Bartók, For Children, “Flower Song”

Béla Bartók, For Children, vol. 1, no. 7 “Flower Song” [“Play Song”]. Note: “Flower Song” is the title in the first edition; “Play Song” is the title in the revised edition from the 1940s.

I examine this piece at length in §1.1 of A Guide to Schenkerian Analysis, Early 20th Century: link to the file. The goal there was to analyze the form design as it guided and was integrated with harmony. The end result of the work was a bass-line sketch.

Here we will work with both bass and upper parts, following the same three-step procedure: form analysis, harmonic analysis, sketch. To begin a form analysis, we recognize phrase and phrase-group levels. Phrases are "2" below, phrase-groups are "1". Note that the melody is identical in bars 1-8 & 9-16.

Since this is a simple melody and I am comfortable using William Caplin's formal-function labels, we identify the phrases as forming a sentence: repetition of a 2-bar motive in the first phrase ("presentation") is followed by a sequence-based continuation (Caplin calls that device "fragmentation"). 

Next we analyze the harmony. For more details and an alternative reading that gives more attention to bar 12, see the Guide narrative. Notice how Bartók makes a sharp contrast in harmony--effectively pedal-point tonic in bars 1-8 but then unstable over V in bar 9ff until the final tonic--while repeating the melody exactly.

 In a traditional Schenkerian analysis, we now try to match the hierarchy of the bass with a line or lines above. In this case, middleground lines (bars 1-8) and a background line (1-16) are the only possibilities, though we are obliged to think differently about priority in the third-motive (I picked the upper note in bars 1 & 3, but traced a line with lower notes in bars 5-7): this kind of inconsistency in tracing through voice-leading is routine in Schenkerian analysis).

Here are some aspects we haven't considered yet that would be obvious to an attentive listener and reasonably could influence our reading.

Fred Lerdahl's reduction method is based on local hierarchies of time/rhythm and harmony that are gradually built up over greater lengths. In bar 1, for example, I have removed the off-beat notes, and in the lower system I have prioritized the tonic on the first beat and connected it to the final tonic note. As I said in a previous post, unlike Schenker, Lerdahl doesn't require a specific shape in the final reductive level.

This is the tree diagram version of the system above, with a little more detail.

And this is a prolongational reading that is similar to the above. I have found such figures based on an upper-register ^8 to be surprisingly common in traditional European tonal music, going back to modal musics as early as the 16th century. See my comments on a similar one by Arthur Komar for a Bach Prelude in my Music Theory Spectrum article (2009).

My proto-background model, discussed at length in the same article and in this essay (link to the essay), has the ^8-^9-^8 figure (marked as ^1-^2-^1) but also preserves the obvious figures of the unfolded thirds, so that the first background is ^1/^3: