The Warning Voice (2 page)

One characteristic of this third volume which will, I suspect, strike many readers is the important part that hitherto unknown or undeveloped minor characters are made to play in it. To the translator, on the other hand, its most striking characteristic is the intensification, to a point at which the novel almost breaks down beneath them, of those textual problems already mentioned in my prefaces to the two earlier volumes. These two characteristics of Volume Three – the development of minor characters and the growing number of contradictions in the text – have a causal connection which I shall presently try to explain.

As a preliminary I should like to examine three statements, each made by one of the people by whom the text was produced. The first is by the author himself.

These words are found in a recension of the text dating from 1754, nine years before the author died. Although the statement is dressed in allegorical terms, there is no uncertainty at all about its meaning:
years
before he died Cao Xueqin completed his novel and subjected it to several revisions, and
at least
as early as 1754 he was engaged in producing what was to be its final version. Whether or not the ‘five times' is to be taken literally does not now concern us – though the fact that this same passage lists five different titles which the novel is said at one time or another to have been given inclines me to think that it is: the most important thing to bear in mind is that there had been several already completed versions prior to the version he was then writing.

The second statement is by Cao Xueqin's kinsman-collaborator Red Inkstone who produced the manuscript recensions of the novel from which most of the still extant manuscripts were copied. It appears in a comment dated September 1764.

Whatever ‘unfinished' means in this second statement, it certainly does not mean that this is a novel like
Edwin Drood
or
Weir of Hermiston
which was never completed because its author was struck down in the midst of writing it. Apart from what Xueqin himself tells us in the first chapter, there is a marginal comment by Odd Tablet dated April 1762 (a few months before the author died) on the manuscript recension of 1760 telling us something about the contents of the final chapter. This proves that there must have been one version at least which the author succeeded in finishing.

The third statement comes in the Prefatory Remarks to the revised edition of the novel printed only a few months after the first edition of 1792. The Prefatory Remarks were published under Gao E's and Cheng Weiyuan's joint signatures, but as it was Gao E who did the editing, it was probably he who wrote the Remarks. (In a short preface to the first edition Cheng Weiyuan had already explained how he spent many years combing the book markets for the missing forty chapters, and how he eventually handed over what he had managed to collect to his friend Gao E to edit.)

What is one to make of these three apparently irreconcilable statements? Merely to say that one of the witnesses,
viz
. Gao E, must be lying leaves too many questions unanswered. Cao Xueqin and Red Inkstone still seem to be contradicting one another. Merely to insist that the last forty chapters of Gao E's edition are spurious still does nothing to explain what happened to the
genuine
last forty chapters. I suggested in my Introduction to Volume One that the family may have suppressed them for political reasons, but I am not at all convinced that that is the correct explanation.

Let me return for a moment to the question of what Red Inkstone and Odd Tablet meant by ‘unfinished'. We know that in some cases it refers to small parts missing from otherwise completed chapters. Red Inkstone wrote a note in 1756 at the end of
chapter 77
to remind himself that he was still waiting for Cao Xueqin to supply the Mid Autumn poems which were to be inserted in that chapter. He was still waiting for them when Xueqin died seven years later. But there are several mentions elsewhere of whole chapters missing. A note dated ‘summer of
ding-hai'
(i.e. early autumn, 1767: four years after the author's death) made by Odd Tablet on the 1760
recension of the first eighty chapters mentions a borrower losing ‘five or six chapters' from the last third of the book ‘when we were making the fair copy'. The date of this loss is impossible to determine, but I should hazard a guess that it occurred not later than 1754 when Red Inkstone was making a fair copy of what was to be the ‘final' version of the novel – the version in which the author refers to his ten years' labour and many revisions of the text. Odd Tablet says that what was lost was the
drafts
of these chapters, i.e. Cao Xueqin's own autograph of them, which had not yet been fair-copied.

A great deal has been made by those wishing to discredit Gao E and Cheng Weiyuan of the fact that the titles or subjects of the lost chapters mentioned by Odd Tablet do not correspond with any of the titles or contents of the last forty chapters of the Gao E-Cheng Weiyuan edition; but it must be remembered that these chapters were lost before Red Inkstone could copy them; and it is by no means obvious that so compulsive a reviser as Xueqin would have reacted to the loss by simply sitting down and rewriting them. In fact he
cannot
have done so if Red Inkstone and Odd Tablet are to be believed, because when Xueqin died, seven years after the recension of 1756, they were still waiting not only for the last forty chapters but even for the pages and the few odd poems that were still missing from the first eighty.

What happened after the ‘five or six chapters' were lost? The answer to that question can only be guessed, and before making the guess, it is necessary to guess the answers to several other questions. First of all, what was the method by which the fifth and ‘final' version was being produced? My guess – it is nothing more – is that Cao Xueqin was rewriting his finished fourth version and sending the manuscript to Red Inkstone for copying in batches of ten chapters. Second question: which were the missing chapters? The titles or descriptions (whatever they are) suggest that they belonged to the section of the novel immediately following the confiscation of the Jia family's estate. In the Cheng-Gao edition we read today the confiscation occurs in chapter 105. Assuming that it came in about the same place in the missing version, my guess would be that Red Inkstone had fair-copied all or
most of the novel up to chapter 100 and had just received the manuscripts of 101–10 when the loss occurred. He and Odd Tablet had already read through all ten chapters and the anonymous friend had already read the first four. He was allowed to take away and read 105-10 while Red Inkstone was busy copying 101-4. After the loss Xueqin asked to have the manuscripts of 101-4 back, as well as everything that had been fair-copied from chapter 81 onwards, in order to help him reconstruct the missing section.

