What did Schubert intend in his accompaniment to Der Leiermann, the last song of Winterreise? A recent recording and two reviews of it raise that question.¹ Schubert wrote out grace-note appoggiaturas for just the first two bars (see the excerpt from the autograph), and most players take a literal approach to the score by continuing with plain fifths in the left hand.

This excerpt from the autograph of Part II of Winterreise, dated October 1827, includes the first few bars of Der Leiermann. The appoggiaturas are clearly shown in the first two bars. The song is given here as no. 12 (i.e. of Part II) and the key is B minor. When Part II was finally published in December 1828, shortly after the composer's death, this song had been transposed to A minor.

Both reviewers criticize the present accompanist, Ulrich Eisenlohr, for his assumption that Schubert implied the instruction simile, meaning to continue adding appoggiaturas for the remainder of the song. I think my offering a dissenting view will not place in jeopardy my friendship with the reviewers.

The first piece of evidence I would bring to bear is obvious: Schubert very often placed certain performance markings (for staccato, accent, portamento, etc.) at the beginning of a song, piece or section, and clearly assumed the player would continue to play simile. He felt no need to give that instruction explicitly. Several examples are to hand in Winterreise: Gute Nacht, Einsamkeit, Die Post, Letzte Hoffnung, Täuschung, Das Wirtshaus. (In some cases certain markings may be intended only for the accompanist's solo passages; Rast might be of that type, but clearly not Der Leiermann.)

The second piece of evidence is equally obvious: that Schubert regularly used the accompaniment to illustrate some feature pertaining to the content of the song's text. Pause from Die Schöne Müllerin provides a clear example, illustrating the blowing of the wind through the strings of a lute. How apt the Leiermann appoggiaturas seem to me in reflecting the sound of the hurdy-gurdy; how lacking seem the plain fifths.

Thirdly, he sometimes represented not only the overt contents but at the same time the state of mind of the subject. Gretchen am Spinnrade was the first such song, reflecting both the spinning and the spinner. In the present case the discords may be taken to represent not only the hurdy-gurdy but at the same time the discordance in the mind of the cycle's subject.

The combination of the above points leaves little doubt in my mind that Eisenlohr was right. But let us consider other objections raised by the reviewers. Michael finds it inappropriate that the cycle would thereby end on a dissonance. However, with appropriate pedalling, as Eisenlohr uses throughout, the final sound is in fact consonant, after its dissonant onset. In this connection, has anyone complained about Schubert's ending the String Quintet D956 on a dissonance? Michael also suggests that we should follow the example of the great accompanists of the past, who did not assume a simile. That seems to me to call into question whether we may think for ourselves while keeping in mind the possibility that something was previously overlooked: without a healthy scepticism and spirit of enquiry little progress could be made, and we would perhaps still believe in a flat earth. Richard finds the discords "at times ... quite obtrusive and distracting", but I find them to be nicely judged in the present performance. Indeed, the dissonances running through Pause might also seem obtrusive until one realises their significance.

A few other points may be noted in Eisenlohr's playing. He replaced the first right-hand note in bar 3 by a rest, in my opinion quite a defensible personal variant. I suppose his idea was to let his appoggiatura in that bar show through clearly so that listeners would know with certainty what he was doing, in case they had not heard that appoggiatura in previous performances. He also quite reasonably emphasized slightly his second left-hand note in bars where it forms a dissonance (bars 17, 18, etc.) and not in bars where it forms a consonance (bars 7, 8, etc.). In various ways Eisenlohr has, then, avoided a literal-minded approach, something which I applaud, for I regard music as one of the greatest opportunities to do just that. Nevertheless, I do not think this performance is beyond criticism in some other ways, especially concerning tempo and ensemble.

The implied simile in Der Leiermann requires, if I am right, a more flexible approach than does one for a simple staccato; if time permitted, I would now look through all Schubert's songs, and his other works as well, for other cases where such flexibility could bring to light an implied simile. I hope an interested reader will take up that research project and let us know the results.

____________________

Footnote:

¹Richard Morris, Review of: "Naxos Deutsche Schubert-Lied-Edition. No. 1: Winterreise. Roman Trekel (baritone) and Ulrich Eisenlohr (piano) Naxos 8.554471", The Schubertian, No. 28, March 2000, p. 19. Michael Lorenz, "Schockierend Kurios...", Schubert durch die Brille, No. 25, June 2000, p. 88.