Review of Elgar, The Starlight Express

Creative Commons Licence
Review of Elgar, The Starlight Express by J. P. E. Harper-Scott is licensed under a Creative Commons Attribution-NonCommercial-NoDerivs 3.0 Unported License.
Based on a work at

Review of Elgar, The Starlight Express, edited by Roger Dubois, Elgar Complete Edition vol. 19 (Rickmansworth: Elgar Society Edition, 2010), Elgar Society Journal 17 no. 1 (2011): 50–53.

There is no such thing as Elgar’s The Starlight Express. Unlike Elgar’s major works, it has no continuous performing history, having never established itself as music for performance beyond its original context in the Kingsway Theatre in 1915–16. It appeared as incidental music to a play of the same name, which was a peculiar stage adaptation by Violet Pearn of an equally peculiar novel by Algernon Blackwood (A Prisoner in Fairyland, 1913). What survives as a record of The Starlight Express – a play typescript, which was changed in the process of composition, and a full score in Elgar’s hand – is testament to a brief performance history after Christmas 1915, as passages were cut and altered and the spoken dialogue bundled about to fit in different ways with the music. As (I guess: the layout of the volume isn’t clear on this point) Jerrold Northrop Moore remarks, among the notes and emendations in the full score ‘it is difficult to distinguish what comes from Elgar, what from the original conductor Julius Harrison, and what from later conductors and perhaps even producers’ (xv).

What did the audience hear on the premiere, 29 December 1915? What did they hear in January 1916, before the end of the play’s run? We can be sure that they were different things. What, then, is the ‘real’ Starlight Express? What should go into an edition? The version heard at the premiere? (Answering Elgarian questions like that in the affirmative would rule out the familiar ending of the Variations, op. 36, which was added later.) Or the last version that Elgar approved for performance later in the run? If the latter, how can we determine what that is among the accretions to the score that are so difficult to date or pin down to a particular person? If we don’t want to privilege Elgar’s ‘original intention’ (which in any case we can never obtain on any point, at least until time-travel and infallible means of mind-reading become possible), and can’t tell what counts as his (as opposed to a later reader’s) last thoughts, what are we left with?

Questions like this are what give all editors ulcers, but they are particularly sharply focused when the music in question is a stage work. Even pieces such as this – relatively small-scale, of very local interest, and with a short (actually, given Elgar’s prestige, embarrassingly short) initial run and no significant full revival – can evolve with drastic speed as a result of input from performers, directors, playwrights, critics, and so on. The first question an editor of a paper edition has to answer, therefore is: ‘do I want to produce a scholarly record of an “original” form of the piece (whatever that means, and for whichever reason I might want that), or do I want to produce an edition that could enable performances in the future, perhaps in the hope that it might give the music a new chance to be heard? ’ (This is a limitation of paper editions: online editions could present multiple readings that are easily switched between at the touch of a button marked performance edition, Urtext, or whatever.) Roger Dubois, the editor of this edition, does not make it clear which side of this question he comes down on (perhaps in conformity with the view of the General Editor, John Pickard, expressed on p. i: ‘The Elgar Complete Edition is intended for both scholarly and practical use’). My own view is that music lives better in performance (real or recorded) than in libraries, and I am happy to say that with the aid of this new edition the possibility of performance is increased, although it would probably only be a performance of excerpts, since the play it accompanies would have negligible appeal today, except for those who are totally immune to innuendo (‘He touches Daddy with his flaming pole’) and sub-Disney saccharine.

