Beethoven's piano sonatas have appeared in innumerable editions – most of them in more than one hundred, as the collection in the library of the Beethoven-Haus in Bonn reveals. The sources for these works have also never been as readily available as they are now, as most first editions can be viewed on the Beethoven-Haus website, which also hosts scans of many important manuscript sources, as well as links to images of source materials on the websites of other archives. Thus, the question must be asked: Is there any scope for another edition of Beethoven's Piano Sonatas?

In 2007, the Associated Board of the Royal Schools of Music's Beethoven: The 35 Piano Sonatas, edited by Barry Cooper, provided a particularly convincing four-part reply to that question.Footnote ¹ First, as the title indicates, the edition included three early compositions, the so-called ‘Kurfürsten’ Sonatas WoO 47, which almost all previous editions had left out. Second, it corrected a number of significant and long-standing textual errors, and explained the editing process behind these corrections in detailed prose. Third, it took into account research on historical performance practice, relying particularly on Carl Czerny's testimony, albeit with a critical eye.Footnote ² Fourth, it included CDs with recorded extracts played on instruments from the composer's lifetime that helped explain some of these last two points. So, Cooper's ABRSM edition provides a suitable baseline from which subsequent editions of this repertoire can be assessed.

Besides ABRSM, the most frequently encountered publishers of Beethoven's piano sonatas are Henle, Bärenreiter and Wiener Urtext, the last two of which are currently in the process of publishing new editions of the entire set. At the time of writing, Bärenreiter has issued editions by Jonathan del Mar of opp. 2, 7, 10, 13, 26, 27, 28, 31, 49, 53, 57, 78, 79, 81a, 90, 101, and WoO 47, while Wiener Urtext has only published an edition of op. 106 by Jochen Reutter in advance of what will be a complete set. In this review, I will give an overview of the editorial choices in these editions, the material they include on performance practice, and whether they add any new insights for performers and scholars.

Del Mar's editions of the piano sonatas can be seen as a continuation of his wider Beethoven project with Bärenreiter, which has already included new editions of the nine Symphonies, which were published between 1996 and 2000, and the String Quartets, which began in 2007 and now includes opp. 18, 59, 74, 95, and 127, with the remaining four presumably to appear in the near future. In addition, Del Mar produced editions of the Cello Sonatas (2004), the Violin Concerto (2009), the Triple Concerto (2012), the five Piano Concertos (2013–2015), and the Variations for Cello and Piano (2012). Consequently, he is one of the most prolific editors of music by Beethoven currently active. His new edition of the piano sonatas includes the three early ‘Kurfürsten’ Sonatas WoO 47 from 1782–83, making Del Mar the second modern editor, after Cooper, to include them in a complete edition of the sonatas.

Most of Del Mar's volumes begin with an introduction by Misha Donat that both provides historical context and draws attention to interesting compositional features. In the case of the WoO 47 Sonatas this is particularly welcome, given their relative obscurity. Here, Donat persuasively refutes the notion that that these early compositions are merely derivative of the style of Beethoven's teacher Neefe, and shows that they even include some aspects that might be called original. These introductions are most useful to those encountering these works for the first time, but some of Donat's insights are also valuable to more seasoned practitioners. His introduction to op. 53, for instance, helpfully observes that the autograph score shows that Beethoven initially planned to repeat both halves of the first movement, and gives adventurous pianists an opportunity to experiment with an earlier version. For reasons of space, however, Donat's introductions are limited, and a fuller discussion of these sonatas can be found in Cooper's recent monograph on the subject.Footnote ³ In each volume, Donat's introduction is followed by a brief preface explaining the sources and the editorial method by Del Mar, and a section on performance practice by both authors to which I will return later.

