There are many people ready for this strange, essayistic phase of Terrence Malick's career to be over, and I get that, though it's not like we're inundated with associatively-edited mood pieces. Three films out in the entire world in a five-year period doesn't seem like too many to deal with. But anyway, Malick has been up to some incredibly weird shit ever since he got The Tree of Life out of his system back in 2011, and not all of it has been consistently successful. According to many standard systems for evaluating narrative cinema, none of it has been at all successful, but I dunno. I was pretty head-over-heels for Knight of Cups, and I'm pretty sure at this point that I need to give To the Wonder another look (as for Voyage of Time, I'll let you know whenever the distributor decide to give that film a comprehensible release outside of the natural history museum in fucking Indianapolis).

Anyway, the director's third (and, to appearances) final film in this funky little sidebar to his career is a certain Song to Song, which pretends to be be a love triangle set against the Austin music scene. And at the level of pure denotative narrative content, that's close enough to accurate, but anybody who wanders into this film on the expectation that they're in for some denotative narrative content is in for an even sadder experience than they had with Knight of Cups. That film at least had a clear-cut, explicit structuring element: it was a narrative interpretation of the Tarot (alright, so maybe not "clear-cut", but definitely explicit). Song to Song, while set in a much more familiar world, and animated by much more familiar emotional themes, genuinely does not seem to have any macro-structure in place at all.

It picks up four different characters - I don't think a single one of them is named in the movie, though all are identified in the credits - and sees how they interact. These are Faye (Rooney Mara), a guitarist who has some fuzzily-defined, probably sexual past history with Cook (Michael Fassbender), a producer; she is, however, currently with struggling songwriter BV (Ryan Gosling), or at least they hook up somewhere early in the film (as far as I can tell, the plot takes place in one singe chronological direction). Cook grows attached to a waitress who wants to be a music star, Rhonda (Natalie Portman), but very quickly starts to emotionally abuse and exploit her. Around them, many musical stars, some apparently playing themselves, put in appearances, and so do the usual gamut of actors who have signed on to be in a Malick film despite the total lack of any assurance that they'll actually end up with screen time in the movie (Val Kilmer and Holly Hunter are the two most noticeable examples - Cate Blanchett also shows up in a large enough role that it doesn't just feel like a cameo - while Christian Bale, of Knight of Cups and before that The New World, was cut out entirely); the music stars generally get more screentime, and are more interesting, most especially Patti Smith playing, basically, herself as Yoda.

As with every Malick film of the modern era - here defined as beginning with 1998's The Thin Red Line - Song to Song was quite obviously found in the editing room by the director and the three editors, Rehman Nizar Ali, Hank Corwin, and Keith Fraase (all of whom have worked with him before). It is, in fact, maybe the most conspicuously-edited of Malick's films, simply because editing is, to a certain extent, the only thing the film consists of. I am certain that I could, if pressed, tell you what happens in terms of plot and story (I've mentioned this before, but not for many years, and it's worth recapping for this film of all films, what the difference between those two things: "story" is everything that happens, "plot" is what is presented onscreen and in what order), because it's pretty damn straightforward, actually: Cook is a bully at beat, actively psychopathic at worst, but the other three can't ever shake his orbit; Faye and BV aren't totally working and so they call it quits, and fall into other relationships. Him with an older woman (Blanchett) who he thinks he adores even though she and everyone else knows he's just happy not to be alone; she with a succession of men and women that she keeps at a friendly remove.

But the plot, even the characters themselves, aren't the interest of Song to Song. Way back up there in the first paragraph, I called it "associatively-edited", and, I think, gets us the the film's actual interest. It's about concepts and emotional states of being, and insofar as it has humanoid characters played by professional famous people, I think that's only because Malick hasn't figured out how to evoke the feeling of disconnection, confusion, and loss that come as the immediate aftermath of a cordial but decisive break-up except through means of building around characters. I think he wants to - much as with Knight of Cups, which very much feels like the first draft of this film (but, like, one of those places where the first draft is actually better than the second, because it has more life and passion even though it's sloppier), I cannot but imagine that a version of Song to Song which full embraced its nature as an experimental film would be more satisfying and powerful by far than the version we've actually gotten which has some lingering feeling of obligation to be a narrative feature even if it is a shockingly dysfunctional and bad one.

The experiment, anyway, is with moving we, the audience, through the emotional states of the characters, tied largely to whomever most recently got to deliver a page or two of Malick's customary Plaintive, Philosophically Abstracted Voice-Over (the content of said voice-over matters much less, I think, than its use to signify the "protagonist" of any given moment) - Faye, BV, Rhonda, and Blanchett's Amanda all get such moments, and Cook pointedly doesn't. We are so moved through the impressionistic connection of moments, with the film generally accreting meaning over the course of sequences rather than generating it at the moment of the cut. I'm doing a shitty job of describing this, I'm sorry. I mean to say that you feel things inย Song to Song because those things iterate over one or two characters and over five or six extended sequences, and so at a certain point you're able to look back and realise that the film has been playing around with a certain kind of emotional state for a good 10 minutes now.

It's all very, very cool, if not necessarily very efficient or effective. The way the film is assembled, through chains of emotional logic rather than narrative clarity, means that Song to Song has no obvious shape, and could be as long or short as it took before the filmmakers got bored; this results in a 129-minute film that I am sorry to report felt a whole hell of a lot longer to me. But you go for the reverie as much as anything - if the film is obviously inspired by music in creating concrete feeling through abstract bits and pieces of art, it's definitely not the kind of music the film itself depicts, with words and short running time and everything; it's more like mid-century experiments in atonality that make a big show of the fact that it takes five uninterrupted hours to play properly. You go as much to lose yourself in your own thought as the art's concepts, and that's not at all without value, though again, Song to Song would benefit from being more conscious about this, and less anxious to pretend that we, any of us, care about Faye or BV (I will say, however, that Mara's performance as an abstracted field of emotions only incidentally required to resemble a human woman is also, by a huge margin, the most excited I've ever been to watch her in a movie).

Outside of the extreme shapelessness of its plot, Song to Song is basically just more of what To the Wonder/Knight of Cups-era Malick has been interested in pursuing: heavily aestheticising profoundly unexceptional contemporary American spaces through the lens of Emmanuel Lubezki's camera (the film is, beyond question, quite gorgeous, though it feels "easy" in a way that Knight of Cups, with its intoxicated exploration of GoPro cameras, never did) and creating a magnificent layer of defamiliarization; attempting to depict through solely visual means basic, essential human emotions about lust and love; people twirling around, though the twirling here is kept to a minimum. I remain fascinated and captivated by this style, which Malick is apparently done with now (his next film, Radegund, is in the can and apparently mostly edited and is not remotely like this trilogy), and which is basically unique in modern narrative cinema, even as I wish that any of the films involved had been unqualified successes, instead of really interesting, inconclusive experiments. If nothing else, there will be no other 2017 film that even remotely resembles Song to Song, and that would be more than enough to make me happy and extraordinarily grateful that it exists.