Giuseppe Tartini » Violin Sonata in G minor, 'Devil's Trill'
Hermann's edition of this sonata, originally issued in 1873, seems to have been the source for Carl Flesch's 1905 recording of the first movement; although Flesch clearly modified a number of Hermann's fingerings, his performance reflects some of them and in general suggests a similar style, and he exactly reproduces Hermann's musical text, which departs from the original in several important respects. Hermann's fingerings appear to have been based upon Joachim's performance of the piece, since they are closely reflected in Joachim's 1905 edition, in which a [footnote](http://mhm.hud.ac.uk/chase/view/pdf/188/15/#page) makes it clear that Hermann prepared his edition after hearing Joachim perform the piece; the footnote refers to the inclusion of Joachim's cadenza in the 1873 Peters edition, which Hermann apparently wrote down after hearing Joachim play it, and which he published with Joachim's permission. Ferdinand David's personal copy of an early edition of the piece, evidently used for performance (since he wrote the word *Umwender* (page turner) on it), contains fewer expressive shifts in the first movement than Hermann's, but relates very closely to it in respect of David's substantial changes to the musical text. David's deletion of many of the original double stops and addition of others (for instance in bars 15 to 17) in the first movement is directly reflected in Hermann's and Joachim's editions, suggesting either that Joachim's conception of the work derived to a considerable extent directly from David or, less plausibly, that David's markings in his personal copy were influenced by Joachim's performance. Alard, in his edition of 1863, retained the double stops in the first movement of the original, though he changed some of the note values. In the other movements of the sonata the connections between the three German violinists are not so neatly sequential. Hermann's edition shows its relationship to David's in several respects and its apparent derivation from Joachim's in others. The successive up bows at the beginning of the final movement evidently reflect a style of performance adopted by Joachim, that appears to differ from David's. The inclusion of a cadenza before the final slow section in the last movement, however, is another connection between them. There is no indication of a cadenza here in the original, and Alard's edition does not contain one. Hermann's cadenza is very similar to the one in Joachim's edition and its derivation from Joachim is confirmed by the footnote in the latter's edition; David's cadenza, is contained in a holograph manuscript attached to his copy. Interestingly, all three cadenzas begin with the same figuration, once more suggesting a palpable connection between the three violinists.
Editions by Vieuxtemps (1855) and Léonard (1865) have still to be acquired.
|Alard, Jean Delphin||Sonate||E. Gérard||1863||View|
|David, Ferdinand||Sonate de Tartini||?||1800||View|
|Hermann, Friedrich||Le trille du diable||Peters||1873||View|
|Joachim, Joseph; Moser, Andreas||Sonate in G mol (Teufelstriller)||Simrock||1905||View|