The musical text in Del Mar's edition generally has a remarkably clean feel to it: in contrast to Cooper's edition, in which most sonatas include editorial fingerings by David Ward in addition to those that can be traced back to the composer, Del Mar has only included the latter. Furthermore, the editorial work remains largely off stage, barring footnote references to the critical commentary that explain complex editorial decisions and/or draw attention to new readings. Two examples that are particularly emblematic of Del Mar's editorial approach are found in the first movements of op. 2 no. 2 (bar 204) and op. 10 no. 1 (bar 161), two hitherto unproblematic spots at the retransitions of these sonatas. According to Del Mar, in early Beethoven sonata forms there is almost always a repeat of the same phrase at this point in the form, albeit with ‘a small but logical tweak’, but in both cases the first edition (the only source for these sonatas) has one or two divergent notes in the first phrase, which Del Mar suspects are erroneous. Here one might ask how small ‘a small but logical tweak’ really is, but it is impressive how Del Mar's thorough knowledge of the repertoire provides plausible alternative readings for pieces with few extant sources.Footnote ⁴ Another interesting new reading occurs at the end of the second movement of op. 28, in which Del Mar argues, on the basis of a reading of Daniel Gottlob Türk's Klavierschule of 1789, for playing the turn in the right hand with a B# rather than the B$ that is in Cooper's edition.

The presentation, however, of these and other less important finds (suspected missing dynamics etc.) is often unnecessarily circumlocutory, with a footnote indicating that the version in the main text is probably erroneous. This is particularly jarring in the first movement of op. 31 no. 3, where the thematic semiquavers in bar 7 and elsewhere in the right hand were probably intended as triplets, an important observation first made by Cooper. Del Mar presents this finding in a footnote as ‘possibly the correct text’; surely it would have been better to include that reading in the main text, if only to draw more attention to this significant interpretation. Similarly, the abovementioned observation in op. 28 is relegated to the critical notes only. The clean presentation is something of a double-edged sword, since it risks the reader overlooking the most valuable findings.

Perhaps Del Mar's most interesting contribution is the discussion of the sources, which are found at the end of each edition. Although the most important new finds are already mentioned in the footnotes in the text, other perhaps less controversial editorial decisions regarding dynamics, slurs and articulation, are discussed here. These, however, hardly change the musical text a great deal (particularly compared to Cooper's edition, which includes almost all of these) and are presented in the form of an appendix with suspected errors, and in the case of some sonatas, a list of alternative readings of the text. The justification for these alternatives is that ‘there are places where a different reading could possibly be advocated on the basis of exactly the same evidence, and all those of any significance are listed here for the benefit of those interpreters who may wish to select the text that best suits their personal preference’.Footnote ⁵

In addition to giving performers more informed options, this approach also has the benefit of undermining some of the common objections against the term Urtext, which is espoused by Del Mar's editions. Cooper distances himself from the label, arguing that ‘claims by some modern editions that they are “Urtext” (literally, “original text”) are misleading, since there probably never was an original text entirely free from errors’.Footnote ⁶ But Del Mar, in a 2003 response to criticism of his Urtext edition of Beethoven's Ninth Symphony, defines the term as

Clearly, Cooper and Del Mar have contrasting definitions of the term, to the extent that some later Bärenreiter editions discussed at the end of this review list the ABRSM edition as Urtext, Cooper's misgivings notwithstanding. So, like the term ‘authenticity’, ‘Urtext’ has lost much of its originally intended meaning, and perhaps its retirement is overdue. But if marketing dictates its use (if that is indeed the underlying reason for its continued existence) Del Mar's fairly innocuous definition might be a sensible compromise to indicate the editor's intention towards the letter of a text.

But what about the spirit? Without some understanding of Beethoven's performance practice, much of the intended effect will be lost, misunderstood or altered, thereby undoing the editorial efforts. Del Mar and Donat's opening paragraph on performance practice, which almost every edition in this series reproduces, bravely sets out to tackle this issue:

The prospect that the reader will be presented with musicological and contrasting alternative ‘internal’ evidence is a welcome one, although the latter kind of evidence is never sufficiently defined. Unfortunately, with the exception of the turn figure in op. 28 discussed above, the only clear evidence of performance practice cited in any edition that includes the above paragraph is Stewart Deas's article on Beethoven's use of ‘Allegro assai’, published in 1950.Footnote ⁹ Deas convincingly argued that, at least in his later compositions, Beethoven's use of Allegro assai constitutes a slower tempo than Allegro, rather than a faster one. Yet in the critical notes to WoO 47 no. 2 – admittedly an edition without the above paragraph – Del Mar writes that

This interpretation raises the question to what extent musical judgements such as the one above can be taken as reliable evidence for what Beethoven had in mind. In recent years, research done on early recordings from the late nineteenth and early twentieth century has seriously undermined our confidence in answering that question in the affirmative. Starting with Robert Phillip, several scholars have pointed to the substantial differences between performance styles as heard on recordings and piano rolls from the late 1800s, and those more commonly heard in the late twentieth and early twenty-first centuries.Footnote ¹⁰ Performance practices that significantly predate recordings, such as those favoured by Beethoven, are likely to have been at least equally contradicting of current practices associated with these works.

