Faculty of Performance, Visual Arts and Communications

School of Music


Cello: Left Arm and Hand - George Kennaway

<p style="         This chapter considers the most basic aspects of the use of the left hand as described by nineteenth-century cellists. In this period, there was some basic common ground concerning the shape of the left hand, with a clear consensus on such simple things as that the fingers curve outwards, the fingertip presses firmly on the string, the thumb pad touches the back of the neck, and the left elbow is low. The last, extremely important, point is the only one which differs substantially from modern practice, and it will be discussed later in this chapter. However, even with the other topics, which to a modern cellist might seem quite obvious, there are several distinctions to be noted. 

This article considers the most basic aspects of the use of the left hand as described by nineteenth-century cellists. In this period, there was a consensus concerning the shape of the left hand, on such topics as that the fingers curve outwards, the fingertip presses firmly on the string, the thumb pad touches the back of the neck, and the left elbow is low. The last, extremely important, point is the only one which differs substantially from modern practice, and it will be discussed later in this chapter. However, even with the other topics, which to a modern cellist might seem quite obvious, there are several distinctions to be noted. 

<p style="         This chapter considers the most basic aspects of the use of the left hand as described by nineteenth-century cellists. In this period, there was some basic common ground concerning the shape of the left hand, with a clear consensus on such simple things as that the fingers curve outwards, the fingertip presses firmly on the string, the thumb pad touches the back of the neck, and the left elbow is low. The last, extremely important, point is the only one which differs substantially from modern practice, and it will be discussed later in this chapter. However, even with the other topics, which to a modern cellist might seem quite obvious, there are several distinctions to be noted. 

Although the need for firm finger pressure on the string was widely acknowledged, there some differences of emphasis. Baillot and Dotzauer said that the left hand finger pressure should be greater than that of the bow on the string (advice which is still given today, even when using metal strings), but several cellists in the later eighteenth and early nineteenth centuries warned about excessive finger pressure and the noise of the finger landing too heavily on the string. Laborde said that the fingers should be rounded:

[…] in order to attack the string, all the time without force or roughness, which is called ‘touch’ [le tact]. One cannot say too often that this is an essential element for playing the instrument well. [‘...afin d'attaquer la corde, toutefois sans force ni roideur, ce qui s'appelle le tact. On ne saurait trop observer que c’est une partie essentielle pour bien jouer l’instrument’. Jean Benjamin Laborde, Essai sur la musique (Paris: Enfroy, 1780), p. 310.]

Dotzauer thought it ‘vicieux’ to have the fingers too high above the string, and Romberg made a rather exaggerated claim that excessive pressure strains the sinews, so ‘that they require whole years of rest before they can again be used for playing’ (Bernhard Romberg, trans. anon., Complete Theoretical and Practical School for the Violoncello (London: T. Boosey & Co., [1840]), p. 97). Although this topic is not often addressed, these sources suggest that the noise of fingers hitting the string was considered unacceptable, and some later teachers seem to have agreed. Herbert Walenn appears to have taught a ‘soft’ left hand technique at the turn of the nineteenth century, as four of his pupils continued to teach it well into the twentieth. These include Valentine Orde, Michael Edmonds, William Pleeth (with all three of whom the writer studied in the 1970s and 80s), and Zara Nelsova. Nelsova eventually rejected the ‘Russian’ approach, with its high left hand fingers and some percussive noise: ‘I learned later that this isn’t the way to create perfect articulation’ (Zara Nelsova, interview with Tim Janof, 2000 - formerly at http://www.cello.org/Newsletter/Articles/nelsova.htm but not currently available) ]. On the whole, the ‘hammered’ approach to the action of the fingers was discouraged.

