Franck's Violin Sonata - which recording is best?

César Franck’s Violin Sonata is a complex piece, and Caroline Gill finds that the numerous recordings fall into two broad camps

There are a number of composers of the French high-Romantic period whose violin sonatas represent some of their greatest artistic achievements: Saint-Saëns, Ravel and Fauré all crystallised their most mature musical ideas into their violin sonatas (many of them writing only a single example). But if one alone of those was not a slow burn, it was the Sonata in A by César Franck. From its first performance (1886) it was lauded and magnified, even though it was given on a borrowed violin and a hotel piano at the wedding of the composer’s friend Eugène Ysaÿe.

The ‘proper’ premiere (also 1886) was hardly less incongruous: also with Ysaÿe, it was given at the Musée Moderne de Peinture in Brussels, in light that was so dim that most of the piece had to be played from memory. Vincent d’Indy, a devoted supporter and pupil of Franck, who chronicled his experiences with him and was present at this performance, described the fading light of the scene: ‘The public was requested to leave, but…refused to budge. Ysaÿe was heard to strike his music stand with his bow, exclaiming “get on, get on”. And then, unheard-of marvel, the two artists, plunged in gloom in which nothing could be distinguished, performed the last three movements from memory… Music, wondrous and alone, held sovereign sway in the darkness of the night. The miracle will never be forgotten.’

Two schools of thought?

Ravel maintained that there were two schools of musical thought before the wars in France: those that aligned themselves with Debussy, and those that followed Franck. In a letter of 1906 he referred to this latter group, which he called ‘scholistes’, as ‘morose followers of a form of neo-Christianity’, only fairly shortly afterwards singling out Franck himself for particular criticism for treating that ‘scholisme’ as a form of dry intellectualism that robbed music of any proper feeling. He complained that all the elements of music in Franck’s work were elaborated separately and united by nothing other than technical exercise, rather than being the product of a simultaneous and instinctive idea. How wrong could Ravel have been (and later Debussy, who also waded into the argument).

It’s difficult not to see the violin sonata as musical poetry of great maturity and depth, written at a time when Franck was still, even into his early sixties, labouring under his deep love for (or, at least, unappeased sexual obsession with) his much younger pupil Augusta Holmès. It cannot be ignored that the majority of his greatest work – late bloomer that he was, despite his father’s hopes for the child prodigy he had attempted to bully into being – was written in the final 10 years of his life. Whether that is to do solely with his continuing artistic development is impossible to know, but it is unassailable that his passion for Augusta (beloved of many of his circle, Saint-Saëns included) informed a lot of that later music – the Piano Quintet in particular.

The Quintet is, though, more raw and emotional than the sonata, which has the poise and elegance of a later work, even though it’s only slightly later than the Quintet. There is much radiance in this sonata – especially in the ravishing final movement, where all Franck’s extraordinary ideas, borne out of a single, tiny motif, come together. It is a paean of such beauty that it is harder to see it as a work borne solely out of human emotion, like the Quintet, than as a joyous equilibrium of emotion and intellectual endeavour.

The complex and minutely wrought structure of the piece is such that – even though the listener isn’t constantly made self-consciously aware of it – its ideas are running beneath the surface in a constant metamorphosis, at the same time as driving the music in a way that makes it stand up to even the most repeated of listenings. And that same complexity makes it impossible for the piece’s lyricism ever to descend into sentimentality. It is a supreme example of head and heart ruling one piece in equal measure – and performers have been in disagreement since the earliest recordings over which faction to give most credence. Very few find a balance, and it is interesting that some of the best recordings remain some of the earliest still available.

