My Canon: "Alice Doesn't Live Here Anymore" (Martin Scorsese, 1974)

Alice Doesn't Live Here Anymore begins with a musical number – an homage to The Wizard of Oz, drenched in Technicolor red and filmed on a set that could only have been built in Hollywoodland. A young Alice belts out a showtune on a ranch somewhere in proverbial Kansas; she declares she can outperform Alice Faye, thus expositing her dream of becoming an elegant singer. Then a sudden smash cut thrusts us into the opening credits, and also into Scorsese territory, as the titles (written in semi-ironic, Blue Velvet-style cursive script) appear over a speeding tracking shot that races over the rooftops of Socorro, New Mexico, 27 years later. The sleek fantasia of classic Hollywood musicals is violently jarred with the fast-paced, no-frills, rough-edged, happy-sad milieu of 1970s New American cinema, a movement of which Scorsese was one of the foremost progenitors. The rapid-fire, profanity-laced dialogue enlivens what could have been a syrupy soap opera, ultimately creating what might be best described as hyper-naturalism.

Still halfway through shooting The Exorcist, Ellen Burstyn was offered the tantalizing opportunity to put another project into production at Warner Bros., with relatively unfettered creative control. As the actress is quoted in Peter Biskind's Easy Riders, Raging Bulls, Robert Getchell's script for Alice Doesn't Live Here Anymore found its way to Burstyn, albeit in slightly rosier, sleeker tones ("in a kind of Doris Day-Rock Hudson kind of way," she explained). She immediately began searching for a rougher-edged director who might inject some much-needed grit and despair into the proceedings; a viewing of Mean Streets convinced her that Scorsese was the right man for the job, though (by his own admission) he knew nothing about women.

Emerging in the middle years of American second-wave feminism, Alice Doesn't Live Here Anymore was both celebrated and decried upon its release for endorsing (or failing to) women's agency in modern American life. Burstyn and Scorsese agreed that they wanted to portray a newly strong and independent woman who came to the realization that she doesn't need the companionship of a man for security and happiness. A powerful indictment of the entitlement and domestic violence perpetrated by men (an indictment all the more sobering because it shares screen time with poignant humanism and breakneck comedy), Alice Doesn't Live Here Anymore follows its titular character after she and her son Tommy abandon their hometown, following the accidental death of her husband – a gruff, emotionally pent-up man who provides money and little else for the family (though it's a sign of the movie's compassion that even he, while unlikeably morose and distant, is hardly the violent monster that a more simplistic movie would have presented him as). Hoping to achieve her dreams of becoming a singer in Monterey, California (a dream she abandoned upon getting pregnant and marrying), the duo's limited funds only get them as far as Phoenix, where Alice stumbles into an affair with the movie's only truly horrible character: Harvey Keitel's Ben, who mercilessly beats his wife in front of Alice when she finds out he's having an affair. Fearfully protecting Tommy from Ben's violence, Alice and her son move on to Tucson, where she lands a job as a diner waitress and reluctantly falls in love with David, the strong-silent type who's given considerable depth and sensitivity by Kris Kristofferson's performance. Having experienced only violence and alienation from the men in her life (aside from Tommy, with whom she has a jokey, intimate, completely naturalistic relationship), Alice holds any kind of relationship with David at arm's length – but, as often happens, their mutual attraction defuses any kind of self-professed insularity.

Certainly Alice Doesn't Live Here Anymore is no Jeanne Dielman or Riddles of the Sphinx, and its allegiance to feminism is shaky in some ways: it was, of course, directed by a man (although apparently the right man for the job, and the director whom Burstyn enlisted), but more debatable is the movie's happy ending. Rather than eschewing any kind of romantic relationship, David decides to follow Alice and Tommy to Monterey, where she can embark on her singing career. (A gorgeous performance she offers at a tiny nightclub in Phoenix reveals how promising her ambition actually is.) In fact, the original ending had Alice abandoning David, embracing her individuality and self-reliance; but Warner Bros., seeking a more satisfying resolution, pleaded for a happy-ending compromise. Though Alice and Tommy's trek to Monterey in some kind of stoic solidarity would have proclaimed a stronger endorsement of feminism, the ending as it currently stands is more humanistic than ideological, asking the equally difficult question of whether those victimized in relationships can or should still find love in the world. Alice Doesn't Live Here Anymore does provide a romantic union between man and woman, but this seems more because Scorsese and his cast care so deeply for these characters, not out of any kind of capitulation to a patriarchal insistence on heteronormative relationships.

