A Dance With Dragons is the fifth volume in the ongoing epic fantasy series A Song of Ice and Fire. And because it’s volume five, I’m somewhat limited in what I can say that’s not a spoiler. But in any case, I recently read and reviewed the first volume, A Game of Thrones, and was rather unimpressed – but between the promising elements of that novel, my memory of enjoying volumes two and three, and my re-awakened desire to find out what happens next after many years away from the series, I decided to buy the latest book and give it a go.
My first impression: it’s a lot, lot better.
The general portrait of the strengths and weakness of the author that I gleaned from the first novel remains true. Martin is an author who is best at the big scale, best at telling an engaging story, best when he has a lot of words to work with, building things (characters, plots, themes, foreshadowing) up with layer after layer of words. He is rather less good at the details – the prose, the incidents, the incisive, delicate sketchwork. That remains true – but the weak points have been considerably strengthened. ADWD feels like a novel written by a man entirely at home in what he’s doing, where AGOT often felt nervous, unsure, as though the author weren’t quite sure what he was doing.
The difference is obvious within a few chapters: the opening chapters of ADWD are far more polished, far more compelling, far more satisfying by themselves, frankly far better written in every way. Within pages I knew that I would never – as I did with AGOT – ask myself whether I could be bothered finishing this novel.
That’s not to say there aren’t problems. Although Martin’s narrative voice, an attempt to merge the fauxdieval intonations of epic fantasy with a more modern, more realistic way of speaking, feels a lot more coherent and settled than it did in the first volume, he’s still not quite in control of it. This leads, for instance, to the repeated use of crutchwords that he seems to have recently discovered and is determined to impress us with – thankfully, the endless “nuncles” that appeared in the fourth volume (Martin evidently learned between books three and four that ‘nuncle’ is an older form of ‘uncle’, though since none of the rest of the book is written in Middle English I don’t see why this matters) have largely, though not wholly, been eradicated, but in their place is an inundation of “neeps” and things that are “leal” and/or “niello”. Thank you George, it’s nice to know you own a dictionary – but a pity that you don’t also own a thesaurus. It’s an even bigger problem on the level of phrases, where Martin’s inability to think of alternative ways of saying things leads to repetitive and boring prose – even if we only talk about the really distinctive, memorable phrases, there are a distracting number of things that are “as useless as nipples on a breastplate” (lampshaded when one character actually finds a breastplate with nipples on – it remains to be seen whether these nipples will prove useful in any way), and there are half a dozen instances where substances (not all the same substance!) “trickle down her thigh”. [Incidentally, many people may find the gratuitous sex and violence an issue – personally, I thought it was most handled fairly well, though it got a bit silly now and then – the implied bestiality, for instance, reads like a very lazy way of making a point]. Sometimes Martin seems to confuse “pervading themes” with “words that are repeated again and again by many different characters” – most notably, rather than (or, charitably, in addition to) demonstrating in many circumstances how stated intentions can be misleading or in vain, he just forces us to listen to a score of characters repeating every few pages that “words are wind”. Likewise, I think it’s been well-enough established that we are dealing with political intrigue without repeatedly being told of a “game of thrones” underway. All these phrases, when occurring in dialogue, would be a lot more convincing as ‘things people in this world say’ if people had been saying them before. If “words are wind” is such a ubiquitous proverb in several different societies on this planet, why weren’t we hearing it in previous novels?
But, despite these problems, the writing here is a lot better than in the first volume, and generally is quite good. There are other things worth praising too. First off has to be the astounding, awe-inspiring scale and ambition of this story. A Dance With Dragons covers events on two different continents (in half a dozen locations), through the eyes of nearly twenty point-of-view characters and a supporting cast of hundreds if not thousands, over a time period of… I’m not sure exactly, but I think many months, if not years. The links between events are tangential or even by now non-existent, as though there are many unrelated stories being told at once (it would be fascinating to diagram the different story threads and how they interacted). This is the biggest and broadest that novels can get. The impression is a sort of controlled confusion, much like listening to a complex fugue – we follow one strand and then another and if we try to grasp to many strands at once we’re overwhelmed by the scale of it. This effect is amplified by the unstable timeline – some threads have so little connexion to anything else that they may be occurring before or after any of the others, and when we can compare timelines directly it seems that later chapters may occur before, or at least begin before, earlier ones.