One set of sources that can give some insight into this difficult matter are the editions and testimonials of his students Carl Czerny and Ignaz Moscheles. Despite his avowed interest in reconstructing performance ‘norms’ of the time, Del Mar and Donat use these sources rather selectively. The metronome marks by Czerny and Moscheles are dismissed altogether on the grounds that they are ‘equally problematic [as Beethoven's own], and in addition worryingly distant from the composer himself’. Yet the earliest metronome marks by Czerny for the sonatas date from 1828, surely close enough to be of some use.Footnote ¹¹ Beethoven's own metronome marks for op. 106, for that matter, are dismissed on the basis that later recordings of composers show that they often played their own works at vastly different speeds than those indicated in the score. (At the time of writing, Del Mar's edition of op. 106 has not yet been published.) Although the extent to which Beethoven and his contemporaries followed these indications in practice is not completely clear, it really is too rash to dismiss the metronome marks altogether. (It should also be noted that in the case of op. 49 no. 2, Czerny's editorial dynamics are listed in the section on performance practice; a curious editorial decision considering that many composers ignore their own dynamics every bit as much as their own metronome markings.Footnote ¹²)

In many cases, Czerny's and Moscheles's editorial speeds are actually perfectly reasonable. The second movement of op. 49 no. 2, for instance, shares a theme with the Septet op. 20, to which Beethoven in 1817 gave a metronome mark of ±=120. Del Mar and Donat criticise this speed for being too fast for the sonata, and cite Donald Francis Tovey's recommendation from 1931 that ‘it needs a tempo slower by one or two metronome-degrees’ in support, but Czerny's speed for this movement in all of his editions is with ±=112 exactly what Tovey proposes.Footnote ¹³ Moscheles, on the other hand, suggested a slightly faster ±=126, the same speed that Beethoven gave to the Tempo di minuetto of the Eighth Symphony. As many recorded performances attest, neither of these speeds is so fast as to lead to problems.

A few of Czerny's metronome marks, however, are indeed problematic, and his speed of °=76 in the Haslinger edition for the Allegretto of op. 54 is a good example. But that is an outlier in context, and Moscheles's roughly contemporaneous suggestion of ±=108 in his edition by Cranz seems a much more reasonable fit. Even Czerny's speed for the abovementioned Allegro assai in WoO 47 no. 2, a work which he in all likelihood did not study with Beethoven, is °=76 – quite fast, as Del Mar seems to think it should be. Whether that speed, which is indeed rather fast compared to Beethoven's metronome marks for similar music, constitutes evidence that Beethoven's intended meaning behind Allegro assai changed in the way that Del Mar describes will probably remain an open question. Either way, it would have been better if the metronome marks by Czerny and Moscheles for these sonatas had been included in this edition, as they provide (partial) insight into the musical norms that the performer needs to understand in order to approach a performance faithful to the composer's intentions, which seems to be the goal here. As it stands, the reader is left without much usable historical information on this aspect of interpretation.

There are also two performance practice issues in specific sonatas that could have been more clearly discussed. One is the effect that Beethoven's indicated fingering has on phrasing; in the case of the first movement of op. 2 no. 2 in particular, this is as vexing and controversial as any tempo issue.Footnote ¹⁴ The fingering in one passage (bars 83–88) is so awkward and counterintuitive that Donat's reference to Czerny's suggested ‘strategic left-hand assistance’ will be a welcome relief, particularly to players with small hands. Nevertheless, it would have still been helpful to discuss the effects of following Beethoven's fingering here, so that the reader would perhaps be able to re-create the effect of the fingering with two hands.

The other problematic passage appears in the Prestissimo finale of op. 53, which contains alternating pianissimo octave glissandi in both hands, with no pedal, while the other hand plays chords. It is rightly stated that ‘on the pianos of Beethoven's day, with their narrower keys and shallower action, their realization would have been a good deal more feasible than it is with a modern instrument, on which the passage confronts the player with an almost unsurmountable hurdle’. Unfortunately, no more information is provided, and any pianist hoping to find a historical alternative to what appears to be the most common solution for this passage – split the glissandi in two hands, and use the pedal to sustain the accompanying chords, resulting in a very different sound – will find themselves disappointed. Here, it might have been good to include Czerny's suggestion to simply play single scales, which eliminates the hurdle and results in a version that is probably closer in sound to the effect of playing glissandi on a period instrument, as it avoids the use of the pedal.