Dubois, who is credited as ‘the editor’, and Moore (who has added only relatively small tweaks) have, as one would expect, taken certain editorial decisions that allow performers room for manoeuvre. For instance, where spoken text has been marked by Elgar to be delivered at a specific point in the music, this is indicated by a solid bracket attached to an arrow. Where Moore has deduced his own links, the bracket is dashed rather than continuous. The editors admit that this might lead to over-precision, and I think their decision tends to overvalue the ‘hand of the master’ here. Viewing this from the perspective of performers, it is obvious that anyone performing this music with spoken text should feel as free as Moore himself has been in attempting to match text-delivery to the music, and should not feel beholden to Elgar’s own designations – which were, let us remember, also just chosen for efficacy in the act of performance (and could have been written at any time during the play’s run – or after – so we’re back to the opening questions about exactly what kind of text is being presented here). From the perspective of a scholar or reader who wishes to access something like an Urtext or a definitive edition of the ‘original’, Moore’s additions are of course totally out of place. This inconsistency of approach, half aimed at performance, half at documentation of an ‘original’, is typical of the edition, and perhaps an error of judgement.

There are also inaccuracies and confusions in both score and annotation. The general practice when Elgar indicates in the play typescript that a character X should sing music Y at point Z is simply to note the fact, but sometimes some musical notation is given. So, on p. xxx, documenting f. 25 of an unbound bundle of sketches (LBL Add.MS 69833), we find a ‘sketch for 9, in ink with pencil amendments; also a complete sketch in ink for 1a in F major’. Various niggles emerge. Nothing in the score is marked as ‘1a’, though the music quoted here is given (split without explanation into two pieces) on pp. 30–1 between nos. 1 and 2 of the full score, and this probably counts as ‘1a’ in the editor’s mind (but he should have made that clear to readers). The music is given in F major, presumably the key of its notation in the sketch, but in no. 41 it is in G major. The text is also different in the two cases (the sketch has ‘Get out, you Morning Spider! You fairy cotton rider With my tiny nets of feather I collect the dust together And on strips of windy weather bring the day’, while the full score has ‘We shall meet the morning spiders the fairy cotton riders, Each mounted on a star’s rejected ray. With the tiny nets of feather They collect our Thoughts together and on strips of windy weather Bring the Day’). Why, then, include the sketch incarnation of the music and text – which flatly contradicts what is presented elsewhere in the volume – when (a) the overwhelming majority of sketches are not included and (b) the express purpose of including this one is to indicate the music and text that Elgar said should be performed at this point in the play? There may be good reasons, but we are not told them. Here in this tiny detail we see the difficulty of taking editorial decisions on music such as this, but unless such things are to appear as editorial blunders the reason for the confusing presentation should be explained. (It would also have been useful to have been provided with a list of sigla used in the notes, since it is sometimes difficult to know exactly what something like ‘TSE1’ signifies without ruffling back through several pages.)

So might a scholar quibble, but any performer will simply perform what is given in the full score. They should pay close attention to the editorial changes documented in the preliminary matter, and may particularly wish to consider their options concerning the use here of ‘Sun Dance’ from The Wand of Youth. Elgar simply indicated that this music should be used in the incidental music, but it was written for a larger orchestra than he used in the Kingsway Theatre, and if Elgar produced a reduced scoring for The Starlight Express it does not survive, so Dubois has had to rescore it. This seems to have been done sensitively but if there are no limitations on any future performance, it is not obvious that the full, original orchestration should not be used. And again the question arises: why change the orchestration at all? This is an entirely speculative reconstruction of music that Elgar might have heard in December 1915 or January 1916, but to what end is it included here? As a pristine (albeit completely inauthentic) record of an ‘original’ or with the intention that a future performance of the play with the same orchestra would require such a reduced instrumentation?

These riddles are difficult to resolve, yet despite them – and chucking scholarly questions aside at last – the final point is this: it is delightful to have at last an edition, very handsomely produced, of what is at times truly delightful music. Although it might never again be performed in its entirety (and Moore’s patient decisions over the exact placement of spoken text might therefore prove a fruitless act of love), and it certainly does not reconstruct an authentic ‘original’ text (since nothing of the sort can honestly be imagined), it is finally somehow appropriate that the conundrums raised by its presentation both slot into a venerable Elgarian tradition of perpetuating enigmas and reflect the weird and ultimately ungraspable nature of this quite unique piece in its composer’s output.

Leave a Reply

Please log in using one of these methods to post your comment: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.