![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
Plot Structures
Jan. 2nd, 2015 02:48 pmImagine that, sometime before all the media hype and before the creation of the derivative fantasy genre (say in the late 1950s), you sat down to read The Lord of the Rings with no spoilers, and you were a few chapters into it (say, at the end of "A Shortcut to Mushrooms"). Would you have any real idea of what was to come next?
Remember, this is before the crowd of derivative works had created a trope of a quest to bring down a Dark Lord. In fact, the quest as a plot model had fallen into relative disuse (Auden has to dust it off in putting it forward as an explanatory mechanism when writing about LOTR).
What was ahead would have been almost entirely unpredictable. Even after the Council of Elrond, when you encounter elements such as Isildur's heir, or hear of Rohan, Orthanc, and Minas Tirith, and know that the announced thrust of the book will be Frodo carrying the Ring to Mordor, the shape of what is in front would be unguessable. (This is not a plot coupon story; more of that later.)
At the same time, there is nothing arbitrary about the events. Within the scope of the story, there is very little that is pure chance ("...if chance you call it") given the motivations and abilities of the characters. Bilbo's picking up the Ring is chance (or divine intervention) in the plot inside the Hobbit, as Gandalf recognizes, but it's already established background for the whole of LOTR. (There's an excellent case to be made that the timing of Merry and Pippin coming to Fangorn is as well, (and that of Frodo and Sam meeting up with Faramir) but in both cases what all the characters are doing is non-arbitrary in terms of their own world-lines.) Although LOTR (like most realistic novels) incorporates some coincidence, it is not a coincidence-driven plot any more than it is a predictable one.
Or consider Anathem. For much of the book, as with LOTR, it is a narrative of discovery for the reader (and for the narrator). And again, what is discovered is not arbitrary, but integral to the world and the people depicted.
This is not confined to SF/Fantasy, where the broad genre expectations are for wonders and the unexpected. Take a look at Gaudy Night, a number of chapters in. We know it's a detective novel, so we have basic expectations of a problem to be solved (successfully -- this is not a subversion like Trent's Last Case), and a perpetrator to be unmasked. But the shape of the plot developments is not predictable for the reader, regardless of the fact that the action of the novel grows integrally out of the characters' motivations. (As far as chance goes, I would have to count only the collision with Lord Saint-George -- and that's a minor coincidence, in a small city like the centre of Oxford.)
Just to extend the genres represented, I'll hold up The Grand Sophy. Again, the genre gives us some very broad expectations (Sophy and Charles will make a couple by the end of the book). But the plot is a set of broadly unexpected twists, and the sequence of events grows organically out of the characters' motivations.
I'm going to make a carefully hedged claim here: in general, barring special cases, most of what we are likely to think of as the best books have both there traits: the plots are at best only very generally predictable, and the events which make them up are neither coincidental nor arbitrary once one accepts the set-up of the book.
Now consider, in contrast, two other sets of novels.
First: Flag in Exile, Murder on the Orient Express, 1636: The Kremlin Games, and A Lady of Quality.
In each of these books, a diligent reader can guess at a good deal of the detailed shape of the book after a few chapters (e.g. Honor will not only come back into action as an Admiral at Grayson, but will have to deal with both a Havenite attack and internal opposition from Grayson conservatives, both of which she will carry off beyond any reasonable expectation; Poirot will examine each potential suspect / witness and gradually put together his theory, keeping it from everyone until a final recognition scene[1]). In the Flint novel, there's a general shape of plot that has become a standard form in the 1632 novels (uptimer is the central figure in a process of changing / revolutionizing another culture which involves conflict with both conservative downtimers and frequently political or personal conflict with other uptimers; if you know anything about Muscovy in the period (say, from reading The Ringed Castle), you can guess a lot more). In the Heyer case, the general nature of the novels' conflicts, both intra-familial and extra-familial, are very similar to those in many of her previous works.
These novels "work", and are in fact all very popular. But the reading experience is conditioned not by a sense of meeting the unexpected, but by navigating a broadly expected course.
This is not simply a contrast of good versus poor craft. (This is one reason I hedged above.) Weber, for example, typically builds his plots by telegraphing broadly what is going to happen long before he actually shows it; this is a deliberate choice of a type of plot construction which is not uncommon in the thriller or military novel: consider Tom Clancy. The reader is in a privileged position of knowing broadly what is going to happen and the pleasure is gained from seeing just how it plays out, or how apparently unbeatable odds are met by the protagonist. (My personal judgement tells me that this is a good way of making good light reads but rarely for even approaching the very good let alone great: one advantage of first-person narrative or tight third person is that it makes this sort of plot far more difficult to write.) And there are great works, usually in a tragic vein, which gain most of their impact from the audience knowing ahead of time what is to happen and seeing the inevitable play out. (The Oedipus Tyrannos is the classic (in both senses) example; Paradise Lost, which announces its plot in the first verse paragraph, is another.