It would of course be possible to imagine something much more sinister: for example that the anonymous borrower or one of his family or acquaintance actually destroyed the manuscripts and gave Red Inkstone to understand, when he informed him of their ‘loss', that they were highly subversive and dangerous and that Xueqin must be urged in the strongest terms to alter that part of the novel.

As for what Xueqin did or did not do during the years which followed: we can guess that he was so disgusted that he did nothing at all, or that he ‘dried up', as authors sometimes will, or that he worked on the last part of the novel intermittently but was too busy scratching a living to do so effectively – there are a hundred compelling reasons for not completing a book. In this mass of guesswork only one thing seems certain: Red Inkstone and Odd Tablet got nothing more out of him until he died.

And when he did die, what remained of that last third part of the book? Presumably all of the last forty chapters
¹
in their fourth version, some twenty of them (81–100) in the fifth version fair-copied by Red Inkstone, and a few chapters (101-4) of Xueqin's autograph of the fifth version which had never been copied. All of this may have been worked on to some extent by Xueqin himself before his death, and it is reasonable to suppose that after his death Red Inkstone or Odd Tablet or someone else may have tried reworking them. The important thing to remember is that if anything emanating
from Cao Xueqin – however much it had been tinkered with by others in the meantime –
did
ever find its way into Cheng Weiyuan's hands (and one must not exclude the possibility that Cheng Weiyuan may himself have made an unsuccessful stab at editing it before handing it over to Gao E) a large part of it would still represent the obsolete fourth version and therefore be at odds in places with the text of the fifth version represented by the 80-chapter Red Inkstone manuscripts, particularly in cases involving the names of minor characters, which, as I attempted to demonstrate in the Preface to the last volume, appear to have remained unstable until a fairly late stage in the novel's development. This in fact is what we seem to find.

Take the case of Cook Liu's consumptive daughter Fivey. In
chapter 77
in the Red Inkstone manuscripts we learn from Lady Wang's lips that Fivey is now dead, yet she appears again in chapter 109 of the novel alive and well. In accordance with the principle enunciated in his Prefatory Remarks, Gao E's solution is to leave the relevant passage in chapter 109 untouched and remove the reference to Fivey's death in
chapter 77
. In the manuscript I have elsewhere referred to as ‘Gao E's draft' we can actually see where he has crossed it out. In an appendix to this volume I have tried to show that something similar to this must have happened in the case of Lady Wang's maid Suncloud, though in her case the confusion in the text was so complete that Gao E failed to spot it.

As a matter of fact, though, the discrepancies between what is found in the last forty and what is found in the first eighty chapters which so exercised Gao E are probably not as numerous as those occurring inside the first eighty chapters themselves. They represent Xueqin and Red Inkstone's failure, even in the ‘final' version, to root out all obsolete survivors from the earlier version.

Such survivals are easiest to spot in the poems. Verse is much harder to alter than prose and would tend to get copied out intact from one revision to another, preserving relics of the earlier versions inside it like flies in amber. The maid Sandal, evidently one of Bao-yu's principal maids in an earlier version of the novel, gradually dwindles out of the prose
narrative in various successive editions and in Gao E's edition has all but vanished, yet we meet her several times as it were mummified inside the poems.

Most of the textual problems of Volume Three occur in that section of the novel which centres on the story of the You sisters,
chapters 63
to
69
.
Chapters 64
and
67
were missing from copies of the Red Inkstone manuscripts circulating in Xueqin's lifetime, and even in Gao E's day, thirty years later, manuscript copies of the first eighty chapters sometimes still lacked
chapter 67
. Two quite different versions of that chapter are now extant. It is generally assumed that the two chapters were omitted because, like the ‘five or six chapters' from the latter part of the novel, they had been lost. I think myself that they were not lost but deliberately held back for recasting because of discrepancies caused by the insertion of new material. In my view the story of San-jie and her tragic betrothal was grafted on to the novel at a very late stage, and the insertion of this sub-plot into the narrative of Jia Lian and Er-jie's secret marriage and Xi-feng's revenge created problems of timing and consistency so great that no amount of tinkering was – or ever has been – able to remove them. It is a measure of Xueqin's genius that he has been able to charm generations of readers into regarding this as one of the most moving and delightful parts of the novel while overlooking the quite extraordinary discrepancies which it contains. Only a spoil-sport, it might be thought – a ‘kill-view', to use the eloquent Chinese expression – would want to break the enchantment by pedantically pointing them all out. I do so only because once or twice, in the interests of clarity and consistency, I have felt obliged to take some trifling liberties with the text (as for instance in the killing-off of Mrs You – not that she was ever very much alive, poor old lady) and hold myself honour bound not only to say what I have done – which I have tried to do in the Appendices – but also to explain, if I can, the circumstances in which I have felt obliged to do it.