Then again, the scale of this book makes the whole notion of timeline, of simultaneity, highly questionable – one of the triumphs of the series is that it deals honestly with the relativity of pre-modern society. The length of time it takes news to travel from one place to another is so long, and so unpredictable, and so dependent on routes of travel and on relative location and on the type of information being relayed, that it is impossible for any character, or for the reader, to have an unambiguous timeline of which events happen before other events. In this case, this effect is made even more prominent by the fact that the first half of the novel (roughly speaking) occurs ‘simultaneously’ with some events in the fourth volume (and some of the fourth volume is ‘simultaneous’ with some of the third volume). This means we have been told about some things, from the perspective of characters appearing in the fourth book, ‘before’ they have ‘actually happened’ from the perspective of the characters who witness them firsthand; while other events from the fourth book have clearly ‘happened’ by the end of the fifth book, even though most characters in the book don’t know that they’ve happened (and likewise, some chapters in the fourth book depict events, first-hand, that we’ve already heard about second-hand in the third volume…). Some might call of this messing about with the timeline ‘lazy’ or ‘confusing’, but I think it’s a brilliant idea – it reflects the actual state of affairs in this sort of society, and is a rich source of dramatic tension and tragic misunderstanding.
Also demanding respect is the greater imagination of this volume. Both in its alien-society-full-of-wonders sections on the continent of Essos and in its dark – sometimes horrifically dark –, fairy-tale-imagery sections in the North, ADWD manages to hit a more fantastic, out-of-this-world note, while still managing to make it seem a natural evolution of what has gone before. Several of the Northern chapters in particular were thrillingly evocative of the long history of dark-forest-in-winter nightmares of European storytelling. And I do mean dark. This novel gets stunningly, overwhelmingly, how-will-they-ever-be-allowed-to-film-this dark – and although it feels a bit artificial in its darkness now and then, a bit sensationalist, that’s actually a fairly rare problem, and most of the time it manages to make journeys through the pits of hell more or less realistic and believable. Remarkably, despite the length and breadth of the narrative threads, many of these chapters felt satisfying and complete in and of themselves, to a degree I don’t recall having encountered in previous Martin novels. Again, though I was dispirited after reading the fourth volume, the long wait for the fifth seems to have sharpened the author’s skills rather than rusting them.
On which note, one last note of acclaim: as always, Martin’s writing here is very cinematic, and the imagery of some chapters is breath-taking – I would say that I couldn’t wait for HBO to get to these chapters just to see what they look like on screen, but unfortunately I know they won’t have the budget to do them justice.
Praise duly given, however, I must return to criticism. I’ll set aside the complaints I’ve already made about Martin’s writing in general and just talk about this book. In this book, the biggest problem is that it’s all pointless. Most of the book – particularly the sections set in the East – feels like filler, and I really don’t understand why either A Feast For Crows or A Dance With Dragons was necessary. It could all have been dealt with via a five-year gap. Even if he didn’t want to give us so many flashbacks, it certainly didn’t need two massive doorstop tomes. The process seem to be that to fill in a section of events he creates characters, and those characters need stories and things to do, and then those ‘things to do’ have causes and consequences, and then those causes and consequences require more characters to be involved and more things to happen and… . Although by the end of the book, Things have started to Happen, most of those Things are really just resolutions for plot threads only introduced in this novel! Very little progress is made on the overarching plot. In fact, much of it feels like a step back, with characters returned to old positions and plot lines reset – near the end one character even muses explicitly that ‘to go forward you must go back’, and this feels like quite a narrative cop-out. I’m left with the inescapable fear that if Martin wants to actually get on with the story he will have to write-off most of what happens in ADWD as ‘learning experiences’ for some of the characters and then ignore the fact that any of it happened. Meanwhile, the fact that it HAS happened will have consequences – so much time has been spent on resolving the relatively minor fallout from the third novel that not having such resolution of the main plot climax of the series will feel unbalanced and deceitful (assuming that the main plot doesn’t end with ‘everybody dies’, which seems to be the only case when Martin will allow ‘and that’s the end of the story’ to be true – everything in his books has consequences, which is an admirable level of realism but will surely make it hard to wrap it all up convincingly). More immediately, while some characters have reverted, others have managed to entangle themselves in very complicated situations, which will surely take a huge number of pages to unravel.