Distinguished early recordings

Lola Bobescu and Jacques Genty in 1951 give a performance of force that is nevertheless evenly balanced, with reciprocity that is hard to find elsewhere. Josef Suk and Jan Panenka (1967) are also surprisingly intimate and unmannered, and very appealing as a result. Mischa Elman, recording with Joseph Seiger at the end of his career in 1955 for Decca, was born in the year after Franck’s death and worked well within the lifetimes of composers directly influenced by Franck, thus carrying with his performance a heritage that is unassailable. Seiger is regrettably more of a passenger than is appropriate for this duo sonata, but Elman’s beautiful warmth of tone (though unfortunately not free of tuning issues when he’s playing in a very high tessitura) creates an intimacy that means that although the movements feel disjointed in terms of structure and relay, there’s nevertheless a strong sense that he is telling a story. There is great artistry in both his playing and his relationship with Seiger, as can also be said of Itzhak Perlman and Vladimir Ashkenazy in 1969 – a recording that, despite a lack of fluidity in the transition between movements and boomy acoustics that don’t protect the sound enough, is one of great authority. Elman never descends into self-indulgence, but displays extraordinary subtlety of articulation, phrasing and even portamento – which are all varied and draw in the listener repeatedly. Elman was overshadowed by Heifetz, and his tone and style were unfashionable for much of his life; but the authenticity he applied to his playing is particularly apparent here, and, surprisingly, his tone and style have aged much more gracefully in comparison with Heifetz’s.

This has much in common with the 1966 recording of Christian Ferras and Pierre Barbizet, which carries an enormous intensity without any hint of tasteless displays of emotion. There is something dreadfully sad about this recording that brings out a sense of pathos that is constantly hovering on the sidelines of this piece, but is rarely invoked with any success. The combination of the Schubert Ensemble’s first violinist Simon Blendis and William Howard in the 2003 recording for Champs Hill creates the same kind of effect, similarly through the ghostly change of tone they adopt when oscillating between statements of musical ideas alternately in the major and minor key (a technique emblematic of Franck’s style, and used frequently in this sonata). Both Blendis and Ferras are steadfast in their support of Franck as a melodist of the most unique type, and thereby they keep their phrasing safely on the right side of heavy-handed: the stately manner in which they play is noticeable in how it opens up the power of the piece.

Fathoming Franck’s intention

It can be difficult to tell what Franck was looking for in any of his music. He was a famously gentle soul, so grateful for any public performance of his work that he would silence anyone making criticisms of any sort as ‘too harsh’; but that is not to say that he was not entirely committed to the music itself. He was an enormously erudite teacher, known to be dedicated and constructive, but if he felt that a pupil had failed to show adequate intensity of effort he would refuse to mark their work. That, combined with his producing four masterpieces, not only of his own, but also of French music in general (the D minor Symphony, the String Quartet, the Three Chorales for organ and the Violin Sonata), in the final years of his life, suggests that to consider only the ‘scholisme’ in his music is to miss the point.

Two modern recordings that choose to follow the emotional route with varying degrees of success are those of Vadim Repin and Nikolai Lugansky (2010) and Joshua Bell and Jean-Yves Thibaudet (1988). Bell adopts a sound and pose with the piece that is noticeably informed by Ysaÿe, and although there is much sense to be had in that, the fundamental basis of the work is to be found – as with the quintet, the quartet, the chorales and the oratorio Les béatitudes – in the music of Beethoven, recreated with supreme imagination by Franck. As a result, the showman position adopted by Bell for the earlier of his two recordings (there is a later one, with Jeremy Denk, made in 2010) is less effective than the almost counter-intuitive introspection of Repin’s performance with Lugansky. It remains a particularly emotional performance, but at the same time holds back so much of Repin’s natural vibrato that it does not overpower – instead creating an overwhelming performance more by stealth than glaring statement.

The 1977 recording of Kyung-Wha Chung and Radu Lupu, however, has an extraordinary fluidity that allows the cyclical ideas of the piece to modify, contort and transform from one to the other with an ease that illustrates with painful beauty the inevitability behind the growth of the musical ideas of the piece. No restatement of a theme is the same, though those tiny shifts are very subtle. There is great fluidity, too, in Lupu’s accompaniment, which ebbs and flows with Chung in a perfect reflection, thus creating great climaxes in the phrasing which completely negate the need to overplay or emphasise the point beyond what’s already written into the music. In contrast, and disappointingly, the live recording of Ruggiero Ricci and Martha Argerich at Carnegie Hall in 1979 starts promisingly but is carried away in the heat of the moment(s), with the tuning (and even, occasionally, the notes themselves) going so awry it is difficult to defend its listenability, even within the framework of what is an overall challenging artistic experience.