Scorsese is typically known for his immersive portrayals of criminal communities and his deconstruction of violent masculinity, yet I've always felt that his non-crime pictures – particularly After Hours, The King of Comedy, The Age of Innocence, Hugo, and of course Alice – are his most interesting. While Scorsese's aesthetic prowess is always on display in Alice Doesn't Live Here Anymore (the mobility of the camera in both the frenetic restaurant scenes and during Alice's nightclub performances are astonishing), even more impressive is his work with this incredible ensemble cast (which also features Diane Ladd as Alice's outspoken coworker and Jodie Foster, in one of her first film roles, as Tommy's cynical, wine-guzzling young companion). Ellen Burstyn achieves a deft balancing act between wisecracking resilience and veiled vulnerability, and her seemingly effortless believability in both this and The Exorcist (which are, of course, completely disparate roles), and her rapport with Alfred Lutter as Tommy (who was cast after auditioning 300-some young actors) has an acrobatic intensity.

If the movie had decided to devote its energies to either bleak working-class suffocation or zippy familial sitcom, it may have been an interesting time-capsule document (the movie's portrayal of southwest America in the mid-1970s is always a wonder to behold); but by deftly infusing its energetic comedy with unflinching portrayals of gender politics and domestic violence, Alice Doesn't Live Here Anymore becomes an overwhelming happy-sad combustion, in the manner of the best Billy Wilder or Ernst Lubitsch. Its methodology is vaguely postmodern: the stylized opening, an homage to classic musicals, reminds us that relationships in real life do not operate according to cinematic fairy-tale splendor. But the movie is too sincere, too compassionate, too in love with the unexpected turns that life takes, to completely deprive its characters of the happiness they so clearly deserve. An unheralded masterpiece in both Scorsese's and Burstyn's filmographies, Alice Doesn't Live Here Anymore is one of the most dazzling and humanistic treasures of 1970s American cinema.

"My Canon" is a series in which I analyze my 100 favorite films in detail, in alphabetical order. Here is my introduction.

My Canon: Introduction

The list is itself a collection, a sublimated collection. One does not actually have to own the things. To know is to have (luckily, for those without great means). It is already a claim, a species of possession, to think about them in this form, the form of a list: which is to value them, to rank them, to say they are worth remembering or desiring.

– Susan Sontag, The Volcano Lover

Do I have a favorite movie?... More important, should I have a favorite movie? The question provokes resistance in me even though I recognize it to be intellectually respectable in a way that a similar but (I assume) distinguishable question – What is the greatest movie ever made? – is not. And yet, in some ways, the silly question would be the easier to answer. 

– William S. Pechter, "These Are a Few of My Favorite Things"

In August of last year, when the British Film Institute (through their publication Sight & Sound) released their seventh once-a-decade Greatest Films poll, it instigated a flurry of articles and responses ranging from the fanboy-enthusiastic to the tactfully dubious. The former camp sees the Sight & Sound poll as an invaluable way of charting and cherishing the most treasured works in the history of cinema; Roger Ebert, for example, has described it as by "far the most respected of the countless polls of great movies – the only one most serious movie people take seriously." The latter camp, on the other hand – exemplified by such critics as Raymond Durgnat and David Thomson – see the poll as elitist, unfairly limiting, or a pointless "children's game" that may be momentary fun but hardly establishes an "official" canon of the best movies ever made, a task which may be impossible from the start. Each worthy selection suggests an equally tragic omission, as the polling is always at least somewhat prone to cultural zeitgeists and viewing availability.