The position of the left thumb

The placing of the left thumb is variable. Most tutors recommend that it lies somewhere between the first and second fingers, but a substantial minority (Bideau, Baillot, Schetky, Crouch, Romberg from the first half of the nineteenth century, and de Swert later) put it opposite the second or even between second and third fingers. Jules de Swert, uniquely, tells the cellist to place the thumb round the neck in fourth position so that it touches the ribs of the instrument (The Violoncello, London: Novello, Ewer & Co., 1882, p. 38). In an illustration, Crouch shows the thumb much further round the neck than most, in the manner depicted in the previous century, and between second and third fingers (Compleat Treatise on the Violoncello, London, Chappell & Co., [1826], unpaginated plate).  

Most tutors recommend keeping the fingers down on the string as much as possible – Eley even gives a substantial exercise for this (Improved Method for the Violoncello, London: Clementi & Co., [1827?], p. 79). Dotzauer is a little more flexible on this point, saying that in some circumstances one need not keep all the fingers down in a short rapid slurred group of stepwise notes, saying ‘It is not vitally necessary for the fingers to remain fixed to string in passages such as these’ (Méthode de Violoncelle. Violonzell-Schule,  Mainz: B. Schott fils, [1825], p. 5). 

The most important aspect of the left hand is its basic shape, and on this issue there are several different views. A small minority of nineteenth-century cellists seem to have adopted a ‘violinistic’ left hand, with the fingers slanted backwards at an extremely oblique angle to the fingerboard and the thumb towards the opposite side of the neck rather than directly beneath it. Romberg is very clear about this in his illustration (figure 1) (Violoncellschule, plate facing p. 6).

Figure 1: Romberg, Violoncellschule,left hand

His left hand leans well back rather than being perpendicular to the fingerboard, and the left elbow is dropped. This is not merely careless drawing, for Romberg also gives a uniquely detailed description of this type of left hand:

The hand should so hold the neck, that the 1st finger should clasp it round, the 2nd should be bent so as to form three sides of a square, the 3rd should be bent half round, and the 4th held straight. The thumb should lie exactly opposite to the 2nd finger […] The palm of the hand should […] be kept hollow, nor must the hollow of the thumb be pressed close to the 1st finger. The neck also must remain quite free in the hand […] [Op. cit., p. 7.]
The thumb must […] fall exactly opposite the 2nd finger. The third joint of the first finger ([…] the joint next the hand), should be laid upon the neck of the Violoncello. The fingers should be held at the distance of at least a thumb’s breadth above the strings, and all of them curved, except the fourth, which should be held straight, but not further removed from the strings […] To [play] B on the A string, the 1st finger (still curved) should be pressed down, without disturbing the position of the other fingers [...] [Op. cit., p. 10]

Romberg’s description of the first finger is exceptional to the point of idiosyncrasy, but his illustration reinforces the point. Keeping the lower joint of the first finger in contact with the neck drastically limits its movement, and makes even occasional ornamental vibrato very difficult (which may explain why Romberg barely discusses it, limiting it in practice to the second finger).

Some cellists played with, and taught, the slanted left hand position, such as Tricklir (1750-1813), Janson (1742-1803), Romberg (1767-1841) and Vaslin (1794-1889) (Valerie Walden, One Hundred Years of Violoncello A History of Technique and Performance Practice 1740-1840, Cambridge: Cambridge University Press, 1998, p. 100). Georg Banger’s Méthode of 1877 also shows a highly slanted left hand with a very similar illustration to Romberg’s but no explanatory text (Georg Banger, Méthode pratique de violoncelle Praktische Violoncell-schule, Offenbach am Main: André, [1877]). However, given the relatively early dates of the first two, the frequently old-fashioned nature of Romberg’s treatise, Banger’s marginal status (he composed several cello works in the 1860s and 1870s, but is not reviewed in the principal journals of the period), and Vaslin’s generally highly idiosyncratic approach, they do not constitute a united group or school, and there are no grounds for suggesting that this left hand shape was in significant use in the 19th century. Vaslin’s case is, nonetheless, interesting. Like Romberg and Lindley, Vaslin did not set down his teaching ideas on paper until near the end of his life, in his case at the age of ninety. He enthusiastically promoted the slanted left hand, which he had first adopted in 1809  aged fifteen. But he used it mainly as a means of compensating for a double-jointed third finger.