The two camps

The balance of this sonata as a piece of intense, Romantic writing on the one hand and carefully wrought polyphony (the canon that forms the final movement is almost as intricate as a Bach fugue) on the other, means that its recordings can fall into two categories: the unintentionally glib, and the overplayed. Performances in the overplayed camp rarely tip over into the magical piece it can be – they largely start out almost matter-of-fact, in a representation of the intellectual side of Franck’s character that was so important. Jennifer Pike and Martin Roscoe, recording in 2010, use the most easily overplayed movement, the second (so fiery that you can be fooled into thinking you haven’t heard any of its material before), to make this point most clearly. Many performances present the music in this movement with a constant accelerating and decelerating surge that is not justifiable by the score, but Pike’s approach is much more deliberate. Rather than making any speed changes at all, she simply changes the articulation, placing an infinitesimal gap between the last two notes of each phrase, which, in turn, propels it into its repeat. It makes her sound considered and modest in a manner that, after many highly emotional performances, is a refreshing relief, particularly in the open-faced simplicity of the vital third movement.

And vital the slow third movement is – as much of the success of any performance hinges on it. It is where the intervals that Franck introduced in the prelude-like first movement, and developed beyond recognition in the heavily disguised scherzo and trio of the second movement, stretch out and blossom. It can range from a breathless staticity to a rippling watery effect to the sinuous bloom of an opening flower. All these approaches can be attractive, but only work if they are presented within a framework of perfect tuning and calm phrasing.

The live recording made at the 1978 Salzburg Festival by Leonid Kogan and Nina Kogan feels the phrasing and the direction, and plainly understands it, but the actual sense is not there, creating a glibness that is difficult to ignore. The 1985 performance by Shlomo Mintz and Yefim Bronfman, though, is almost unequalled as an example that propels itself forwards simply by virtue of the notes on the page. There’s a warmth of tone all the way through the build-up of this crucial juncture of the piece – where the music starts to come out of the complexity of its development and propel the listener towards the radiance of the final movement. It also stands in stark contrast to the 1995 recording of Anne-Sophie Mutter and Lambert Orkis, in which the ravishing central phrase is swooped over with such extreme vibrato that the bow is almost bounced off the string, leaving the sound so undermined that it is impossible to be comforted by it. Mutter even occasionally changes the rhythm slightly on a restatement (although far without the bounds of rubato), super librum, to make her point – in stark contrast to the manner in which, for instance, Chung goes about it, and there is rarely a sense that Orkis, an artist of profound thought, is more than a hitchhiker on Mutter’s journey. Not so in the 2014 recording by Renaud Capuçon and Khatia Buniatishvili, which is not so much driven by the assertive accompaniment as held together as a completely credible whole by it. Buniatishvili communicates with Capuçon’s calm self-possession as the musical ideas are passed between the two with sensitivity as much as intent to emphasise the role of the accompanist.

A triple from Dumay

Augustin Dumay’s second phrase of the opening movement isn’t as ghostly in his 2012 recording with Louis Lortie as it is in his previous two recordings, but he nevertheless brings about a contrast by intensifying its sweetness. In all his recordings, Dumay plays like a classical violinist of the concerto school, and his multiple recordings of this piece are dramatically informed by his accords with the various partners with whom he has worked. In the performance with Lortie it’s hard to escape the sense that it is the Dumay Show, and that the slightly combative relationship they have is not what the work needs. The final section of the second movement, on this basis, becomes relentless fairly early on, and the sense of duality that has been lost by that prevails over the rest of the performance. All the tempo changes are driven by the violin, with no sense of mutuality, which is conspicuous and stressful. The second movement requires the pianist to step up, even if he hasn’t previously and doesn’t subsequently, and there is no question that Lortie does this admirably – but Dumay doesn’t let him get a word in edgeways, and that is frustrating.

Far less frustrating, though, is Augustin Dumay’s sublime 1995 recording with Maria-João Pires. Here, there is a far greater sense of reciprocity, and no feeling at all that the players are locked in a competition. It is more a meeting of the minds of two great artists who are at all times engaged on an intellectual as well as emotional level. There is all the beauty of Dumay’s tone to be found here (as there is also in the recording with Lortie), but there is less stress and less ruthlessness.

It is, though, in Augustin Dumay’s earliest recording, from 1989, with Jean-Philippe Collard, that pianist and violinist are pitched with the greatest equality. Although the violin is recorded more closely than the piano, there is never the sense that Dumay is pushing himself too far forwards: there is a constant, knife-edge balance between the two instruments that supports the overall construction of the sonata and never loses track of how a single germinative idea is the expressive basis of the entire musical cycle. Further, though, there are passages of great subtlety that give life to the unusual cadences and Gregorian chant-style modality that perfume Franck’s style – something that is missing in so many other recordings.