The truth, as it often does, lies somewhere in the middle. Yes, polls of this kind are inevitably exclusionary: advising audiences which 50 (or even 250) films should be their viewing priorities unfairly sequesters other titles that might, in fact, end up being those viewers' favorites. On the other hand, semi-official canons are a valuable way to conceive of the entire history of an art form in a synoptic view, offering a shorthand guide to the most culturally vital, artistically innovative, emotionally powerful, and thematically complex works of cinematic art that have been made throughout history. When the majority of mainstream criticism today treats its vocation as a sort of Consumer Reports for moviegoers – judging movies based on the satisfaction they provide to audiences who are merely looking for a good story, as commodities which provide entertainment – this kind of well-intentioned, perspicacious critical overview is incredibly valuable. Furthermore, as Sight & Sound explains, the 2012 poll expanded its breadth to include 846 responses from 73 different countries entailing 2,045 films listed; if you're going to attempt a semi-official canon, it would be difficult to achieve one that's more eclectic and far-reaching than this one.

Which isn't to say there aren't problems: the results are still predominantly Anglocentric, with few female directors listed. Furthermore, as Michael Atkinson points out, the ability to seriously judge every movie in history by our own well-intentioned criteria is becoming increasingly difficult in a digital age, inundated with critical opinion: "As our digital intercourse about all things continues to grow like kudzu, threatening to involve practically every human being on Earth in open conversation, the feedback loops surrounding cultural investigation and appraisal of all kinds will get so pervasive that it may well become impossible, some day soon, to arrive at a truly singular and independent perspective on a film – much less hope that that perspective is attained by others independently as well, and might therefore constitute a valuable consensus about what that film is and how good it actually might be. Is such a questionable thing even possible, or are our poll-taking endeavours destined, in a fondly Camusian way, to long for a singular ‘truth’ that we know cannot exist, under any conditions?"

Andrew Sarris' "ballot" for the 1962 Sight & Sound poll

This being the case, one might assume that it's more valuable to attempt a personal canon – in other words, a list of "favorite" movies rather than "greatest" movies. As William Pechter suggests in one of the epigraphs above, the "favorite" question is more intellectually sound (though difficult to answer) than the "greatest" question, precisely because it depends on one's subjective response rather than a semi-objective range of established criteria. If forming and elucidating a subjective response is increasingly difficult (nigh impossible) given a digital discourse that provides critical and public opinion before many moviegoers even have a chance to see the movie, it becomes all the more valuable to formulate an idiosyncratic canon.

An even more pertinent question: why write about this six months after the poll was released? Am I not entering the fray a little too late? The answer is yes, but in my defense: before starting to build my personal canon, I wanted a little critical distance from the Sight & Sound poll, a little time to ponder its inclusions and exclusions, and needed a while to decide what are my favorite films of all time. This lengthy deliberation process may seem self-defeating: shouldn't a list of favorites be somewhat spontaneous, without an excessive amount of reconsideration? Yes, but the opposite pitfall is also dangerous: selecting films impulsively may reflect a little too strongly the cultural opinions of the critical tastemakers who cobbled together the Sight & Sound poll in the first place.

Hereby, then, is the beginning of a new project: My Canon, or my 100 favorite movies, analyzed one-by-one in alphabetical order. It should be noted that some of these movies I haven't seen in a matter of years, so it's conceivable that, after rewatching them for this project, I might in fact revise my original opinion and regard them with diminished esteem this time around. So be it: if this happens, it will merely prove the point that any kind of subjective list-making is prone to impulse and reformed opinions. At the end of the project, I may have a slightly recalibrated Top 100, and will be able to whittle it down to a Top Ten with more decisiveness (presumably, at least, though Top Tens are even more unfair and impossibly exclusionary than Top 100s). It should also be noted that, while I am attempting to establish my own personal canon, there will be considerable overlap with the Sight & Sound poll; while I don't want to merely regurgitate critical favorites, it's undeniable that some of those titles reappear frequently because they do hold seemingly inexhaustible narrative, aesthetic, emotional, and thematic rewards.

To return, finally, to the first epigraph above: Sontag's insightful remark about list-making seems absolutely appropriate – critics and art-lovers enjoy making lists because it offers an ardent way to "possess" the memories we have of our favorite films. To rank these movies is to remember and desire them, to travel back in time to the overwhelming initial experience I had with them. This might be the significant difference between favorite films and greatest films: favorites continue to exert a resonant hold over our memories, sometimes exuding an emotional power equivalent to our own real-life remembrances. Making a canon like this is, in other words, a way to possess (maybe meekly, desperately) the feelings we initially had about the movies we include.