[In the Orchestre des Varietés] I met M. Ropiquet père, a modest violinist  […] He was struck by the weakness of my left hand, and he had little trouble in making me understand that the size of this large instrument need not preclude the logical, rational principles of the small one, of the same family. So I abandoned the position of the thumb relative to the second finger in order to obtain this end, that of fingers which held on to the string. I had at the same time to work on the difficult correction of a third finger whose nature was to flex […] [‘…j’entrai à 15 ans à l’Orchestre des Varietés. […] Là je rencontrai Monsieur Ropiquet père, modeste violoniste […] La défectuosité de ma main gauche le frappe il n’eut pas de peine à me faire comprendre que la dimension du gros instrument n’était pas une raison d’exclure les principes logiques et rationnels du petit, son congénère. Donc j’abandonnai la pose du pouce vis-à-vis du second doigt, afin d’obtenir que ce fut le bout, mais bien le bout des doigts qui portât sur la corde. J’eus en même temps à opérer la rectification pénible d’un troisième doigt dont la nature était de fléchir […]’.  Olive Vaslin, L’art du violoncelle (Paris: Richault, 1884), p. 1.]

Figure 2 shows him using this left hand shape, in an illustration dating from shortly after Romberg’s death. In this illustration he is not in the act of playing, but in his treatise he describes how this left hand shape is that by which one takes hold of the instrument:

It is easy to understand and obtain if one can carefully keep the hand in the position which it takes to grasp the neck at the moment of placing the cello between the legs […] the thumb goes around the neck and the fingers are found to be arranged curved outwards, offering only the tips to the strings. [‘Ceçi est facile à comprendre et à obtenir si l’on veut bien conserver à la main la position qu’elle prend pour saisir le manche au moment de placer le violoncelle entre les jambes […] le pouce embrasse le manche  et les doigts se trouvent tous disposes à se placer en arc-boutants et par conséquent à ne présenter aux cordes que leur extrémité’. Vaslin, op. cit., p. 3.]

Figure 2: Emil Lassalle, lithograph portrait of Olive Vaslin, 1842 (Bibliothèque national français, Richelieu Musique fonds estampes Vaslin)

John Gunn had already dismissed the slanted left hand in the first edition of his tutor (c.1787), saying that the modern hand has a ‘great advantage […] over that formerly in use’ (John Gunn, The Theory and Practice of Fingering the Violoncello (London: the Author, 1st edn. [1789]), p. 61). He put the point even more firmly in the second edition:

The position […] formerly much in use, and originating probably from the position of the hand on the Violin, in which it is the best practicable, is given as a beacon to avoid; the fingers tending to an oblique direction, as expressed by the dotted lines, cannot be corrected without very long practice…  [Gunn, Theory and Practice, 2nd ed. [1793?], p. 6. Emphasis added.]

Gunn’s illustration of the violin hold is very similar to Romberg’s, especially in the extreme curling of the first finger (figure 3). Something very like this hand shape is seen in Gainsborough's portrait of the Rev. Chafy from 1750-52 (figures 4, 4a), where the player's left thumb is just visible.

Figure 3: Gunn, Treatise, LH posture – left to right: closed, extended, and the violin hold

Figure 4: Thomas Gainsborough, The Rev. John Chafy Playing the Violincello in a Landscape (circa 1750-52), Tate Gallery

Figure 4a: The Rev. Chafy, detail of LH showing thumb

Duport had likewise rejected the violin hold:

