1 Schoenberg, Arnold, ‘Composition with Twelve Tones (I)’ (1941)Google Scholar. in Style and Idea: Selected Writings of Arnold Schoenberg, ed. Stein, Leonard with transl. Black, Leo (London: Faber, 1975), p. 215 Google Scholar.

2 Webern, Anton, letter to Reich, Willi, 6 August 1943, in The Path to the New Music, ed. Reich, Willi, transl. Black, Leo (London: Presser, 1963), p. 64 Google Scholar.

3 Webern, p. 18.

4 Perle, George, Serial Composition and Atonality: An Introduction to the Music of Schoenberg, Berg, and Webern, 3rd ed. (Berkeley: University of California Press, 1972), pp. 111–116 Google Scholar.

5 Krieger, Georg, Schoenberg Werke für Klavier (Göttingen: Vandenhoeck & Ruprecht, 1968), pp. 98–106 Google Scholar.

6 Krieger, p. 101.

7 Krieger, p. 104. See also p. 106.

8 Graebner, Eric, ‘An Analysis of Schoenberg's Klavierstück, op. 33a’, Perspectives of New Music 12 (1973–1974): 128–140 CrossRefGoogle Scholar.

9 E.g.: On pp. 129–130, the decision as to which section m. 29–32 belongs in, is reversed several times. If the form had not been unjustifiably called ‘ternary’ in the first place, the problem need never have arisen. On p. 132, m. 23–25 are adduced to show ‘that the movement to the second hexachord can be performed entirely by semitonal steps’. But the appended table of such steps shows: (1) that G and A♭ have somehow been reversed, spoiling the scheme, and (2) that E♭ moves to E, which it does not do in the same voice in the music. On p. 139, the description of a supposedly elegant group of relationships is inexplicably marred by the phrase: ‘except for the non-corresponding E’.

10 Schoenberg, p. 226.

11 Babbitt, Milton, ‘Twelve-Tone Invariants as Compositional Determinants’, Musical Quarterly 46 (1960), p. 247 Google Scholar.

12 Perle, p. 112; Krieger, p. 102. Graebner, p. 128, gives it in general form: 0 7 2 1 11 8 3 5 9 10 4 6.

13 Cf Krieger, p. 103: ‘However, it would be justifiable to consider the three-note chords to be more important, since actually, except for the Motto, the four notes are never sounded together’. (My translation).

14 Perle, George, Twelve-Tone Tonality (Berkely: University of California Press, 1977), p. 23 Google Scholar. The footnote states: ‘This section is taken from Serial Composition and Atonality, 2nd ed. (1968), pp. 104–106’. As pointed out above, Perle's actual analysis of Opus 33a employs the opposite option, being entirely based on a row.

15 This actually happens in m. 28–29.

16 Babbitt, p. 247. See also the second sentence in the previous quotation from Perle.

17 Cf. Graebner, p. 136: ‘…it would hot be perverse in the present connection to think of m. 1–2 as up-beat to m. 3–4’.

18 Cf. MacOonald, Malcolm, Schoenberg (London: Dent, 1976), p. 163 Google Scholar: ‘Opus 33a, for instance, announces its row not as a melodic idea but as a lush succession of four-note chords, which are only given melodic identity towards the end of the piece’.

19 Krieger, p. 103, also describes major and minor triads but one fails to find them in the music.

20 In this example, many dynamic markings have been erased for the sake of legibility.

21 In the light of this scheme, several of Graebner's explanations of relationships would seem to be unnecessarily complicated, e.g.: the ‘interchordal connections’, p. 133; or, ‘Measures 8–9 can be described as a variation on m. 3–5 backwards’, p. 135.

22 See Perle, , Serial Composition and Atonality, p. 113, n. 2Google Scholar.

23 Perle, p. 115 suggests that since the missing notes are the fifths from the beginning of the row, their omission ‘gives this interval additional weight upon its eventual appearance in bars 23–24’. But there are several other fifths in m. 20–22.

24 See Graebner, p. 130, n. 4.

25 Graebner's attempt, on pp. 130–131, to establish a corresponding transition from ABC to WXYZ is less convincing. He is not sure whether the overlapping occurs in m. 6–7 or m. 8–9. But neither is relevant: the transition occurs at m. 13–14.

26 Cf. Krieger, p. 103, where he describes seventh-chords present in the work as ‘foreign bodies’.

27 See Krieger, p. 105. Graebner's, ‘dyad-pair pattern’, p. 134 Google Scholar, is merely one instance of this, whose specific simultaneities are never repeated. Krieger's, ‘bitonal complex’, p. 103 Google Scholar, is surely an abuse of the term.

28 Aa and Cc are the same pitch-class set transposed a tritone. Graebner states, p. 135: ‘…m. 8–9 can equally be regarded as T6 of m. 3–5 (disregarding intra-tetrachordal ordering)’. He need not have looked so far: m. 5 (Aa) is T6 of m. 3 (Cc).

29 Krieger, p. 99, calls the fifth the ‘characteristic sonority’. In Perle's, analysis, it is a basic premise that the fifth is ‘the dominating intervallic relationship’ (p. 115)Google Scholar, and that the fifth-chords, the two fifths at the beginning of the row, the transpositions of one and two fifths, and the employment of I5 and RI5, are all related.

30 See Perle, p. 115, Krieger, p. 105.

31 Cf. Perle, p. 115. He describes ‘integrative rhythmic procedures’ but cannot identify any consistent motives.

32 Cf. Krieger, p. 103, suggesting that this has ‘systematically utilized consequences’.

33 Leibowitz, René, Schoenberg and his School: The Contemporary Stage of the Language of Music, transl. Newlin, Dika (N.Y.: Philosophical Library, 1949), pp. 275–276 Google Scholar.

34 Perle, p. 115.

35 Schoenberg, p. 226. (My italics).

36 Perle, p. 113; Krieger, p. 98–99. Graebner remarks, p. 128, that ‘the first big break in textural consistency occurs at m. 14’, but never mentions it again.

37 See: Clark, Timothy Vincent, A Notational Problem in Schoenberg's Opus 33a, Doctoral Dissertation, Brandeis University, 1981 Google Scholar. The author discusses unclear notation in m. 24, which raises doubts as to whether the B-flat is restruck or sustained. He finds ‘a structural principle … which provides a resolution of the notational difficulty’. In our scheme, where the whole point here is the slowing down of the circulation of the twelve pitches, this detail would make not the slightest difference.

38 Cf. Krieger, p. 100, on the subject of m. 30.

39 Graebner, p. 132 and passim.

40 Perle, p. III.

41 See Perle, p. 113.

42 Cf. Cone, Edward T., ‘Analysis Today’, Musical Quarterly 46 (1960), p. 178 Google Scholar. Although he defines a slightly different linear contour, I am indebted to him for the idea.

43 Cf. Krieger, p. 104, pointing out that the melodic line of m. 32–34 RH is repeated in m. 37 RH, but ignoring the problem of its difference from m. 1–2 and m. 10–11.

44 It might be claimed that the fifth is now at the end of the melody: the F–B at the end of m. 36 RH. Not the same thing at all!

45 Krieger, p. 100.

46 Schoenberg, , ‘Composition with Twelve Tones (2)’ (c. 1948), in Style and Idea, p. 246 Google Scholar.

47 Schoenberg, p. 218.

48 Perle, , Twelve-Tone Tonality, pp. 23–24 Google Scholar.