contains multiple instances of a particular harmony or motive, or so that is uses as many different intervals as possible, or I could roll dice, or I could quote a famous row from a different piece, or I could more-or-less intuitively build it simultaneously as I compose a melody, or, or, or, etc. ad infinitum. And even after I’ve built or chosen a row, I still have infinitely many choices for how I actually use it. Maybe I’ll use it to build a melodic line, or maybe it will help me build a series of harmonies. Maybe it will only be a background structural element that is never heard as a melody at all.

Your question makes it sound as if you just chose a row or built a row without thinking about its melodic consequences, and then have been frustrated that it doesn’t sound melodically interesting to you, but that’s a kind of backwards approach. I mean, it can be kind of fun to pick some row at random and then find out what it sounds like later, but that’s not a common or particularly effective approach in most circumstances. If you want to use your row as a melody (which, again, is not something that has to be done in a serial piece!), then you should probably construct your row as a melody from the very beginning. If you definitely want to use a classic 12-tone row without structural repetitions (which is also not a requirement!), then you can make a list of all 12 notes and cross them off as you write, but otherwise approach the writing of the melody purely in terms of what sounds appealing to you. Once you’ve used all 12 pitches, you have your row, and you can now use it to insure intervallic and structural consistency in other parts of the piece. Pick an inversion of your row and see if it yields a new melodic idea that’s sounds good to you. It doesn’t? Fine, then don’t use it! Explore your row and see what ideas it helps you generate.

And the instant you have an idea that “breaks the rule?” Do it! There are no serialism police that will arrest you. A tone row is only useful inasmuch as the composer finds it to be useful, so use it to generate cool ideas but abandon it when it doesn’t. If you look at the great serial pieces even from the earliest days, there are constant “exceptions” and “violations” of the “rules” because there aren’t any. Some composers, in some pieces, liked the music they created when they adhered to strict pre-compositional structural ideas, and so they stick to a row rather literally. Others just used the general idea to generate some initial ideas, and then kind of forgot about it. By all means, do some exercises for yourself where you impose different kinds of arbitrary limitations, but always remember that you’re the one imposing the restrictions, so you can abandon them as soon as you realize something better is possible.

Answer 3

...the melody I wrote just seemed like random notes of the chromatic scale...

In a way you are describing in a nutshell what 12 tone music is. Of course it isn't really random. The constraint to use all 12 in sequence is meant to equalize their importance and thereby eschew tonality, and that is not random.

I suppose you could describe the resulting atonal effect as feeling random in that there is no tonal center. In other words, the perception of randomness is the result of the 12 tone technique which produces the effect through a non-random, but equalized series of tones.

...put to a rhythm.

This I think is true - you apply rhythms to the tone rows - and in fact by my understanding this is one key points of 12 tone style: to focus composition on a new kind of rhythmic freedom.

Tonal music has been described as a "tyranny of the barline" meaning that tonal grammar it linked with rhythm and barlines: chord changes at bar lines, resolutions tending to long rhythm values, basically tonality and metrical feeling strongly linked.

12 tone music broke that rhythmic tyranny through atonality, the absence of the tonal grammar and it's rhythmic expectations.

In 12 tone style you are supposed to have more freedom to use whatever rhythm you want without the expectation to reconcile it with tonal/metrical expectations.

In that sense putting the tone rows to rhythm is exactly where the attention should go.

There are lots of ways to manipulate segments of a tone row. But IMO they will more or less sound harmonically equal. You can play with the density of "chords" and the octaves where the tones go, etc. but it will all sound atonal. All the same by a certain reckoning.

Rhythm, tone color, dynamics: those are especially important aspects.

I think the usual approach in 12 tone is pointillism (or punctualism) instead of traditional melodic contour.

Of course with matters of style there are varying degrees. Alban Berg was a 12 tone composer considered more traditional than Schoenberg. If these are not familiar to you, try comparing Berg's Violin Concerto - which is relatively lyrical - with Schoenberg's Serenade - which sounds more frenetic. You can hear there is a range of possibilities in 12 tone style.

...pleasing to the ears

If you really want 12 tone music, I think you don't want to get trapped thinking about how to soften traditional dissonance. You can try to handle the tone row to 'soften' dissonances, but I think that is just swimming against the current. If it's 12 tone, it's gonna sound 12 tone harmonically and melodically.

What other musical resources can be used to make it 'pleasing to the ear?' Pleasing in what way? Whose ears? Yours? What audience? In the end, you have to explore that and come up with your own answer.

Answer 4

What 12-tone music have you listened to? Do you like it?

What's your favourite passage of Schoenberg? Study it, see how he gets his effects.

Or maybe you just agree with a whole lot of musicians that serialism - though an interesting and maybe necessary reaction to centuries of Common Practice writing - doesn't actually sound very good.

Answer 5

The chromatic scale is a kind of superset of diatonic scales. It is therefore, by design, a tool for creating tonal music in a flexible way. Because of this, there is arguably a self-contradiction in the idea of 12-tone technique in that it aims to get away from tonality, and yet aims to do so using a musical framework that has evolved with tonal music in mind.

This means any 12-tone composition has a high risk of falling between two stools - you might end up with something that neither fully escapes the harmonic relationships present in tonal music, nor takes advantage of them.

In my opinion - and this is only an opinion - the most pleasing 12-tone music does not seem (on listening) to attempt to escape tonality fully. Rather, it remains an exploration of the harmonic relationships in the universe it inhabits (the chromatic scale).

If you really want to escape as many aspects of tonality as possible, I would suggest stepping away from the chromatic scale. Come up with your own sets of frequency ratios and use those as a framework for working in the pitch space.