The plot coupon novel is a special case of the predictable plot. The hero (this is frequently a "quest" or heroic fantasy) is told of a set of tasks which have to be accomplished, leading up to a final objective, at the outset of the novel, and he (usually he) or his companions set about collecting the coupons. Sometimes even the order is decreed in advance. Eddings' series are built this way; so are Riordan's. (Internally, these are justified by "prophecy". but it always strikes me as lazy plot-building.) One trait which makes this a particularly weak plot type is that these coupons are typically arbitrary, independent of one another, and the novel becomes a series of disconnected episodes rather than an organic whole. It's popular in YA fantasy, though, possibly because it provides a simple structure for the reader to measure the progress through the novel by.
Consider, in contrast, another set of novels: The Three Hostages, The Shambling Guide to New York City, and Diplomatic Immunity. These are not entirely predictable novels, but they are driven by entirely too much coincidence. DI not only begins with the coincidence of Miles contact with Bel, but by the fact that he's basically the only person in the galaxy who has the contacts and the knowledge to deal with a situation he's assigned to simply because he's nearby to it. (I mean, would anyone else within several parsecs even have recognized the sigil for the Star Creche?) Lafferty's novel -- which got her plenty of attention, and a nomination for the Campbell award -- hinges on a whole set of coincidences, the biggest being that the conflict which drove the heroine to New York in the first place, and which has no logical connection to the world she has fallen into, turns out to involve a major villain within that world. Buchan's (deservedly classic) thriller is largely worked out organically, but begins with just too may coincidences to be true. Hannay just happens to be able to make contact with Medina just at the same time as he has been handed the kidnapping to solve (and have Medina decide that he would be a useful tool), just happens to be a poor hypnotic subject, and just happens to be a good friend of Sandy, who is Medina's ideal antagonist. (Buchan plays games with the plot coupon theme by deliberately having a character talk about the technique of linking three arbitrary topics together, choosing three images which turn out to be critical to the whole plot.)
Coincidence is one aspect of the more general case of the arbitrarily driven plot. In a plot driven by coincidence, the arbitrariness accompanies things which are intrinsically improbably and advance a plotline; more generally, episodes which are introduced with no organic connection to the central drivers of the plot may not be "coincidence" -- they advance no plot, they're there just for themselves -- but they are surely random.
Again, this can be used in great works as well as mediocre ones. In tragedy, coincidences which keep shutting down options for the protagonist create a sense of struggling vainly against fate (consider Romeo and Juliet). In comedy, coincidence becomes a technique of its own, precisely because expectations in comedy are explicitly removed from those in reality. (And frequently the point of a coincidental episode in a comedy is not to advance a plot arc but to provide an incongruous episode which is humorous for its own sake: consider the scene in Crispin's Holy Disorders with the pet raven.) Once it becomes a structure in itself, coincidence can work very well: the picaresque novel is basically a set of random encounters strung along the thread of the central character's experience; so is Christian's pilgrimage in Bunyan, or Quixote's journey in Cervantes. (A fully coincidence-driven plot moves away, again, from the predictable, frequently in the direction of the absurd.)
So where am I going here?
I realized, as I was looking back on last year's reading, that the books which I rated most highly were those which had an organic development of a set of initial conditions (minimal coincidence -- not necessarily none -- you can get away with one or so simply because that's sometimes just how life is and it can, used carefully, increase rather than weaken verisimilitude) combined with an essential unpredictability in the reading experience: Hild, City of Stairs, Ancillary Sword, The March North, Three Parts Dead.
Neither of these structural principles is exactly unobvious, but it can be interesting to sort one's favourite (or not-so-favourite) books into heaps based on these criteria. Sometimes books by the same author fall into different heaps: Bujold loves coincidence-driven plots, but Memory is far less coincidence-driven than her other books; Heyer is all over the map; Pride and Prejudice is much less coincidence-driven than Persuasion, and Mansfield Park even less so (but MP telegraphs its plot more than do the others).
