Art can be broadly classified into two groups: narrative and non-narrative. Within the narrative group, there is a further division between those art-works, which I will call closed, that are complete in and of themselves, and those, which I will call open, that were made deliberately incomplete, so that they would force their consumer to take up the work the artist has started, whether it be by exercising their imagination, or by analysing issues raised within the art-work (it is worth noting that by this I do not mean analysis along the lines of ‘But what would Fanny Price have done had not Henry Crawford committed adultery with Maria Rushworth?’ but rather consideration of questions deliberately posed by the art-work but left unanswered).
My purpose in this essay is simple, but ambitious: to attempt to understand what characteristics there are of a narrative art-work that mark out to which of these two classes it belongs. My conclusion is simultaneously negative (there are no overt characteristics of an art-work that allow one to point at them and say ‘That means this is open’) and positive (everything depends on the attitude that the artist brings to the act of creation).
This may seem a rather abstract endeavour; it is not. In fact, it is a necessary preface to a wider work I am undertaking which considers the translations of narrative art from one medium to another: most obviously filmic adaptations of books, plays, opera, etc. The emphasis in this wider work will be on filmic adaptations, hence the reason why the analysis below uses examples drawn from the medium of moving pictures.
For the purposes of this piece I will limit myself to narrative art-forms: novels / stories, theatre / opera, film, narrative painting. It is arguable (and I will argue it elsewhere) that my definitions could be extended to other art-forms, e.g. non-narrative painting, but it simplifies the argument if the art-work being analysed has the intention of telling (in however complex a manner) a story. My examples will (nearly) all be taken from the world of film, but it will not be hard to find examples in other media.
The definition, then, is quite simple. A piece of narrative art is open if it is deliberately constructed so as to provoke its consumer’s imagination. So it does not provide a simple story that answers all questions at the end. Another way of saying this is that an open art-work is not complete in and of itself: it requires the active collaboration of its consumer’s intellect.
A piece of narrative art is closed if it is not open. So it aims only to entertain or inform; that doesn’t mean it can’t provoke, but it intends to provoke specific actions that it sets out, whereas an open art-work merely provides the starting-point for a train of thought that could lead anywhere.
Another way of saying that is this: closed art wants to see the world in (figurative) black and white. It knows exactly what its consumer should think, and proceeds to communicate that to her. Open art has no answers; some ideas may be preferred to others, but there is no certitude, so it knows only a multitude of shades of grey.
What I want to do here is to see if there is any way of determining from external formal characteristics of an art-work whether it is open or closed. So, let me begin with one obvious candidate: the ending. At the end of the narrative is everything neatly tied up, so the consumer understands exactly what happened, and why it happened (which I will call a resolved ending), or is the consumer left with questions regarding the narrative: why events happened, exactly what was going on, etc (which I will call an unresolved ending).
So, consider Birth. In this movie the heroine’s dead former husband may or may not have been reincarnated in the form of a small boy. At the end of the movie, the heroine appears to separate herself from the boy and remarries, but the final scenes give no resolution. Moreover, the movie is very finally balanced so that it is quite possible to argue with equal plausibility that the boy is and isn’t the dead husband. The film, very delicately, opens up questions about identity, what it means to be an adult or a child, the nature of love and the possibility of sexual love between adults and children, but it only pushes at the door: it gives no answers. So it has an unresolved ending and is open.
Now consider The Big Sleep. In this case, notoriously, a scene in which Marlowe attempted to explain whatever it is that was going on was actually cut from the film to make way for more footage of Bacall and Bogart doing their thing. The plot is so complex as to be incoherent: there is no logic; the only thing tying events together is that one happens after the other. At the end we know that Marlowe loves Vivian Rutledge, and that’s about it; everything else is a mystery. So we have a very unresolved ending, but the movie itself is clearly closed (and here one must also mention the Coen brothers’ homage, The Big Lebowski, which is cheeky enough to have an unresolved ending, but to pretend to have resolved it with a resolution that is, on examination, totally nonsensical).
Now consider Secretary. The heroine is a troubled young woman who gets a job as a lawyer’s secretary. She falls in love with him, and gradually persuades him to realise that he loves her. They marry. The end. About as resolved a plot as one can get: it could have come from a Fred and Ginger movie. But now we get on to the bit that would have sent Fred and Ginger scurrying away. The heroine has a history of self-harming; what starts off the process of her falling in love with her boss is his introduction of her to his little kink: sadomasochism. And their subsequent marriage has at its core a sadomasochistic relationship that they both find fulfilling and exciting. So what is generally a taboo subject is forced on our attention: when does something cease to be loving and become abusive? What is permissible within the bounds of a loving relationship? The movie provides no answers, but it forces us to ask these questions, so it is open.
