W. A. Mozart: Fantasia in D minor for Piano (cont'd)

Paradoxes of Tempo Interpretations in the Fantasia

“In Mozart’s own words, written to his father on October 24, 1777, tempo is ‘the most difficult and most important, and the main thing in music’. For Mozart, tempo was by no means the mere equivalent of time.”7 Indeed, the “tempo” problem is crucial and the most difficult interpretive issue in the Fantasia. The tempo markings are quite consistent in the urtext and all other editions. The introduction (first 11 measures) is marked Andante Alla breve, the major section (starting from measure 12) is marked Adagio (there is no indication of a time change, so one can assume it is still alla breve); the cadenza passages (bridges) are marked Presto and the D Major section is marked Allegretto (time 2/4). Score editors avoid any metronome markings in this piece: from all scores I have observed only one (Maurice Hinson’s edition) has exact metronome indications: for Andante a quarter note equals 60, for Adagio56 (I really love this!), for Allegretto120 (when I first saw the Hinson’s tempos, I was very glad that I was not alone in my perception of these tempos). Michael Davidson’s suggestions for tempos: 88, 66 and 108, respectively. Jean-Pierre Marty’s suggestion for the tempo of the Adagio section: a quarter note equals 60.

As one can notice, there is a paradoxical difference in tempo interpretations by different editors (critics), especially for the first section! This can be explained by the statement, “Andante is the most problematic tempo marking in Mozart’s music.”8 The problem of interpreting Andante is intensified even more by the Alla breve time associated with it. “The alla breve sign in Mozart’s music, in addition to pointing to the fundamental pulsation in two, is also the key to a particular reading of the musical notation.”9 Since there is no indication by Mozart of the time change, it can be assumed that Adagio is also associated with Alla breve time. This makes the Adagio section the most problematic in this fantasia in terms of choosing the right tempo (Alla breve suggests that the tempo should be somewhat faster).

As was shown above, the Adagio section consists of the three contrasting episodes (A-B-C), and paradoxically all of them should be played in the same tempo! Apparently, reserving some freedom for other aspects, many performers do try to keep Tempo primo the same and that usually causes some distortion of images: the primary theme (A) is often played too fast in order to keep the following episodes (B and C) from dying tempos; if (A) is played slower, the agitation episode (C) becomes too slow losing its agitated nature which is so essential for this episode. My suggestion would be quite simple: the tempo of episode (C) can be a little faster than (A) (for example, if the starting tempo for Adagio is 56-60 (for the quarter note), the agitation episode (C) can be performed at about 70-75). The difference is not so noticeable, but it makes episode (C) sound much more “appealing.”

Presto for the bridges sometimes becomes misleading too: some pianists consider them just virtuosic passages and play too fast. I am doubtful that Mozart meant these passages to be that fast. I think the quick tempo should not be the most important issue here; most importantly, the music should be very dramatic, with all notes distinctively heard. I am completely in agreement with Michael Davidson’s statement that “this frenzied outburst should not be played so fast that it becomes impossible to hear all the notes.”10 These episodes really require a certain level of maturity from players for they have to restrain themselves from just showing virtuosity and rather express the dramatic essence of all these passages.

Allegretto is definitely the most agreeable and “easy-to-pick” tempo of the Fantasia (probably because the D Major section is in the “regular” classical style). Still, some pianists consider this tempo to be somewhat faster than it is originally assumed to be. Fast tempo, in my perception, obscures the astonishing elegance and gracefulness of the main melody. I can support my opinion by the following statement: "Many reports suggest that Mozart tended to take his allegro movements at a moderate speed. If Mozart wanted a movement played really fast, then he marked it presto or allegro assai. An unadorned allegro carries the original meaning of the word – gay, cheerful.11 If Mozart’s Allegro is to be played moderately, what to say about Allegretto?

There is still another question with tempo interpretation: how steady should the tempo be in this piece? Paradoxically, despite the nature of the genre (fantasy - freedom), there is just one indication of gradual slow-down in the urtext (rallentando) in measure 86, before the recapitulation of Allegretto. (Tempo primo means the return to Adagio after the Presto passages.) In other editions, we can see the short poco rit. at the end of the introduction (Maurice Hinson’s Edition, Alfred Masterwork Edition) and another poco rit. at the end of the cadenza-bridge before the recapitulation of Adagio (Hinson’s edition). Additionally, there is an odd tempo marking in the Alfred’s edition (accelerando!? in a short arpeggio before the end of the Adagio section). Again, I can support my observations by the Michael Davidson’s statement, “…the slightest rubato in the first four measures (of Adagio theme – S.G.) would be sentimental, not tragic.”12 [Bolded - S.G] This obviously means – rhythm and tempo should be kept steady.