By holding the hand faultily we mean the manner and habit of holding the neck of the instrument as is done with the violin, in the palm of the hand; by this the fingers are shortened, and the stretch from the 1. to the 4. finger […] will be found almost impossible […] Those therefore who have adopted this vicious manner, are obliged continually to change the holding of the hand, even while playing one and the same position. [‘Ce que nous appelons mauvaise position de la main, est d’empoigner le manche comme on fait sur le Violon, cela raccourcit les doigts et rend presque impossible, l’écart du premier au quatrième […] ce qui fait que les personnes qui jouent avec cette position du main, sont obliges de sauter la main a tout moment, même en jouant la même position …’. Jean Louis Duport, Essai sur le doigté (Paris: Imbault, [1806]), p. 8. The English translation above, from the later trilingual edition, is much more strongly worded than the original. Duport, trans. August Lindner, Anleitung zum Fingersatz auf dem Violoncell und zur Bogenführung (Offenbach: Jean André, Philadelphia: G. André & Co., Frankfurt: G. A. André, London: Augener & Co., [1864]), p. 7.]

Duport and Kummer each give an example of a passage which cannot be played without moving the hand if it is slanted (Duport, Essai, p. 8 ; Kummer, Violoncelloschule, p. x). Duport points out that backward extensions become difficult, giving the notes E flat, F, and G in first position on the D string; Kummer finds that it inhibits the independence of the fingers, especially the second and third when alternating between F and F# in the same position. They fundamentally agree: the violinistic left hand hampers the freedom of the fingers to move, and forces a constant readjustment of the whole hand.

The modern, 'square', shape of the left hand is asserted particularly firmly by some, and quite early in this period – Gunn has already been mentioned. According to Bideau it is vital for good tone:

Execution and accuracy depend on the position of the [left] hand. This point is so essential, that it is necessary to work for a long time before becoming able to place the hand on the instrument. […] One must put the four fingers on the fingerboard, two inches from the nut, rounding them as much as possible. It is essential in order to produce a good sound to press them firmly on the string, and at the tip. One must then place the thumb behind the neck without holding it, and in the middle, so that it is between the middle and ring fingers. [‘De la position de la main dépend l’exécution et la justesse. Cet article est si essentiel, qu’il faut travailler longtems avant que de pouvoir parvenir à fixer la main sur cet instrument. […] Il faut poser les quatre doigts sur la touche, a la distance de deux pouces du sillet, les arrondir autant qu’il est possible. Il est essentiel pour tirer un beau son de les appuyer fortement sur la corde, et de l’extrémité, il faut placer ensuite le pouce derrière le manche sans le tenir, et au milieu, de sorte qu’il se trouve entre le doigt du milieu et l’annulaire…’ Dominique Bideau, Grande nouvelle méthode (Paris: Naderman, [c. 1802]), p. 3.]

In his version of Bréval, Peile manages to criticise the violin hold by implication:

The Learner is then to bring the left hand to the neck of the Instrument, by placing the Thumb without pressure on the back of the Neck, and bend the fingers in an arch like form over the Strings, stretch’d from each other about an Inch, the first joints of which from their points being nearly perpendicular to the strings, which position must be particularly observed, as any other would be bad, that is to say, the hand must be square with the fingerboard. [J. B. Bréval, trans. J. Peile, Bréval’s New Instructions for the Violoncello (London: C. Wheatstone & Co., [1810]), p. 6. Emphasis added.]

Crouch gives a particularly good illustration (figure 5) of the square left hand.

Figure 5: Crouch, left hand (Compleat Treatise, unpaginated plate) 

Significantly, whereas de Swert quoted Romberg’s description of posture without a tail-pin and then went on to say that it was quite out of date, he omitted Romberg’s violinistic left hand shape entirely, without comment, simply teaching the square left hand, implying that Romberg's left hand shape was not even worth discussing (Bernhard Romberg, rev. Jules de Swert and Heinrich Grünfeld, Violoncelloschule, Berlin: E. Bote & G. Bock [1888], p. 4). However, although the violin hold did not last, not everyone advocated a strictly square left hand either. Later in the nineteenth century a third shape appears, somewhere between the two. Carl Schroeder illustrates it well (figure 6).[24] His left hand is somewhat slanted, but not as much as Romberg’s.