Until he wrote his Violin Sonata, Franck had lacked pure lyricism in his melodic writing – it had only shown itself a tiny amount, even in the slow movement of the quintet. With the Violin Sonata, though, he finally found a perfect virtuous circle of lyricism, melodic beauty and technical perfection, and created a piece that has enjoyed almost ubiquitous popularity since its first performance. To find a recording that can sit proud of the enormous catalogue of performances of this piece is to find one that can capture its classical serenity without making it waffle or sound pretentious; apprehend its improvisatory character without losing the sense that everything in it appears out of an action and a reaction, and, most importantly, appreciates that even though Franck was closely acquainted with Classical techniques, choosing to remove himself so far from them does, in the end, make him his own kind of revolutionary, whatever Ravel said.

Best Archive Recording

Ferras; Barbizet
Brilliant Classics

Christian Ferras has a natural tone that suits this work particularly well. It is impassioned, but pure. He also had a direct line back to the in-fighting and pettiness that ran between Franck and his contemporaries – his father was taught by Marcel Chailley, a devotee of Saint-Saëns.

Best Modern Recording

Capuçon; Buniatishvili


The balance of serenity in the violin with the fieriness of the accompaniment, rather than the usual other way round, works beautifully in this recording. The piece makes perfect sense with this equity, and the fact that Capuçon and Buniatishvili communicate so effectively with so little obvious wilfulness is testament to their sophistication, both as individual artists and as a duo.


Dumay; Pires


The glow of Dumay’s tone and Pires’s extraordinary subtlety of phrasing would have made this recording a worthy winner, if it hadn’t been for Collard’s comfort in this repertoire. The final movement on this disc, when you arrive at the end of it, provides a feeling of total expansiveness in its simplicity, with all the radiant breadth of a Sibelius symphony.

Top Choice

Dumay; Collard


The main criticism often levied against this recording is that it can be too subtle at times, but for me the contrasts in it suggest a vocal element to the musical line, which gives it an almost human voice. This is particularly appealing, because Franck was so uninterested in words – it was the music in its purest form that concerned him; and his phrases take on an almost spoken element as a result.

Selected Discography

Date / Artists / Record company (review date)

1951 Bobescu, Genty Testament SBT1360

1955 Elman, Seiger Testament SBT4 1344

1966 Ferras, Barbizet Andromeda ANDRCD5140; Brilliant Classics 93791; DG 480 6655 (12/67R)

1967 Suk, Panenka Supraphon SU4075-2 (11/68R)

1969 Perlman, Ashkenazy Decca 475 8246DOR (5/69R)

1977 K-W Chung, Lupu Decca 421 154-2DM (9/80R, 1/89); 460 006-2DM; (20 discs) 478 7611DB20

1978 L & N Kogan Orfeo C657 051B (2/06)

1979 Ricci, Argerich Etcetera KTC1038 (3/86R, 9/86)

1985 Mintz, Bronfman Brilliant Classics 94160 (9/86R); DG 477 5448GTA2 (9/86R)

1988 Bell, Thibaudet Decca 475 6709DF2 (11/89R)

1989 Dumay, Collard EMI/Erato 381783-2 (2/90R)

1995 Dumay, Pires DG 445 880-2GH (10/95)

1995 Mutter, Orkis DG 445 826-2GH (12/96)

2003 Blendis, Howard Champs Hill CHRCD004 (4/06R)

2010 Pike, Roscoe Chandos CHAN10667 (6/11)

2010 Repin, Lugansky DG 477 8794 (3/11)

2012 Dumay, Lortie Onyx ONYX4096 (6/13)

2014 R Capuçon, Buniatishvili Erato 2564 62501-8 (11/14)

Gramophone Print

  • Print Edition

From £67/year


The Gramophone Digital Club

  • Digital Edition
  • Digital Archive
  • Reviews Database
  • Events & Offers

From £90/year


Gramophone Reviews

  • Reviews Database

From £67/year


Gramophone Digital Edition

  • Digital Edition
  • Digital Archive

From £67/year


If you are a library, university or other organisation that would be interested in an institutional subscription to Gramophone please click here for further information.