In sum, the impression one gets from the material on performance practice in these editions is that although they provide some helpful information – albeit less than they could have – there is very little that challenges any preconceived ideas a reader might have, which seems a missed opportunity. The authors explicitly state that ‘perhaps, different eras demand different solutions’, and although their reluctance to insist that these sonatas be performed in a historical way is admirable and worth contemplating, to some extent this attitude undermines the raison d’être of their entire project. If different eras indeed demand different, presumably non-historical solutions, why bother at all with coming up with a performance text that approaches the composer's intentions as closely as possible?

So the justification of these new editions is largely left to Del Mar's editing of the text, and whether that is enough depends very much on the particular sonata in question. On the one hand, there are a few sonatas with new and interesting textual readings, although they affect only a handful of notes overall; in addition, the inclusion of an appendix with alternative readings, although not containing any major revelations, is a welcome and enlightening feature. On the other hand, almost all of the unambiguous editorial improvements in the musical text were already present in Cooper's edition, which also included a more elaborate discussion on performance practice.

The final point concerns the pricing of these editions, which currently range from €6.95 for opp. 54 and 79, relatively short single sonatas, to €17.95 for the three sonatas op. 10, to a surprisingly high €21.95 and €25.95 for three sonatas opp. 31 and 2, respectively. Although the less expensive volumes are easy to recommend, minor shortcomings notwithstanding, one has to wonder about the economic viability of those on the upper end of that scale, especially since the three volumes of Cooper's ABRSM edition are available for just £20 each.

***

Much like the Bärenreiter editions, the single edition of op. 106 by Wiener Urtext under review here can best be seen as part of a wider enterprise, namely the upcoming publication of Jochen Reutter's revision of all the sonatas. The approach taken in this single edition is therefore worth considering in detail, starting with the definition of the term Urtext. The publisher's website lists ten ‘Principles of Urtext’, the first five of which are relevant here:

The first principle defines the term, similarly to Del Mar, as a statement of intent, rather than an achieved goal. The other four principles also conform rather closely to the editions by Bärenreiter, and one would expect similar results on that basis.

The overall design of the Wiener Urtext edition of op. 106 Bärenreiter editions discussed above, with the introductory material divided between an exploration of the historical context and a separate section on performance practice issues. The former is written by Reutter and is largely concerned with an account of the sources, with abundant references to the critical notes in the back of the volume. Uniquely among editions discussed in this review, the critical notes in the Wiener Urtext edition appear in both English and German.

The sources for op. 106 are problematic. Not only is the autograph score lost, the London and Vienna first editions have different readings for some passages. Dispute also exists over when the London edition was published, as there are two entries in Stationers’ Hall, on 1 October and 24 December 1819 respectively, only the second of which Reutter mentions.Footnote ¹⁶ Most significantly, the London edition includes adaptations for the smaller range (CC–c4) of most English pianos, presumably made by Ferdinand Ries, who was Beethoven's representative in London at the time. Strangely, these adaptations only appear in the first movement, and in later passages in which the piece exceeds the range of English instruments – bar 112 of the scherzo and bars 102–103 of the adagio, in which the right hand goes to f4 and c#4, respectively – no alternatives are offered, something which goes unmentioned in the introduction. The London edition also includes several corrections, presumably also by Ries, of errors that appear in the Viennese edition. So the editor's task is a complicated one, and involves making judgements about the nature of the differences between the Viennese and London editions, which may well contain (possibly unauthorized) corrections by Ries, printing errors that slipped through, or even last-minute decisions by Beethoven that caused the ever so slightly later Viennese edition to have corrections not incorporated in the London publication.