Figure 6: Schroeder, Catechism, left hand

Indeed, Schroeder’s fingers look almost as if pointing vertically to the floor, as opposed to perpendicular to the fingerboard. This may well be what Junod intends when he says that the first finger should land ‘perpendicularly on the string’, rather than to the string – in other words, vertically (New Concise Method, p. 3). However, Schroeder’s left hand attracts a dissenting footnote from the translator Matthews, who himself quotes the cellist Edward Howell at some length:

Considerably greater variety exists in the manner of holding and the playing the Violoncello than the violin, and the following observations upon this point by the well known English violoncellist, Mr. Edward Howell, will be read with interest: — “The English (really Duport’s) style consists of holding the fingers stretched out over the finger-board in the first position, with every finger over its proper note in the scale of C [sic – Howell means semitones]. Under the French system, the fingers are not stretched out at all, but are held sloping back as in playing the violin. The advantage of [the English hold] is obvious. The English method […] keeps the fingers and hand always in readiness; the fingers have only to be dropped on the note required and with a large amount of certainty. Moreover, a firmer pressure is obtained upon the strings as the flat of the finger is used. Added to this is the certainty and ease with which the hand can be shifted, and an enormous amount of pressure to be gained when using the thumb. The French style of fingering is illustrated by playing with the tip of the finger, each finger being shifted with each note of the scale. The result of this arrangement, which necessitates the bringing forward of the finger for each note, is a loss of power of grip, and a perpetual glissando effect. The labour of the performer is increased to a large degree, with results scarcely satisfactory, or even pleasing.” If the English method is adopted, the left arm must be held out straighter than as shown in the engraving. Tr. [Schroeder, trans. Matthews, Catechism (1889, trans. 1893), pp. 22-23.]

Although confusingly expressed, this interesting comment shows that a sloping left hand was perhaps more widespread than other evidence would suggest, in spite of the objections given here. Given that a moderately sloped hand need not, in fact, necessitate a continual readjustment of the fingers (see Becker below), it may be that Howell’s criticisms are directed at a more extremely sloped hand than that, say, of Schroeder himself. Howell (a pupil of Piatti) must have known of Romberg’s violinistic hand, since he used Romberg’s method, simplified and drastically abridged, as the basis of his own. However, Howell’s method omits all reference to the violinistic hand, replacing Romberg’s detailed explanation with two simple sentences:

The hand should hold the neck so that the thumb may be exactly opposite to the second finger. The palm of the hand should not be pressed close to the neck, but should be kept hollow. [Edward Howell, First Book for the Violoncello Adapted from Romberg’s School (London: Boosey & Co, [1879]), p. 1.]

It is not clear why Howell should call the violin hold ‘French’, as there is no evidence for this in French cello methods, apart from the special case of Olive Vaslin. Indeed, Duport, whom Howell sees as the founder of the ‘English’ hold, advocated playing as close to the fingernail as possible, and not, as Howell would have it, with ‘the flat of the finger’. The translator’s own addition concerning the different angle of the left arm if using the square hand exemplifies a trend towards holding the arm further away from the body, discussed below.

Diran Alexanian (1881-1954) who also supplies exercises to train the spacing of the fingers, gives a much more detailed version of the slightly sloped left hand (Traité théoretique et pratique du violoncelle, Paris: A. Z. Mathot, 1922, p. 25). The additional photographs supplied by Becker for his revision of Kummer’s method also show this sloped hand clearly, although as with Alexanian it would seem that Becker had large hands and in particular a long fourth finger (figure 7) (Friedrich August Kummer, rev. H. Becker, Violoncelloschule op. 60, Leipzig: C. F. Peters, 1909, p. Vb).