After reading the above comments, one may be stunned - why, with almost no agogic deviations, is there a strong feeling of spontaneity in this piece? As Michael Davidson noticed (possibly, with a little exaggeration), “the fantasy seems to evolve haphazardly, almost somnambulistically.”13 Here another paradox appears - Mozart creates a feeling of a spontaneous evolving of music with…fermatas! Does anyone still think the piece is not Mozart’s? Who else would be able to reach horizons of such ingenious paradoxicality?

There are 11 fermatas in the Fantasia (between all major sections of the piece!) – same in the urtext and all later editions. I want to do justice to Michael Davidson who mentioned the role of fermatas in the Fantasia: “the two fermate in measure 11 indicate freedom from the meter14 [bolded - S.G.] Yes, this is where the feeling of freedom comes from! This is where the performer and listener can breathe! This is what I tell my students who work on this piece: breathe in fermata breaks! Take your time to breathe! Take advantage of being able to breathe here! [I cannot resist sharing one interesting observation here. Only 15 years later (1802), in the Finale of his prominent Moonlight Sonata (Quasi Una Fantasia!), Beethoven used fermatas in a similar way (between major sections of the sonata form)! There are fermatas between the primary and the transitional sections in the exposition; between the transitional and the secondary theme sections in the recapitulation; and a very important one – at the end of the most fantasia-like episode in the Finale (transition again) -between the recapitulation and the second development. See Example 8. As in the Fantasia, the use of fermatas creates a clear feeling of slowing down, without any other indication of a tempo change. As in the Mozart’s piece, these fermatas are the “breathing” moments, where the performer can take a short breath in this almost non-stop Presto Agitato movement. (After the cadenza at the end of the second development, Beethoven indicates Adagio instead of fermatas for the two measures precluding the coda that is the other way of making the two transitional octaves longer). This observation is another weight on a scale demonstrating the obvious parallels in applying the dramatic sonata principles in the short Mozart’s fantasia and the large-scale Beethoven’s work.]

It is necessary to bring up another question here – how long should the fermatas be in different sections of the piece? Especially problematic are those above the whole rests, meaning the rest should last for two full measures! (As generally accepted, fermata means approximately double-prolongation of the note to which it is attached.) The answer can be found in measure 11- two fermatas are placed there by Mozart – one above the half note in bass and one above the whole rest (see Example 1, the last measure). If we simply calculate the length of a bass note plus a following half rest, we will get an actual length of this measure equal to one and a half measures. Therefore, the answer for the length of the upper fermata is obvious – it is one and a half measures long. The other place where double-fermatas appear is in measure 43 – end of the agitation (C) episode. See Example 7. Interestingly, in the urtext the upper fermata has two dots one under another! Does this mean that this fermata should be even longer (like in the case of any double-dotted note)? Possibly, but the second dot is omitted in all other editions. In addition to that, I do not think that performers are so exact in the execution of all the fermatas – it mainly depends on everyone’s inner feeling of how long the specific rest should be.

Example 8. Beethoven. Sonata No. 14 (“Moonlight”),
Finale Transition to the Second Development.

Beethoven. Sonata No. 14 (“Moonlight”), Finale Transition to the Second Development

To finish with the tempo problems, one more question needs to be answered – how does the rhythm in the piece affect the final tempo realization? A good starting point would be from the Michael Davidson‘s statement, “during the entire piece, strict observance of the rests is as important as the shaping of the notes15 [bolded – S.G.] Yes, if we take a close look at the score, it becomes obvious that the observance and execution of rests throughout the piece, especially in the Adagio section, is crucially important for the correct interpretation of rhythm in the piece. The Fantasia is quite unique in terms of a number and usage of different kinds of rests. The list of all rests used in the Adagio section (from shortest to longest) is as follows:

  • Sixteenth rest – regular and with the fermata (measure 54; see Example 9);
  • Eighth rest, regular and with the fermata (measure 34);
  • Dotted eighth rest, regular (throughout the main theme);
  • Quarter rest, regular and with the fermata (measure 54; see Example 9);
  • Half rest, regular;
  • Whole rest, regular and with the fermata (actually, breve rest - measure 28; see Example 3).