The musical text of this new edition is strongly reminiscent of the 2001 Wiener Urtext edition by Peter Hauschild, and at first sight the differences are hardly noticeable: the page layout is unchanged and the fingerings by Alexander Jenner have also been maintained. The alternative passages in the first movement that appear in the London edition are also still reproduced in footnotes; this seems a strange decision, as Reutter ascribes these to Ries's attempt to come to terms with the range of the English pianos, which somewhat contradicts Wiener Urtext's first principle cited above. The awkward original orthography in the bars leading up to the first movement's arrival in B-flat in bar 17, in which the melody soars five leger lines high above the staff, is also maintained, instead of adding an ottava sign that would have significantly improved the legibility of this passage.

Nevertheless, there are several important improvements compared to the 2001 edition. Wrong notes have been corrected, and many editorial slurs not in the sources have been removed or indicated as such, although all of these were already correct in Cooper's ABRSM edition.Footnote ¹⁷ Ultimately, the most considerable difference between the ABRSM and Wiener Urtext edition in terms of musical text is found in the notorious ambiguity concerning the A/A# in the passage immediately preceding the recapitulation of the first movement in bar 224. Here, Reutter relies on research by Paul Badura-Skoda from 1980 that concludes that the pitch should probably be A#.Footnote ¹⁸ In 2012, however, Baruda-Skoda argued in a follow-up article, similarly to Cooper, for A, based on Beethoven's wider compositional practices, evidence from the sketches, and the fact that A# creates a progression that would have been unheard of in that time.Footnote ¹⁹ So despite claims to the contrary, the critical notes do not contain the most recent research.

But by far the most surprising and unusual aspect of this edition is the inclusion of an article by Johann Sonnleitner about Beethoven's metronome marks for this sonata, instead of more general historically informed notes on interpretation. Since this article is written not just to inform, but also to persuade the reader of the historical basis of a particular interpretation, it requires close scrutiny. Sonnleitner argues that some but not all of Beethoven's metronome marks for op. 106 were intended to be interpreted differently than they are today. According to this interpretation, the indicated speeds for the first and second movements and the fugue in the fourth movement should be halved compared to the common readings, while the speed for Largo introduction should remain as it is. About the third movement Sonnleitner is agnostic, and he lists both the literal and the halved speed as possibilities. The justification for this interpretation is presented as seven mutually supporting pieces of evidence: Mälzel's protest against the supposed ‘incorrect’ use of the metronome in an article in the AMZ of September 1821; the notion in nineteenth-century literature that tempo was often considered to be consisting of two parts, a heavy and a light part, each of which according to Sonnleitner need an audible tick from the metronome; the markings on some later metronomes that suggest multiple audible ticks per beat; the physical limitations of historical fortepianos; diverging metronome marks in the early editions; the perceived contradictions between character and metronome marks; and some reported durations of op. 106. It is probably also worth stating that the adherence to a composer's intended practice does not guarantee critical or musical success, and there are well-known successful performances of some of Beethoven's works at completely different tempo, or even half the indicated tempo that the composer indicated.

Making Beethoven's piano music more accessible to a modern audience is a laudable goal in itself, but the fringe theory presented here, first proposed at length by Willem Retze Talsma in 1980, is insufficiently grounded historically, and although it has been refuted several times since its first presentation, no rebuttal is offered by Sonnleitner.Footnote ²⁰ The principle objection involves the lack of unambiguous supporting evidence: if the practice of halving some but not all speeds was as widespread in the nineteenth century, as claimed by Sonnleitner, one would expect to see at least some explicit discussion of it in the contemporary literature, both during its heyday and during its supposed phasing out. Furthermore, a justification for a deliberately ambiguous two-tier system is also lacking.

The aspects of the article that apply more specifically to op. 106 are likewise problematic. First, although the metronome mark for the first movement in particular is remarkably fast – indeed it was considered ‘impossible’ by Donald ToveyFootnote ²¹ – there are several recordings at or close to the indicated speed.Footnote ²² Second, an edition by Holle that Sonnleitner uses to argue that Moscheles sometimes indicated a significantly slower speed has been shown to be a forgery, and Moscheles was not involved in the editing at all.Footnote ²³ Third, although the supposed limits of keystroke repetition of Beethoven's instruments is an interesting (albeit not unambiguous) angle, there are problems with the evidence, which consists of out of date references to websites, and one examination of historical instruments in their current condition, rather than their condition two centuries ago.Footnote ²⁴ Fourth, the author simply asserts that it is ‘inconceivable’ that Czerny and Hummel would not have halved some of their speeds, without providing further evidence. And last, the core of the argument, the notion that Beethoven is supposed to have halved some but not all of his metronomic speeds too, is supported merely by two references to the Sonnleitne's own unpublished writings, which raises questions about the historical validity of this claim, and in turn about editorial standards. So unlike the Bärenreiter editions discussed above, this edition is all too than willing to challenge preconceived ideas about performance practice, but its ambitions are lamed by its biased and inaccurate scholarship.