Figure 7: Kummer rev. Becker, Violoncelloschule,photograph of left hand, closed (L) and extended (R)

Becker wrote about this hand shape as well:

It should be noted that the thumb should exert an opposing pressure (Kummer speaks of a “fulcrum”) in a diagonal direction; more specifically: with the fingering on the A and D strings the thumb lies more on the inner part of the neck (thus, under the covered strings); with the fingering on the G and C strings, however, more to the outside. The hand is correctly positioned if the channel created by the finger placement runs across (not parallel, but in a sharp angle to the nails) the fingertips. The first joint of the first, second, and third fingers remains almost vertical on the string. The fourth finger, however, is more extended, due to its shortness. With the fingerings in extended positions (two consecutive whole tones) only the second finger is curved; the two others are extended. As a general rule, one can say that the finger position, wherever practicable, should be curved; but where a longer reach is required, extended. To achieve proper position of the left arm the elbow should be held far enough from the body so that the upper arm creates an angle of 45 degrees with the torso, and the forearm is a direct continuation of the back of the hand. The wrist, consequently, should be neither raised nor lowered. [‘Hierzu ist zu bemerken, daß der Daumen den Gegendruck (Kummer spricht von Stützpunkt) in diagonaler Richtung ausüben soll; besser: Bei den Griffen auf a- un D-saite lege man den Daumen mehr an der inneren Teil des Halses (also unter die besponnenen Seiten), bei Behandlung von G- und C-Seite hingegen mehr an den äußeren. Steht die Hand rightig, so läuft die durch das Aufdrücken der Finger entstehende Rinne quer (nicht parallel, sondern in einem spitzen Winkel zu den Nägeln) über die Fingerspitzen. Das erste Glied des ersten, zweiten und dritten Fingers steht dabei fast senkrecht auf der Saite. Der vierte Finger hingegen wird, seiner Kürze halber, mehr gestreckt. […] Bei Griffen in weiten Stellungen (zwei aufeinanderfolgende ganze Töne) […] wird nur der zweite Finger rund aufgesetz; die beiden anderen sind auszustrecken. Dies bezieht sich jedoch nur auf die unteren Positionen. Als allgemeine Regel mag dienen, die Finger, wo immer angägig, rund aufsetzen, wo großere Spannungen zu bewältigen sind, aber auszustrecken. […] Um eine günstige Stellung des linken Armes zu erreichen, halte man den Ellbogen so weit vom Körper entfernt, daß der Oberarm zum Oberkörper einen Winkel von etwa 45° und der Vorderarm die Fortsetzung des Handrückens in gerarder Richtung bildet. Das Handgelenk darf also weder gehoben noch gesenkt werden.’  [Kummer, rev. Becker, Violoncelloschule, p. x]

Becker shows here that the partly sloped hand need not involve continual readjustment for passages requiring extensions, in the way that the ‘violinistic’ hand did. The importance of this passage lies in the fact that Becker does not advocate a square hand, even for cellists with smaller hands. All those cellists who recommend a moderately sloped left hand place the thumb centrally behind the neck (something similar is true of the illustrations in Maurice Eisenberg’s Cello Playing of Today, London: Novello, 1959, p. 14-15).

Several short Pathé newsreel films from the late 1920s/early 1930s also show cellists using a sloped hand (see films nos. 3254:12,  1612:23,  1163:05,  at http://www.britishpathe.com). This hand shape was advocated in modern times by William Pleeth, among others who explicitly related it to that of the violin (this, for Pleeth, being a good thing). He criticised the square hand as making the fourth finger become ‘bandy-legged in order to reach the fingerboard’ (W. Pleeth, Cello, London: MacDonald, 1982, p. 159). The evidence is clear that the square hand was the most widely advocated throughout the nineteenth century, but that towards the end of this period and continuing into the first decades of the twentieth century there was interest in a sloped hand that has persisted to modern times.