In short, this novel not only doesn’t get us much closer to the end of the series, it actually makes the end look further away. Given the length-to-content ratio of this book, it’s hard to imagine Martin finishing in only two more volumes. In some ways, the constantly-increasing attention to detail is a good thing – it’s a big part of why this is a more satisfying book than AGOT – but the combination of obsessive detail with vainglorious breadth means that this could take half a century to complete.
Myself, particularly for Dany’s storyline, I think a cold opening going straight from the end of A Storm of Swords to the end of A Dance With Dragons, missing out everything that happened in-between, would have been the greatest opening to a fantasy novel ever… but I realise that that’s probably a minority taste!
So, it’s a fun read but doesn’t get us very far. But some of it isn’t even all that fun. Personally, I didn’t find any of it intolerable, but it does go on for a very long time, and many chapters are devoted to world-building with little or no action in them. I can’t fathom why Martin thought certain characters needed so many chapters of exposition – of the three main POVs in this book, only one didn’t feel as though it was killing time, and not much even happened in that one! The vast bulk of the book was made up of these three characters not doing very much, and I was desperate to have some more from the ‘minor’ characters, whose chapters were actually both exciting and plot-advancing! Page allocations seem to have been done more on the basis of which characters are the designated stars and less on the basis of who has most to say and do. Meanwhile, other characters just pop in once or twice for little “I’m still alive! Don’t forget that I exist! I’ll probably be important in the future! Maybe!” cameos. Nor does the weird ‘book four characters merge in halfway through except for the one that’s there all along due to the timeline being strange’ structure work all that well. These characters got only one or two chapters each, wrapping up the cliffhangers set up at the end of the last book, and reminding us how pointless the last book was. The main effect of these was a slap in the face, returning us to a world that felt very different – and more boring – than the two worlds explored in the rest of the novel.
The length and dullness of some parts of the book was ameliorated for me by the exciting plot twists. Or, I should say, by the expectation of exciting plot twists, since Martin’s style is as much about building tension as it is about shocks. So actually, although there were slow patches, there were enough thrilling sections to make it, on balance, pretty exciting. Unfortunately, there are very, very few actual shocks in this book, and although I really enjoyed reading it this time, I fear that on a re-read, now that I know what does and doesn’t happen, I would probably find much of it quite boring and hard to chisel through.
On the subject of shocks: Martin came to fame as a ‘brutal’ author, somebody who could kill off anybody, no matter how important they were, no matter how loved they were. Where did that Martin go? I understand that it’s harder to kill characters now, because now that we’re closer to the end of the story, the surviving characters are more likely to actually be important, rather than being decoys. And yet Martin retains his obsessive need for cliffhangers and life-in-the-balance thrills – which works a lot less well once you realise that everyone’s safe. I lost track of how many dead, believed-dead, probably dead or almost dead characters turned out to be alive in this book. It’s ridiculous. This meant that when, toward the end of the novel, there were what looked to be a major plot twist, I didn’t really care, because I knew that it wouldn’t really matter. Everyone affected will turn out to be never-dead, not-quite-dead or dead-but-not-dead in some way or other. So what’s the point?