Example 9. Transition to the D Major Section.

Transition to the D Major Section

Execution of rests (and, in general, a rhythmic accuracy) in the Fantasia appears to be the most challenging interpretation issue, especially in the two episodes – (A) and (C); in these two episodes the most challenging rests to execute are the dotted eighth (episode A, see Example 2) and the sixteenth (episode C, see Example 3). In my view, the rhythmic challenge is the main reason why the piece is much harder to perform than it looks – the ‘hidden’ challenge, invisible at the first glance, becomes evident when the performer is required a very precise counting and execution of all note and rest values.

For instance, in the Episode A many students have problems with execution of the sixteenth notes following the dotted eighth rests – they tend to make them shorter as if they were the thirty-second notes; and vice versa – the last eighth note of the three-note ‘short-of-breath’ motive followed by the quarter rest is tended to be performed shorter (as if it were the sixteenth note). Execution of all the sixteenth rests in the Episode C is quite a challenge too because of polyrhythm in the left and right hand parts: in the left hand part, there are eighth notes throughout the ‘agitation’ episode; in the right hand part, there are sixteenth notes followed by the sixteenth rests. Apparently, students tend to ignore the rests and play the same eighth notes in both hands (this, unfortunately, completely ruins the short-of-breath intonation, the most important motive in the entire episode).

Paradoxically enough, the Presto passages are somewhat ‘simpler’ for the performer who possesses scaling or arpeggio technique – there is no much counting, nor much of the challenging articulation issue, as in the main theme and the agitation episode (the short-of-breath motives should be performed in two different articulation styles – staccato (the first note) and legato (the second and third notes)). Thus, the sections that look hard (quick passages) are simpler in reality than those looking simpler but having some ‘hidden’ (rhythm and articulation) challenges in them. Quite a paradox, isn’t it? (It seems interesting to compare the rhythmic patterns of the Adagio and Allegretto sections. In Allegretto, rhythm becomes more regular, with no extreme challenges as in the Adagio section.)

As was noted above, the rhythmic challenges in the piece affect the tempo realization, namely: the sections with more complicated rhythm (such as Episodes A and C) place the tempo limits on the performance, for there is a “speed limit” at which all the rhythmic and articulation challenges can be well executed. For instance, the main theme should not be performed too fast – otherwise, there is a high chance of loosing some important rhythmic and articulation details there. The same is true about the agitation episode – speeding up over the limit leads to the indistinctiveness of very important interpretative components – namely: rhythm, phrasing, dynamics and articulation.

We are ready now to listen to different recordings and judge them in terms of tempo realizations. As was mentioned before, much like the music editors and critics, every performer interprets tempos differently. This is what I have found from listening with a metronome.

Recording No. 1 (Andrei Gavrilov)

  • Andante – ~55 per quarter note, slows down in measures 7-8, some rubato at the end;
  • Adagio – ~55 (same tempo!), agitation episode seems slow at this tempo;
  • Presto – really fast, recapitulation of Adagio – many rubatos;
  • Allegretto - ~115, seems too slow.

Recording No. 2 (Christian Zacharias)

  • Andante: - 85 – seems fast, with many rubatos;
  • Adagio: ~65; Agitation episode: 75 (faster than (A), (A) and (C) in different tempos!)
  • Presto: fast; recapitulation of (A): many rubatos at the end of Adagio;
  • No Allegretto section! (The only recording without Allegretto!)

Recording No. 3 (Jorg Demus)

  • Andante – about 65, slowing down at the end;
  • Adagio – almost 70, seems a little rushing; Agitation episode – 65 (slower than Adagio!)
  • Prestoseems too fast; recapitulation of Adagio - almost 70, rubato in transition passage;
  • Allegretto125130 – seems too fast.

Recording No. 4 (Lili Kraus)

  • Andante85 (seems too fast), almost no rubatos (slightly at the very end of introduction);
  • Adagio65, episode B – faster; Agitation episode – 65same as adagio (seems a little slow), Presto – fast; recapitulation of Adagio - rubatos, fast transition passage;
  • Allegretto135-140 (seems too fast).