Finally, consider The Thin Man. This is clearly a closed movie: a (relatively) straightforward comedy detective story. It doesn’t pretend to be about anything more than entertainment. And what’s more, at the end Nick Charles very kindly summons all the characters (well, those who are still alive, that is) together and carefully explains exactly what happened. A perfectly resolved ending.
So, we have one open-unresolved movie, one closed-unresolved movie, one open-resolved movie and one closed-resolved movie. Therefore we conclude that the nature of the ending of an art-work’s narrative cannot be used as an indicator of whether it is open or closed.
This shouldn’t be necessary, but there will be a natural tendency to assume that open movies are somehow better than closed ones. Let us examine the evidence.
Consider Jean Cocteau’s Orphee. This is almost an exemplar of the open movie: it portrays a mysterious world in which mysterious events (such as a voice reading very strange experimental poetry over the radio, but such that only one receiver can pick it up) are taken as normal by the inhabitants. Some aspects of the plot are obviously derived from the Orpheus myth (the crowd of vengeful women are clearly the Furies), some have relatively simple symbolism (the cafe is the world), but on the whole we are left to attempt to make sense of what we see with no help from the movie. It is also, by common consent a good, no a great movie.
On the other hand, consider Glen or Glenda. This is definitely thought-provoking: it brings the whole cross-dressing / trans-gender issue to the viewer’s attention, and inevitably provokes consideration of the nature of gender, and how it relates to sex. It also contains some provocative material relating to the nature / nurture debate. So it is open. It is also, by common consent, one of the worst movies ever made.
Now consider Bringing up Baby. It need scarcely be said that this movie is closed (the only real thought it provokes being to wonder what it would be like being married to a woman that deranged): it is pure entertainment and has no intention or desire to be anything more. And it is a great movie, with brilliant performances by the whole cast: Katharine Hepburn’s almost unhinged performance may stand out, but there are no weak links.
Finally consider Batman and Robin. Alas, one cannot say the same of it. Its (mis)casting is extraordinary, and it is deservedly considered quite dreadful. And it is the epitome of closedness; its plot can be summed up by saying that a bloke wearing a funny costume beats up other blokes wearing equally funny costumes, while in the background things explode and women are helpless (though, this is, of course, the plot summary of every Batman movie ever made). There is nothing there to provoke the intellect (I do not count wondering why anyone ever green-lighted this disaster).
So, we have one good-open movie, one bad-open movie, one good-closed movie and one bad-closed movie. Therefore we conclude that there is no direct relationship between openness and quality. I would, however, be prepared to posit (without being able to prove it) that in a statistical sense there may be such a relationship, i.e. that an open movie has a higher probability of being good than a closed movie.
Now consider plot. Is there a relationship between the nature of the plot and whether the art-work is open or closed? This can be broken down into several sub-questions, so let us start by asking whether complexity of plot is indicative of being open.
So consider Georges Franju’s Nuits Rouges. This has a plot so complex that I will make no effort to summarise it; suffice it to say that there are many mysterious and secretive powerful groups that exist in parallel with the official executive powers of the State. And we are left with all kinds of questions about who they are, what they are, and what exactly was going on, as well as any number of startling images. Is the overt machinery of the state a sham, with real power wielded by one or more shadowy organisations locked in eternal conflict? Is there an esoteric world parallel to the exoteric one, the two meeting only occasional? So it’s open with a complex plot (the same is true of Orphee).
Now consider Ocean’s Eleven (either version, though especially the remake). This has an extremely complex plot, and yet is a pure piece of entertainment. There is no intellectual agenda, or intention of provoking thought, only a desire to shock, surprise and amuse. So it is closed with a complex plot.
What may be, in many ways, the ultimate open movie is Synecdoche New York, which achieves it not through plot (the plot, in as far as there is one, is actually quite banal) but through a complex multi-layering that means that we are never sure whether the events we see are real or not, and how many degrees of separation there are between them and reality, which means that our whole concept of reality is called into question. Something similar happens with Dogville. The plot is very simple, and at the end we know exactly what happened. We just have no idea what it signifies: it is clear that something larger, beyond the mere plot, is going on, that the actions are symbols rather than mere acts, and we are told (next to) nothing about what that larger something might be. So these movies are open with simple plots.
Finally consider The Muppet Movie. This is pure entertainment, and it has a very simple linear plot: the classic quest plot, in fact. So it is closed with a simple plot.