[1]The contrast with Gaudy Night is not in recognizing that the solution will be reached and the reasoning revealed, as this is a demand of the genre itself; it's that the nature of the plot advancement is generally foreseeable in Christie's case (because of her previous use in other novels) whereas this is not the case with GN by reference to prior works of Sayers, or anyone else.
Remember, this is before the crowd of derivative works had created a trope of a quest to bring down a Dark Lord. In fact, the quest as a plot model had fallen into relative disuse (Auden has to dust it off in putting it forward as an explanatory mechanism when writing about LOTR).
What was ahead would have been almost entirely unpredictable. Even after the Council of Elrond, when you encounter elements such as Isildur's heir, or hear of Rohan, Orthanc, and Minas Tirith, and know that the announced thrust of the book will be Frodo carrying the Ring to Mordor, the shape of what is in front would be unguessable. (This is not a plot coupon story; more of that later.)
At the same time, there is nothing arbitrary about the events. Within the scope of the story, there is very little that is pure chance ("...if chance you call it") given the motivations and abilities of the characters. Bilbo's picking up the Ring is chance (or divine intervention) in the plot inside the Hobbit, as Gandalf recognizes, but it's already established background for the whole of LOTR. (There's an excellent case to be made that the timing of Merry and Pippin coming to Fangorn is as well, (and that of Frodo and Sam meeting up with Faramir) but in both cases what all the characters are doing is non-arbitrary in terms of their own world-lines.) Although LOTR (like most realistic novels) incorporates some coincidence, it is not a coincidence-driven plot any more than it is a predictable one.
Or consider Anathem. For much of the book, as with LOTR, it is a narrative of discovery for the reader (and for the narrator). And again, what is discovered is not arbitrary, but integral to the world and the people depicted.
This is not confined to SF/Fantasy, where the broad genre expectations are for wonders and the unexpected. Take a look at Gaudy Night, a number of chapters in. We know it's a detective novel, so we have basic expectations of a problem to be solved (successfully -- this is not a subversion like Trent's Last Case), and a perpetrator to be unmasked. But the shape of the plot developments is not predictable for the reader, regardless of the fact that the action of the novel grows integrally out of the characters' motivations. (As far as chance goes, I would have to count only the collision with Lord Saint-George -- and that's a minor coincidence, in a small city like the centre of Oxford.)
Just to extend the genres represented, I'll hold up The Grand Sophy. Again, the genre gives us some very broad expectations (Sophy and Charles will make a couple by the end of the book). But the plot is a set of broadly unexpected twists, and the sequence of events grows organically out of the characters' motivations.
I'm going to make a carefully hedged claim here: in general, barring special cases, most of what we are likely to think of as the best books have both there traits: the plots are at best only very generally predictable, and the events which make them up are neither coincidental nor arbitrary once one accepts the set-up of the book.
Now consider, in contrast, two other sets of novels.
First: Flag in Exile, Murder on the Orient Express, 1636: The Kremlin Games, and A Lady of Quality.
In each of these books, a diligent reader can guess at a good deal of the detailed shape of the book after a few chapters (e.g. Honor will not only come back into action as an Admiral at Grayson, but will have to deal with both a Havenite attack and internal opposition from Grayson conservatives, both of which she will carry off beyond any reasonable expectation; Poirot will examine each potential suspect / witness and gradually put together his theory, keeping it from everyone until a final recognition scene[1]). In the Flint novel, there's a general shape of plot that has become a standard form in the 1632 novels (uptimer is the central figure in a process of changing / revolutionizing another culture which involves conflict with both conservative downtimers and frequently political or personal conflict with other uptimers; if you know anything about Muscovy in the period (say, from reading The Ringed Castle), you can guess a lot more). In the Heyer case, the general nature of the novels' conflicts, both intra-familial and extra-familial, are very similar to those in many of her previous works.
These novels "work", and are in fact all very popular. But the reading experience is conditioned not by a sense of meeting the unexpected, but by navigating a broadly expected course.
This is not simply a contrast of good versus poor craft. (This is one reason I hedged above.) Weber, for example, typically builds his plots by telegraphing broadly what is going to happen long before he actually shows it; this is a deliberate choice of a type of plot construction which is not uncommon in the thriller or military novel: consider Tom Clancy. The reader is in a privileged position of knowing broadly what is going to happen and the pleasure is gained from seeing just how it plays out, or how apparently unbeatable odds are met by the protagonist. (My personal judgement tells me that this is a good way of making good light reads but rarely for even approaching the very good let alone great: one advantage of first-person narrative or tight third person is that it makes this sort of plot far more difficult to write.) And there are great works, usually in a tragic vein, which gain most of their impact from the audience knowing ahead of time what is to happen and seeing the inevitable play out. (The Oedipus Tyrannos is the classic (in both senses) example; Paradise Lost, which announces its plot in the first verse paragraph, is another.