Lastly: the problem I identified in AGOT regarding background characters is still there. Martin’s background characters seem to fall into two categories: Boba Fetts, who, it is suggested, are Deep and Complicated and Mysterious and Cool and Dangerous even if it’s not always clear how exactly, and Other People, who are mostly cartoons. Both tendencies become annoying over time.
By now, the criticisms have probably lead you to think that this isn’t a book worth reading. I don’t know whether that’s what you should think or not. By itself, it stood alone as an enjoyable reading experience, but of course it cannot actually stand alone, inasmuch as there are four books before it and at least two more to come. As I say, I enjoyed reading this: but given its length, it was also a major investment. Does that investment pay off? I think we have to wait and see. If Martin manages to pull off the end of this cycle well, this novel will stand proudly as a strong, enjoyable, well-written part of the whole – not perhaps the most plot-advancing part, but an interesting and engaging slower middle movement. If, on the other hand, the end of the series doesn’t succeed, this novel will stand sadly as an overlong, over-ambitious symptom of a promising writer losing their way. If the series fails as a whole, there’s not enough in this volume to be worth wading your way through it all; if it succeeds, there’s enough in it to be a valuable part of the whole that will really add to the experience of it all. I really don’t know which way it will turn out. I think primarily this novel has heightened the ambition – I’m now less confident that it will end well, but if it does end well, it’ll be even better than I thought it would be.
Finally, of course the higher quality of this novel relative to the first in the cycle should stand as a proxy for the quality of the intervening novels, which I have not yet re-read (and won’t for a while – too many pages read for now).
Adrenaline: 4/5. I may be being charitable, because there were long, boring bits. On the other hand, there were a lot of non-boring bits as well. Some bits were genuinely heart-pounding.
Emotion: 2/5. I did connect with some of the characters, but few of them had really emotional stories in this instalment – I guess it’s a slow period of set-up and construction, rather than the part of the story that tugs at the heart-strings. There’s certainly emotional potential here…
Thought: 3/5. It’s a vast and complicated plot, which kept my mind very active. On the other hand, no great mysteries really grabbed me, and for the most part I was willing to sit back and watch things unfold.
Beauty: 4/5. The prose is adequate, though there are nice bits here and there. The strong point in this respect is the scenery, which is strikingly filmic at many points.
Craft: 4/5. Assuming that the entire novel doesn’t turn out to be a mistake – which will depend on what happens next – there are no massive blunders here. There are certain repetitions in the prose, and certain plot twists feel too sudden and unjustified (in one case it seems as though the author expects readers to have been reviewing internet speculation for the last five years, since one familiar-to-scholars-of-the-novels development (no, not THAT one!) is treated as though fairly obvious and uninteresting, but to virgins in the series it’ll be baffling, shocking and entirely unforeshadowed). There are slow periods, and page-time is misallocated. But nothing really hurts too much. On the other hand, some scenarios are built up very well, there is an astonishing control over a staggeringly huge cast (you could spend hours reminding yourself where various minor characters have been glimpsed before, or who they’re related to and why that matters), and some chapters could stand well as polished short stories in their own right.
Endearingness: 3/5. Some POVs I really enjoyed; others I survived. I could see myself selectively reading the good ones again, but the bad ones I don’t like the thought of having to re-read. So it balances out, I think.
Originality: 3/5. Bound by the limitations of its setting and structure; nonetheless, its sprawling plot and travelogue elements do make it a bit more exotic, and unusual, than earlier books in the series.
Overall: 5/7. Good. It’s definitely a step up from the first book, and as far as I remember it’s at least equal to, and possibly superior to, the second and third volumes in terms of writing quality – although it can’t match the third volume for thrills, as rather less happens here. As I say above, the ultimate verdict will have to wait until the series is completed, but this book was certainly good enough to make me want to read The Winds of Winter when (if!) it comes out, and it gives me confidence that the second and third volumes may actually be as good as I remember them being (which is very good indeed).