So, we have open-complex, closed-complex, open-simple and closed-simple moves, so there is no direct relationship between being open or closed and the complexity of the plot. The next question is, does plot matter at all: that is to say, does the plot necessarily influence whether a movie is open or closed? Note the use of the word ‘necessarily’: obviously there will be such an influence in some cases; what I am asking is whether there is a universal functional relationship between the two parameters of plot and open / closed.
First some antinomic pairs of films which are either similar or identical in terms of ‘plot’, but which differ as to whether they are open or closed. So, two films that (entirely in one case, partially in the other) deal with penguins. March of the Penguins is strongly closed: it knows what it wants you to think and tells you what that is. Encounters at the end of the World, even if we limit ourselves to only the section relating to penguins, is open: it presents some startling images and facts, such as a penguin deliberately setting off on a journey that must inevitably lead to its death, and wants you to think about them yourself.
Now consider Solaris. Tarkovski’s film is wide open: though famously Lem did not like it, and though it changes the ending of the novel, it produces a sense of wonder, because we feel always the looming presence of the ocean; and we never really understand what the ‘visitors’ were: all we learn is how to get rid of them. And Tarkovski’s ending raises many questions regarding just how real what we have been witnessing actually is: were the scenes on Earth real? Was any of it real? What is ‘real’? Tarkovski also uses images to create a sense of wonder and provoke the imagination, e.g. in the sequence of driving through the streets of Tokyo. The Soderbergh film, on the other hand, chooses to remove the ocean entirely, concentrating instead almost entirely on the psychology of Kelvin and his love for ‘Rheya’. Rather surprisingly, for a movie ostensibly about a planet-wide sentient ocean, there is no sense of wonder, or of the alien; ‘Rheya’ behaves just like a woman, not like a construct made in the form of a woman. The movie is closed. So two movies with the same ‘plot’ manage to differ as to whether they are open or closed.
Now some other examples. First two Terry Gilliam movies. Both Brazil and Time Bandits present fragmentary views of very complex worlds, and force us to guess what is going on in order to understand the movie. Time Bandits is open. The Supreme Being says that Kevin must stay on to continue the struggle, but the struggle against whom: Evil or the Supreme Being? Does Kevin become the new embodiment of Evil, or a knight in shining armour? Did any of it even happen? And if it didn’t, what on Earth is going on in the final scene? Is Evil necessary to the functioning of the Universe? And what about the question, the answer to which is ‘something to do with free will’? Brazil, on the other hand, is closed. One might, at a very low level, discuss how such a society might work, but there are no big ideas and little ambiguity. So despite have a similar approach, these two movies differ as to whether they are open or closed.
And finally, an example to demonstrate just how disconnected openness can be from plot. Barbarella has either an insanely complex plot, or no plot at all, depending on how you look at it, but that doesn’t matter, because plot is not what the movie is about. It’s a succession of startling images and ideas (feeding prisoners on orchids?) that can’t fail to provoke the imagination simply because they are so bizarre. It is therefore open and the plot (if there is one) is no more than a clothes-line on which to hang interesting events.
Thus it appears that subject-matter of a movie does not necessarily influence whether it is open or closed. Rather, it is beginning to look as if what matters is the attitude that the creators brought to the process of turning raw material into a movie.
But there is one final pair of parameters: age and whether or not the film is mainstream or ‘art house’. I am not certain, but my suspicion is that ‘old’ movies could be open and yet still remain part of the mainstream (e.g. Woman of the Year, The Birds, The Forbidden Planet, Fantasia), while now we find that such open movies as there are (and there clearly are some, as can be seen from the comments above) are most likely to be niche products (though South Park: Bigger, Longer and Uncut is an honourable exception). I find that hypothesis somewhat disturbing, and rather hope that it isn’t true.
Finally, note the interesting phenomenon of accidental openness. I very much doubt that Billy Wilder intended Some Like it Hot to provoke intellectual debate, and yet it can form the vehicle for discussion of the relationship (mentioned above) between sex and gender. Who is more real: Jerry or Daphne (Jerry seems to take to being a woman far more than does Joe)? Likewise, My Man Godfrey was almost certainly intended to make one laugh, not think, and yet if one is in the mood, there is all kinds of interesting material to be mined: the life of the idle rich vs the working poor; love vs infatuation, etc.
So, it seems that there is nothing inherent in the structure of the art-work itself that dictates whether it is open or closed. All we can say is that it’s something about the attitude of the art-work’s creators. This chimes with Roger Ebert’s observation (in his review of the remake of Psycho, where he noted that while the remake was not very good, the original is a masterpiece) that anauteur‘s genius is to be found ‘between the frames’, because it seems to be independent of the material being filmed.