The plot coupon novel is a special case of the predictable plot. The hero (this is frequently a "quest" or heroic fantasy) is told of a set of tasks which have to be accomplished, leading up to a final objective, at the outset of the novel, and he (usually he) or his companions set about collecting the coupons. Sometimes even the order is decreed in advance. Eddings' series are built this way; so are Riordan's. (Internally, these are justified by "prophecy". but it always strikes me as lazy plot-building.) One trait which makes this a particularly weak plot type is that these coupons are typically arbitrary, independent of one another, and the novel becomes a series of disconnected episodes rather than an organic whole. It's popular in YA fantasy, though, possibly because it provides a simple structure for the reader to measure the progress through the novel by.
Consider, in contrast, another set of novels: The Three Hostages, The Shambling Guide to New York City, and Diplomatic Immunity. These are not entirely predictable novels, but they are driven by entirely too much coincidence. DI not only begins with the coincidence of Miles contact with Bel, but by the fact that he's basically the only person in the galaxy who has the contacts and the knowledge to deal with a situation he's assigned to simply because he's nearby to it. (I mean, would anyone else within several parsecs even have recognized the sigil for the Star Creche?) Lafferty's novel -- which got her plenty of attention, and a nomination for the Campbell award -- hinges on a whole set of coincidences, the biggest being that the conflict which drove the heroine to New York in the first place, and which has no logical connection to the world she has fallen into, turns out to involve a major villain within that world. Buchan's (deservedly classic) thriller is largely worked out organically, but begins with just too may coincidences to be true. Hannay just happens to be able to make contact with Medina just at the same time as he has been handed the kidnapping to solve (and have Medina decide that he would be a useful tool), just happens to be a poor hypnotic subject, and just happens to be a good friend of Sandy, who is Medina's ideal antagonist. (Buchan plays games with the plot coupon theme by deliberately having a character talk about the technique of linking three arbitrary topics together, choosing three images which turn out to be critical to the whole plot.)
Coincidence is one aspect of the more general case of the arbitrarily driven plot. In a plot driven by coincidence, the arbitrariness accompanies things which are intrinsically improbably and advance a plotline; more generally, episodes which are introduced with no organic connection to the central drivers of the plot may not be "coincidence" -- they advance no plot, they're there just for themselves -- but they are surely random.
Again, this can be used in great works as well as mediocre ones. In tragedy, coincidences which keep shutting down options for the protagonist create a sense of struggling vainly against fate (consider Romeo and Juliet). In comedy, coincidence becomes a technique of its own, precisely because expectations in comedy are explicitly removed from those in reality. (And frequently the point of a coincidental episode in a comedy is not to advance a plot arc but to provide an incongruous episode which is humorous for its own sake: consider the scene in Crispin's Holy Disorders with the pet raven.) Once it becomes a structure in itself, coincidence can work very well: the picaresque novel is basically a set of random encounters strung along the thread of the central character's experience; so is Christian's pilgrimage in Bunyan, or Quixote's journey in Cervantes. (A fully coincidence-driven plot moves away, again, from the predictable, frequently in the direction of the absurd.)
So where am I going here?
I realized, as I was looking back on last year's reading, that the books which I rated most highly were those which had an organic development of a set of initial conditions (minimal coincidence -- not necessarily none -- you can get away with one or so simply because that's sometimes just how life is and it can, used carefully, increase rather than weaken verisimilitude) combined with an essential unpredictability in the reading experience: Hild, City of Stairs, Ancillary Sword, The March North, Three Parts Dead.
Neither of these structural principles is exactly unobvious, but it can be interesting to sort one's favourite (or not-so-favourite) books into heaps based on these criteria. Sometimes books by the same author fall into different heaps: Bujold loves coincidence-driven plots, but Memory is far less coincidence-driven than her other books; Heyer is all over the map; Pride and Prejudice is much less coincidence-driven than Persuasion, and Mansfield Park even less so (but MP telegraphs its plot more than do the others).
[1]The contrast with Gaudy Night is not in recognizing that the solution will be reached and the reasoning revealed, as this is a demand of the genre itself; it's that the nature of the plot advancement is generally foreseeable in Christie's case (because of her previous use in other novels) whereas this is not the case with GN by reference to prior works of Sayers, or anyone else.