The left arm

Whereas the left hand was employed in a shape, or shapes, still recognisable today, the left arm as a whole, rather like the bowing arm (see ARTICLE), was generally lower. The consistent advice to keep the left elbow low is striking. Many tutors ignore the left arm altogether, concentrating entirely on the placing of the fingers. This is true of basic, cheaper, methods, and most of the anonymous ones, but it applies to Gunn, Duport and Lindley as well. Some illustrations are quite clear, such as Bréval’s, with a dropped left shoulder and elbow and a smooth wrist, and Romberg’s similar picture. Reinagle puts the ball of the left hand close to the neck, which effectively lowers the arm (Joseph Reinagle, A Concise Introduction to the art of Playing the Violoncello, London: Goulding, Phipps, and d’Almaine, [1800]). Crouch, differing from the Paris method, says the upper left arm should be close to the body, and Kummer is quite clear: ‘The left elbow must not be raised’ (Crouch, op. cit., p.7; 'Der linke Ellbogen darf nicht gehoben werden' - Kummer, op. cit., p. x). Cellists in the second half of the nineteenth century move slightly away from this position. Junod puts the arm ‘in an easy position and at some distance from the body’, with the left elbow neither raised nor resting against the instrument (op. cit., p. 3). Carl Schroeder keeps the upper arm and elbow a little away from the body (op. cit., p. 22). Van der Straeten describes the arrangement of the left arm, not in the context of a more or less static position, but one which allows the greatest freedom of movement:

The upper arm should therefore be kept as steady as possible, so as to allow perfect freedom to the left hand and its movements. To find out the proper position of the left hand and arm, stretch out the latter straight from the shoulder. Then stretch out the fingers […] and bend the first and second joints, as if for the purpose of scratching. Now, turn the forearm towards you from the elbow joint, and, without altering the relative distance of the fingers, place their tips on the A string […] the left hand standing almost at right angles to the fingerboard. […] The position of the left arm must of course be modified for comfort’s sake; but on no account should the elbow hang quite down, as that would cause the left hand to turn too much sideways [backwards], and when shifting beyond the fourth position, the arm would have to be brought forward […] If the left hand and arm are placed in the proper manner […] the latter can shift right up the fingerboard without the least change in position of the upper arm. [Edmund van der Straeten, Technics of Violoncello Playing (London: 'The Strad' Office, 1898), pp. 69-71]

Later cellists raise the left arm still more. Schroeder places the upper arm away from the body and not touching it, with a wrist slightly curved outwards (more so on the C string); Rabaud says it should be neither raised nor rested against the cello, and kept some distance from the body, and Becker indirectly contradicts Kummer:

To achieve a proper placing of the left arm the elbow should be held far enough from the body so that the upper arm creates an angle of 45 degrees with the torso, and the forearm is a direct continuation of the back of the hand. The wrist, consequently, should be neither raised nor lowered. [‘Um eine gunstige Stellung des linken Armes zu erreichen, halte man den Ellbogen so weit vom Korper entfernt, dass der Overarm zum Oberkorper einen Winkel von etwa 45° und der Vorderarm die Fortsetzung des Handruckens in gerarder Richtung bildet. Das Handgelenk darf also weder gehobennoch gesnkt werden.’ F. A. Kummer, rev. Hugo Becker, Violoncelloschule, p. x.]

Becker’s version was to become the twentieth-century standard. Schroeder’s outwardly curved wrist looks like an element of an older playing style when compared with Becker’s forearm continuing in the same line as the back of the hand. In this context, Casals’s account of his early tuition with José Garcia (in 1888, at the age of twelve) at the Municipal School of Music in Barcelona may reflect an unusually rigid approach:

We were taught to play with a stiff arm and obliged to keep a book under the armpit! [J. M. Corredor, trans. André Mangeot, Conversations with Casals (London: Hutchinson, 1956), p. 25.]

No cellist in the nineteenth century advocated a stiff arm, even if there was a general view that the left elbow should be kept low, so Casals’s teacher may have adopted an extreme version of this posture.