In summary, this edition only partially lives up to the standards that the publisher has set, with the promises to provide critical notes based on the most recent scholarly research as well as historically informed notes on interpretation being most clearly violated. It is thus difficult to recommend this edition without significant reservations, as the musical text offers little to improve on Cooper's edition, and the notes on interpretation are so problematic. Nevertheless, those who prefer Wiener Urtext's design – the iconic red cover has its own attraction – might welcome Reutter's corrections of Hauschild's 2001 edition.

***

None of the editions discussed so far have managed to offer new robust findings in terms of both musical text and introductory material, although Del Mar's editions have made a notable contribution in the former category. But there is a hitherto undiscussed edition of one of the sonatas, the Bärenreiter edition of op. 13 published in 2012, that comes closest to doing so. This was the first piano sonata by Beethoven that Del Mar edited, and it contains two features not present in subsequent editions that I hope will be re-introduced in the future.

The most conspicuous difference is probably the material on performance practice in the introduction, which is provided by Mario Aschauer. Much like Del Mar's later introductions, it is divided into sections on individual aspects (instruments, articulation, pedal, embellishments and tempo), but these draw on a much wider range of scholarly literature, with primary sources informing the readings. Aschauer also includes music examples that help performers engage with and incorporate the scholarly material in their practice: the section on embellishments, in particular, is much strengthened by short examples showing appoggiaturas being played on the beat, although the length of the notes and the placement of the accents could perhaps have been more clearly discussed.Footnote ²⁵ And unlike the later editions, Czerny's and Moscheles's metronome marks are also included, with most of the relevant scholarship appearing in the footnotes. In the end, the reader comes away from this introduction with a range of ideas about historical performance practice specifically applicable to this sonata, which might inspire historically informed choices that would not otherwise be made. The other part of the introduction is by Hartmut Hein, and covers the genesis, aesthetic, and formal designs of the sonata in significant detail.

The second remarkable aspect of this edition – the musical text is as good as in other editions by Del Mar – is found in the critical notes. Here, rather than merely listing the differences between the sources, the critical commentary also discussed the variants in the editions by Henle, Wiener Urtext, and ABRSM, thereby making explicit to what extent this publication has anything new to add. (The editions of opp. 27 and 57 also include this feature, but it is absent from the editions published from 2016 onwards.) Along with the aforementioned practice of including alternative readings in the appendix, this is another feature that should be more widely adopted, as it explicitly answers the question stated at the start of this review. An edition with all these features would be a very welcome addition indeed, and had some of these features not been phased out it would have existed by now.

***

The above review of Urtext editions has highlighted a number of characteristics that good editions of this repertoire (should) have, as well as some pitfalls that are best avoided, whether they aim to represent the composer's intentions, as far as they can be determined, or something else entirely. Drawing on these observations, the following list of guiding principles can be synthesized. (Of course, these principles are not exactly new, but they are worth rearticulating clearly considering how few editions adhere to them.)

First, editions of Beehoven's piano sonatas should include (and almost all already do) an introduction detailing the genesis of the music and any noteworthy structural aspects or other particulars. Second, any discussions of performance practice should offer a summary of the current state of relevant scholarship in order to inform and therefore conceivably change the way that the reader plays these sonatas. To many players, this scholarship is unfamiliar or even inaccessible, and audio examples, perhaps on a companion website, can be of great value here. Third, the musical text should clearly indicate the most likely reading, with the editorial justification in the critical notes and preferably alternative readings indicated in a separate appendix. Fourth, considering the number of editions available of this repertoire, some account needs to be given of the differences with other editions, including any new sources that might have been consulted, and/or new editorial interpretations. This is in part to justify the production of yet another edition, but also to pre-empt any potential accusations of breach of copyright; with multiple editors pursuing the same goal, it is not inconceivable that this will be a problem in the future, although it has not been so far. Ideally, new publications should clearly show what they have to add, so that the question posed at the start of this review can always be readily answered.