Recording No. 5 (Robin Alciatore)

  • Andante - ~60, rubatos starting from measure 7;
  • Adagio - ~40 (seems too slow); Agitation episode – much faster than (A) (70+);
  • Presto – not as fast as others;
  • Allegretto - ~100 – 105 (seems too slow).

Summary of Tempo Observation

As expected, the recordings demonstrate a variety of different interpretations of this fantasia and a great deal of freedom in these interpretations. This is what Badura-Skoda was talking about: “The individual musician retains a certain freedom which should not be undervalued – within a certain range, any [italicized by me – S.G.] tempo can be artistically valid.”16

The range of tempos for Andante – 55 (recording No. 1) 85 (recordings 4 and 2) - 30! Usually with rubatos and slowing down in measures 7-8 where the harmonic pulse is becoming twice faster (there is no indication in the urtext of tempo change though). For Andante, my FAVORITE RANGE would be ~ 55 – 65 (in my perception, in a faster tempo, the depth and the quality of tone, and the coherence of phrasing are significantly lost).

The range of tempos for Adagio40 (recording No. 5) – 70 (recording No. 3) – 30 again! 70 seems a little rushing, 40 seems too slow, so for the Adagio, my FAVORITE RANGE would be 55 – 65 for episode (A) (main theme) (same as for the Andante).

For agitation episode (C) the range is 55 (recording No. 1) – 75 (recording No. 2) – 20! In my perception, any tempo slower than 70 is too quiet for this episode (the feeling of agitation is lost). For the episode (C), my FAVORITE RANGE would be 70 – 75.

The range of tempos within Adagio (episodes (A) – (C)):

  • Recording No. 1: 55 - 55 - same for both (A) and (C) – good for (A) and slow for (C).
  • Recording No. 2: 65 – 75 - (C) is faster – a little too fast for (A) and perfect for (C).
  • Recording No. 3: 70 – 65 - (C) is slower! – fast for (A) and a little slow for (C).
  • Recording No. 4: 65 – 65 - same for both (A) and (C) – a little fast for (A) and a little slow for (C)
  • Recording No. 5: 40 – 70+ - (C) is almost twice faster! – too slow for (A) and perfect for (C)!


Only two out of the five recordings keep tempo the same in both (A) and (C) episodes (Recordings Nos. 1 and 4). In the other two – episode (C) is faster (Recordings Nos. 2 and 5); and in one (Recording No. 3, unusual) – (C) is slower than (A).

The range of tempos for Allegretto: 115 (recording No. 1) – 140 (recording No. 4) – 25! In my perception, 115 is a little slow, 140 is too fast. For Allegretto, my FAVORITE RANGE would be 120 – 125.

The above observation explains why, from the five recordings, there are none that I would entirely enjoy (my FAVORITE RANGES are different from any of these recordings). This also illustrates how challenging the piece is for any performer (from a student to a world-class pianist), no matter how easy the score might look.

« Previous: Interpretative Paradoxes in the Fantasia


7 Jean-Pierre Marty. “Mozart’s Tempo Indications and the Problems of Interpretation”. Essay from “Perspectives on Mozart Performance”. Edited by R. Larry Todd and Peter Williams. Cambridge University Press, 1991.

8 Eva and Paul Badura-Skoda. “Interpreting Mozart on the Keyboard.” Translated by Leo Black, St. Martin Press, New York, 1962, page 30.

9 Jean-Pierre Marty, page 59.

10 Michael Davidson, page 231.

11 Eva and Paul Badura-Skoda, page 30.

12 Michael Davidson, page 230.

13 Michael Davidson, page 227.

14 Michael Davidson, page 230.

15 Michael Davidson, page 230.

16 Badura-Skoda. Page 27.


March 29th, 2014

Dear Ms. Gorlin,

I so enjoyed reading your analysis of the Fantasy in D minor. Thank you!
Warmest regards,
Beth Levin

January 6th, 2012

Alan Beggerow:

I enjoyed reading your essay concerning the Fantasia in D minor by Mozart. Very well done, very informative! I have linked to your essay from my blog Musical Musings at http://muswrite.blogspot.com/2012/01/mozart-fantasia-in-d-minor-for-piano.html

Thanks for the great read and analysis!

« return to "Music Criticism" « back to top